Você está na página 1de 553


Guide to
Guide to

Edited by

Alan J. Gelenberg, M.D.

University of Arizona College of Medicine
Tucson, Arizona


Ellen L Bassuk, M.D.

The Better Homes Fund
Newton Centre, Massachusetts
and Harvard Medical School
Boston, Massachusetts


Library off Congress Catalog1ng-1n-Publ1cat1o n Dat a

The practitioner' s guid e t o psychoactiv e drug s / edite d by Ala n J .

Gelenberg and Elle n L . Bassuk. — 4t hed .
p. en .
Include s bibliographica l reference s and Index .
ISBN978-1-4613-7695- 8 ISBN978-1-4615-5877- 4(eBook)
DOI 10.1007/978-1-4615-5877-4
1. Mental 111ness--Cheiotherapy . 2 . Psychopharnacology.
3. Psychoactiv e drugs . I . Gelenberg, Ala n J . I I . Bassuk, Elle n
L. . 1945-
[DNLM: 1 . Mental Disorders—dru g therapy . 2 . Psychotropi c Drugs.
WM402 P8952 1997]
RC483.P72 6 1997
615* .78—dc2 1
fo r Librar y o f Congress 97-29885

The information in this book is based on the experience s and research of the editors and authors.
This material is for informational purpose s only and should not be construe d as prescribing
information for individual patients. The editors, authors, and publisher assum e no responsibility for
any treatmen t undertake n by the practitioner with individual patients. Companies , interventions,
and products are mentioned without bias to increase your knowledge only.

ISBN 978-1-4613-7695-8
© 1997 Springer Science+Busines s Media New York
Originall y published by Plenum Publishing Corporation in 1997, 1991, 1983, 1977
Softcover reprint of the hardcover 4th edition 1997

Al l rights reserve
No part of this book may be reproduced , stored in a retrieval system, or transmitted in any form or
by any means , electronic, mechanical
, photocopying , microfilming, recording, or otherwise,
without written permission from the Publisher
To Sara and Rebecca Gelenberg

To my father Irving Bassuk

For your unwavering support of my work

Lori Altshuler, M.D. Associate Professor, Department of Psychiatry; Director,

Mood Disorders Program, UCLA Neuropsychiatric Hospital and Center for Behav-
ioral Sciences, Los Angeles, California 90024; VA Medical Center, West Los
Angeles, Los Angeles, California 90073

Joseph Biederman, M.D. Chief, Joint Program in Pediatric Psychopharmacology,

Massachusetts General Hospital, Boston, Massachusetts 02114; McLean Hospital,
Belmont, Massachusetts 02178; Professor of Psychiatry, Harvard Medical School,
Boston, Massachusetts 02115

Andrew W. Brotman, M.D. Chief of Psychiatry, Deaconess Hospital, Boston, Mas-

sachusetts 02215; Associate Professor of Psychiatry, Harvard Medical School,
Boston, Massachusetts 02115

Lee S. Cohen, M.D. Director, Perinatal and Reproductive Psychiatry Clinical Re-
search Program, Clinical Psychopharmacology Unit, Massachusetts General Hos-
pital, Boston, Massachusetts 02114; Associate Professor of Psychiatry, Harvard
Medical School, Boston, Massachusetts 02115

Pedro L. Delgado, M.D. Director, Affective Disorders Program; Associate Pro-

fessor of Psychiatry, Department of Psychiatry, University of Arizona Health Sci-
ences Center, Tucson, Arizona 85724

Francisco Fernandez, M.D. Professor and Chairman, Department of Psychiatry,

Stritch School of Medicine, Loyola University Medical Center, Maywood, Illinois

Alan J. Gelenberg, M.D. Professor and Head, Department of Psychiatry, University

of Arizona Health Sciences Center, Tucson, Arizona 85724

Shelly F. Greenfield, M.D., M.P.H. McLean Hospital, Belmont, Massachusetts

02178; Department of Psychiatry, Harvard Medical School, Boston, Massachusetts

Thomas G. Gutheil, M.D. Program in Psychiatry and the Law, Massachusetts Men-
tal Health Center, Boston, Massachusetts 02115

Vicki L. Heller, M.D. Associate in Obstetrics and Gynecology, Brigham and Wom-
en's Hospital, Boston, Massachusetts 02115

David B. Herzog, M.D. Director of the Eating Disorders Unit, Massachusetts Gen-
eral Hospital, Boston, Massachusetts, 02114; Associate Professor of Psychiatry,
Harvard Medical School, Boston, Massachusetts 02115

Heather S. Hopkins Associate Editor, Department of Psychiatry, University of Ari-

zona Health Sciences Center, Tucson, Arizona 85724

David C. Jimerson, M.D. Director of Research, Department of Psychiatry, Beth

Israel Hospital, Boston, Massachusetts 02215; Associate Professor of Psychiatry,
Harvard Medica. School, Boston, Massachusetts 02115

Samuel Keith, M.D. Chair, Department of Psychiatry, University of New Mexico

School of Medicine, Albuquerque, New Mexico 87131

Jorge Maldonado, M.D. Clinical Fellow, Psychosomatic Medicine & Consultation

Liaison Psychiatry, St. Luke's Episcopal Hospital, Houston, Texas 77030; Baylor
College of Medicine, Houston, Texas 77030

Steven M. Mirin, M.D. McLean Hospital, Belmont, Massachusetts 02178; Depart-

ment of Psychiatry, Harvard Medical School, Boston, Massachusetts 02115

Eric M. Reiman, M.D. Department of Psychiatry, University of Arizona; Scientific

Director, Samaritan PET Center, Good Samaritan Regional Medical Center, Phoe-
nix, Arizona 85006

Jerrold Rosenbaum, M.D. Director, Outpatient Psychiatry Division, Massachusetts

General Hospital, Boston, Massachusetts 02114

Carl Salzman, M.D. Department of Psychiatry, Harvard Medical School, Boston,

Massachusetts 02115

Paul H. Soloff, M.D. Professor of Psychiatry, University of Pittsburgh Medical

Center, Western Psychiatric Institute and Clinic, Pittsburgh, Pennsylvania

Thomas Spencer, M.D. Assistant Director, Pediatric Psychopharmacology Unit,

Massachusetts General Hospital, Boston, Massachusetts 02114; Assistant Pro-
fessor, Harvard Medical School, Boston, Massachusetts 02115

Jeffrey B. Weilburg, M.D. Director, Neuropsychiatry Section, Psychopharmacology

Unit, Massachusetts General Hospital, Boston, Massachusetts 02114; Harvard
Medical School, Boston, Massachusetts 02115

Roger D. Weiss, M.D. Alcohol and Drug Abuse Program, McLean Hospital, Bel-
mont, Massachusetts 02178; Department of Psychiatry, Harvard Medical School,
Boston, Massachusetts 02115

Timothy Wilens, M.D. Staff Psychiatrist, Pediatric Psychopharmacology Unit,

Massachusetts General Hospital, Boston, Massachusetts 02114; Assistant Pro-
fessor, Harvard Medical School, Boston, Massachusetts 02115

Janet Wozniak, M.D. Staff Psychiatrist, Pediatric Psychopharmacology Unit, Mas-

sachusetts General Hospital, Boston, Massachusetts 02114; Instructor in Psychia-
try, Harvard Medical School, Boston, Massachusetts 02115

Once again, in their new edition of a classic American handbook of clinical psycho-
pharmacology, Drs. Gelenberg and Bassuk and their colleagues have produced a master-
work of sound clinical guidelines for the use of medicines as an increasingly central
component of contemporary psychiatric practice. They have recruited an outstanding
panel of coauthors, but have managed to maintain a high level of consistency of quality
and style throughout the many chapters on specific classes of psychiatric illnesses and
their corresponding treatments. The book continues to encapsulate the clinically rele-
vant essentials of the pharmacology of each major class of psychotropic agents in a
way that clinicians can easily grasp. Moreover, it provides sound and carefully consid-
ered specific guidelines to diagnosis, drug selection, and dosing and patient assessment,
with particularly rich presentations on the recognition and management of adverse
effects. Inclusion of chapters on pediatric and geriatric populations and on topics
pertaining to substance abuse disorders and medicolegal aspects of the field are unusual
in books of this kind, and so are particularly valuable. Students and trainees will
appreciate the handbook's well-organized and readable style, and practicing clinicans
should be satisified with its balanced consideration of older, standard treatments as well
as the latest medicines and trends in clinical psychopharmacology, with glimpses into
future developments. The fresh and up-to-date quality of this edition is particularly
attractive in this era of progress and appearance of new treatments that is unprece-
dented since the early decades of our field. Sensitivity to the art of the practice of
clinical psychopharmacology and specific consideration of cost-effectiveness are also
very welcome in a modem textbook. Finally, the format and presentation of the book
continue to be attractive, easily handled, and especially appealing to busy trainees and
clinicians. This is a book to be used by students, trainees, and clinicians from many
disciplines, not only by psychiatrists and mental health professionals.

Ross 1. Baldessarini, M.D.

Professor of Psychiatry and in Neuroscience

Harvard Medical School
Director, Laboratories for Psychiatry Research
and the Bipolar & Psychotic Disorders and Psychopharmacology Programs
McLean Division of Massachusetts General Hospital

Since the third edition of this guide was published in 1991, clinical psychopharmacol-
ogy has made a leap forward that rivals the rapid advances of the 1950s. In the
intervening years, the role of serotonin-selective reuptake inhibitors (SSRIs) has ex-
panded, and their efficacy has been affirmed in the treatment of depression, obsessive-
compulsive disorder and related conditions, and panic and other anxiety disorders, and
perhaps in the adjunctive treatment of substance abuse. The greater tolerability of these
agents and their relative safety in overdose have benefited hundreds of thousands of
people who suffer from these conditions. Other antidepressants with unique phar-
macologic and clinical profiles have been introduced during this time, and their place
in psychiatric pharmacotherapy is currently being investigated.
As we write this Preface, scientists are studying the pharmacologic treatment of
psychoses. We are learning more about the potential and the limitations of the landmark
drug clozapine, and several new and novel antipsychotic agents are being tested and
probably will be released soon. At the same time, gradual progress is being made in the
use of drugs to treat addictive disorders.
Biological research is likely to inform clinical practice in greater and more mean-
ingful ways over the next few years. We may soon understand the genetic origins of
some psychiatric diseases, such as bipolar disorder. Studies of the P450 liver enzymes
have given us greater insight into interindividual differences in the pharmacokinetic
handling of medications and also the potential for drug-drug and drug-food interac-
tions. The tryptophan depletion paradigm has held up in multiple replication studies
following the innovative research of Delgado and others (1994), and it has been
extended from depression studies to studies of other conditions as a way to explore the
mechanism of antidepressant agents. Functional and structural imaging technologies
promise to reveal even more about how psychotropic drugs work.
In parallel with the unprecedented discoveries and imminent breakthroughs in
psychiatric pharmacotherapy, a virtual revolution in healthcare is occurring in the
United States. Practices have changed and continue to be modified at a dizzying pace,
but the ultimate shape and structure of our healthcare system, including that portion
devoted to the treatment of mental and emotional disorders, remains unclear. Although
cost drove the equation in the early 1990s, we hope quality will have a growing
influence in the closing of this millennium. We anticipate that practitioners conversant
with both pharmacologic and psychosocial approaches to behavioral conditions will
function prominently in the new system.
The editors appreciate the important practical role The Practitioner's Guide has
come to play for many clinicians. We have kept that in mind in planning and executing
this edition. We hope it will continue to assist practitioners in their daily work and
ultimately benefit many patients.

Dr. Gelenberg appreciates the ongoing support of Ms. Josie Riley, who keeps his
daily schedule rational, and Ms. Heather Hopkins, whose writing and research assis-
tance with this and countless other projects has been invaluable. The University of
Arizona Department of Psychiatry and College of Medicine have provided a supportive
and intellectually stimulating milieu. Finally, he wants to acknowledge and thank his
wonderful daughters, whose love has provided vital sustenance through the most
difficult times. Dr. Gelenberg is particularly honored that his first teacher in psycho-
pharmacology, an internationally renowned and award-winning expert in the field,
Professor Ross Baldessarini, has agreed to write the Foreword to this book.
Dr. Bassuk wants to thank Jayne Samuda for her good-humored support and
assistance in completing this manuscript. Special thanks go to her two wonderful
children, Sarah and Danny, for being part of her life.

Delgado P. L., Price, L. H., Miller, H. L., Salomon, R. M., Aghajanian, G. K., Heninger, G. R., Charney, D. S.
1994, Serotonin and the neurobiology of depression. Effects of tryptophan depletion in drug-free
depressed patients. Arch Gen Psychiatry 51 :865-874.

I. State of the Art

1. Introduction: The Practice of Pharmacotherapy

Alan 1. Gelenberg, M.D.
I. Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
II. Current Trends in Psychopharmacology ........................ 5
A. Definition of Clinical Syndromes .......................... 5
B. Development of Brain Chemistry Models and New
Pharmacological Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
C. Design of Drug Studies .................................. 6
III. Perspectives on Clinical Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
A. Matching Medications with Drug-Responsive Syndromes . . . . . . . 7
B. Pharmacotherapy and Psychotherapy ....................... 8
IV. General Clinical Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
A. Patient Evaluation ...................................... 9
B. Principles of Drug Use .................................. 10
1. Understand the Pharmacokinetics of Psychotropic Agents . . . 10
2. Prescribe the Simplest Drug Regimen to
Increase Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3. Use the Fewest Number of Drugs to Achieve
the Desired Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4. Provide the Most Cost-Effective Treatment. . . . . . . . . . . . . . . 13
5. Exercise Special Care with Medically III Patients ......... 14
V. Conclusion ............................................... 14
References ............................................... 15

II. Major Psychiatric Disorders

2. Depression
Alan 1. Gelenberg, M.D., and Pedro L. Delgado, M.D.
I. Introduction ............................................. . 19
II. Causal Models ........................................... . 20
A. Early Models and Historical Perspective 20

B. Contemporary Models ................................... 20

1. Psychological . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2. Biological ......................................... 21
3. Integrative ......................................... 22
III. Diagnostic Considerations ................................... 24
A. Clinical Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
B. Classifications.......................................... 26
C. Biological Diagnosis .................................... 28
IV. General Therapeutic Measures ............................... 28
A. Milieu and Crisis Techniques ............................. 29
B. Psychotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
C. Electroconvulsive Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1. Introduction ........................................ 31
2. Mechanism of Action ................................ 32
3. Indications ......................................... 32
4. Contraindications and Adverse Reactions ................ 33
5. Technique ......................................... 33
V. Pharmacotherapy .......................................... 34
A. Heterocyclic Antidepressants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1. Chemistry ......................................... 38
2. Mechanism of Action ................................ 41
3. Adverse Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
a. Gastric ......................................... 43
b. Hematological ................................... 44
i. Leukocytic Effects and Purpura ................. 44
ii. Agranulocytosis and Other Hemolytic Effects ...... 44
c. Hepatic ......................................... 44
d. Endocrine ....................................... 44
e. Ocular . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
f. Cardiovascular ................................... 45
i. Hypotension and Hypertension .................. 45
ii. Heart Rate Changes ........................... 46
iii. Conduction Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
iv. Antiarrhythmic/ Arrhythmogenic Activity .......... 47
g. Neurological. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
i. Sedation/Stimulation .......................... 49
ii. Headaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
iii. Nocturnal Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . 50
iv. Seizures..................................... 50
v. Anticholinergic Effects . . . . . . . . . . . . . . . . . . . . . . . . . 50
vi. Movement Disorders .......................... 50
vii. Withdrawal Reactions ......................... 51
Vlll. Psychiatric Side Effects ........................ 52
h. Cutaneous. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
i. Autonomic ...................................... 52
j. Sexual Side Effects ............................... 55

k. Drug Interactions ................................. 56

1. Weight Changes .................................. 57
4. Precautions for Administration during Pregnancy . . . . . . . . . . 57
5. Acute Toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
a. Intoxication Syndromes ............................ 58
b. Anticholinergic Syndromes ......................... 62
i. Description .................................. 62
ii. Anticholinesterase Therapy ..................... 62
6. Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
7. Plasma Levels ...................................... 66
8. Preparations and Dosage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
B. Monoamine Oxidase Inhibitors ............................ 69
1. Chemistry ......................................... 69
2. Mechanism of Action ................................ 70
3. Adverse Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
a. Hypertensive Crisis ............................... 72
b. Hypermetabolic Crisis ............................. 73
c. Postural Hypotension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
d. Cardiac Effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
e. Sexual Dysfunction ............................... 75
f. Edema and Weight Gain ........................... 75
g. Neuropsychiatric ................................. 76
i. Sleep Disturbances ............................ 76
ii. Psychiatric .................................. 76
iii. Other Neurological Effects ..................... 76
h. Hepatic ......................................... 76
i. Other Adverse Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
j. Withdrawal Syndrome ............................. 77
4. Contraindications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5. Use during Pregnancy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6. Toxicity ........................................... 77
7. Pharmacokinetics.................................... 78
8. Preparations and Dosage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
C. Stimulants ............................................. 79
D. Evaluation of the Depressed Patient ........................ 79
E. Drug Therapy for the Depressed Patient . . . . . . . . . . . . . . . . . . . . . 81
1. Initiating Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
2. Completing a Medication Trial ........................ 83
3. Continuation Therapy ................................ 85
4. Maintenance Therapy for Chronic or
Recurrent Depressions ............................... 85
5. Depressive Episodes in Bipolar Disorders . . . . . . . . . . . . . . . . 86
6. Some Dichotomies: Mild versus Severe, Acute versus Chronic,
Agitated/Anxious versus Retarded . . . . . . . . . . . . . . . . . . . . . . 87
7. "Atypical" Depression ............................... 88
8. Major Depressive Episode with Psychotic Features ........ 89

9. Depression in Schizophrenic Patients ................... 89

10. Resistant Depressions ................................ 90
VI. Conclusion ............................................... 92
References ............................................... 93

3. Bipolar Disorder
Heather S. Hopkins and Alan 1. Gelenberg, M.D.

I. Introduction .............................................. 99
A. Clinical Presentation .................................... 100
B. Course of the Illness .................................... 101
II. General Measures for Treating Bipolar Disorder ................. 102
A. Electroconvulsive Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
B. Milieu Therapy. . . . . . . . . . . . . .. . . . . . . . . .. . . . . . . . . . . . . . . . . 103
C. Group Therapy. . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . 103
D. Psychotherapy and the Therapeutic Relationship .............. 104
III. Phannacotherapy .......................................... 104
A. Lithium ............................................... 105
1. Phannacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
2. Adverse Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
a. Endocrine ....................................... 108
i. Thyroid ..................................... 108
ii. Diabetes Mellitus ............................. 108
b. Renal .......................................... 109
i. Decreased Concentration ....................... 109
ii. Structural Changes and Glomerular
Filtration Alterations .......................... 109
iii. Renal Function ............................... 109
iv. Other Kidney Effects .......................... 110
c. Hematological ................................... 110
d. Cardiovascular ................................... 110
e. Cutaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
f. Gastrointestinal .................................. 111
g. Central Nervous System and Neuromuscular . . . . . . . . . . . 111
h. Ocular. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
i. Weight Gain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
j. Other Adverse Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . 113
3. Drug Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
4. Lithium Toxicity .................................... 115
5. Pregnancy ......................................... 117
6. Preparations and Dosage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7. Other Possible Indications for Lithium .................. 118
B. Valproate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
1. Phannacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

2. Adverse Reactions .................................. . 119

a. Gastrointestinal ................................. . 119
b. Central Nervous System Effects .................... . 120
c. Dermatological .................................. . 120
d. Hematological .................................. . 120
e. Hepatic ........................................ . 121
f. Other .......................................... . 121
3. Drug Interactions ................................... . 121
4. Overdose ......................................... . 122
5. Pregnancy ........................................ . 123
6. Preparations and Dosage ............................. . 123
C. Carbamazepine ........................................ . 123
1. Pharmacokinetics ................................... . 123
2. Adverse Reactions .................................. . 124
a. Hematological .................................. . 125
b. Dermatological .................................. . 125
c. Cardiovascular .................................. . 125
d. Hepatic ........................................ . 125
e. Neurological .................................... . 125
f. Gastrointestinal ................................. . 126
g. Ocular ......................................... . 126
h. Metabolism ..................................... . 126
i. Renal ......................................... . 126
j. Genitourinary ................................... . 126
k. Other .......................................... . 126
3. Drug Interactions ................................... . 127
4. Overdose ......................................... . 129
5. Pregnancy ........................................ . 130
6. Preparations and Dosage ............................. . 130
D. Antipsychotics ........................................ . 130
E. Antidepressants ........................................ . 131
F. Other ................................................ . 132
N. Clinical Applications ...................................... . 133
A. Diagnostic Evaluation .................................. . 133
B. Medical Evaluation .................................... . 135
C. Treatment ............................................ . 136
1. Acute Manic Episode ............................... . 138
2. Hypomania ....................................... . 141
3. Acute Depressive Episode ........................... . 141
4. Maintenance Therapy ............................... . 142
5. Rapid Cyclers ..................................... . 143
D. Mixed or Dysphoric Mania .............................. . 144
E. Schizoaffective Disorder ................................ . 144
V. Conclusion 145
References .............................................. . 145

4. Psychoses
Alan J. Gelenberg, M.D., and Samuel Keith, M.D.
I. Introduction .............................................. 153
II. Other Treatments .......................................... 154
A. Nonbiological .......................................... 154
B. Nondrug, Biological . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
1. Electroconvulsive Therapy ............................ 156
2. Psychosurgery ...................................... 156
III. Antipsychotic Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
A. Introduction and Terms .................................. 156
B. Effects on Behavior and the Nervous System ................ 157
C. Mechanism of Action. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
D. Classes and Chemistry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
E. General Principles of Use ................................ 166
F. Pharmacokinetics ....................................... 168
G. Adverse Effects and Toxicity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
1. Neurological ....................................... 170
a. Extrapyramidal Syndromes ......................... 170
i. Acute Dystonic Reactions ...................... 170
ii. Akathisia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
iii. Parkinson's Syndrome ......................... 173
iv. Tardive Dyskinesia and Tardive Dystonia. . . . . . . . . . 176
b. Sedation ........................................ 185
c. Seizures ........................................ 185
d. Neuroleptic Malignant Syndrome .................... 186
2. Anticholinergic ..................................... 187
a. Peripheral ....................................... 187
b. Central ......................................... 188
c. Serious Toxicity .................................. 188
3. Cardiovascular and Respiratory ........................ 188
a. Hypotension ..................................... 188
b. Cardiac ......................................... 189
4. Ocular ............................................ 190
5. Cutaneous ......................................... 191
6. Hormonal, Sexual, and Hypothalamic Reactions .......... 191
7. Hepatic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
8. Hematological ...................................... 193
9. Pregnancy and Lactation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
10. Withdrawal Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
11. Overdose .......................................... 195
H. Drug Interactions and Combinations . . . . . . . . . . . . . . . . . . . . . . . . 196
I. Laboratory Tests and Monitoring .......................... 198
J. Clinical Uses of the Antipsychotic Drugs. . . . . . . . . . . . . . . . . . . . 199
1. Acute Treatment .................................... 199

2. Transition and Continuation Therapy . . . . . . . . . . . . . . . . . . . . 204

3. Maintenance Therapy ................................ 205
a. Definition ....................................... 205
b. Chronic Schizophrenia. . . . . . . . . .. . . . . . . . . .. . .. . . . . . 205
c. Other Diagnostic Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
IV. Conclusion ............................................... 208
References ............................................... 209

5. Anxiety
Eric M. Reiman, M.D.
I. Introduction .............................................. 213
II. Current Concepts/Etiological Theories ......................... 213
A. Symptoms ............................................. 213
B. Substrates ............................................. 214
1. Neuroanatomical Substrates ........................... 214
2. Neurochemical Substrates ..................... ~ . . . . . . . 215
3. Psychological Substrates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
C. Integrating Biological and Psychosocial Contributions
to the Problem ......................................... 217
D. Differential Diagnosis ................................... 218
1. The Anxious Response to Everyday Stressors ............ 219
2. Anxiety Disorders ................................... 219
a. The Physiological Effect of a Nonpsychiatric
Medical Condition ................................ 219
b. The Physiological Effect of a Substance or
Its Discontinuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
c. Panic Disorder ................................... 220
d. Agoraphobia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
e. Social Phobia .................................... 221
f. Specific Phobia .................................. 222
g. Obsessive-Compulsive Disorder .................... 223
h. The Response to a Catastrophic Stressor .............. 225
i. Generalized Anxiety Disorder . . . . . . . . . . . . . . . . . . . . . . . 226
3. Other Psychiatric Disorders ........................... 227
III. Nonpharmacological Treatment Measures ...................... 227
A. Supportive Therapy ..................................... 227
B. Cognitive and Behavioral Therapies ........................ 228
C. Neurosurgical Treatments ................................ 229
IV. Pharmacotherapy .......................................... 229
A. Benzodiazepines .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
1. Introduction ........................................ 229
2. Mechanism of Action ................................ 231
3. Chemistry ......................................... 231

4. Principles of Benzodiazepine Use ...................... 232

5. Pharmacokinetics.................................... 233
a. Onset .......................................... 233
b. Administration Route . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
c. Duration ........................................ 235
d. Potency. .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . 237
6. Adverse Effects ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
a. Cognitive and Performance Effects. . . . . . . . . . . . . . . . . . . 237
b. Uncommon Effects ............................... 238
c. Dependence ..................................... 239
d. Pregnancy-Related Risks. . . .. . . . . . . . . . . . . . .. . . . . . . . 240
e. Risks in the Elderly ............................... 241
f. Toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
7. Drug Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
8. Laboratory Tests .................................... 242
B. Other Sedative-Hypnotics .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
C. Buspirone ............................................. 243
D. ~-Blockers ............................................ 245
E. Antidepressants and MAO Inhibitors ....................... 246
V. Clinical Considerations ..................................... 248
A. General Principles ...................................... 248
B. Adjustment Disorder .................................... 249
C. Panic Disorder ......................................... 251
D. Agoraphobia ........................................... 254
E. Social Phobia, Circumscribed Type . . . . . . . . . . . . . . . . . . . . . . . . . 255
F. Social Phobia, Generalized Type . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
G. Specific Phobia. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . 256
H. Obsessive-Compulsive Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
I. Acute and Posttraumatic Stress Disorders ................... 259
J. Generalized Anxiety Disorder ............................. 260
VI. Conclusions .............................................. 262
References ............................................... 262
Selected Readings ......................................... 262

6. Insomnia
Jeffrey B. Weilburg, M.D., and Heather S. Hopkins

I. Introduction .............................................. 265

A. Normal Sleep .......................................... 266
B. Sleep Disorders ........................................ 266
1. Sleep Apnea ....................................... 267
2. Nocturnal Myoclonus ................................ 267
3. Phase Shifts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
4. Psychophysiological Insomnia ......................... 268
5. Primary Idiopathic Insomnia .......................... 268
II. Diagnostic Considerations ................................... 268

III. General Therapeutic Measures ............................... 270

IV. Pharmacotherapy .......................................... 271
A. Benzodiazepines .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
1. Mechanism of Action ................................ 272
2. Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
3. Onset and Duration of Action ......................... 273
4. Adverse Effects ..................................... 273
a. Next-Day Sedation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
b. Rebound Insomnia and Anxiety ..................... 274
c. Cognitive and Psychomotor Impairment .............. 274
d. Behavioral Alterations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
e. Respiratory and Cardiovascular ..................... 275
f. Elderly Patients .................................. 275
5. Drug Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
6. Dependence and Withdrawal .......................... 277
7. Toxicity ........................................... 277
8. Pregnancy and Breast Feeding ......................... 277
9. Preparations and Dosing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
B. Zolpidem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
1. Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
2. Adverse Effects ..................................... 279
3. Withdrawal ........................................ 279
4. Drug Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
5. Overdose .......................................... 279
6. Pregnancy ......................................... 280
7. Preparations and Dosing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
C. Antihistamines ......................................... 281
D. Barbiturates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
E. Chloral Hydrate ........................................ 281
F. Ethchlorvynol .......................................... 282
G. Over-the-Counter Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
H. Other Psychotropics for Insomnia . . . . . . . . . . . . . . . . . . . . . . . . . . 283
1. Antidepressants ..................................... 283
2. Antipsychotics ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
3. Buspirone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
4. Clonidine .......................................... 284
V. Clinical Application Summary ............................... 284
VI. Conclusion ............................................... 286
References ............................................... 286

III. Psychoactive Substance Abuse

7. Psychoactive Substance Use Disorders
Shelly F. Greenfield, M.D., M.P.H., Roger D. Weiss, M.D., and Steven M. Mirin, M.D.
I. Introduction and Definitions ................................. 291

II. Opioid Abuse and Dependence ............................... 293

A. Classification .......................................... 293
B. Pharmacology........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
C. Acute Effects .......................................... 294
1. Central ............................................ 294
2. Peripheral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
D. Tolerance and Physical Dependence. . . . . . . . . . . . . . . . . . . . . . .. 295
E. Acute Intoxication and Overdose .......................... 295
1. Etiology ........................................... 295
2. Clinical Manifestations ............................... 295
3. Management ....................................... 296
a. General Life-Support Measures ..................... 296
b. The Use of Naloxone ............................. 296
c. Other Considerations .............................. 296
F. Opioid Withdrawal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
1. Clinical Manifestations ............................... 297
2. Management ....................................... 297
a. Medical Detoxification ............................ 297
b. Detoxification with Methadone . . . . . . . . . . . . . . . . . . . . . . 298
i. Pharmacology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
ii. Principles of Use ............................. 298
iii. Clinical Application ........................... 298
c. The Use of Clonidine in Opioid Detoxification. . . . . . . . . 299
d. Clonidine/Naltrexone Rapid Detoxification ............ 300
e. The Use of Buprenorphine in Opioid Detoxification. . . . . 300
G. Treatment of Chronic Opioid Abuse. . . . . . . . . . . . . . . . . . . . . . . . 301
1. Medical Complications ............................... 301
2. Treatment Approaches ............................... 302
a. General Considerations ............................ 302
b. Methadone Maintenance ........................... 302
i. Background and Theory . . . . . . . . . . . . . . . . . . . . . . . . 302
ii. Clinical Application ........................... 302
iii. Detoxification from Methadone . . . . . . . . . . . . . . . . . . 303
iv. Current Issues in Methadone Treatment ........... 304
v. LAAM Maintenance .......................... 304
c. The Use of Narcotic Antagonists .................... 304
i. Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
ii. Clinical Use ................................. 305
d. Outpatient Psychosocial Treatments .................. 306
e. Treatment of Co-occurring Psychiatric Disorders ....... 307
f. Therapeutic Communities .......................... 307
III. Sedative-Hypnotic and Benzodiazepine Abuse and Dependence. . . . 307
A. Introduction ........................................... 307
B. Tolerance and Physical Dependence. . . . . . . . . . . . . . . . . . . . . . . . 308
C. Classification .......................................... 308
1. Benzodiazepines .................................... 308

2. Barbiturates ........................................ 309

3. Other Sedative-Hypnotics ............................ 309
D. Abuse of Sedative-Hypnotics and Benzodiazepines ........... 310
E. Acute Intoxication and Overdose .......................... 311
1. Etiology .......................... . . . . . . . . . . . . . . . . . 311
2. Clinical Manifestations ............................... 311
3. Management ....................................... 312
F. Withdrawal from Benzodiazepines and Sedative-Hypnotics. . . . . 312
1. Clinical Manifestations ............................... 312
a. High-Dose Benzodiazepine Withdrawal Syndrome 313
b. Low-Dose Benzodiazepine Withdrawal Syndrome ...... 313
2. Treatment of Benzodiazepine and Sedative-Hypnotic
Withdrawal Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
a. Treatment of High-Dose Benzodiazepine and Sedative-
Hypnotic Withdrawal .................. , . . . . . .. . . . . 314
i. Phenobarbital Substitution ...................... 314
ii. The Phenobarbital Tolerance Test. . . . . . . . . . . . . . . . 315
iii. Gradual Dose Reduction ....................... 316
b. Treatment of Low-Dose Benzodiazepine Withdrawal .... 317
3. Mixed Opioid-Sedative Dependence . . . . . . . . . . . . . . . . . . . . 317
IV. Alcohol Abuse and Dependence .............................. 317
A. Introduction ........................................... 317
B. Pharmacology.......................................... 318
C. Tolerance and Physical Dependence. . . . . . . . . . . . . . . . . . . . . . . . 318
D. Acute Intoxication ...................................... 319
1. Simple Type ....................................... 319
2. Alcohol Intoxication Delirium ......................... 320
E. Chronic Intoxication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
1. Peripheral and Central Nervous System Effects ........... 320
2. Effects on Other Organ Systems ....................... 321
F. Alcohol Withdrawal ................................. . . . . 321
1. Minor Abstinence Syndrome .......................... 321
a. Clinical Manifestations ............................ 321
b. Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
2. Major Abstinence Syndrome (Alcohol
Withdrawal Delirium) ............ . . . . . . . . . . . . . . . . . . . . 323
a. Clinical Manifestations ............................ 323
b. Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
3. Alcohol-Induced Psychotic Disorder .................... 324
G. Treatment of Chronic Alcohol Use Disorders ................ 324
1. Introduction ........................................ 325
2. Biological Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
a. Disulfiram ...................................... 325
b. Naltrexone ...................................... 326
c. Psychotropic Drugs ............................... 326
3. Aftercare .......................................... 327

4. Cognitive-Behavioral and Behavioral Approaches

to Treatment ....................................... 327
a. Cognitive-Behavioral Techniques. . . . . . . . . . . . . . . . . . . . 327
b. Other Behavioral Techniques ....................... 327
5. The Psychotherapies ................................. 328
6. Brief Interventions .................................. 328
7. Alcoholics Anonymous and Other Self-Help Groups. .. . . . . 328
8. Adjunctive Services ................................. 329
9. Summary.......................................... 329
V. Central Nervous System Stimulant Abuse and Dependence ........ 329
A. Cocaine ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
1. Epidemiology ...................................... 329
2. Pharmacology and Methods of Use ..................... 330
3. Acute Intoxication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
4. Abstinence Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
5. Tolerance and Physical Dependence .................... 331
6. Medical Complications ............................... 332
7. Treatment of Cocaine Abuse .......................... 332
B. Amphetamines ......................................... 333
1. Introduction ........................................ 333
2. Pharmacology ...................................... 333
3. Acute Effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
4. Tolerance and Physical Dependence .................... 334
5. Patterns of Abuse ................................... 334
6. Acute Intoxication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
a. Clinical Manifestations ............................ 335
b. Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
7. Amphetamine Withdrawal ............................ 336
8. Amphetamine- or Similarly Acting Sympathomimetic-Induced
Delusional Disorder ................................. 336
VI. Hallucinogen Abuse ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
A. Introduction ........................................... 337
B. Classification .......................................... 337
C. d-Lysergic Acid Diethylamide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
1. Pharmacology ...................................... 338
2. Tolerance and Physical Dependence .................... 338
3. Acute Intoxication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
D. Adverse Reactions following Use of LSD or
Other Hallucinogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
1. Panic Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
2. Toxic Delirium ..................................... 341
3. Drug-Precipitated Functional Psychosis. . . . . . . . . . . . . . . . . . 341
4. "Flashbacks" - Hallucinogen Persisting
Perception Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
E. Sequelae of Chronic Hallucinogen Use ..................... 342
VII. Phencyclidine Abuse ....................................... 342

VIII. Inhalant Abuse 343

A. Introduction .......................................... . 343
B. Patterns of Abuse ...................................... . 343
C. Acute Intoxication ..................................... . 343
D. Sequelae of Chronic Inhalant Use ......................... . 344
E. Treatment Approaches .................................. . 344
IX. "Designer" Drugs: MDMA and MDEA ....................... . 344
X. Marijuana Use and Abuse .................................. . 346
A. Introduction .......................................... . 346
B. Patterns of Use and Abuse .............................. . 346
C. Tolerance and Dependence .............................. . 347
D. Pharmacology ......................................... . 347
E. Acute Intoxication ..................................... . 348
1. Subjective Effects .................................. . 348
2. Physiological Changes .............................. . 348
3. Cognitive and Psychomotor Effects .................... . 348
F. Adverse Reactions following Marijuana Use ................ . 349
1. Acute Panic ....................................... . 349
2. Toxic Delirium .................................... . 349
3. Recurrent Reactions ("Flashbacks") .................... . 349
G. Sequelae of Chronic Use ................................ . 350
1. Cannabis Psychosis ................................. . 350
2. Amotivational Syndrome ............................ . 350
3. Depression ........................................ . 351
4. Medical Problems .................................. . 351
H. Treatment of Chronic Marijuana Abuse .................... . 351
XI. Substance Abuse in Perspective ............................. . 352
Selected Reading ......................................... . 353

IV. Special Topics

8. Geriatric Psychopharmacology
Carl Salzman, M.D.

I. Introduction .............................................. 367

II. Adverse Reactions ......................................... 367
A. Central Nervous System ................................. 368
1. Sedation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
2. Confusion ......................................... 372
3. Extrapyramidal ..................................... 373
4. Other Central Nervous System Symptoms ............... 373
B. Cardiac ............................................... 374
1. Effects on Blood Pressure ............................ 374
2. Effects on Rate, Rhythm, and Contractility. . . . . . . . . . . . . . . 374
C. Other Organ Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

III. Clinical Application ........................................ 375

A. Principles of Medication Use to Treat Elderly Patients Safely ... 375
B. Clinical Use ........................................... 376
1. Agitated Behavior and Psychotic Thinking ............... 376
a. Neuroleptic Treatment of Acute Psychosis or
Behavior Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
b. Non-Neuroleptic Treatment of Acute Psychosis or
Behavior Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
c. Treatment of Extrapyramidal Symptoms .............. 378
2. Dementia .......................................... 378
3. Anxiety ........................................... 379
4. Insomnia .......................................... 380
5. Depression......................................... 381
6. Bipolar Affective Disorder ............................ 383
IV. Conclusion ............................................... 384

9. Pediatric Psychopharmacology
Janet Wozniak, M.D., Joseph Biederman, M.D., Thomas Spencer, M.D.,
and Timothy Wilens, M.D.

I. Introduction .............................................. 385

II. Attention-Deficit Hyperactivity Disorder ....................... 386
III. Major Depression ........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
IV. Bipolar Disorder. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
V. Anxiety Disorders ......................................... 404
VI. Obsessive-Compulsive Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
VII. Tic Disorders ............................................. 406
VIII. Psychosis ................................................ 408
IX. Developmental Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
X. Sleep Disturbances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
XI. Enuresis ................................................. 412
XII. Combined Pharmacotherapy ................................. 412
XIII. Conclusions .............................................. 413
References ............................................... 413

10. Psychotropic Drug Use in Pregnancy

Lee S. Cohen, M.D., Lori Altshuler, M.D., Vicki L. Heller, M.D.,
and Jerrold Rosenbaum, M.D.

I. Introduction .............................................. 417

II. Assessment of Women of Childbearing Potential . . . . . . . . . . . . . . . . . 418
A. Risks of Pharmacotherapy ................................ 418
B. Risks Associated with Psychiatric Illness. . . . . . . . . . . . . . . . . . . . 419

III. Psychotropic Drugs in Pregnancy ............................. 420

A. Antipsychotics ......................................... 420
B. Antidepressants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
1. Tricyclic Antidepressants ............................. 421
2. Serotonin-Selective Reuptake Inhibitors ................. 422
3. Other Antidepressants .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
C. Electroconvulsive Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
D. Mood Stabilizers ....................................... 423
1. Lithium ........................................... 423
2. Anticonvulsants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
E. Benzodiazepines ........................................ 425
IV. Treatment Guidelines .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
A. Psychosis ............................................. 426
B. Mood Disorders ........................................ 427
1. Major Depression ................................... 427
2. Bipolar Disorder .................................... 428
C. Anxiety Disorders ...................................... 430
1. Panic Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
2. Obsessive-Compulsive Disorder ....................... 431
V. Breast Feeding ............................................ 431
VI. Conclusion ............................................... 431
References ............................................... 432

11. Psychopharmacology of HIV-Related Psychiatric Disorders

Francisco Fernandez, M.D., and Jorge Maldonado, M.D.

I. Introduction .............................................. 441

II. Laboratory Diagnosis of HIV / AIDS Disease .................... 442
III. Neurobehavioral Manifestations of HIV Disease ................. 443
A. General Considerations .................................. 443
B. Neuropsychiatric Complications ........................... 443
1. Dementia .......................................... 445
a. Epidemiology .................................... 445
b. Clinical Features ................................. 445
c. Differential Diagnosis ............................. 445
d. Evaluation ...................................... 446
2. Depression ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
a. Epidemiology .................................... 446
b. Clinical Features ................................. 446
c. Differential Diagnosis ............................. 448
d. Evaluation ...................................... 448
3. Anxiety ........................................... 449
a. Epidemiology ............................... . . . . . 449
b. Clinical Features ................................. 450

c. Differential Diagnosis ............................. 450

d. Evaluation ...................................... 450
4. Delirium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
a. Epidemiology .................................... 450
b. Clinical Features ................................. 450
c. Differential Diagnosis ............................. 451
d. Evaluation ...................................... 451
5. Psychosis.......................................... 451
a. Epidemiology .................................... 451
b. Clinical Features ................................. 452
c. Differential Diagnosis ............................. 453
d. Evaluation ...................................... 453
6. Insomnia .......................................... 453
a. Epidemiology ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
b. Clinical Features ................................. 453
c. Differential Diagnosis ............................. 454
d. Evaluation ...................................... 454
IV. Treatment of HIV-Related Neuropsychiatric Syndromes ........... 454
A. Dementia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
1. Primary Therapy .................................... 454
2. Adjunctive Therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
3. Palliative Therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
4. Nonbiological Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
B. Depression ............................................ 456
1. Heterocyclic Antidepressants .......................... 456
2. New Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
3. Other Agents ....................................... 457
4. Nonbiological Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
C. Anxiety ............................................... 458
1. Anxiolytics ........................................ 458
2. Nonbiological Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
D. Delirium .............................................. 459
E. Psychosis ............................................. 460
1. Neuroleptics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
2. Lithium ........................................... 461
3. Anticonvulsants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
4. Nonbiological Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
F. Insomnia .............................................. 462
1. Pharmacotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
2. Nonbiological Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
V. Conclusions .............................................. 463
Selected Reading .......................................... 463

12. Eating Disorders

Andrew W. Brotman, M.D., David B. Herzog, M.D., and David C. Jimerson, M.D.
I. Introduction .............................................. 467

II. Diagnostic Considerations .................................. . 468

A. Anorexia Nervosa ...................................... . 468
B. Bulimia Nervosa ....................................... . 468
C. Binge Eating Disorder .................................. . 470
III. General Approach to Treatment ............................. . 471
A. Outpatient Treatment ................................... . 471
B. Inpatient Treatment .................................... . 471
IV. Pharmacotherapy for Eating Disorders ........................ . 471
A. Anorexia Nervosa ...................................... . 471
1. Use of Medications in AN ........................... . 471
2. Clinical Recommendations ........................... . 473
3. Outcomes ......................................... . 474
B. Bulimia Nervosa ....................................... . 474
1. Use of Medications in BN ........................... . 474
2. Clinical Recommendations ........................... . 476
C. Binge Eating Disorder .................................. . 477
Use of Medications in BED ............................ . 477
V. Conclusion 477
References .............................................. . 478

13. Borderline Disorders

Paul H. Soloff, M.D.
I. Introduction .............................................. 483
A. Psychobiologic Models .................................. 484
B. Psychosocial Models .................................... 485
II. Diagnostic Considerations ................................... 486
A. Diagnostic Criteria ...................................... 486
B. Differential Diagnosis ................................... 488
C. Course of the Illness .................................... 488
III. General Therapeutic Measures ............................... 489
A. General Nonbiological Therapeutic Approaches: Psychotherapy .... 489
B. General Issues Related to Pharmacotherapy .................. 490
1. Attrition ........................................... 490
2. Hospitalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
3. Special Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
IV. Pharmacotherapy .......................................... 491
A. Approach to Treatment .................................. 491
B. General Clinical Considerations in Pharmacotherapy of BPD . . . . 492
C. Main Effects of Medications on Target Symptoms in BPD . . . . . . 492
1. Low-Dose Neuroleptic Strategy ........................ 492
2. Antidepressants ..................................... 493
3. Anticonvulsants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
4. Lithium Carbonate .................................. 495
5. Benzodiazepines .................................... 495
V. Clinical Considerations/Applications .......................... 496
A. The Globally SymptomaticlDepressed Borderline Patient . . . . . . . 496

B. The Hostile/Depressed Borderline Patient ................... 496

C. The Schizotypal Borderline Pattern . . . . . . . . . . . . . . . . . . . . . . . . . 497
D. The Impulsive Borderline Symptom Pattern. . . . . .. . . . . . .. . . . . 497
VI. Conclusion ............................................... 497
References ............................................... 498

14. Medicolegal Psychopharmacology

Thomas G. Gutheil. M.D.
I. Introduction ............................................. . 501
II. Informed Consent ......................................... . 501
A. Consent Forms and Alternatives .......................... . 502
B. Exceptions to Informed Consent .......................... . 503
III. Liability ................................................ . 503
A. Definition ............................................ . 503
B. Some Common Forms of Liability ........................ . 504
1. Lack of Informed Consent ........................... . 504
2. Misdiagnosis ...................................... . 505
3. Inadequate Treatment ............................... . 505
4. Side Effects and Bad Reactions ....................... . 505
5. Tardive Dyskinesia ................................. . 505
6. Special Problems with Tricyclic Antidepressants ......... . 506
C. Fundamentals of Liability Prevention ...................... . 507
1. Sharing Uncertainty ................................ . 507
2. The Therapeutic Alliance ............................ . 508
3. Selection ......................................... . 508
4. Documentation and Consultation ...................... . 508
IV. The Right to Refuse Treatment .............................. . 508
A. Basic Issues .......................................... . 508
1. Emergencies ...................................... . 509
2. Vicarious Consent .................................. . 509
B. Some Modem Developments ............................. . 509
V. Guidelines for Clinicians ................................... . 510
VI. Conclusion 510
References .............................................. . 510

Index ....................................................... . 513

State of the Art
Introduction: The Practice
of Pharmacotherapy



The use of psychotropic drugs-chemicals that affect the mind-probably dates from
the dawn of humankind. Early people turned to plant products such as alcohol, opiates,
and hallucinogens to relieve pain, quell anxiety, promote sleep, or facilitate religious
experiences. Although many of these uses fell outside the realm of medical healing,
many others were among the physicians' more potent and commonly employed treat-
ment options.
Modem psychopharmacology did not begin until the middle of the 20th century.
Before then, there were few chemical compounds used to treat psychiatric symptoms
and many potential pitfalls. In the l840s, clinicians prescribed bromides as sedatives.
In the latter part of the 19th century, other sedative agents were synthesized-includ-
ing paraldehyde and chloral hydrate. During this same period, Sigmund Freud sug-
gested that cocaine was a useful psychoactive drug, and Emil Kraepelin started the first
laboratory for testing drugs in humans. At the beginning of the 20th century, re-
searchers introduced the barbiturates.
In 1940, lithium chloride was given to cardiac patients as a salt substitute, but
many developed severe toxic reactions. Cade gave lithium to animals to determine if it
increased the solubility of uric acid. Noting its sedative properties, he then gave the salt
to several manic and agitated patients. l However, because of continued concerns about
toxic reactions, the drug was not approved for use in the United States until 1970.
The modem revolution in psychopharmacology began in 1949, when Charpentier
synthesized chlorpromazine (Thorazine@ and others), and by 1952, several individuals
had noted its beneficial effects on mania, paranoia, and other psychoses. 2 They observed

Alan J. Gelenberg, M.D. • Department of Psychiatry, University of Arizona Health Sciences Center,
Tucson, Arizona 85724.
The Practitioner's Guide to Psychoactive Drugs (Fourth Edition), Gelenberg and Bassuk, eds.
Plenum Publishing Corporation, New York, 1998


that patients improved primarily because of a decrease in disorganized thoughts, feelings,

and behaviors rather than because they were sedated. Chlorpromazine was introduced for
general use in the United States in 1954 and created a revolution in patient care. 3
Once researchers identified the "neuroleptic" properties of chlorpromazine, they
began to develop similar antipsychotic agents. In 1958, while attempting to find a more
effective analgesic agent, Janssen synthesized haloperidol (Haldol® and others), a
butyrophenone with strong neuroleptic properties. Since then, other researchers have
discovered additional classes of antipsychotic drugs, including the thioxanthenes, the
dibenzoxazepines, and the dihydroindolones.
In the 1950s, meprobamate (Miltown® and others) became more popular than the
barbiturates for relieving anxiety and promoting sleep. However, physiological toler-
ance, dependence, and severe withdrawal reactions were as much of a problem as they
had been with the barbiturates. In 1957, Sternbach synthesized the benzodiazepine
chlordiazepoxide (Librium® and others).4 The finding that it calmed animals without
producing marked sedation spurred clinical trials in humans. Since then, various ben-
zodiazepines have largely replaced other drugs for the treatment of anxiety and insom-
nia owing to their clinical effectiveness and their broad margin of safety.
As with most other psychotropic agents, the antidepressant drugs were identified
fortuitously. For example, clinicians noticed that tuberculosis patients treated with
iproniazid experienced elevated mood; this led to further investigation of the mono-
amine oxidase inhibitors (MAOIs). In 1952, these agents were administered to patients
with depression, and by the late 1950s, this class of drugs became part of common
clinical practice. However, when severe hypertension developed in a few patients,
MAOIs fell into disfavor. Better knowledge about the "MAOI diet," clearer clinical
guidelines, and evidence of their effectiveness in depression and phobic states have
rejuvenated their use.
The tricyclic antidepressant imipramine was initially synthesized in the 1940s as a
promazine analogue. In 1958, Kuhn reported its effectiveness as an antidepressant. 5
Soon afterward, other tricyclic antidepressants with similar properties were synthe-
sized. The 1980s and 1990s have seen the introduction of several chemically and
pharmacologically distinctive antidepressants, including the serotonin-selective reup-
take inhibitors (SSRIs).
In recent years, syndromes previously resistant to any form of treatment (e.g.,
obsessive-compulsive disorder) have yielded to new drugs [e.g., clomipramine (An-
afranil®) and, more recently, SSRI antidepressants]; syndromes thought not to be
responsive to medication (e.g., social phobia, dysthymia) have responded to previously
available agents; and even long-standing behavior problems previously thought to be
immutable characteristics of temperament (e.g., shyness, inhibition, chronic depressive
attitudes) have responded to pharmacotherapy. These discoveries have occurred, in
part, because investigators have not allowed terminology to limit the use of various
agents (Le., by using antidepressants for disorders other than depression). And seren-
dipity continues to playa role, as new drugs are found to be efficacious for disorders
beyond their initial indications. In Listening to Prozac, Kramer contends that fluoxetine
can change components of personality, making people feel "better than well."6 Wheth-
er this is true, and whether psychotropic drugs should be prescribed for this purpose,

may soon be determined, as investigators construct new and more complex models of
how psychotropic drugs work in the brain, and as clinical researchers examine the
interface of pharmacotherapy and psychotherapy and the applications of both in the
long-term treatment of chronic disorders.


A. Definition of Clinical Syndromes

Although many diverse diagnostic frameworks exist, they often do not help the
practitioner decide whether or not to prescribe drugs. In the past, diagnosis usually was
based on presumed cause (e.g., endogenous versus reactive depressions). More recent-
ly, the third edition of the Diagnostic and Statistical Manual of Mental Disorders
(DSM-III) was developed to define discrete diagnostic groups according to symptom
profiles and inclusion and exclusion criteria. This framework, carried forward through
DSM-III-R to the current DSM-IV, provides reliable diagnoses and consistent criteria
for research and encourages accurate communication among professionals.? However,
it does not specify drug-responsive disorders. Some researchers, studying the outcome
of specific drug treatment, have delineated symptom profiles that respond to phar-
macotherapy. Unfortunately, despite these advances, many patients still do not receive
effective medication trials. Some individuals with similar presenting symptoms have
disorders requiring different drugs, whereas other patients with disparate symptoms
may respond to similar treatments. To remedy this problem, investigators are trying to
widen their knowledge of drug-responsive presentations and ideally to develop biolog-
ical methods of diagnosis.

B. Development of Brain Chemistry Models and New

Pharmacological Agents
A significant advance in psychopharmacology has been the evolution of biolog-
ical models of disease based on increased knowledge of brain chemistry. Despite their
limitations, the dopamine hypothesis for schizophrenia and the catecholamine hy-
pothesis for depression have provided useful starting points for research. For example,
biochemical models of depression have undergone various modifications to account for
evidence that norepinephrine (NE), serotonin (5-HT), dopamine (DA), and acetylcho-
line may each have a role in depression. The brain chemistry models have stimulated
the search for drugs that produce hypothetically effective neurohumoral changes and
fewer adverse reactions. On the other hand, new compounds whose effects are not
consistent with postulated biochemical alterations [such as the antidepressant bupro-
pion (Wellbutrin®), which does not block the reuptake of either 5-HT or NE] promote
the revision of existing hypotheses. In the pharmacotherapy of depression, research is
focusing on variable agonist and antagonist properties at serotonin receptor subtypes
and on drugs' abilities to affect these presumed sites of efficacy while minimizing the
receptor-blocking properties that led to so many side effects in the tricyclic antidepres-

sants. In the treatment of psychosis, pharmaceutical companies attempt to create drugs

that have the enormous benefits of clozapine (Clozaril®) without its toxicity. These
attempts have led to a focus on neurotransmitter systems other than dopamine (most
notably, serotonin and its receptor subtypes) and the relative ratios of a drug's effects
on these different systems. At present, these advances have only limited clinical useful-
ness, but the future promises the development of more definitive approaches.

C. Design of Drug Studies

The methodology of clinical investigations in pharmacology has improved vastly
over the past few decades. The introduction of double-blind techniques, blind raters,
"active placebos," and sophisticated analyses has brought the light of science to
previously insoluble clinical dilemmas. Nevertheless, many methodological problems
still beset clinical pharmacology. Drug studies are expensive: it may take as much as
$100 million to bring a drug to market. This means that if a drug does not appear to
have potential for profit, it may never be adequately studied. Because pharmaceutical
companies sponsor a large majority of drug trials, most studies are designed to show
efficacy (Le., superiority of the new compound over placebo) and comparability to
existing medications, whereas clinicians and patients may be at least as interested in
differences (both positive and negative) from existing treatment options.
The validity of a clinical study depends on the diligence of an investigator in
selecting appropriate patients and adhering rigorously to a study protocol. However,
even careful investigators can get different results when studying the same drug be-
cause many psychiatric disorders, such as unipolar depression, are heterogeneous. An
additional concern is that trial populations are often comparatively "pure" -e.g., pa-
tients at either end of the age spectrum, pregnant women, and patients who have
comorbid medical problems, use other medications, or abuse drugs or alcohol are
excluded. In the "real world," these patients, as well as those with less pure psychiatric
diagnoses, will probably make up a majority of those for whom physicians prescribe
psychotropic drugs.
Another problem is that of chronicity. Until recently, the majority of medication
trials were short-term, whereas pharmacotherapy for most psychiatric disorders lasts
for years. Rare but serious adverse drug effects or those more likely after long-term
treatment tend not to be detected until a medication has been released for general
clinical use. Because there are few systematic techniques for accurately identifying and
quantifying such reactions, one drug may be falsely incriminated as causing a serious
toxic reaction while another may be suspected of causing an effect but may not be
accurately labeled because insufficient evidence exists to force the manufacturer to do
so. Despite these difficulties, however, a surprising consensus has emerged from re-
search in recent years, and if sufficient funding is available for future studies, prospects
appear bright.


The clinical practice of psychopharmacology represents the traditional blend of art

and science in medicine. The art entails, among other things, the ability to establish

rapport with a patient. This increases the quantity and reliability of information from
the patient, leading to a diagnosis, and solidifies the relationship on which compliance
with the medication regimen will be built. The clinician's "third ear" will help detect
inconsistencies, disparities, and missing pieces of data that might change diagnostic
impressions or highlight problems in following a medication regimen. A sophisticated
practitioner will recognize the nuances among different personalities and be sensitive
to their interactions with medication-responsive psychiatric syndromes.
At the same time, the science that grounds the clinician's practice continues to
expand and demands attention to a growing body of knowledge. The up-to-date practi-
tioner must recognize the increasing number of psychiatric syndromes identified as
responsive to medication and those drugs currently indicated for such disorders. He or
she must know which features of an illness are likely to respond to drug treatment and
be aware of adjuncts or second- or third-line treatments in case the first fails to produce
expected improvement. The pharmacokinetics of drugs and their potential interactions
are becoming more important, and clinicians have to maintain familiarity with medical
conditions and substance abuse problems (and their treatments) and how these might
interface with the presentation of psychiatric disorders and their treatment.

A. Matching Medications with Drug-Responsive Syndromes

Like most other medical specialties, clinical psychopharmacology tends to be
empirically derived. Thus, the clinician should learn which clinical syndromes are apt to
be responsive to medications and remain abreast of the developing literature on phar-
macotherapeutics. Fifty years ago, there was no psychiatric syndrome known to respond
to medicine. By the 1960s, it was becoming apparent that medications were indicated in
the treatment of bipolar disorder, most psychotic disorders, and many forms of depres-
sion, as well as for symptomatic relief of anxiety and insomnia. More recently, medica-
tion appears to be a primary treatment for obsessive-compulsive disorder, panic disor-
der, social phobia, and chronic and milder forms of depressive illness.
In the mid-1990s, psychiatry remains a field based largely on descriptive phenom-
enology. We define a syndrome based on history and mental status, and we select a
drug treatment based on the fact that many other people with a similar constellation of
symptoms have responded significantly better to the medicine than to placebo. Some
patients who meet diagnostic criteria of these sorts will fail to respond to even the most
aggressive and progressive pharmacotherapy. Others may improve with nondrug treat-
ments or possibly even without any treatment. Someday, we may have biological tests
that can dissect these syndromes into homogeneous disease states. Ideally, we will be
able to tailor a particular drug to a patient's underlying or biological deficit, just as the
infectious disease specialists can select an antimicrobial based on a culture sensitivity.
But today we select agents and sequential medication trials based primarily on clinical
investigation and statistical probabilities, taking into account a patient's history or the
family history of treatment response. Although there will always be room in therapeu-
tics for clinical intuition and the practitioner's personal experience, one must be leery
of idiosyncratic practices based on ill-informed impressions without reasonable scien-
tific validation. The rationale for innovative and unusual treatments should be detailed
in the clinical record as should an ongoing and interactive process of informed consent.

B. Pharmacotherapy and Psychotherapy

When the impact of the modem psychopharmacology revolution became obvious

in the 1960s, many perceived a threat to the psychoanalytic schools of thought that
largely held sway in American psychiatry. At that time, many contended, practiced, and
taught that virtually all psychopathology-from schizophrenic psychosis to low-grade
anxiety-was caused by early life experiences and definitively remediable only
through insights gained in long-term analytic therapy. Medication was considered at
best palliative, treating the symptoms, not the cause. At worst, it was thought to be
counterproductive, diminishing the anxiety necessary for insight. For many practi-
tioners, dogma took the place of science. To question was to doubt, which was almost
equivalent to religious heresy. The treatment a patient received was more likely to be
determined by which doctor's door he or she entered than by science and diagnosis.
Such dogmatism was likely the product of natural human tendencies to resist
change, eschew ambiguity, and embrace simplistic answers. As in many areas of
medicine, schools of psychiatric thought sometimes became rigidified into belief sys-
tems unapproachable by the empirical questioning of science. At a more personal level,
clinicians trained in one theoretical model and set of practices felt threatened by newly
emerging treatment technologies.
A fundamental conceptual misunderstanding often underlies the artificial dichot-
omization inherent in the question of whether a patient should have psychotherapy or
medication treatment. Too often, laypeople and even professionals assume the answer
lies in whether a symptom or syndrome is "biological" or "psychological." In fact, each
person exists in a multitude of abstractions: for example, as a psychological, spiritual,
biological, and social being. (For that matter, our physical existence can be abstracted
from the atomic level to the complex systems.) No psychiatric syndrome exists on only
one of these levels: they exist on all. One may genetically inherit, for example, a
tendency to mood swings or paranoid delusions. Yet whether, when, and how severely
the syndrome of behavior pathology gets expressed may be aggravated or attenuated by
early (and later) life experiences. The genetic endowment, prenatal environment, and
physical and psychological environmental postnatal exposures affect the brain's struc-
ture and function; resultant behaviors, in tum, affect the environment, producing com-
plex and dynamic feedback loops. A baby's temperament, for example, affects parental
behavior, which, in tum, affects the translation of temperament into character.
Therapeutic intervention, similarly, occurs at multiple levels. Talking to someone,
whether casually or in formal psychotherapy, affects neurobiological as well as psycho-
logical processes-just at different levels of abstraction. For example, a medication
can affect perceptions and feeling states, which in tum may lead to different assump-
tions, behaviors, and interpersonal interactions.
These complex multidimensional conceptualizations can help the clinician avoid
simplistic thinking and artificial dichotomies about treatment choices. As noted earlier,
pharmacotherapy can never totally exclude psychotherapy, as the very interaction
between a clinician and a patient must be therapeutic for the patient to comply with the
biological regimen. Rather, one must determine the nature of the interaction and wheth-
er some form of theoretically derived and formal psychotherapy is appropriate.

A growing scientific literature defines and assesses the roles of various forms of
rigorously described psychotherapies in the treatment of phenomenologically defined
psychiatric syndromes. Interpersonal and cognitive-behavioral therapies are the best
studied. Both were initially derived from psychoanalytic constructs and clinical obser-
vations. Unfortunately, more formal psychoanalytic treatments have not been systemat-
ically studied as much, although many experienced psychiatrists value the theories and
The current literature suggests that for some patients and certain syndromes (as
will be described in detail in the subsequent chapters), medication combined with
minimal psychotherapy is sufficient for symptomatic relief. Increasingly, though, expe-
rienced clinicians recognize that long-term compliance with treatment requires a re-
spectful, interactive relationship, often with significant others as well as the patient, and
abundant opportunities for exchange of information that allow the patient to be an
informed consumer. For some conditions, psychotherapy is as effective as phar-
macotherapy, particularly for those at the milder end of symptomatic spectra. To date,
no research has established superior efficacy for one type of psychotherapy over
another in any defined disorder. What the effective components of psychotherapy are-
such as warmth, support, insight, behavior change, or altered cognitive schemata-
remain to be understood. At a minimum, practitioners should attempt to identify
psychological and interpersonal misunderstandings and dysfunctional patterns of
thought and behavior that increase stress and lower resistance, leading to symptomatic
exacerbation of long-term psychiatric disorders.
In summary, many patients with psychiatric disorders can be treated with psycho-
therapy alone, but the decision not to prescribe medication should be based on up-to-
date knowledge about the patient's diagnosis and condition. Moreover, prolonged
psychotherapy alone should not be used in disorders known to be responsive to medica-
tion without considering drug therapy. More severe disorders known to respond to
medication should be considered for medication therapy, often along with psycho-
therapy, from early clinical contact. Of course, patient preference must be taken into
account, but the clinician is obligated to provide up-to-date information about treat-
ment options for the condition. The clinician is similarly obligated to remain abreast of
the latest knowledge about types of therapy known to be helpful for specific psychiatric
disorders. For many psychiatric patients, particularly those with serious and chronic
mental illnesses, some form of psychotherapy in tandem with medication treatment is
worthwhile, whether provided by the same clinician or two different ones.


A. Patient Evaluation
Evaluating a patient for psychoactive medication is inherently similar to any other
medical assessment. The clinician should begin with a comprehensive history, includ-
ing a careful review of medical history, other medications, and recent medical changes
and-very important-a review of systems. Observations of family members or other

observers of the patient's behavior can be highly valuable, particularly for a psychiatric
diagnosis. A family history of medical, neurological, and psychiatric disorders often
will affect diagnostic considerations and treatment decisions. Knowledge of a patient's
past treatment successes and failures will be important, not only from a biological
perspective but also to determine a patient's comfort with various kinds of interperson-
al interactions. The patient should undergo a physical examination, unless one has been
done recently, and various laboratory tests, as determined by the clinician's differential
diagnosis. Although there are no laboratory tests that specifically rule in a psychiatric
diagnosis, many-such as the assessment of endocrine function or brain imaging-can
rule out other diagnoses that could produce emotional and behavior changes. As in all
medical diagnoses, hypotheses tend to be framed in the course of history taking, and
these should be worked and reworked as more information becomes available.

B. Principles of Drug Use

1. Understand the Pharmacokinetics of Psychotropic Agents
The rate of drug absorption from the gastrointestinal tract largely determines the
speed of onset of action after a single oral dose. The length of time a drug's effects last
after a single oral dose is determined primarily by the rate and extent of the drug's
distribution. For example, highly lipid-soluble compounds tend to be rapidly and
extensively distributed throughout the body's tissues, which indicates a relatively brief
duration of clinical effects following a single dose.
After repeated dosing. however, the elimination half-life becomes clinically impor-
tant in determining the duration of a drug's effects. The elimination half-life determines
the rate at which a drug accumulates in body tissues. Drugs with longer half-lives
accumulate more gradually (steady-state concentrations are achieved after approx-
imately four half-lives) and disappear from the body gradually following discontinua-
tion. Those drugs with shorter half-lives, on the other hand, are eliminated more rapidly
(which for some drugs can result in the rapid appearance of an intense withdrawal
syndrome). The half-life also can guide a clinician in choosing the frequency of dosing
intervals: drugs whose half-lives exceed 24 hr usually can be administered once each day.
With the exception of lithium. virtually all psychoactive drugs undergo bio-
transformation by the liver's hepatic microsomal enzymes. The presence of liver dis-
ease can affect the rate at which a drug is broken down and hence the level of the drug
and its metabolites in the blood. The recent recognition of which isoenzymes are
involved in the metabolism of different drugs helps in the prediction of potential drug
interactions that can have clinically important effects on blood levels of the different
compounds (see Table 1). The hepatic isoenzymes appear to be under genetic control,
and growing evidence suggests that some populations have a tendency to be slow
metabolizers of certain agents. This can have profound effects on dosing and body
Virtually all psychotropic drugs and their breakdown products are excreted pri-
marily by the kidneys. For these drugs, diminished kidney function can have clinically
important and serious implications for drug dosing and adverse effects.

TABLE 1. P450 Isoenzymes-Potential Drug InteractionS"

Isoenzyme Metabolizes Inhibited by Other features

206 Alprenolol, amiflamine, Amitriptyline, clomipramine, 5-10% of Caucasians are

amitriptyline,b desipramine, fluoxetine, slow metabolizers (do not
brofaromine, bufarolol, fluphenazine, haloperidol, have 206 gene); 5-20% of
clomipramine, b clozapine, paroxetine, quinidine, Caucasians are fast
codeine, debrisoquin, sertraline (less potent than metabolizers (have two
desipramine, other SSRIs), thioridazine copies of 206 gene)
encainide, ethylmorphine,
guanoxan, haloperidol,
idoramin, imipramine, b
metoprolol, mexiletine,
paroxetine, perhexiline,
perphenazine, phenformin,
propafenone,b propranolol,
reduced haloperidol,
remoxipride, risperidone,
Sertraline, sparteine,
terfenadine, b thioridazine,
timolol, tomoxetine,
trazodone, trimipramine,b
IA2 Amitriptyline, caffeine, Caffeine, Auvoxamine,
clomipramine, clozapine, grapefruit juice
fluvoxamine, haloperidol,
imipramine, paracetamol,
phenacetin, propranolol,
R-warfarin, tacrine,
theophylline, b verapamil
3A3/4 Alprazolam, amitriptyline, Astemizole, fluoxetine, Potentially dangerous
astemizole,b bupropion?, fluvoxamine,c grapefruit arrhythmias for those
carbamazepine, cisapride, juice, itraconazole,c.d taking some nonsedating
clomipramine, clozapine, ketoconazole, c.d antihistamines; tricyclics
cortisol, cyclosporine, nefazodone, c.d sertraline and venlafaxine less likely
dexamethasone, to inhibit; carbamazepine
dextromethorphan, induces


TABLE 1. (Continued)

Isoenzyme Metabolizes Inhibited by Other features

diltiazem, Erythromycin,
ethinylestradiol, felodipine,
imipramine, lidocaine,
midazolam, nefazodone,
propafenone, quinidine,
sertraline, tamoxifen,
terfenadine, b triazolam,
venlafaxine, verapamil,
2C Amitriptyline, citalopram, Auoxetine, fluvoxamine, 18% of Japanese, 19% of
clomipramine, sertraline African Americans, 8% of
desmethyldiazepam, Africans, and 3-5%
diazepam, hexobarbital, Caucasians are poor
imipramine, metabolizers
moclobemide, omeprazole,
phenytoin,b proguanil,
tolbutamide, warfarinb

"The information in this table is gleaned from both in vitro and in vivo studies and case reports, which are not always consistent
with each other. The interactions of drugs with the P450 isoenzymes are currently the subject of rigorous clinical investigation.
In particular. there is controversy over which serotonin-selective reuptake inhibitors (SSRIs) inhibit which isoenzymes at what
doses and to what degree.
bHas a narrow therapeutic index. Monitor plasma levels closely and consider decreasing dose when coadministered with one of
the drugs that inhibits this enzyme.
<Contraindicated with the nonsedating antihistamines terfenadine and astemizole.
dContraindicated with cisapride.

2. Prescribe the Simplest Drug Regimen to Increase Compliance

For a drug to be effective, a patient must take it as prescribed and promptly report its
effects, both positive and negative, to the clinician. Education-about illness and
medicine-is a crucial factor toward establishing an interactive relationship. Another
important factor is the practitioner's respect for the patient and capacity to build trust,
individualized to the patient's temperament and sophistication. Beyond these essential
elements, compliance may be additionally enhanced by simplifying the medication
regimen. For most psychotropic agents, once-daily dosing is possible and easier to
remember. Some medications need to be given in divided doses, however, to limit the
risk of adverse effects. When multiple daily doses or combinations of medications
become part of a treatment regimen, compliance can be enhanced by the use of
medication boxes, calendar notations, or other mnemonic devices.

3. Use the Fewest Number of Drugs to Achieve the Desired Result

The larger the number of medications in a regimen, the greater is the potential for
interactions, adverse effects, and confusion. Although clinicians sometimes develop
their own impressions of what works best for whom, combining medicines within the
same class-e.g., antipsychotics, tricyclic antidepressants-is seldom indicated and has
not been proven beneficial. Sometimes, medicines are added as physicians respond to the
desperation or frustration of patients, other caregivers, or family members. Often it is
better to counsel patients about the gradual onset of a drug's effects. Like the tortoise who
defeated the hare, slow and steady sometimes wins the race.
Occasionally, the use of multiple medications can be avoided by carefully docu-
menting treatment results. Identifying target symptoms that may be attenuated by a
medication and quantifying the symptoms and the treatment effect can go a long way
toward answering the questions: has this drug been helpful and, if so, how much? That
way, a new drug can be substituted for an ineffective or suboptimal medication rather
than the too-frequent practice of adding yet another medication to a regimen.
Having stated those caveats and qualifiers, there are many instances in modem
psychopharmacology in which polypharmacy is just what the doctor should order. The
following are illustrations:
1. When a potentially beneficial synergistic effect outweighs the risk of an
adverse reaction. For example, sometimes antidepressants alone do not allevi-
ate depression but may relieve the patient's symptoms with an adjunct such as
2. When the diagnosis calls for a particular combination. For instance, the
clinician may administer an antidepressant and an antipsychotic agent to
depressed patients with psychotic symptoms.
3. When adverse reactions caused by a psychotropic drug require treatment. For
example, antipsychotic-drug-induced extrapyramidal reactions may be treated
with antiparkinson drugs.
4. When changing from an acute regimen to maintenance treatment, during which
change the use of two preparations may overlap. In schizophrenia, a depot
antipsychotic may be started while the patient is still receiving an oral anti-
5. When a patient has two significant symptoms that do not respond to one drug.
Sleeping medication may be indicated in depressed patients with persistent
insomnia, even after they have been started on an antidepressant regimen.

4. Provide the Most Cost-Effective Treatment

Theoretically, the clinician should help patients save money by prescribing drugs
by their generic names. Unfortunately, the bioavailability (i.e., the amount of drug
absorbed into the plasma) of different preparations may vary significantly. In practice,
the clinician should try to prescribe the least expensive and most appropriate drug first
and then discuss possible problems of bioavailability with the patient. In addition, he or
she should encourage the patient (and/or family) to inquire about possible changes in the

brand when prescriptions are refilled and to monitor closely the appearance of adverse
reactions and the reemergence or worsening of clinical symptoms. Because of the
possible differences in effectiveness of the preparations administered, the practitioner
may also have to adjust the dosage.
In addition to being easier for the patient, once-a-day dosing can also be more cost-
effective, because it allows the use of higher-strength tablets or capsules. A lOO-mg
tablet, for example, tends to cost little more than a 50-mg tablet. Therefore, a single daily
dose of 100 mg can be more cost-effective than two 50-mg doses. If a patient is
hospitalized, less frequent dosing also means less staff time preparing and administering
the compound. If the dose must be divided, a lOO-mg tablet can be split in half to keep
expenses down.
A therapy's cost-effectiveness cannot always be measured strictly by the cost of the
medication doses. If side effects are unacceptable to a patient, for example, and if he or
she therefore stops taking the medication, it will obviously be ineffective. In such cases,
the cost of the illness is usually higher than the price of the medicine. If a medication
produces side effects such as sedation or ataxia, which lead to the risk of automobile
accidents or falls, the costs of possible injuries must be factored into the overall cost of
the medication. The savvy practitioner, therefore, becomes familiar not only with the
price of drugs but also with their relative value to a patient, based on treatment
effectiveness and adverse events.

5. Exercise Special Care with Medically III Patients

Medically ill patients often are older and therefore have an increased risk of adverse
reactions and a decreased tolerance to medication. The elderly often are taking medica-
tions that may interact with psychoactive agents. The medical condition itself may
represent a relative contraindication to drug treatment or may obligate the clinician to
alter the dosage or methods of administration significantly. Because of these limitations,
the clinician should:
1. monitor possible drug interactions, adverse and toxic reactions, and special
metabolic problems;
2. assess the role of the medical condition in the etiology of the psychological
symptoms; and
3. administer the lowest effective dose.


As drug-responsive syndromes become more clearly defined, practitioners increas-

ingly use pharmacotherapy as the primary treatment for many psychiatric disorders.
Based on clinical experience and new research findings, medication options have greatly
increased. Many new agents with fewer adverse reactions, lower toxicity, and somewhat
different effects on brain chemistry have become available. These advances have
improved patient care, and the future holds even greater promise. Increased understand-
ing of brain chemistry and pharmacokinetics will significantly change clinical practice,

and, as investigators define the interactions between biological and nonbiological

interventions, more comprehensive treatment approaches will evolve.

1. Cade J. F. J.: Lithium salts in the treatment of psychotic excitement. Med J Aust 2:349-352, 1949.
2. Laborit H., Huguenard P., Alluaume R.: Un nouveau stabilisateur vegetatif, Ie 4560 RP. Presse Med 60:206-
3. Lehmann H. E., Hunrahan G. E.: Chlorpromazine, a new inhibiting agent for psychomotor excitement and
manic states. Arch Neurol Psychiatry 71 :227 -237, 1954.
4. Sternbach L. H., Randall L. 0., Gustafson, S. R.: 1,4-Benzodiazepines: Chlordiazepoxide and related
compounds, in Gordon M. (ed): Psychopharmacological Agents. New York, Academic Press, 1964, vol 1,
pp 137-224.
5. Kuhn R.: The treatment of depressive states with G22355 (imipramine hydrochloride). Am J Psychiatry
6. Kramer P.O.: Listening to Prozac. New York, Viking, 1993.
7. Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, D.C., American Psychiatric
Association, 1994.
Major Psychiatric Disorders




Major depression is a common, serious, and costly disorder. More than just feeling sad,
a depressed person may be plagued by self-doubt, guilt, and recriminations and lose
interest in formerly pleasurable activities (including eating and sex). Sleep, weight, and
other bodily functions may be altered, and the patient might lose touch with reality,
often attempting or actually succeeding in committing suicide. A recent survey found
the lifetime prevalence of major depression to be extremely variable among different
geographic locales, ranging from 1.5% in Taiwan to 19% in Beirut.) In this survey, the
rate in the United States was 5.2%. The Epidemiologic Catchment Area study in five
U.S. communities in the early 1980s reported a I-month prevalence of 1.6% for men
and 2.9% for women. 2 More recently, the National Comorbidity Survey found a 30-day
prevalence of 4.9% and a lifetime prevalence of 17.1 % for major depressive episode. 3
Unfortunately, in a survey it is impossible to distinguish a true difference reflecting
culture or genetics from artifact and diagnostic variables. Less in question is the
female-to-male ratio, which was approximately 2: 1 in all countries surveyed. For at
least half of those who experience a major depressive episode, similar episodes will
recur in the future, at times frequently and to a debilitating degree, and in some taking
on a chronic course without periods of remission. Fortunately, the increasing availabili-
ty of modem therapeutics, both biological and nonbiological, allows most people who
suffer from depression the chance to be effectively treated.

Alan J. Gelenberg, M.D., and Pedro L. Delgado, M.D. • Department of Psychiatry, University of
Arizona Health Sciences Center, Tucson, Arizona 85724.
The Practitioner's Guide to Psychoactive Drugs (Fourth Edition), Gelenberg and Bassuk, eds.
Plenum Publishing Corporation, New York, 1998



A. Early Models and Historical Perspective

Melancholia has accounted for severe human suffering throughout history. From
the time of Hippocrates and Galen until the late 17th century, melancholia was ex-
plained by an excess of black bile and treated with emetics, purgatives, and bloodlet-
ting. In the 17th and 18th centuries, models of depression emphasized hydrodynamic
imbalances of blood affecting portions of the brain. During the 19th century, both
heredity and life experiences were considered factors in the etiology of melancholia.
Treatment approaches focused on attempts to distract and positively engage patients in
physical, interpersonal, and intellectual pursuits.
In the late 19th and early 20th centuries, Emil Kraepelin believed that heredity
was the main cause of depression. However, he also noted that "psychic" influences
such as the death of a spouse and other stresses could precipitate episodes and that
"environmental" influences such as syphilis could also be a cause.

B. Contemporary Models
1. Psychological

In his now classic paper "Mourning and Melancholia," Freud emphasized intra-
psychic conflict and life experience in the etiology of depression. 4 Abraham postulated
intense oral fixation in depressed individuals and differentiated grief from morbid
depression. 5 Moreover, he viewed depression as aggression turned toward the self.
Freud embraced these ideas but elaborated the concept of object loss. 4 He contended
that the real or fantasied loss of an ambivalently loved person precipitates an ego
regression with introjection and intense anger toward the lost object. This process
results in self-deprecating, guilty feelings. Bibring viewed depression as a discrete ego
state resulting from an individual's inability to achieve his ego idea1. 6 He believed that
aggression was not the driving force behind depression; depleted self-esteem was the
primary issue and secondarily led to hostility.
Other researchers stressed the role of deprivation during critical developmental
stages. The Harlows, Spitz, Ainsworth, Bowlby, the Robertsons, and Mahler empha-
sized early attachments and the process of separation-individuation.7. 8 Early depriva-
tion resulting from the loss of or separation from a major attachment (i.e., mothering
figure) may have permanent ramifications if it occurs during a critical period and is not
replaced by a meaningful relationship. Spitz observed that infants separated from their
mothers during the first year of life developed various symptoms that progressed from
apprehension and crying to withdrawal, motor slowing, despair, and, finally, detach-
ment and even retarded physical and emotional development. This "anaclitic depres-
sion" was also described by Robertson and Bowlby in older children and by Harlow in
primates. Clinical studies have since corroborated the importance of developmental
deficits in the etiology of depression.
From the perspective of self-psychology, patients who have a poor sense of self
and a lack of soothing self-objects frequently feel overburdened, helpless, worthless,

and depressed. 9 In clinical practice, those with severe narcissistic disorders and bor-
derline states often develop adult "anaclitic" conditions marked by feelings of empti-
ness, loneliness, and lowered mood, even in the presence of others. 10
From foundations in Ellis's rational emotive therapy and ego psychology, Beck et
al. envision depression as primarily a disorder of thinking. I I Like the psychoanalysts,
Beck postulates that distorted views of the self, the world, and the future result from
early learning. This negative set consists of disordered cognitive "schemata" and
cognitive "errors" that are stimulated by events in the patient's life and cause the
patient to feel hopeless and helpless in the face of life's difficulties. Beck, in a way,
bridges the psychodynamic and behavioral models.
Seligman's concept of "learned helplessness"12 is based on the hypothesis that
people get depressed when they feel a loss of control over positively reinforcing
experiences. Individuals exposed to unavoidable aversive stimuli may respond with
passivity and helplessness to similar "stresses" in the future. In addition, this reaction
can become a generalized personality trait; the individual believes that all adaptive
efforts will eventually fail. Lewinsohn et al., 13 like Seligman, view the lack of reinforc-
ing stimuli as central to the development of depression. However, they state that the
important deficit is the low number of positive reinforcements "contingent" on the
efforts of the individual. Moreover, according to this model, patients exposed to pun-
ishing or nonreinforcing responses from their environment may stop behaving in ways
that will yield rewarding responses.
Some authors have focused on social and existential themes. They emphasize not
only the importance of relationships but also the broader social context. Problems with
role status, individual purpose and meaning, and cultural myths may contribute to a
person's sense of alienation and despair. 14

2. Biological

Biological hypotheses of the etiology of depression have evolved over the past 40
years, paralleling the growth of the neurosciences and the development of effective
medication treatments. Theoretical formulations followed the fortuitous discovery of
agents that alleviated severe depression.
Initial assumptions underlying early neurobiological models of depression were
greatly influenced by the discovery of the dopamine deficiencies in the central nervous
system (eNS) of patients suffering from Parkinson's disease l5 and by the remarkable
therapeutic effects of the dopamine precursor L-dopa. 16 These groundbreaking discov-
eries were reported two years after dopamine was first identified in the basal ganglia. 17
It was in this historical context that the therapeutic and pharmacological effects of
antidepressant drugs were discovered and the first neurobiological theories of depres-
sion established.
The catecholamine deficiency hypothesis of depression was based on the observa-
tion that many antidepressant drugs increased synaptic concentrations of norepineph-
rine, whereas the catecholamine-depleting drug reserpine seemed to cause depres-
sionlike symptoms. 18, 19 This hypothesis proposed that depression was caused by a
deficiency of norepinephrine, and mania by an excess. The indoleamine hypothesis

stated that a deficit of brain serotonin was responsible for depression, whereas drugs
that increased synaptic serotonin, such as monoamine oxidase (MAO) inhibitors or
serotonin precursors such as 5-hydroxytryptophan (5-HTP) and L-tryptophan, relieved
depression. 2o- 22
The deficiency hypotheses have been only partially supported. 23 Deficiencies of
norepinephrine or serotonin or their metabolites in cerebrospinal fluid (CSF), blood, or
urine have not been consistently demonstrated in depressed patients despite intensive
efforts to do SO.24 A subgroup of depressed patients with a history of impulsivity or
suicide do appear to have decreased CSF levels of the primary metabolite of serotonin,
5-hydroxyindoleacetic acid (5-HIAA). However, decreased CSF 5-HIAA is also found
in subjects with other diagnoses. 25 .26
As new data emerged, it became clear that alterations in CNS neurotransmitter
system function could occur by changes in presynaptic and postsynaptic receptor
sensitivity without the amount of the neurotransmitter itself altering. Thus, the defi-
ciency hypotheses were modified, and the "receptor sensitivity hypothesis" of anti-
depressant action was proposed. This hypothesis stated that the delayed therapeutic
effects of antidepressant treatment were related to time-dependent alterations in cate-
cholamine and indoleamine receptor sensitivity and implied that the pathophysiology
of depression may be more related to abnormal regulation of receptor sensitivity than to
deficiencies of a neurotransmitter. 24 The "dysregulation hypothesis," subsequently pro-
posed by Siever and Davis,27 stated that in affective disorders, regulatory or homeosta-
tic mechanisms controlling neurotransmitter function were dysregulated and effective
pharmacologic agents would restore normal regulation to these systems. 27
The dysregulation and receptor sensitivity hypotheses went beyond neuro-
transmitter deficiencies, proposing that functional deficits in neurotransmission could
occur with normal neurotransmitter content. These hypotheses specifically focused on
the monoamine systems and, until recently, tended to conceptualize depression as a
dysfunction in a single neurotransmitter system. These theories assumed that the neuro-
biological basis of depression involved the opposite neurochemical state to that caused
by antidepressant drugs and, therefore, antidepressant treatments worked by reversing
an actual or functional monoamine deficit state.
These models are limited because they are based on simplistic models of the
nervous system, focusing on a single neurotransmitter system. They are also limited by
the assumption that the neurobiological cause of depression is a mirror deficit of the
pharmacologic effects of antidepressants. This is comparable to assuming that the
cause of rheumatoid arthritis is a corticosteroid deficiency.

3. Integrative
More recent models of the neurobiological basis of depression have attempted to
integrate data from a variety of sources, including clinical observations, epidemiologi-
cal studies, functional brain imaging studies, human neurobiological challenge studies,
and animal models. These new models have tried to account for the diversity of clinical
presentation and course of illness in the context of what is known about the brain
systems that modulate affective and emotional states.

A consensus has now developed that major depression is not a single disorder or
disease but a syndrome. 28 This is based on the notable variation in the age of onset,
course of illness, and symptom profile of major depression. 29 A variety of structured
diagnostic instruments and classification nomenclatures have been developed to aid in
standardizing diagnosis. The concept of "major depression" has emerged from this
process as the umbrella diagnosis under which this broad spectrum of symptoms have
been grouped. As described in the fourth edition of the Diagnostic and Statistical
Manual of Mental Disorders (DSM-IV), the diagnosis of major depression is made
when a mixture of symptoms are present for most of the day, nearly every day during a
2-week period. 3o These symptoms include depressed mood as well as significant
changes in sleep, appetite, mental concentration, energy, ability to experience pleasure,
self-esteem, sense of hope, guilt, and desire to live.
Extremes of this diagnostic grouping are seen daily in clinical practice and can
range from a patient who describes a depressed mood distinctly different from sadness
and is experiencing insomnia, loss of appetite and weight, decreased libido, energy,
interest, and concentration, and suicidal ideas to another patient who describes a
slightly depressed mood with prominent irritability, loss of energy and mental concen-
tration, decreased libido, moderate generalized anxiety, and inappropriate excessive
worry but normal appetite, weight, and sleep. Yet another patient may describe daily
crying spells and intense sadness and severe inappropriate guilt, along with increased
appetite, weight, and sleep (sleeping 12-14 hours/day).
While the etiology of major depression may be heterogeneous, the brain systems
involved in determining specific emotions, cognitions, and somatic processes are likely
to be similar across patients. There is considerable evidence that specific brain net-
works and the neuroanatomical circuits that join them underlie emotional experience,
cognition, and somatic function. 31 While our current understanding of the role and
function of these circuits and networks remains primitive, a rough outline is beginning
to emerge. 32 As most antidepressant drugs are comparably effective across patients-
in spite of the sometimes dramatic differences in individual symptoms between pa-
tients with major depression-these drugs probably cause widespread effects in many
brain regions.
Akiskal and McKinney have described depression as a psychobiological "final
common pathway of the various interlocking processes at chemical, experiential, and
behavioral levels."33 They suggest that many factors-including biological vul-
nerability, developmental deficits, and physiological and psychosocial stressors-can
contribute in varying degrees to functional impairment of the diencephalic centers that
help to maintain mood, motor activity, appetite, sleep, and libido.
If these hypotheses are correct, then depressed patients suffering from similar
symptoms most likely have similar dysfunction of the same networks and circuits. The
dysfunction's etiology may differ between patients, but the net functional impact is
similar. An analogous situation would be a cardiac arrhythmia leading to an increase in
heart rate. In some situations this could be caused by an electrolyte imbalance, and in
others it might arise from infarction of a critical region of the heart. The same symptom
arises from different pathological processes. These types of integrative models begin to
account for the diversity of clinical presentations and course and offer explanations of

how neurotransmitter systems, brain circuits, and the life experiences that affect the
development of these systems could interact to produce vulnerabilities to depression
and how life stress or neurobiological dysfunction could lead to a similar clinical


A. Clinical Presentations
Whether or not to treat a patient with an antidepressant medication is a common
and important question for mental health clinicians and primary-care doctors. Some-
day, biological tests may give us clues or even crisp answers to this question. For now,
however, psychiatric treatments are based on verbal reports and behavioral observa-
tions, coupled with knowledge gained from clinical research.
Several theoretical and practical issues remain extant and troubling to our field.
The very term "antidepressant" is unsatisfying. When a class of agents can treat
anuresis and chronic pain, panic and phobias, obsessions and compulsions, and other
symptoms, does this mean that these other conditions are somehow linked to mood
pathology, that they share a common brain center or pathway, or simply that the
pharmacologic action of the drugs is salutary for a heterogeneous group of conditions?
And what of the increasingly common observations of comorbidity? Patients with
depression frequently have coexisting anxiety disorders, substance abuse conditions,
obsessive-compulsive disorder, eating disorders, and more. Does one condition render
them vulnerable to others? Is a vulnerable brain more vulnerable to other conditions
(coincidentally or causally)? Or are the observations of frequent comorbidity simply
artifacts of our observational, rating, and diagnostic techniques?
If you are sitting with a patient who claims or appears to be depressed, what
should lead you to recommend or prescribe an antidepressant? The first question is
whether we are observing normal sadness or grief in reaction to a real, perceived, or
feared loss. Such losses can include not only a person (through separation or death), but
a pet, job, dream, wish, or plan. In psychologically healthy individuals, time, thought,
alternatives, and caring others can promote healing and restoration in most circum-
stances. Clinicians tend to judge the appropriateness of grief or "reactive" depression
by its duration and magnitude in response to the loss: greater losses are expected to
produce psychological responses of comparatively greater magnitude and duration.
Because normal grief responses can vary considerably among people from different
cultural backgrounds, the further your own cultural background diverges from your
patient's, the more circumspect you should be in drawing conclusions about the mean-
ing of the loss and the appropriateness of the reaction. Others more familiar with the
patient's culture should be consulted. The process of grieving is an iterative, dynamic,
progressive, and variable one. Most important is that the individual not become "stuck"
in the process: unable to move, to utilize the resources of others, or to experience the
cleansing relief of the grieving process and the waves of alternating sadness and
distraction that characterize this normal human experience. When symptoms and dys-

function persist, are severe, or threaten life, medical or psychological intervention may
become necessary. In some cases, the stress of grief can lead to the morbid mood
reaction of depression.
Mild, nonpsychotic depression has been shown to benefit from carefully defined
and studied forms of psychotherapy, such as interpersonal and cognitive-behavioral
techniques. Psychotherapy may also be useful when the patient engages in behavior
patterns that lead to loss and bereavement. Some patients, for example, habitually put
themselves psychologically "in harm's way"-for example, by pursuing or selecting
inappropriate mates or work situations. In these circumstances, successful psycho-
therapy might clarify misperceptions and lead to behavioral changes that can prevent
repetition of these patterns and recurrence of depression.
When symptoms of depression are severe and persistent, or when psychosocial
intervention has failed to yield notable relief and change over a matter of months,
antidepressant drugs usually are indicated. Particularly with the newer generation of
antidepressants, the risk-benefit ratio almost always favors at least a few therapeutic
trials of these agents.
There are several circumstances in which long-standing behavior patterns affect the
likelihood of response to antidepressant pharmacotherapy. One is a coexisting borderline
personality. Here the work of Soloff and others has shown that, although patients with
borderline personality disorder often qualify for diagnosis of major depression, their
response to antidepressant drugs tends to unsatisfactory (see Chapter 13).
On the other hand, when a patient has a long-standing depression, clinicians,
friends and family, and even the patient may come to believe that this is just "who they
are," in other words, that depression is part of the individual's character. A series of
recent studies, however, has indicated that chronic depression-whether dysthymia,
"double depression" (major depression superimposed on dysthymia), or chronic major
depression-is as responsive to antidepressant pharmacotherapy as depressions of
shorter duration. 34 Thus, chronicity should not in and of itself dampen clinicians' ardor
for prescribing. If, however, the patient has had a reasonable number of adequate
antidepressant trials and has failed to respond, the clinician may want to inquire again
about life circumstances, unrecognized stressors, behavior patterns, and occult medical
conditions that may suggest alternate approaches to the problem.
What can we learn from the pattern of depressive symptoms? The existence of a
DSM-IV diagnosis of depression does not automatically indicate the prescription of an
antidepressant medication. As noted above, mild depression may respond to psychoso-
cial intervention. Conversely, as noted above, depression that falls below the formal
symptomatic threshold of major depression should not automatically preclude the
prescription of an antidepressant drug-particularly if the symptoms of depression are
chronic. Cluster typing of depression symptoms may give some guidance. Depression
with melancholic features may be more likely to respond to antidepressants in general
or electroconvulsive therapy. Atypical depression has been shown to respond better to
monoamine oxidase inhibitor (MAOI) antidepressants than to tricyclics. 35
The concomitant presence of psychotic features roughly halves the estimated
response rate to an antidepressant drug (when used alone), whether a heterocyclic agent
or an MAOI. Certain types of personality features, such as hypochondriasis and histri-

onic traits, bode poorly for antidepressant response. Prior resistance to biological
therapy in multiple previous depressive episodes may also be a poor prognostic sign.36
As long as psychiatry remains empirical and phenomenologically based, we will
be treating syndromes that can reflect diverse etiologies. Undoubtedly, this accounts
for the variable responses to psychotropic medications in different settings and studies.
The syndrome of depression is no different. It can be caused or mimicked by a variety
of metabolic, endocrine, toxic, pharmacologic, and neurologic conditions. It makes no
sense for clinicians to memorize all such possible "organic" causes of depression, and it
would be frivolous to conduct laboratory testing on a routine basis to rule them out.
Rather, by conducting a review of systems and medical history, an astute clinician with
sensitive medical "antennae" can detect and become suspicious of symptoms, signs, or
medical history that could indicate further testing and broaden the differential diag-

B. Classifications
Historically, depressions have been subtyped according to dichotomous categories
(e.g., endogenous-reactive and neurotic-psychotic). Recently, this view has been re-
fined and broadened to include primary-secondary and unipolar-bipolar. In general,
endogenous depressions correspond to the melancholic subtype of major depressive
illness. The abandonment of the endogenous-reactive dichotomy has been important,
since it implied that some depressions do not have exclusively environmental causes
and require pharmacotherapy whereas others result from life stresses and should be
"worked through" psychotherapeutically. This clinical stance would preclude the use of
pharmacotherapy for patients with "reactive depressions" and the use of psychotherapy
for "endogenous" depressions. In fact, the decision to medicate with antidepressants
relies not so much on presumed causality but on the presence of particular symptom
clusters. The neurotic-psychotic classification of depressions is equally misleading,
since the definitions of these terms are variable and do not correlate well with drug
More recently, diagnosticians have reclassified depressions. The primary-second-
ary dichotomy avoids some of the problems of the older continua. In primary depres-
sions, there is no previous history of psychiatric illness except for episodes of depres-
sion or mania; secondary depressions occur in patients with a previously documented
psychiatric disorder (other than an affective episode) or a medical illness. Investigators
originally thought that primary depressions might preferentially respond to antidepres-
sants. However, secondary depressions generally have similar symptoms and responses
to medication. Clinically, patients with secondary depressions have a somewhat greater
incidence of somatic and psychotic symptoms, phobias, anger, and suicidal behaviors.
Winokur, 115 in particular, believes that secondary depressions represent an acute pro-
cess superimposed on a chronic set of problems.
The unipolar-bipolar dichotomy represents a further subclassification of primary
mood disorders. Bipolar disorders must have a history of both depression and mania. In
contrast, a unipolar depressive disorder must include a history of depressions only,
although the number and frequency of depressions require further definition. By study-

TABLE 1. Diagnostic Criteriaa

Major depressive episode

At least five of the following for at least 2 weeks (at least one must be depressed mood or loss of interest
or pleasure):
Depressed mood most of the day. nearly every day
Diminished interest or pleasure in activities most of the day, nearly every day
Poor appetite or weight loss or increased appetite or weight gain
Insomnia or hypersomnia '
Loss of energy or fatigue
Psychomotor agitation or retardation
Feelings of worthlessness or excessive or inappropriate guilt
Diminished ability to think or concentrate
Recurrent thoughts of death or suicide
Dysthymic disorder
Depressed mood for most of the day, for more days than not, for at least 2 years
Presence, while depressed, of two (or more) of the following:
Poor appetite or overeating
Insomnia or hypersomnia
Low energy or fatigue
Low self-esteem
Poor concentration or difficulty making decisions
Feelings of hopelessness
Never without symptoms for more than 2 months during 2-year period

"From Ref. 30.

ing families, Winokur and Coryell have further subdivided primary depression into
pure depressive disease, depressive spectrum disease, and sporadic depressive illness
(nonfamilial).28 Pure depressive disease occurs equally in men and women, but the
subgroup in which it occurs most often is older males who have one or more first-
degree relatives with depressions. Depressive spectrum disease, in contrast, occurs
most frequently in women under 40 years of age. Their female relatives have an
increased frequency of depression, and first-degree relatives have a high frequency of
alcoholism or antisocial characteristics. Males with depressive spectrum disease usu-
ally have alcoholic or antisocial symptoms.
Sporadic depressive disorders comprise the largest group of unipolar depressions
(about 40%). Although unipolar illnesses can be divided into relatively homogeneous
groups, the clinical ramifications of these findings await clarification. In addition,
researchers must further correlate epidemiologic, familial, and clinical data with bio-
chemical findings.
The DSM-IV is the currently adopted diagnostic framework for psychiatric disor-
ders. However, because DSM-IV must ensure a high degree of reliability and validity,
some more atypical presentations of depression are excluded. Despite limitations, it
defines a group of major depressive disorders (see Table 1) that are generally:

1. recurrent and severe (70-80%); and

2. responsive to antidepressant medications.

C. Biological Diagnosis
Since antidepressant drugs were first found to be effective in treating depression,
there has been hope that a biological test could be discovered to identify patients who
were more likely to improve on a specific drug. As new antidepressant drugs and
combination and augmentation strategies have been discovered, this hope has further
increased. Without doubt, treating depression with drugs will change remarkably if and
when we have biological tests similar to those in other fields of medicine: these will
enable us to subdivide an illness by etiology and select an appropriate treatment or
combination of treatments. Indeed, as Klerman has quipped, this is psychiatry's "search
for the holy grail." It stands to reason that there may be biological subtypes of depres-
sion and that these will prove differentially responsive to pharmacologic therapies.
When practical markers become available, they will represent a great leap forward,
shortening the lag time before remission and making a series of empirical trials of
antidepressant drugs a historical curiosity.
To date, unfortunately, we have no such biological testing available to practi-
tioners,36 but it has not been for lack of effort from investigators, attention in the
scientific literature, or research funding. Many hopes have risen and been subsequently
dashed. Candidates for biological markers have included urinary metabolites of central
monoamine neurotransmitters [e.g., 3-methoxy-4-hydroxyphenylglycol (MHPG)], pro-
vocative challenge tests (e.g., amphetamines), neuroendocrine markers [e.g., the Dexa-
methasone Suppression Test (DST) and the thyrotropin-releasing hormone (TRH) stim-
ulation test], the prolactin response to intravenous tryptophan, and the growth hormone
response to clonidine administration. Neurophysiological markers, such as the latency
before the first rapid eye movement (REM) period in the sleep cycle, also are being
actively studied. Although some markers have shown promise in predicting lack of
placebo response (e.g., failure to suppress cortisol appropriately during DST testing),
none of the markers studied to date nor the clinical subtypes identified have been
shown to have sufficient sensitivity or specificity to be useful as a marker of specific
medication response or nonresponse.
Attention is currently being focused on a variety of markers for specific neuro-
transmitter systems using positron-emission tomography (PET) technology. Radio-
labeled markers for specific components of the serotonergic, noradrenergic, and dopa-
minergic systems as well as markers of the functional state of these systems are
actively being investigated. Also being investigated are clinical neurotransmitter deple-
tion strategies. Identifying the reasons why some patients respond to an antidepressant
while others do not and why some patients respond to certain antidepressants and not
others continues to be just beyond our grasp. As our knowledge of basic neuroscience
increases and as new technologies are developed, we continue to hope that these
findings will lead to tests or procedures that can be successfully used to improve the
effectiveness of our treatments.


A broader understanding of the relationships among brain chemistry, behavior,

thinking, and feeling has fostered a more open environment for investigation. In fact, as

we clarify the multidetermined nature of depressive illness, we have also started to

develop a range of therapeutic techniques for depression. Research has demonstrated
the effectiveness, if not the necessity, of using somatic treatments in severe mood
disorders. However, recent studies have also shown that various psychological tech-
niques may represent an effective alternative for treating some depressed patients. I I ,37

A. Milieu and Crisis Techniques

Many depressed patients manifest intense suffering, suicidal impulses, and poor
self-care, requiring immediate and intensive care. Others require hospitalization be-
cause of uncertain diagnosis or medical problems that may confound somatic treat-
ment. In these cases, inpatient care offers various therapeutic advantages and possi-
bilities. At the most basic level, the ward milieu contains symptoms and provides
Frequently, depressed individuals exhibit self-destructive thoughts and behaviors
that develop, in part, as a result of self-punitive thinking and disengagement from
social supports. A therapeutic milieu can interrupt the depressive cycle and provide at
least some measure of protection. Patients intent on dying, however, usually succeed
unless immediately protected against self-harm. Because severe suicidal impulses often
last only a few days, close observation and enforced human contact often help the
patient live through the acute crisis. An inpatient milieu offers other therapeutic advan-
tages as well, including relief from self-demeaning or disorganizing interactions with
persons close to the patient; emotional support and self-validation; structured interper-
sonal interactions; correction of cognitive distortions, maladaptive patterns, and self-
punishing rituals; and the opportunity to practice new behaviors. These psychological
interventions can also be provided in other settings (e.g., day care) and by other
personnel (e.g., home visit and crisis teams). In fact, most intensive care settings
employ crisis techniques that have a focused, here-and-now, time-limited, reality-based
orientation. The therapeutic approaches include mobilizing the patient's natural sup-
ports; adopting an active and supportive (and sometimes directive) role for the thera-
pist; reviewing, labeling, and correcting maladaptive behaviors and cognitive distor-
tions; and teaching new coping and problem-solving techniques. In many ways, this
approach integrates aspects of cognitive therapy, behavioral techniques, and the short-
term psychodynamic therapies.

B. Psychotherapy
Depression is a set of symptoms with complex biopsychosocial determinants. For
the most severe depressions, biological factors rather than environmental or psycho-
logical factors seem to determine the course of the illness. These syndromes require
pharmacotherapy and may be only slightly responsive to specific psychotherapeutic
interventions. However, for depressions of moderate severity, various types of short-
term psychotherapy can play an important role in treatment.
Weissman et aI., in the Boston-New Haven Collaborative Studies, found that both
antidepressant medication and structured, short-term interpersonal psychotherapy
worked better than a mock treatment for patients with nonpsychotic major depressive

illness. 38 Both drug therapy and combined drug and psychotherapy provided equal
symptom relief and prevention of relapse. Psychotherapy alone offered somewhat less
symptom relief but more improvement in social adjustment.
As part of the research effort, the investigators developed a method of therapy that
has been specified in detail, tested in adequately controlled studies, and applied consis-
tently in concert with pharmacotherapy.39.40 The technique, including 12-16 once-
a-week sessions focusing on current interpersonal relationships, is comprised of
several structured phases. Initially, the clinician uses a semistructured evaluation inter-
view to clarify symptom history and to review typical depressive symptoms and the
natural course of depressive illness with the patient. The goal of the assessment is to
define interpersonal problems of four distinct types: grief reactions, interpersonal con-
flicts, role transitions, and interpersonal deficits.
The core treatment sessions focus on present, not past, relationships and behav-
iors. During the process, therapists actively help patients understand their range of
feelings, delineate past model relationships that were positive experiences, and select
and refine potentially helpful interpersonal strategies and actions. The clinician uses
both directive and interpretive techniques. However, the focus of intervention is on
here-and-now interpersonal relationships, not historical, intrapsychic, or ego-structural
problems. Perhaps most important, interpersonal psychotherapy has demonstrated an
important synergy between biological and psychosocial interventions for moderately
severe depressions. In concert with drugs, this psychotherapy seems to decrease symp-
toms better than drugs alone during the acute illness, increase social functioning during
recovery, and produce better treatment outcomes overall. 41
Cognitive therapy, developed by Beck and Rush, represents another important
contribution to the treatment of depressive illness. Instead of an interpersonal relation-
ship focus, cognitive therapy confronts the distorted cognition of depressed patients. I I
The therapy aims to alter patients' punitive, undermining negative thoughts, images,
and underlying assumptions about themselves and others. Cognitions are those
thoughts and images elicited in specific situations. In depressed individuals, cognitions
include several kinds of logical errors, such as overpersonalization, ignoring positive
parts of situations, and overgeneralizing or exaggerating negative aspects of situations.
Beck et a1. 11 also emphasize that patients' hopelessness and helplessness stem, at
least in part, from long-standing negative views of themselves, the world, and the
future (cognitive triad) and that depressed individuals habitually perpetuate those
views, called schemata. These chronic underlying assumptions, derived from early
developmental experience, color the way cognitions are used day to day. In the case of
depressed patients, undermining schemata and negative cognitions fuel deep, self-
denigrating perspectives.
Therapy to correct these misattributions is generally short-term, consisting of one
or two sessions per week for 14 weeks. Initially, treatment focuses on helping the
patient specify and record cognitions and schemata that are self-deprecating. Subse-
quently, the therapist outlines possible new or revised cognitions that are more self-
empowering and more aligned with objective reality. Later in therapy, the clinician
helps the patient identify the schemata, or hidden assumptions, that underlie depressive

The final part of therapy focuses on generating new cognitions and schemata and
applying them to actual and anticipated life events. Patients may complete this phase of
care during the last session of the initial 14-week therapy or may require additional
once- or twice-a-month sessions extending over several months. The most important
goals of cognitive therapy include helping the patient recognize self-denigrating cogni-
tive structures and develop practical alternative viewpoints and actions. Increasingly
studied and applied, cognitive therapy appears to be effective for moderately severe
depression and may have a synergistic effect when used together with antidepressants.
Therapists who are more behaviorally oriented use similar techniques to treat
depression. 41 In addition to identifying and confronting cognitive structures, behavioral
therapists employ formal cognitive restructuring, skills training (including assertive-
ness training), and environment manipulation.
In general, evidence indicates not only that structured psychotherapies, including
various forms of short-term dynamic psychotherapy and crisis intervention, help many
depressions but also that their effectiveness might be equal to that of pharmacotherapy
for some milder depressions. However, we must be careful in drawing definitive
conclusions pending further investigation. Studies often have employed antidepres-
sants and specific psychotherapeutic methods inadequately. No investigations compar-
ing psychotherapy and pharmacotherapy have addressed the problem of "fit" between a
patient and a particular drug or a specific model of therapy. In addition, the accumulat-
ing evidence in psychotherapy research indicates that many factors other than the type
of therapy significantly promote positive outcome, including a patient's positive atti-
tude toward therapy and the therapist, a therapist's positive feeling about the patient,
and the patient's motivation.
Which depressed patients should receive psychotherapy? Every patient should be
involved in a therapeutic relationship. Individuals taking drugs should receive emotion-
al support and encouragement and the opportunity to address their feelings about the
depressive episode. Moreover, for selected individuals with significant character pa-
thology (e.g., borderline states, narcissistic personality disorders, masochistic or de-
pressive characters) who develop mild to moderately severe depressions in response to
chronic disappointment and low self-esteem, long-term insight-oriented psychotherapy
might be the treatment of choice. These patients, however, frequently develop more
severe and autonomous depressive symptoms accompanied by vegetative signs; such
symptoms may improve with medication. In addition, some therapies may be equal to
or better than medication for individuals who become depressed in response to severe
stress, life crises, or real or fantasied losses or from a severely negative view of the self,
world, or others.

C. Electroconvulsive Therapy
1. Introduction
Although its mechanism of action remains unknown, electroconvulsive therapy
(ECT) effectively treats most patients with severe depression. In fact, 60,000 to
100,000 individuals receive this treatment each year, most without significant mor-

bidity. However, questions of efficacy, concerns about the development of an organic

brain syndrome, and occasional abuses have stigmatized this procedure. 42 In addition,
antidepressant medication offers an effective and often more acceptable alternative to
ECT. Thus, some patients do not receive ECT even when it is the treatment of choice.
In the last decade, however, investigators have attempted to define more precisely the
various clinical effects and indications and contraindications of ECT.

2. Mechanism of Action

Although no consensus exists about the therapeutic mechanism of ECT, re-

searchers have documented profound central nervous system changes resulting from its
use. Some experts believe that it alters neurotransmitter availability similarly to the
effects of antidepressant medications. ECT increases the sensitivity of postsynaptic
neurotransmitter receptors and enhances serotonin function after repeated, but not
single, administration. 43 ,44

3. Indications

Numerous controlled studies demonstrate that ECT works better than placebo and
at least as well as heterocyclic antidepressants in the treatment of severe depression. In
addition, ECT may be more effective than heterocyclics alone for treating psychotic
depressions, particularly when they are accompanied by agitation. ECT also alleviates
the symptoms of extreme agitation and catatonia in patients with schizophrenia who
respond poorly to medication, making these patients more amenable to other treatment.
ECT offers significant advantages over medication in certain clinical situations.
The strongest indications include:
1. Previous good response to ECT without significant untoward effects.
2. Strong family history of positive response to ECT.
3. Depression characterized by psychotic features, particularly somatic or ni-
hilistic delusions.
4. Relative or absolute contraindication to medication, such as in patients with
significant cardiovascular disease or other medical conditions that preclude
an adequate trial of drugs, including pregnancy or history of neuroleptic
malignant syndrome.
S. Need for immediate containment of symptoms (e.g., severely suicidal pa-
tients, severe mania, extremely agitated or assaultive patients, and individuals
with severe retarded depression) that impair self-care and feeding.
6. Poor response to trials of medication.
7. Severe, persistent behavior problems or mildly disorganized thinking in
schizophrenic patients.
ECT should be strongly considered for psychotically depressed patients and
those who have not responded to medication. In this situation, the clinician may start
or continue an antidepressant drug, since ECT and heterocyclics may work syner-
gistically. However, the anesthesia procedure occasionally may produce additional risk
of cardiotoxicity from the combination of ECT with antidepressants and anti psycho-

tics. Therefore, a clinical decision should be made together with an anesthesiologist. In

addition, patients receiving lithium have a greater risk of developing significant
central nervous system (CNS) impairment during a course of ECT; therefore,
lithium treatment should be discontinued before ECT and restarted after treatment has
ended. 45 ,46 Many patients require a psychoactive drug after ECT to prevent relapse.
Although fewer than a quarter of patients on a heterocyclic antidepressant or lithium
relapse after 6 months, more than half of individuals without maintenance phar-
macotherapy subsequently have a recurrence. However, the guidelines for starting
long-term drug therapy following ECT require further clarification. In occasional
cases, maintenance ECT is needed and helpful.
Sometimes ECT does not provide adequate relief from depressive symptoms.
Factors that correlate with poor outcome include hypochondriasis, hysterical character
structure, absence of vegetative symptoms (i.e., changes in sleeping, eating, bowel
function, or motor activity), absence of a family history of depression, absence of
delusions, presence of reactive factors, and presence of a marked personality disorder.
Clinically, however, this list of negative criteria does not provide much guidance, since
individuals with these attributes also respond poorly to other treatment modalities.

4. Contraindications and Adverse Reactions

Several medical conditions may limit the use of ECT. Increased risk generally
stems from the anesthesia, the induced seizure, or the transient increase in CNS pres-
sure and blood flow. Relative contraindications include recent myocardial infarction,
active pulmonary inflammation, CNS tumors, increased intracranial pressure,
and recent carbon monoxide poisoning. (ECT occasionally has resulted in brain
damage in these cases.)
The risks that attend preparatory anesthesia include significant respiratory depres-
sion, allergy to premedications, and cardiac failure. The main adverse effects of ECT
are confusion and short-term memory loss (which may last five times as long in elderly
patients). Both risks are significantly reduced with unilateral treatments. 47

5. Technique

The clinician should inform patients about what they will experience during and
after the procedure: they will be asked to lie down, a sedative will be administered, and
they may wake up with some discomfort (and rarely with a sense of being paralyzed).
In addition, each ind\vidual should know the possible risks (particularly memory loss)
and treatment alternatives. The practitioner should obtain informed consent (from a
relative or guardian, if necessary) and should carefully document discussions and
The procedure begins with the administration of an anticholinergic agent
(scopolamine or atropine), a short-acting anesthetic agent (usually a briefly acting
barbiturate), and a muscle relaxant [succinylcholine (Anectine@ and others)]. The
barbiturate usually takes about 1 min to work. During anesthesia, because of muscle
paralysis, assisted ventilation with oxygen is required. A physician then applies a brief
electrical stimulation. The electrical impulse is titrated to produce a generalized CNS

seizure of about 40- to 60-sec duration. The strong muscle relaxant protects the patient
from physical injury but paralyzes the respiratory and limb muscles. Therefore, to
monitor the seizure, the clinician can tighten a blood pressure cuff on one arm, which
will prevent the succinylcholine from affecting that limb and allow the seizure activity
to be observed. Alternatively, the clinician may use an electroencephalogram (EEG)
tracing to observe the spike- and-wave pattern. After the seizure, recovery usually takes
about 10-20 min.
The number and frequency of treatments vary with the clinical needs of the
patient. Therapy should be administered until the patient experiences complete recov-
ery. The typical regimen nsually consists of every-other-day or twice-weekly treat-
ments for a total of 5-10 sessions.
Clinical research and enhanced ECT technology have improved our understanding
and technique in recent years. Newer ECT devices make it easier to monitor seizure
duration electroencephalographically, and noninvasive fingertip cuffs allow anesthe-
siologists to monitor arterial blood gases throughout the treatment. For all patients and
particularly for those receiving unilateral ECT, it is important to verify that a seizure
has occurred. Moreover, when the unilateral method is used, a suprathreshold stimulus
provides better efficacy than a lower-energy stimulus that barely triggers a convulsion.
Some practitioners locate the threshold stimulus needed to trigger a seizure on a
patient's first treatment and then subsequently administer stimulation at 100% greater
energy for bilateral electrode placement and 150-200% greater energy for unilateral
electrode placement. 48 An alternative approach is to monitor the seizure duration,
which optimally should be between 40 and 120 seconds. The energy of stimulation is
then adjusted upward or downward to keep the seizure duration within that time range.
Increasingly, bilateral stimulation is being recognized as more effective and more rapid
for most patients (particularly the elderly), with unilateral treatments reserved for those
particularly sensitive to the cognitive side effects of ECT or who have already shown
deterioration in cognition and memory.


Twin coincident discoveries in the mid-1950s spawned two distinct classes of

pharmaceuticals that, for the first time, could reach to the core of the depressive
syndrome, alleviating the feelings of helplessness, hopelessness, and worthlessness that
often drive patients to suicide. Earlier, the only drugs appropriate for treating patients
with major depression were those that alleviated symptoms more peripheral to the
depressive syndrome. Barbiturates and bromides, for example, could be prescribed to
alleviate depressive symptoms such as anxiety, agitation, and insomnia. In a search for
new antituberculous agents, investigators discovered that iproniazid (withdrawn from
clinical use), a structural analogue of isoniazid (INH® and others), has mood-elevating
properties. It also inhibits the enzyme monoamine oxidase (MAO), which catabolizes
the monoamine neurotransmitters dopamine, norepinephrine, and serotonin. Thus, a
new group of antidepressants was born: the MAO inhibitors. They not only represented
a therapeutic breakthrough but also gave rise to speculation about mechanism of action
and, reasoning backwards, the pathophysiology of depression itself.

Several years later, research with structural analogues of the newly discovered
phenothiazines gave rise to imipramine (Tofranil®). Although its two-dimensional
configuration closely resembles that of its phenothiazine parents, minor changes in the
central "bridging" ring in this tricyclic compound, which gave the class its name,
resulted in alterations in its three-dimensional configuration. This, in tum, resulted in
major pharmacologic and clinical differences from the phenothiazines.
The remainder of the 1950s and the following two decades saw no comparable
breakthrough in antidepressant pharmacotherapy. Rather, a parade of "me-too" anti-
depressants entered the American market, driven largely by the financial needs of their
manufacturers. Discovered through the use of preclinical paradigms that virtually
ensured similarity to the prototypes, these agents possessed only minor to moderate
differences in side-effect profiles.
As the years went by, other countries began to introduce substantially different
drugs for treating depression. Then, in 1979, trimipramine (Surmontil®) was marketed
in the United States. There was some suggestion that it possessed a unique neurochemi-
cal and clinical profile, but the manufacturer has failed to pursue these possibilities
with definitive research; the best "hard" evidence suggests that trimiprarnine is just a
seventh "me-too" tricyclic. The years 1980, 1981, and 1982 saw the sequential intro-
duction of three somewhat different antidepressants: amoxapine (Asendin®),
maprotiline (Ludiomil® and others), and trazodone (Desyrel® and others). Although
each was introduced with considerable advertising ballyhoo, subsequent research and
clinical impressions suggest that these agents offer less than initially met our hopeful
eyes. By now, amoxapine and maprotiline are infrequently prescribed; trazodone is
used largely for its sedating properties.
The subsequent introductions into the U.S. market of nomifensine (no longer
available) and then bupropion (Wellbutrin®) were fraught with even greater complica-
tions. After less than a year, nomifensine was withdrawn worldwide because of its
association with a high rate of immunologic abnormalities, most notably hemolytic
anemia, which is potentially fatal. Although this experience highlighted potential pit-
falls in introducing new drugs, many clinicians have grieved the total withdrawal of
this agent, because some patients appeared uniquely responsive to it. It is possible that
neurochemical features distinguishing nomifensine from older compounds accounted
for these cases of distinct efficacy: most commonly mentioned is its enhancement of
central dopaminergic activity.
Bupropion spent a short time on the American market before it was withdrawn
because of its association with a high incidence of epileptic seizures. In 1989 it was
reintroduced, but with warnings about its effects on the seizure threshold. Since then, it
has demonstrated substantial efficacy "in the field," the seizure incidence appears
manageable, and its lack of sexual side effects has been a distinguishing feature.
Fluoxetine (Prozac® and others) was introduced into the United States at the
beginning of 1988 and rapidly enjoyed remarkable commercial and clinical success. In
addition to its apparently comparable efficacy to older antidepressants, two factors
were largely responsible for fluoxetine's popUlarity: fluoxetine has fewer side effects
than tricyclics, and it has an apparently simpler dosing schedule. Both factors made it
easier for primary-care physicians to prescribe this new agent.
As every educated layperson knows by now, fluoxetine began what is no less than

a second antidepressant "revolution," becoming the first in a series of new antidepres-

sants that became widely known and widely used. Most tricyclic antidepressants inhibit
the presynaptic reuptake of serotonin, but because they also inhibit the reuptake of
norepinephrine (and sometimes other neurotransmitters), they are considered to be
nonselective uptake inhibitors. Fluoxetine, by contrast, inhibits only the uptake of
serotonin, leading to the term serotonin-selective reuptake inhibitor (SSRI). In 1992,
sertraline (Zoloft®) became the second SSRI introduced, followed in 1993 by parox-
etine (Paxil®). Fluvoxamine (Luvox®), another SSRI, was introduced in 1995, but with
a labeled indication only for obsessive-compulsive disorder, not depression.
Have the SSRls brought us enhanced efficacy? Probably not. In the absence of
definitive data, debates continue as to whether these drugs are as efficacious for
patients with more severe depression. At least as troubling are suggestions that the
efficacy of SSRls appears to wane over months and years-during the phases known
as continuation and maintenance therapy, which are often so important in the manage-
ment of this typically recurrent illness.
For most patients, the side-effect profiles of SSRls are "kinder and gentler" than
those of tricyclic antidepressants, and these drugs are certainly less cumbersome and
less potentially hazardous than MAOIs. They are also much less hazardous in overdose
than older agents, which is a distinct advantage. They are not without side effects and
complications, however, and inhibitions of sexual desire and function are troublesome
in many patients taking them. In addition, they often cost more than older agents in
generic formulations.
The dose-response relationships of these drugs are not as simple as was originally
believed (one pill on one day for one patient), and proper dosing strategies remain a
matter of debate and require additional study.
In 1994, venlafaxine (Effexor®) was approved by the U.S. Food and Drug Admin-
istration (FDA). Inhibiting the reuptake of both serotonin and norepinephrine, ven-
lafaxine appears to have a pattern of side effects distinctive from both those of older
agents and those of the SSRls. Some clinicians find it helpful for patients previously
unresponsive to other antidepressants. Nefazodone (Serzone®) was introduced in 1995.
It simultaneously antagonizes the type 2A serotonin receptor and modestly inhibits the
uptake of serotonin and norepinephrine presynaptically. Nefazodone is less sedating
than tricyclic antidepressants or SSRls (although sedation does bother some patients)
and causes fewer anticholinergic effects than tricyclics. It also causes less insomnia
than SSRls, does not appear to affect sleep architecture or the electrocardiogram
(EKG), and appears not to cause the sexual dysfunction so common with other anti-
depressants. For both venlafaxine and nefazodone, higher doses may be positively
correlated with superior efficacy.
Mirtazapine (Remeron), introduced in 1996, has a unique pharmacologic profile
and is structurally unrelated to any antidepressant so far available in the United States.
By selectively blocking presynaptic <x 2 -adrenergic receptors, mirtazapine enhances
both noradrenergic and serotonergic transmission, which probably underlies its anti-
depressant properties. It is also a potent antagonist of serotonergic 5-HT2 and 5-HT3
and histaminic H I receptors but does not affect norepinephrine, serotonin, or dopamine
reuptake. Its weak inhibition of peripheral <X I-adrenergic and muscarinic receptors may

explain its rare association with orthostatic hypotension and its relatively low incidence
of anticholinergic side effects, respectively. Published reports do not show greater
efficacy for mirtazapine over previously available antidepressants.
Clomipramine (Anafranil®), introduced in the United States in 1990 after many
years of availability abroad, is a potent serotonin reuptake inhibitor (SRI) but a non-
selective one. It was the first drug labeled for the treatment of obsessive-compulsive
disorder, and although SSRIs also appear efficacious for this condition, there are some
suggestions that clomipramine may be superior. On the "downside," its side effects-
including gastrointestinal upset, sexual dysfunction, and seizures-tend to be more
problematic. Because of diverse reactions and toxicity, the drug is not labeled for the
treatment of depression and is seldom used to treat depression in the United States,
although it probably is at least as efficacious against depression as other tricyclics.
Selegiline (Eldepryl®), a selective inhibitor of the B type of MAO, is available
now in the United States and labeled for the treatment of Parkinson's disease. Although
some have reported apparent efficacy of selegiline in the treatment of depression, this
has never been conclusively demonstrated. Moreover, at doses that may be efficacious
in depression, selegiline loses the specificity that makes it relatively immune from the
"cheese effect," by which exogenous tyramine can cause hypertensive reactions. Pa-
tients taking selegiline at any dose are at risk to develop the hypermetabolic reaction
("serotonin syndrome") if they concomitantly take a drug like meperidine (Demerol®
and others) or an SSRI. Serious and sometimes fatal reactions have also occurred when
patients have taken selegiline in combination with a tricyclic antidepressant.
Several reversible inhibitors of type A MAO (RIMAs) are currently in use abroad
and are being tested in the United States. These agents should have less risk of the
"cheese effect" than MAOIs and greater tolerance. However, there have been reports of
excitatory effects with the RIMA moclobemide and adverse interactions with cimetidine
(Tagamet®) and meperidine. RIMAs should elevate brain levels of the catecholamines
norepinephrine and dopamine, which could explain stimulantlike properties. They
should also raise brain serotonin levels, and the "serotonin syndrome" has been associ-
ated with moclobemide, especially in combination with the SRI clomipramine.
From the time the third edition of The Practitioner's Guide went to press until the
time of this writing, the new generation of antidepressants have substantially changed
the prescribing patterns of American physicians. Unless new (and unexpected) toxicity
emerges, or suggested problems of tolerance are confirmed and serious, the new
generation of antidepressants are likely to be employed more and more as first-line
antidepressant agents. Without doubt, they give us more options, and most patients can
take them with a greater degree of comfort, leading primary-care doctors in particular
to be more willing to prescribe them, and patients more willing to take them regularly.
Beyond the availability of more and different drugs, which has led to an increased
number of options for patients refractory to initial trials, practitioners who treat depres-
sion will only get to "the promised land" when we will have tests like the culture
sensitivity available in the treatment of infectious disease-tests that can tailor treat-
ment to the underlying problem. Only then will we be able to select a drug to treat a
patient's presumed neurochemical defect, thus sparing people unnecessary drug expo-
sure and prolonged empirical trials. For now, however, the greatest challenge lies in

convincing the public and health care givers to recognize depression, appreciate its
profound human and economic costs, and learn about the high degree of efficacy of
currently available antidepressant drugs.

A. Heterocyclic Antidepressants
1. Chemistry
The chemical structures of the heterocyclic antidepressants currently available in
the United States are shown in Figs. 1-3. The chemical structure of the tricyclic
antidepressants (Fig. 1) differs from that of the phenothiazines only in the substitution of
a carbon-carbon double bond for the sulfur atom in the bridge between the two benzene
rings. This minor difference, however, changes the three-dimensional conformation of
the molecules, which in tum alters their pharmacological and clinical properties.
With two methyl groups on the terminal nitrogen atom of the side chain, im-
ipramine, amitriptyline (Elavil® and others), trimipramine, doxepin (Sinequan® and
others), and clomipramine are called tertiary amines. The demethylation of imipramine
and amitriptyline results in, respectively, desipramine and nortriptyline (Pamelor® and
others), which also are active antidepressants. With a hydrogen substituted for one of
the methyl groups on the terminal nitrogen, desipramine and nortriptyline, along with
protriptyline (Vivactil®), are known as secondary-amine tricyclics. By and large, the
secondary-amine tricyclics have weaker anticholinergic properties and other neuro-
chemical differences that affect their clinical profile and possibly their blood level
relationship as well. Much as the addition of a chlorine atom to the phenothiazine
promazine (Sparine®), resulting in chlorpromazine (Thorazine® and others), produced
an agent with distinctive pharmacologic and clinical properties, so the addition of a
chlorine atom to imipramine, producing clomipramine, has altered its pharmacologic
and clinical properties. This has resulted in an agent that is more highly serotonergic
and possibly has dopamine-blocking properties, and one that appears to be effective in
alleviating obsessions and compulsions.
Amoxapine also has a three-ring structure, but it is a demethylated metabolite of
the antipsychotic loxapine, which it resembles pharmacologically and clinically; both
are termed dibenzoxazepines. Maprotiline has a fourth ring and is, therefore, termed a
tetracyclic, but it has few pharmacologic or clinical differences from the tricyclics.
Trazodone is a phenylpiperazine derivative of triazolopyridine-chemically, phar-
macologically, and clinically distinct from the tricyclic antidepressants.
As noted above, there are four SSRI antidepressants (Fig. 3). Fluoxetine's chemi-
cal structure makes it a straight-chain phenylpropylamine. Sertraline is a
naphthylamine derivative, paroxetine is a phenylpiperidine, and fluvoxamine is one in
a series of 2-aminoethyl oximes of aralkyl ketones. Although each SSRI is structurally
unique, they all have an electron-dense phenyl ring four to five carbons over from a
nitrogen. It may be this structural similarity that enables the inhibition of serotonin
Different from other antidepressant drugs is bupropion. This drug is structurally
related to phenylethylamine, which probably accounts for some of its stimulant proper-

090 CHCH2CH2N(CH3)2 • HCI

~I I ..........-
0c:J0 I
Amitriptyline Hydrochloride Clomipramine Hydrochloride B Desipramine Hydrochloride

0c:J0 I

Doxepln Hydrochloride Imipramine Hydrochloride Nortriptyline Hydrochloride


0r:J0 I
Protriptyline Hydrochloride Trlmlpramlne
Figure 1. Tricyclic antidepressants. Clomipramine is labeled in the United States only for the treatment of obsessive-compulsive disorder. ~
OC-PCI Y HNC(CH 3)3 cr;:o
eN H
Amoxapine Bupropion Hydrochloride Maprotlline Hydrochloride

N) )=~
o-OCH.cH2~N""CH2CH2CH2"N~N<O(· HCI
otb CI

Mlrtazaplne Nefazodone Hydrochloride

CH 3
-\ HCI CI ;o45 3


Trazodone Hydrochloride Venlafaxine Hydrochloride

Figure 2. Other antidepressants.

F3CQO-CH-<CH2)2-NH-CH3 • HCI <::OOCH2 ... ...

Fluoxetine Hydrochloride

Sertraline Hydrochlooride

Figure 3. Serotonin-selective reuptake inhibitor antidepressants.

ties. Relatively weak in blocking neuronal reuptake of serotonin and norepinephrine,

bupropion does inhibit dopamine uptake. Venlafaxine is a phenethylamine, nefazodone
is a phenylpiperazine derivative, and the tetracyclic mirtazapine is a piperazinoazepine.

2. Mechanism of Action
All effective antidepressant medications and ECT have potent pharmacologic
effects on monoamine neurotransmitter systems. While some drugs inhibit the reuptake
of monoamines, other such as mianserin* and nefazodone primarily block neuro-
transmitter receptors. Potent serotonin reuptake (fluoxetine), norepinephrine reuptake
(desipramine), and dopamine reuptake (nomifensine*) inhibitors effectively treat de-
pression. Current thinking is that these effects on monoamine systems underlie the
initiation of antidepressant action. Through diverse pharmacologic properties, this
group of drugs increases net transmission through either serotonergic, noradrenergic, or
dopaminergic systems. This is the case even for drugs that act as receptor antagonists
(such as nefazodone). Because stimulation of serotonin 2A receptors inhibits the effects
of serotonin on the serotonin lA receptor, when nefazodone blocks the serotonin 2A
receptor, it actually potentiates the effects of serotonin on the serotonin lA receptor.
Recent evidence from clinical studies suggests that while increased synaptic levels
of monoamines is not temporally correlated with therapeutic response, it is an essential
aspect of the mechanism of action of some medications. Selective depletion of seroto-
nin49 ,5o or norepinephrine and dopamine 51 causes a rapid relapse of depression in
depressed patients who have had therapeutic responses to and are being maintained on
antidepressant drugs. The relapse is qualitatively similar to the depression prior to
treatment, it correlates with the time course of monoamine depletion, and the patients

*Mianserin and nomifensine are not available in the United States.


improve rapidly as monoamine levels return to baseline levels. Depleting 5-HT causes
a rapid return of depression in depressed patients who have responded to fluoxetine,
fluvoxamine, MAOIs, and imipramine but not those who have responded to desi-
pramine, nortriptyline, or bupropion. 50,52 Depleting norepinephrine and dopamine
causes a rapid return of depression in those depressed patients who have improved on
desipramine but not in those who have improved on fluoxetine. 51 These studies under-
score the importance of monoamine reuptake inhibition and suggest that more than one
mechanism of action exists.
Monoaminergic systems in the brain modulate many behaviors, cognitions, and
somatic functions. By altering the monoaminergic systems, antidepressant drugs poten-
tly affect the core brain circuits involved in modulating stress responses. Antidepres-
sant drugs could increase neurotransmission in normal monoaminergic systems, lead-
ing to a beneficial impact on dysfunctional brain networks and restoring more
appropriate overall function. This may account for the fact that antidepressants are
effective in the acute treatment of milder mood disorders such as dysthymia, as well as
in generalized anxiety disorder, panic disorder, social phobia, obsessive-compulsive
disorder, bulimia and anorexia nervosa, posttraumatic stress disorder, and possibly
alcohol abuse. Perhaps antidepressant drugs "nonspecific ally" restore function irre-
spective of the underlying pathology, analogously to corticosteroids reducing inflam-
mation regardless of etiology.
We probably will not find a single mechanism for all antidepressants. In addition
to the existence of depressive SUbtypes, each distinct syndrome may result from a
complex interaction of presynaptic, neurohumoral, and postsynaptic receptor properties
of specific neural pathways. As Maas postulated more than 20 years ago, antidepres-
sants may act in various complicated ways to restabilize an unbalanced system. 53

3. Adverse Reactions
The clinician should know a drug's adverse reactions and a patient's suscep-
tibilities. Some of these effects cause minor discomfort, but others may interfere with
drug trials or produce serious morbidity. Of the common reactions, cardiac, anti-
cholinergic, and CNS effects produce the most clinical morbidity. Untoward effects
occur more often when cyclics are coadministered with other drugs or when a patient
has a coexisting physical disorder.
Overall, tricyclic antidepressants resemble one another in their adverse-effect
profile. In general, secondary-amine tricyclics are less anticholinergic and sedating and
produce less weight gain and postural hypotension. Maprotiline's and amoxapine's
adverse effects are comparable to those of the tricyclics. However, amoxapine also
resembles the neuroleptics in producing both extrapyramidal reactions and neuroen-
docrine changes.
Trazodone presents a somewhat distinct clinical profile, certainly different from
that of the tricyclic antidepressants. Highly sedating, trazodone produces postural
hypotension and some gastric irritation. It also is associated with occasional cases of
priapism and can exacerbate preexisting myocardial irritability. It is unlikely, however,
to produce anticholinergic effects and appears very safe in overdose. Chemically re-

lated to trazodone, nefazodone differs somewhat in its side-effect profile. The most
common adverse effects with nefazodone are headache, dry mouth, nausea, drowsi-
ness, dizziness, constipation, lack of energy, insomnia, diarrhea, and light-headedness.
Nefazodone does not appear to affect sexual function adversely nor to cause priapism.
It may cause postural hypotension. Initial activation is a problem for some patients, and
cognitive dulling can be bothersome, especially in the elderly.
The SSRIs cause nausea, headache, nervousness, and insomnia more commonly
than the tricyclics but generally do not cause acute weight gain, are less likely to cause
anticholinergic effects or orthostatic hypotension, and are much safer in overdosage.
With prolonged use, some patients do gain weight on SSRI treatment. Agitation and
increased anxiety may occur in the first weeks of treatment and subside later. All SSRIs
appear to cause a high incidence of anorgasmia and other sexual dysfunction. Akathisia
can be troublesome and may be more common with fluoxetine. Many antidepressants
are metabolized by the liver's P450 enzyme systems and can alter the pharmacokinetics
of other drugs (see Table 1 in Chapter 1 and Section V.A.3.k). SSRIs, in particular, can
raise blood levels of many other pharmaceuticals, thus magnifying their possible side
Bupropion may cause agitation and, rarely, seizures but causes fewer anti-
cholinergic effects than the tricyclic antidepressants, little if any orthostatic hypoten-
sion, and a relatively low incidence of sexual dysfunction. Venlafaxine generally is
similar to the SSRIs in its adverse effects, but it has been associated with a sustained
increase in diastolic blood pressure. Nausea may limit rapid dose increases.
Transient somnolence occurs in more than 50% of patients taking mirtazapine,
and increased appetite and weight gain are also common problems. Bone marrow
suppression has occurred in a few patients. There are unsubstantiated suggestions that,
unlike other antidepressants, mirtazapine has a low incidence of adverse sexual side
effects. However, this has not been systematically studied.

a. Gastric. Irritation of the gastric mucosa, resulting in heartburn, nausea, and,

rarely, vomiting, is a relatively uncommon side effect of tricyclic antidepressants,
which usually counteract excessive stomach acidity (via anticholinergic and anti-
histaminic activity). Serotonergic antidepressants (including clomipramine and ven-
lafaxine), however, frequently cause stomach upset. Patients troubled by gastric irrita-
tion can be advised to take their medicine with food, milk, or an antacid. Also, the daily
dose might be divided rather than administered at one time, and dosage increments can
be made very gradually.
Although some of the gastric effects of serotonergic antidepressants are likely due
to increased levels of serotonin in the gut itself, nausea and related symptoms may
reflect central actions of the drugs. While tolerance may develop over time in some
patients, in others the daily dose may have to be reduced. Preliminary case reports
suggest that cisapride (Propulsid®), an agent indicated for the treatment of nocturnal
heartburn due to gastroesophageal reflux, may be an effective antidote for nausea
associated with early therapy with serotonergic antidepressants, both SSRls and non-
specific agents such as venlafaxine. (Cisapride is contraindicated with nefazodone
because of an interaction that can lead to dangerous cardiac ventricular arrhythmias.)

Gastrointestinal side effects are seldom a major cause of noncompliance or drug

discontinuation for patients taking antidepressants.

b. Hematological
i. Leukocytic Effects and Purpura. Heterocyclic antidepressants only rarely pro-
duce alterations in white cell count. Clinicians have reported leukocytosis, leukopenia,
Loeffler's syndrome, eosinophilia, thrombocytopenia, and purpura. These conditions
should be medically evaluated, although they seldom cause significant morbidity. The
patient can usually continue the medication at the same dosage.
ii. Agranulocytosis and Other Hemolytic Effects. Both the heterocyclic anti-
depressants and the phenothiazines can cause agranulocytosis, but the incidence is
much lower with antidepressants than with antipsychotics. Death from infection sel-
dom occurs if the drug is discontinued immediately and the patient placed immediately
in reverse isolation. White blood cell numbers usually return to normal within several
weeks of discontinuing the drug. Most reported cases have occurred with imipramine
(although this may be an artifact, since it is the drug most often studied).
Agranulocytosis seems to be an allergic response of sudden onset that usually
appears 40-70 days after initiation of the medication. Patients have a greater risk with
advancing age and concomitant physical illnesses. The clinical syndrome is charac-
terized by a low white count (composed almost completely of lymphocytes), normal
red count, and infection (usually involving the oropharynx) accompanied by fatigue
and malaise.
Routine blood studies do not help with early detection, since the syndrome devel-
ops very rapidly. Instead, the clinician should evaluate and treat any sign of infection,
especially of the pharynx. He or she should discontinue the heterocyclic and never
administer it to that patient again. The practitioner might try an antidepressant with a
different structure and follow the patient closely for adverse hematological effects.
It is unclear whether antidepressants ever cause aplastic anemia. Nomifensine was
withdrawn from the market worldwide because of its association with hemolytic ane-
mia. The tetracyclic mianserin, widely used in England, has been associated with a
higher incidence of bone marrow suppression than tricyclic antidepressants.

c. Hepatic. Fewer than 1% of patients treated with heterocyclic antidepressants

develop liver toxicity. This adverse reaction seems to be a hypersensitivity response
similar to that caused by the phenothiazines.
The toxicity usually produces abnormal liver function tests without clinical jaun-
dice. Mild, transient jaundice preceded by abdominal pain, anorexia, fever, and transi-
tory eosinophilia may occur during the first 2 months of treatment. In these cases, the
patient has abnormal liver function tests. Most often, individuals have a high conju-
gated bilirubin and congested bile canaliculi without hepatocellular damage. After the
drug is discontinued, recovery from this syndrome usually takes several weeks. When a
toxic reaction occurs, the clinician should lower the dosage or switch to another type of

d. Endocrine. Heterocyclic antidepressants have fewer endocrine effects than the

phenothiazines. Unlike the phenothiazines, amoxapine, and possibly clomipramine, the

other cyclics have produced no clear-cut cases of amenorrhea or galactorrhea. Rare

patients, however, may develop menstrual irregularities.
Antidepressants may lower blood glucose in patients with diabetes mellitus but do
not seem to provoke or worsen diabetes. In rare cases, idiosyncratic nephrogenic
diabetes insipidus has occurred.
Patients with preexisting thyroid disease may show altered tolerance to cyclics.
Euthyroid patients who are given tricyclics and triiodothyronine (T 3' Thyrolar®), thy-
roxine (T4 , Synthroid® and others), or thyroid-stimulating hormone (TSH) simul-
taneously sometimes show quicker resolution of the depression while still remaining
euthyroid. In women, estrogen supplementation also may have a salutary adjunctive
role along with antidepressants.

e. Ocular. Unlike the phenothiazines, the heterocyclic antidepressants do not

produce ocular pigmentation. Tricyclics do, however, commonly cause blurred vision.
This is caused by a failure of accommodation, which usually leaves distant vision
intact. This anticholinergic effect results from ciliary muscle relaxation. Tolerance
sometimes develops over the first few weeks of drug treatment.
In patients with narrow-angle glaucoma, tricyclics theoretically could cause
significant damage. Anticholinergic properties foster pupillary dilation, which can
precipitate an acute episode. The problem arises with undiscovered, unreported, or
untreated cases. If open-angle glaucoma, however, is being treated, the patient can take
tricyclics. To detect patients at risk for the narrow-angle variety, the clinician can
perform an examination with a penlight and take a good history, including questions
about blurred vision, halos around lights, and eye pain. To do a flashlight test, the
practitioner should shine a light laterally across the iris. If it illuminates the entire iris,
the patient probably does not have glaucoma. If angle closure exists, the iris bunches
and will block the laterally shining light. Angle-closure glaucoma usually requires
surgery, whereas the common open-angle type is usually well treated with drugs.
Antidepressants with weak or no anticholinergic properties, including SSRIs and
other new agents, do not cause these ocular problems.

f. Cardiovascular. Cardiovascular effects have been among the most trouble-

some and also among the most misunderstood reactions to tricyclic antidepressants.
Fortunately, research in recent years has clarified our understanding of these problems,
and some of the newer heterocyclic antidepressants have expanded treatment options
by offering different cardiovascular profiles.
i. Hypotension and Hypertension. Tricyclic antidepressants interfere with the nor-
mal hemodynamic changes that maintain cerebral perfusion pressure on rising from a
recumbent position. As a result, with postural change, blood pressure falls, and the
patient becomes dizzy or, worse still, develops a syncopal episode. Particularly in the
elderly, a fall may cause a bone fracture.
Tricyclics vary along a spectrum in their tendency to cause this side effect, which
is possibly related to their blockade of a-noradrenergic receptor sites; however, the
position of each agent on this axis has not yet been systematically worked out. We do
know, though, that imipramine causes a substantial degree of postural hypotension,

whereas nortriptyline causes much less. It appears that amoxapine, maprotiline,

trazodone, and nefazodone similarly cause a fair degree of postural blood pressure
decline. In contrast, SSRIs, venlafaxine, and bupropion rarely cause postural changes
in blood pressure.
The elderly are at high risk for the consequences of orthostatic hypotension
(namely, falls). It is less clear if they are more likely to develop hypotension than
younger patients. People who have had heart failure, though, are at greater risk, along
with patients who are volume depleted or who are taking other drugs that might cause
hypotension, such as antihypertensives. Patients with impaired cardiac conduction also
may be at higher risk, along with those who have dysautonomias such as that caused by
diabetic neuropathy.
Many management strategies have been suggested to help patients control anti-
depressant-induced orthostatic hypotension, but their number indicates that none is
universally successful. Advising patients to arise slowly (particularly when getting out
of a warm bath), to sleep with the head of the bed tilted upward, and to avoid prolonged
bed rest can sometimes be useful. Adequate fluid and salt intake (or even salt supple-
ments) are useful if they are not medically contraindicated. Support stockings help
some patients. Caffeine occasionally is a useful antidote, and some doctors have
reported success with yohimbine (Yocon® and others). Switching to another anti-
depressant less likely to cause hypotension is an option, and even lowering the dose of
the original antidepressant might help. Most complicated, but sometimes successful, is
the use of the mineralocorticoid fludrocortisone (Florinef® and others).
While tricyclic antidepressants most commonly are associated with orthostatic
hypotension, in occasional patients they may cause hypertension. This may be more
likely in younger patients. Venlafaxine is known to produce hypertension in a minority
of patients in a dose-dependent fashion: In premarketing studies, the incidence of
hypertension with venlafaxine was 3% at doses of less than 100 mg/day, 5% at 101-
200 mg/day, 7% at 201-300 mg/day, and 13% at doses greater than 300 mg/day.54
Overall, about 3% of patients receiving venlafaxine experienced blood pressure eleva-
tion, which was defined as ~ 10 mm Hg above baseline and ~90 mm Hg for three
consecutive visits.
ii. Heart Rate Changes. As do any strongly anticholinergic agents, tricyclic anti-
depressants cause a statistically significant increase in heart rate. For most patients, this
is clinically insignificant. For a patient with marginally compensated cardiac output,
however, the increased pulse rate can trigger or exacerbate heart failure. If tachycardia
becomes a problem, switch to drugs with minimal anticholinergic activity, such as
MAO inhibitors or any of the newer agents.
SSRI antidepressants, by contrast, cause a slight but clinically insignificant slow-
ing of the heart rate in many individuals. In a rare patient, however, clinically signifi-
cant bradycardia may occur. This appears more likely in the elderly, those with preex-
isting cardiac disease, or patients simultaneously taking other drugs that may slow the
heart rate. 55 ,56 Neither venlafaxine nor nefazodone appears to cause any change in
heart rate.
iii. Conduction Delay. Tricyclic antidepressants slow electrophysiological con-
duction velocities within the heart. This is most evident in the lower segment of the

atrioventricular conduction system, between the bundle of His and the ventricles, and is
clinically inconsequential for a majority of patients. Patients at risk include those with
second-degree heart block or with right or left bundle branch block, who may develop a
brady arrhythmia when treated with a tricyclic antidepressant. If the physician wishes to
prescribe a tricyclic for such a patient, the initial dose should be low, and dosage should
be increased very gradually. A cardiologist should follow the patient and should con-
duct Holter monitoring before and after each dose increment. Newer antidepressants
and MAO inhibitors do not slow cardiac conduction.
iv. Antiarrhythmic!Arrhythmogenic Activity. In overdose, with toxic blood levels,
tricyclic antidepressants have long been associated with malignant ventricular ar-
rhythmias, often leading to cardiac arrest. Within a broad range of therapeutic concen-
trations, however, tricyclic antidepressants possess a quinidinelike (type IA) antiar-
rhythmic activity. Thus, in patients with frequent atrial or ventricular premature
contractions, a tricyclic antidepressant will usually result in diminished extra beats. The
physician should be wary, of course, when coprescribing a tricyclic antidepressant with
another type IA antiarrhythmic, such as quinidine, procainamide (Pronestyl® and oth-
ers), or disopyramide (Norpace® and others), lest additive toxicity occur.
Evidence from a recent cardiac arrhythmia suppression trial (CAST), which found
increased mortality among cardiac patients on long-term treatment with type I antiar-
rhythmics, has raised new concerns about the use of tricyclic antidepressants in cardiac
patients. 57 Although these agents do, in fact, suppress extra cardiac beats, when cardiac
ischemia and anoxia occur, patients taking type I antiarrhythmics appear to die at a
significantly greater frequency than those taking placebo. This has led Glassman and
Roose, who have conducted research on the cardiac effects of tricyclics for many years,
to caution about the long-term use of tricyclic drugs in patients vulnerable to suffering
myocardial ischemia. 57 Newer antidepressant agents do not appear to carry this risk,
but, as Glassman and Roose have noted, their efficacy in depression of greater severity
may not be comparable to that of the tricyclics in some patients. ECT may be consid-
ered for some patients in this category.
There is no evidence that nontriCYclic antidepressants possess antiarrhythmic
capacities. Trazodone may trigger ventricular tachycardia in patients with preexisting
myocardial irritability. Because most drugs introduced to the market have been studied
only in patients with excellent health status, it may be some time before possible risks
of arrhythmogenicity are fully identified in newer antidepressant drugs. 58 ,59 However,
the SSRIs and bupropion have been used now for a long time in a lot of patients,
including patients with various cardiac conditions, without evidence of either antiar-
rhythmic or arrhythmogenic effects.
Guidelines for Clinical Use
1. Take a good medical (including cardiac) history, as well as a review of
2. Discuss possible cardiovascular reactions with the patient.
3. In young adult patients with no apparent cardiac problems, the choice of
drug need not take into account cardiovascular side effects. If these occur,
however, they should affect choice of subsequent treatments.

4. In older patients or those with manifest cardiac problems or history of

orthostatic hypotension:
a. Select a drug with few hypotensive properties. Less likely to cause
postural hypotension are bupropion, SSRI antidepressants, and ven-
lafaxine. Among tricyclics, nortriptyline is the least likely to cause
orthostatic hypotension, although it still has that potential. MAOI anti-
depressants and nefazodone have a fair likelihood of causing orthostatic
hypotension. ECT is an alternate option for patients with symptomatic
b. Advise patients with symptomatic postural hypotension to change from
lying or sitting to vertical position slowly and to try sleeping with their
head elevated.
c. Prescribe adequate fluid and salt intake and/or support stockings.
5. In older patients or those with manifest or potential tachycardia, use drugs
with minimal anticholinergic activity, such as SSRls, bupropion, venlafax-
ine, nefazodone, or MAOls.
6. In patients with conduction delays who require an antidepressant, select a
medication that does not alter cardiac conduction (e.g., SSRls, bupropion,
venlafaxine, nefazodone, or MAOIs) or ECT.
7. In older patients, those with known cardiac histories, and patients taking
SSRI antidepressants, be sensitive to treatment-emergent symptoms such as
dizziness, faintness, or syncope. These may reflect episodes of symptomatic
bradycardia. If this occurs (or is suspected), switch to a non-SSRI anti-
depressant and consult with a cardiologist.
8. As noted earlier, the findings from the cardiac arrhythmia suppression trial
have led to concern about the tricyclic antidepressants, with their type I
antiarrhythmic activity, in patients with ischemic heart disease and at risk for
possible infarctions. 57 If possible, treat these patients with antidepressants
without antiarrhythmic properties, e.g., SSRIs, bupropion, venlafaxine, or
nefazodone. ECT is another option. If tricyclics are warranted based on a
previous favorable response or lack of improvement with or tolerance to
other drugs, initiate treatment with low doses, monitor the EKG carefully
(consider Holter monitoring), work closely with a cardiologist, and monitor
blood antidepressant levels at steady state after each dosage adjustment.
9. For patients taking other medications that may have cardiovascular effects,
be sensitive to the possibility of pharmacodynamic or pharmacokinetic inter-
actions. A tricyclic antidepressant, for example, may act additively with a
drug that affects cardiac rhythm (e.g., an antiarrhythmic) or slows conduc-
tion [e.g., a ~-blocker or nonsedating antihistaminic, such as terfenadine
(Seldane®»). An SSRI may act additively with a drug that slows cardiac
conduction (e.g., a ~-blocker). Nefazodone is contraindicated with ter-
fenadine, astemizole, and cisapride, as it can cause increased concentrations
of these compounds, leading to a potentially lethal cardiac arrhythmia. Con-
current use of fluoxetine with tricyclics or other drugs that inhibit the cyto-
chrome P450 hepatic enzyme system can lead to increased concentrations of

the other drugs, sometimes to toxic levels. Interactions may also occur with
paroxetine but seem to be less of a problem with sertraline, at least at the
lower end of its dosage range (see Chapter 1, Table 1).
10. When treating depression in a patient with significant cardiovascular dis-
ease, who is likely to be taking one or more cardiotonic medications, work
closely with the patient's cardiologist. Consider consultation with a psychia-
trist well versed in psychopharmacology.

g. Neurological
i. Sedation/Stimulation. Tricyclic antidepressants tend to be sedating; tertiary
amines are even more so than their secondary-amine metabolites. Doxepin and amitrip-
tyline are particularly sedating, as is the nontricyclic trazodone. While SSRIs and other
newer antidepressants cause stimulation more often than sedation, many patients do
experience the latter, sometimes to a troublesome degree. Among newer antidepres-
sants, sedation may be problematic with nefazodone and mirtazapine. In some cases,
sedation becomes more of a problem after a patient has been taking an antidepressant
for weeks or several months.
If sedation is a problem during the first few hours following ingestion of an
antidepressant, advise the patient to take it at or soon before bedtime. Morning-after
sedation (hangover) may be a problem initially, but tolerance usually develops to the
sedative side effects of any drug. If reassurance and the use of bedtime dosing do not
alleviate sedation sufficiently, try lowering the dose. Since it is unclear which patients
will experience significant sedation from an antidepressant (except for the tertiary-
amine tricyclics, which are sedating for just about everyone), consider having the
patient take the first dose or two of an antidepressant on a weekend or holiday, when he
or she does not have to go to work and can experiment with the drug's side effects. Be
certain also that the patient has a phone number to reach a knowledgeable clinician
about early and possibly idiosyncratic reactions to the medicine. (Easy access to
information, reassurance, and guidance can enhance compliance and, therefore, effec-
tiveness of medications and decrease liability exposure.)
When a patient has been taking an antidepressant for a number of weeks or a few
months and then experiences a gradual increase in sedation and decrease in energy, a
downward adjustment in dose may be called for. Particularly with a drug with a long
half-life, such as fiuoxetine, steady-state levels (which typically take more than a
month to achieve) may be associated with the side effects, which signal the need to
lower the dose for longer-term treatment.
Fluoxetine (and to a lesser degree other SSRIs), bupropion, venlafaxine, and
nefazodone stimulate many patients, particularly initially. In such cases, ask the patient
to take more of the daily dose earlier in the day. Begin with the lowest possible dose
and raise the daily dose slowly. If the patient has difficulty falling asleep, the last dose
should be in the late afternoon. As with sedation, tolerance tends to develop to the side
effect of unwanted stimulation and difficulty falling asleep. Consider a dosage reduc-
tion if necessary, with very gradual dosage increments. In some cases, consider adding
a benzodiazepine or trazodone at bedtime for a short period of time to facilitate sleep.
A benzodiazepine during the daytime may help diminish excessive feelings of agitation

and anxiety. Benzodiazepines in general should be avoided in patients with histories of

sedative-hypnotic or other drug abuse.
ii. Headaches. Headaches sometimes accompany the use of some of the newer
antidepressants such as SSRIs, bupropion, and venlafaxine. Rarely, it may necessitate
change of drug. Headache is also a common symptom of the withdrawal reaction from
tricyclic antidepressants (see below).
iii. Nocturnal Reactions. Myoclonus, an abrupt jerking of a muscle group either
when conscious or during sleep, is an occasional side effect of tricyclic antidepressants.
Rarely, patients taking tricyclics report hypnagogic hallucinations, which may be alle-
viated by dividing the daily dose rather than administering the drug entirely at bedtime.
iv. Seizures. Many antidepressants seem to lower the seizure threshold. This is
rarely a problem clinically, although in patients with a preexisting epileptic condition,
the physician should try to keep the dose low and make increments slowly. The
clinician should also be aware of possible interactions with concurrent anticonvulsant
drugs. Patients withdrawing from alcohol or sedatives and those with eating disorders
may be at increased risk for seizures.
Infrequently, patients will experience seizures de novo when taking an antidepres-
sant drug. The risk is idiosyncratic and dose related, but with tricyclic antidepressants it
probably occurs in fewer than 1% of patients. Clomipramine presents a greater risk
than other tricyclics. Some newer antidepressants may carry greater risks of seizures.
With maprotiline, the risk appears to be four times higher than with tricyclics and to be
greater at higher doses. Amoxapine also may have an increased risk of seizures, which
are most frequent and severe in overdose. For bupropion, seizure incidence has been
calculated to be 0.44% for patients taking up to 450 mg per day, rising to 2.19% for
patients at higher dosages. 6o Venlafaxine has a lower rate of seizures than imipramine,
trazodone, and clomipramine, and nefazodone has not been shown to increase the risk
of seizures.
v. Anticholinergic Effects. Anticholinergic agents (i.e., tricyclics) interfere with
the encoding of new memories. Thus, although remote memories will remain intact, a
patient may have difficulty recalling events earlier in the day or learning new informa-
tion. The more anticholinergic the specific agent, the higher the dose, and the more
anticholinergic agents the patient is taking, the greater the memory defect is apt to
become. The elderly and the demented are at greatest risk. Rarely, a patient develops an
atropine-type delirium (see Section V.A.5).
vi. Movement Disorders. With the exception of amoxapine, antidepressants do not
cause the types of extrapyramidal reactions associated with neuroleptic drugs (see
Chapter 4). Tricyclics, however, cause a fine "action" or postural tremor, most readily
observed when a body part is kept in a sustained posture (e.g., hands outstretched).
Functionally, this tremor-an exaggeration of a physiological tremor-is apt to be-
come manifest when a patient drinks from cup and saucer. Finer and more rapid than a
parkinsonian tremor (which occurs at rest), this action tremor is made worse by anti-
cholinergic agents (do not use antiparkinson drugs), anxiety, and stimulants but might
improve with alcohol (not recommended therapeutically) or l3-blockers. It is akin to the
tremor induced by lithium61 or associated with thyrotoxicosis. It probably has a heredi-
tary component and worsens with aging.

Amoxapine is a neuroleptic drug and, as such, has been associated with the full
range of neuroleptic extrapyramidal reactions (see Chapter 4). These include acute
dystonias and dyskinesias, akathisia, Parkinson's syndrome, and late-onset and with-
drawal dyskinesias and dystonias.
There have been suggestions that fluoxetine may cause akathisia, although there is
some controversy as to whether the motor restlessness associated with nonsedating
antidepressants may not be a distinct condition. 62 The highly controversial (and highly
public) reports suggesting that fluoxetine might cause or exacerbate thoughts of suicide
(now largely disbelieved among psychiatrists) have been linked in at least some cases
by some authors to akathisia. 63 It is conceivable that some patients develop akathisia
while taking SSRI antidepressants (analogous to the same syndrome caused by neuro-
leptics) and experience feelings of desperation as a result of these feelings of wanting
to "jump out of their skin." There are also anecdotal data suggesting that when com-
bined with a neuroleptic, fluoxetine might enhance the probability of extrapyramidal
reactions. 62
vii. Withdrawal Reactions. Tricyclic antidepressants are associated with a distinct
withdrawal syndrome. As with most psychoactive compounds, the likelihood of a
withdrawal reaction and its severity vary as a function of the dose and the length of
time the patient has taken the drug at the time of withdrawal. It may be mitigated in part
by very gradual withdrawal.
Symptoms typically begin when the dose has been reduced to a particularly low
level or the drug is stopped entirely. Clinical manifestations that have been attributed to
anticholinergic rebound include both peripheral and central effects. Autonomic reac-
tions, the opposite of so-called anticholinergic symptoms, include hypersalivation,
nausea, vomiting, abdominal cramps, and diarrhea. Central manifestations typically
include headaches, vivid dreams, and sleep disturbances. Especially when patients have
taken these drugs chronically, withdrawal symptoms may persist for 2 months or
Symptoms associated with withdrawal from SSRIs include dizziness, light-head-
edness, headaches, confusion, memory difficulties, decreased energy, weakness, ex-
tremity tingling, chest tightness, tremor, nausea, anorexia, insomnia, nightmares, cold
hands, and sweating. The proposed mechanism is a decrease in serotonin. There is
insufficient data to describe withdrawal syndromes with some of the newer antidepres-
sants, but the likelihood is that they exist. Withdrawal symptoms with venlafaxine may
be particularly severe.
Gradual withdrawal of the antidepressant is the best preventive and management
strategy. At the lower end of the dosage scale, consider dividing tablets in half. An
alternative (or additional) strategy is to end the drug by administering a dose every
other day, then every 3 days, before discontinuing it entirely. If there is no rush to
discontinue the agent, tapering of the drug can be protracted over several months,
partjcularly in patients who have previously taken the antidepressant for years.
When symptoms are particularly uncomfortable, the patient can be returned to a
higher dose of the antidepressant and then the taper made more gradual. Antidotes to
tricyclic withdrawal can include a sedating antihistamine with anticholinergic proper-
ties [e.g., diphenhydramine (Benadryl® and others)], a less sedating centrally active

anticholinergic agent [e.g., benztropine (Cogentin® and others)], or, if peripheral auto-
nomic manifestations predominate, a peripherally active anticholinergic drug [e.g.,
methscopolamine (Pamine® and others)].
viii. Psychiatric Side Effects. It seems that virtually any effective antidepressant
can precipitate mania in some patients. Many such patients have a prior history of
mania and qualify for a bipolar diagnosis. Occasional patients, however, experience
mania only on "provocation" with an antidepressant drug, and some authorities have
suggested that these patients fall clinically along a bipolar spectrum. In DSM-IV,
however, a patient who has experienced an episode of mania or hypomania only while
taking an antidepressant drug does not qualify for a bipolar (or bipolar II) diagnosis.
Instead, the diagnosis is of mania or hypomania "due to" an antidepressant.
The psychotic symptoms of patients with schizophrenia may be made worse by
treatment with an antidepressant drug. When a full-blown depression is present in a
schizophrenic patient, however, it is acceptable to administer an antidepressant drug
along with an antipsychotic agent. Bupropion has been reported to exacerbate psycho-
tic symptoms in patients with schizoaffective disorder; this may be related to its
inhibition of neuronal dopamine uptake. 64
Patients with organic brain syndromes may become increasingly confused when
treated with a strongly anticholinergic drug such as a tricyclic antidepressant. Use of
antidepressants with minimal anticholinergic activity, however, may be considered.
A controversial article published in 1990 suggested that, in some patients, anti-
depressant treatment with fluoxetine could lead to de novo thoughts of suicide. 65 Many
other case series and retrospective analyses have failed to confirm an association
between fluoxetine or other antidepressants and new or exacerbated thoughts of sui-
cide. The prevailing wisdom is that untreated (or undertreated) depression carries a
much greater risk of suicide. Nonetheless, as noted earlier (see Section V.A.3.g.vL),
there may be occasional patients who react to side effects of a drug (such as akathisia)
by becoming deeply distressed and thinking about suicide. Clinicians must be continu-
ously alert to how patients may react to a new factor in their lives, such as a drug.
Common side effects should be predicted, and any patient newly starting on a medica-
tion should be instructed to call immediately and report any distressing reactions. A
knowledgeable clinician should be available 24 hours a day. Some patients are dis-
tressed that they do not feel better immediately and can be reassured. Others are deeply
troubled by one or another side effect of a new medicine and can be comforted by
information about the drug.

h. Cutaneous. Similar to most pharmaceutical compounds, antidepressants can

cause allergic skin reactions. Reintroducing the drug typically results in the abrupt
onset of a worse reaction. Therefore, if a patient with a history of a cutaneous allergy
requires an antidepressant drug, it is best to choose one as dissimilar chemically as
possible from the previously offending agent.

i. Autonomic. Like the phenothiazines, tricyclic antidepressants produce numer-

ous autonomic effects, predominantly anticholinergic (see Table 2). Patients most
commonly report dry mouth (usually early in treatment). Generally, physiological
tolerance develops to this adverse reaction. However, in some, dry mouth persists and

TABLE 2. Pharmacologic Properties of

Heterocyclic Antidepressants

Sedation Anticholinergic

High High
Amitriptyline Amitriptyline
Trimipramine Imipramine
Doxepin Trimipramine
Trazodone Doxepin
Moderate Moderate
Imipramine Amoxapine
Amoxapine Nortriptyline
Nortriptyline Maprotiline
Maprotiline Low
Sertraline Desipramine
Low Protriptyline
Desipramine Very low
Protriptyline Trazodone
Fluoxetine Fluoxetine
Bupropion Bupropion
Paroxetine Nefazodone
Fluvoxamine Venlafaxine
Venlafaxine Sertraline

may cause discomfort, mouth infections, and dental caries (sometimes rampant). In
severe cases of xerostomia, the clinician should intervene by:
1. Encouraging adequate hydration.
2. Lowering the dosage of medication if possible.
3. Discouraging the use of sugar-laden drinks, chewing gum, or candy.
4. Considering switching to a less anticholinergic (i.e., nontricyclic) antidepres-
5. Providing alternative means of mouth lubrication:
a. "Sugarless" gums (although these actually have sugars that in high doses
can contribute to the formation of caries).
b. Fluoride lozenges.
c. Direct lubricants (for a list of saliva substitutes, see Table 3).
6. Administering bethanechol chloride (Urecholine® and others), a peripherally
active cholinergic agent, 25 mg up to 3 times per day, in patients whose
symptoms do not respond to conservative measures, who would not be able to
complete a drug trial because of the severity of symptoms, and who are not at
increased risk from cholinergic stimulation (e.g., patients with asthma, poten-
tial cardiac compromise, or ulcer disease).
The major cardiovascular anticholinergic effect is tachycardia. Tolerance to this
reaction usually develops quickly. Blurred vision, resulting from failure of accom-

TABLE 3. Some Saliva Substitutesa

Drug Form

Carboxymethylcellulose or hydroxyethylcellulose solutions

Moi-Stir® (Kingswood Labs) Spray, swabs
Orex® (Young Dental) Spray
Sal-Ese® (North Pacific Dental) Liquid
Saliment® (Richmond) Aerosol
Saliv-Aid® (Copley) Liquid
Salivart® (Westport)b Aerosol
Saliva Substitute® (Roxane) Liquid
VA Oralube® (Oral Dis Res Lab)" Liquid
Xero-Lube® (Scherer)" Spray
Mucin-containing solution
Saliva Orthana® (A/S Orthana, Denmark) Spray, lozenges, gum
Mucopolysaccharide solution
MouthKote® (Parnell) Liquid
Glycerate polymer
Oral Balance® (Laclede) Gel

aReprinted from The Medical Letter l13 with permission of the publisher.

modation, occurs in up to 20% of patients. This symptom appears at the beginning of

treatment and rapidly resolves without dosage adjustment. If a patient has severe
blurring, and the dosage cannot be lowered, the clinician can give I % pilocarpine
nitrate eyedrops for a short time.
By blocking the muscarinic SUbtype of acetylcholine receptor, tricyclic and other
anticholinergic antidepressants (i.e., maprotiline, amoxapine) decrease motility in the
gastrointestinal and urinary systems. In most patients, this leads to some degree of
constipation and possibly urinary hesitancy. The elderly, those taking other anti-
cholinergic agents, and those with certain medical conditions may be at increased risk.
Paralytic ileus is the extreme case of intestinal hypomotility; patients who are recently
postoperative are at greater risk of developing this problem. A man with a prostate
condition, as another example, is at greater risk of developing acute urinary retention
when treated with strongly anticholinergic drugs.
Beyond being aware of these most serious situations, the physician will want to
help patients deal with milder but more common versions of these problems. Constipa-
tion often can be alleviated by dietary counseling, especially by encouraging an ade-
quate intake of fiber and fluids. In addition, a stool softener may help. Sugarless
candies and gum that contain sorbitol or xylitol serve a laxative function as well as
lubricating the mouth. In some cases, oral bethanechol can alleviate both urinary
retention and severe constipation, but it is important to avoid excessive cholinergic
In addition to atropine-type side effects, heterocyclic antidepressants cause other
autonomic imbalances. Although excessive sweating seems paradoxical in view of the

tricyclics' anticholinergic actions, many patients treated with imipramine, in particular,

experience hyperhidrosis. It is usually more annoying than dysfunctional. Newer anti-
depressants, such as SSRIs, venlafaxine, and bupropion, have minimal anticholinergic
potency but cause autonomic symptoms such as headache, anorexia, and nausea. As
noted earlier (see Sections Y.A.3.f.i and ii), some cause hypertension and bradycardia.

j. Sexual Side Effects. Psychoactive drugs frequently interfere with sexual desire
and function, and antidepressant agents are no exception. The mechanism of action
remains obscure, although the interplay of various autonomic effects must bear some
Delayed, incomplete, or altered orgasm is reported occasionally by both men and
women. Decreased libido also is sometimes described by both sexes. In addition, men
may experience difficulties achieving or maintaining a penile erection and/or seminal
ejaculation. Rates of sexual side effects seem to be highest with SSRIs and
clomipramine, followed by MAOIs and tricyclics, and least common with bupropion
and possibly nefazodone. In one study, clomipramine was found to be associated with
sexual side effects in over 90% of patients.66 The incidence of sexual dysfunction with
fluoxetine has been variably reported as 8% to 75%,67-70 although prerelease studies
found an incidence of only 1.9%.71 Reported rates of sexual dysfunction with sertraline
are 15.2% for men and 1.7% for women; for paroxetine, they are 23% for men, 13% for
women. Due to underreporting and based on clinical experience, however, the "true"
incidences are probably much higher.
Strategies for treating adverse sexual effects include (1) waiting for tolerance to
develop, (2) reducing the antidepressant dose, (3) changing to a different antidepressant
(especially bupropion72), and (4) adding a new medication. A weekend without the
antidepressant can restore sexual functioning temporarily for short-half-life agents (not
fluoxetine), but with a risk of increased symptoms of depression.?3
Adjunctive medications found to be helpful include the peripherally active para-
sympathomimetic bethanechol (Urecholine® and others), the serotonin antagonist cy-
proheptadine (Periactin® and others), the a-2 antagonist yohimbine (Yocon® and oth-
ers), amantadine, bupropion, and buspirone. None of these has been systematically
studied in controlled trials, but single cases and small, open series have reported
varying degrees of success. Cyproheptadine has been reported to reverse antidepres-
sant-induced anorgasmia in both sexes and impaired ejaculation in men. 74 - 81 An-
orgasmia from tricyclic antidepressants, one MAOI, and fluoxetine has been suc-
cessfully treated. In the majority of successful cases, patients used cyproheptadine on
an as-needed basis, taking 2-16 mg a few hours before sexual activity. It was also
effective when taken in a daily dose. Cyproheptadine has reversed the effects of the
antidepressant in seven cases. Yohimbine has been reported to ameliorate decreased
libido, erectile dysfunction, and anorgasmia caused by fluoxetine and clomipra-
mine. 67 .82 ,83 Effective doses have ranged from 5.4 to 16.2 mg taken as needed 2-4 hr
before sexual activity or 5.4 mg taken 3 times daily. Bethanechol has reversed the
sexual side effects of anorgasmia, impotence, and impaired ejaculation caused by both
tricyclic antidepressants and MAOIs with doses of 10-40 mg as needed or 30-100 mg
daily.84-88 Amantadine reversed fluoxetine-induced anorgasmia in five patients taking

loo-200 mg daily.89 Some clinicians are employing low doses of bupropion as an

adjunct to an SSRI to restore orgasmic capacity.90
Although the incidence is unclear, there have been case reports of women and,
more recently, men who have experienced an egosyntonic increase in libido while
taking trazodone. More troublesome with trazodone is priapism, a dysfunctional and
often painful erection of the penis not associated with sexual desire that requires
emergency urological intervention and can lead to permanent impotence. Eighty per-
cent of all psychotropic medication-induced cases of priapism have been linked to
trazodone, although the actual incidence of priapism in men taking trazodone is proba-
bly well under 1%.

k. Drug Interactions. Drug interactions fall into two broad categories: phar-
macokinetic and pharmacodynamic. The former occur when a drug affects another
drug's absorption, distribution, biotransformation, or excretion. Pharmacodynamic in-
teractions, by contrast, occur when two drugs have additive (or antagonistic) effects at
a target site. Thus, drugs with anticholinergic properties can be additive with the
anticholinergic properties of tricyclics. Hypotension can be a problem if another hypo-
tensive agent is added to an antidepressant that can lower blood pressure-e.g., tri-
cyclics or nefazodone. Other side effects increased by pharmacodynamic interactions
between drugs include sedation, stimulation, and sexual dysfunction.
Virtually all drugs with serotonergic properties-SSRls, venlafaxine, nefazodone,
and strongly anticholinergic tricyclics such as clomipramine-can produce a poten-
tially fatal hypermetabolic ("serotonin") syndrome when coadministered with (or pre-
scribed in close temporal proximity to) a MAOI. (For more on this syndrome, see
Section Y.B.3.b.)
A number of pharmacokinetic interactions are observed with modem antidepres-
sants. A growing understanding about the liver's P450 drug-metabolizing system (see
Table 1 in Chapter 1) has helped to clarify the nature of many of these. All SSRIs
except fluvoxamine are potent inhibitors of the P4502D6 isoenzyme. (Sertraline is less
potent at the lower end of its dosage range-50 mg daily.) Therefore, when SSRls are
coadministered with tricyclic antidepressants and some anti psychotics and antiar-
rhythmic agents, the levels of these other drugs may rise. The package inserts for
fluoxetine and sertraline caution against administering these SSRls with drugs that are
metabolized by P4502D6 and have a low therapeutic index, such as the tricyclics,
propafenone, and flecainide. Fluvoxamine is a potent inhibitor of the P4501A2 iso-
enzyme and can raise the levels of coadministered theophylline, haloperidol, clozapine,
some other antipsychotics, and tricyclics. Nefazodone inhibits the P4503A4 iso-
enzyme. It can thereby raise blood levels of cisapride and the nonsedating anti-
histamines terfenadine and astemizole and is considered contraindicated with these
three agents, because elevated blood levels can slow cardiac conduction and produce
potentially fatal ventricular arrhythmias. Other drugs whose levels can be raised by
nefazodone through its inhibition of the 3A4 system include the benzodiazepines
alprazolam, triazolam, and midazolam, as well as carbamazepine, quinidine, and lido-
caine. Many clinicians will lower the dose of coadministered drugs in anticipation of
this effect. Although it has not been definitely demonstrated that fluvoxamine inhibits

P4503A4, a substantial interaction has been observed between fluvoxamine and al-
prazolam, a drug that is known to be metabolized by the 3A4 isoenzyme. Therefore,
like nefazodone, fluvoxamine is contraindicated with terfenadine and astemizole and
should be used with caution with cisapride.
Mirtazapine is metabolized by several cytochrome P450 isoenzymes, including
2D6, lA2, and 3A4. Concomitant use of an inhibitor of one of these isoenzymes could
increase mirtazapine levels. Because mirtazapine does not inhibit P450 isoenzymes, it
probably has no effect on other drugs metabolized by them. If mirtazapine is combined
with alcohol or benzodiazepines, it may produce additive effects on sedation and
cognitive and motor impairment. Although there are no data on the combination of
mirtazapine with MAOIs, the manufacturer recommends that mirtazapine not be used
in combination with an MAOI or within 14 days of discontinuing one.
Fluoxetine, sertraline, and paroxetine (but not fluvoxamine and venlafaxine) are
highly bound to plasma proteins and can displace other drugs from their protein
binding sites, thus increasing their blood levels. This may effect the clinical activity of
such diverse agents as warfarin and carbamazepine.
Having available an increased number of antidepressants with different phar-
macologic profiles brings many benefits but also increases the potential for drug
interactions. Knowing the underlying basis for these potential interactions can help
clinicians to rationalize treatment regimens. In general, the addition or deletion of any
drug from a medication regimen should raise interaction as a possibility. Among
available antidepressants, as of now, venlafaxine appears the least likely to cause
interactions with other medicines.

I. Weight Changes. Patients taking tricyclic antidepressants often gain weight.

This may be a result of constipation or, less frequently, fluid retention. Most common-
ly, weight gain reflects an increase in body fat, the mechanism of which is a subject of
ongoing debate and investigation. Patients taking tricyclics may have an increased
appetite and sometimes a craving for sweets. Some researchers have suggested that
tricyclics decrease the resting metabolic rate.
For some of the newer antidepressants, changes in body weight are less obvious.
With SSRIs, many patients initially lose weight, although the long-term net changes are
uncertain, and some patients taking SSRIs report long-term weight gain. Bupropion
may also cause weight loss, possibly related to its stimulant properties. Weight gain
associated with tricyclic antidepressants tends to be persistent, while weight loss
caused by some of the newer agents may not last. Nonetheless, this difference in weight
change tendency among classes of antidepressants gives clinicians more flexible op-
tions in helping patients with these side effects.

4. Precautions for Administration during Pregnancy

No psychiatric drug should be routinely prescribed during the first trimester of

pregnancy. Tricyclic antidepressants, however, have been given during this period
without harmful effects on the fetus. Occasionally, the newborn of a mother who took
tricyclics may exhibit irritability, hyperhidrosis, tachycardia, tachypnea, and cyanosis

for several days following delivery. These symptoms do not seem to cause long-term
morbidity. Therefore, the pregnant patient with a severe depression accompanying a
serious suicidal risk or an inability to care for herself might receive a trial of a tricyclic
antidepressant. The clinician and patient should carefully weigh the risks and benefits
(see Chapter 10). The effects of MAOI and newer antidepressants on the developing
fetus are less clear; if possible, these drugs should be avoided in a pregnant woman.
ECT is an option to be seriously considered when a pregnant patient develops a
depression of moderate or greater severity.

5. Acute Toxicity

a. Intoxication Syndromes. Individual patients show a variable tolerance to anti-

depressants. Usual clinical doses may occasionally cause toxic responses resulting
from special sensitivity or from interaction with coadministered medications. Most
often, however, patients develop CNS depression and/or cardiotoxicity from an acute
or chronic overdose of tricyclics or similar antidepressants (i.e., maprotiline, amox-
apine) or MAOIs. The clinician should treat each overdose aggressively because of the
possibility of refractory cardiac arrhythmias and circulatory or respiratory collapse.
Prolonged observation is mandatory, since recovery from acute ingestions can be
delayed because of rapid and extensive tissue distribution.
Symptoms develop 1-4 hr after the overdose and depend on the age of the patient,
individual tolerance, and dose. Sometimes a large number of antidepressant tablets
taken at the same time clump together in the stomach and enter the small intestine as an
agglomeration. There may be a sudden release of material from this bezoar, which can
allow rapid absorption of great quantities of the drug. Clinically, the patient may
progress from ambulation and clear consciousness to coma and cardiorespiratory arrest
over a period of minutes. With tricyc1ics, doses equivalent to 1.2 g of imipramine
frequently are toxic (although doses of <1.0 g occasionally cause severe reactions
and death), and more than 2.5 g is commonly fatal. Prior to a marked decrease in
consciousness, hallucinations, sensitivity to sounds, delirium, or agitation may occur. A
"hyperactive" coma may develop and progress to a deeper nonreactive coma. Hypoten-
sion or hypertension, dilated, sluggishly reactive pupils, and hypothermia also occur.
Patients sometimes die from severe hyperpyrexia secondary to central anticholinergic
effects. Myoclonic seizures and bilateral plantar responses on neurological examination
are common. Arrhythmias, which are difficult to control, include ventricular tachycar-
dia, atrial fibrillation, and atrioventricular and intraventricular block. Manifestations of
cardiac effects appear on EKG as bundle branch block, varying degrees of heart block,
ventricular extrasystoles, and bizarre QRS complexes.
A large acute overdose may have a somewhat different character. Deep coma
often appears rapidly. Reflexes decrease, and bilateral plantar responses occur. Patients
frequently exhibit seizures, arrhythmias, and respiratory arrest. The pupils may be of
normal size, but they will react sluggishly or not at all. The pulse rate increases, and
blood pressure and body temperature decrease.
There is considerable controversy about the correlation of QRS duration, tricyclic
antidepressant blood level, and clinical outcome in overdose. One study failed to find a

statistically significant association between serum drug levels and the occurrence of
seizures or ventricular arrhythmias but did observe a correlation between QRS duration
and overdose outcome. 91 Patients with QRS durations of <0.1 sec had neither seizures
nor ventricular arrhythmias, whereas those with QRS 2:::0.1 sec had a 50% incidence of
ventricular arrhythmias. By contrast, another study failed to find either blood anti-
depressant concentrations or the EKG helpful in predicting seizures or ventricular
arrhythmias. 92 Still, all patients in this latter study with tricyclic antidepressant serum
levels of 2:::1000 ng/mL had QRS intervals of 2:::0.1 sec, and 2 cases that resulted in
fatalities of 102 consecutive tricyclic overdoses met these criteria.
The authors of the latter study recommend that an antidepressant level of 2::: 1000
ng/mL with a QRS interval of 2:::0.1 sec should be taken as a danger sign but that a
lower blood level or a normal QRS interval should not be a license to relax. Rather,
antidepressant blood level and QRS interval should be considered together with the
duration of time that has elapsed since the drug ingestion, the level of the patient's
consciousness, the presence of seizures, and the patient's vital signs in deciding on the
length of observation and the intensity of management.
Management of serious tricyclic overdoses poses serious difticulties, since
cardiorespiratory collapse, arrhythmias, and coma present both an immediate
and a persistent life threat (see Table 4). Any patient with a suspected tricyclic
antidepressant overdose-no matter how small the amount alleged to have been taken
or how alert the patient seems-should be rushed to an emergency room. Many
fatalities occur before a patient even reaches medical help. In the hospital, an IV line
should be established immediately, and the stomach emptied by gastric aspiration and
lavage. Next, a physician should administer 50-100 mg of activated charcoal, either as
an oral slurry mixed with water or by orogastric tube; it is too thick to administer via a
nasogastric tube. The clinician should avoid mixing it with milk, jelly, or other food-
stuffs, which can interfere with the charcoal's adsorptive capacity. Repeated adminis-
tration of charcoal every 4 hr can help remove additional quantities of the antidepres-
sant, which is recirculated via the enterohepatic system.
Catharsis can further speed elimination of the drug. Some preparations of charcoal
come mixed with sorbitol, which serves this purpose. An alternative cathartic is magne-
sium sulfate, 250 mg/kg per dose as a 20% solution.
An EKG should be performed immediately when the patient arrives at the emer-
gency room. If the patient has altered consciousness, arterial blood gases should be
drawn. Blood should also be drawn to establish antidepressant concentrations (which
may later rise if absorption from the gastrointestinal tract continues), to complete tests
to rule out the presence of other toxins (most overdoses are multiple), and to identify
any coincident medical conditions or complications.
The crucial time for observation in antidepressant overdoses is the first 6 hours. If
a patient has no alterations in consciousness, EKG abnormalities, respiratory depres-
sion, seizures, or hypotension, he or she can be given a final dose of charcoal and be
transferred to a psychiatric unit or given an appointment for psychiatric follow-up. On
the other hand, the presence of any clinical sign of toxicity should prompt immediate
admission to a medical intensive-care unit for at least 48 hours of cardiac monitoring.
Treatment of life-threatening consequences of heterocyclic antidepressant over-

TABLE 4. Management of Tricyclic Overdose

Initial measures
Induce emesis and catharsis (magnesium sulfate, 250 mg/kg in 20% solution)
Implement gastric aspiration and lavage with activated charcoal (Charcodote® powder), 50-100 mg,
every 4 to 6 hr for 24 to 48 hr
Complete physical examination and obtain vital signs
Adequately ventilate; monitor by EKG
Insert an intravenous line with cardiac pacing capacity
Insert pulmonary artery catheter
Obtain blood chemistries including antidepressant serum level
Maintenance procedures
Monitor vital signs, electrolytes, serum heterocyclic levels, and EKG
Maintain fluid and electrolyte balance
Ensure adequate ventilation and skin care if the patient is comatose
Elevate legs
Adequately hydrate
Administer phentolamine (Regitine®), 5 mg (if persistent and severe)
Administer diazepam (Valium® and others), 5-10 mg 1M or IV pm (avoid barbiturates)
Central and peripheral anticholinergic syndrome
Slowly inject physostigmine (Antilirium®), 1-2 mg IV every 30 to 60 min pm
Use ice mattress, ice packs, or cold sponges
Cardiac interventions
Avoid type A antiarrhythmics [procainamide (Pronestyl® and others), quinidine, and disopyramide
(Norpace® and others) I
Use lidocaine (Xylocaine® and others) for ventricular arrhythmia
Use lidocaine or phenytoin (Dilantin® and others) for heart block arrhythmia
Use physostigmine for supraventricular arrhythmia; avoid in cases of cardiovascular or respiratory
Implement volume expansion for decreased left atrial pressure
Use dopamine (Intropin®) for increased left atrial pressure
Alkalinize the urine by giving IV NaHC0 3 (acid imbalance may predispose to ventricular irritability)
Persistent arrhythmia with heart failure
Use cardiac pacing or cardioversion

doses is complex and incompletely formulated. Because of their high degree of li-
pophilicity and extensive tissue and protein binding, heterocyclic antidepressants are
minimally removed by dialysis. The key medical treatments are symptomatic and hinge
on treating (or preventing) seizures, respiratory arrest, cardiac arrhythmias, and resul-
tant complications, such as renal failure and brain damage.
Seizures are often treated with intravenous diazepam, 5-10 mg repeated as
needed. Seizures may be more common with maprotiline and amoxapine overdose.
Amoxapine-related seizures, in particular, appear difficult to treat. Very high doses of
diazepam might be helpful.
If a patient lapses into stupor or coma, respiration should be monitored continu-
ously, respiratory arrest anticipated, and ventilatory assistance readily accessible. Ap-

propriate treatment of hypertension, or more commonly, hypotension must depend on

individual circumstances.
Treating cardiac arrhythmias that result from antidepressant overdoses presents a
thorny clinical agenda. Alkalinization, whether by intentional hyperventilation or by
infusion of sodium bicarbonate, may improve outcome, particularly in the face of
acidosis. Other approaches include electrical pacing, j3-blockers, j3-stimulating agents,
and phenytoin (Dilantin® and others). Physostigmine (Antilirium®), which can reverse
some of the eNS features of atropine-type toxicity (Le., delirium), is likely to worsen
cardiac rhythm, may possibly interfere with respiration, and may lower the seizure
threshold, so it should not be used in circumstances of cardiovascular and respiratory
compromise. Type IA antiarrhythmic drugs-Le .• procainarnide, quinidine, or dis-
opyrarnide-should be avoided entirely in cases of tricyclic overdoses. Their cardiac
effects are apt to be additive with those of the antidepressants.
When treating previously healthy patients suffering from even devastating anti-
depressant overdoses. intensive-care personnel should not give up hope prematurely.
With adequate external cardiovascular perfusion and oxygenation, patients can survive
intact after even several hours of resuscitation efforts.
Postmortem blood levels of tricyclics cannot be assumed to indicate overdose or
even intoxication, due to postmortem redistribution of these drugs. The release of
tricyclics from antemortem binding sites at death can lead to massive increases in
tricyclic blood levels. especially if there is a delay between death and the autopsy.
Samples taken from the liver may be more reliable than blood samples for differentiat-
ing acute overdose from therapeutic usage.
The incidence of fatalities following tricyclic overdoses probably lies between
0.5% and 4%, depending on the amount ingested, the presence of other drugs, and the
patient's prior medical status. Lethality associated with MAO! overdoses is less clear
but should be assumed to be high. Maprotiline may cause a higher incidence of seizures
than tricyclics in overdose, but its mortality figures appear comparable. Amoxapine
overdoses appear to yield a higher rate of fatalities, certainly produce more seizures,
and have been associated with acute tubular necrosis of the kidney and brain damage
following recovery. Trazodone, when taken alone in overdose, appears comparatively
One of the most encouraging findings in antidepressant pharmacotherapy in recent
years has been the accumulating evidence that SSRI and other new antidepressants are
truly less hazardous when taken in overdose. In a comparative study, fluoxetine was
found to be two and a half times safer than imipramine. 93 With venlafaxine, the 3
deaths that occurred in 30 reported cases of overdose all involved multiple drug
ingestion and/or alcohol. Most patients recover without sequelae from overdoses with
bupropion, although seizures occur in approximately one-third of cases. The rare
fatalities have been associated with massive doses of the drug. Multiple uncontrolled
seizures, bradycardia, cardiac failure, and cardiac arrest prior to death were reported in
these patients.
In premarketing clinical trials of nefazodone, there were no deaths associated with
overdose in doses up to 11,200 mg. Of course, prerelease studies tend to be much more
selective, and the patient population much more restricted. Fluvoxamine appears to be

relatively safe in overdosage: in over 300 case reports involving doses of up to 10 g,

there has been only one reported fatality in a patient taking only fluvoxamine. Since
suicide is a risk for many patients suffering from depression, drugs that are less likely
to be lethal if a patient impulsively decides to take the drug in overdose have a distinct
advantage, and a physician may often consider lethality in overdose in deciding on an
antidepressant of first choice for a patient.

b. Anticholinergic Syndromes
i. Description. Many psychoactive drugs have both peripheral and central anti-
cholinergic effects. Anticholinergic effects occur with tricyclics, maprotiline, and
amoxapine, antipsychotics, some hypnotics, antihistamines, and anti parkinson agents.
Combinations of these drugs, such as a psychoactive drug (antidepressant or antip-
sychotic or both) and an antiparkinson agent, may produce additive anticholinergic
effects. Older patients seem particularly sensitive. Acute overdoses of the above-
mentioned agents also frequently cause anticholinergic crises.
The anticholinergic syndrome may present with a mixture or a predominance of
either peripheral or central symptoms. In its florid state, the CNS picture consists of
confusion, delirium with disorientation, agitation, visual and auditory hallucina-
tions, anxiety, motor restlessness, pseudoseizures (myoclonic jerks and choreo-
athetoid movements with EEG seizure activity), and a thought disorder (e.g.,
delusions). The peripheral syndrome may be manifested by decreased bowel
sounds and constipation, urinary retention, anhidrosis, mydriasis, dry mouth,
cycloplegia (decreased accommodation), increased body temperature, motor inco-
ordination, flushing, and tachycardia. When these syndromes are caused by hetero-
cyclic antidepressants, there is also a high risk of life-threatening arrhythmias. Alipha-
tic and piperidine phenothiazines [especially thioridazine (Mellaril® and others)] and
the tricyclics amitriptyline, doxepin, and imipramine are the most anticholinergic. A
combination of the above drugs or of these drugs with other anticholinergic agents
greatly increases the risk of an anticholinergic syndrome.
ii. Anticholinesterase Therapy. Anticholinesterase therapy has proven very effec-
tive for the treatment of anticholinergic syndromes. All anticholinesterases counteract
the peripheral manifestations of the syndrome. However, clinicians should use either
physostigmine or pyridostigmine because they cross the blood-brain barrier and
therefore counteract the central symptoms. Most clinical studies have focused on
physostigmine, but equivalent doses of pyridostigmine also are effective. 94 •95 In gener-
al, the clinician should avoid using physostigmine in patients with unstable vital
signs. In these cases, cardiac arrhythmias from cholinergic stimulation are common.
Cholinergic effects can also produce seizures (particularly with rapid injection of
physostigmine) and respiratory arrest in selected patients. Although practitioners have
tried many different regimens, physostigmine salicylate, 1-2 mg 1M or IV, will
relieve symptoms dramatically. The clinician should infuse the drug very slowly (e.g.,
1 mg over 2 min), monitor the cardiac status, and have means for respiratory
support available. If no improvement occurs within 15-20 min, another dose of 1-
2 mg should be given. Up to 4 mg may be administered over 10-15 min. The body
degrades physostigmine almost completely within 11/2-2 hr. Since the toxic agents may

disappear more slowly, additional 1- to 2-mg doses at 30-min intervals may be neces-
sary, even if the initial treatment is successful.
Although physostigmine treatment provides dramatic and sometimes lifesaving
relief from heterocyclic-induced arrhythmias, it also has risks from cholinergic stimula-
tion. Excessive acetylcholine can result in tearing, salivation, rhinorrhea, sweating,
pallor, bronchial constriction, hypotension, muscle weakness and fasciculations, nausea
and vomiting, abdominal cramps, urinary frequency, and bradycardia. The antidote is
SSRIs and other new antidepressants have negligible anticholinergic effects. This
allows clinicians considerable flexibility in avoiding anticholinergic toxicity where it
has been or may become a problem.

6. Pharmacokinetics
Tricyclic antidepressants (except protriptyline) are generally rapidly absorbed
from the gastrointestinal tract, typically peak in the plasma within 2-8 hr, and remain
unbound for about 30 min. Between 80 and 90% of the drug is protein bound, although
individual differences in binding can produce a fourfold variation in the amount of free
drug. Tissue distribution and the first step in metabolism of tertiary cyclics, demethyla-
tion, occur rapidly. Heterocyclics typically have a high volume of distribution because
they are quite soluble and actively bind to various tissues. Patients usually achieve a
steady state for each dosage after 1 to 4 weeks (depending on the drug used). Liver
metabolism, including demethylation, hydroxylation, and glucuronide conjugation,
also accounts for large variations in serum levels. Individual differences in patients'
microsomal enzyme activity produce steady-state plasma concentrations (bound and
unbound) that may vary as much as 40-fold. These characteristics suggest that individ-
ual patients may require widely varying doses to produce clinically effective serum
During the initial pass through the liver, 30-70% of the drug is degraded ("first-
pass effect"), as only 30-70% of orally administered antidepressant is "biodegrad-
able." Doxepin may produce lower steady-state plasma levels than other tricyclic
antidepressants per milligram ingested. This might reflect lower plasma protein bind-
ing and greater first-pass degradation. Protriptyline, by contrast, might produce higher
plasma levels per milligram, possibly because of comparatively low first-pass metabo-
lism. The first demethylation of tertiary amines results in an active compound, whereas
a second demethylation and hydroxylation at the 2-position of the central ring of
tricyclics produce degradation products (probably therapeutically inactive but possibly
with cardiotoxic effects). Glucuronidation occurs after hydroxylation and makes the
derivative water-soluble. Initially, about 50% of tricyclics are excreted through the bile,
but because of an active enterohepatic circulation, two-thirds of the drugs are eventu-
ally eliminated in the urine. Clearance of the cyclics is relatively slow, but they have a
wide range of half-lives (see Table 5).
Fluoxetine is well absorbed with a small first-pass effect. Food does not affect the
extent of absorption, although the rate may be slightly decreased. Fluoxetine is metabo-
lized by demethylation in the liver by the cytochrome P4502D6 isoenzyme to the active

TABLE 5. Mean Half-Life of Heterocyclicsa

Drug Time (hr)

Amitripty line 16
Amoxapine 30
Bupropion 14
Clomipramine 24
Desipramine 22
Fluoxetine 48
Fluvoxamine 20
Imipramine 16
Maprotiline 47
Nefazodone 3
Nortripty line 24
Paroxetine 24
Protriptyline 126
Sertraline 25
Trazodone 5
Venlafaxine 4

aHalf-life varies greatly among patients and tends to increase

significantly with age.

metabolite, norfluoxetine, and other unidentified metabolites. As discussed earlier,

fluoxetine inhibits the 2D6 isoenzyme, causing elevated levels of any coadministered
drugs also metabolized by this enzyme. Elimination half-life is 2-3 days for fluoxetine
and 7-9 days for norfluoxetine. The onset of action is between I and 4 weeks. It takes
6-8 hr after a single oral dose of 40 mg to reach a peak serum concentration of 15-55
ng/mL. After multiple doses (40 mg/day for 30 days), the peak concentration of
fluoxetine is 91-302 ng/mL and that of norfluoxetine is 72-258 ng/mL. Eighty per-
cent of fluoxetine is excreted in the urine, and 15% in the feces. Fluoxetine is not
dialyzable because of high protein binding (94.5%).
Absorption of sertraline is slow but consistent. Bioavailability is increased if
sertraline is taken with food, possibly due to extensive first-pass metabolism in the
liver. Both sertraline and its metabolites are extensively distributed into tissues. Protein
binding is very high (98%). The primary initial pathway is N-demethylation to form
N~desmethylsertraline, which is substantially less active than the parent compound.
Both sertraline and N-desmethylsertraline undergo oxidative deamination and subse-
quent reduction, hydroxylation, and glucuronide conjugation. Sertraline is also metabo-
lized by and inhibits the P4502D6 isoenzyme. Sertraline's elimination half-life is 24-
26 hr; N-desmethylsertraline's is 62-104 hr. Sertraline's onset of action occurs within
2-4 weeks. The time to reach mean peak plasma concentration following administra-
tion of 50-200 mg of sertraline once daily for 14 days ranges from 4.5 to 8.4 hr. About
40-45% of sertraline is excreted in the urine within 9 days after a single dose, with less
than 0.2% recovered unchanged. Another 40-45% is excreted in the feces over the
same time period, including 12-14% unchanged sertraline.
Paroxetine is well absorbed, with bioavailability ranging from 50 to 100%. Bio-
availability increases after multiple dosing owing to partial saturation of extensive first-

pass metabolism in the liver. Absorption is not influenced by food. Paroxetine is

extensively distributed into tissues, with only 1% remaining in the systemic circulation.
As with the other SSRIs, protein binding of paroxetine is very high (95%). At least
85% of a paroxetine dose is oxidized to a catechol intermediate that undergoes subse-
quent methylation and conjugation to clinically inactive glucuronide and sulfate met-
abolites. Metabolism is accomplished in part by cytochrome P4502D6; partial satura-
tion of this enzyme at clinical doses appears to account for the nonlinear kinetics
observed with increasing dose and duration of paroxetine treatment. Paroxetine inhibits
cytochrome P450 enzymes to a lesser degree than fluoxetine, but more than sertraline.
In healthy adults, the elimination half-life is about 24 hr with a range of 3-65 hr. Half-
life is prolonged in patients with severe hepatic or renal function impairment and
possibly in the elderly. The onset of action is within 1-4 weeks. The range of time to
peak concentrations is 2-8 hr. Steady-state serum concentrations are achieved by 7 -14
days in most patients. In the 1O-day period following administration of 30 mg of a
paroxetine solution, approximately 64% of the dose was excreted in the urine, of which
2% or less was the parent compound. In the same period, about 36% of the dose was
excreted in the feces, of which unchanged paroxetine comprised less than 1%.
Well absorbed after oral administration, fluvoxamine reaches peak plasma con-
centrations in 2-8 hr. Its absorption is not significantly affected by food. Fluvox-
amine's mean elimination half-life is about 19 hr after a single dose and about 22 hr
following multiple doses. Fluvoxamine is extensively metabolized in the liver, but its
metabolites are not believed to be active in the brain. More than 90% of the drug is
eliminated in the urine as metabolites. Cirrhosis increases plasma concentrations and
half-life, suggesting the need for lower doses in patients with hepatic impairment.
Limited data in the elderly indicate a pharmacokinetic profile comparable to that in
younger patients.
Bupropion is rapidly absorbed from the gastrointestinal tract. Protein binding is
high (75-85%). There is extensive pre systemic or first-pass metabolism, with the
metabolites possibly having lesser therapeutic activity than the parent drug. The elim-
ination half-life is approximately 14 hr with a range of 8-24 hr. Onset of action is 1-3
weeks. The time to peak concentration is within 1-3 hr, followed by biphasic decline.
Less than 1% of bupropion is excreted in the urine unchanged. Over 60% is excreted as
metabolites within 24 hr, and over 80% within 96 hr. Less than 10% of bupropion and
metabolites is excreted in the feces.
Venlafaxine is well absorbed from the gastrointestinal tract, metabolized by cyto-
chrome P4502D6 in the liver, and excreted in urine. The elimination half-life of the
drug is about 5 hr; the half-life of its major active metabolite, O-desmethy1venlafaxine,
is 9-11 hr. Approximately 87% of a dose is excreted in the urine within 48 hr. Only 2%
of a dose is recovered in feces after 35 days. Unlike the SSRIs, venlafaxine is only 27%
protein bound. Venlafaxine and O-desmethylvenlafaxine reach steady-state concentra-
tions in plasma within 3 days.
Nefazodone is rapidly and completely absorbed after oral administration with
peak plasma concentrations 1-3 hr after dosing. Because it is subject to extensive first-
pass metabolism, its absolute bioavailability is about 20%. Nefazodone is extensively
metabolized in the liver via the P4503A4 isoenzyme to its three metabolites. Highly

protein bound, nefazodone could increase the free concentrations of other highly pro-
tein-bound drugs if coadministered. After oral or IV administration, urinary excretion
of the parent drug is negligible. Steady-state plasma levels are achieved by the fourth or
fifth day of multiple dosing. At steady state, nefazodone half-life is dependent on dose:
the mean value varies from about 2 hr at twice-daily doses of 50 mg to between 4 and 5
hr at twice-daily doses of 300 mg. Decreased clearance of nefazodone has been ob-
served in the elderly, in hepatic ally impaired individuals, and in subjects who are poor
metabolizers of dextromethorphan. A moderate decrease in nefazodone plasma levels
occurs when the drug is administered with food.
Mirtazapine is rapidly and completely absorbed following oral administration and
extensively metabolized in the liver. Peak plasma levels occur within about 2 hr after
an oral dose. Food has a minimal effect on the rate and extent of absorption. Mir-
tazapine is excreted mainly in urine and has a half-life of 20-40 hr. Steady-state levels
are reached within 5 days. Mirtazapine is approximately 85% bound to plasma pro-

7. Plasma Levels

Knowledge about the correlation between plasma levels of antidepressant drugs

and their clinical effects remains limited. Correlations of efficacy are fairly good with
nortriptyline, reasonable with imipramine, and limited or controversial with the re-
maining heterocyclic compounds.
For nortriptyline, research evidence suggests a curvilinear ("therapeutic window")
correlation, with maximum antidepressant efficacy achieved between approximately 50
and 150 ng/mL; above and below that "window," patients are less likely to benefit.
With imipramine, there is evidence of a straight linear correlation: maximum benefit
comes with levels of imipramine and desipramine above 200 ng/mL; there is no
obvious upper limit at which efficacy begins to decline, short of toxicity. In both dose-
response studies and blood level studies, with imipramine the rule seems to be that
more is better. With other tricyclic and newer antidepressants, the meaning of blood
levels is insufficiently understood to recommend routine monitoring (Tables 6 and 7).
Generally, however, there are broad "intuitive" ranges of antidepressant levels that
clinicians can use for guidance in selected situations. With most tricyclic antidepres-
sants, for example, a concentration much under 50 ng/mL probably is subtherapeutic,
whereas one much above 500 ng/mL could be toxic. Very low levels could signal that a
patient is inconsistently compliant with the medication regimen, or kinetic factors
could be at fault: the patient may have a problem with absorption, for example, or may
be a particularly rapid metabolizer. At the other end of the kinetic spectrum, a patient's
sensitivity to side effects at relatively modest drug doses may mean that he or she is an
unusually slow metabolizer, achieving high blood levels that can be measured despite
comparatively low doses.
At times, a physician may want to increase the daily dose of an antidepressant
beyond usual ranges. Thus, a patient who has responded minimally to imipramine up to
300 mg/day but whose side effects are minor may benefit from gradual dosage incre-
ments, perhaps 50 mg/day every week. In such a case, the doctor may find it reassuring

TABLE 6. Reasons for Monitoring Plasma Concentrations of Tricyclics

To document serum levels of cyclics in "nonresponders" after a few weeks of generally adequate doses
of medication
To determine if a reticent patient is taking medication
To monitor plasma levels in overdose cases
To monitor plasma levels in medically ill patients
To document serum levels within the "therapeutic window" for nortriptyline (i.e., 50-ISO mg/mL)
To determine if a potentially therapeutic dose of imipramine has been reached (Le., >200 mg/mL)
To minimize dosage levels in elderly patients
To determine plasma level in the context of a severe adverse reaction or persistent, bothersome side

to order a plasma level determination (of imipramine and desipramine) along with an
EKG to ensure against unanticipated toxicity. If the blood level is still comparatively
low (say, under 200 ng/mL), the physician is justified in this instance in gradually
elevating the imipramine dose until either the depression remits or the patient is unable
to tolerate side effects. With each successive increment in dose, it would be prudent to
order a repeated blood level and EKG. Some patients may require (and tolerate)
imipramine doses in excess of 500 mg/day.
In treating toxicity, plasma level determinations can help to establish when the
peak occurred and how high the peak was and to follow the egress of the drug. Plasma
tricyclic antidepressant levels in excess of 1000 ng/mL have been correlated with
severe and life-threatening toxicity, but lower levels should not be taken as a guarantee
of safety, since absorption may still be progressing in some cases. There is no clear
correlation between blood levels and toxicity because of interindividual differences in
sensitivity and also the fact that a patient's response to a given blood level when
concentrations are rising may not be identical to what happens at the same blood level
on the way down.
Correlations between plasma concentrations and therapeutic response or toxicity
remain obscure with SSRIs and newer antidepressants.

8. Preparations and Dosage

The wide range of effective clinical doses of heterocyclic antidepressants results
in part from individual differences in absorption, plasma binding, first-pass metabo-

TABLE 7. Precautions in Obtaining Heterocyclic

Plasma Levels

Generally use Venoject® (Kimble-Terumo, Inc.) or glass syringes

Do not use rubber stoppers
Take blood samples 12 hr after the last drug dosage
Promptly centrifuge sample to avoid hemolysis
Use a laboratory with proven reliability

lism, and liver enzyme activity. The tricyclic compounds amitriptyline, imipramine,
and doxepin have usual dose ranges of 150-300 mg/day. The tetracyclic maprotiline
has comparable potency, but because of its dose-related tendency to provoke seizures,
its daily dose should be capped at 225 mg. Nortriptyline is about twice as potent,
whereas protriptyline is about four times as potent. Doxepin is somewhat less potent
than the other dimethylated tricyclics. Amoxapine and trazodone are about half as
potent as imipramine, with clinically effective doses ranging from 150 to 600 mg/day,
although there is a hint that trazodone is more effective at lower than at higher doses.
When fluoxetine was introduced in 1988, it appeared to have a very simple dose-
response relationship. The prevaling wisdom at the time was that most patients bene-
fited optimally from a daily dose of 20 mg. Since then, the picture has become
complicated. As noted in other chapters (see Chapters 5, 12, and 13), dose-response
relationships may vary according to the condition being treated. Even in a single
condition (at least as homogeneous as we can become in modem psychiatry), such as
depression, the dose-response relationship appears to vary widely among different
patients, presumably due to interindividual differences in pharmacokinetics but also to
possible differences in actual effects in brain.
Most healthy adults of average size will still start treatment with 20 mg of
fluoxetine daily. If side effects are a problem, the dose can be reduced, with the
possibility of gradual increases. For the first few years of its availability, fluoxetine was
available only in 20-mg capsules. To take daily doses below 20 mg, the patient had to
break the capsule open and dissolve the contents in liquid, thereby creating a prepara-
tion that could be dispensed in lower dosages. More recently, fluoxetine has become
available in a liquid preparation and in a lO-mg capsule. Although data gleaned from
large, controlled trials still suggest an optimal daily dose for depressed adults of 20 mg,
with no apparent benefit from increased dosages, clinical observations suggest that
some patients may benefit from daily doses up to and beyond 100 mg. 96 Still other
patients are unusually sensitive to fluoxetine's side effects and might tolerate the drug
better at daily doses below 20 mg, with possible increments as tolerance develops. As
noted earlier, some patients become sedated, fatigued, and apathetic after weeks or a
few months of treatment with fluoxetine. 97 Such individuals may benefit from a reduc-
tion in the daily dose. Given the long half-lives of fluoxetine and norfluoxetine, the
reduction in average daily dose can be achieved (with improved daily cost of the
medication) by the use of 20-mg fluoxetine capsules in staggered daily dosing sched-
ules: e.g., one capsule every other day, one or two capsules in 3 days, etc.
Dose-response relationships for SSRI antidepressants introduced more recently
than fluoxetine are even less clear-cut. With sertraline, many patients appear to reach
optimum therapeutic benefit at a daily dose of 50 mg. Some tolerate the drug better
when started at 25 mg/day, at least for the first few days. Depending on therapeutic
response, many clinicians increase the dosage of sertraline by 50-mg/day increments
every 2 weeks up to a maximum of 200 mg daily. The dose-response curve for
sertraline remains a matter of active debate.
For paroxetine, the recommendation is to begin adult patients at a daily dose of 20
mg and increase the dose every 2 weeks up to a maximum daily dose of 50 mg.
Occasional patients benefit from even higher doses. Older patients should be started at

10 mg/day. The effective antidepressant dose of fluvoxamine is 100-200 mg daily,

with some patients requiring up to 300 mg/day. Initiating treatment with 50 mg/day for
a week may minimize nausea. The manufacturer recommends divided doses for dos-
ages greater than 100 mg/day. Although the elderly do not require lower doses, they
may benefit from a more gradual titration. Patients with hepatic or renal insufficiency
also should have lower initial dosages with careful monitoring of laboratory indices.
Results of dose-response studies suggest that nefazodone should be initiated at a
dose of 100 mg bj.d. in most patients, with an increase to 400 mg/day within 1-2
weeks based on clinical response. Elderly patients and those who cannot tolerate an
initial dose of 100 mg bj.d., often because of stimulation, can be started at lower doses,
such as 50 mg bj.d. Some patients may need a subsequent increase to 500 mg daily.
The highest recommended dose is 600 mg/day. For patients 65 years or older, the
therapeutic range may be 100-400 mg daily. There is a suggestion of a positive
relationship between dose and probability of response: that is, more is better.
The initial dosage of venlafaxine recommended by the manufacturer is 75
mg/day, divided into two or three doses. If necessary, the dosage can be increased by
75 mg/day at 4-day intervals, up to a usual maximum of 225 mg/day. Some hospi-
talized patients may only respond to higher doses, up to a maximum of 375 mg/day
(see Table 8). With venlafaxine, higher doses appear more likely to be effective. Side
effects like nausea often limit the rate of dosage increase.
Like venlafaxine and nefazodone, bupropion must be administered in divided
daily doses. A typical starting dose for bupropion is 100 mg bj.d., raised after several
days to 100 mg tj.d. With gradual dosage increments, the total daily dose may go as
high as 150 mg t.i.d., which should not be exceeded to avoid a high risk of seizures.
The manufacturer recommends that no individual dose be above 150 mg and the total
daily dose not exceed 450 mg.
The recommended starting dose of mirtazapine is 15 mg/day, given in a single
dose at bedtime. The dose should not be increased more than once every 1 to 2 weeks.
The maximum recommended dosage is 45 mg/day. Elderly patients and those with
moderate to severe renal or hepatic disease may require lower doses to minimize

B. Monoamine Oxidase Inhibitors

1. Chemistry

Based on their chemical structure, monoamine oxidase inhibitor (MAOI) anti-

depressants are subdivided into hydrazines and nonhydrazines (Fig. 4). The one hy-
drazine marketed in the United States as an antidepressant is phenelzine (Nardil®).
[Isocarboxazid (Marplan®), another hydrazine, was withdrawn by its manufacturer in
1994 but is still available by special request.] Structurally, phenelzine and isocarbox-
azid resemble the prototype MAOI, iproniazid, which was withdrawn from the market
because of hepatocellular toxicity. The one nonhydrazine MAOI antidepressant is
tranylcypromine (Parnate®), which resembles amphetamine structurally and clinically.

TABLE 8. Preparations and Dosage of Heterocyclic Antidepressants

Initial daily therapeutic
Heterocyclic Trade name Preparations dosage dosage

Amitriptyline Elavilill> and others 10-, 25-, 50-, 75-, 100-, and 25-75 mg 150-300 mg
150-mg tablets; 25- and 20-
mg capsules; vials of 10 ml
with 10 mg/ml
Imipramine Tofranilill> and 10-, 25-, 50-, and lOO-mg 25-75 mg 150-300 mg
others tablets; 75-, and 150-mg
capsules; ampules of 20 ml
with 25 mgl2 ml
Doxepin Sinequanill> and 10-, 25-, 50-, 75-, 100-, and 25-75 mg 150-300 mg
others 100- and 150-mg capsules;
solution of 10 mg/ml
Nortriptyline Pamelor® and 10- and 25-mg capsules; 20-40 mg 75-150 mg
others solution of 10 mg/5 ml
Desipramine Norpramin® and 25-, 50-, 75-, 100-, and 150- 25-75 mg 75-200 mg
others mg tablets; 25- and 50-mg
Protriptyline Vivactil® 5- and IO-mg tablets 10-20 mg 20-60 mg
Trimipramine Surmontilill> 25- and 50-mg capsules 25-75 mg 75-300 mg
Amoxapine Asendinill> 30-, 100-, and 150-mg tablets 50-ISO mg 150-600 mg
Maprotiline Ludiomilill> and 25- and 50-mg tablets 25-75 mg 75-225 mg
Trazodone Desyrelill> and others 25-, 50-, and 1000mg tablets 50-100 mg 150-600 mg
Auoxetine Prozac® 10- and 20-mg capsules; liquid 20mg 20mg
Bupropion Wellbutrin® 75- and loo-mg tablets 100 mg bj.d. 450 mg
or 75 mg
Sertraline Zoloft® 50- and loo-mg tablets 25-50 mg 50-250 mg
Paroxetine Paxil® 20- and 30-mg tablets 20mg 50 mg
Auvoxamine Luvox® 50- and loo-mg tablets 50 mg 100-300 mg
Nefazodone Serzoneill> 100-, 150-, 200-, and 250-mg 100 mg bj.d. 200-600 mg
Venlafaxine Effexor® 25-, 37.5-, 50-, 75-, and 100- 75 mg 75-225 mg
mg tablets

2. Mechanism of Action

Monoamine oxidase inhibitors inhibit the various subtypes of MAO throughout

the body (gut, liver, brain, platelets, and blood vessels). They elevate body levels of
epinephrine, norepinephrine, 5-HT, and dopamine by irreversibly blocking the degra-
dation of these substances. Investigators hypothesize that the increased availability of
CNS norepinephrine and serotonin may account for the antidepressant activity of
Recently, two types of monoamine oxidase have been identified (types A and B).
Type B (accounting for about 80% of CNS MAO) degrades mostly dopamine and


Figure 4. Monoamine oxidase inhibitor antidepressants.

phenylalanine, whereas type A (accounting for 20% of CNS MAO but most of the
gastrointestinal MAO) primarily degrades serotonin and norepinephrine. Although the
antidepressant MAOIs that are currently available affect both enzymes, experimental
agents (such as clorgyline) that selectively inhibit type A MAO (MAO-A) seem to be
more effective in treating depression than selective inhibitors of type B MAO [such as
selegiline (Eldepryl®) or some available drugs used for other purposes, including
pargyline (Eutonyl®), procarbazine (Matulane®), or furazolidone (Furoxone®)]. Se-
legiline is available by prescription for the treatment of Parkinson's disease as an
adjunct to levodopa/carbidopa therapy. At its recommended dose, 10 mg/day, it does
not appear liable to the food, beverage, and medication interactions of the nonselective
MAOIs. If it possesses antidepressant activity at all (an uncertain assumption), how-
ever, it is probably at a daily dose higher than that recommended for the treatment of
Parkinson's disease, at which its specificity for MAO-B is lost.
Several reversible inhibitors of MAO-A (RIMAs) are currently in use abroad, but
one such agent, brofaromine, has been removed from testing by its manufacturer, and
the chances of another one, moclobemide, coming to the United States is doubtful at
this point. Ostensibly, RIMAs should provide less risk of the "cheese effect" than
nonselective MAOIs and greater tolerance. The literature on controlled studies with
moclobemide, however, shows a relatively high incidence of excitatory effects such as
insomnia/restlessness, manialike symptoms, irritability/excitation, agitation, and anxi-
ety. There may also be the risk of the "serotonin syndrome," a potentially lethal
hypermetabolic reaction, occurring with RIMAs. A RIMA should elevate brain levels
of the catecholamines norepinephrine and dopamine, which could explain stimulantlike
properties. It should also raise brain serotonin levels, and the serotonin syndrome
sometimes includes symptoms of CNS hyperexcitability. In addition to reporting ad-
verse effects with moclobemide, investigators have also questioned whether its efficacy
equals that of older antidepressants. The role of RIMAs in the treatment of depression
remains in question.

3. Adverse Reactions

The MAOI antidepressants produce a wide range of distinctive adverse effects.

The hypertensive crisis that results from their interaction with cheese, other foodstuffs,
and a number of drugs has been the most feared. However, it is the less dramatic side
effects, such as orthostatic hypotension, weight gain, sexual dysfunction, sleep distur-
bances, and edema, that most often lead to intolerance and drug discontinuation.

a. Hypertensive Crisis. Dramatic and life-threatening hypertensive crises in pa-

tients taking MAOI drugs generated a fear of these agents among physicians and their
patients. In some countries, they were taken off the market entirely. In the United
States, MAOIs were eclipsed by the tricyclic antidepressants. Hypertensive crises are
well described in patients on MAOI therapy who ingest either pressor substances (in
food or beverages) or pressor drugs. Spontaneous hypertension has been described in
patients taking tranylcypromine without evidence of food or drug interaction. 98
Better understanding of the factors that can precipitate a hypertensive crisis in an
MAOI-treated patient has allowed safer administration of these drugs. Usually, a sharp
rise in blood pressure follows the ingestion of an exogenous pressor substance. In the
face of MAO inhibition in the gut, blood vessels, and liver, the exogenous pressor is
free to gain entry to the systemic circulation. At the same time, peripheral cate-
cholamine stores are fuller than usual and, therefore, available for sudden release,
which can lead to hypertension. How high a patient's blood pressure will go and
whether or not a stroke may result are functions of several variables: the amount of
pressor ingested, the distensibility of the patient's vascular tree, and the vulnerability of
the patient's cerebral vasculature to a sharp rise in pressure. One patient may be more
vulnerable than another, and a given patient can "cheat" on an MAOI diet one time
with impunity yet suffer catastrophic consequences on a subsequent occasion.
Table 9 details our current knowledge about foods most likely to contain high
levels of pressor substances. The most common of these is tyramine, a degradation
product of the amino acid tyrosine. Because a majority of the hypertensive crises
precipitated by such food-medication interactions have been attributed to the ingestion
of cheeses, this is often termed the "cheese effect." Many pressors, however, lurk in
soups, sauces, gravies, and dressings. It is prudent to warn patients to avoid foods
prepared by others as they will be unable to ensure the freshness and contents of what
they consume. In general, the fresher and less processed a food, the less likely it is to
contain pressors, which usually are degradation products of protein and increase with
the age of the food.
Various drugs (outlined in Table 10) can cause hypertension in an MAOI-treated
patient. These include prescription and over-the-counter medications as well as drugs
of abuse. Some patients have developed hypertensive crises when switched abruptly
from one MAOI to another: most notably from phenelzine to tranylcypromine. There-
fore, a waiting period of 1-2 weeks is recommended.
Although most cases of hypertensive crises in MAOI-treated patients are attribut-
able to the ingestion of exogenous pressors, as detailed above, occasional patients have
developed spontaneous hypertensive episodes shortly after a dose of phenelzine or

TABLE 9. Food and Beverage Products that Can Cause Hypertension with MAOIsa

1. High tyramine content: boursault, Camembert, cheddar, Gruyere, Stilton. For all fermented cheeses,
tyramine content may be highest near rind and fermentation holes.
2. Moderate tyramine content: Gouda, Parmesan
3. Low tyramine content: American, ricotta, cottage cheese, cream cheese
Other foods
I. High tyramine content: aged and processed meats, sausages, lox, pickled herring (but a recent study
found no tyramine in pickled herring excepting brine), bean curd, possibly any overripe produce
2. Moderate tyramine content: salted herring, figs, raisins, broad beans (fava beans), concentrated yeast
extracts (but not Brewer's tablets or flakes), pickles, sauerkraut, coffee, chocolate, cocoa, soy sauce,
sour cream, snails, avocado, banana peels, licorice
3. Questionable tyramine content: liver (fresh liver probably safe), protein extracts (often in soups),
oriental foods. Beware of any gravy, soup, sauce, or dressing of unknown content.
I. High tyramine content: some imported beers and ales
2. Moderate tyramine content: most sherry and beer
3. Low tyramine content: champagne, most Italian red wines, Riesling, Sauterne, "hard" liquor (but
reactions have occurred with whiskey; cause unknown)
4. Questionable tyramine content: Chianti, liqueurs, nonalcoholic beers
1. Questionable effects: cyclamates, monosodium glutamate

aGeneral rule of thumb: Try to eat only fresh foods; avoid overripe or prepared foods; and favor foods that have been
prepared simply.

tranylcypromine. 99 The proximity to dosing in reported cases suggests a direct phar-

macological effect, perhaps resulting from a sudden release of catecholamines.
A severe, sudden headache typically heralds a hypertensive crisis (although pa-
tients taking MAOIs sometimes experience headaches unrelated to blood pressure
rises). Accompanying signs and symptoms include flushing, palpitations, retro-orbital
pain, nausea and vomiting, and photophobia. If intracerebral bleeding occurs, the
patient may collapse suddenly. Traditional treatments for MAOI-associated hyperten-
sive crises have involved a.-blocking agents such as phentolamine (Regitine®) or
chlorpromazine. [3-Blockers also have been used.
As noted earlier, when prescribed at the daily dose recommended for the treatment
for Parkinson's disease (10 mg/day), selegiline does not appear to be liable to cause a
hypertensive crisis in association with the ingestion of pressors and foods, beverages,
or medications. However, as also noted earlier, at this dose it may not be an effective
antidepressant, and at higher doses, it becomes a nonselective MAOI, which appears
similar to phenelzine and tranylcypromine in its liability to cause hypertensive crises in
association with foods, beverages, and other medications. RIMAs also are relatively
free from the food and dietary restrictions required with irreversible MAOIs.

b. Hypermetabolic Crisis. A different type of profound medical crisis can occur

in MAOI-treated patients who ingest certain other substances that may elevate brain
levels of serotonin (Table 10). This syndrome has been frequently observed with the
older, nonselective, irreversible MAOIs phenelzine and tranylcypromine. It may also

TABLE 10. Drugs that Can Cause Adverse Interactions with MAOIsa

Drugs that may cause hypertension with MAOls

Amphetamine (Benzedrine® and others)
o-Amphetamine (Dexedrine® and others)
Methylamphetamine (Desoxyn®)
Methylphenidate (Ritalin® and others)
Cocaine and possibly other illegal drugs
Procaine preparations (Novocain® and others, which often contain epinephrine)
Ephedrine (Tedral® and others)
Epinephrine (Adrenalin® and others)
Over-the-counter preparations that contain phenylpropanolamine, including
Dimetane-DC Cough
Vicks Formula 44D Decongestant Cough Mixture®
Alka-Seltzer Plus Cold Medicine® (Alka-Seltzer® acceptable)
Cheracol Plus® (Cheracol D Cough Formula® acceptable)
Over-the-counter preparations that contain pseudoephedrine, including
CoTylenol® (Tylenol® acceptable)
Dimetapp Sinus Caplets®
Vicks Formula 44M®
Vicks Formula 44D®
Vicks Nyquil®
Robitussin-PE® (Robitussin® acceptable)
Tylenol Maximum-Strength Sinus Medication® (Tylenol® acceptable)
Over-the-counter preparations that contain phenylephrine, including
Dimetane Decongestant®
Dristan Advanced Formula® (tablets and coated caplets)
Vicks Sinex®
Robitussin Night Relief®
Another MAOI
Drugs that can cause other severe adverse reactions with MAOls
A number of drugs cause a syndrome that can include such features as hyperpyrexia, hyperreflexia,
muscle rigidity, seizures, hypotension, and death. Meperidine (Demerol® and others) is particularly
dangerous. Some cyclic antidepressants [most notably clomipramine (Anafranil®) and possibly
fluoxetine (Prozac®)] have caused death. It is less clear but possible that the syndrome occurs with the

TABLE 10. (Continued)

over-the-counter antitussive dextromethorphan (contained in many combination products) and

Surgical precautions
I. If possible, discontinue MAOI 2 weeks prior to elective surgery.
2. When a MAOI is administered, warn the anesthesiologist about possible complications regarding the
use of pressor agents, although many MAOI-treated patients have received general anesthesia
without complications.

aFrom Harrison et al l14 with pennission. Copyright 1989, Physicians Postgraduate Press.

occur with a relatively selective MAOI such as selegiline and with RIMAs. Clinical
manifestations include profound hyperthermia, neuromuscular irritability, altered con-
sciousness (through coma), seizures, and possibly death. Offending agents have in-
cluded meperidine (Demerol® and others), SSRIs, and some tricyclic antidepressants,
most notably clomipramine. A nonselective SRI such as venlafaxine has to be treated
as if it too can cause this reaction. Because of its very long half-life, fluoxetine's
manufacturer recommends an interval of 5 weeks between the discontinuation of
fluoxetine and initial treatment with an MAO I. For sertraline and paroxetine, the
manufacturers recommend discontinuing the SSRI for 2 weeks before initiating an
MAOI. Milder syndromes have been associated with the antitussive dextromethorphan
(contained in many over-the-counter "cold" preparations) and tryptophan. The anti-
spasticity agent dantrolene (Dantrium®), useful in malignant hyperthermia of anesthe-
sia and possibly in the neuroleptic malignant syndrome, may be considered an antidote.

c. Postural Hypotension. Symptomatic orthostatic hypotension may be more

troublesome to patients taking MAOIs than to those receiving tricyclic antidepressants.
Contraactive strategies can include those mentioned earlier for hypotension associated
with heterocyclic antidepressants. Additional strategies include the use of triiodothyro-
nine, methylphenidate (Ritalin® and others), D-amphetamine (Dexedrine® and others),
or salt tablets.

d. Cardiac Effects. The MAOIs do not have direct cardiac effects or slow cardi-
ac conduction. Thus, they may be considered alternatives to tricyclic antidepressants
for patients with bundle branch block or second-degree atrioventricular (A-V) block.

e. Sexual Dysfunction. Sexual psychology and physiology are probably impaired

more frequently by MAOIs than by tricyclics, although less frequently than by SSRIs.
Problems include decreased libido and inhibited orgasm in both sexes and impaired
erection and ejaculation in males. Proposed antidotes have included neostigmine (Pros-
tigmin® and others), bethanechol, and cyproheptadine.

f. Edema and Weight Gain. Weight gain, often associated with increased food
intake, is sometimes observed in MAOI-treated patients. This is more common with
phenelzine and isocarboxazid than with the amphetaminelike tranylcypromine. If diet

management is insufficient, the dose of the antidepressant may have to be lowered or

the patient switched to a different agent.
Bipedal edema also is a common side effect of MAOIs. A thiazide or loop diuretic
may be tried, but if this side effect is refractory, the patient may need to be changed to a
different agent.

g. Neuropsychiatric
i. Sleep Disturbances. Frequently patients discontinue MAOI agents because of the
combination of restless nocturnal sleep plus irresistible afternoon somnolence. Possibly
more common in patients with "bipolar spectrum" depressions, these side effects have
been observed with phenelzine and tranylcypromine. There is a suspicion that they may
be less common with isocarboxazid.lOo This syndrome tends to resist attempts at
management, whether with sleeping pills at night or stimulants during the day.
ii. Psychiatric. Occasional patients develop manic episodes during MAOI thera-
py. More common in patients with histories of mania, such pathological euphoria
occasionally develops de novo on provocation with an antidepressant. Some patients
treated with MAOIs also develop confusional states, sometimes while taking the drugs
and occasionally as part of a withdrawal syndrome.
iii. Other Neurological Effects. Exaggerated nocturnal myoclonus in patients tak-
ing MAOIs may respond to a change in medication schedule from bedtime to daytime
or to divided doses or occasionally to clonazepam (Klonopin®), 0.5-2 mg at bedtime.
Paresthesias, often experienced as "pins-and-needles" sensations, are sometimes sec-
ondary to pyridoxine (vitamin B6 ) deficiency induced by MAOIs and may respond to
supplementation with pyridoxine, 50-150 mg daily by mouth.

h. Hepatic. Liver dysfunction caused by antipsychotic drugs (and more rarely by

tricyclic antidepressants) typically is of the cholestatic type, with a clinical and labora-
tory profile resembling that of outflow obstruction. In contrast, hepatic injury caused
by MAOIs is characteristically hepatocellular, carrying a potential for severe liver
necrosis. The prototype MAOI, iproniazid, was withdrawn from clinical use because of
hepatotoxicity affecting about 1% of patients, with a mortality among those afflicted
that may have exceeded 20%.99 This reaction is more likely with the two hydrazines,
phenelzine and isocarboxazid, than with tranylcypromine. Although phenelzine and
isocarboxazid is are less likely than iproniazid to cause hepatotoxicity, the mortality
remains comparably high.
After 1 to 6 months of drug treatment, patients developing MAOI-induced hepato-
toxicity typically experience anorexia, weakness, malaise, and insidious jaundice, with
elevated bilirubin and transaminases 8-100 times the upper limit of normal. Patients
with markedly elevated bilirubin are at greater risk of succumbing to the reaction.
Although recovery usually is complete, in rare cases chronic disease and cirrhosis
result. The only safeguard is a clinical index of suspicion: if prodromata appear, the
MAOI should immediately be discontinued while the physician orders liver function

i. Other Adverse Reactions. Some MAOI-treated patients report feeling more

constipated, whereas others develop anorexia and symptoms of gastrointestinal irrita-
tion. Leukopenia, skin eruptions, and photosensitivity are reported rarely.

j. Withdrawal Syndrome. After abrupt MAOI withdrawal, patients have mani-

fested agitation, irritability, pressured speech, insomnia or excessive sleepiness, visual,
olfactory, or tactile hallucinations, disorientation, labile mood, paranoid delusions,
aggressiveness, slurred speech, myoclonus, hyperreflexia, ataxia, choreoathetosis, de-
lirium, and catatonia. The proposed mechanism has been adrenergic overdrive (Le.,
excessive noradrenergic activity as a result of the withdrawal). Dilsaver has pointed out
the similarity between this syndrome and that associated with the abrupt discontinua-
tion of long-term psychostimulant treatment. 101

4. Contraindications
The clinician should administer MAOIs only after a careful medical and psychi-
atric history and review of systems. In various medical situations, however, the clini-
cian should use MAOIs with caution. These include liver disease, advanced renal
disease, pheochromocytoma, cardiovascular disease, hypertension, asthma, or
chronic bronchitis (since pressor agents like epinephrine or theophylline may be
necessary). In addition, MAOIs should seldom be given to patients who take drugs that
can produce dangerous synergisms.

5. Use during Pregnancy

The safety of MAOIs during pregnancy is unsure. The practitioner should consid-
er alternatives when an antidepressant is required (see Chapter 10).

6. Toxicity
Experience with overdoses of MAOI drugs remains limited but is beginning to
expand as their use increases. Mortality has been reported following doses of phe-
nelzine, 375-1500 mg, and tranylcypromine, 170-650 mg. Because clinical signs may
not become manifest for 12 hr or more after ingestion, patients should be intensively
monitored for an extended period in a medical emergency setting.102
Headache, dizziness, and precordial pain may be early symptoms. Signs of hyper-
metabolism predominate and can include hyperthermia, increased heart and respiratory
rates, muscular rigidity, metabolic acidosis, decreased plasma oxygen, and increased
carbon dioxide. Irritability of the nervous system may be manifested by restlessness,
brisk reflexes, and, at times, seizures. Hypertension may occur, but hypotension, some-
times occurring late in the course, appears to be a poorer prognostic sign.
The MAOIs may delay gut motility, so gastric lavage should be performed even
hours after the ingestion. Acidification can speed MAOI elimination and increase the
amount of MAOI eliminated in the urine. In contrast, with tricyclic antidepressants,
hemodialysis can effectively remove much of the MAOI. Similar to the treatment of
heterocyclic overdoses, activated charcoal and magnesium citrate can help reduce

In one case, IV dantrolene produced prompt and dramatic improvement in muscu-

lar rigidity, trismus, tachycardia, hyperthermia, and metabolic acidosis that resulted
from a massive phenelzine overdose.103 In another report, physicians in Australia
recommended pulmonary artery catheterization to monitor hemodynamic response and
the use of selected a and 13 blockade to treat hyperadrenergic signs. 104 These authors
also recommended treatment of hyperthermia with active cooling techniques and fur-
ther pointed out that prolonged observation and support may be required for the
approximately 2 weeks it takes for a patient's supply of the enzyme MAO to regenerate
after discontinuation of the MAOI.
Overdoses with MAOI antidepressants tend to be very serious. Suggestions that a
patient may have taken an overdose of an MAOI drug should prompt an immediate and
urgent medical response.

7. Pharmacokinetics
Monoamine oxidase inhibitors are rapidly and fully absorbed from the gastroin-
testinal tract. They undergo biotransformation in the liver and are excreted very quickly
through the intestinal tract and, to a lesser extent, via the kidneys. The time to peak
plasma concentration is 2-4 hr after an oral dose for phenelzine and 1-3.5 hr for
tranylcypromine. The onset of action may be as early as 7-10 days with appropriate
dosage in some patients, but it may take up to 4-8 weeks to achieve full therapeutic
effect. It takes at least 10 days for MAO activity to be recovered due to irreversible
binding of MAOIs. The half-life of MAOIs in the body is very short. However, the
traditional MAOIs have long-lasting pharmacologic effects, since they permanently
inactivate enzymes. Therefore, MAO inhibition continues for up to several weeks after
discontinuation of the drug. The body must resynthesize the enzymes before normal
metabolism of body amines resumes, a process taking 1-2 weeks.
Little is know about the metabolic fate of the three MAOI antidepressants, either
in animals or in humans. Some research has focused on acetylation of hydrazines as a
possible predictor of clinical response. At least for phenelzine, the primary degrading
process appears to be oxidation rather than acetylation. Furthermore, the genetic poly-
morphism of the liver enzyme system that determines whether individuals are rapid or
slow acetylators appears to have no clinical usefulness in predicting treatment response
or side effects. Although patients with a degree of platelet MAO inhibition greater than
80% by 2 weeks of therapy are more likely to have a favorable treatment outcome with
phenelzine, ensuring adequate dosage is probably equally valuable and certainly less
expensive. There is no evident correlation between platelet MAO inhibition and clini-
cal response with either tranylcypromine or isocarboxazid.

8. Preparations and Dosage

Each MAOI is produced in a single preparation (see Table 11). Usual doses vary
widely among patients but generally are about 1 mg/kg of body weight per day for
phenelzine but less for isocarboxazid and tranylcypromine.

TABLE 11. Preparations and Dosages of Monoamine Oxidase Inhibitors

Usual dosage range

Generic name Trade name Preparation (mg/day)

Isocarboxazida Marplan® 10-mg tablets 10-30

Phenelzine Nardil® 15-mg tablets 45-90
Tranylcypromine Parnate® lO-mg tablets 10-30

alsocarboxazid is only available in the United States through special request from the manufacturer.

c. Stimulants
Psycho stimulants, such as the amphetamines, were used before the tricyclics and
MAOIs to treat depressed patients. Often, they did little more than convert psychomo-
tor retardation to anxiety or agitation. Additionally, tolerance developed to their effects,
anorexia and insomnia were unwanted reactions, and these drugs could promote depen-
dency and abuse. Furthermore, there was little scientific validation of their antidepres-
sant efficacy, and for this reason, coupled with increasing government regulation over
the prescription of these substances, they fell into disfavor and disuse among the
medical community.
More recently, however, there has been heightened interest in the use of stimu-
lants, growing largely out of the experience of consultation-liaison psychiatrists. Be-
cause of their very different profile of clinical response and adverse reactions, these
drugs have been found safer and more useful than standard antidepressants, particularly
in the medically ill and the elderly. Low doses of methylphenidate or D-amphetamine,
administered for up to several weeks, can be helpful in turning the ~orner in depressive
symptoms in some such patients. Despite the paucity of scientific validation of these
clinical impressions, many psychiatrists practicing in general hospitals are pleased to
have them available.
Earlier research suggesting that stimulant response could be used to predict anti-
depressant response has not withstood the tests of science and time. In some instances,
clinicians have used stimulants to get a "leg up" on the lag time while waiting for a
standard antidepressant to work or to "jump start" a treatment-resistant patient. Anoth-
er use of stimulants as adjuncts to antidepressant medications is in patients suffering
from drug-induced postural hypotension.

D. Evaluation of the Depressed Patient

The depressed patient can present with many different clinical symptoms and
signs. Mood disturbances frequently accompany major losses and other stressful
events, particular character styles, and certain psychodynamic constellations. More
often, depressed patients come to both medical and psychiatric clinicians with physical

complaints. These may take many fonns, from a request for a physical examination,
complaints of fatigue, weight loss, or insomnia to specific manifestation of somatic
symptoms (often careful questioning reveals that several organ systems are involved).
Along with or instead of physical symptoms, the patient frequently has both emotional
(e.g., lowered mood, anhedonia, negativism) and psychological symptoms [e.g., irri-
tability, loss of interest, ruminations, poor concentration, suicidal preoccupations, and
even pseudodementia (see Chapter 8)]. In addition to defining the specific nature and
history of the mood disorder, the clinician should also evaluate the other important
components of the patient's life that might affect treatment: his or her environmental
supports and important relationships, family history, coping style, and ego strengths.

In Each Depressed Patient, the Practitioner Should:

1. Look for life-threatening symptoms (i.e., poor self-care, suicidal and homici-
dal impulses).
2. Take appropriate measures to address suicidal or homicidal impulses.
3. Assess the patient's diet.
4. Define medication-responsive symptom clusters.
5. Obtain a history of the present illness, focusing on possible situational,
constitutional, and biological contributions to the mood disturbance.
6. Inquire about a history of previous mood episodes, their course, treatment,
and response(s) (including adverse reactions).
7. Inquire about a family history of mood episodes, their course, treatment, and
8. Define the patient's strengths, including coping mechanisms and environ-
mental supports.
9. Determine the patient's attitude toward help and medication.
10. Take a complete medical history (including a review of organ systems and
documentation of both prescribed and nonprescribed drugs; see Table 12).
11. Complete or arrange a physical examination (and perhaps other baseline
studies such as an EKG when clinically indicated).
12. Avoid the nonselective use of extensive medical screening and biochemical
tests for major depressive illness.
13. Get appropriate consultation if psychological or medical symptoms are not
clarified by the evaluation.
14. Infonn the patient (and perhaps persons close to him or her) about the
findings of the evaluation, the recommended treatment(s), benefits, risks,
and possible alternatives.
15. Document conversations and recommendations.
16. Establish rapport with the patient and enlist him or her (and often family
members) in the treatment.

After completing a careful evaluation, the clinician should decide how to manage
acute symptoms, possible precipitants, and sequelae of the affective episode (e.g.,
stigmatization, lowered self-esteem, alienation from social supports).

TABLE 12. Drugs Associated with Depressiona

Antihypertensives Anti-infectious agents (cant.)

Guanethidine (Ismelin® and others) Metronidazole (FJagyl®)
Methyldopa (Aldomet® and others) Antineoplastic agents
Reserpine (Serpasil® and others) Plicamycin (Mithracin®)
Hydralazine (Apresoline® and others) Azathioprine (Imuran®)
Propranolol (lnderal® and others) Bleomycin (Blenoxane®)
Sedative-hypnotic agents L-Asparaginase (Elspar®)
Alcohol Cardiac drugs
Chloral hydrate (Noctec® and others) Digitalis (Crystodigin® and others)
Benzodiazepines Procainamide (Pronestyl® and others)
Steroids Propranolol (lnderal® and others)
Oral contraceptives Clonidine (Catapres®, Combipres®)
Cortisol (Solu-Cortef® and others) Stimulants
ACTH (Acthar®) Amphetamine (Benzedrine® and others)
Antipsychotic medications Fenfluramine (Pondimin®)
Analgesics Other drugs
Opiates L-Dopa (Larodopa® and others)
Anti-inflammatory agents Amantadine (Symmetrel® and others)
Phenacetin (Emprazil® and others) Methysergide (Sansert®)
Phenylbutazine (Butazolidin® and others) Acetazolamide (Diamox® and others)
Pentazocine (Talwin®) Carbamazepine (Tegretol® and others)
Estrogen withdrawal (Premarin® and others) Choline (Trilisate® and others)
Anti-infectious agents Disulfiram (Antabuse®)
Sulfonamindes Physostigmine (Antilirium®)
Clotrimazole (Lotrimin®, Mycelex®) Ethambutol (Myambutol®)
Ethionamide (Trecator-SC®) Indomethacin (Indocin® and others)
Griseofulvin (FJuvicin® and others)

a Association does not mean causality. Some agents (e.g., reserpine and steroids) have a clear relationship with depression,
whereas many other drugs do not.

E. Drug Therapy for the Depressed Patient

1. Initiating Treatment
After reaching a diagnostic decision and assessing the physical status of the
depressed patient, how does the practitioner choose a drug for an individual? If the
patient or a close relative has previously responded to a particular medication, use that
drug (ifit can be given safely). Conversely, poor tolerance or lack of adequate response
to a previous trial suggests the wisdom of avoiding that particular agent.
Since the last edition of The Practitioner s Guide, clinical practice patterns in
antidepressant prescribing have changed considerably, owing to the nature and number
of new antidepressants available. As noted earlier, the newer agents have distinctly
different side-effect profiles and in general are better tolerated. As a group, they are less
hazardous in overdose. Therefore, for a patient with mild to moderately severe depres-
sion, without delusions or other psychotic symptoms, without a prior history of favor-
able or unfavorable response to an antidepressant, and without medical complications,

many practitioners now choose a "second generation" antidepressant as a first-line

choice. The three antidepressant SSRIs are typically tried first. A second SSRI may be
used if the first fails. Venlafaxine and nefazodone are usually tried later. The proper
place of mirtazapine in a clinical algorithm has not been established as of this writing.
Because of its increased seizure risk, most clinicians reserve bupropion for subsequent
trials. Similarly, MAOI antidepressants, with their increased risks and complicated diet
and medication prohibitions, are also reserved for later trials. Tricyclics may still be
considered for first-line treatment, particularly in more severe depression, but because
they tend to be less well tolerated and are associated with more profound toxicity in
overdose, they too have been falling to the category of second-line agents. However, as
noted at the beginning of this chapter, there is currently some concern that SSRIs (and
perhaps other newer antidepressants) may not possess the same "staying power" as
tricyclics in long-term treatment. If this concern is validated by subsequent clinical and
research evidence, the pendulum may swing back in favor of tricyclics as drugs of first
choice in the treatment of depression.
If the clinician selects one of the newer antidepressants, there is no compelling
evidence for the preference of one over the other. Instead, practitioners typically
become familiar with one drug in a category, knowing its "quirks" better than others
and developing an intimate familiarity with its clinical profile. Some tailoring can be
made for individual patients based on different adverse-effect profiles and phar-
macokinetics among different agents.
Among tricyclics, too, the choice of agent should follow the guidelines above. In
recent years, clinicians have come to favor the secondary-amine tricyclics nortriptyline
and desipramine over the tertiary amines owing to their more favorable side-effect
As noted earlier, dosing patterns vary considerably among different agents and
especially among different categories of antidepressant drugs. This is one reason to
favor physicians becoming familiar with one drug per category. In addition, patients
tend to vary widely in their responses to the same antidepressant: whatever the "usual
adult dose" of a drug, there will likely be patients who will need less and others who
will benefit from more than the standard range. Awareness and sensitivity to side
effects and close contact with the patient and "significant others" can allow individual-
ization of dosage regimens.
Some drugs, such as bupropion, venlafaxine, and nefazodone, are supposed to be
administered in multiple daily doses. Sustained-release preparations are being studied,
however, and nefazodone probably can be used in once-daily bedtime dosing. Tri-
cyclics and mirtazapine can be administered once daily, which makes it easier for
patients (or caregivers) to comply with the prescribed regimen. Once-daily dosing,
particularly if it can be achieved at bedtime. often can allow greater tolerance to
adverse effects, which can occur during sleep.
For patients who have failed to show optimal response with an antidepressant drug
at the top of the dosage range recommended by the manufacturer, yet who have not
manifested limiting side effects, clinicians may sometimes elect to prescribe dosages
above that range. This often is a good time to obtain a consultation from a psychiatrist
specializing in pharmacotherapy. Obtaining plasma levels of the antidepressant and

other laboratory tests, such as an EKG, also may be worthwhile in such circumstances.
When a patient has been partially responsive to standard doses of an antidepressant, an
alternate strategy to raising the antidepressant dose or switching antidepressants is to
use an adjunctive drug, such as lithium or triiodothyronine (T 3). When the antidepres-
sant is an SSRI, some clinicians favor the addition of low doses of a tricyclic (see also
Section V.E.I0).
As noted earlier, overdoses with tricyclic and MAOI antidepressants produce a
substantial risk of lethality. Prescribing these drugs, therefore, to a patient with the risk
of suicide demands careful thought and discussion with the patient and friends or
family members. If the risk of suicide is considered substantial, the practitioner will
want to hospitalize the patient or arrange for other circumstances in which the patient
can be monitored in an ongoing fashion for the patient's protection. If the risk of
suicide is not considered to be at that level, and a potentially hazardous agent such as a
tricyclic or MAOI is to be prescribed, the patient and concerned others must be advised
about the lethal potential of the drugs (to avoid an unintended "gesture" that inadver-
tently becomes lethal). Consider limiting the patient's access to potentially lethal
quantities of the drug (for tricyclics, as little as one gram). Along these lines, when
children may have access to these hazardous drugs, the drugs should be kept in safe
places and in childproof containers.

2. Completing a Medication Trial

Although many antidepressants have been introduced with the promise of faster
onset of action, these claims have not withstood the test of time. It seems that all
antidepressant drugs currently known-tricyclics, MAOIs, and newer agents-take
some weeks to become maximally effective. The conventional wisdom is that the lag
time is 2-3 weeks, but this is oversimplified.
Prospective studies have found differences in symptom responses between drug
responders and nonresponders as early as the first week of treatment. In other words,
patients who will go on to become responders show the beginnings of improvement in
a wide range of depressive symptoms by the seventh day of antidepressant drug
therapy. Typically, patients may be the last to know (or to acknowledge) that they are
better; a significant other, or the clinician, may be more perceptive. Sleep, by the way,
is not a useful symptom for distinguishing responders from nonresponders; the non-
specific sedating effects of these drugs can improve sleep (or cause excessive drowsi-
ness) independent of treatment response.
Some patients show an early and dramatic improvement in depressive symp-
toms-perhaps a complete remission within the first week. Others show a variable and
stuttering course of symptom improvement. Both patterns may be more likely to be
placebo responses than a true biological antidepressant reaction. The caveat here is that
these patients are more likely to lose their apparent benefit from the antidepressant.
Anticipating this, the clinician can be prepared to counteract a tendency to frustration,
despair, or anger by suggesting the need to reevaluate the drug and its dose and also to
emphasize the role of interpersonal and intrapsychic factors.
While awaiting treatment response, a knowledgeable clinician should be in fre-

quent contact with the patient, and possibly family members, to provide support and
reassurance and to assess suicide potential. Family and patients also should be advised
from the outset about the lag time before the drugs become effective and should be
informed about common side effects. Education about unwanted effects of these drugs
can tum an annoyance into "cement" for the therapeutic alliance: the patient can treat
dry mouth, for example, as evidence that the drug is in his or her system and working; it
also affirms the knowledge of the doctor.
The maximal efficacy of an antidepressant may not plateau for 6 to 8 weeks. Thus,
if a patient's improvement has progressed from 80% to 90% between weeks 4 and 5 of
treatment, the clinician can stay with the same dose for at least another week, awaiting
continued improvement. On the other hand, a plateau in improvement at anything less
than 100%, or a return to status quo ante, should prompt a reassessment of the drug
The goal of antidepressant drug therapy should be complete remission of symp-
toms. Needless to say, character problems will not usually improve, but sometimes
symptoms attributed to a patient's personality may reflect untreated depression. Still,
for patients with dysthymia who later develop major depression (so-called "double
depression"), antidepressant drugs may reverse the major depression but return the
patient to a stable depressed character type.
The clinician would do best to continue adjusting the dose of the original anti-
depressant drug until either the acute depression remits completely or the patient is
unable to tolerate side effects. If side effects are the limiting variable, and dosage
reduction brings loss of efficacy, then the physician must switch to an alternate anti-
depressant. In this circumstance, one chooses a drug with fewer of the offending side
effects, for example, one that is less anticholinergic, less hypotensive, less likely to
promote weight gain, or less likely to impair sexual functioning.
Failure of an antidepressant to treat symptoms effectively despite adequate dosing
for a sufficient time period is more troublesome. There is no evidence that switching
among tricyclics is apt to produce efficacy when one tricyclic has failed, although "the
jury is still out" about differences among SSRIs. In some patients who have achieved
partial but incomplete remission, lithium, often starting with as little as 300 mg a day,
may provide a needed "jump start." For a patient partially responsive to an SSRI, the
addition of a low-dose tricyclic, such as desipramine, may close the therapeutic gap.
Perhaps this will not be necessary with combined serotonin and norepinephrine reup-
take inhibitors such as venlafaxine. Sometimes two, three, or even more drugs are
being combined in increasingly creative ways by clinicians specializing in hard-to-treat
When one agent fails to induce remission, switching to an antidepressant with
substantially different pharmacologic characteristics (e.g., tricyclics, MAOIs, SSRIs,
bupropion, venlafaxine, nefazodone) makes rational sense. Before starting an MAOI
after discontinuing a different antidepressant, the patient should be free of the prior
antidepressant for at least a week after a tricyclic, 2 weeks after another MAOI or an
SRI, and 5 weeks to 2 months (depending on duration of treatment) after fluoxetine.
Electroconvulsive therapy always remains an important backup option.
When a patient has proved refractory to several adequate trials of biological

antidepressant treatments, a diagnostic reassessment is in order. Covert medical and

neurological factors, substance abuse, character and intrapsychic elements, and inter-
personal contributions all must be weighted in the balance. When continuing anti-
depressant drug treatment appears indicated, one of a number of combinations may be
considered (see Section V.E.1O).

3. Continuation Therapy
Having successfully ameliorated the acute symptoms of depression, the practi-
tioner should now embark on what is termed continuation therapy. Typically lasting
about 6 months, continuation therapy protects against a relapse of the depression.
Presumably, we are "holding the fort" against depressive symptoms while the underly-
ing process runs its course.
For continuation treatment, the clinician should usually continue the same dose
that was effective during the acute phase of therapy. If, however, the patient is plagued
by annoying but persistent side effects, the physician may try to lower the dose very
slowly, although both patient and doctor should be acutely aware that this enhances the
risk of relapse. At times, a switch to a different agent might be considered, but again the
benefits in comfort must be weighted against the risk of recrudescent symptoms.
After about 6 months of successful antidepressant therapy, the physician evaluates
the patient for possible discontinuation of the drug. If depressive symptoms have been
successfully and completely suppressed for at least 4 months, a gradual taper may
be attempted. If, however, there has been a "flickering" of symptoms (whether during a
missed dose, a time of stress, or for no apparent reason), it is most prudent to continue
the continuation therapy for at least another few months. Tapering the medication after
an acute episode of depression should not take place until and unless depression has
been virtually absent for at least 4 months.
To avoid antidepressant withdrawal symptoms (see Section V.A.3.g.vii), tapering
should be very gradual. Withdrawal symptoms after only 6 months of therapy are not
likely to be as intense as those following many years of treatment. Still, if there is no
medical urgency about discontinuing the antidepressant, prolong the taper over a month
or more. Toward the end of the taper, tablets can be broken in half or administered
every other day to make the final phases even more gradual.

4. Maintenance Therapy for Chronic or Recurrent Depressions

The course of bipolar illness is, by its very nature, recurrent. Among patients with
unipolar depression (i.e., major depression with no history of mania), a recurrent course
may affect between 70% and 85%. Probably worse still, 10-20% of patients with
unipolar depressions develop a chronic course.
Depression is a serious illness. It carries a high morbidity of economic losses and
disruption of families as well as other vital interpersonal linkages. Also, depression
carries a substantial risk of mortality, usually from suicide, but also from a worsening
of coexisting medical disorders and less commonly from neglect. For these reasons,
and because of the intense pain associated with depressive episodes, some patterns of

mood disorder require maintenance therapy rather than intermittent treatment with
Whether early or late in the course of a depressive illness, if the physician cannot
wean the patient from antidepressant medication at the end of continuation therapy
because of reemerging symptoms, this suggests that a chronic course is developing. In
such cases, if antidepressant medication has successfully suppressed all or most of the
depressive symptoms, there is no reason not to continue it indefinitely, possibly with
periodic attempts to taper. Obviously, the physician seeks to find the lowest effective
dose of the antidepressant, but because most patients can take these agents chronically
with impunity, there is no pressing need to compromise efficacy in the interest of
lowering the dose. (This situation is different from the concern with neuroleptics that
higher doses increase the risk of tardive dyskinesia during the course of maintenance
treatment.) In fact, recent research suggests that the most effective doses for mainte-
nance antidepressant therapy, at least with tricyclics, are the same as those effective in
the acute treatment of depression. I 05 Therefore, if at all possible, maintain the patient
on maintenance therapy at the same doses that were effective (and presumably toler-
ated) in the acute and continuation phases of treatment. Occasional patients may have
required ECT during an acute episode but will benefit from maintenance antidepressant
For a patient with multiple recurring episodes of depression, the question of when
to consider maintenance therapy will hinge on the number of episodes as well as their
frequency and severity. To be sure, the more numerous, frequent, and severe the
depressive episodes have been, the earlier and more vigorously the doctor should
pursue a course of maintenance. The need for such an approach should be discussed
actively with the patient and possibly with significant others. The benefits in terms of
prevention of relapse, with all the attendant morbidity and mortality, are obvious; the
risks are few and usually center around uncomfortable rather than hazardous side
There are no known risks that increase with long-term exposure to these drugs.
However, as a patient ages, it may be necessary to reassess the dose of the drug and its
interaction with illnesses that may have developed or with medications taken to treat
If an antidepressant is ineffective in the prophylaxis of recurrent depressive epi-
sodes, the practitioner may consider maintenance lithium therapy. This is especially
true if the patient has a history of episodes suggesting hypomania or a family bipolar
history. Most nonbipolar patients, however, will not require lithium, which probably is
less effective than antidepressants in maintenance therapy. There are no good studies
about the optimal plasma range of lithium in depression. We prefer the dose levels most
effective in bipolar patients, 0.8-1.0 mEq/L. Of course, for bipolar patients, lithium is
the maintenance therapy of choice and should be augmented as needed with antidepres-
sant drugs.

5. Depressive Episodes in Bipolar Disorders

Bipolar patients usually are taking lithium (Eskalith® and others) or other mood-
stabilizing maintenance therapy, which tends to be more effective in preventing manic

than depressive recurrences. When minor depressive symptoms emerge, time, support,
and psychotherapy may suffice. Emergence of a full-blown episode of major depres-
sion, however, can be serious and usually requires prompt and vigorous intervention.
For a major depressive episode in a bipolar patient, an antidepressant may be
added to the maintenance lithium regimen. Despite the lithium, however, there is a
finite risk of triggering a switch into mania, although in some patients this would
happen whether or not an antidepressant was prescribed. To minimize this risk, discon-
tinue the antidepressant drug soon after the patient has achieved a remission from
depression. This also lowers the possibility that prolonged antidepressant therapy can
speed up the rate of a bipolar patient's mood cycles.
There is limited evidence that the MAOI tranylcypromine is particularly effective
in treating the bipolar depressed patient, especially when depression is accompanied by
psychomotor retardation, apathy, and withdrawal.! 06 There is less experience with
some of the newer antidepressants in bipolar patients, although there have been sugges-
tions that bupropion may be beneficial for this population. Bupropion and SSRIs may
cause switches into mania less often than tricyclics.107 ECT is highly effective and
should particularly be considered in emergency cases. (Most psychiatrists discontinue
lithium during the course of ECT to avoid neurotoxicity.)

6. Some Dichotomies: Mild versus Severe, Acute versus Chronic,

Agitatedl Anxious versus Retarded

When the symptoms of major depression are present consistently over a period of
weeks, most psychiatrists will consider antidepressant drug therapy. If the patient
simultaneously has a number of melancholic symptoms, the likelihood of improvement
with an antidepressant drug (or ECT) becomes even better. These predictive factors
tend to be independent of the degree of severity or longevity of the depression. Neither
mildness nor chronicity should preclude consideration of an adequate antidepres-
sant trial, nor do they suggest use of lower doses. The only way to know if a
patient's depressive symptoms will remit on antidepressant drug therapy is to try
it. Obviously, if depressive symptoms yield to psychodynamic interpretations or
changes in living or work arrangements, antidepressant drug therapy may become
moot. However, analogously, if symptoms remit on prescription of an antidepressant
drug, major life changes or long-term psychotherapy may be unnecessary. Until a
vigorous and adequate antidepressant trial has been attempted and failed, it is
impossible to say whether a patient's symptoms are "characterologic" or ''neurot-
ic" rather than drug responsive.
An appropriate role for antidepressants in treating the patient with dysthymia
remains to be elucidated. Indeed, dysthymia most probably reflects a range of disor-
ders, including both those that are responsive to medication and those that are not.
Again, the only way to distinguish the two is empirically: most, if not all, dysthymic
patients deserve at least one vigorous antidepressant trial.
In the early days of modem psychopharmacology, there were hopes and expecta-
tions that antidepressants that were more or less sedating could be tailored to depres-
sive subtypes such as agitated, anxious, or retarded. This has not proved to be the case.
Nevertheless, some patients may appreciate the sedation provided by drugs such as

amitriptyline, doxepin, or trazodone-usually prescribed at bedtime-whereas others

will prefer the more stimulating effects of fluoxetine or tranylcypromine. Sometimes
the use of an adjunctive benzodiazepine for the first few weeks of antidepressant
therapy can enhance sleep or help a patient relax during the day. It is unclear whether
the triazolobenzodiazepine alprazolam has intrinsic antidepressant properties in addi-
tion to its proven anxiolytic activity, but some patients with milder depressions might
be tried on it. The clinician must remember, however, that abuse and dependency have
been reported as problems with benzodiazepines, and particularly with alprazolam.
Patients with borderline personality disorder often experience episodes of major
depression. Biologically oriented psychiatrists then pursue treatment of the depression
vigorously, only to be surprised and perplexed by the patient's apparent refractoriness.
Although one or even several antidepressant trials may be warranted in such patients,
there is evidence that, by and large, they are less responsive to typical antidepressant
therapy (see Chapter 13). For borderline patients, other medications may be in order,
along with long-term supportive psychotherapy.

7. "Atypical" Depression

The term "atypical" is used variably in clinical psychopharmacology. Clinicians

usually employ this concept to refer to major depression with a cluster opposite to that
of melancholia. Thus, patients with atypical depression experience more depression
early in the day, have greater difficulty falling asleep, and may experience food (often
carbohydrate) craving with accompanying weight gain. They can experience "leaden
paralysis," in which their limbs feel excruciatingly heavy and hard to move. Psycho-
logically, they are often rejection-sensitive, anxious, and phobic and have a more
"reactive" (rather than autonomous) mood. Investigators at Columbia University have
found that atypical patients (per their definition; see Table 13) may do better on MAOI
than tricyclic therapy.35
Nonetheless, it is still easier to start such patients on an SSRI or other antidepres-
sant because of the greater ease of the regimen. An MAOI should certainly be consid-
ered in patients who do not respond to or who cannot tolerate other antidepressants.

TABLE 13. Atypical Depressiona

Mood reactive
Has exhibited at least two of the following:
I. Increased appetite or weight gain while depressed (defined as present if appetite markedly increased
or at least a 4.5-kg weight gain)
2. Oversleeping (at least 10 hr/day)
3. Severe fatigue that creates a sensation of leaden paralysis or extreme heaviness of arms or legs
(defined as present if a marked decrease in energy or if heaviness of limbs is a prominent symptom)
4. Sensitivity to rejection as a trait throughout adulthood (defined as present if rejection results in
depression with functional impairment, e.g., missing work or school)

"Ref. 35.

How the newer antidepressants will compare in efficacy for atypical depression is
unclear, although patients for whom weight gain is a substantial problem may appreci-
ate the weight loss often associated with SSRIs. Electroconvulsive therapy is not
usually thought of in this condition.

8. Major Depressive Episode with Psychotic Features

Major depression accompanied by delusions, hallucinations, or other failures in
reality testing is a serious and often life-threatening condition. In a majority of cases, it
is unresponsive to tricyclic or other antidepressants alone. Electroconvulsive therapy
produces substantial clinical improvement in over 80% of cases. A combined regimen
of a tricyclic and an antipsychotic agent (doses adjusted independently and clinically)
tends to have a success rate not much lower than that of ECT.108 Amoxapine may also
be efficacious in this condition. 109 The combination of fluoxetine and an antipsychotic
agent has been reported to be as efficacious as the tricyclic/antipsychotic combination,
with fewer side effects. I 10 There are few data on continuation and maintenance therapy
for this condition, but it tends to be recurrent.

9. Depression in Schizophrenic Patients

Schizophrenic patients sometimes develop depressive symptoms and syndromes.
These require a careful psychiatric differential diagnosis.
What appears (and even feels to the patient) like depression can at times be a
reflection of neuroleptic-induced bradykinesia. The motor slowing of this parkinsonian
symptom may be interpreted by the patient as low motivation and low spirits. To assess
this possibility empirically, the practitioner can prescribe an antiparkinson drug for a
few days or a week and perhaps lower the dose of the patient's antipsychotic agent.
From a psychosocial perspective, the clinician also must assess the apparently
depressed schizophrenic patient regarding such issues as autistic withdrawal or demor-
alization. Most commonly, these reactions are apt to be responses to the patient's
interpersonal milieu, and sensitive inquiry might reveal emotional issues with which
the patient is grappling. Such problems are best handled by counseling, emotional
support, and structured psychotherapy. We believe that schizophrenic patients often
benefit from a frank acknowledgment of their cognitive and interpersonal limitations,
assistance in grieving these realities, and then support in making the most of their
assets and strengths.
For many schizophrenic patients, the chronic course of their illness is charac-
terized by interpersonal withdrawal as part of a "defect state" or negative syndrome.
Some such negative symptoms have been reported to improve more on treatment with
clozapine than with standard neuroleptics. Perhaps other new antipsychotics will also
share this benefit.
When a schizophrenic patient develops a full-blown major depressive disorder,
antidepressant drugs can sometimes be helpful, used in the same fashion prescribed for
any patient with this syndrome. There is a finite risk of worsening the underlying
psychosis, but this should be mitigated by the coadministration of a neuroleptic. I I I

10. Resistant Depressions

The key to success in antidepressant pharmacotherapy is adequacy in time and

amount: enough drug for a long-enough period. How much is enough? As noted
previously, with a positive, linear correlation between dose and efficacy, as in the case
of imipramine and possibly venlafaxine and nefazodone, the clinician should "push"
the dose until meeting the limits of patient tolerance or enjoying the fruits of success.
With nortriptyline, for which there is reasonable evidence of a curvilinear "therapeutic
window" relationship, a treatment-resistant patient can have blood drawn for a serum
nortriptyline assay, and the clinician may adjust the dose to bring the blood level within
the "therapeutic range" of 50-150 ng/mL. For all other antidepressant drugs, the best
that a blood level has to offer is a broad "intuitive" sense of whether or not a patient
may be extremely high or low. Moreover, with SSRIs and other new antidepressants, as
noted earlier, dose-response relationships are unclear, and individual titration is re-
How long is long enough? As we discussed earlier, the clinician should see some
improvement within a week or two. Often, improvement is noted by others before
being observed by the patient. Conversely, the absence of at least some improvement
should prompt a medication adjustment-usually a dosage increase. If benefits are
being accrued, the practitioner may stay with the same dose of the same drug for a few
weeks. A plateau in benefits may not occur for as long as 8 weeks at an optimal dose. If
a patient has stayed with the same dose of the same drug for more than several weeks
without accrual of additional benefits, yet is still not fully recovered, then additional
steps should be taken-dosage increase, addition of an adjunctive medication, or
change of antidepressants.
As noted earlier, partial improvement on an antidepressant often can be changed
to full improvement by the addition of an adjunctive medication (to be discussed later).
Lack of significant improvement suggests a change of antidepressant. There is no
absolute algorithm at this time for changing antidepressants. If a patient has been tried
on an SSRI and has had limiting side effects, a different SSRI can be tried. If a patient
has failed to respond, it is probably wiser to switch to an antidepressant of a different
category. Venlafaxine and nefazodone are different enough from SSRIs to be consid-
ered in SSRI nonresponders. Bupropion also is substantially different. Beyond doubt, a
patient cannot be considered treatment refractory if he or she has not also had trials of
one tricyclic antidepressant and one MAOI. Within the MAOI category, there is even
evidence that patients nonresponsive to phenelzine may respond to tranylcypromine
and vice versa. Patients who have failed to respond to four or more trials of mono-
therapy with antidepressants in the above categories should be considered for a trial of
ECT. A patient who has had adequate trials of the aforementioned and still has not
improved can be categorized as truly treatment resistant. In such cases, a diagnostic
reappraisal is in order
From a psychosocial perspective, a number of factors may complicate or antago-
nize response to antidepressant therapy. Patients with some forms of character pathol-
ogy, such as the borderline personality, tend to respond poorly to antidepressant medi-
cation; alternate medications and verbal interactions may be preferable (see Chapter

13). Individuals for whom depressive behaviors are part of their lifestyle and interper-
sonal coping strategies also may do poorly on antidepressant therapy, although they
should still be given at least several trials of adequate pharmacotherapy. A sensitive
clinician will always inquire about neurotic and character mechanisms that may be
"driving" depressive symptoms. An interview with a significant other can shed new
light on the meaning of an individual's symptoms and on the accuracy of the history.
Unresolved grief is another phenomenon that sometimes appears to be a treatment-
refractory depression.
Many biological factors can similarly render a patient unresponsive to antidepres-
sant medicines. Consider possible adverse interactions with other medications a patient
may be taking: prescribed, over-the-counter, or substances of abuse. A wide range of
medical illnesses-especially endocrinopathies, infections (including AIDS), brain
diseases, or neoplasms-can similarly produce a depression that might fail to respond
to antidepressants. To consider all of these and many other biological possibilities,
return to the medical history and review of systems, obtain additional data from past
records, other treating physicians, or family members, see that an adequate physical
examination has recently been performed, and order any laboratory testing indicated by
the foregoing.
What if nothing new turns up from either a psychosocial or a biological perspec-
tive? At this point, the physician can consider a wide variety of add-ons and medication
combinations, unfortunately few of which have been studied systematically. Probably
best studied and most frequently used today in clinical practice is adjunctive lithium.
Many knowledgeable clinicians add lithium to a first or second antidepressant trial in
an incompletely responsive patient. Some patients appear to respond dramatically and
quickly to minimal doses of lithium, whereas others respond more gradually to some-
what higher doses. Perhaps partial responders are most likely to progress to complete
response, or maybe some patients are hastened to what would ultimately be a complete
recovery anyway. Despite these doubts, the usual augmentation doses are between 300
and 1200 mg daily, maintained for several weeks. One approach is to begin with 300
mg bj.d. and, if there is no response within a week, to increase to 900 mg daily. Some
clinicians prefer to bring patients to a "standard" blood lithium level approaching 1.0
mEq/L. Most physicians try to stop the dose of lithium after several weeks, whether it
is effective (in which case perhaps no longer needed) or ineffective (in which case
something else should be tried).
Adjunctive thyroid hormones have been tried, although our reading of the data
suggests that, in patients chemically euthyroid, this approach is less likely to meet with
success. The usual strategy is to initiate treatment with 10-25 j.1g of liothyronine (T3 ,
Cytomel®), increasing the dose every few days to a maximum of 50-75 j.1g/day. Some
recent studies suggest that thyroxine (T4 ) may be an effective supplement in treatment-
resistant depressed patients. Other adjunctive strategies with even weaker data bases
include the coadministration of reserpine (Serpasil® and others), an antipsychotic drug,
conjugated estrogens in women, or methylphenidate with an SSRI. Tryptophan and
tyrosine, the amino acid precursors of serotonin and norepinephrine, respectively, also
have been used adjunctively with varying reports of success. For patients with seasonal
affective disorder, spending upwards of 2 hr a day in front of high-intensity full-

spectrum lights has been reported to yield dramatic results, although the scientific
validation of phototherapy as an antidepressant treatment remains insufficient. We have
already mentioned the combination of an SSRI and a tricyclic, and some clinicians are
combining bupropion with other agents.
There is no scientific evidence that the combination of a tricyclic and an MAOI
antidepressant provides more efficacy than either drug alone. Still, there is a wealth of
anecdotal reports suggesting that some patients appear to respond better to the combi-
nation than to monotherapy. Many patients have taken this combination with impunity,
although a number of caveats are in order. Obviously, as with any regimen involving an
MAOI, the usual food, beverage, and medication prohibitions apply. Certain medica-
tions are best avoided in this combination, more to prevent the hypermetabolic than the
hypertensive crisis. Imipramine, desipramine, and especially clomipramine are best
avoided as the tricyclic. Tranylcypromine may be chancier than phenelzine on the
MAOI side. It is best to stop previous medications for at least a week and then to begin
a low dose of a tricyclic, e.g., 50 mg of amitriptyline, trimipramine, or doxepin.
Isocarboxazid or phenelzine is then added in low doses, and, if tolerated, the doses of
the tricyclic, alternating with that of the MAOI, are "walked up" every few days
gradually and with cautious monitoring of the patient's symptoms and vital signs.
Patients with difficult-to-treat depressions are typically managed with different
drugs sequentially or often with combination therapies. This means a need for height-
ened awareness of potential interactions. Switching between heterocyclics and MAOIs,
or between two different MAOIs, presents potential hazards. At least a week's washout
should intervene when changing from a tricyclic to an MAOI, and probably 2 weeks
between an MAOI and a tricyclic. The very long half-lives of fluoxetine and its active
metabolite suggest the need for an even longer washout between fluoxetine and an
MAOI, since cases of potentially lethal hypermetabolic syndrome (presumably a seroto-
nin syndrome) have occurred following this switch of medicines. When patients have
been switched from phenelzine to tranylcypromine without sufficient washout, cases of
hypertensive crises and strokes have developed. A pharmacokinetic interaction has been
described between SSRls and tricyclic antidepressants, in which SSRI coadministration
markedly elevates the plasma level of the tricyclic (see Chapter 1, Table 1).
For further discussion and details on the treatment of the treatment-resistant
depression, the reader is referred to Ref. 112.


The 1980s saw the introduction of a number of new antidepressant drugs. In the
early 1980s, the only thing new that they seemed to provide was toxicity. Newer drugs,
however, have broadened our range of therapeutic options, both in terms of conditions
that are treatable and different adverse-effect profiles.
Research also has told us more about the appropriate use of old drugs, such as
tricyclics and MAOIs. One of the most important and troubling findings in recent years
is that a majority of depressed patients go untreated or undertreated with antidepressant

medications, making vigorous pharmacotherapy a wat~hword for the tum of the mil-
As we look to the future, health care professionals anticipate the advent of biolog-
ical tests that will allow us to select treatments with greater specificity. We also look
forward to new agents greater in efficacy but safer in practice.
Whatever the fruits of neuroscience, there always will be a role for sensitive and
skillful interpersonal interactions in the approach to depressed human beings. The
response of a patient to an antidepressant pharmaceutical is independent of whether the
etiology of that patient's condition is biological, psychosocial, or some combination of
the two. To paraphrase Maimonides, seeing in the patient the suffering human being
often is the greatest challenge for the clinician, yet the greatest blessing we may

I. Weissman M. M., Bland R. C., Canino G. 1., et al: Cross-national epidemiology of major depression
and bipolar disorder. lAMA 276:293-299, 1996.
2. Regier D. A., Myers J. K., Kramer M., et al: The NIMH Epidemiologic Catchment Area program:
Historical context, major objectives, and study population characteristics. Arch Gen Psychiatry
41:934-941, 1984.
3. Blazer D. G., Kessler R. C., McGonagle K. A., Swartz M. S.: The prevalence and distribution of major
depression in a national community sample: The National Comorbidity Survey. Am 1 Psychiatry
151 :979-986, 1994.
4. Freud S.: Mourning and melancholia, in Anonymous: Collected Papers. London, Hogarth Press, 1946,
5. Abraham K.: Notes on the Psychoanalytic Investigation and Treatment of Manic-Depressive Insanity
and Allied Conditions (1911). New York, Basic Books, 1960, p. 137.
6. Bibring E.: The mechanism of depression, in Greenacre P. (ed): Affective Disorders. New York,
International Universities Press, 1965, pp. 13-48.
7. Harlow H., Harlow M.: Learning to love. Sci Am 54:244-272, 1966.
8. Mahler M. S.: On sadness and grief in infancy and childhood: Loss and restoration of the symbiotic
love object. Psychoanal Study Child 16:332-351, 1961.
9. Kohut H.: The Analysis of the Self, New York, International Universities Press, 1971.
10. Kernberg 0.: Borderline Conditions and Pathological Narcissism. New York, Jason Aronson, 1975.
II. Beck A. T., Rush A. J., Shaw B., et al: Cognitive Therapy of Depression. New York, Guilford Press,
1979, p. I.
12. Seligman M., Maier S., Geer J.: The alleviation of learned helplessness and the dog. 1 Abnorm Soc
PsychoI73:256-262, 1968.
13. Lewinsohn P., Biglar A., Zeiss A.: Behavioral treatment of depression, in Davidson P. (ed): The
Behavioral Management of Anxiety, Depression and Pain. New York, Brunner-Maze!, 1976, pp. 91-
14. Decker E.: The Revolution in Psychiatry. London, Collier-Macmillan, 1964, p. 108.
15. Ehringer H., Hornykiewicz 0.: Verteilung von noradrenalin and D opamin (3-hydroxytyramin) im
gehirn des menschen und ihr verhalten bei erkrankungen des estrapyramidalen systems. Klin
Wochenschr 38: 1236-1239, 1960.
16. Cotzias G. c., Van Woert M. H., Schiffer L. M.: Aromatic amino acids and modification of parkinson-
ism. N Engl 1 Med 276:374-378, 1967.
17. Carlsson A., Lindquist M., Magnusson T., Woaldeck B.: On the presence of 3-hydroxytyramine in
brain. Science 127:471, 1958.

18. Schildkraut J. 1.: The catecholamine hypothesis of affective disorders: A review of supporting evi-
dence. Am J Psychiatry 122:509-522, 1965.
19. Bunney W. E. 1., Davis J. M.: Norepinephrine in depressive reactions: A review. Arch Gen Psychiatry
13:483-494, 1965.
20. Coppen A: The biochemistry of affective disorders. Br J Psychiatry 113:1237-1264, 1967.
21. Lapin I. P., Oxenkrug G. E: Intensification of the central serotonergic processes as a possible determi-
nant of the thymoleptic effect. Lancet 1:132-136,1969.
22. Murphy D. L., Cambell I., Costa 1. L.: Current status of the indoleamine hypothesis of the affective
disorders, in Lipton M. A, DiMascio A, Killam K. E (eds): Psychopharmacology: A Generation of
Progress. New York, Raven Press, 1978, pp. 1235-1247.
23. Heninger G. R., Charney D. S., Delgado P. L.: Neurobiology of treatments for refractory depression, in
Tasman A, Goldfinger S. M., Kaufman C. A. (eds): Review of Psychiatry. Washington, D.C., American
Psychiatric Press, 1990, pp. 33-58.
24. Charney D. S., Menekes D. B., Heninger G. R.: Receptor sensitivity and the mechanism of action of
antidepressant treatment. Arch Gen Psychiatry 38: 1160-1180, 1981.
25. Asberg M., Schalling D., Traskman-Bendz L., Wagner A.: Psychobiology of suicide, impulsivity, and
related phenomena, in Meltzer H. Y. (ed): Psychopharmacology:The Third Generation of Progress.
New York, Raven Press, 1987, pp. 655-668.
26. Roy A., Virkkunen M., Linnoila M.: Serotonin in suicide, violence, and alcoholism, in Coccaro E. E,
Murphy D. L. (eds): Serotonin in Major Psychiatric Disorders. Washington, D.C., American Psychi-
atric Press, 1990, pp. 187-208.
27. Siever L. 1., Davis K. L.: Overview: Toward a dysregulation hypothesis of depression. Am J Psychiatry
142: 1017-1031, 1985.
28. Winokur G., Coryell w.: Familial subtypes of unipolar depression: A prospective study of familial pure
depressive disease compared to depression spectrum disease. Bioi Psychiatry 32:1012-1018, 1992.
29. Bech P.: Symptoms and assessment of depression, in Paykel E. S. (ed): Handbook of Affective Disor-
ders. New York, Churchill Livingstone, 1992, pp. 3-13.
30. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed.
Washington, D.C., American Psychiatric Association, 1994.
31. Panksepp 1.: Towards a general psychobiological theory of emotions. Behav Brain Sci 5:407-467,
32. Halgren E., Marinkovic K.: Neurophysiological networks integrating human emotions, in Gazzaniga
M. S. (ed): The Cognitive Neurosciences. Cambridge, MIT Press, 1995, pp. 1137-1151.
33. Akiskal H. S., McKinney W. I., Jr.: Overview of recent research in depression; integration of ten
conceptual models into a comprehensive clinical frame. Arch Gen Psychiatry 32:285-305, 1975.
34. Keller M. B., Harrison W., Fawcett 1. A., et al: Treatment of chronic depression with sertraline or
imipramine: Preliminary blinded response rates and high rates of undertreatment in the community.
Psychopharmacol Bull 31:205-212, 1995.
35. Quitkin EM., McGrath P. 1., Stewart J. w., et al: Atypical depression, panic attacks, and response to
imipramine and phenelzine. Arch Gen Psychiatry 47:935-941, 1990.
36. Brotman A. w., Falk W. E., Gelenberg A. 1.: Pharmacologic treatment of acute depressive subtypes, in
Meltzer H. (ed): Psychopharmacology: The Third Generation of Progress. New York, Raven Press,
1987, pp. 1031-1040.
37. Weissman M. M.: The psychological treatment of depression. Arch Gen Psychiatry 36:1261-1268,
38. Weissman M. M., Prusoff B. A., DiMascio A., et al: The efficacy of drugs and psychotherapy in the
treatment of acute depression episodes. Am J Psychiatry 136:555-558, 1979.
39. Herceg-Baron R. L., Prusoff B. A., Weissman M. M., et al: Pharmacotherapy and psychotherapy in
acute depressed patients. A study of attrition patterns in a clinical trial. Compr Psychiatry 20:315-325,
40. Weissman M. M., Klerman G. L., Prusoff B. A., et al: Depressed outpatients: Results one year after
treatment with drugs and/or interpersonal psychotherapy. Arch Gen Psychiatry 38:51-55, 1981.

41. Kovacs M.: The efficacy of cognitive and behavior therapies for depression. Am J Psychiatry
137:1495-1501, 1980.
42. Frankel F. H.: Current perspectives on ECT: A discussion. Am J Psychiatry 134:1014-1019, 1977.
43. Grahme-Smith D. S., Green A. R., Costain D. W.: Mechanism of the antidepressant action of electro-
convulsive therapy. Lancet 1:254-256, 1978.
44. Blier P., de Montigny C., Chaput Y.: A role for the serotonin system in the mechanism of action of
antidepressant treatments: Preclinical evidence. J Clin Psychiatry 51: 14-20, 1990.
45. Small 1. G., Kellams 1. 1., Milstein v., et al: Complications with electroconvulsive treatment combined
with lithium. Bioi Psychiatry 15:103-112, 1980.
46. Mandel M. R., Madsen J., Miller A. L., et al: Intoxication associated with lithium and ECT. Am J
Psychiatry 137: 1107 -1109, 1980.
47. Squire L. R., Slater P. C.: Bilateral and unilateral ECT: Effects on verbal and nonverbal memory. Am J
Psychiatry 135:1316-1320, 1978.
48. Fink M.: An adequate treatment? Editorial. Convulsive Ther 5:311-313, 1989.
49. Delgado P. L., Charney D. S., Price L. H., Aghajanian G. K., Landis H., Heninger G. R.: Serotonin
function and the mechanism of antidepressant action: Reversal of antidepressant induced remission by
rapid depletion of plasma tryptophan. Arch Gen Psychiatry 47:411-418, 1990.
50. Delgado P. L., Price L. H., Miller H. M., et al: Rapid serotonin depletion as a provocative challenge test
for patients with major depression: Relevance to antidepressant action and the neurobiology of depres-
sion. Psychopharmacol Bull 27:321-330, 1991.
51. Delgado P. L., Miller H. M., Salomon R. M., Licinio 1., Gelenberg A. 1., Charney D. S.: Monoamines
and the mechanism of antidepressant action: Effects of catecholeamine depletion on mood in patients
treated with antidepressants. Psychopharmacol Bull 29:389-396, 1993.
52. Delgado P. L., Gelenberg A. J.: Decision making in the use of antidepressants. in Directions in
Psychiatry, Vol. 14, New York, Hatherleigh, 1994, pp. 1-8.
53. Maas J. W.: Biogenic amines and depression: Biochemical and pharmacological separation of two types
of depression. Arch Gen Psychiatry 32:1357-1367, 1975.
54. Feighner 1. P.: The role of venlafaxine in rational antidepressant therapy. J Clin Psychiatry 55 (Suppl
9A):62-68, 1994.
55. Ellison 1. M., Milofsky J. E., Ely E.: Fluoxetine-induced bradycardia and syncope in two patients. J
Clin Psychiatry 51 :385-386, 1990.
56. Feder R.: Bradycardia and syncope induced by fluoxetine. J Clin Psychiatry 52:139, 1991.
57. Glassman A. H., Roose S. P., Bigger 1. T., Jr.: The safety of tricyclic antidepressants in cardiac patients:
Risk/benefit reconsidered. JAMA 269:2673-2675, 1993.
58. Jefferson J. W: Cardiovascular effects and toxicity of anxiolytics and antidepressants. J Clin Psychiatry
50:368-378, 1989.
59. Glassman A. H., Roose S. P.: Cardiovascular effects of tricyclic antidepressants. Psychiatr Ann
17:340-347, 1987.
60. Davidson J.: Seizures and bupropion: A review. J Clin Psychiatry 50:256-261, 1989.
61. Gelenberg A. J., Jefferson J. W: Lithium tremor. J Clin Psychiatry 56:283-287, 1995.
62. Lipinski J. F., Mallya G., Zimmerman P., Pope H.: F1uoxetine-induced akathisia: Clinical and theoreti-
cal implications. J Clin Psychiatry 50:339-342, 1989.
63. Rothschild A. J., Locke C. A.: Reexposure to fluoxetine after serious suicide attempts by three patients:
The role of akathisia. J Clin Psychiatry 52:491-493, 1991.
64. Preskorn S. H., Othmer S. c.: Evaluation of bupropion hydrochloride: The first of a new class of
atypical antidepressants. Pharmacotherapy 4:20-34, 1984.
65. Teicher M. H., Glod c., Cole 1. 0.: Emergence of intense suicidal preoccupation during fluoxetine
treatment. Am J Psychiatry 147:207-210, 1990.
66. Monteiro W.O., Noshirvani H. E, Marks 1. M., et al: Anorgasmia from clomipramine in obsessive-
compulsive disorder: A controlled trial. Br J Psychiatry 151: 107-112, 1987.
67. Jacobsen F. M.: F1uoxetine-induced sexual dysfunction and an open trial of yohimbine. J Clin Psychia-
try 53:119-122, 1992.

68. Herman J. B., Brotman A. W., Pollack M. H., Falk W. E., Biederman J., Rosenbaum 1. E: F1uoxetine-
induced sexual dysfunction. J c/in Psychiatry 51:25-27, 1990.
69. Musher 1. S.: Anorgasmia with the use of fluoxetine. Am J Psychiatry 147:948, 1990.
70. Patterson W. M.: F1uoxetine-induced sexual dysfunction. J Clin Psychiatry 54:71, 1993.
71. Physicians' Desk Reference, Montvale, New Jersey, Medical Economics, 1994, p. 879.
72. Walker P. w., Cole J. 0., Gardner E. A., et al: Improvement in fluoxetine-associated sexual dysfunction
in patients switched to bupropion. J c/in Psychiatry 54:459-465, 1993.
73. Rothschild A. 1.: Selective serotonin reuptake inhibitor-induced sexual dysfunction: Efficacy of a drug
holiday. Am J Psychiatry 152:1514-1516, 1995.
74. Sovner R: Treatment of tricyclic antidepressant-induced orgasmic inhibition with cyproheptadine. J
Clin Psychopharmacol4:169, 1984.
75. Decastro R M.: Reversal of MAOI-induced anorgasmia with cyproheptadine. Am J Psychiatry
142:783, 1985.
76. Riley A. 1., Riley E. 1.: Cyproheptadine and antidepressant-induced anorgasmia. Br J Psychiatry
142:217 -218, 1986.
77. Kahn D. A.: Possible toxic interaction between cyproheptadine and phenelzine. Am J Psychiatry
144:1242-1243, 1987.
78. Steele T. E., Howell E. E: Cyproheptadine for imipramine-induced anorgasmia. J c/in Psychophar-
macoI6:326-327, 1986.
79. Goldbloom D. S., Kennedy S. H.: Adverse interaction of fluoxetine and cyproheptadine in two patients
with bulimia nervosa. J c/in Psychiatry 52:261-262, 1991.
80. Zajecka J., Fawcett 1., Schaff M., Jeffriess H., Guy C.: The role of serotonin in sexual dysfunction:
F1uoxetine-associated orgasm dysfunction. J Clin Psychiatry 52:66-68, 1991.
81. Cohen A. 1.: F1uoxetine-induced yawning and anorgasmia reversed by cyproheptadine treatment. J c/in
Psychiatry 53: 174, 1992.
82. Price 1., Grunhaus L. J.: Treatment of clomipramine-induced anorgasmia with yohimbine: A case
report. J Clin Psychiatry 51 :32-33, 1990.
83. Hollander E., McCarley A.: Yohimbine treatment of sexual side effects induced by serotonin reuptake
blockers. J c/in Psychiatry 53:207-209, 1992.
84. Gross M. D.: Reversal by bethanechol of sexual dysfunction caused by anticholinergic antidepressants.
Am J Psychiatry 139: 1193-1194, 1982.
85. Sorscher S. M., Dilsaver S. c.: Antidepressant-induced sexual dysfunction in men: Due to cholinergic
blockade? J Clin Psychopharmacol 6:53-55, 1986.
86. Yager J.: Bethanechol chloride can reverse erectile and ejaculatory dysfunction induced by tricyclic
antidepressants and mazindol: Case report. J c/in Psychiatry 47:2\0-211, 1986.
87. Segraves R T.: Reversal by bethanechol of imipramine-induced ejaculatory dysfunction. Am J Psychia-
try 144:1243-1244, 1987.
88. Pollack M. H., Rosenbaum J. E: Management of antidepressant-induced side effects: A practical guide
for the clinician. J c/in Psychiatry 48:3-8, 1987.
89. Balogh S., Hendricks S. E., Kang J.: Treatment of fluoxetine-induced anorgasmia with amantadine. J
c/in Psychiatry 53:212-213, 1992.
90. Labbate L. A., Pollack M. H.: Treatment of flouxetine-induced sexual dysfunction with bupropion: A
case report. Ann Clin Psychiatry 6: 13-15, 1994.
91. Boehnert M. T., Lovejoy E H., Jr.: Value of the QRS duration vs the serum drug level in predicting
seizures and ventricular arrhythmias after an acute overdose of tricyclic antidepressants. N Engl J Med
313:474-479, 1985.
92. Foulke G. E., Albertson R. E.: QRS interval in tricyclic antidepressant overdose: Inaccuracy as a
toxicity indicator in emergency setting. Ann Emerg Med 16:160-163, 1987.
93. Kapur S., Mieczkowski T., Mann 1. J.: Antidepressant medications and the relative risk of suicide
attempt and suicide. JAMA 268:3441-3445, 1992.
94. Crome P., Dawling S., Braithwaite R. A., et al: Effect of activated charcoal on absorption of nortrip-
tyline. Lancet 2:1203-1205, 1977.

95. Granacher R. D., Baldessarini R. J.: Physostigmine: Its use in acute anticholinergic syndrome with
antidepressant and antiparkinson drugs. Arch Gen Psychiatry 32:375-379, 1975.
96. Stoll A. L., Pope H. G., McElroy S. L.: High-dose fluoxetine: Safety and efficacy in 27 cases. J Clin
Psychopharmacol 1\ :225-226, 1991.
97. Beasley C. M., Jr., Sayler M. E., Weiss A M., Potvin J. H.: Auoxetine: Activating and sedating effects
at multiple fixed doses. J Clin Psychopharmacol 12:328-333, 1992.
98. Keck P. E., Jr., Yuckovic A, Pope H. G., Jr., Nierenberg A. A., Gribble G. w., White K.: Acute
cardiovascular response to monoamine oxidase inhibitors: A prospective assessment. J Clin Psycho-
pharmacoI9:203-206, 1989.
99. Gelenberg A. J.: MAOIs: Assorted concerns. Bioi Ther Psychiatry 11:33-36, 1988.
100. Teicher M. H., Cohen B. M., Baldessarini R. J., et al: Severe daytime somnolence in patients treated
with an MAOL Am J Psychiatry 145:1552-1556, 1988.
101. Dilsaver S. C.: Monoamine oxidase inhibitor withdrawal phenomena: Symptoms and pathophysiology.
Acta Psychiatr Scand 78:1-7, 1988.
102. Tollefson G. D.: Monoamine oxidase inhibitors: A review. J Clin Psychiatry 44:280-288, 1983.
103. Kaplan R. E, Feinglass N. G., Webster w., et al: Phenelzine overdose treated with dantrolene sodium.
JAMA 255:642-644, 1986.
104. Breheny EX., Dobb G. J., Clarke G. M.: Phenelzine poisoning, case report. Anesthesia 41:53-56,
105. Frank E., Kupfer D. J., Perel J. M., et al: Three-year outcomes for maintenance therapies in recurrent
depression. Arch Gen Psychiatry 47:1093-1099, 1990.
106. Himmelhoch J. M., Thase M. E., Mallinger A G., Houck P.: Tranylcypromine versus imipramine in
anergic bipolar depression. Am J Psychiatry 148:910-916, 1991.
107. Sachs G. S., Lafer B., Stoll A. L., et al: A double-blind trial ofbupropion versus desipramine for bipolar
depression. J Clin Psychiatry 55:391-393, 1994.
108. Spiker D. G., Weiss J. C., Dealy R. S., et al: The pharmacological treatment of delusional depression.
Am J Psychiatry 142:430-436, 1985.
109. Anton R. E, Sexauer J. D.: Efficacy of amoxapine in psychotic depression. Am J Psychiatry 140:1344-
1347, 1983.
110. Rothschild A J., Samson J. A, Bessette M. P., Carter-Campbell J. T.: Efficacy of the combination of
fluoxetine and perphenazine in the treatment of psychotic depression. J Clin Psychiatry 54:338-342,
HI. Siris S. G., Bermanzohn P. C., Gonzalez A., Mason S. E., White C. Y., Shuwall M. A.: The use of
antidepressants for negative symptoms in a subset of schizophrenic patients. Psychopharmacol Bull
27:331-335, 1991.
112. Roose S. P., Glassman A. H.: Treatment Strategies for Refractory Depression. Washington, D.C.,
American Psychiatric Press, Inc., 1990.
113. Anonymous: Treatment of xerostomia. Med Lett 30:74-76, 1988.
114. Harrison W. M., McGrath P. J., Stewart J. W., Quitkin E: MAOIs and hypertensive crises: The role of
OTC drugs. J Clin Psychiatry 50:64-65, 1989.
115. Winokur G.: The concept of secondary depression and its relationship to comorbidity. Psychiatr Clin
North Am, 13:567-583, 1989.
Bipolar Disorder


Bipolar disorder is one of the most common and treatable of the mood disorders,
affecting over I % of adults in industrialized nations over the course of a lifetime. l The
essential feature of bipolar disorder is the presence of one or more manic episodes;
usually, there is also a history of major depressive episodes. Bipolar II disorder is
characterized by the occurrence of one or more major depressive episodes and at least
one hypomanic episode.
The clinician can use the severity and pattern of a patient's illness as a baseline for
monitoring the effectiveness of pharmacotherapy. Because of the range of clinical
presentations and the wide variation in the course of bipolar illness, the clinician must
follow each individual carefully.
The successful management of bipolar disorder usually requires early, aggressive,
and long-term treatment. Lithium (Eskalith® and others) has had an important impact
on the treatment of bipolar disorder, literally saving the lives of many patients. Without
treatment, mortality is two to three times higher for bipolar patients than it is for the
general population; with lithium treatment, mortality is not significantly higher in
bipolar patients. 2 However, lithium does not work for every patient. An estimated 20-
40% of people with bipolar disorder do not adequately respond to or are unable to
tolerate lithium. Lithium's drawbacks-including side effects, the risk of toxicity, and
limited efficacy in some patients-have encouraged research into other treatments.
Divalproex sodium (Depakote®) is now approved by the u.s. Food and Drug Adminis-
tration (FDA) as a treatment for acute mania, and another anticonvulsant, car-
bamazepine (Tegretol® and others), has shown promise in clinical trials as an adjunct or
alternative to lithium or valproate treatment. Antipsychotic, anxiolytic/hypnotic, and

Heather S. Hopkins and Alan J. Gelenberg • Department of Psychiatry, University of Arizona Health
Sciences Center, Tucson, Arizona 85724.
The Practitioner's Guide to Psychoactive Drugs (Fourth Edition), Gelenberg and Bassuk, eds.
Plenum Publishing Corporation, New York, 1998


antidepressant medications are used as adjuncts to the mood stabilizers for episodes of
acute mania and acute depression, and nonpharmacological treatments such as electro-
convulsive therapy and psychotherapy can also play important roles in the treatment of
this disorder.

A. Clinical Presentation
Bipolar disorder occurs in about 1.6% of the general population, with an equal
incidence in men and women.) In a cross-national collaborative study, the lifetime rates
of bipolar disorder were found to be more consistent across countries than the rates for
major depression (0.3% in Taiwan to 1.5% in New Zealand).3 Patients usually develop
their first mood episode (most often mania or hypomania) between the ages of 20 and
40 years (average 32), although presentation during the teenage years is not uncom-
mon. Late onset of mania should prompt a neurologic and medical differential diag-
nosis, since mania may be secondary to many other conditions. Gender mayor may not
be a factor in the age of onset of bipolar disorder. 4 - 7 There is some evidence that
female bipolar patients have more depressive episodes than male bipolar patients do,
but again the data are contradictory. 4
In mania, elevated mood is common but not universal. Sometimes, depressive
symptoms break through a facade of emotional lability. Pressure of speech, flight of
ideas, increased motor activity, and decreased sleeping time often accompany the sense
of elation and well-being. These patients are very self-involved but often can sense the
vulnerabilities of others and become inappropriate and intrusive. The manic patient
often gets too involved in pleasurable activities, with painful consequences in family
relationships, social interactions, economic stability, and health. If mania progresses,
the patient's sense of humor may border on anger and irritability. Pressured speech can
deteriorate into fragmented sentences and clang associations, while a sense of well-
being may escalate into grandiosity. In the most extreme form, the patient becomes
psychotic. Auditory hallucinations, delusions, and paranoia are frequent, and incoher-
ence, sensory distortions, agitation, and combativeness often accompany the most
disorganized states. Manic patients frequently require hospitalization to prevent them
from harming themselves or others.
Clinicians have described some cases of prolonged mania or intermittent mania
without depression. Most commonly, however, the illness is bipolar, and the first
hospitalization is usually for mania. Bipolar depression typically is indistinguishable
from other forms of major depression, except that episodes may be more frequent and
shorter in duration. The fourth edition of the Diagnostic and Statistical Manual of
Mental Disorders (DSM-IV) defines the symptom criteria, which include persistently
lowered mood; psychomotor agitation or retardation; feelings of worthlessness or guilt;
suicidal ideation or attempts; and various somatic concerns, such as weight loss,
insomnia, and fatigue. 8
Occasionally, a mood episode in a bipolar patient is not clearly mania or depres-
sion, but a mixture of both. Recent evidence suggests that this mixed picture may be
less treatment-responsive and have a poorer prognosis than simple mania or depres-

sion.9 Similarly, rapid cycling, in which a patient has 4 or more episodes in one year,
and "cycling episodes," in which patients move from one mood pole to the other
without an intervening period of euthymia, also have a more guarded prognosis.to,1I

B. Course of the Illness

The manic and depressive episodes of bipolar illness are usually time limited
(typically 3 to 9 months)-most often shorter than those of pure mania or major
depression. As patients grow older, however, the episodes may occur more frequently
and last longer. Most bipolar individuals become relatively symptom-free between
episodes. However, one-quarter to one-third of patients exhibit significant symptoms or
decreased functioning between episodes, which worsens with recurrences,l2 Although
general trends exist, the course of bipolar disorder varies considerably in different
individuals. Many patients have episodes every few years, while others have clusters of
episodes. Women with bipolar disorder have an increased risk of relapse during the
postpartum period. 13,14 In fact, the risk of a mood episode is higher after childbirth than
in any other period during the life of a male or female bipolar patient. 4
In general, recurrent episodes are the rule rather than the exception: at least 80%
of patients who have an initial episode of mania will have one or more future epi-
sodes. 15 Psychosocial stressors and mood episodes may affect gene expression, leaving
a neurobiological trace that makes patients more vulnerable to recurrent episodes. 16
After a series of episodes, a patient's illness may become more "malignant," develop-
ing an autonomy that might require different medications. For example, a patient who
experiences three or more prior episodes may be less responsive to standard-range
lithium treatment,17 In addition, bipolar episodes might be associated with neuro-
anatomic and cognitive changes I8: some bipolar patients have shown cognitive impair-
ment that persists beyond the resolution of a mood episode,I9,2o and others have
exhibited differences in brain structure from normal controls. 21 - 24 In particular, bipolar
patients-more commonly male and older patients-have been found to have enlarged
ventricles. 25 ,26
Bipolar disorder appears to be an inherited illness, and studies have indicated that
anticipation-a decrease in age at onset and an increase in disease severity in succeed-
ing generations-may occur.27 Investigators are attempting to identify the specific
genes responsible for causing bipolar disorder,28 which would allow for much earlier
detection and treatment of this illness. Findings so far suggest a complex mode of
inheritance: there may be several genes that make a person susceptible to this disorder,
rather than a single dominant gene. Of course, there are likely to be some cases of
bipolar disorder that are not transmitted genetically.
The effects of untreated bipolar disorder can be devastating. An estimated one-
quarter of bipolar patients attempt suicide (a figure exceeded only in schizophrenia and
schizoaffective disorder),29 and probably many more engage in other hazardous behav-
iors. In addition, work, family, and social contacts deteriorate, initially around episodes
but later even between episodes. A U.S. Public Health Service study reported that
without adequate treatment, the average woman whose bipolar disorder begins at age

25 can expect to lose approximately 9 years of life, 14 years of major activities such as
work, school, or child rearing, and 12 years of normal health.30 With optimal treatment,
however, the woman will regain 6.5 years of life, 10 years of effective major activity,
and 8.5 years of normal health. Poorly controlled mania can lead to family dysfunction,
economic collapse, auto accidents, and the contraction of sexually transmitted diseases.
Although the trends gleaned from longitudinal studies can provide a general set of
expectations about the course of bipolar disorder, the clinician must assess the nature
and course of each patient's illness individually. An appropriate treatment plan can be
developed only after evaluating the impact and meaning of the illness in the patient's
life, along with objective clinical symptoms.


Many nonpharmacologic approaches are used to treat bipolar illness, including

electroconvulsive therapy (ECT), milieu therapy, group therapy, and various individual

A. Electroconvulsive Therapy
Electroconvulsive therapy is an effective treatment for mania and for major de-
pression (see Chapter 2). When an acutely manic patient is unresponsive or intolerant
to medication treatment or medication presents a threat to the patient's well-being (e.g.,
during pregnancy), ECT should be seriously considered and may be lifesaving. Al-
though we have no coherent theories about why ECT is effective in acute mania, there
are widespread clinical impressions and evidence from the literature of its safety and
efficacy. Reports in the medical literature over the past 50 years show ECT to be
associated with remission or marked clinical improvement in 470 (80%) of 589 manic
patients: 78% of manic patients in early studies, 85% of patients in six retrospective
studies reported since 1976, and 77% of patients in two prospective studies. 31
Because there is the impression in the field that ECT works faster than phar-
macologic treatment, it is the preferred treatment for patients who are homicidal,
psychotic, or severely suicidal. Unfortunately, there is little experimental evidence to
support this belief. One study found ECT to be superior to lithium during the first 8
weeks of treatment, especially for severely manic patients and those with mixed
states. 32 After 8 weeks, however, there were no significant differences between the
lithium- and ECT-treated patients on ratings of manic and depressive symptoms or in
the incidence of relapses, recurrences, or rehospitalizations.
As in acute mania, ECT is the first choice for bipolar depressed patients who
require a rapid response, cannot tolerate or be exposed to medication, or do not respond
to medication. There is some evidence that ECT has a more rapid therapeutic onset
than antidepressant drugs. 33 - 35 Bilateral ECT may be more efficacious than unilateral
ECT for bipolar patients, and patients who receive a dosage over twice the initial
seizure threshold, with either unilateral and bilateral treatment, may show more rapid
improvement of their depressive symptoms.

If a bipolar patient to be treated with ECT is taking lithium, it is best to discon-

tinue lithium during the treatment series. Although controversy remains, there is some
evidence that coadministering lithium and ECT leads to an increased incidence of
confusion. In patients receiving neuromuscular blockers for ECT or surgical proce-
dures, lithium might sometimes prolong the recovery period for muscle function.
Circumstances that Suggest the Use of ECT
1. Previous good response to ECT.
2. Medical contraindication to drug therapy.
3. Immediate threat of suicide or homicide, or serious withdrawal or regression
that interferes with self-care.
4. Failure of successive trials of pharmacotherapy.
5. Acute disorganization or combativeness that requires immediate behavioral

B. Milieu Therapy
To ensure a manic patient's safety and to contain symptoms, inpatient care may be
necessary. These patients, however, frequently disrupt a ward milieu. The patient's
infectiously elevated mood may be reinforced by staff and other patients, and irri-
tability and hostility can be highly provocative. Structuring daily activities, correcting
cognitive distortions, and limiting external stimulation by staff and patients may reduce
the patient's excitement. Caretakers should remember that milieu therapy and phar-
macotherapy are synergistic. A safe and stable environment promotes medication com-
pliance, and the salutary effect of drugs increases insight, emotional availability, under-
standing, and formation of social attachments.

c. Group Therapy
Fieve has described "lithium groups" designed to discuss issues related to mainte-
nance therapy.36 This approach is not only time efficient for clinicians but also advan-
tageous for patients. Sharing similar concerns about medication usage can be both
supportive and growth-producing and can increase compliance. Studies have shown
that group therapy, in addition to a medication regimen, can decrease rates of hospital-
ization and improve social and occupational functioning. 37 - 39
Families and significant others can also benefit from learning about mood disor-
ders and sharing common approaches to dealing with their loved ones. Support and
practical counseling are valuable, and people close to the patients can learn the prodro-
mal signs of impending relapse, noninflammatory ways of coping with symptoms, how
to predict side effects, the importance of avoiding sleepless periods, and more. Bipolar
patients who have received family therapy in addition to pharmacologic treatment in
several different studies have shown significantly better social functioning and family
interactions and required fewer hospitalizations than those who have received only
standard pharmacotherapy.4O-43
Although group therapy can be helpful during the maintenance phase of treatment

for patients with mood disorders, it can be counterproductive during episodes of acute
illness, particularly for hospitalized patients. A depressed patient might become irrita-
ble with forced socialization. A manic patient, in contrast, can become more excited
with interpersonal stimulation and, in tum, may be provocative to others.

D. Psychotherapy and the Therapeutic Relationship

An appropriate role for psychotherapy in the treatment of patients with bipolar
illness is a subject of ongoing research. Even assuming that disorders of mood, partic-
ularly bipolar illness, reflect genetic diatheses, nonbiological treatments can augment
pharmacotherapy.44 Perhaps psychotherapy acts synergistically to diminish the likeli-
hood of recurrences. Lithium might diminish the amplitude of reaction to life events
and stresses, while psychotherapy may allow a patient to avoid interpersonal crises and,
when they occur, to deal with them more adaptively.
Data about whether or not psychoanalytically oriented therapies are appropriate
for bipolar patients are inconclusive. One study randomly assigned 26 bipolar patients
maintained on lithium to standard clinical care or individual cognitive therapy interven-
tion. 45 Cognitive intervention produced slightly better compliance at the end of treat-
ment and, at 6-month follow-up, therapy patients were significantly less likely to have
major adherence problems, discontinue lithium against medical advice, have episodes
brought on by not taking lithium, or be hospitalized. For most patients, here-and-now
approaches combining cognitive, interpersonal, behavioral, and psychoeducational
components with appropriate balancing of individual, group, and family treatments can
be useful.
Although larger, better controlled, and methodologically rigorous studies must be
performed before we can draw definitive conclusions about the benefits of psychoso-
cial treatments in bipolar disorder, we do know that a sustained relationship with a
psychotherapeutically oriented clinician can improve the outcome of drug treatment.
Patients will be more likely to tolerate adverse effects and to work with the doctor to
find optimal drug levels and contraactive options. Family members will be better
educated to spot early signs of toxicity and to report prodromata of relapse or recur-
rence. Patients in regular contact with a therapist are less likely to drop out of treatment
and more likely to be picked up early in the course of a new mood episode.

Pharmacologic agents have markedly decreased the morbidity associated with
bipolar illness, recurrent depressions, and several other cyclic psychiatric disturbances.
Antipsychotic drugs provide an effective means of controlling the behavioral symp-
toms and thought disorder that accompany mania and hypomania. Antidepressant
medications usually reverse depressive episodes. Lithium can diminish the symptoms
of acute mania and, over the long term, often suppresses both manic and depressive
episodes. Valproate is now considered a first-line treatment for acute mania. Car-
bamazepine, clozapine (Clozaril®), and various other agents are being tried as adjuncts
or second-line options for patients who do not respond to lithium satisfactorily.

A. Lithium
Naturally occurring salts of the chemical element lithium have been used to treat
various medical conditions since the 1800s. Lithium currently is used to treat acute
hypomanic or manic episodes and recurrent mood disorders. Despite its effective-
ness, its mechanism of action remains unclear. One theory is that lithium alters neuro-
nal function by affecting the distribution and kinetics of other ions that have similar
properties, such as sodium, potassium, magnesium, and calcium. 46 However, abnor-
malities in these systems do not appear to be specific to mood disorders. Lithium's
mechanism of action may be related to its effects on the neurotransmitters serotonin,
dopamine, norepinephrine, or acetylcholine. A third possibility is that lithium stabilizes
mood through its effects on second-messenger and signal-transduction systems 46 - 50 or
by its actions on signal-transducing guanine nucleotide-binding (G) proteins. 51 The G
proteins help regulate neuronal function and maintain the functional balance between
neurotransmitter systems in the central nervous system (eNS) and may be the molecu-
lar sites underlying the pathophysiology of this illness.
In acute hypomanic or manic episodes, lithium is frequently effective within 1-2
weeks but may require several more weeks or even a few months to contain the episode
fully. Sometimes an antipsychotic agent [such as haloperidol (Haldol® and others)]
should be used acutely to manage behavioral excitement and acute psychotic symp-
toms. After the acute symptoms are diminished, lithium maintenance therapy decreases
the number, severity, and frequency of mood episodes. Even with drug treatment,
however, some patients experience various symptoms, intermittent periods of distress,
or unwanted effects. To achieve optimal results, the clinician may need to experiment
with various dosage levels and continue maintenance therapy for a year or longer. In
addition, patients with acute depressive symptoms tend not to respond to lithium alone.
Eliminating significant mood swings is the optimal clinical goal of lithium main-
tenance therapy. However, mood fluctuations, and even the recurrence of a major
episode, do not indicate that lithium treatment has failed. Some clinicians feel that
manic or depressive symptoms that occur within a few months of initiating treatment
represent a relapse of the initial episode and may require medication adjustment.
Patients on a maintenance lithium regimen also may develop mild manic and depres-
sive symptoms that subside spontaneously or respond to interpersonal support or a
transient increase in lithium. Hypomania appears more likely than minor depression to
be a prodrome of a major episode. Optimal dosing of lithium can diminish subsyndro-
mal symptoms. 52
Long-term studies have consistently demonstrated that lithium maintenance sup-
presses mood swings in most bipolar patients. 53 However, individuals must take appro-
priate amounts of medication consistently. Most failures of lithium therapy occur
during the first year of treatment and result from inadequate resolution of the initial
episode, inadequate regulation of dosage, or lack of patient compliance. About 20-
30% of patients discontinue lithium therapy on their own.

The Most Common Reasons for Stopping Antimanic Medication

1. Denial of the illness.

2. Unawareness of the consequences of stopping medication.

3. Feeling no effects from the drug.

4. Intolerable adverse reactions.
5. Longing to be "high."
6. Fear of diminished competence, productivity, or creativity. 54
If these and other concerns are discussed with the patient, the number of treatment
failures attributed to poor patient compliance (three-fourths of the total) should be

1. Pharmacokinetics

Lithium is administered clinically as the citrate or carbonate. In the body, lithium

circulates as a small ion with a single positive electrical charge. It is rapidly and usually
fully absorbed after oral administration, but extended-release preparations also are
available. Although the exact sites of absorption are unclear, blood levels after single
oral doses peak in 1-4 hr (6-8 hr after delayed-release dosage), and complete absorp-
tion takes about 8 hr (see Table 1). In practice, the clinician should obtain serum
lithium levels 12 hr after the last dose: standard recommendations are based on this
reference time.
Unbound to plasma proteins, the lithium ion initially is distributed in the extra-
cellular fluid and then enters different tissues at different rates. A dynamic equilibrium
is established between plasma and cells, and at steady state-usually achieved between
3 and 8 days-the plasma lithium will accurately reflect total body lithium and can be
used to monitor treatment. The apparent volume of distribution of lithium is about the
same as total body water (approximately 70% of body weight), but at equilibrium
different organs concentrate lithium to differing degrees. In the liver, the lithium
concentration is lower than that in extracellular fluid, whereas in muscle, bone, and
thyroid, it is two to four times higher, and in brain about the same as in extracellular
fluid-roughly 40% of the concentration in plasma. 55
Lithium is not metabolized. About 95% is excreted by the kidneys; small amounts
appear in sweat and feces. Lithium is fully filtered and about 80% reabsorbed in the
proximal tubules. Small amounts are reabsorbed in the loop of Henle, but most of the
remaining is excreted in the urine. Under normal body conditions, the amount of
lithium filtered and reabsorbed is constant. Lithium excretion is, therefore, proportio-

TABLE 1. Pharmacokinetics

Protein Peak plasma Steady

Agent Absorption binding levels state Half-life

Lithium 6-8 hr 0% 1-4 hr 4-5 days 18-36 hr

Valproate Rapid 90-95%a 2-4 hr 1-4 days 9-16 hr
Carbamazepine Slow 55-76% \.5-6 hr 4-5 days or 25-65 hr or
3-4 weeks 8-29 hr

a At serum concentrations up to 50 IJ.g/mL.

b At serum concentrations of 50-100 IJ.g/mL.

nate to plasma concentrations, and lithium clearance is relatively constant for each
patient. With a single dose of lithium, 50% is excreted within 5-8 hr. In patients on
maintenance lithium, the half-life is about 24 hr. However, depending on the age and
kidney status of patients, the half-life may vary widely (approximately 18 hr in young,
healthy individuals to 36 hr in older patients).
Effective regulation of lithium depends on the sodium and fluid balance of the
body. When lithium is initially administered, a cation balance among lithium, sodium,
and potassium occurs; sodium and potassium excretion increases. Clinically, the bal-
ance between lithium and sodium is important for the following reasons:
1. Sodium depletion can result in marked lithium retention and possible toxicity.
2. High levels of lithium can lead to increased sodium excretion.
In cases of toxicity, the loss of body sodium obligates more lithium retention by the
kidney, prolonging the toxicity. Decreased fluid intake also can diminish lithium excre-
tion and could lead to intoxication.

2. Adverse Reactions

Lithium causes a wide range of unwanted effects (see Table 2). Many appear at the
onset of therapy, but patients soon become tolerant to them, and they cease to be a
problem. As in all clinical pharmacology, patients' sensitivity to a side effect at a given

TABLE 2. Adverse Reactions to Lithium

Nausea and vomiting
Muscle weakness
Fine tremor
Increased thirst (polydipsia)
Increased urination (polyuria)
Potentially indicative of toxicity
Nausea, vomiting, diarrhea
Drowsiness and mental dullness
Slurred speech
Coarse tremor and twitching
Muscle weakness
Persistent but benign
Increased thirst
Increased urination
Fine tremor
Weight gain

serum lithium concentration will vary. For some patients, a concentration necessary for
an effective result will produce unacceptable adverse reactions. For most patients, side
effects are tolerable at therapeutic levels, but the appearance or reappearance of certain
problems-such as tremor, gastrointestinal disturbances, or increased urination-
could herald an increase in blood level to toxic concentrations (see Section III.AA).

a. Endocrine. The long-term use of lithium may cause various endocrine abnor-
i. Thyroid. Lithium therapy can produce significant changes in thyroid function-
ing. It inhibits several steps in the process of hormone synthesis and degradation,
including iodine uptake by the thyroid gland, iodination of tyrosine, the release of
triiodothyronine (T3) and thyroxine (T4), the peripheral degradation of thyroid hor-
mones, and the stimulating effects of thyroid-stimulating hormone (TSH). Patients can
usually compensate for the initial decrease in thyroid hormones. However, approx-
imately 5% of patients treated with lithium develop signs of hypothyroidism; another
3% develop a diffuse nontender goiter. 56 This latter effect occurs more commonly in
women, between 5 months and 2 years after starting treatment. In addition, about 30%
of patients develop consistently increased levels of TSH indicated by elevated hormone
or abnormal clinical signs and symptoms. 57 Each of these thyroid abnormalities typ-
ically remits when lithium is stopped.
Hypothyroidism can be confused with a depressive disorder. Therefore, the practi-
tioner should be sensitive to clinical manifestations and laboratory findings indicative
of thyroid disease. 58
Since lithium affects thyroid function so profoundly, the clinician should take the
following measures during the various stages of maintenance therapy:
Before Starting Treatment
1. Screen for signs and symptoms of preexisting thyroid disease.
2. Obtain laboratory tests, including TSH, thyroxine (T4) T4I (free thyroxine
index), and T3RU (resin uptake).
During Treatment
1. Obtain a TSH measurement every 6 months during the first year and yearly
2. Follow up an elevated TSH with a complete battery of thyroid tests.
3. Maintain a clinical index of suspicion for thyroid dysregulation.
4. Initiate thyroid replacement therapy if hypothyroidism or elevated TSH oc-
curs (discontinuing lithium is not necessary).
In rare instances, lithium appears to trigger cases of hyperthyroidism. A high
index of suspicion is the best guideline, with appropriate testing of thyroid function to
confirm the clinician's impressions.
ii. Diabetes Mellitus. Lithium has many effects on carbohydrate metabolism,
including altered glucose tolerance, that are incompletely understood. Rarely, patients
may develop diabetes mellitus. 59 In these cases, the clinician must decide whether or
not the risk from diabetes outweighs the morbidity of the individual's affective disor-

der. Generally, diabetic patients tolerate lithium well. Sometimes, however, this agent
causes altered sensitivity to insulin, requiring a dosage readjustment.

b. Renal. As noted earlier, the body depends almost entirely on the kidneys to
eliminate lithium. In a patient with no glomerular filtration, who is totally dependent on
dialysis, a dose of lithium will remain in the body from one dialysis until the next.
Patients with diminished renal filtration will have prolonged elimination half-lives of
Aside from the kidney being the primary organ of lithium excretion, it is also the
site of some of the most common side effects, polydipsia and polyuria, which affect
roughly 60% of patients taking lithium. Although these effects usually are reversible, in
occasional patients they may persist after lithium has been discontinued, suggesting
some degree of structural impairment. There have also been rare reports of nephrotic
syndrome associated with lithium treatment.
i. Decreased Concentration. Presumably by counteracting the effects of anti-
diuretic hormone (ADH or vasopressin) on the renal tubule, lithium decreases the
kidney's maximal concentrating capacity, analogous to nephrogenic diabetes insipidus.
The obligatory water loss goes up, often to 3 liters or more per day. Patients must not
restrict their water intake, for they will be subject to dehydration with attendant lithium
intoxication and potential renal shutdown. Instead, they must drink as much fluid as it
takes to maintain hydration.
For some patients, once-a-day dosing of lithium can reduce polyuria. 60 For others,
a lower dose of lithium may help, but lower levels can bring reduced protection against
relapse. The use of diuretics (paradoxical though it sounds, with the mechanism poorly
understood) and a diet low in protein and salts are alternative treatment options. 61
(Some diuretics can markedly increase serum lithium levels. If a diuretic is used to
decrease urine volume, serum lithium levels should be closely monitored and the
lithium dose reestablished.)
ii. Structural Changes and Glomerular Filtration Alterations. In 1977 reports
began to appear of kidney structural damage in patients treated with lithium. The
incidence of this finding varied depending on patients studied, and so did the nature of
the lesions and relation to therapy. The changes described were mostly of a tuberoin-
terstitial nature. Although similar changes have been described in patients with bipolar
illness before treatment with lithium, they probably occur at a greater frequency in
patients who have been so treated. Mellerup et al. estimate a 10% incidence of struc-
tural changes in lithium-treated patients. 6o Unquestionably, lithium intoxication can
lead to acute functional and structural alterations.
Fortunately, the most vital kidney function, glomerular filtration, is minimally
impaired by long-term lithium treatment absent toxic levels. Even patients treated for
several decades appear to have negligible decreases in glomerular filtration rates be-
yond the decline expected with age. Only a few cases of renal failure have been
apparently linked to therapeutic lithium. 63
iii. Renal Function. Earlier recommendations for extensive testing of kidney
function-including 24-hr urine collections and concentrating capacity tests-have
proven unnecessary. 58 Instead, pretreatment assessment should include a history and

necessary laboratory testing to rule out renal pathology or, if present, to assess its
nature and degree. First, the clinician should complete personal and family medical
histories and review of systems. Laboratory tests should include serum creatinine and
blood urea nitrogen (BUN) determinations along with routine urinalysis.
Once lithium therapy has been initiated, a serum creatinine determination every 6
months should suffice. Beware, however, of progressive rises in serum creatinine, even
within the normal range, as these may herald a decline in glomerular filtration rate.
Similarly, a rise in serum lithium without an increase in dose may reflect decreasing
glomerular filtration. Patients on maintenance therapy should be questioned peri-
odically about changes in urine volume, and their families should be counseled about
the importance of adequate hydration-especially under conditions of heat, increased
physical activity, gastroenteritis, or other situations that may alter fluid and electrolyte
iv. Other Kidney Effects. Rare cases of nephrotic syndrome have been described
in patients treated with lithium. Although this condition is usually detected by asymp-
tomatic proteinuria on routine urinalysis, occasional patients develop edema. Almost
all cases have been reversible after discontinuing lithium, although at least one patient
required continued steroid therapy.64
Because lithium and sodium are handled similarly in renal reabsorption mecha-
nisms, sodium depletion can lead to increased lithium reabsorption and, consequently,
elevated blood levels. Thus, using salt-lowering diuretics or instituting a restricted-
sodium diet can raise blood lithium levels in a lithium-treated patient. Limited research
on strenuous exercise suggests that, contrary to initial expectations, blood levels of
lithium are more likely to decrease than to increase, presumably because more lithium
than sodium is lost in sweat.
c. Hematological. Mild to moderate elevations of the white blood cell count
(typically 12,000-15,000 cells/mm3 ) occur commonly with both acute and chronic
administration of lithium. 65 Neutrophils are most affected. Lithium leukocytosis is
clinically benign and reversible and can develop at various doses and at any stage of
d. Cardiovascular. Lithium has significant effects on the heart. However, ad-
verse reactions in clinical practice are rare, even in patients with known heart disease.
Perhaps because of its effects on potassium balance, lithium causes electrocardiogram
(EKG) T-wave flattening or inversion in 20% or more of individuals. 66 Sometimes with
normal doses, U waves also appear. Since these are common changes, the clinician
should obtain a baseline EKG prior to starting therapy.
Several types of conduction problems have appeared during lithium therapy,
including first-degree atrioventricular (A-V) block, irregular or slowed sinus node
rhythms (particularly in elderly patients), and increased numbers of ventricular prema-
ture contractions. 67 •68 Some individuals have also developed severe congestive heart
failure, cardiomyopathy, and ventricular tachycardia. However, the degree to which
lithium therapy contributed to the development of these disorders was unclear. During
toxic states, some patients have developed persistent ventricular tachycardia, atrial
fibrillation, advanced A-V block, and vascular collapse. 69

Because Cardiac Reactions May Produce Serious Morbidity in

an Occasional Patient:
1. Take a careful cardiac history.
2. Obtain a baseline EKG.
3. Follow the patient's cardiac status at least by monitoring clinical signs and
4. Fully evaluate any conduction abnonnalities that occur during therapy.
5. Consider lowering dosage in patients with conduction changes if continued
lithium therapy is a high priority. Similarly, in individuals with irregular or
slowed sinus rhythm, consider a pacemaker if other measures fail to provide
cardiac stability.
6. Discontinue lithium in those who develop heart failure, ventricular tachycar-
dia, or cardiomyopathy.

e. Cutaneous. Lithium occasionally causes bothersome skin problems. A pruri-

tic, maculopapular rash may appear during the first month of therapy. This allergic
response has a variable course that sometimes progresses to a serious dennatitis. Other
observed dennatological effects include acneifonn lesions, hyperkeratotic papules,
xerosis cutis, cutaneous ulcers, thinning and drying of scalp hair, exacerbation or
appearance of psoriasis, chronic folliculitis, and anesthesia of the skin. The usual
dennatologic treatments for these conditions can be used when they are caused or
exacerbated by lithium. Many of these reactions respond to conservative treatment,
remit over time, or may not reappear if lithium is stopped and then restarted. Only the
more serious responses require the discontinuation of lithium.

f. Gastrointestinal. Gastrointestinal (GI) reactions occur frequently during initial

lithium therapy or as a result of dosage increase and include gastric irritation, anorex-
ia, abdominal cramps, nausea, vomiting, and diarrhea. Gastrointestinal symptoms
sometimes herald impending toxicity. Individuals, however, vary greatly in their sensi-
tivity to changes in dosage or preparation and in the persistence of their symptoms.
To Cope with Gastrointestinal Reactions
1. Check the serum lithium level.
2. Treat gastric irritation early in lithium therapy by administering lithium after
meals, lowering the dose, further dividing dosages, or using a sustained-
release preparation.
3. Treat other GI symptoms by changing or further dividing the dose or switch-
ing to another preparation (tablets, capsules, elixir, or sustained release).
4. If a patient develops a viral gastroenteritis, monitor lithium blood levels
closely and be prepared to discontinue lithium temporarily.

g. Central Nervous System and Neuromuscular. Lithium commonly produces

CNS and neuromuscular effects at therapeutic doses. The most frequent reactions occur
at the beginning of treatment and include mental dullness, decreased memory and
concentration, headache, fatigue and lethargy, muscle weakness, and tremor.

These symptoms most often remit quickly and usually do not require changing the
A fine rapid hand tremor accompanies lithium therapy in 15-70% of patients,
depending on dose. It is an action or postural tremor (made worse by sustained posture
or purposeful movements) and may persist from the beginning of treatment, appear at
any point, or disappear and then recur.70 However, it usually appears when therapy
starts and decreases over time. CNS stimulants (including caffeine), anxiety, muscle
tension, and, occasionally, antiparkinson drugs worsen the tremor, whereas sedative-
hypnotic agents may improve it. Patients with other forms of action tremor (i.e.,
familial, idiopathic, senile) develop lithium-induced tremors more frequently. Rarely,
tremulousness may involve the upper extremities, face, or eyelids. Lowering the dose
of lithium can provide relief. However, if the tremor still persist, propranolol (Inde-
ral. and others), 20-160 mg per day, may help.71 Other remedies that have been
studied include other l3-blockers, primidone, safflower oil or evening primrose oil,
inositol, and potassium, but so far results are inconclusive. 70
Lithium also causes various other troublesome effects, including dysarthria, ver-
tigo, ataxia, tinnitus, nystagmus, autonomic slowing of bladder and bowel function,
visual distortion, muscle irritability (e.g., twitching, fasciculations, facial spasm, in-
creased tendon reflexes, clonus, and choreoathetosis), and even an organic brain syn-
drome. When any such signs appear abruptly in the course of lithium therapy, they may
signal toxicity and should prompt a blood level assay.
Extrapyramidal reactions, although sometimes occurring without any apparent
predisposition, are more likely in the elderly, those taking antipsychotic medication,
and those with toxic blood levels of lithium. Cogwheeling and generalized muscle
rigidity are more common signs.72 Sometimes these reactions appear during the first
months of therapy at therapeutic levels of lithium. However, they more often accom-
pany serum lithium levels greater than 1.5 mEq/L. With levels above 3.0 mEq/L,
patients usually show profound neurotoxic effects that can progress to seizures and
incontinence, stupor, coma, brain damage, and death.
In addition to the adverse reactions already mentioned, rare cases of brain damage
have been reported with usual (i.e., nontoxic) doses of lithium (see Chapter 4). This
neurotoxicity occurs most often in patients who are elderly, schizophrenic, or already
neurologically impaired.
To Minimize CNS Effects
1. Take an appropriate history for schizophrenia, possible CNS disease, tremors,
and antipsychotic medication usage; these will dictate greater caution in
administering lithium.
2. Note and carefully follow any mild, initial neurological signs (i.e., mental
dullness, poor concentration, weakness, lethargy, or tremors).
3. Discontinue lithium and obtain a serum level in patients who develop any
significant neurological symptoms (particularly an organic brain syndrome).
4. Treat hand tremors by reassuring the patient or administering 20-160 mg of
propranolol per day. Failing those, consider lowering the dose.

h. Ocular. Lithium therapy rarely causes adverse reactions involving the eyes.
Sometimes tearing, itching, burning, or blurring may occur during the first few weeks
of treatment. With decongestant eyedrops and time, most of these reactions abate.
Occasionally, more significant reactions develop, including exophthalmos and worsen-
ing of cataracts, and eight cases of bilateral papilledema have been reported. 73 - 77
However, the role of lithium in these disorders remains unclear.

i. Weight Gain. Reports about weight gain during lithium therapy vary widely.
Although the etiology of this effect is unclear, many causes have been hypothesized,
including altered carbohydrate and lipid metabolism, increased fluid intake and reten-
tion, improved appetite with the resolution of affective episodes, diminished thyroid
functioning, and increased intake of high-caloric fluids. Since this adverse reaction is
frequent, patients should be warned and instructed about possible dietary restrictions.
For those who gain weight, limiting calories usually results in appropriate weight loss.
Thirst should be quenched with no-calorie beverages.

j. Other Adverse Reactions. These include a persisting metallic taste and pre-
tibial edema.

3. Drug Interactions
Lithium and other drugs, particularly the diuretics, have various synergistic effects
(see Table 3). Thiazide diuretics commonly increase lithium levels by decreasing
clearance. This reaction can occur quickly, resulting in significant increases in blood
levels and potential toxicity. Some other diuretics, notably osmotic drugs and carbonic
anhydrase inhibitors, decrease lithium levels by increasing excretion.
Antipsychotic agents are commonly prescribed with lithium for acutely disor-
ganized patients. This practice is usually safe. However, some individuals have devel-
oped acute neurotoxicity and/or permanent brain damage (described most often with
haloperidol). These reactions occur rarely and may reflect lithium toxicity only or a
neuroleptic malignant syndrome (see Chapter 4). Whether or not a combined toxic
effect occurs is still unclear. See Table 3 for a list of other medications that, when
combined with lithium, can affect lithium levels or increase adverse reactions.
As mentioned earlier, lithium should be discontinued when bipolar patients are
treated with ECT to avoid a possible increase in confusional states.
To Minimize Possible Drug Synergisms
1. Rationalize polypharmacy.78
2. Lower the dose of lithium in patients starting on drugs known to elevate blood
lithium levels.
3. Monitor serum lithium levels more often in patients on osmotic diuretics or a
carbonic anhydrase inhibitor.
4. Administer furosemide (Lasix@ and others) as a diuretic rather than others if it
fits the patient's medical needs. (This drug may not affect blood lithium

TABLE 3. Drug Interactions with Lithium

Increase lithium levels:

Alprazolam (Xanax<BI)
Amiloride (Midamor<Bi and others)
Angiotensin-converting enzyme (ACE) inhibitors
Antipsychotic agents?
Ethacrynic acid (Edecrin<BI)
Fiuoxetine (Prozac<BI)
Indapamide (Lozol<BI)
Mefenamic acid (Ponstel<BI)
Metronidazole (Fiagyl<BI and others)
Nonsteroidal anti-inflammatory drugs (NSAIDs)a:
Carprofen (Rimadyl<BI)
DicJofenac (Cataflam<BI, Voltaren<BI, Voltaren Ophthalmic<BI)
Diflunisal (Dolobid<BI)
Etodolac (Lodine<BI, Ultradol<BI)
Fenoprofen Calcium (Nalfon<BI)
Fiurbiprofen (Ansaid<BI)
Ibuprofen (Advil<BI, Motrin<BI, and others)
Indomethacin (Indocin<BI and others)
Ketoprofen (Orudis<BI, Oruvail<BI)
Ketorolac Tromethamine (Acular<Bl, Toradol<BI)
Meclofenamate Sodium (MecJomen<BI)
Mefenamic Acid (Ponstel<BI)
Mesalamine (Asacol<BI, Pentasa<BI, Rowasa<BI)
Nabumetone (Relafen<BI)
Naproxen (EC-Naprosyn<BI, Naprosyn<BI)
Naproxen Sodium (Anaprox<BI, Anaprox DS<BI)
Oxaprozin (Daypro<BI)
Phenylbutazone (Azolid<BI, Butatab®, Butazolidin®, Butazone-lOO<BI)
Piroxicam (Feldene®, Lampoflex<BI)
Sulindac (Clinoril<BI)
Tolmetin Sodium (Tolectin<BI)
Phenylbutazone (Butazolidin® and others)
Potassium-retaining diuretics
Some antibiotics
Spironolactone (Aldactone<BI and others)
Sulindac (Clinoril®)
Thiazide diuretics
Triamterene (Dyrenium® and others)
Zomepirac (Zomax<BI)
Decrease lithium levels:
Aminophylline (Aminophyllin<BI and others)
Carbonic anhydrase inhibitors
Dyphylline (Lufyllin<BI, Dilor<Bi)
Osmotic diuretics
Oxtriphylline (Choledyl<BI)
Sodium-containing medications
Theobromine (Athenol<BI and others)
Theophylline (Tedral<BI and others)

TABLE 3. (Continued)

Adverse reactions:
Antipsychotics: extrapyramidal symptoms, neurotoxicity, somnambulism
Atracurium (Tracium®): prolonged neuromuscular blocking effects
Hydroxyzine (Atarax®, Vistaril®, and others): cardiovascular toxicity
Iodine: antithyroid effects
Methyldopa (Aldomet® and others): hypertension, toxic symptoms at normal blood levels
Pancuronium (Pavulon®): prolonged neuromuscular blocking effects
Succinylcholine (Anectine® and others): prolonged neuromuscular blocking effects
Verapamil (Calan® and others): neurotoxicity and bradycardia

"Many of these are now available without a prescription.

5. Discontinue lithium and obtain a blood level in any patient who develops
signs of an organic brain syndrome while taking antipsychotic medication.
6. Avoid abruptly discontinuing caffeinated beverages (e.g., coffee, tea, colas)
and monitor blood levels more closely in patients who markedly change their
consumption of caffeine.
7. Avoid the use of methyldopa and hydroxyzine with lithium.
8. Warn patients about taking nonsteroidal anti-inflammatory drugs (NSAIDS;
see Table 3 for a list of NSAIDs).

4. Lithium Toxicity

Toxic blood levels of lithium can occur acutely by accidental or intentional over-
dose or chronically through excessive dosing, decreased renal function, drug interac-
tions, diminished salt intake, or other fluid and electrolyte abnormalities. Some patients
will develop toxic symptoms and signs at doses usually considered therapeutic, where-
as other patients will tolerate fairly high plasma concentrations without ill effect. Most
probably, these individual variations reflect differences in brain levels achieved at a
given serum concentration, but they also may reflect differences in end-organ sensi-
Patients should not have serum levels above 1.5 mEq/L, and certainly alarm bells
must sound at 2.0 rnEq/L and above. Twelve-hour levels above 3.0 mEq/L are an
emergency and should be dealt with immediately.
Gastrointestinal symptoms may initially appear, followed or accompanied by CNS
depression. This may include somnolence, sluggishness, the various hallmarks of an
organic brain syndrome, dysarthria, seizures, choreoathetoid movements, increased
muscle tone, and increased deep tendon reflexes. Cardiovascular collapse marked by
lowered blood pressure, irregular cardiac rhythm, decreased urine output, and conduc-
tion abnormalities (and EKG changes) can be life-threatening.
Acute intoxication causes significant CNS depression. Patients typically develop
pyramidal tract signs and impaired consciousness or coma. Individuals who chronically
take too much medication develop CNS impairment more gradually. Sluggishness and
drowsiness may progress over a period of days. Often, gastrointestinal symptoms,

slurred speech, ataxia, and coarse trernor accornpany these changes. If the initial signs
of chronic intoxication are overlooked, a rnore florid CNS syndrorne rnay develop,
rnost often rnanifested by hyperpyrexia and stupor or corna. It can also include neuro-
logical asymmetries, nystagrnus, stiff neck, and hyperextension of the extrernities.
The possibility of lasting neurological irnpairment as a result of lithiurn intoxica-
tion probably is a product of the rnaxirnal blood concentration and the duration of toxic
levels. Various permanent neurological sequelae are possible, rnost commonly involv-
ing cerebellar and cognitive dysfunctions,?9
Lithiurn toxicity is generally rnanaged by supportive rneasures. If the toxicity
occurs as part of an acute rnedication regirnen or rninor overdose, blood levels never
exceed 3.0 rnEq/L, and kidney function is intact, careful observation usually suffices.
Since lithiurn is excreted rapidly, the syndrorne rnost often abates within a few days.
However, in a large or chronic overdose, very large total-body lithiurn stores rnay
accurnulate. In these cases, the patient often suffers persistent (several days or longer)
life-threatening CNS depression and cardiovascular irnpairment.

Reasons for Obtaining a Lithium Level

1. Syrnptorns and signs of toxicity, including nausea, vorniting, diarrhea, in-

creasing fatigue, rnental dullness, trernor, or ataxia.
2. Signs of rnania, including euphoria, irritability, hyperactivity, inappropriate
actions, inability to cornplete tasks, persistent insomnia.
3. Syrnptorns and signs of depression, including lowered rnood, self-deprecat-
ing behavior, guilt, decreased activity, insornnia, weight loss, poor appetite,
suicidal thoughts.
4. Changes in dose initiated by the clinician or patient. (Obtain level about 5
days after dosage is changed to allow steady state.)
5. The developrnent of a rnedical disease, particularly those that cause fever or
6. Significant increases in sweating (e.g., rnove or visit to a warmer clirnate or
rnarked increase in exercise).
7. The institution of steroids, diuretics, antipsychotic drugs, sodiurn bicarbo-
nate, angiotensin-converting enzyrne (ACE) inhibitors, or nonsteroidal anti-
inflamrnatory drugs.
8. Question of pregnancy.
9. Change in salt intake or diet.
10. Signs of thyroid deficiency.
11. Routine rnonitoring.

Management of Serious Toxic States

1. Rapidly assess (including clinical signs and syrnptorns, serurn lithiurn levels,
electrolytes, and EKG), rnonitor vital signs, and rnake an accurate diagnosis.
2. Discontinue lithiurn.
3. Support vital functions and rnonitor cardiac status.

4. Limit absorption
a. If the patient is alert, provide an emetic.
b. If the patient is obtunded, intubate and suction nasogastrically (prolonged
suction may be helpful, since lithium levels in gastric fluid may remain
high for days).
5. Prevent infection in comatose patients by body rotation and pulmonary toilet.
6. In all cases, vigorously hydrate (ideally, 5-6 liters per day); monitor and
balance the electrolytes.
7. In moderately severe cases
a. Implement osmotic diuresis with urea, 20 g IV 2 to 5 times per day, or
mannitol, 50-100 g IV per day.
b. Increase lithium clearance with aminophylline, 0.5 g up to every 6 hr, and
alkalinize the urine with IV sodium lactate.
c. Ensure adequate intake of NaCI to promote excretion of lithium.
8. Implement hemodialysis in the most severe cases. These are characterized by
a. Serum lithium levels between 2.0 and 4.0 mEq/L with severe clinical
signs and symptoms (particularly decreasing urinary output and deepen-
ing CNS depression).
b. Serum lithium levels greater than 4.0 mEq/L. Most patients completely
recover from lithium toxicity; several may die; some develop permanent
neurological damage.

5. Pregnancy

Various congenital abnormalities have been reported in babies exposed to lithium

in utero, particularly anomalies of the heart and great vessels-most commonly Ebs-
tein's anomaly. However, the risk of these occurring now appears to be much less than
was originally suspected, i.e., lower than 0.1 %.80 Ideally, lithium (and any other
psychotropic medication) should be avoided during pregnancy, especially during the
first trimester, but the risk of congenital malformations should be weighed against the
risks of untreated episodes of bipolar disorder (see Chapter 10). If a woman does
decide to take lithium during pregnancy, a higher dose will be necessary to maintain
the same blood levels. Conversely, the dose should be lowered postpartum. The anti-
convulsants valproate and carbamazepine appear more teratogenic then lithium, but
ECT is a relatively safe alternative. Infants born to mothers taking lithium occasionally
develop temporary adverse reactions postpartum. Also, lithium's concentration in
breast milk can cause side effects in nursing infants. Long-term effects of lithium on
the brain of fetuses and neonates who have been exposed are unknown.

6. Preparations and Dosage

In the United States, lithium comes mostly as a carbonate salt in oral preparations
of 300 mg (8.12 mEq). This dosage and composition usually meet most clinical needs.
However, a few patients exhibit sensitivities to various preparations and can benefit
from alternate forms, such as slow-release tablets or lithium citrate syrup.

In Individuals Who Develop Adverse Reactions

1. Use a once-a-day regimen (or even an alternate-day regimen in elderly pa-
2. Further divide doses by administering portions of scored tablets.
3. Switch between tablets and capsules (or vice versa).
4. Administer lithium citrate syrup (to try a different salt) or give smaller doses.
5. Give slow-release tablets to minimize serum level fluctuations.

7. Other Possible Indications for Lithium

The U.S. Food and Drug Administration accepts acute mania and the prophylaxis
of recurrent episodes of mania as indications for prescribing lithium. The scientific
literature also supports a role for lithium in attenuating recurrent episodes of depression
among bipolar patients and perhaps among some unipolar patients as well. Lithium is
the best-studied adjunct to antidepressants in treatment-resistant patients during acute
depressive episodes, and its efficacy is now generally accepted. 81
Beyond the treatment of primary disorders of mood, data supporting a role for
lithium in other conditions become thin, yet the number of applications proposed in the
literature and used in practice is extremely broad. For example, some patients who
meet criteria for schizophrenia appear to benefit from lithium therapy, most often as an
adjunct to neuroleptics. 82 Unfortunately, clinical criteria are not particularly helpful in
predicting who among a schizophrenic population is likely to respond to lithium, so
lithium trials are most commonly conducted among treatment-resistant schizophrenic
patients, as well as in those with suggestions of a mood disorder (or schizoaffective
disorder) in the patient or family. Likewise, aggressive or cyclic behavior among
patients with frank brain damage or mental retardation sometimes improves on treat-
ment with lithium, but, again, empirical trials in individual cases are the only sound
way to know if this approach will be effective.
Other periodic disorders with an affective component that may be related to manic-
depressive illness (e.g., schizoaffective disorder) can sometimes be treated effectively
with lithium. Less clearly defined periodic disturbances are also occasionally responsive
to lithium and include periodic catatonia,83 periodic alcoholism (particuarly accom-
panied by depressive affect),84 emotionally unstable character disorders,85 cyclothymic
personality disorders,86 and obsessive-compulsive states. 87

B. Valproate
Valproic acid, an eight-carbon, branched-chain fatty acid, was synthesized in 1882
and recognized to have antiseizure activity in 1963. Valproate's efficacy in stabilizing
mood disturbances was first reported in 1966, and in 1995, divalproex sodium, an
enteric-coated, stable, equimolar combination of sodium valproate and valproic acid,
was approved by the U.S. Food and Drug Administration (FDA) for treating acutely
manic bipolar patients. 88
Valproate's antimanic mechanism of action has yet to be established but may be
related to its increase of 'Y-aminobutyric acid (GABA) concentrations, by decreasing
either GABA's metabolism or its reuptake in brain tissues. Another theory is that

valproate acts on postsynaptic receptor sites to mimic or enhance GABA's inhibitory

action. Valproate's effect on the neuronal membrane is not completely understood.
Some studies suggest a possible direct effect on membrane activity related to changes
in potassium conduction.

1. Pharmacokinetics
All valproate preparations are rapidly absorbed after oral administration, reaching
peak plasma levels within 2-4 hours of ingestion (see Table 1). Food may delay
absorption but does not affect bioavailability. Valproate is rapidly distributed and
reaches the CNS within minutes of administration. It is highly bound (90-95%) to
plasma proteins at serum concentrations up to 50 j.Lg/mL. As the concentration in-
creases from 50 to 100 j.Lg/mL, the percentage bound decreases to 80-85% and the free
fraction becomes progressively larger, thus increasing the concentration gradient into
the brain and also increasing its metabolism. Valproate's half-life ranges from 5 to 16
hr, depending on whether it is taken alone or with other medications, and it takes 1-4
days to attain steady state. The half-life may be considerably longer in patients with
impaired hepatic function, in the elderly, and in children up to 18 months of age. It may
be much shorter in patients concomitantly taking hepatic enzyme-inducing anticonvul-
Valproate is primarily metabolized in the liver through a variety of conjugative
and oxidative processes and produces many metabolites, some of which may have
pharmacologic or toxic activity. The increased free serum fraction at higher blood
levels can lead to faster metabolism. Therefore, some patients require a greater dosage
increment to raise a blood level from 100 to 125 j.Lg/mL, for example, than to go from
75 to 100 j.Lg/mL. The rate of metabolism is faster in children and in patients concur-
rently using enzyme-inducing medications, such as phenytoin, phenobarbital, pri-
midone, and carbamazepine. Val pro ate is eliminated by the kidneys, mainly as the
glucuronide conjugate.

2. Adverse Reactions

A summary of adverse reactions to valproate is presented in Table 4.

a. Gastrointestinal. The most commonly reported side effects at the start of

valproate therapy are nausea, vomiting, and indigestion. These effects usually diminish
over time or with a reduction in dose and rarely require discontinuation of therapy.
(Some patients have less nausea with a once-a-day dose of valproate at bedtime; for
others, once-daily dosing increases gastrointestinal symptoms.) Some patients experi-
ence diarrhea, abdominal cramps, and constipation. The administration of a hista-
mine-2 antagonist such as famotidine (Pepcid®) or cimetidine (Tagamet®), which are
now available over the counter, or the antiulcer medication sucralfate (Carafate®) may
alleviate persistent gastrointestinal problems. Anorexia with weight loss and increased
appetite with weight gain have both been reported. Weight gain can be aggravated by
the coadministration of other medications with the same side effect, such as lithium,
antipsychotics, and some antidepressants.

TABLE 4. Valproate-Adverse Effects

Common Rare (cont.)

Alopecia Eosinophilia
Gastrointestinal upset Erythema multiforme
Sedation Galactorrhea
Tremor Hallucinations
Weight loss Headache
Weight gain Hepatoxicity
Rare Hypofibrinogenemia
Acute pancreatitis Hyponatremia
Anemia (including macrocytic with or Incoordination
without folate deficiency) Irregular menses
Asterixis Leukopenia
Ataxia Macrocytosis
Bone marrow suppression Nystagmus
Breast enlargement Parotid gland swelling
Coma Photosensitivity
Dermatitis Porphyria, acute intermittent
Diplopia Pruritus
Dizziness Relative lymphocytosis
Dysarthria Secondary amenorrhea
Edema of the extremities Stevens-Johnson syndrome
Encephalopathy with fever SIADH
Enuresis Thrombocytopenia
Thyroid function test abnormalities

b. Central Nervous System EtTects. Sedation has occurred in patients receiving

valproate alone but occurs most often in patients receiving combination therapy. Other
eNS effects include hallucinations, ataxia, headache, nystagmus, diplopia, asterixis,
"spots before eyes," dysarthria, dizziness, and incoordination. An essential-type tremor
is a common neurologic adverse reaction to valproate and may be dose and blood-level
related. If valproate is administered together with lithium, which can produce the same
type of tremor, this may be more of a problem. Rare cases of coma have been reported
in patients receiving valproate alone or in combination with phenobarbital. Encepha-
lopathy with fever has occurred in occasional patients just after the introduction of
valproate monotherapy, without evidence of hepatic dysfunction or inappropriate plas-
ma levels; all patients recovered after the drug was withdrawn.

c. Dermatological. Some patients develop alopecia, thinning, or changes in hair

texture while taking valproate. Since lithium also can affect skin and hair, there is an
additive potential. Other dermatological effects include skin rash, photosensitivity,
generalized pruritus, erythema multiforme, and Stevens-Johnson syndrome.

d. Hematological. Thrombocytopenia and inhibition of the secondary phase of

platelet aggregation may be reflected in altered bleeding time, petechiae, bruising,
hematoma formation, and frank hemorrhage. Platelet count usually recovers with a
dosage decrease, but the occurrence of thrombocytopenia or leUkopenia may necessi-

tate the discontinuation of valproate. Evidence of hemorrhage, bruising, or a disorder of

hemostasis/coagulation would also necessitate either a reduction of dosage or with-
drawal of therapy. Other hematological problems associated with valproate therapy
include relative lymphocytosis, macrocytosis, hypofibrinogenernia, leukopenia, eo-
sinophilia, anemia (including macrocytic with or without folate deficiency), bone mar-
row suppression, and acute intermittent porphyria.

e. Hepatic. Serum hepatic transaminase elevations are common, dose related, and
usually self-limiting and benign. Occasionally, laboratory test results indicate increases
in serum bilirubin and abnormal changes in other liver function tests. These results may
reflect potentially serious hepatotoxicity. Serious or fatal hepatotoxicity may be pre-
ceded by nonspecific symptoms such as malaise, weakness, lethargy, facial edema,
anorexia, and vomiting. Patients on multiple anticonvulsants, children, those with
congenital metabolic disorders, and those with organic brain disease may be at particu-
lar risk. Fatal hepatotoxicity is extremely rare and usually develops within the first 6
months of valproate therapy. Valproate should be discontinued immediately in the
presence of suspected or apparent significant hepatic dysfunction or if hyperam-
monemia occurs, with or without lethargy or coma.

f. Other. Patients on valproate therapy have also reported irregular menses, sec-
ondary amenorrhea, breast enlargement, galactorrhea, parotid gland swelling, abnormal
thyroid function tests, acute pancreatitis (including rare fatalities), hyponatremia, inap-
propriate ADH secretion, edema of the extremities, and enuresis. Studies of women
taking valproate for epilepsy show an increase in reproductive endocrine disorders,
especially in those who began taking valproate before age 20 years. 4

Drug Interactions
Valproate is metabolized in part by the liver's cytochrome P4502D6 isoenzyme
system, but this is a relatively minor secondary metabolic pathway compared to glu-
curonidation and [3-oxidation. Drugs that inhibit the cytochrome P450 isoenzymes,
therefore, may have little effect on valproate clearance. Drugs that induce hepatic
enzymes, on the other hand, particularly those that elevate levels of glucuronosyl
transferases, may increase the clearance of val pro ate and thus decrease plasma levels.
For example, phenytoin, carbamazepine, and phenobarbital (or primidone) can double
the clearance of valproate, thereby diminishing its efficacy. Serum valproate concentra-
tions should be monitored as a guide to dosage when valproate is prescribed with
carbamazepine or any other enzyme-inducing drug, especially when one of these
medications is added to or withdrawn from an existing regimen.
Although valproate does not induce hepatic metabolism in general, it does appear
to inhibit drug oxidation and may increase serum levels of concomitantly administered
drugs that are oxidatively metabolized, such as phenobarbital, phenytoin, and tricyclic
antidepressants. See Table 5 for a list of other drugs that interact with valproate.

TABLE 5. Valproate-Drug Interactions

Valproate increases levels of the following drugs:

Diazepam (Valium® and others)
Ethosuximide (Zarontin®)
Lamotrigine (Lamictal®)
Tolbutamide (Orinase®)
Tricyclic antidepressants
Zidovudine (Retrovir®)
The following drugs increase valproate levels:
Felbamate (Felbatol®)
The following drugs decrease valproate levels:
Carbamazepine (Tegretol® and others)
Mefloquine (Lariam®)
Rifampin (Rifadin® and others)
Valproate increases the CNS-depressant effects of the following drugs:
Clozapine (Clozaril®)
Haloperidol (Haldol®)
Loxapine (Loxitane®)
Maprotiline (Ludiomil®)
Monoamine oxidase inhibitors
Tricyclic antidepressants
There is an increased risk of hemorrhage when valproate is combined with:
Anticoagulant therapy
Anti-inflammatory analgesics
Sulfinpyrazone (Anturane)
Thrombolytic therapy

4. Overdose
Overdosage with valproate may result in a deep coma. Since valproic acid is
absorbed very rapidly, gastric lavage or emesis will be more effective the sooner after
ingestion it is used. General supportive measures should be applied with particular
attention to maintaining adequate urinary output. Naloxone has been reported to re-
verse the eNS depressant effects of valproate overdosage.

5. Pregnancy
Valproate should not be used in pregnant women. It crosses the placenta and has
been reported to cause teratogenic effects, including neural tube defects, in the fetus.
The risk of spina bifida in children born to women who took valproate during the first
trimester of pregnancy is 1-2%, and other congenital anomalies have also been report-
ed. When valproate is taken later in pregnancy, it may affect the development of the
central nervous system. Even though these data are from patients with epilepsy rather
than bipolar disorder, caution dictates the same avoidance. Valproate also should be
avoided in nursing mothers, as it is excreted in breast milk in concentrations of 1-10%
of the maternal serum concentration.

6. Preparations and Dosage

Valproate is available in the United States as divalproex sodium or valproic acid.
Divalproex sodium comes in an enteric coated sprinkle capsule of 125 mg or enteric
coated tablets of 125, 250, and 500 mg. Valproic acid is available in 250-mg elastic
capsules and a 250-mg/5-mL syrup.

C. Carbamazepine
Carbamazepine is a dibenzazepine derivative structurally related to the tricyclic
antidepressants. Although some double-blind studies have suggested that carbamaz-
epine is effective in treating acute mania, the methods in most of these studies have
been criticized. Large, carefully designed, prospective, controlled trials would be
needed to establish carbamazepine's efficacy and safety, its specific indications, and its
range of side effects. There is also some evidence for carbamazepine's efficacy as a
long-term treatment for preventing recurrences in bipolar disorder, but, again, better
studies are needed to elucidate its role for this indication.
The exact mechanism by which carbamazepine might alleviate mania in bipolar
disorder is unknown. Its antimanic properties could be related to its anticonvulsant or
antineuralgic effects or to its effects on one of the neurotransmitter or second-messen-
ger systems. Carbamazepine increases acetylcholine in the striatum, decreases dopa-
mine turnover, decreases the release of norepinephrine and cerebrospinal nor-
epinephrine, decreases the activity of adenylate and guanylate cyclase, and decreases
GABA turnover; any or all of these actions could be relevant to carbamazepine's
effects in mania.

1. Pharmacokinetics
The pharmacokinetics of carbamazepine are summarized in Table 1. Carbamaz-
epine is almost completely absorbed from the gastrointestinal tract, but its absorption is
slow and variable. Protein binding is moderate: 55-59% in children and 76% in adults.
Peak plasma levels occur within 4-6 hr after ingestion of tablets and within 1.5 hr
following chronic administration of the suspension. Bioavailability is estimated at 85%
but may be less when the drug is taken with meals. Carbamazepine's half-life ranges

from 25 to 65 hr after an initial single dose. With chronic dosing, it may decrease to 8-
29 hr (average 12-17) because of autoinduction of metabolism (see below). Its active
metabolite, carbamazepine-10,11-epoxide, has a half-life of 5-8 hr. Carbamazepine is
eliminated primarily through the kidneys.
Like valproate, carbamazepine is metabolized by the hepatic P4502D6 isoenzyme.
It causes an induction of the P450 enzymes, resulting, in some patients, in an increase
in the rate of its own metabolism over time as well as that of other drugs metabolized
by the P450 system. This often results in having to raise the dose after 2-4 months of
treatment. Steady state may be attained within 4-5 days or, in patients in whom
clearance increases as a result of autoinduction, steady state may not be achieved until
3-4 weeks after treatment initiation.

2. Adverse Reactions
Adverse reactions to carbamazepine are listed in Table 6. The most common side
effects of carbamazepine in the first few weeks of treatment are drowsiness, dizziness,

TABLE 6. Carbamazepine-Adverse Effects

Common (diminish over time or with temporary reduction in dose)

Blurred vision
Less common
Cardiovascular complications
Gastrointestinal upset
Skin reactions (if severe, may require discontinuation of carbamazepine)
Cognitive impairment
Genitourinary effects
Increased intraocular pressure
Jaundice, cholestatic and hepatocellular
Liver function test abnormalities
Renal damage leading to oliguria and hypertension
Transient leukopenia (carbamazepine may be continued unless infection develops)
Water intoxication
Very rare
Aplastic anemia
Lupus erythematosus-like syndrome
Pulmonary hypersensitivity

ataxia, diplopia, nausea, blurred vision, and fatigue. These tend to diminish over time
or to respond to a temporary reduction in dose. Less common reactions include gastro-
intestinal upset, hyponatremia, and a variety of skin reactions, some of which are
severe enough to require discontinuation of carbamazepine. About 10% of patients
experience transient leukopenia, but, unless infection develops, carbamazepine may be
continued. More serious hematopoietic reactions, including aplastic anemia and
agranulocytosis, are very rare. Gradual buildup of carbamazepine levels and smaller,
more frequent dosing can help to minimize side effects, particularly in patients con-
comitantly taking other drugs such as lithium. The most severe adverse reactions have
been observed in the hematopoietic system, the skin, and the cardiovascular system.

a. Hematological. Carbamazepine usually produces small decreases in erythro-

cyte, leukocyte, and platelet counts. This routine neutropenia is not correlated with
agranulocytosis. Severe or fatal aplastic anemia, agranulocytosis, or other blood dys-
crasias are rare, appearing in approximately lout of 125,000 cases. Early cases of
agranulocytosis may have been caused by impurities, which have been reduced as
manufacturing techniques have improved. Concomitant use of myelosuppressive
agents is, however, not recommended. Other hematological side effects that have been
reported include bone marrow depression, leukocytosis, eosinophilia, and acute inter-
mittent porphyria.

h. Dermatological. Approximately 5-15% of patients taking carbamazepine de-

velop a skin rash. Minor, nonspecific rashes may be observed or treated with anti-
histamines, but carbamazepine should be stopped immediately if a patient develops
urticaria, exfoliative dermatitis, or toxic epidermal necrolysis. Stevens-Johnson syn-
drome, lupuslike reaction, or any skin problem associated with fever, chills, arthralgia,
myalgia, lymphadenopathy, and eosinophilia may indicate impending bone marrow
suppression. The drug should be stopped and the patient closely monitored if these
problems appear. Patients taking carbamazepine may also have an increased sensitivity
to sunlight.

c. Cardiovascular. Although cardiovascular side effects of carbamazepine are

uncommon, there have been reports of congestive heart failure, edema, aggravation of
hypertension, hypotension, syncope and collapse, aggravation of coronary artery dis-
ease, arrhythmias and A-V block, primary thrombophlebitis, recurrence of throm-
bophlebitis, and adenopathy or lymphadenopathy. Some of these cardiovascular com-
plications have resulted in fatalities. Carbamazepine's potential to induce cardiac
conduction abnormalities may warrant baseline and periodic EKGs in some patients.

d. Hepatic. Patients taking carbamazepine have experienced abnormalities in

liver function tests, cholestatic and hepatocellular jaundice, and hepatitis.

e. Neurological. Sedation, dizziness, and ataxia may occur at the start of car-
bamazepine therapy unless starting doses are low and the dose is raised slowly. Fatigue,
headache, dizziness, ataxia, and diplopia may appear during maintenance therapy in
some patients. There may be little to no correlation between these problems and dose or

blood level. Cognitive slowing, perhaps secondary to sedation, may occur in some
patients. The cognitive impairments produced by carbamazepine are usually less than
those produced by the barbiturates or phenytoin.
Other neurological effects that have been associated with carbamazepine treat-
ment include: confusion, blurred vision, visual hallucinations, oculomotor distur-
bances, nystagmus, speech disturbances, abnormal involuntary movements, peripheral
neuritis and paresthesias, depression with agitation, talkativeness, tinnitus, and hyper-
acusis. There have been reports of associated paralysis and other symptoms of cerebral
arterial insufficiency, but the exact relationship of these reactions to carbamazepine has
not been established.

f. Gastrointestinal. Gastrointestinal side effects of carbamazepine include nau-

sea, vomiting, gastric distress and abdominal pain, diarrhea, constipation, anorexia, and
dryness of the mouth and pharynx, including glossitis and stomatitis. Nausea is some-
times a problem at the initiation of carbamazepine therapy, but it may not appear until
higher blood levels have been reached. Hepatitis is rare but if symptoms appear, serum
glutamic oxalacetic transaminase (SGOT), alkaline phosphatase, and bilirubin levels
should be tested.

g. Ocular. Increased intraocular pressure may occur with carbamazepine treat-

ment, so patients with glaucoma should have an ophthalmological exam before and
during treatment. Scattered punctate cortical lens opacities, as well as conjunctivitis,
have been reported.

h. Metabolism. The syndrome of inappropriate secretion of antidiuretic hormone

(SIADH), fever, chills, and frank water intoxication, with decreased serum sodium
(hyponatremia) and confusion, have been reported in association with carbamazepine

i. Renal. Carbamazepine has a vasopressinlike effect on the kidney. This may

cause the serum sodium to fall to low or slightly subnormal values in many patients,
but this rarely causes clinical difficulty. In rare cases, renal damage leading to oliguria
and hypertension has been reported after carbamazepine use.

j. Genitourinary. Genitourinary effects in patients taking carbamazepine include

urinary frequency, acute urinary retention, oliguria with elevated blood pressure,
azotemia, renal failure, and impotence. Albuminuria, elevated BUN, and microscopic
deposits in the urine have also been reported.

k. Other. Isolated cases of lupus erythematosus-like syndrome and pulmonary

hypersensitivity characterized by fever, dyspnea, pneumonitis, or pneumonia have been
reported in association with carbamazepine. There have been occasional instances of
anticonvulsant-treated patients developing elevated levels of cholesterol, HDL choles-
terol, and triglycerides. Cross-sensitivity to the allergic effects of other tricyclics may
occur when a patient is taking carbamazepine. Other side effects include aching joints
and muscles and leg cramps.

3. Drug Interactions
Carbamazepine induces the cytochrome P4502D6 system, which may increase the
metabolism, decrease serum concentrations, and reduce the elimination half-lives of
medications metabolized by this system when used concurrently. (See Table 7 for a list
of drugs that interact with carbamazepine.) Valproate may prolong the half-life and
reduce the protein binding of carbamazepine as well as increase the concentration of its
active 1O,1l-epoxide metabolite. Carbamazepine may increase levels of valproate's
hepatotoxic metabolite, 2-propyl-4-pentanoic acid.
The primary step in carbamazepine's metabolism may be via the P4503A4 sys-
tem, so carbamazepine levels may be increased when carbamazepine is used concur-
rently with 3A4 inhibitors, such as danazol (Danocrine® and others), diltiazem (Cardi-
zem® and others), verapamil (Calan® and others), or the influenza virus vaccine.
Carbamazepine toxicity may be delayed for several weeks after initiation of danazol
therapy and 1-2 weeks after the influenza virus vaccination. It is recommended that
nifedipine (Adalat® and others) be used as an alternative to verapamil or diltiazem.
Carbamazepine also increases the metabolism of estrogen-containing oral contra-
ceptives, so women taking these medications may need higher doses to prevent break-
through bleeding and to obtain adequate protection. * Parenteral medroxyprogesterone
or nonhormonal methods of birth control may be considered as alternatives.
Chronic use of a hepatic enzyme-inducing agent such as carbamazepine prior to
anesthesia may increase the metabolism of enflurane (Ethrane®), halothane (Fluo-
thane®), or methoxyflurane (Penthrane®), leading to an increased risk of hepatotoxi-
city. It may also increase the formation of nephrotoxic metabolites of methoxyflurane,
leading to increased risk of nephrotoxicity. In addition, cardiac arrhythmias may occur,
possibly due to sensitization of the myocardium resulting from increased concentra-
tions of norepinephrine.
Carbamazepine has an antidiuretic-hormone-like effect, which may counteract the
opposite effect of lithium (which produces polyuria) when the two drugs are used
together. By itself, carbamazepine has the potential to produce hyponatremia. Concur-
rent use of chlorpropamide (Diabinese®), clofibrate (Atromid-S®), desmopressin,
lypressin, posterior pituitary, thiazide diuretics (when used for their paradoxical anti-
diuretic activity in the treatment of diabetes insipidus), or vasopressin with carbamaze-
pine may potentiate the antidiuretic effect, leading to a lower sodium concentration and
causing adverse effects that include increased seizure activity. Coadministration of
carbamazepine and lithium may increase neurotoxic side effects even at nontoxic blood
concentrations of both drugs.
Carbamazepine can slow atrioventricular cardiac conduction and should be
avoided in patients who have heart block or prolonged QT and administered cautiously
in conjunction with other drugs that also can increase heart block, such as tricyclic
antidepressants and J3-blockers, among others.
The risk of hepatotoxicity with single toxic doses or prolonged use of high doses

*Failure to make this dosage increase in oral contraceptives can lead to an unwanted pregnancy, with the
fetus exposed to a teratogen!

TABLE 7. Drug Interactions with Carbamazepine

The following drugs increase levels of carbamazepine:

Cimetidine (Tagamet®)
Clarithromycin (Biaxin®)
Danazol (Danocrine® and others)
Diltiazem (Cardizem® and others)
Erythromycin (Emgel® and others)
Fluoxetine (Prozac®)
Fluvoxamine (Luvox®)
Influenza virus vaccine
Isoniazid (Nydrazid® and others)
Nefazodone (Serzone®)
Propoxyphene (Darvon® and others)
Troleandomycin (TAO®)
Verapamil (Calan® and others)
The following drugs decrease levels of carbamazepine:
Phenytoin (Dilantin® and others)
Carbamazepine decreases levels of:
Benzodiazepines metabolized via hepatic microsomal enzymes
Coumarin-derivative anticoagulants
Dacarbazine (DTIC-Dome®)
Digitalis glycosides [except possibly digoxin (Lanoxin®, Lanoxicaps®)]
Disopyramide (Norpace®)
Doxycycline (Vibramycin®)
Estrogen-containing oral contraceptives
Hormonal contraceptives
Hydantoin anticonvulsants
Indandione-derivative anticoagulants
Lamotrigine (Lamictal®)
Levothyroxine (Synthroid® and others)
Mexiletine (Mexitil®)
Succinimide anticonvulsants
Thyroid hormone
Tricyclic antidepressants

of acetaminophen may be increased in patients taking carbamazepine, and therapeutic

effects of acetaminophen may be decreased. When carbamazepine is coadministered
with aminophylline, oxtriphylline, or theophylline, it may stimulate hepatic metabolism
of the xanthines (except dyphylline), resulting in increased theophylline clearance.
The CNS depressant effects of carbamazepine may be enhanced when it is coad-
ministered with tricyclic antidepressants, haloperidol, loxapine, maprotiline, molin-
done, phenothiazines, pimozide, or thioxanthenes. This may lower the seizure thresh-
old. Anticholinergic effects may be potentiated, leading to confusion and delirium.
Also, concurrent use of haloperidol, and possibly other neuroleptics, with car-
bamazepine may decrease plasma concentrations of the neuroleptic by about 60%, with
or without adverse clinical effects.
Carbamazepine has been shown to lower plasma mebendazole (Vermox) concen-
trations in patients receiving high oral doses of mebendazole for treatment of tissue-
dwelling organisms and to impair the therapeutic response. Concurrent use of car-
bamazepine with monoamine oxidase inhibitors (MAOIs), including the antibacterial
drug furazolidone (Furoxone®) and the antilymphomic agent procarbazine (Matu-
lane®), has resulted in hyperpyretic crises, hypertensive crises, severe convulsions, and
death. A medication-free interval of at least 14 days is recommended between discon-
tinuation of MAOI therapy and initiation of carbamazepine therapy or vice versa. The
risk of carbamazepine-induced osteopenia is increased when carbamazepine is used
concurrently with carbonic anhydrase inhibitors.

4. Overdose
Symptoms of carbamazepine overdose include irregular breathing, respiratory
depression, tachycardia, hypotension or hypertension, shock, conduction disorders,
impairment of consciousness, convulsions (especially in small children), motor rest-
lessness, muscular twitching, tremor, athetoid movements, opisthotonus, ataxia,
drowsiness, dizziness, mydriasis, nystagmus, diadochokinesia, ballism, psychomotor
disturbances, dysmetria, initial hyperreflexia (followed by hyporeflexia), nausea, vom-
iting, anuria or oliguria, and urinary retention. Signs and symptoms begin to appear
after 1-3 hr, with neuromuscular disturbances being the most prominent. When very
high doses of carbamazepine (>60 g) have been ingested, severe cardiac complications
may occur, but cardiovascular problems are generally mild. When an overdose of
carbamazepine is taken in conjunction with alcohol, tricyclic antidepressants, barbitu-
rates, or hydantoins, the signs and symptoms of acute poisoning may be aggravated or
There is no specific antidote for carbamazepine overdose. The prognosis in cases
of severe poisoning depends on prompt elimination of the drug through induced vomit-
ing' stomach irrigation, or other appropriate measures that diminish absorption. Even
when more than 4 hr have elapsed following ingestion of the drug, the stomach should
be repeatedly irrigated, especially if the patient has consumed alcohol. Activated char-
coal and laxatives can be used to reduce absorption, and forced diuresis may accelerate
elimination. Dialysis is indicated only in severe poisoning associated with renal failure.
Replacement transfusion should be used in severe poisoning in small children.

The airways should be kept free to combat respiratory depression. Endotracheal

intubation, artificial respiration, and/or administration of oxygen can be used if neces-
sary. In cases of hypotension or shock, the patient's legs should be raised and a plasma
expander administered. A vasopressor may be used if blood pressure fails to rise
despite measures taken to increase plasma volume.
Diazepam or barbiturates should be used for convulsions. However, these medica-
tions may aggravate respiratory depression (especially in children), hypotension, and
coma. Barbiturates and benzodiazepines should not be used if drugs that inhibit mono-
amine oxidase have also been taken by the patient either in overdosage or in recent
therapy (within one week).
After a carbamazepine overdose, respiration, cardiac function (EKG monitoring),
blood pressure, body temperature, pupillary reflexes, and kidney and bladder function
should be monitored for several days.

S. Pregnancy
Carbamazepine crosses the placenta and should not be used in pregnant women.
Although adequate and well-controlled studies in humans have not been done, there
have been reports of babies (usually of mothers with epilepsy) prenatally exposed to
carbamazepine having small head circumferences, low birth weights, craniofacial de-
fects, fingernail hypoplasia, developmental delays, and spina bifida. Carbamazepine
also is distributed into breast milk. Concentrations in breast milk and in the plasma of
nursing infants have been reported to reach 60% of the maternal plasma concentration.
Therefore, the possibility exists that carbamazepine may cause adverse effects in a
nursing infant.

6. Preparations and Dosage

Carbamazepine is available in a lOO-mg chewable tablet, a 200-mg uncoated
tablet, and a 100-mg/5-ml suspension.

D. Antipsychotics
Antipsychotics, particularly haloperidol, can rapidly diminish increased motor
activity and also suppress other symptoms of mania. Lithium, in contrast, has more
specific antimanic effects but may take from several days to several months to control
manic symptoms. In general, lithium is more effective for stabilizing mood and ide-
ation while neuroleptics are superior in controlling hyperactivity. The course of clinical
response to antipsychotics is usually faster than that to lithium. Therefore, based on
research 89 - 92 and clinical experience, anti psychotics are usually used in addition to
lithium (or another antimanic agent) to gain more rapid relief of severe psychotic
symptoms and acute behavioral excitement, typically during an acute manic episode.
Although neuroleptics may have specific antimanic properties in addition to their
sedative and antipsychotic effects, they carry the risk of extrapyramidal symptoms and
may not be as effective over time as other agents. There is also concern about possible
neurotoxicity from the combination of lithium and a neuroleptic. 93 ,94

Recent data suggest that the atypical drug clozapine may have particular benefit
for acutely manic and psychotic patients,95 especially those who are treatment resistant
and/or intolerant of other agents. 96 - 98 Clozapine has also been reported to be effective
in a few cases of rapid-cycling bipolar disorder (see Section IV.C.5).99 Although
clozapine is free from most of the extrapyramidal side effects associated with typical
neuroleptics and may not cause tardive movement disorders, it does cause a number of
serious toxic reactions-including agranulocytosis, seizures, and cardiorespiratory
complications (see Chapter 4). Whether clozapine is superior to traditional neuroleptics
in the treatment of bipolar disorder remains to be seen. There is also preliminary
evidence 100. 10 I that another atypical neuroleptic, risperidone, may be beneficial in the
treatment of manic patients, but as several reports lO2 - 104 have found an association
between risperidone and new or increased manic symptoms, the jury is still out as to its
effectiveness in this condition.
Several studies have evaluated the use of depot antipsychotics for maintenance
treatment of bipolar disorder, with mixed results. In general, depot antipsychotics-
including flupenthixol, fluphenazine, haloperidol, pipothiazine, and zuclopenthixol-
reduce manic but not depressive episodes,l°5-107 This seems to be the case whether
depot antipsychotics are used to augment lithium or replace it, although some studies
have found no benefit to depot antipsychotic augmentation,l°8 One study reports a
decrease in relapses and hospitalization for manic, depressive, and mixed episodes
during treatment with depot antipsychotics,109 whereas another found a decrease in
manic episodes but an increase in depressive episodes in patients nonresponsive to or
intolerant of lithium. I 10
Prescription surveys have shown that many bipolar patients are maintained rou-
tinely and for long periods on antipsychotics in addition to lithium or other mood
stabilizers. I I I Reported rates for any antipsychotic use during treatment for bipolar
disorder range from 75% to lOO%, with 40-95% of patients continuing to take an
antipsychotic during maintenance treatment,l12-114 Beyond doubt, antipsychotic drugs
have a role to play in augmenting other antimanic agents during the acute treatment of
excited or psychotic mania, but their efficacy in maintenance treatment has never been
conclusively demonstrated. In addition, the risk of tardive movement disorders with
long-term antipsychotic use is 20-30%; the risk is probably even higher in bipolar
patients. On the other hand, there is a strong feeling in the field that some patients need
antipsychotic drug treatment, along with (or in some cases instead of) mood-stabilizing
agents to remain free of recurrent episodes of mania. To what extent some patients may
be unnecessarily exposed to long-term treatment with these agents is unknown. For the
clinician, the current state of knowledge translates into: Use maintenance antipsychotic
treatment for a bipolar patient if it really helps and the patient gives informed consent.
Document both of these facts.

E. Antidepressants
If mild, a depressive episode in a bipolar patient may be "ridden out" with
continuing lithium or other maintenance therapy and psychological support. More
severe depressions, however, require biological treatment. Early studies found com-

plete or partial antidepressant response to lithium in approximately 80% of pa-

tients.115-119 When lithium alone is not effective for acutely depressed bipolar patients,
an antidepressant should be added. If divalproex and carbamazepine are shown to be
efficacious as prophylactic agents for bipolar patients, it is likely that they will be more
effective in preventing mania than depression.
Studies on the use of antidepressants to treat the depressed phase of bipolar
disorder are woefully inadequate. Nonetheless, the few well-designed trials that have
been conducted can give us some guidance, and definite clinical impressions are
beginning to emerge from experts in the field. In particular, studies are examining the
controversy over whether antidepressants induce mania and rapid cycling in bipolar
patients and, if so, which antidepressants are more likely to have this effect. In one
group of patients with treatment-refractory bipolar disorder, 35% had mania likely
induced by an antidepressant, and 26% experienced cycle acceleration associated with
antidepressant treatment. 120 In another group, antidepressant-induced mania was found
to be a milder and more time-limited syndrome than spontaneous mania, and mania
induced by a tricyclic or by fluoxetine was rated as slightly more severe than that
induced by MAOIs or bupropion.l2l Twenty percent of patients with rapid-cycling
bipolar disorder experience rapid cycling only in association with the use of antidepres-
sants. 122 Of patients with spontaneous rapid cycling, 95.8% also experience some
antidepressant-induced or antidepressant -exacerbated rapid cycling.
The tricyclics are least favored among antidepressants for bipolar patients because
of suggestions that they may speed up cycling and also may be more likely than other
antidepressant groups to trigger switches into mania.
Although evidence is far from conclusive, many experts are turning to bupropion
for the depressed bipolar patient. Whether or not it is more efficacious than other
antidepressants, there are some suggestions that it may be less likely to trigger a switch
into mania. 123 Analyses of multiple studies comparing serotonin-selective reuptake
inhibitor (SSRI) antidepressants to tricyclics suggest that SSRIs also may be less likely
than tricyclics to precipitate mania in a bipolar patient. Since they generally are better
tolerated than tricyclics, SSRIs could be a good first- or second-line treatment. One
study suggests the possible superiority of tranylcypromine (Pamate®) to imipramine in
anergic depressed bipolar patients. 124 In cases of psychotic depression in a bipolar
patient, an antipsychotic drug may be necessary-usually in addition to an antidepres-
sant and a mood-stabilizing agent.

F. Other
The anticonvulsant drug lamotrigine (Lamictal®) has not been studied systemat-
ically in patients with bipolar disorder, but clinicians have had some success with this
agent in individual cases and one open trial. 125 Lamotrigine may have antidepressant as
well as mood-stabilizing properties. It appears to prevent antidepressant-induced cy-
cling and to be beneficial for patients with rapid-cycling bipolar disorder or mixed
states. 126 There have been several reports of serious adverse reactions associated with
lamotrigine, including multiple organ failure,127 encephalopathy,128 Stevens-Johnson
syndrome, 129 fatal toxic epidermal necrolysis, 130 and fulminant hepatic failure. 13 I Less

serious side effects include dizziness, headaches, diplopia, ataxia, nausea, vomiting,
somnolence, asthenia, and rash. Rash is the side effect that most often causes the drug
to be discontinued and is more likely to develop when initiallamotrigine doses are high
or when lamotrigine is started too rapidly in someone taking valproate. When
lamotrigine is introduced gradually, the risk of rash does not exceed that with car-
bamazepine. In adults, the initial dose of lamotrigine is usually 25 mg b.i.d. with
increases of 25 or 50 mg/day every week or two up to 50-100 mg b.i.d. In patients
taking carbamazepine, larger initial doses and more rapid dose increases are possible,
because carbamazepine-induced enzymes facilitate the metabolism of lamotrigine,
lowering serum lamotrigine levels. Phenobarbital and primidone also lower plasma
lamotrigine levels (by about 40%). As mentioned earlier, valproate increases serum
levels of lamotrigine, and a dosage reduction should be considered if the two medica-
tions are used concomitantly.132 In patients taking valproate, the initial dose of
lamotrigine is often 12.5 mg/day, and the drug is increased by 12.5 or 25 mg/day every
2 weeks.
There have been anecdotal reports of another new anticonvulsant, gabapentin
(Neurontin), being used to treat mania, but no data are yet available on its safety and
efficacy. It seems to have a more benign side-effect profile than lamotrigine, with fewer
reports of serious adverse reactions.
A few, small, double-blind trials have found the calcium channel blocker ve-
rapamil to be equal in efficacy to lithium 133 ,134 and superior to placebo135 in patients
with acute mania. Since there have been reports of neurotoxicity and bradycardia from
the combination of lithium and verapamil, these two medications should be combined
only with great caution. Because verapamil raises carbamazepine serum levels, this
combination should be monitored carefully.


A. Diagnostic Evaluation
Barring contraindications to its use, lithium should be the first agent tried for the
treatment of classic bipolar disorder, although valproate is also an acceptable first-line
alternative for acute mania. Before employing either of these agents for an acute
episode, however, the clinician should have a working diagnosis compatible with this
therapy. As noted earlier, the diagnoses of bipolar, manic, and schizoaffective disorder
clearly justify lithium treatment during the acute phase, whether lithium is used alone
or in conjunction with a neuroleptic. At times, the physician may prescribe lithium on a
trial-and-error basis for another psychiatric indication, usually based on the patient's
insufficient response to another treatment. At other times, the overlapping symptoms of
several possible psychiatric diagnoses-for example, acute mania versus schizo-
phrenia-will lead the physician to try lithium on an empirical basis.
In addition to making a primary diagnosis, the practitioner should become familiar
with factors that help predict outcome. He or she also should assess current stressors,
precipitating events, ego strengths, environmental resources, and psychodynamic, char-
acterologic, and interpersonal issues. Family and friends can be educated and serve as

allies. A careful history of the patient's illness over a lifetime is important. So too are
medical and neurologic histories and evidence of current or past substance abuse. The
development of an initial rapport that contributes to a growing therapeutic relationship
occurs simultaneously with this information-gathering process.
If the diagnosis of bipolar disorder is unclear, the clinician might justifiably
hesitate before recommending long-term maintenance lithium therapy. An acute epi-
sode with manic features could also be a brief reactive psychosis, a drug-precipitated
reaction, the beginning of schizophrenic illness, or a manifestation of a medical or
neurologic disorder.136 Especially when this is a first episode in a relatively young
patient without a family history of mood disorders, the practitioner may elect to follow
the patient carefully over time to chart out the course of future episodes, if any.
However, since recurrences cause a worsening offunction between episodes l2 and may
make a patient less likely to respond to medication, 17 intervention needs to be early and
aggressive. Therefore, contacts should be regular; psychosocial, medical, and neuro-
logic data gathering should be open and ongoing; and the physician should be available
between visits, in case intercurrent symptoms arise.
The diagnosis and treatment of bipolar disorder is often complicated by comorbid
substance abuse. While substance use disorders are prevalent among patients with all
psychiatric illnesses, bipolar disorder appears to be the most likely to co-occur with
alcohol or drug abuse.137 In fact, only 39% of bipolar patients do not have comorbid
substance abuse at some time during their lives. 138 Bipolar patients with substance use
disorders tend to have an earlier onset and worse course of illness compared to bipolar
patients who do not abuse drugs or alcohol, and many have a poor response to lithium
treatment. Preliminary data suggest that valproate may be effective in the acute treat-
ment of bipolar episodes complicated by concomitant substance dependence.139
When a bipolar diagnosis is clear-cut, maintenance lithium therapy might be
considered from the very first episode, especially if it is manic and severe-even more
in the presence of a family history of bipolar disorder. Although the course of episodes
in bipolar disorder varies from patient to patient, episodes tend to become more
frequent and severe over time, recurrences are the rule, and it is a lifetime disorder.
Furthermore, each episode carries a risk of mortality and a substantial likelihood of
psychosocial and economic morbidity. Worse, there is some suggestion that the occur-
rence of each episode might actually enhance the risk of future episodes, analogous to
the "kindling" model of epileptic seizures. All of these facts and hypotheses add up to a
strong argument for early maintenance treatment in the course of a bipolar disorder,
once the diagnosis is apparent and other possible causes (e.g., medical, neurologic,
substance abuse) are ruled out. Few clinicians would not consider maintenance treat-
ment after three episodes.
Factors Predicting Lithium Responsiveness 140,141
1. Positive response
a. Family history of mania or depression
b. Prior positive response to lithium
c. Positive response of family member to lithium
d. Mild and uncomplicated illness

2. Negative response
a. Rapid cycling (four or more episodes/year)
b. Mixed states or dysphoric mania
c. History of alcohol or drug abuse
d. Noncompliance with previous maintenance treatment
e. Cycle pattern of depression-mania-euthymia
f. Personality disturbance
g. History of poor interepisodic functioning
h. Poor social support system
i. Three or more prior episodes

B. Medical Evaluation
Various medical evaluations should be performed before beginning treatment with
any of the mood-stabilizing drugs used for the treatment of bipolar disorder (see Table
8). Before starting lithium therapy, the clinician should complete a medical history,
including a review of organ systems and a physical examination. The patient should be
screened for pregnancy, thyroid disease, epilepsy, renal disease, cardiovascular disease,
and evidence of brain damage. If the patient has any of these conditions, lithium should
be administered more cautiously. Appropriate laboratory screening prior to lithium
therapy should include a complete blood count; an EKG; urinalysis, BUN, and electro-
lytes; serum creatinine; and serum TSH, T4, T3RU (or other measure of binding
globulin), and free T4 (measured or calculated).
Clinicians should intermittently evaluate thyroid and renal functioning dur-
ing lithium therapy. A consensus about appropriate tests and their timing does not
exist. However, we recommend a serum creatinine determination every 6 months and a
serum TSH determination at 6 months and then annually thereafter. Beware of gradual
creatinine increases, even within the normal range, as these may indicate a gradual
deterioration in glomerular filtration. Similarly, a gradual increment in lithium level

TABLE 8. Pretreatment Tests

Before starting any mood stabilizer Comprehensive medical history

Physical examination
Pregnancy test
Before starting lithium Complete blood count
Electrocardiogram (EKG)
Renal panel (BUN, creatinine, and routine urinalysis)
Thyroid panel plus TSH
Before starting valproate Liver function test
Hematologic function test
Before starting carbamazepine Blood and platelet counts
Hepatic function test
Renal function test

despite a constant dose may reflect a similar renal deterioration. If TSH is elevated, a
full thyroid panel is indicated, and, together with an internist, the psychiatrist may
consider thyroid supplementation. We hasten to emphasize that, despite periodic labo-
ratory screening, the clinical index of suspicion must remain foremost as a means to
detect treatment-emergent medical conditions. Routinely questioning the patient about
physical symptoms, along with psychiatric and other mood changes, may prompt
targeted laboratory tests outside of the schedule we recommend. Any patient-and
particularly those taking maintenance medications-deserves appropriate annual
physical examinations, which, depending on a patient's age, may be supplemented with
an EKG or additional laboratory testing.
Before initiating valproate, the physician should obtain a comprehensive medical
history and perform a physical examination, paying particular attention to suggestions
of liver disease or bleeding abnormalities. Liver and hematologic functions should be
measured before treatment, every 1 to 4 weeks for the first 6 months, and then every 3
to 6 months. Serum ammonia concentrations should be monitored during treatment,
and bleeding time determinations, blood cell counts, and renal function determinations
are recommended prior to treatment and periodically during therapy. Patients should be
screened for pregnancy prior to treatment with valproate, as it is teratogenic.
Before carbamazepine is started, baseline blood and platelet counts, urinalysis,
and hepatic and renal function tests are in order. Although earlier guidelines called for
routine monitoring of some or all of these indices, and some physicians still perform
blood counts once or twice during the first few months of treatment and when plasma
concentrations are drawn, a more general consensus at present is to instruct patients
and family members to contact the doctor immediately if petechiae, pallor, weakness,
fever, or infection occurs. At that time, the physician should order relevant tests. As
with valproate, women of childbearing potential should be tested for pregnancy prior to
treatment with carbarnazepine, another known teratogen.
A first onset of mania after age 40 should prompt a thorough medical and neuro-
logic differential diagnosis. Primary mania seldom appears de novo in midlife or later,
but a wide range of brain insults can precipitate secondary mania, often indistinguish-
able by behavior alone from a primary mood disorder. At a minimum, the physician
should ensure a thorough physical and neurologic examination and routine laboratory
and possibly neuropsychological testing. An electroencephalogram (EEG) and head
computerized tomography (CT) or magnetic resonance imaging (MRI) should be con-
sidered. When a specific and treatable condition is found, it should be addressed
specifically, and the manic syndrome should be treated on a symptomatic basis: often
antipsychotic drugs will be satisfactory. If, however, there is no specific treatment for
the underlying organic condition, a psychiatrist may treat the manic symptoms much as
he or she would primary mania, i.e., with lithium, valproate, neuroleptics, and, if
necessary, other alternatives. Valproate might be superior to lithium for patients whose
mania is secondary to a neurologic condition.

C. Treatment
The therapeutic relationship provides a foundation for all drug therapies. In acute
situations, a caretaker may use his or her position to establish quick rapport and, when

necessary, to confront or set limits. His or her initial approach should include various
crisis intervention techniques, such as mobilization of environmental resources. In-
volvement and education of persons close to the patient generally increases medication
compliance. In addition, maintenance therapy requires the formation of a trusting
collaboration between clinician and patient. With many patients, open discussions
about the illness set the tone for the ongoing dialogue. Most individuals have some
trouble accepting their illness and its possible ramifications. Therefore, the clinician
should try to understand how mood episodes affect the patient's life, exploring how
each episode has interfered with the patient's relationships and major roles.
During the early phase of treatment, patients with bipolar illness most often want
to know:
1. Where did I get it?
2. Is it transmitted to my children?
3. Is my problem permanent?
4. What is the treatment?
5. What are the risks of drug therapy?
6. How effective is therapy?
These queries frequently encompass more fundamental, but less obvious, concerns
about control, dependency, personal defectiveness, and "lovableness" in personal rela-
tionships. Therefore, the clinician should try to clarify both the explicit and implicit
concerns and requests before providing answers that might close off continued discus-
sion. No matter what course is taken, however, discussion of the potentially inherited
nature of bipolar illness and its possible course should be coupled closely with explana-
tions about the effectiveness of lithium maintenance and the importance of the thera-
peutic relationship. Establishing an open relationship may take many months. The
clinician should encourage the patient to become an active partner in maintenance
therapy. For example, the patient and family members should be sensitized to emerging
hypomanic and depressive symptoms. The clinician, patient, and family together
should delineate a list of symptoms that are early manifestations of the patient's
affective illness and require immediate clinical contact (such as sleeplessness, spending
extra money, or anorexia). Also, the patient and family should be enlisted to detect
medical conditions that might affect the lithium level and signs and symptoms of
The discussion of these aspects of treatment often stimulates other questions,
1. How long will treatment last?
2. How will we know when to stop lithium?
3. What are the risks of treatment?
4. How will I feel on the medicine?
5. Does lithium interact with other drugs?
Again, these questions should be answered only after determining what they mean to
the patient and how he or she will use the information. In addition, the clinician has a
human and legal responsibility to discuss adverse reactions (see Tables 2, 4, and 6),
potential toxicity, alternative treatments, limitations of maintenance therapy, and long-

tenn prognosis. Although medication-free periods may be tried, there is a real risk of
recurrences, and no one seems to "lose" a mood disorder over time; if anything, the
disorder worsens.

1. Acute Manic Episode

Lithium remains the first choice of most clinicians for a classic manic episode;
valproate is an acceptable first-line alternative. Although valproate has appeared in
recent studies to be equal in efficacy to lithium for acute mania, lithium has been used
clinically for decades, has a more extensive supporting literature, and costs less than
valproate. For patients who have problematic side effects, medical complications,
previous nonresponse, rapid cycling, dysthymic mania, or other factors suggesting an
alternative to lithium, valproate often is the first choice. Carbamazepine, combination
therapies, and ECT are alternatives for patients who do not respond to first-line treat-
Hypomanic symptoms can often be treated in an outpatient setting, sometimes by
psychosocial support, appropriate family intervention, and environmental manipula-
tions, with lithium as the sole medication. If the patient's excitement, poor judgment,
and embarrassing or dangerous behaviors mount, and if frank mania with psychotic
features develops, hospitalization will usually be required to protect the patient and
initiate, restart, or adjust the medication regimen. In such cases, it is probable that the
patient will require an antipsychotic drug in addition to a mood stabilizer, even though,
as we mentioned earlier, this might increase the risk of neurotoxicity in occasional
patients. Haloperidol works well for acute behavioral excitement and the disorganized
thinking of manic psychosis, usually takes effect promptly, and can be parenterally
administered. However, all antipsychotic drugs can probably contain acute psychotic
As in the treatment of schizophrenic psychoses (see Chapter 4), the usual effective
daily dose for haloperidol is between 5 and 15 mg. A coadministered mood stabilizer
probably will act synergistically to nonnalize a patient's mood, but only after a week or
more. What should the clinician do if a patient is highly excited and possibly danger-
ous? Although in the 1970s it was commonplace to use "rapid neuroleptization"-that
is, high, frequent, and parenteral doses of a high-potency antipsychotic drug-in
attempts to suppress psychotic symptoms (sometimes in excess of 100 mg of halo-
peridol daily), this is seldom necessary, it may be counterproductive, and it can in-
crease the risk of acute cardiovascular complications, seizures, acute dystonia, and
possibly neuroleptic malignant syndrome. For most acutely manic patients, 5-15 mg of
haloperidol daily should suffice, augmented by a benzodiazepine to increase sedation.
In emergent situations, intramuscular administration of the neuroleptic can achieve
higher blood levels more rapidly than oral preparations.
For the highly excited manic patient, clinicians tum more now to adjunctive
benzodiazepine medications. Some have claimed clonazepam (Klonopin®) to be an
effective antimanic drug in its own right, but scientific data are limited and inconclu-
sive, and our own clinical impression has been that clonazepam functions more as a
highly sedating agent and, therefore, a good adjunct to a neuroleptic. Lorazepam
(Ativan® and others) has the advantage of availability in parenteral fonn, which is

distinctive in its rapid and predictable absorption when administered intramuscularly.

Whichever benzodiazepine is chosen, it will allow enhanced behavioral control of the
acutely excited manic patient without the risks of high antipsychotic doses. At the same
time, benzodiazepines raise rather than lower the seizure threshold, do not add cardio-
vascular complications, treat rather than exacerbate extrapyramidal reactions, and will
not engender the neuroleptic malignant syndrome.
For the acutely excited manic patient, physical restraints and ECT, preferably
bilateral, can be used. Inpatient staff should be encouraged to minimize stimulation of
these patients, even to the point of isolating them from other patients.
In general, start withdrawing the neuroleptic gradually after therapeutic serum
levels of lithium (i.e., from 0.8 to 1.3 mEq/L) have been achieved and the patient has
been euthymic and nonpsychotic for several weeks. Tapering the antipsychotic usually
requires 1-2 weeks. The clinician can assess whether or not manic symptoms are
reappearing by carefully evaluating the patient's feelings and thoughts and soliciting
information from family or hospital staff. Many manic patients, in fact, will report
racing thoughts even when they appear calm.
Initially, practitioners should prescribe lithium in divided doses, follow the serum
levels, and then increase the amount as necessary. Although various formulas have been
proposed by which steady-state lithium plasma levels can be predicted based on the level
after a single lithium dose,142.143 they have not been found consistently to be accu-
rate. 144.145 Instead, the physician may begin therapy with a modest dose, such as 300 mg
oflithium carbonate b.i.d., and then measure the plasma level after 5 to 7 days' treatment.
Of course, in an elderly, frail, medically ill, or renally impaired patient, the dose can be
started with as little as 300 mg daily (or even every other day). If the plasma level after a
week of therapy at the initial dose is below 0.8 mEq/L, the clinician can increase the dose
by one capsule daily and check the level again after another week. If symptoms or signs
suggesting toxicity occur, a 12-hr plasma level should be checked sooner to ensure that
toxic blood levels have not been reached.
The above gradual approach to initiating lithium therapy should mitigate adverse
effects and make the patient's introduction to lithium a gentler experience. Also,
although this approach takes more time, the slow onset of lithium's therapeutic effects
usually means that it is not the critical variable in "breaking the back" of a manic
episode. Having said this, there are times when a clinician may feel desperate to move
speedily in controlling a florid and dangerous manic episode. If such is the case in a
young, healthy, and robust patient, one may start with 900 or 1200 mg of lithium daily
in divided doses and check the blood level after only 4 or 5 days. Such a strategy may
lead to "overshoot" in plasma lithium level, and if any l2-hr blood level comes back
above 1.5 mEq/L, the doctor should promptly lower the lithium dose. With a more
rapidly escalating regimen, medical and nursing staffs must be doubly sensitive to early
signs of toxicity, in which case the next dose of lithium should be held while awaiting
an immediate blood level determination.
Many clinicians have opined that acutely manic patients require a much higher
dose of lithium to achieve a given blood level than the same patients do when they have
returned to euthymia. The validity of this observation remains untested scientifically,
and if it is true, it is unclear whether or not it may simply be an epiphenomenon of
increased motor activity. Clinically, the prudent response is to use sufficient lithium to

achieve the desired blood level but to be observant during the postmanic phase for
signs of toxicity that might accompany a rise of blood lithium level into a toxic range.
After the first week, lithium levels should be determined weekly for a few weeks
and then about once a ,month. Once steady state has been achieved, the frequency of
dosing may be reduced to just once or at most twice daily, even with standard lithium
carbonate preparations. When the manic episode resolves, the clinician must make a
decision about maintenance therapy.
For a patient with acute mania who cannot tolerate lithium or does not respond to
lithium despite adequate levels and appropriate neuroleptic therapy, valproate should
be substituted for lithium. If the patient responds partially to lithium therapy, valproate
may be added to lithium therapy. In each case, carbamazepine should be tried instead of
valproate if valproate also appears to be inefficacious or intolerable to the patient.
When these medications are combined, they may be effective at lower dosages than
those required in monotherapy.
The typical starting dose of valproate for healthy adults is 750 mg daily in divided
doses. (Equivalent oral doses of divalproex sodium and valproic acid capsules deliver
systemically equivalent quantities of the valproate ion.) If helpful in selected cases,
valproate is available as "sprinkles," which can be mixed with food. The dose should
be increased as rapidly as possible to achieve the lowest therapeutic dose that produces
the desired clinical effect or a l2-hr serum valproate concentration between 50 and 125
IJ.g/mL. As serum levels rise, the likelihood of efficacy increases, but so do side effects.
The time of dosing is determined by possible side effects, and, if tolerated, a once-
a-day dosing schedule can be employed. As with lithium, the antimanic response to
valproate typically occurs after 1-2 weeks, although some experts think valproate
works several days faster than lithium. A still faster response may be achieved by oral
loading. Patients given divalproex sodium at a dosage of 20 mg/kg per day from day 1
often attain serum concentrations greater than 50 mg/L by the second day. The maxi-
mum recommended dosage is 60 mg/kg per day. The rapid onset of antimanic and
antipsychotic response to this method may be comparable to that of haloperidol. 146
Geriatric patients taking valproate tend to have elevated free, unbound valproic
acid concentrations and lowered intrinsic clearances. This indicates a decrease in
valproate-metabolizing capacity and a fall in serum albumin. Therefore, these patients
should receive a lower daily dosage, and the serum concentrations should be kept in the
lower therapeutic range.
For carbamazepine used as a monotherapy, the typical starting dose is 200 to 400
mg/day in three to four divided doses, increased to 800-1000 mg/day by the end of the
first week. If clinical improvement is insufficient by the end of the second week, and
the patient has not developed intolerable side effects to the drug, increases to as high as
1600 mg daily may be considered. Although a correlation between blood level and
clinical response has yet to be established, most clinicians are guided by the antiepilep-
sy range: usually 4-15 ng/mL. If carbamazepine is combined with lithium or neurolep-
tics, physicians often prefer to use lower doses and blood levels. If valproate and
carbamazepine are administered simultaneously, blood levels of each should be moni-
tored carefully because of complex interactions between the two agents. When the dose
of carbamazepine is built up rapidly, side effects are more likely.

2. Hypomania
Clinicians often can begin lithium treatment in hypomanic patients without hospi-
talization. Lithium is relatively easy to start once the patient has been medically
cleared. Give 300 mg of lithium carbonate 2 or 3 times a day for 1 week and then draw
a serum level. Inform the patient about the blood-drawing procedure and instruct him
or her to allow 12 hr between the last dose and the serum measurement. For each
dosage change, a steady-state level is achieved after about 1 week. After drawing the
first serum level, the clinician can usually approximate a maintenance dose, which can
then be tried and tested with the next blood level. In maintenance therapy, 0.8-1.0
mEq/L may be the most effective range. When side effects are intolerable, reduced
plasma concentrations of lithium may diminish them. However, reduced levels de-
crease protection against relapse and recurrence, particularly for mania. 17 Few, if any,
patients will experience prophylactic benefit from lithium at blood levels below 0.4
During the first week of therapy, the administration of lithium has few serious
risks in healthy patients because of rapid clearance and low total dosage. Therefore,
telephone contact or short office visits usually suffice unless the patient requires
additional support and reassurance. Gastrointestinal symptoms are common during
initial treatment and during dosage increases. If these symptoms are severe, however,
smaller doses or a different preparation may help. If the dose is changed, another serum
level should be obtained a week later. Once the desired level is reached, monthly
checks are adequate. The clinician should remember that patients might retain more
lithium when they are hypomanic or manic than when they are euthymic. Therefore,
before the number of serum level checks is diminished, the patient's mood should be
stable. For fully stabilized patients, the clinician can check lithium levels as infre-
quently as every 3 months. In any patient taking lithium, and particularly in those
stopping medication, the caretaker should educate the patient and relatives about the
signs of incipient mania. Some patients may want to carry a wallet card or an ID
necklace, bracelet, or anklet indicating their condition and medication.
Lithium can provide dramatic relief for both manic and hypomanic symptoms.
However, a maintenance regimen for the prevention of future episodes may require the
administration of lithium for 12-18 months before the clinician can fully determine its
Antipsychotics are seldom needed in treating hypomania, but a benzodiazepine at
bedtime often promotes sleep. Although the efficacy of valproate and carbamazepine in
hypomania is incompletely studied, we may assume efficacy for valproate and possibly
carbamazepine, which can be used in a manner similar to what we have described
under mania.

3. Acute Depressive Episode

As mentioned earlier, there is a wid~spread clinical impression that administering

an antidepressant to a bipolar patient can trigger a switch into mania. Some experts also
believe that antidepressants speed up mood cycles, although this point is more contro-
versial. Because of both concerns, some psychiatrists prefer to be cautious about

prescribing an antidepressant at the first sign of depression in a bipolar patient who is

being maintained on lithium, an anticonvulsant, or a combination. If symptoms are
mild and short-lived, therefore, the clinician might consider not prescribing medication
right away, but offering psychological support, checking drug levels, and maintaining
close contact to detect any more severe deterioration. When biological intervention is
necessary-based on severity of symptoms, dysfunction, and possibly threat to life-
an antidepressant must be considered.
In a recent survey,147 clinicians ranked bupropion and SSRIs as first-line treat-
ments for a severe bipolar depression, with MAOIs as the next preferred option,
followed by venlafaxine, tricyclic antidepressants, and nefazodone. These medications
were also recommended in the same sequence for a milder depressive episode. Admin-
istration of an antidepressant to reverse an acute depression in a bipolar patient should
always be used together with one of the mood-stabilizing agents-usually lithium,
valproate, carbamazepine, or a combination. Mood stabilizers may enhance the effec-
tiveness of the antidepressant and might protect against the possibility of a switch into
mania. Psychotherapy plays a second-line role as an adjunct to medication. Electrocon-
vulsive therapy is a highly effective biological antidepressant for patients insufficiently
responsive to drugs or incapable of tolerating them or for whom depression is of life-
threatening severity.
Once symptoms of depression have completely remitted and the patient has re-
turned to euthymia, the antidepressant can be tapered and discontinued over a matter of
weeks-a much shorter period than the 6 months typically recommended for continua-
tion therapy after recovery from a nonbipolar depression. If a patient switches into
mania, the antidepressant should be discontinued immediately.

4. Maintenance Therapy
During mania or depression, a patient may not fully understand the illness and
circumstances or have insight and judgment about treatment options. When mainte-
nance therapy is contemplated, in contrast, the patient should be competent to give
informed consent and must be an active partner in treatment. In fact, only the patient
can control many of the vital aspects of therapy. Therefore, the clinician should openly
discuss with the patient the nature of the mood illness and the clinical effects and
pharmacokinetics of lithium or other medications to be employed. He or she should
also inform the patient about possible adverse reactions, normal precautions, and the
importance of monitoring the serum drug levels. For example, a patient taking lithium
should be advised to maintain a stable salt intake and contact the physician when there
are significant changes in salt balance [e.g., after marked sweating, diarrhea from an
illness, institution of a low-sodium diet, or when another medication is coadministered
(particularly diuretics)]. The patient also should be told about common adverse effects:
hand tremor, gastrointestinal upset, polydipsia and polyuria, and weight gain. Patients
should know that changes in caffeine intake may alter lithium levels. In addition,
patients should be instructed to stop lithium and contact the physician if any of these
symptoms persist or get worse or if dizziness, drowsiness, slurred speech, or ataxia
appears. Patients taking other mood stabilizers need analogous information about their

regimen. Sometimes we give patients a "log" to highlight important information,

reinforce their role, and provide a means of following their treatment. The development
of a collaborative relationship usually instills feelings of confidence and competence
and improves compliance.
At least in some individuals, symptoms are precipitated by stressful events. There-
fore, many patients benefit from treatment that helps to alter a negative self-image or a
destructive relationship to their environment. Patients and their families should know
that sleep loss can precipitate mania, and they should be prepared to take appropriate
steps as needed. In all instances, a trusting, constant, collaborative, helping relationship
is prerequisite. Moreover, formal psychotherapy may sometimes offer the patient a
more self-directed means of preventing, suppressing, understanding, or resolving some
of the affective symptoms.
If a patient does well on lithium maintenance treatment but develops cognitive
dulling, partial or full substitution of divalproex for lithium appears to improve adverse
cognitive effects while maintaining a stable mood. 148 A small pilot study comparing
lithium plus divalproex with lithium alone for maintenance treatment found increased
efficacy for the combination but, not surprisingly, increased side effects-such as
gastrointestinal distress, tremor, cognitive impairment, and alopecia-as well. 149
When valproate is used for maintenance therapy, it is best to start with a low dose, such
as 250-500 mg daily, building the dose gradually to attain a plasma level between 50
and 125 IJ.g/mL.
Carbamazepine for long-term therapy should be started at a low dose, 100
mg/day, with increments of 100 mg/day every 4 to 5 days. Dosing is usually two to
four times daily. The target range is the same as that used in acute mania: 4-15 ng/mL.
When carbamazepine is used to control seizures, concentrations as high as 17 ng/mL
are only rarely required and do not necessarily produce unacceptable side effects or
toxicity. Most patients taking carbamazepine for bipolar disorder are maintained on
doses between 400 and 1800 mg daily. Because of enzyme induction, it is often
necessary to increase the dose after 2 to 3 weeks on treatment to maintain the same
blood level.

5. Rapid Cyclers
Patients with rapid-cycling bipolar disorder-defined as four or more affective
episodes in one year, with or without an intervening period of euthymia-tend to be
less responsive to lithium treatment. lO For many patients, rapid cycling is a phase in
their lifelong illness. Whether rapid cycling is a natural progression of the illness or a
separate disorder has yet to be determined. 122 The onset of rapid cycling has been
associated with antidepressant drugs (especially tricyclic antidepressants) and hypo-
thyroidism. 150,151 Some people also experience ultrarapid cycling, switching between
moods over days or even hours.
Rapid cycling tends to be more prevalent in women, who make up approximately
75% of this population. 4 Premenstrual syndrome (PMS) has been related to major
depressive disorder, and one study found that 15 (60%) of 25 female rapid-cycling
patients had severe PMS symptoms as opposed to 5 (20%) of 25 control subjects. 152 In

addition, bipolar patients with severe PMS had more affective episodes in one year than
those with moderate or mild PMS symptoms.
Various therapeutic approaches have been investigated for treating patients with
rapid-cycling bipolar disorder. Although at this time the scientific literature does not
point to one standard protocol, experts rate valproate as the treatment of choice for a
patient with acute mania and rapid cycling.147 Carbamazepine is considered an alter-
nate first-line treatment, whereas lithium, alone or in combination with other drugs
(such as choline 153 ), is considered a second-line treatment. Before beginning any
medication for a rapid-cycling patient, the clinician should obtain a careful history,
searching for possible provocative factors, such as a precipitating effect of antidepres-
sants or the presence of endocrinopathies, other medications, seasonal factors, psycho-
social stressors, and so on. If there is not a specific provocation that can be removed or
remedied, the clinician should attempt to stabilize the patient pharmacologically.
Patients with rapid-cycling bipolar II disorder may require lower doses of valpro-
ate than those currently recommended for bipolar I disorder: 125-500 mg/day, corre-
sponding to a mean serum valproate level of 32.5 j.Lg/mL, was effective in 69% of
patients in one trial. 154 Clozapine and nimodipine may be effective in patients with
rapid-cycling bipolar disorder nonresponsive to conventional treatments. 99 ,155,156
There are also suggestions that sleep deprivation may help alleviate depressive epi-
sodes in rapid-cycling patients, but it also may precipitate mania. 157,158

D. Mixed or Dysphoric Mania

Forty percent of bipolar patients experience mixed or dysphoric mania, in which
symptoms of depression and mania occur simultaneously.9 Dysphoric mania may
represent a distinct clinical state. It appears to be more prevalent in women (although
data are contradictory4), to have a poorer prognosis, and to respond differently to
medication. Response to lithium is poor in 60-70% of these patients. As with rapid
cycling, patients with dysphoric mania do better with valProate or carbamazepine than
with lithium, and valproate is now considered the first choice for this indication, ahead
of lithium and carbamazepine, which are also first-line altematives. 147

E. Schizoaffective Disorder
The boundaries between schizophrenia and schizoaffective disorder and between
the latter and bipolar disorder are blurry at best and nonexistent at worst. Most proba-
bly, the schizoaffective label includes patients from both the schizophrenic and mood-
disorder categories, as well as others who might have a distinct condition. Since there
are no valid predictors-clinical or biological-of pharmacological response, the best
strategy is empirical. With lithium as a less hazardous option than neuroleptics for
long-term therapy (no tardive dyskinesia), patients with a suggestion of mood disorder
should be tried on it first. In some cases, the clinician may conclude that combining
lithium with an antipsychotic drug is the maintenance treatment of choice, but because
of the additive risks and adverse effects, the clinical record should clearly justify this
polypharmacy. As clinical experience increases, roles for valproate and other mood
stabilizers and for atypical neuroleptics should become clearer.


Although less common than recurrent major depression, bipolar illness is impres-
sive in its manifestations, more frequently requires hospitalization, and carries consid-
erable hazards in morbidity and mortality, economic and social stability, and risk of
substance abuse. Its response to biological treatment, however, is equally impressive,
and for many patients with this psychiatric condition, lithium is little short of a god-
send. As noted earlier, though, for patients who cannot tolerate or do not benefit fully
from lithium, valproate has been found to be equally effective in treating acute mania.
Perhaps it will tum out to be a good treatment for maintenance treatment also. Car-
bamazepine and clozapine may also be effective for some patients.
All patients with bipolar disorder require a stable relationship with the treating
physician. This is no different from any other chronic medical illness. Many patients
also require some form of psychosocial intervention to enjoy maximum mood stability
over the lifetime course of this condition. Approached sensitively and knowledgeably,
treating patients with bipolar disorder can be both successful and gratifying.

1. Kessler R C., McGonagle K A., Zhao S., et al: Lifetime and 12-month prevalence of DSM-III-R
psychiatric disorders in the United States. Arch Gen Psychiatry 51:8-19, 1994.
2. Muller-Oerlinghausen B., Ahrens B., Grof E., et al: The effect of long-term lithium treatment on the
mortality of patients with manic-depressive and schizoaffective illness. Acta Psychiatr Scand 86:218-
222, 1992.
3. Weissman M. M., Bland R c., Canino G. 1., et al: Cross-national epidemiology of major depression
and bipolar disorder. JAMA 276:293-299, 1996.
4. Leibenluft E.: Women with bipolar illness: Clinical and research issues. Am J Psychiatry 153:163-173,
5. Krasa N. R.: Sex differences in age at first affective episode. Am J Psychiatry 152:960, 1995.
6. McMahon E 1., Stine O. C., Chase G. A., Meyers D. A., Simpson S. G., DePaulo J. R., Jr.: Influence of
clinical subtype, sex, and lineality on age at onset of major affective disorder in a family sample. Am J
Psychiatry 151:210-215, 1994.
7. Goodwin E K, Jamison K R: Manic-Depressive Illness. New York, Oxford University Press, 1990.
8. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Washing-
ton, D.C., American Psychiatric Association, 1994.
9. Calabrese J. R, Woyshville M. 1., Kimmel S. E., Rapport D. J.: Mixed states and bipolar rapid cycling
and their treatment with divalproex sodium. Psychiatr Ann 23:70-78, 1993.
10. Dunner D. L., Fieve R R.: Clinical factors in lithium carbonate prophylaxis failure. Arch Gen Psychia-
try 30:229-233, 1974.
11. Keller M. B., Lavori P. W., Coryell W., et al: Differential outcome of pure manic, mixed/cycling, and
pure depressive episodes in patients with bipolar illness. JAMA 255:3138-3142, 1986.
12. Prien R. E, Gelenberg A. J.: Alternatives to lithium for preventive treatment of bipolar disorder. Am J
Psychiatry 146:840-848, 1989.
13. Cohen L. S., Sichel D. A., Robertson L. M., Heckscher E., Rosenbaum J. E: Postpartum prophylaxis
for women with bipolar disorder. Am J Psychiatry 152:1641-1645, 1995.
14. Kendell R E., Chalmers J. C., Platz C.: Epidemiology of puerperal psychoses. Br J Psychiatry
150:662-673, 1987.
15. Winokur G., Clayton P. J., Reich T.: Manic Depressive Disease. SI. Louis, C.V. Mosby, 1969.
16. Post R. M.: Transduction of psychosocial stress into the neurobiology of recurrent affective disorder.
Am J Psychiatry 149:999-1010, 1992.

17. Ge1enberg A. 1., Kane 1. M., Keller M. B., et al: Comparison of standard and low serum levels of
lithium for maintenance treatment of bipolar disorder. N Engl J Med 321:1489-1493, 1989.
18. Altshuler L. L.: Bipolar disorder: Are repeated episodes associated with neuroanatomic and cognitive
changes? Bioi Psychiatry 33:563-565, 1993.
19. Friedman M., Culver C., Ferrell R.: On the safety oflong-term treatment with lithium. Am J Psychiatry
134:1123-1126, 1977.
20. Savard R. J., Rey A. C., Post R. M.: Halstead-Reitan category test in bipolar and unipolar affective
disorders: Relationship to age and phase of illness. J Nerv Ment Dis 168:297-304, 1980.
21. Hauser P., Altshuler L. L., Berrettini W., Dauphinais I. D., Gelernter J., Post R. M.: Temporal lobe
measurement in primary affective disorder by magnetic resonance imaging. J Neuropsychol 1:128-
134, 1989.
22. Altshuler L. L., Conrad A., Hauser P., Li X., Guze 8., Post R. M.: Temporal lobe volume in bipolar
disorder: A preliminary report. Arch Gen Psychiatry 48:482-483, 1991.
23. Coffman, J. A., Bornstein R. A., Olson S. C., Schwarzkopf S. B., Nasrallah H.: Cognitive impairment
and cerebral structure by MRI in bipolar disorder. Bioi. Psychiatry.
24. Swayze V. w., II, Andreasen N. C., Alliger R. J., Yuh W. T. C., Ehrhardt J. C.: Subcortical and temporal
structures in affective disorder and schizophrenia: A magnetic resonance imaging study. Bioi Psychia-
try 31:221-240, 1992.
25. Dolan R. J., Calloway P., Mann A. H.: Cerebral ventricular size in depressed subjects. Psychol Med
15:873, 1985.
26. Andreasen N. C., Swayze V. w., II, Aaum M., Alligen R., Cohen G.: Ventricular abnormalities in
affective disorder: Clinical and demographic correlates. Am J Psychiatry 147:893-900, 1990.
27. McInnis M. G., McMahon F. J., Chase G. A., Simpson S. G., Ross C. A., DePaulo J. R. J.: Anticipation
in bipolar affective disorder. Am J Hum Genet 53:385-390, 1993.
28. Clift E.: Data Consortium to search for genes causing manic-depressive illness. Psychiatric Times
29. Weissman M. M., Leaf P. 1., Livingston B. M., et al: The epidemiology of dysthymia in the community:
Rates, risks, comorbidity and treatment, in 140th Annual Meeting of the American Psychiatric Associa-
tion. Chicago, American Psychiatric Association, 1987.
30. Anonymous: Medical Practice Project: A State-of the-Science Report for the Office of the Assistant
Secretary for the U.S. Department of Health, Education and Welfare, 1979 (Abstract).
31. Mukherjee S., Sackeim H. A., Schnur D. B.: Electroconvulsive therapy of acute manic episodes: A
review of 50 years' experience. Am J Psychiatry 151:169-176, 1994.
32. Small J. G., Klapper M. H., Kellams 1. J., et aI: Electroconvulsive treatment compared with lithium in
the management of manic states. Arch Gen Psychiatry 45:727-732, 1988.
33. Coffey C. E., Winer R. D.: Electroconvulsive therapy: An update. Hosp Community Psychiatry
41:515-521, 1990.
34. Consensus Development Conference: Electroconvulsive therapy. JAMA 254:2103-2108, 1985.
35. Segman R. H., Shapira B., Gorfine M., Lerer B.: Onset and time course of antidepressant action:
Psychopharmacological implications of a controlled trial of electroconvulsive therapy. Psychophar-
macology 119:440-448, 1995.
36. Fieve R. R.: The lithium clinic: A new model for the delivery of psychiatric services. Am J Psychiatry
132:1018-1022, 1975.
37. Shakir S. A., Volkmar F. R., Bacon S., Pfefferbaum A.: Group psychotherapy as an adjunct to lithium
maintenance. Am J Psychiatry 136:455-456, 1979.
38. Kripke D. F., Robinson D.: Ten years with a lithium group. McLean Hosp 10: I-II, 1985.
39. Volkmar F. R., Bacon S., Shakir S. A., et al: Group therapy in the management of manic-depressive
illness. Am J Psychother 35:226-234, 1981.
40. Davenport Y. B., Ebert M. H., Ad1and M. L., Goodwin F. K.: Couples group therapy as an adjunct to
lithium maintenance of the manic patient. Am J Orthopsychiatry 47:495-502, 1977.
41. Glick I. D., Clarkin J. F., Spencer J. H., et aI: A controlled evaluation of inpatient family intervention:
Preliminary results of the six-month follow-up. Arch Gen Psychiatry 42:882-886, 1985.

42. Miklowitz D. J., Goldstein M. 1.: Behavioral family treatment for patients with bipolar affective
disorder. Behav Modif 4:457-489, 1990.
43. Miller 1. w., Keitner G. 1., Epstein N. B., Bishop D. S., Ryan C. E.: Families of bipolar patients:
Dysfunction, course of illness, and pilot treatment study. 1991 (Abstract).
44. Scott J.: Psychotherapy for bipolar disorder. Br J Psychiatry 167:581-588, 1995.
45. Cochran S. D.: Preventing medical noncompliance in the outpatient treatment of bipolar affective
disorders. J Consult Clin Psycho I 52:873-878, 1984.
46. Price L. H., Heninger G. R.: Lithium in the treatment of mood disorders. N Engl J Med 331 :591-598,
47. Manji H. K., Potter W. Z., Lenox R. H.: Signal transduction pathways: Molecular targets for lithium's
actions. Arch Gen Psychiatry 52:531-543, 1995.
48. Pollack S. J., Atack J. R., Knowles M. R., et al: Mechanism of inositol monophosphatase, the putative
target of lithium therapy. Proc Natl Acad Sci U.S.A. 91 :5766-5770, 1994.
49. Berridge M. J., Downes C. P., Hanley M. R.: Neural and developmental actions of lithium: A unifying
hypothesis. Cell 59:411-419, 1989.
50. Baraban J. M.: Toward a crystal-clear view of lithium's site of action. Proc Natl Acad Sci U.S.A.
91:5738-5739, 1994.
51. Manji H. K., Chen G., Shimon H., Hsiao J. K., Potter W. Z., Belmaker R. H.: Guanine nucleotide-
binding proteins in bipolar affective disorder: Effects of long-term lithium treatment. Arch Gen Psychi-
atry 52:135-144, 1995.
52. Keller M. 8., Lavori P. W., Kane 1. M., et al: Subsyndromal symptoms in bipolar disorder: A compari-
son of standard and low serum levels of lithium. Arch Gen Psychiatry 49:371-376, 1992.
53. Fieve R. R.: Overview of therapeutic and prophylactic trials with lithium in psychiatric patients, in
Gershon S., Shopsin B. (eds): Lithium. Its Role in Psychiatric Research and Treatment. New York,
Plenum Press, 1973, p. 336.
54. Van Putten T.: Why do patients with manic-depressive illness stop their lithium? Compr Psychiatry
16:179-183, 1975.
55. Cooper T. B.: Pharmacokinetics of lithium, in Meltzer H. Y. (ed): Psychopharmacology: The Third
Generation of Progress. New York, Raven Press; 1987, pp. 1365-1376.
56. Schou M., Anderson A., Eskajaer-Jensen S., et al: Occurrence of goiter during lithium treatment. Br J
Med 3:710-713,1968.
57. Emerson C. H., Dyson W. L., Utiger R. D.: Serum thyrotropin and thyroxin concentration in patients
receiving lithium carbonate. J Clin Endocrinol Metab 36:338-346, 1973.
58. Jefferson 1. w.: Lithium carbonate-induced hypothyroidism-its many faces. JAMA 242:271-272,
59. VanderVelde C. D., Gordon M. W.: Manic-depressive illness, diabetes mellitus and lithium carbonate
treatment. Arch Gen Psychiatry 21 :478-485, 1969.
60. Mellerup E. T., Plenge P., Rafaelson O. J.: Renal and other controversial adverse effects of lithium, in
Meltzer H. Y. (ed): Psychopharmacology: The Third Generation of Progress. New York, Raven Press;
1987, pp. 1443-1448.
61. Gelenberg A. J.: Lithium-induced polyuria: Approaches to management. Bioi Ther Psychiatry Newslett
6:17-19, 1983.
62. Gelenberg A. J., Wojcik J. D., Falk W. E., et al: Effects of lithium on the kidney. Acta Psychiatr Scand
75:29-34, 1987.
63. Gitlin M. 1.: Lithium-induced renal insufficiency. J Clin Psychopharmacol 13:276-279, 1993.
64. Gelenberg A. 1.: Lithium and the nephrotic syndrome. Bioi Ther Psychiatry 12:24, 1989.
65. Shopsin B., Gershon S.: Pharmacology-toxicology of the lithium ion, in Gershon S., Shopsin B. (eds):
Lithium: Its Role in Psychiatric Research and Treatment. New York, Plenum Press; 1973, pp. 107-146.
66. Schou M.: Electrocardiographic changes during treatment with lithium and with drugs of the im-
ipramine type. Acta Psychiatr Scand 169:258-259, 1963.
67. Roose S. P., Nurnberger J. L., Dunner D. L., et al: Cardiac sinus node dysfunction during lithium
treatment. Am J Psychiatry 136:804-806, 1979.

68. Jaffe C. M.: First-degree atrioventricular block during lithium carbonate treatment. Am 1 Psychiatry
134:88-89, 1977.
69. Worthley L. 1. C.: Lithium toxicity and refractory cardiac arrhythmia treated with intravenous magne-
sium. Anesth Intens Care 2:357-360, 1974.
70. Gelenberg A. 1., Jefferson 1. w.: Lithium tremor. 1 Clin Psychiatry 56:283-287, 1995.
71. Jefferson 1. W., Greist 1. H.: Adverse reactions-neurological tremor, in Jefferson 1. W., Greist 1. H.
(eds): Primer of Lithium Therapy. Baltimore, Williams & Wilkins, 1977, pp. 139-150.
72. Ghandirian A. M., Lehman H. E.: Neurological side effects of lithium: Organic brain syndrome,
seizures, extrapyramidal side effects, and EEG changes. Compr Psychiatry 21:327-335, 1980.
73. Levine S. H., Puchalski c.: Pseudotumor cerebri associated with lithium therapy in two patients. 1 Clin
Psychiatry 51 :251-253, 1990.
74. Alvarez-Cermeno J. c., Fernandez 1. M., O'Neill A., Moral L., Saiz-Ruiz J.: Lithium-induced head-
ache. Headache 29:246-247, 1989.
75. Saul R. E, Hamburger H. A., Selhorst J. B.: Pseudotumor cerebri secondary to lithium carbonate.
lAMA 253:2869-2870, 1985.
76. Pesando P., Nuzzi G., Maraini G.: Bilateral papilloedema in long term therapy with lithium carbonate.
Pharmakopsychiatr Neuro-Psychopharmakol 13:235-239, 1980.
77. Lobo A., Pilek E., Stokes P. E.: Papilledema following therapeutic dosages of lithium carbonate. 1 Nerv
Ment Dis 166:526-529, 1978.
78. Gelenberg A. 1.: Polypharmacy in psychiatry: Balancing caution with daring. CNS Drugs 4:410-411,
79. Schou M.: Long-lasting neurological sequelae after lithium intoxication. Acta Psychiatr Scand 70:594-
602, 1984.
80. Cohen L. S., Friedman J. M., Jefferson 1. w., Johnson E. M., Weiner M. L.: A reevaluation of risk of in
utero exposure to lithium. lAMA 271:146-150, 1994.
81. Thase M. E., Rush A. J.: Treatment-resistant depression, in Bloom E E., Kupfer D. 1. (eds): Psycho-
pharmacology: The Fourth Generation of Progress. New York, Raven Press; 1995, pp. 1081-1097.
82. Donaldson S. R., Gelenberg A. J., Baldessarini R. 1.: The pharmacologic treatment of schizophrenia: A
progress report. Schizophrenia Bull 9:504-527, 1983.
83. Gjessing L. R.: Lithium citrate loading of a patient with periodic catatonia. Acta Psychiatr Scand
43:372-375, 1967.
84. Kline N. S., Wren J. C., Cooper T. B., et al: Evaluation of lithium therapy in chronic and periodic
alcoholism. Am 1 Med Sci 268:15-22, 1974.
85. Rifkin A., Quitkin E, Carrillo R., et al: Lithium carbonate in emotionally unstable character disorder.
Arch Gen Psychiatry 27:519-522, 1972.
86. Gottfries C. G.: The effect of lithium salts on various kinds of psychiatric disorders. Acta Psychiatr
Scand 203:157-167, 1968.
87. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, DSM-Ill-R.
Washington, D.C., American Psychiatric Association, 1987.
88. Bowden C. L., Brugger A. M., Swann A. c., et al: Efficacy of divalproex vs. lithium and placebo in the
treatment of mania. lAMA 271:918-924, 1994.
89. Prien R. E, Caffey E. M., Klett C. J.: A comparison of lithium carbonate and chlorpromazine in the
treatment of acute mania. Arch Gen Psychiatry 26:146-153, 1972.
90. Shopsin B., Gershon S., Thompson H., Collins P.: Psychoactive drugs in mania: A controlled compari-
son of lithium carbonate, chlorpromazine, and haloperidol. Arch Gen Psychiatry 32:34-42, 1975.
91. Post R. M., Jimerson D. c., Bunney W. E., Jr., Goodwin E K.: Dopamine and mania: Behavioral and
biochemical effects of the dopamine receptor blocker pimozide. Psychopharmacology (Berlin) 67 :297-
305, 1980.
92. Cookson J. C., Silverstone T., Wells B.: A double-blind controlled study of pimozide versus chlo-
rpromazine in mania. Psychopharmacol Bull 16:38-41, 1980.
93. Kessel J. B., Verghese C., Simpson G. M.: Neurotoxicity related to lithium and neuroleptic combina-
tions? A retrospective review. 1 Psychiatry Neurosci 17:28-30, 1992.

94. Spring G., Frankel M.: New data on lithium and haloperidol incompatibility. Am J Psychiatry 138:818-
821, 1981.
95. Zarate C. A., Tohen M., Baldessarini R. J.: Clozapine in severe mood disorders. J Clin Psychiatry
56:411-417, 1995.
96. Suppes T., McElroy S. L., Gilbert J., Dessain E. c., Cole J. 0.: Clozapine in the treatment of dysphoric
mania. Bioi Psychiatry 32:270-280, 1992.
97. Kimmel S. E., Calabrese J. R., Woyshville M. J., Meltzer H. Y.: Clozapine in treatment-refractory mood
disorders. J Clin Psychiatry 55:91-93, 1994.
98. Calabrese J. R., Kimmel S. E., Woyshville M. 1., et al: Clozapine for treatment-refractory mania. Am J
Psychiatry 153:759-764, 1996.
99. Calabrese J. R., Meltzer H. Y., Markovitz P. J.: Clozapine prophylaxis in rapid cycling bipolar disorder.
J Clin PsychopharmacoI1l:396-397, 1991.
100. Goodnick P. J.: Risperidone treatment of refractory acute mania. J Clin Psychiatry 56:431-432,1995.
101. Tohen M., Zarate C. A., Jr., Centorrino E, Hegarty J., Froeschl M., Zarate S. B.: Risperidone in the
treatment of mania. J Clin Psychiatry 57:249-253, 1996.
102. Dwight M. M., Keck P. E., Jr., Stanton S. P., Strakowski S. M., McElroy S. L.: Antidepressant activity
and mania associated with risperidone treatment of schizoaffective disorder. Lancet 344:554-555,
103. Koek R. J., Kessler C. C.: Probable induction of mania by risperidone. J Clin Psychiatry 57: 174-175,
104. Tomlinson W. c.: Risperidone and mania. Am J Psychiatry 153:132-133, 1996.
105. Naylor G., Scott c.: Depot injections for affective disorders. Br J Psychiatry 136:105, 1980.
106. Lowe M., Batchelor D.: Depot neuroleptics and manic depressive psychosis. Int J Clin Psychophar-
macoll:53-62, 1986.
107. White E., Cheung P., Silverstone T.: Depot antipsychotics in bipolar affective disorder.lnt Clin Psycho-
pharmacoI8:1l9-122, 1993.
108. Esparon J., Kalloor J., Naylor G. J., et al: Comparison of the prophylactic action of flupenthixol with
placebo in lithium treated manic depressive patients. BrJ Psychiatry 148:723-725,1986.
109. Littlejohn R., Leslie E, Cookson 1.: Depot antipsychotics in the prophylaxis of bipolar affective
disorder. Br J Psychiatry 165:827-829, 1994.
110. Ahlfors U. G., Baastrup P. C., Dencker S. J., et al: Aupenthixol decanoate in recurrent manic depressive
illness. Acta Psychiatr Scand 64:226-237, 1981.
111. Sachs G. S.: Adjuncts and alternatives to lithium therapy for bipolar affective disorder. J Clin Psychia-
try 50:31-39, 1989.
112. Sernyak M. J., Woods S. W.: Chronic neuroleptic use in manic-depressive illness. [Review]. Psycho-
pharmacol 8uIl29:375-381, 1993.
113. Sernyak M. J., Griffin R. A., Johnson R. M., Pearsall H. R., Wexler B. E., Woods S. W.: Neuroleptic
exposure following inpatient treatment of acute mania with lithium and neuroleptic. Am J Psychiatry
151:133-135, 1994.
114. Vestergaard P., Schou M.: Prospective studies on a lithium cohort. Acta Psychiatr Scand 78:421-426,
115. Goodwin E K., Murphy D. L., Bunney W. E., Jr.: Lithium carbonate treatment in depression and mania:
A longitudinal double-blind study. Arch Gen Psychiatry 21:486-496, 1969.
116. Goodwin E K., Murphy D. L., Dunner D. L., Bunney W. E., Jr.: Lithium response in unipolar versus
bipolar depression. Am J Psychiatry 129:44-47, 1972.
117. Noyes R., Jr., Dempsey G. M., Blum A., Cavanaugh G. L.: Lithium treatment of depression. Compr
Psychiatry 15:187-193, 1974.
118. Baron M., Gershon E. S., Rudy V., Jonas W. Z., Buchsbaum M.: Lithium carbonate response in
depression: Prediction by unipolar/bipolar illness, average-evoked response, catechol-O-methyl trans-
ferase, and family history. Arch Gen Psychiatry 342:1107-1111, 1975.
119. Mendels J.: Lithium in the treatment of depressive states, in Johnson E N. (ed): Lithium Research and
Therapy. New York, Academic Press; 1975, pp. 43-62.

120. Altshuler L. L., Post R. M., Leverich G. S., Mikalauskas K., Rosoff A., Ackerman L.: Antidepressant-
induced mania and cycle acceleration: A controversy revisited. Am J Psychiatry 152: 1130-1138, 1995.
121. Stoll A. L., Mayer P. Y., Kolbrener M., et al: Antidepressant-associated mania: A controlled compari-
son with spontaneous mania. Am J Psychiatry 151:1642-1645, 1994.
122. Bauer M. S., Calabrese J., Dunner D. L., et al: Multisite data reanalysis of the validity of rapid cycling
as a course modifier for bipolar disorder in DSM-IY. Am J Psychiatry 151:506-515, 1994.
123. Sachs G. S., Lafer B., Stoll A. L., et al: A double-blind trial ofbupropion versus desipramine for bipolar
depression. J Clin Psychiatry 55:391-393, 1994.
124. Himmelhoch J. M., Thase M. E., Mallinger A. G., Houck P.: Tranylcypromine versus imipramine in
anergic bipolar depression. Am J Psychiatry 148:910-916, 1991.
125. Bowden C. L.: Role of newer medications for bipolar disorder. J Clin Psychopharmacol 16:48S-55S,
126. Calabrese J. R, Fatemi S. H., Woyshville M. 1.: Antidepressant effects of lamotrigine in rapid cycling
bipolar disorder. Am J Psychiatry 153: 1236, 1996.
127. Gregersen H., Nielsen J. S., Peterslund N. A.: [Acute porphyria and multiple organ failure during
treatment with lamotrigine). Ugeskr Laeg 158:4091-4092, 1996.
128. Hennessy M. 1., Wiles C. M.: Lamotrigine encephalopathy. Lancet 347:974-975, 1996.
129. Campistol J., Geli M., L1istosella E., Molins 1., L10bert M.: [Stevens-Johnson syndrome after
lamotrigine treatment). Rev NeuroI23:1236-1238, 1995.
130. Sterker M., Berrouschot J., Schneider D.: Fatal course of toxic epidermal necrolysis under treatment
with lamotrigine. Int J Clin Pharmacol Ther 33:595-597, 1995.
131. Makin A. J., Fitt S., Williams R, Duncan J. S.: Fulminant hepatic failure induced by lamotrigine. Br
Med J 311:292, 1995.
132. Anderson G. D., Yau M. K., Gidal B. E., et al: Bidirectional interaction of valproate and lamotrigine in
healthy subjects. Clin Pharmacol Ther 60: 145-156, 1996.
133. Giannini A. 1., Houser W. L., Jr., Loiselle R H., Giannini M. C., Price WA: Antimanic effects of
verapamil. Am J Psychiatry 141:1602-1603, 1984.
134. Garza-Trevino E. S., Overall 1. E., Hollister L. E.: Verapamil versus lithium in acute mania. Am J
Psychiatry 149:121-122, 1992.
135. Dubovsky S. L., Franks R D., Allen S., Murphy J.: Calcium antagonists in mania: A double-blind study
of verapamil. Psychiatry Res 18:309-320, 1986.
136. Popkin M. K.: Patients with multiple sclerosis presenting to psychiatric hospitals. J Clin Psychiatry
56:307-308, 1995.
137. Brady K. T., Sonne S. c.: The relationship between substance abuse and bipolar disorder. J Clin
Psychiatry 56: 19-24, 1995.
138. Regier D. A., Farmer M. E., Rae D. S., et al: Comorbidity of mental disorders with alcohol and other
drug abuse: Results from the Epidemiologic Catchment Area (ECA) study. JAMA 264:2511-2518,
139. Brady K. T., Sonne S. C., Anton R., Ballenger 1. c.: Valproate in the treatment of acute bipolar affective
episodes complicated by substance abuse: A pilot study. J Clin Psychiatry 56:118-121, 1995.
140. Prien R E, Potter W. Z.: NIMH workshop report on treatment of bipolar disorder. Psychopharmacol
Bull 26:409-427, 1990.
141. Bowden C. L.: Predictors of response to divalproex and lithium. J Clin Psychiatry 56:25-30, 1995.
142. Zetin M., Garber D., De Antonio M., et al: Prediction of lithium dose: A mathematical alternative to the
test-dose method. J Clin Psychiatry 47: 175-178, 1986.
143. Zetin M., Shokooh S., Daghighi-Asi A., et al: Prediction of lithium dose: A follow-up study. Lithium
1:182-183, 1990.
144. Cummings M. A., Haviland M. G., Wareham J. G., et al: A prospective clinical evaluation of an
equation to predict daily lithium dose. J Clin Psychiatry 54:55-58, 1993.
145. Terao T., Oga T., Nozaki S., et al: A further prospective evaluation of an equation to predict daily
lithium dose. J Clin Psychiatry 56: 193-195, 1995.
146. McElroy S. L., Keck P. E., Stanton S. P., Tugrul K. C., Bennett 1. A., Strakowski S. M.: A randomized

comparison of divalproex oral loading versus haloperidol in the initial treatment of acute psychotic
mania. J CUn Psychiatry 57:142-146, 1996.
147. Frances A., Docherty J. P., Kahn D. A., Carpenter D.: The expert concensus guideline series: Treatment
of bipolar disorder. J Clin Psychiatry, 57 (Suppl 12A):1-88, 1996.
148. Stoll A. L., Locke C. A., Vuckovic A., Mayer P. V.: Lithium associated cognitive and functional deficits
reduced by a switch to divalproex sodium: A case series. J Clin Psychiatry 57:356-359, 1996.
149. Solomon D. A., Ryan C. E., Keitner G. I., et al: A pilot study of lithium carbonate plus divalproex
sodium for the continuation and maintenance treatment of patients with bipolar I disorder. J Clin
Psychiatry 58:95-99, 1997.
150. Wehr T. S., Goodwin F. K.: Can antidepressants cause mania and worsen the course of affective illness?
Am J Psychiatry 144:1403-1411, 1987.
151. Roy-Byrne P. P., Joffe R. T., Uhde T. w., Post R. M.: Approaches to the evaluation and treatment of
rapid-cycling affective illness. Br J Psychiatry 145:543-550, 1984.
152. Price W. A., DiManio L.: Premenstrual tension syndrome in rapid-cycling bipolar affective disorder. J
Clin Psychiatry 47:415-417, 1986.
153. Stoll A. L., Sachs G. S., Cohen B. M., Lafer B., Christensen 1. D., Renshaw P. F.: Choline in the
treatment of rapid-cycling bipolar disorder: Clinical and neurochemical findings in lithium-treated
patients. Bioi Psychiatry 40:382-388, 1996.
154. Jacobsen F. M.: Low-dose valproate: A new treatment for cyclothymia, mild rapid cycling disorders,
and premenstrual syndrome. J Clin Psychiatry 54:229-234, 1993.
155. Privitera M. R., Lamberti J. S., Maharaj K.: Clozapine in a bipolar depressed patient. Am J Psychiatry
150:986, 1993.
156. Goodnick P. J.: Nimodipine treatment of rapid cycling bipolar disorder. J Clin Psychiatry 56:330, 1995.
157. Benjamin 1., Zohar J.: Sleep deprivation in rapid-cycling bipolar affective disorder: Case report. Eur
NeuropsychopharmacoI2:463-465, 1992.
158. Kasper S., Wehr T. A.: The role of sleep and wakefulness in the genesis of depression and mania.
Encephale 18:45-50, 1992.




The focus of this chapter is on antipsychotic agents-drugs also commonly called

neuroleptics and, less commonly (and less appropriately), antischizophrenic agents or
major tranquilizers (a misnomer). This pharmacologic family has become a mainstay in
the treatment of aspects of schizophrenia and other psychotic disorders as well as many
nonpsychotic conditions. The chapter also describes other approaches to treating psy-
choses and concludes by discussing future directions.
The most firmly grounded indication for using antipsychotic drugs is in the
treatment of the more florid and acute symptoms of schizophrenic psychosis such as
hallucinations, delusions, other aspects of thought disturbance, and excited, aggressive
behavior. In addition, antipsychotic drugs may alleviate similar symptoms in patients
with other syndromes, such as paranoid disorders, schizophreniform disorder, brief
psychotic disorder, schizoaffective disorder, atypical psychosis, and psychosis associ-
ated with mood disorders such as melancholia and mania. Furthermore, antipsychotic
drugs may alleviate psychotic symptoms or excited and assaultive behavior in patients
with organic mental disorders, retardation, and childhood psychoses. At times, these
ubiquitous chemicals have helped in the treatment of patients with severe pain syn-
dromes and difficult personality disturbances.
The almost never-ending list of disorders treated with antipsychotic agents, which
practically spans the fourth edition of the Diagnostic and Statistical Manual of Mental
Disorders (DSM-IV), far exceeds the experimental data base that should ideally bolster
clinical practice. The most convincing data support the use of these drugs to treat
schizophrenic patients. Some experimental evidence suggests that they are effective for
other disorders, such as mood-related psychoses, but for many illnesses, the literature is

Alan J. Gelenberg, M.D.· Department of Psychiatry, University of Arizona Health Sciences Center,
Tucson, Arizona 85724. Samuel Keith, M.D. Department of Psychiatry, University of New Mexico
School of Medicine, Albuquerque, New Mexico 87131.
The Practitioner's Guide to Psychoactive Drugs (Fourth Edition), Gelenberg and Bassuk, eds.
Plenum Publishing Corporation, New York, 1998


merely anecdotal. However, even when scientific research has not kept pace with
clinical practice, patients continue to require medication; therefore, this chapter dis-
cusses the most up-to-date information on the use of these compounds, blending data
from scientific research with clinical experience.


A. Nonbiological
Psychosocial treatments for schizophrenia have gone through a number of phases
relative to their usefulness in this disorder. This evolution began in the 1960s, when it
was felt unethical not to provide psychotherapy, progressed through a period of high
enthusiasm for pharmacologic treatments during which psychotherapy was considered
useful only if it encouraged compliance, and has reached a point today where specific
therapies are considered based on their specific efficacy. It is safe to say that enthusi-
asm for intensive individual psychotherapy either on the inpatient unit or during fol-
low-up with outpatients has relatively little support from the research literature. Begin-
ning with the Camarillo State Hospital Study, by May et al.,) which examined five
therapeutic modalities-milieu, psychotherapy alone, ECT, psychotherapy with medi-
cation, and medication alone-the major studies have failed to find additional benefit
from psychotherapy, nor have they found that psychotherapy can substitute for medica-
tion. The results of the May study indicated that those patients who were treated with
medication did better than those who were not and that the addition of the psycho-
therapeutic component to medication did not add appreciable benefits. The second
major study, by Stanton, Gunderson, Knapp, et al., 2 was designed to determine if there
were any benefits from intensive psychotherapy over supportive psychotherapy. The
results of this study were twofold: first, the sicker patients dropped out of the more
intensive treatment, but not out of the supportive treatment; and, second, despite this
differential attrition, there were no differences between treatment outcomes on a wide
variety of measures. This does not mean that the person with schizophrenia can be
ignored; it only means that there is little support for using substantial resources on
intensive individual psychotherapy.
The data on group psychotherapy for acutely psychotic inpatients with schizo-
phrenia are even more sobering. 3 There is no evidence to suggest that group therapy is
beneficial, and, from what we know about the sensitivity of people with schizophrenia
to complex, multifocal, intensive, intrusive, critical stimuli, there is little theoretical
support for it either. Therefore, the role of inpatient group meetings should be reconsid-
ered for this population. In the outpatient setting, where the goal changes from remis-
sion of psychosis to the development of interactional skills that will allow patients to
resume some level of productivity, theoretical support and a marginal data base of
empirical evidence finds a role for low-intensity group therapy.
The major change in psychosocial treatments for schizophrenia over the last decade
has been the refinement of treatments that involve the family. Eschewing the historically

and painfully incorrect ideas of family pathogenesis of schizophrenia, the new forms of
family treatment encourage the family to become a member of the clinical treatment
team. This rationale is based not on what a family does wrong but on what the family is in
a position to do correctly and better than anyone else in the patient's environment. Sixty-
five percent of schizophrenic patients discharged from a hospital return to their families
of origin, and even more in the early stages of the illness. In these days of declining
resources, the family represents a natural support system that would be very difficult to
replicate. However, the family should be a short-term resource and not be expected to
become the new back wards of the community. Further, many families have suffered
under the threefold punishment of having a child with schizophrenia, being accused of
causing the illness, and, despite being excluded from the treatment process, being asked
to pay for it. It is no wonder that families are skeptical over a "new" approach to family
The common features of new approaches to family treatment of schizophrenia
include (1) a positive clinical alliance, (2) psychoeducation, (3) teaching general man-
agement skills for the illness, and (4) expanding social networks of the family. The
strategies have differed in location of delivery (home or clinic), in mode (multiple
families, single families; with or without the patient), in timing (in hospital or outpa-
tient), and even in principles (behavioral, educational, or psychodynamic). What has
not differed is the results. Each of the five studies done in the 1980s showed a
remarkable effect of family management programs when compared with other forms of
treatment, including, for example, individual psychotherapy.4,5 The treatment princi-
ples include education of the patient and family, management of stress (particularly
interpersonal stress), general case management, and a special emphasis on early recog-
nition of relapse.
Family management, like all psychosocial approaches, should not be seen as
competitive with pharmacologic treatment but instead as an additive or synergistic
approach to the illness of schizophrenia.

B. Nondrug, Biological

1. Electroconvulsive Therapy
Electroconvulsive therapy (ECT) can playa dramatic, even lifesaving, role in the
treatment of psychotic depression. It also suppresses acute mania. A wealth of clinical
anecdotes testify to similarly impressive results in many cases of catatonic schizo-
phrenia (however, not all nor even most catatonia is schizophrenia, which reduces the
confidence in the anecdotal data6 ).
At the 1985 Consensus Development Conference on ECT sponsored by the Na-
tional Institutes of Health and Mental Health, Dr. Joyce Small reported that ECT can
playa role in treating schizophrenic disorders that are relatively acute and marked by
intense affective symptoms.? By contrast, patients who have been ill for 5 years or
more are unlikely to improve. Antipsychotic drugs appear superior in efficacy to ECT
for the treatment of acute schizophrenia, but the combination of drugs and ECT might

be better still. Electroconvulsive therapy is clearly superior to placebo in patients who

have been ill for less than 2 years. The treatment of acute schizophrenia with ECT
seems to require a greater number of sessions than does the treatment of acute depres-

2. Psychosurgery
The frontal lobotomy, used to treat many schizophrenic patients earlier in the
century, has largely been abandoned because of an unfavorable risk/benefit ratio. More
restrictive psychosurgical procedures, such as cingulotomy, may effectively treat pa-
tients with chronic, severe depressions, obsessive-compulsive disorder, and intractable
pain syndromes, but no substantial evidence supports the use of psychosurgery in the
treatment of psychosis. 8


A. Introduction and Terms

The introduction of the first antipsychotic drugs-reserpine (Serpasil® and others)
and the phenothiazines-in the early to mid-1950s revolutionized psychiatry. For the
first time, chemotherapy offered more than sedation to acutely disturbed and psychotic
patients. Ultimately, antipsychotic drugs combined with a changing fiscal philosophy
helped to diminish the previously sharp increase in the number of hospital beds in the
United States occupied by schizophrenic patients. Many of these individuals could now
function in their communities with periodic admission to psychiatric units, often in
general hospitals. Psychiatric wards became quieter and less violent. The subspecialty
of psychopharmacology was born, and many other important classes of psychophar-
maceuticals have emerged. Unfortunately, the advent of antipsychotic drugs has not
eliminated the problem of chronic schizophrenia. With the passage of the Community
Mental Health Centers Act in 1963, President John F. Kennedy mandated a "bold new
approach" to treating the mentally ill. However, because of political and fiscal pres-
sures, many severely ill patients were "deinstitutionalized" to nonexistent community
programs. The sad result has been the re-creation of back wards in the community, with
many chronic schizophrenic patients poorly medicated, hardly managed, and home-
less. 9
The rauwolfia alkaloid reserpine caused a number of problems, but the effects of
the phenothiazine chlorpromazine (Thorazine® and others) were impressive. From this
prototypical phenothiazine emerged a host of sister compounds with similar actions
and effectiveness but with different potencies (i.e., milligram dosages) and unwanted
effects. During the past three decades, various classes of antipsychotic compounds
have been synthesized. They sometimes differ structurally from chlorpromazine but are
pharmacologically and clinically similar.
Clinicians and researchers have labeled antipsychotic drugs with various syn-
onyms, but the terms often have resulted in both semantic and conceptual confusion.
When they were first used, these pharmaceuticals often were called "major tran-

quilizers." This term developed from the observation that chlorpromazine and reser-
pine produce somnolence and relaxation and the consequent misbelief that their major
action was sedative. However, relatively non sedating antipsychotic compounds just as
effectively combat psychotic symptoms. Moreover, patients often become tolerant to
the sedating effects of antipsychotic drugs but not to the antipsychotic effects them-
selves. Finally, because of the implication that "major tranquilizers" are on a spectrum
with, but more powerful than, sedative-hypnotic and antianxiety agents (sometimes
called "minor tranquilizers"), this term is best avoided. Our concern is that patients
may have these medications prescribed for "tranquilization" and be at risk for long-
term side effects.
Because antipsychotic drugs frequently produce signs of neurological dysfunc-
tion, most notably Parkinson's syndrome and other extrapyramidal reactions, the term
"neuroleptic" was coined. In fact, researchers originally believed that (1) any drug
effective in combating psychosis must produce extrapyramidal effects and (2) in any
given patient, the induction of extrapyramidal signs indicated an optimal therapeutic
dose. However, based on the following evidence, these assumptions are incorrect:
1. The piperidyl phenothiazine thioridazine (Mellaril® and others) produces a
relatively low incidence of short-term extrapyramidal effects yet is as effec-
tive as any other antipsychotic agent.
2. Clozapine (Clozaril®), a dibenzodiazepine compound, is an effective anti-
psychotic agent that produces few, if any, extrapyramidal effects.
3. Many patients show marked clinical improvement in psychotic symptoms
without experiencing parkinsonian signs or related reactions.
Therefore, although all antipsychotic agents on the U.S. market are, in fact, neurolep-
tics, an antipsychotic drug need not be neuroleptic-a fact presenting a challenge for
future research. Furthermore, when treating individual patients, clinicians should
try to avoid the emergence of neurological effects. It is beginning to appear that there
are several dopamine systems in the brain, with the nigrostriatal being responsible for
the extrapyramidal symptoms and those projecting from the ventral tegmental area
being responsible for antipsychotic effects (see Fig. 1). If this is confirmed, it may
mean that it is not necessary for extrapyramidal symptoms to appear at all. Some of the
antipsychotic drugs currently under investigation purport to have this differential ef-
fect. Recent research suggests the intriguing possibility that the first appearance of
slight hypokinesia and rigidity might signal the minimally effective neuroleptic dose
for the treatment of an acutely psychotic patient I 0, II; this remains to be explored more
fully. The introduction of a growing number of "atypical neuroleptics," which are
supposed to be less likely than typical neuroleptics to cause extrapyramidal effects, has
opened new clinical vistas and promises enhanced insights into the nature of these
drugs and their actions.

B. Effects on Behavior and the Nervous System

In animals, antipsychotic drugs inhibit conditioned avoidance behavior, suppress
electrical intracranial self-stimulation, block vomiting and aggression produced by the

---:;;,:.0+--- Frontal & cingulate cortex

~..----- Nucleus accumbens

- - - - - Mesocortical pathway
-"+-t......f - - - - Caudate & putamen
"'fT;I~- Mesolimbic pathway
1 - - - - Median eminence
Tubero-infundibular pathway
Arcuate nucleus of hypothalamus
......--.,;:~-- Nigrostrtatal pathway
- - - Amygdala
1---- Substantia nigra (A9)
- - - Ventral tegmental area (A10)

Figure 1. Brain dopamine tracts.

dopamine agonist apomorphine, and produce cataleptic immobility resembling human

catatonia. In early testing, researchers observed that antipsychotic drugs potentiated
anesthesia and produced a state called "artificial hibernation." Chlorpromazine, the
prototype phenothiazine, did not by itself induce anesthesia but rather promoted sleep
and diminished interest in the environment; animals required increased stimulation or
motivation to perform tasks. Antipsychotic drugs have relatively little tendency to

suppress vital centers in the brainstem: coma, respiratory depression, and cardiovascu-
lar collapse are rare, even at very high doses.
In the electroencephalogram (EEG), phenothiazines and other antipsychotic
chemicals produce slowing and synchronization and a decrease in arousal-induced
changes-effects that are reversed by dopamine agonists. The low-potency agents
(e.g., chlorpromazine) also tend to lower the seizure threshold. Clinically, this effect is
particularly important in patients predisposed to seizures, such as those with epilepsy,
and in individuals undergoing withdrawal from sedative-hypnotic drugs (including
alcohol) (see Chapter 7).
Most antipsychotic agents (with thioridazine as an interesting exception) have
antiemetic effects. They can protect against the nausea and vomiting that usually follow
administration of apomorphine, presumably by blocking the latter's dopamine agonis-
tic effects in the chemoreceptor trigger zone of the medulla.

C. Mechanism of Action
Antipsychotic drugs block dopamine receptors in various pathways within the
brain, which probably accounts for their therapeutic effectiveness as well as for some
of their more prominent unwanted effects. According to widely held theories, antip-
sychotic activity depends on the blockage of postsynaptic receptors in dopamine-
mediated pathways that run from the midbrain to the limbic system (septal nucleus, the
olfactory tubercle, and the amygdala) (see Fig. 1) and to the temporal and frontal lobes
of the cerebral cortex. In fact, the effectiveness of antipsychotic drugs in blocking the
D2 subtype of dopamine receptors correlates with their clinical potency (i.e., usual
daily doses). Presumably, tolerance to the dopamine-blocking action of antipsychotic
drugs does not develop in these mesolimbic and mesocortical pathways, explaining the
impression that tolerance does not develop to their antipsychotic efficacy. [Although
some authors have postulated the development of tolerance to the antipsychotic effec-
tiveness of these drugs ("tardive psychosis"),12 at best this is rare, and most clinicians
and scholars remain to be convinced that it is a valid phenomenon.]
Antipsychotic drugs also block dopamine receptors in the pathways from the
substantia nigra in the midbrain to the head of the caudate nucleus in the basal ganglia
(see Fig. 1). Interruption of communication in this nigrostriatal pathway is thought to
account for Parkinson's syndrome-bradykinesia, rigidity, and tremor. In this neuronal
network, tolerance to the dopamine-blocking action of the drugs does seem to develop.
Chemically blocked dopamine receptors are initially underactive and then become
normally active to overactive, developing what is analogous to denervation supers en-
sitivity. Underactivity of the striatal dopamine receptors presumably results in parkin-
sonian signs; overactivity has been postulated to cause tardive dyskinesia, a syndrome
of abnormal involuntary movements, although this is being called into question, as we
discuss later. Acetylcholine and O-aminobutyric acid (GABA) mediate transmission
within adjacent connecting neuronal systems. Treatment of both Parkinson's syndrome
and tardive dyskinesia may involve drugs purported to act on any or all of these three
neurotransmitters (see Section III.G.1.a).
Researchers believe that a third important dopamine pathway, the tubero-infun-

dibular system, is affected by antipsychotic drugs. It projects from the arcuate nucleus
of the hypothalamus to the median eminence, where it acts to inhibit (directly or
indirectly) the release of prolactin from the anterior pituitary (see Fig. I). By blocking
dopamine neurotransmission in this system, antipsychotic drugs cause increased pro-
lactin secretion and hyperprolactinemia, producing unwanted effects and possible long-
term toxicity (see Section III.G.6). Although researchers have believed that tolerance
does not develop to the prolactin-elevating effect of antipsychotic drugs, it now appears
that at least partial tolerance may develop over a period of many months or even years.
Some have suggested that the elevation of prolactin in plasma could be used as an
index of the clinical effectiveness of antipsychotic drugs; however, maximum prolactin
elevation occurs at much lower doses than those usually required to treat a psychotic
Our recent understanding of dopamine autoreceptors provides a possible clue to
the behavior of the various major dopamine pathways. Autoreceptors are present on the
body or axon of a nerve cell and respond to the cell's own neurotransmitter by decreas-
ing its synthesis and release. Thus, they serve as part of a single-cell feedback inhibi-
tion loop. Interestingly, dopamine autoreceptors do not occur in the mesocortical
pathways but are present in the nigrostriatal and mesolimbic pathways.13 The former
does not exhibit tolerance to antipsychotic drugs, but the latter two do. Similarly, we
think tolerance does not develop to the antipsychotic actions of these drugs but does to
their extrapyramidal and prolactin-raising actions. Thus, it may be the presence or
absence of the dopamine autoreceptors within a system that determines whether toler-
ance to a clinical effect will or will not occur.J3
Our understanding of the mechanisms of action of antipsychotic drugs has ex-
panded in recent years with receptor subtype cloning, new imaging technology, and the
advent of agents with novel and atypical activities. Clozapine, synthesized in the late
1960s and introduced into clinical research in the 1970s (although not approved for
general clinical use in the United States until 1990), showed that it was possible for an
agent to be an effective antipsychotic and have minimal extrapyramidal and prolactin-
elevating effects. This suggested that clozapine might have more selective dopamine-
blocking properties in pathways such as the mesolimbic and mesocortical, with compar-
atively mild effects on nigrostriatal pathways (where Parkinson's syndrome is generated)
and the tubero-infundibular pathway (where prolactin secretion is regulated). Recent
research has shown that both typical antipsychotics and clozapine decrease spontaneous
activity in single dopamine cells in the mesolimbic pathways. Although typical anti-
psychotics similarly suppress spontaneous activity of neurons in the nigrostriatal path,
clozapine does not. 14 An anatomic correlate of this pharmacologic effect has emerged in
the past few years: patients treated with typical antipsychotic drugs have increased
caudate nuclei volumes, but caudate nuclei volume decreases when patients are switched
to clozapine.
Also of interest, clozapine decreases extracellular dopamine concentrations in rat
striatum and nucleus accumbens but increases extracellular dopamine in the medial
prefrontal cortex-a possible explanation for its diminution of negative as well as
positive schizophrenic symptoms. 15 Recent studies also have shown that while all
anti psychotics block the D2 and D3 subtypes of dopamine receptors, clozapine has a

uniquely high affinity for the D4 receptor, found to be increased in postmortem brains
of schizophrenics. 16 Scientists have shown that drugs that more selectively block DJ
receptors activate exploratory locomotor activity in laboratory animals, while D 2 -
receptor blockers, which include most classical neuroleptics, tend to produce more
sedation. 16 These findings might suggest means to alleviate both negative and positive
symptoms of schizophrenia. Neuroscientists also have paid attention to the fact that
clozapine has high affinity for the 5-HT2 type of serotonin receptor. Some of the newer
agents, such as risperidone (Risperdal®), which have comparatively low tendencies to
produce extrapyramidal reactions at low doses, have similarly high ratios of blocking
ability at the 5-HT2 receptor versus the D2 receptor.
Olanzapine (Zyprexa®) has a relatively high affinity for muscarinic, anti-
cholinergic, 5-HT2, and dopamine D 1, D2, and D4 receptors. Its blockade at a 2 recep-
tors is weaker than that of clozapine or risperidone. Like typical neuroleptics, but
unlike clozapine, it produces a modest rise in prolactin levels. Like olanzapine, sertin-
dole (Serlect®) is a potent blocker of cortical 5-HT2 and dopamine receptors. Other
atypical neuroleptics currently being studied have different properties (see Table 1).
The potent effects of antipsychotic agents on the autonomic nervous system
explain many of their adverse reactions. They block a-noradrenergic receptors, which
probably accounts for their hypotensive action, particularly on a postural basis. At
common clinical doses, low-potency antipsychotic drugs tend to be more potent at
a-adrenergic receptors and to produce a greater drop in orthostatic blood pressure. The
same low-potency agents also cause more sedation, which may reflect a-adrenergic
antagonism at brain receptors. Central effects on noradrenergic systems might be
involved in antipsychotic activity as well.t7
Antipsychotic agents block muscarinic acetylcholine receptors, producing other
autonomic effects (see Section III.G.2). As a group, antipsychotic drugs tend to be
much less potent as anticholinergic agents than the tricyclic antidepressants (see Chap-
ter 2). The most potent antimuscarinic drug among typical neuroleptics is the phe-
nothiazine thioridazine, which approaches the tricyclics in anticholinergic activity.
Clozapine is even more potent in vitro, but in humans it manifests such seemingly
cholinergic effects as increased salivation and enuresis. High-potency antipsychotic
agents are relatively weak blockers of cholinergic receptors.
Typical antipsychotic drugs produce other effects on neurotransmission, although
their significance is unclear. Similarly to tricyclic antidepressants, they block the reup-
take of norepinephrine, but this effect is probably outweighed by their tendency to
antagonize norepinephrine receptors. Other antipsychotic drugs also block serotonin
and histamine receptors; the latter might account for some of their sedative and appe-
tite-increasing tendencies. Other effects involve r3-adrenergic receptors and GABA,
serotonin, and peptide neurotransmitters.
The hypothalamus mediates various antipsychotic drug effects. In addition to
effects on prolactin, antipsychotic drugs inhibit the release of growth hormone (which
might have implications for their use in children). Some of their effects on autonomic
activity may also result from actions within the hypothalamus. Furthermore, they
impair the temperature-regulating mechanisms by making the normally homeothermic
mammalian system poikilothermic (i.e., the body temperature drifts toward that of the

TABLE 1. Types of Atypical Antipsychotic Drugsa

Selective D z antagonists Sulpiride


Partial D2 agonists SDZ HDC 912

B-HT 920

D I antagonists SCH 23390

SCH 39166

5-HTz antagonists Ritanserin

5-HT 3 antagonists Ondansetron


5-HTz' Dz antagonists Risperidone

ICI-204, 636
ORO 5222



a Adapted from Marder. 86

environment), which has resulted in cases of hypo- and hyperthermia. In the extreme,
antipsychotic drugs have created what has been termed the neuroleptic malignant
syndrome (see Section III.G.l.d), also thought to involve the hypothalamus. Increased
appetite, yet another hypothalamic effect, often results in weight gain.

D. Classes and Chemistry

Table 2 presents a list of antipsychotic drugs, grouped by chemical classes, that
are currently available by prescription in the United States. The first group, the phe-
nothiazines (chlorpromazine is the prototype), are three-ring (tricyclic) molecules
made up of two benzene rings linked by a sulfur and a nitrogen atom. The nitrogen
atom, which is attached to a carbon side chain, determines the phenothiazine subtype.
A straight chain of carbon atoms attached to the nitrogen indicates an aliphatic phe-
nothiazine (e.g., chlorpromazine). When the amino nitrogen at the end of the chain is
incorporated into a cyclic structure, the molecule is a piperidine (e.g., thioridazine). A
TABLE 2. Currently Available Antipsychotic Drugs"

Chemical structurel
Nonproprietary name Trade name Approximate potencyb Available as injectable representative agent

Chlorpromazine Thorazine® and others 1:1 Yes
s~ tH3
Triflupromazine Vesprin® 4:1 Yes

~ CH 3
Thioridazine Mellaril® and others 1:1 No

Mesoridazine Serentil® and others 2:1 Yes

Trifluoperazine Stelazine® and others 30:1 Yes

,,&; N ............... N
~ 1
s~ ~N,CH
I 3
Fluphenazine Prolixin® 50-100:lc Yesd

Perphenazine Trilafon® and others 10:1 Yes

Prochlorperazine Compazine® and others 7:1 Yes

Chemical structure!
Nonproprietary name Trade name Approximate potencyb Available as injectable representative agent

Chlorprothixene Taractan® and others 1:1 Yes
Thiothixene Navane® and others 25:1 Yes
.-!!P t;Yl
"1 C~'CH
I 3 3


%J cf'b 3

Haloperidol Haldol® 50:1 Yes

CI 0
Pimozide Orap® -' No

HN3 ~ --
o ~ b
Dibenzodiazepine F
Clozapine Clozaril® 2:1 No

Olanzapine Zyprexa® -' No
Loxapine Daxolin®, Loxitane®, 7:1 Yes
and others Nf"'.NCH3

Molindone Moban® 10:1 No
o CH 3
o V~CH3
Risperidone Risperdal® _e No
~CH3 H

,----QCH,CH,()( ~
~ -~F
Sertindole Serlect® _e No

I 0

0 H N /CH".)lNH
......... CH 2 V
a Adapted from Gelenberg. 87
hMiIligram equivalence to chlorpromazine.
"Oral only. Potency equivalences not established for long-acting injectable forms. Usual dosage for fluphenazine decanoate is 0.5-2.5 ml every 3 to 5 weeks.
d Available both as short-acting fluphenazine Hel injection and as long-acting enanthate and decanoate esters.
eNot established.

somewhat different cyclic structure results in the piperazine phenothiazines [e.g., tri-
fluoperazine (Stelazine® and others)]. When a piperazine phenothiazine has a terminal
hydroxyl (OH) group, esterifying it with a fatty acid results in a highly fat-soluble
hybrid that diffuses into the body's adipose tissue, releasing the parent phenothiazine
over a period of weeks. Examples are the enanthate and decanoate esters of the
piperazine phenothiazine fluphenazine (Prolixin® and others).
Replacing the nitrogen atom in the central ring with a carbon atom produces a
second group of effective antipsychotic substances, the thioxanthenes. They have either
aliphatic [e.g., chlorprothixene (Taractan®)] or piperazine [e.g., thiothixene (Navane®
and others)] structures and are chemically and pharmacologically similar to the phe-
The butyrophenones, developed by Janssen, appear structurally quite different
from the phenothiazines but are pharmacologically very similar to the piperazines. The
only butyrophenone labeled for antipsychotic use in the United States is haloperidol
(Haldol® and others), although the anesthetic agent droperidol (Inapsine® and others)
also appears to have antipsychotic properties. Closely related to the butyrophenones are
the diphenylbutylpiperidines, which are undergoing experimental investigation. This
class currently includes only pimozide (Orap®), an extremely potent blocker of the
neurotransmitter dopamine.
The dibenzoxazepine drugs are a fourth antipsychotic group with a three-ring
structure. The only member currently labeled as an antipsychotic agent in the United
States is loxapine (Loxitane® and others). [The demethylation of loxapine has resulted
in amoxapine (Asendin®), a compound with antidepressant properties.) Closely related
to these compounds are clozapine and olanzapine, which are dibenzodiazepines.
The sixth antipsychotic group, the dihydroindolones, are solely represented by
molindone (Moban®). Introduced into clinical use in 1994, risperidone is a novel
antipsychotic agent of a new chemical class, the benzisoxazole derivatives, with a
molecular structure distinctive from all other currently available antipsychotic classes.
Another structurally unique agent is sertindole, an imidazolidinone. Rauwolfia alka-
loids, such as reserpine and tetrabenazine (not available in the United States), are
infrequently used as antipsychotic agents and are primarily of historical interest.
For a tricyclic antipsychotic to be effective, three carbon atoms must lie between
the amino nitrogen and the nitrogen of the center ring. The addition of an electronega-
tive substituent to the benzene ring (e.g., CI, SCH3 , CF3 ) enhances its efficacy, whereas
the piperazine group on the side chain increases its potency. As a rule, molecules of
typical neuroleptics with greater milligram potency produce less sedation and
hypotension but more acute extrapyramidal reactions (see Table 3).

E. General Principles of Use

How do clinicians choose among the many typical antipsychotic agents? Most
importantly, they should review the patient's medication history. If the patient has
previously responded favorably to a given agent, that agent should be tried again. In
fact, if a patient has been taking a drug for maintenance therapy, and an acute exacerba-
tion occurs during a stressful period, then raising the dose of the drug may diminish the

TABLE 3. Spectrum of Adverse Effects Caused by Antipsychotic Drugs

Low-potency drugs
Fewer extrapyramidal reactions (especially thioridazine)
More sedation and postural hypotension
Greater effect on the seizure threshold and electrocardiogram (especially thioridazine)
More likely skin pigmentation and photosensitivity
Occasional cases of cholestatic jaundice
Rare cases of agranulocytosis

High-potency drugs
More frequent extrapyramidal reactions
Less sedation and postural hypotension
Less effect on the seizure threshold, less cardiovascular toxicity
Fewer anticholinergic effects
Occasional cases of neuroleptic malignant syndrome

symptoms. However, if a patient has responded unfavorably to a given drug because of

either lack of efficacy or unacceptable adverse effects, avoid that agent. On the other
hand, if a patient has had no prior experience with antipsychotic drugs, the experiences
of a family member could be used as a guide. Failing all of these suggestions, the
clinician is free to choose among the agents on the basis of his or her own experience
and the spectrum of adverse effects. In general, the high-potency antipsychotic drugs
are less sedating, produce less hypotension, and have less effect on the seizure thresh-
old, fewer anticholinergic effects, less cardiovascular toxicity, less weight gain, and
very little effect on the bone marrow and liver. On the negative side, high-potency
antipsychotic drugs have a greater incidence of acute extrapyramidal effects. The
converse is true of low-potency agents (see Table 3).
The introduction of atypical neuroleptics has expanded therapeutic options and, at
the same time, created practical challenges for clinicians, patients, and family members
in the choice of a neuroleptic. Since these drugs are most commonly used in patients
with chronic psychotic illnesses, the treatment systems in which they are prescribed
and administered tend to be in the public sector. Protocols for choosing agents tend to
be heavily influenced by the price of medication (as opposed to the overall "costs" of
treating a patient), which means that to receive one of the newer agents, a patient
commonly has to be tried first on treatment with a typical agent, which must be shown
to be either ineffective or poorly tolerated. However, this strategy may not be optimal
for patient care or even be the most cost-effective method.
If a drug causes fewer complications and has a higher rate of patient acceptance
and compliance, its true costs may be lower than those associated with drugs that cost
less at the pharmacy. Even more, if an agent has demonstrated superior efficacy and
can produce improved functioning and enhanced quality of life-all having been
demonstrated with clozapine-there is strong evidence that the decreased costs associ-
ated with these benefits will more than offset the increased price of medicine (and
attendant laboratory testing in the case of clozapine). Regarding clozapine, it is reason-

able to predict that in the not-distant future, the scheduled weekly venopunctures and
white blood cell counts required at the time of this writing may be decreased, and with
increased competition from newer agents, the manufacturer may decrease the actual
price of the medicine.
Rather than considering clozapine a "last-ditch" option for difficult-to-treat pa-
tients, there is an argument for its use as a first-line antipsychotic. True, it has a high
incidence of troublesome and even life-threatening complications. However, chronic
psychosis is a debilitating disease that robs the soul if not physical life itself. With
careful management, the life-threatening complications of clozapine can be kept to a
minimal risk, while the agent itself clearly has superior efficacy to traditional neurolep-
tics. Moreover, its capacity to treat the negative symptoms of schizophrenia, which
cause so much functional debility and suffering, are clearly superior to that of tradition-
al agents, along with much greater normalization of a patient's life.
There is some argument that newer atypical agents, with gentler side-effect pro-
files, should be tried before clozapine. The other atypical agents available in the United
States are risperidone and olanzapine. (Sertindole may be introduced soon.) Ris-
peridone has not been shown to be a superior antipsychotic nor to be as effective as
clozapine in patients who have not responded to traditional agents. In addition, ris-
peridone's pharmacologic profile resembles that of typical antipsychotics in that it
induces hyperprolactinemia, produces extrapyramidal symptoms at higher doses, and
has shown inconsistent effects on the negative symptoms of schizophrenia. As newer
atypical agents are introduced, they must be held to clozapine's standards as we
struggle to create a set of clinical priorities and a rank ordering of medications to try in
patients with schizophrenia and other chronic psychotic disorders.

F. Pharmacokinetics
The rate and completeness of absorption of drugs determine the rapidity and
intensity of onset of their clinical effects. Orally administered antipsychotic drugs are
variably absorbed. When given intramuscularly, the drugs produce higher (2-10 times)
and more reliable blood levels at a more rapid rate. Antipsychotic agents are highly
lipophilic [with the apparent exceptions of thioridazine and its metabolite mesoridazine
(Serentil®)] and bind tightly to protein and body membranes; they are poorly dialyzed.
As with other drugs that cross the blood-brain barrier, they also cross the placental
membrane to enter the fetal circulation and are transported into mammalian milk.
The half-life of a drug determines the time required to achieve a steady-state
concentration in the body (i.e., the point at which tissue concentrations become stable).
Steady state is reached after approximately four half-lives. Drugs with longer half-lives
may be administered less frequently, and they disappear from the body more gradually
when discontinued following chronic treatment. The half-life of an antipsychotic drug
in plasma following a single dose is usually 10-20 hr (see Table 4). However, the
drug's effects typically persist much longer, presumably because the brain half-life is
longer. When these drugs are administered chronically, the brain and body adipose
tissues become saturated; these supplies are then released and excreted very slowly. In
addition, when an antipsychotic drug has many metabolites (e.g., chlorpromazine), they
TABLE 4. Pharmacokinetics of Antipsychotic Agentsa

Time to peak Elimination Percent Tentative

after oral dose haIf-Iife protein therapeutic serum
Antipsychotic agent (hr) (hrY' binding Active metabolites conCentrationc

Chlorpromazine 2-4 16-37 (but active metabs 98-99 Many, unclear which of > 160 30-350 ng/mL?
probably much longer) metabs active
Fluphenazine (oral) 12-24 None? 0.5-3.0 ng/mL?
Mesoridazine -6 -25
Perphenazine 8-21 None 0.8-2.4 ng/mL?
Thioridazine 2 (concentrate); 7-42 96-99 Mesoridazine, 1-1.5 ng/mL?
4 (tablets) sulphoridazine
Trifluoperazine 3-6 17-18 Unknown Uncertain (possible demethylated Not established (1-2.3 ng/mL?)
and hydroxylated metabolites)
Haloperidol 2-6 12-40 -90 Reduced haloperidol? 5-20ng/mL?
Loxapine IIMI2PO 3.4 (loxapine) [9 (8-0H -90 8-0H loxapine, 8-OH amoxapine 30-100 ng/mL (pamlt loxapine
loxapine), 30 (8-0H plus 8-0H loxapine)
Clozapine 1-4 6-33 >90 Uncertain (demethylated Not established (100-600
metabolites may be active) ng/mL?)
Olanzapine 4-6 20-70 93 None
Molindone 0.5 1.5 None ?
Thiothixene 1-3 a = 3.5;~ = 34 Not studied None identified 2-15ng/mL?
Risperidone 3.6 90 9-Hydroxyrisperidone Not established
Sertindole 7-13 24-200 >99 Uncertain (dehydrosenindole and
hydroxylated metabolites may
be active)

"From Taylor and Caviness."

"Brain elimination half-life may be more germane clinically but none established; brain half-lives probably much longer, accounting for prolooged action after discontinuation.
<None well established.

are detected in the urine for many months after discontinuation of the drug following
chronic administration.
Attempts to correlate plasma concentrations of antipsychotic drugs with their
clinical efficacy have yielded contradictory results. This unsatisfying situation reflects
problems in research methodology, but also problems in biochemical assays and ques-
tions about active drug metabolites. Although it was once thought that a radioreceptor
assay measuring the level of dopamine-blocking activity in serum would replace direct
chemical measurements of drug concentrations, this is less clear-cut at present. 18 ,19
Available evidence hints at the existence of a curvilinear ("therapeutic window")
relationship for the butyrophenone haloperidol, a drug with no (or possibly one) active
metabolite(s), although the precise upper and lower boundaries on this "window"
remain to be delimited.
Regardless of any given antipsychotic drug's specific half-life, virtually all of
these agents can be administered in a once-daily dose, usually at bedtime. The clinical
effectiveness of this regimen presumably reflects the drugs' prolonged brain effects and
gradual elimination from the body. After chronic dosing, biological effects of these
agents persist for many weeks. After prolonged use of decanoate preparations, clinical
and biological effects can extend beyond 6 months!
As with most psychotropic drugs, lipophilic parent compounds are oxidized to
inactive hydrophilic metabolites largely within microsomal enzymes in the liver. These
metabolites are excreted primarily in the urine and to a lesser extent in bile, Metabo-
lism and excretion of antipsychotic drugs are greatest in healthy young people and less
at either end of the age spectrum.
The blood levels achieved with antipsychotic drugs following a standard dose
vary widely among individuals, which may account for some of the differences in
clinical effects.

G. Adverse Effects and Toxicity

1. Neurological
a. Extrapyramidal Syndromes. Antipsychotic drugs may cause four types of
extrapyramidal syndromes: acute dystonic reactions, akathisia, Parkinson's syndrome,
and, after longer-term use, tardive movement disorders. This section describes each
syndrome and its treatment.
i. Acute Dystonic Reactions. Acute dystonic reactions, including acute dyskin-
esias (i.e., abnormal involuntary movements of various types) and oculogyric crises,
typically occur during the early hours or days following the initiation of antipsychotic
drug therapy (seldom after 10 days) or after a marked dosage increment. Involuntary
muscle contractions are common, particularly about the mouth, jaw, face, and neck.
The symptoms are episodic and recurrent, lasting from minutes to hours. There may be
trismus ("lockjaw"), dystonia or dyskinesias of the tongue, opisthotonus (spasms of the
neck that arch the head backward), or eye closure. In oculogyric crises, there is a
dystonic reaction of the extraocular muscles, and gaze is fixed in one position.
Acute dystonic reactions are distressing, particularly to the patient, family mem-

ber, or clinician who is unfamiliar with them. They may be uncomfortable. They are
rarely dangerous. However, in rare cases there can be respiratory compromise with the
potential for a fatality. The diagnosis of acute dystonic reaction is usually not difficult if
it is clear that a patient has recently begun taking an antipsychotic drug or has had a
switch in the type or dosage of medication. At times, however, eliciting this informa-
tion may be difficult, especially with patients who are taking prochlorperazine supposi-
tories (and say they are not taking any tranquilizer pills), who do not wish to acknowl-
edge use of antipsychotic drugs, or who have sought antipsychotic drugs for illicit use.
Among the many neuropsychiatric syndromes that must be considered in the differen-
tial diagnosis of acute dystonic reactions are tetanus, seizures, and conversion reac-
The likelihood of producing acute dystonias is the greatest with the highest-
potency antipsychotic drugs, much less with the low-potency agents, and the least with
thioridazine and clozapine. Young people are at greater risk of developing this syn-
drome than the elderly. We used to think males experienced it more frequently than
females, but this is no longer clear. Patients who have previously experienced acute
dystonia are at greater risk of its recurrence upon rechallenge with a neuroleptic. When
the acute dystonic reactions are frequent or severe in a patient, it may be worthwhile to
assay the serum concentration of calcium, since rare cases of hypocalcemia have been
detected in this way.
The mechanism underlying acute dystonic reactions is unclear. One hypothesis is
that the syndrome reflects an acute increase in dopamine neurotransrnission in the basal
ganglia, which transiently supervenes the blockade of dopamine receptors brought
about by the same drugs. Although dystonic reactions do not occur in naturally occur-
ring Parkinson's disease (paralysis agitans), they are observed in postencephalitic
Parkinson's syndrome, which shares other features with antipsychotic drug-induced
extrapyramidal reactions.
Although the mechanism may be unclear, the treatment of an acute dystonic
reaction is straightforward, readily available, and usually dramatically successful.
Parenteral treatment is preferred for initiating drug therapy in more severe
cases, with intravenous being more rapid than intramuscular. The intravenous
injection of contraactive medication provides both immediate relief and confirmation
of the diagnosis. For this purpose, a number of classes of agents have been employed
with considerable success. An injectable anticholinergic antiparkinson drug such as
benztropine (Cogentin Gt and others), 1 mg, or biperiden (Akineton Gt and others),
2 mg, may be used. Other clinicians prefer to administer an antihistamine drug such
as diphenhydramine (BenadrylGt and others), 50 mg. Some doctors use a ben-
zodiazepine intravenously, such as diazepam (Valium® and others), up to 10 mg ad-
ministered over at least 2 minutes. This drug is relatively safe and does not add
additional anticholinergic effects. As with the injection of any other central depres-
sant compound, equipment for support of the airway should be immediately available
in case an emergency occurs, and care must be taken to avoid accidental intraarterial
Following immediate relief of acute dystonic signs, the clinician may wish to
begin oral administration of one of these drugs, using the lowest effective dose (for

TABLE 5. Antiparkinson Agents Used in the Treatment of Neuroleptic-Induced

Extrapyramidal Syndromesa

Usual dose range

Generic name Trade name Type of drug (mg per day) Injectable

Amantadine Symmetrel® and others Dopamine agonist 100-300 No

Benztropine Cogentin® and others Antihistamine and 1-6 Yes
Biperiden Akineton® and others Anticholinergic 2-6 Yes
Diphenhydramine Benadryl® and others Antihistamine and 25-100 Yes
Ethopropazine Parsidol® Antihistamine and 50-600 No
Orphenadrine Norflex® and others Antihistamine 50-300 Yes
Procyclidine Kemadrin® and others Anticholinergic 6-20 No
Trihexyphenidyl Artane® and others Anticholinergic 1-10 No

aAdapted from Wojcik.23

range, see Table 5). If this successfully prevents additional reactions, the contraactive
medication can usually be tapered and discontinued within several weeks. Recent data
suggest that coadministering an antiparkinson drug from the beginning of antipsychotic
drug therapy in the hope of avoiding an acute dystonic reaction may offer partial
protection to some patients. 19 The physician may want to consider this approach for
patients at highest risk (i.e., young patients receiving high-potency antipsychotic
agents) and for those in whom a reaction is likely to be clinically disruptive. For other
patients, it makes sense to avoid a drug that may be unnecessary, reserving treatment
until signs of dystonia appear. As atypical antipsychotics are used more often and
typical agents are used at lower doses, concomitant antiparkinson therapy should be
needed less often.
Physicians, nurses, family members, and anyone else who may observe a patient
receiving an antipsychotic medicine should be aware of the possible occurrence of a
dystonic reaction early in therapy. Treatment should be readily available. For most
patients, the risk of dystonic reactions appears to wane with continued antipsychotic
drug therapy.
ii. Akathisia. Akathisia is another extrapyramidal reaction associated with both
antipsychotic drugs and postencephalitic Parkinson's syndrome (rarely with Parkin-
son's disease). Akathisia is a symptom (i.e., subjective) defined as a compulsion to be
in motion. Patients describe an inner restlessness, an intense desire to move about
simply for the sake of moving. Persons who are virtually paralyzed by the akinesia of
postencephalitic Parkinson's disorder have been known to ask other people to move
their limbs, just to relieve this intense compulsion. Patients suffering from akathisia are
often observed to pace aimlessly, fidget, and be markedly restless. At times, akathisia
may cause a worsening of psychosis.20
Akathisia can occur early in the course of drug treatment, or it may not appear for
several months. It, too, appears to be more prevalent with high-potency drugs. The
natural course of akathisia is less clear than that of acute dystonic reaction: at times it

appears to wane, yet some patients are troubled by it for a long time. Its mechanism is
obscure but may involve a blockade of mesocortical dopamine receptors.
Treatment responses are variable; they include attempting to lower the dose of the
antipsychotic drug, switching to a lower-potency or atypical agent, or adding a contra-
active drug. The same drugs discussed for the treatment of acute dystonic reactions-
anticholinergic antiparkinson agents, antihistamines, and benzodiazepines-also may
be tried orally in cases of akathisia, although results are less universally successful (see
Section III.G.l.a). Whether the dopamine agonist antiparkinson drug amantadine
(Symmetrel® and others) may playa role in the treatment of this disorder is unclear.
Recent research has shown that the l3-adrenergic-blocking drug propranolol (Inde-
ral® and others) can effectively treat both subjective and objective manifestations of
akathisia while producing an acceptable profile of side effects; other l3-blockers may be
safe and effective as well. 19 Another agent that diminishes central noradrenergic activ-
ity, the antihypertensive clonidine (Catapres®), also may be effective but appears to
cause an unacceptable incidence of sedation and hypotension. 2o Akathisia is an impor-
tant syndrome to recognize, as it may severely complicate a patient's response to
antipsychotic drug therapy; perhaps most important, it makes patients extremely un-
iii. Parkinson's Syndrome. Parkinson's syndrome is characterized by a triad of
signs: tremor, rigidity, and akinesia (or bradykinesia). The tremor (by definition, a
regular and rhythmic oscillation of a body part around a point) is approximately 4-8
cycles per second and is greater at rest than during activity. A parkinsonian tremor is
often observed in the hands; the thumb rubbing against the pad of the index finger may
produce a characteristic "pill-rolling" appearance. The tremor also can involve the
wrists, elbows, head, palate, or virtually any body part. In neuroleptic-induced Parkin-
son's syndrome, the tremor is typically bilateral; unilateral tremors should raise ques-
tions about the etiology. Although tremor is very common and may be one of the earlier
signs in naturally occurring Parkinson's disease, in antipsychotic drug-induced Parkin-
son's syndrome it is less common than rigidity and akinesia, and it may not appear until
relatively late in the drug-related syndrome.
Rigidity is an increase in the normal resting tone of a body part. It is detectable
only by palpation on physical examination. (In other words, a patient does not look
rigid; he or she must feel rigid.) In testing for rigidity, the physician asks the patient to
relax completely and allow body parts to be manipulated without moving them. The
examiner then rotates the head on the neck, moves the major joints in the upper and
lower extremities, and raises each extremity to note the rapidity with which it falls by
gravity. An increased resistance to passive motion and a slow return from a raised
position denote the presence of rigidity. When tremor coexists with rigidity, the rigidity
may take on the feel of a "cogwheel." In extreme forms of the condition, rigidity may
mimic (or actually become) the waxy flexibility with sustained postures characteristic
of catatonia. 21 In antipsychotic drug-induced Parkinson's syndrome, rigidity tends to
be more common than tremor but less common than the third sign of the triad-
Akinesia is a disinclination to move in the absence of paralysis. Literally, akinesia
means the absence of motion, whereas bradykinesia, truer to the clinical reality in most

patients, implies a slowness of motion. The bradykinetic patient frequently shows a

masklike facies, with diminished expressiveness and less frequent eye blinking.
The bradykinetic patient turns his or her body "en bloc." Instead of turning first
with the eyes, then with the neck, followed in tum by shoulders, hips, and lower
extremities, the Parkinsonian patient turns as if he or she were one solid block of wood,
without joints. A typical stance includes a flexing of elbows and wrists together with a
stooped posture. The patient's gait is typically inclined forward, and he or she may
walk with small, rapid steps (a festinating gait, or marche apetit pas). In the extreme
forms, bradykinesia, as well as rigidity, may shade into catatonic immobility.21
In a less severe manifestation, the slowed movements of the parkinsonian patient
may appear primarily as apathy, boredom, and a "zombielike" appearance. If other
signs of Parkinson's syndrome are not prominent, this social akinesia may be misdiag-
nosed as depression. 22 Together with the major triad of Parkinson's syndrome (tremor,
rigidity, and bradykinesia), associated signs often include seborrhea and drooling.
Parkinson's syndrome usually occurs within weeks to months after the beginning
of antipsychotic drug therapy. Although tolerance develops in many patients, the disor-
der may persist and require ongoing treatment in others. Women and the elderly are
affected more commonly. Again, the high-potency agents appear more likely to pro-
mote this disturbance. Probably a late-occurring variant of Parkinson's syndrome is the
so-called rabbit syndrome, in which a rapid tremor of the mouth and jaw is reminiscent
of the facial expression of a rabbit. The rabbit syndrome appears to respond phar-
macologically similarly to other forms of Parkinson's syndrome. Parkinson's syndrome
reflects diminished dopaminergic input from the substantia nigra in the midbrain to the
head of the caudate nucleus in the basal ganglia (see Fig. 1). In this extrapyramidal
network, neuronal systems employing dopamine as a neurotransmitter are in balance
with other systems mediated by acetylcholine and GABA. In naturally occurring forms
of Parkinson's syndrome, the dopaminergic defect is a result of the destruction of
dopamine-containing neurons in the substantia nigra. This destruction may occur as a
result of viral infection (e.g., encephalitis) or poisoning (e.g., carbon monoxide, man-
ganese) or for unknown reasons (i.e., Parkinson's disease). In the drug-induced variety,
a blockade of the dopamine receptors within the caudate nucleus appears to be respon-
sible for the disturbance.
The treatment of Parkinson's syndrome reflects our understanding of the patho-
physiology. One approach is to decrease the dose of the antipsychotic drug, presumably
thereby decreasing the degree of dopamine blockade at the synaptic receptor. Alter-
natively, a less potent antipsychotic agent or one of the new, novel compounds may be
Among the original neuroleptics, thioridazine has the lowest incidence of parkin-
sonian reactions. The reason for this differential effect of antipsychotic drugs in caus-
ing Parkinson's syndrome may be related to variations in activity within the several
dopamine systems. An alternative explanation is that the agents that produce the fewest
extrapyramidal reactions, such as thioridazine, have the most potent effect in blocking
acetylcholine, thus acting as if they had inherent anti parkinson activity. Clozapine has
by far the lowest incidence of parkinsonism. At lower doses, risperidone also has a low
incidence of parkinsonism, but with increasing doses it begins to resemble more typical

neuroleptics. As new atypical agents come to the market, it will be important to assess
whether they are truly like clozapine or whether they are more like thioridazine in their
likelihood of generating extrapyramidal effects.
Pharmacological contraactive therapy for parkinsonism consists of attempts to
counterbalance the decreased dopamine neurotransmission by either blocking acetyl-
choline transmission [anticholinergic antiparkinson drugs include benztropine and tri-
hexyphenidyl (Artane® and others)] or by increasing dopamine neurotransmission
(e.g., by the use of amantadine and others). Other dopamine agonist drugs such as
L-dopa and bromocriptine (Parlodel®) may be useful in other forms of Parkinson's
syndrome but are seldom employed for the drug-induced variety. Table 5 lists available
drugs that are useful for the treatment of drug-induced Parkinson's syndrome. With
constant awareness of the pleomorphic and often subtle manifestations of extrapyrami-
dal reactions, the physician prescribes one of these contraactive drugs as needed for a
patient's comfort and optimal functioning, always seeking the lowest effective dose.
Doctors should periodically attempt to taper and discontinue a concomitant antiparkin-
son drug, although recent evidence has shown that many patients on maintenance
neuroleptic therapy require parallel maintenance with antiparkinson drugs; for most,
this can be carried out with impunity.20
With the exception of amantadine, all of the antiparkinson drugs listed in Table 5
have anticholinergic and/or antihistaminic properties. Among these drugs, trihex-
yphenidyl has a relatively short half-life, whereas benztropine has a relatively long
half-life. Amantadine is a dopamine agonist that has few anticholinergic effects. This
feature may make it preferable in patients particularly sensitive to anticholinergic
reactions. 23 For many patients, the strongly anticholinergic antiparkinson drugs can
impair memory, time perception, and cognition: these central nervous system anti-
cholinergic side effects also might prompt the use of amantadine. 24 Amantadine, 100
mg b.i.d. to 100 mg t.i.d., may be effective in some cases of drug-induced Parkinson's
syndrome, particularly the more severe variety, when anticholinergic antiparkinson
drugs have been ineffectual.25 Amantadine has a relatively long half-life (approx-
imately 24 hr). It is excreted in the urine unchanged and may lead to toxicity (including
psychiatric symptoms) in patients with impaired renal function. It is also particularly
hazardous and potentially lethal in overdose.
Antiparkinson drugs, as well as other chemicals with anticholinergic activity (e.g.,
belladonna-containing compounds), have been used for "recreational" purposes. 26 It
appears that the anticholinergic activity provides a mood-elevating effect for many
people, creating a feeling of euphoria or a "high." Some aficionados of the drug culture
actually use the drugs to create a toxic delirium, which they perceive as pleasurable.
Many schizophrenic patients treated with antiparkinson drugs become more attached to
these drugs than to their antipsychotic agents, possibly because of the relief of extra-
pyramidal symptoms or perhaps because of directly pleasurable effects. Trihexypheni-
dyl has been reported to be more commonly abused than other antiparkinson agents. If
this is true, it could be related to the pharmacology of the drug or simply to its
popularity and, therefore, availability. Anticholinergic toxicity and its treatment are
described in Section III.G.2.
For the last 10 to 15 years, there has been a growing awareness that lower

maintenance doses of antipsychotic drugs can provide effective prophylaxis against

recurrent psychotic episodes for many patients with schizophrenia. At the same time,
low doses are more tolerable, thereby increasing patient comfort and encouraging
compliance. As other side effects decrease in incidence and severity, so does the
incidence and severity of Parkinson's and other extrapyramidal reactions-and with
them the need for maintenance anti parkinson drug treatment. We then not only have
fewer antipsychotic-induced side effects but, concomitantly, no need for antiparkinson
drugs and their side effects. If the new generation of atypical neuroleptics, whose
hallmark is a relatively low incidence of extrapyramidal effects, lives up to its promise
of high efficacy and low toxicity, this long section on neuroleptic-induced Parkinson's
syndrome and its treatment may become a historical curiosity in future editions of this
and other texts on psychopharmacology.
iv. Tardive Dyskinesia and Tardive Dystonia. Tardive dyskinesia is characterized
by abnormal, involuntary, choreoathetotic movements involving the tongue, lips, jaw,
face, extremities, and occasionally the trunk. Not many years after the introduction of
antipsychotic agents, patients were described who displayed these movements after a
period of drug treatment. However, many questioned the association between these
signs and the use of antipsychotic drugs. Instead, they attributed the movements to
senile choreas, other brain disorders, schizophrenic stereotypies and mannerisms, or
adventitious mouth movements associated with poor dentition and dry mouth. For the
most part, clinicians generally ignored the existence of tardive dyskinesia.
By the early 1970s, however, tardive dyskinesia was recognized as more wide-
spread than hitherto appreciated, and a 1981 review even suggested that its prevalence
has progressively increased over the years since the introduction of antipsychotic
medication. 27 Moreover, epidemiologic evidence and clinical opinion now clearly
support a drug-related etiology for tardive dyskinesia, which has become a serious
concern among psychiatrists, other physicians, allied mental health professionals-and
Tardive dyskinesia typically involves orobuccolingual masticatory movements.
These may include lip smacking, chewing, puckering of the lips, protrusion of the
tongue, and puffing of the cheeks. A common early sign is wormlike movements of the
tongue. Other movements of the face can be observed, including grimacing, blinking,
and frowning. Another late-onset condition, presumably caused by antipsychotic drug
exposure, is tardive dystonia, characterized by slow, sustained, involuntary twisting
movements of limbs, trunk, neck, or face, including involuntary eye closure-
In younger patients, movements sometimes begin in the distal extremities and
may consist of rapid, purposeless, quick, jerky movements distally-chorea ("dance")-
or more sinuous, writhing movements proximally-athetosis. Occasionally, abnormal
movements involve the trunk and pelvis. The signs of tardive dyskinesia can range in
intensity from minimal to severe. Patients' awareness of the movements similarly
varies. Institutionalized chronic schizophrenic patients may deny even very severe
movements, whereas a highly functioning patient with a mood disorder could be
extremely troubled by the most minimal symptom. The movements may embarrass the

patient and interfere with important activities such as eating, talking, and dressing. In
rare instances, tardive dyskinesia can impair breathing and swallowing.
A metaanalysis of 21 studies found that exposure to neuroleptics multiplies the
risk of developing dyskinesia by 2.9 times. 29 It is generally believed that for tardive
dyskinesia to occur, a patient must have been exposed, more or less continuously, to an
antipsychotic drug for at least 3-6 months, although it is probable that a rare, highly
sensitive patient may develop this syndrome after an even briefer period. Presumably,
increasing exposure to antipsychotic drugs increases the risk of developing tardive
dyskinesia, although if a patient has not developed it after some length of time, he or
she is probably less vulnerable.
Patients who are older, female, black, or right-handed have an increased risk of
developing tardive dyskinesia on neuroleptic exposure. Other factors associated with
an increased risk are psychiatric diagnosis of mood disorder, negative schizophrenic
symptoms, a past history of early and severe acute extrapyramidal reactions, and type II
diabetes. Among typical neuroleptics, there is no evidence that anyone drug, chemical
class, or potency grouping is more or less likely to produce tardive dyskinesia. Recent
evidence suggests that clozapine is more likely than traditional neuroleptics to be
associated with a decrease in abnormal involuntary movements in a patient with preex-
isting movements following previous treatment with typical neuroleptic drugS. 30,31 In
addition, clozapine is less likely than typical neuroleptics to be associated with new-
onset cases of tardive dyskinesia in patients with previous histories of exposure to
traditional neuroleptics. 31 Coadministration of anticholinergic agents can exacerbate
existing movements of tardive dyskinesia but does not appear to increase the risk of it
developing. It appears that "drug holidays," i.e., regular, abrupt discontinuation of
antipsychotic drug therapy, do not reduce the risk of tardive dyskinesia. Although it has
been suggested that drug holidays may even increase the risk,27 this is controversial.
Prospective studies suggest a link between cumulative dosage and the development of
tardive dyskinesia, underscoring the wisdom of maintaining patients on the lowest
effective dosage over prolonged periods of time.
Estimates of the prevalence of tardive dyskinesia have varied widely-from 1%
to over 50% of patients currently taking neuroleptic drugs. 32 However, when the most
minimal cases are removed from consideration, an attempt is made at differential
diagnosis, and a careful and reliable screening procedure is used, the prevalence of
tardive dyskinesia in a group of patients maintained on antipsychotic drugs for a
variable period of time (i.e., a typical outpatient schizophrenic population) is likely to
be 20-30%.33 (A comparable group of patients not treated with antipsychotic drugs
may have a prevalence rate of 5%.) A more important question concerns the incidence
of tardive dyskinesia. A review of seven tardive dyskinesia incidence studies by Glazer
et al. in 1993 revealed average rates of new cases per year from approximately 4% to
11 %, with 5-year risks between 17.5% and 42.1 %.34 The prevalence of tardive dyston-
ia may be 1-2%.33
Although the prevalence of tardive dyskinesia is considerable (and actually may
be growing, as the population at risk takes neuroleptics for longer and longer periods),
for most patients the neurological manifestations are comparatively mild. Many pa-

tients deny even an awareness of the dyskinetic movements, and dysfunction is rare.
The most severely afflicted patients, and those likely to suffer the most, are patients
with tardive dystonia.
The movements of tardive dyskinesia may appear during treatment when a patient
is taking a constant dose of an antipsychotic, or they may initially appear when the
dosage of the drug is lowered or stopped entirely. Conversely, if movements are
present, increasing the dose of an antipsychotic drug can make the movements cease.
This latter phenomenon has been referred to as "masking" of the movements and gives
rise to the concept of "covert dyskinesia" -in other words, dyskinesia observed only
when an antipsychotic drug has been discontinued. 35 On the other hand, when an
antipsychotic drug is discontinued abruptly, a patient may show transient dyskinetic
movements that can disappear in a matter of days or weeks. This phenomenon has been
known as withdrawal dyskinesia, and it might conceivably indicate that a patient is
vulnerable to a persistent dyskinesia if the medication is reinstituted and continued.
Tardive dyskinesia itself is believed to be persistent, although evidence suggests that if
the movements are detected early and medication is discontinued, many patients will
show gradual improvement over time. 36
Anticholinergic drugs frequently exacerbate the movements of tardive dyskinesia
once they are present, and the administration of anticholinergic agents may "unmask"
latent movements. Conversely, discontinuing an anticholinergic drug in a patient with
tardive dyskinesia may improve the movements. Tardive dystonia, by contrast, might
actually improve in some cases on administration of an anticholinergic drug. Anti-
cholinergics do not seem to increase the likelihood that a patient will develop tardive
The most widely held hypothesis about the mechanism of tardive dyskinesia has
involved the nigrostriatal dopamine pathway. Chronic blockade of dopamine receptors
within the basal ganglia leads first to receptor underactivity, then to overactivity. In
other words, the initial receptor blockade from antipsychotic drugs leads to hypoac-
tivity at the receptor neurons. This state is believed to underlie the genesis of Parkin-
son's syndrome. Prolonged blockade, however, results in a situation analogous to
denervation supersensitivity. In fact, chronic neuroleptic administration to animals has
been shown to cause an actual multiplication in the number of postsynaptic receptor
sites. More recent magnetic resonance imaging (MRI) studies in humans have found
that prolonged exposure to typical neuroleptics increases the size of the caudate (possi-
bly correlating with the small duplication in receptor sites)37-39; switching a patient
from a typical neuroleptic to clozapine or possibly discontinuing a typical neuroleptic
results in a reduction in the increased size of the caudate. 40 A recent computerized
tomographic (CT) study found structural abnormalities in the caudate nucleus and
temporal lobe in patients with tardive dyskinesia-more marked in the left hemi-
sphere. 41 Thus, long-term administration of antipsychotic drugs leads ultimately to an
increase in activity at postsynaptic dopamine receptors. This state of excessive dopa-
minergic activity is believed to result in the abnormal movements of disorders such as
tardive dyskinesia and Huntington's chorea. 42
The "dopamine excess" hypothesis of tardive dyskinesia is not completely satisfy-
ing, however. For example, prolonged exposure to neuroleptic drugs results in dopa-

mine receptor supersensitivity in virtually all experimental animals, but only some
patients develop tardive dyskinesia. Also, such supersensitivity is less likely to develop
in older animals, but tardive dyskinesia is more likely to develop in older patients.
These and other objections have led many scholars to conjecture that the dopamine
supersensitivity hypothesis of tardive dyskinesia is too simplistic to be universally
applicable. Perhaps such a mechanism is valid in some patients, but alternative neuro-
transmitter systems that may go awry and contribute to the development of tardive
dyskinesia include those involving GABA, acetylcholine, and norepinephrine.
Tardive dyskinesia (and Huntington's chorea) is in many ways the opposite of
Parkinson's syndrome: the former is a disorder of extraneous movements, and the latter
a paucity of movements; the former is thought to reflect dopaminergic overactivity
(and relative cholinergic underactivity), whereas the latter is caused by too little dopa-
mine neurotransmission (and relatively too much cholinergic) (but see above reserva-
tions). Pharmacologic strategies that tend to make one better frequently make the other
Treatment approaches to tardive dyskinesia are based on our understanding (and
beliefs) about the nature of antipsychotic drugs and tardive dyskinesia. First, of course,
is primary prevention. This means avoiding unnecessary exposure of patients to anti-
psychotic drugs; in particular, avoiding the use of these agents to treat relatively benign
conditions or those that could respond equally well or better to other agents (e.g.,
anxiety, nonpsychotic depression). It also means using the lowest dose for the shortest
period of time when it is necessary to employ antipsychotic drug therapy. Of course,
most schizophrenic patients will require prolonged therapy with these agents, but the
clinician should find the lowest effective maintenance dose. Similarly, in the treatment
of patients with other chronic disorders such as mental retardation and organic brain
syndrome, the use of drugs should be minimized and constantly reevaluated.
Secondary prevention of a disorder means early detection. For tardive dyskinesia,
this entails routine screening and monitoring of patients for the presence of abnormal
movements. A standard neurological examination can be employed for this purpose, or
a clinician may want to use a specific examination procedure and rating scale such as
the Abnormal Involuntary Movement Scale designed by the National Institute of Men-
tal Health (see Fig. 2). It is best to perform an examination and note the results prior to
the initiation of antipsychotic drug therapy (or before too long into the course) and then
to repeat the examination every 6 to 12 months while the patient remains on drug
When the presence of abnormal involuntary movements is unequivocal, consider
the diagnosis (see Table 6). Is any other diagnosis likely? Are the movements charac-
teristic of tardive dyskinesia, or might they be tremors or other movements? Are they
chronic mannerisms and stereotypies characteristic of the underlying psychiatric disor-
der, or have these movements only appeared following chronic drug treatment? Family
history should be explored to rule out the presence of hereditary disorders of the central
nervous system such as Huntington's chorea or dystonia. The patient's history also
should be reexplored to establish a reasonable linkage between drug therapy and the
appearance of abnormal movements. Movements that inexorably progress are more
likely caused by a degenerative neurological disorder than by tardive dyskinesia. A

INSTRUCTIONS: Complete Examination Procedure CODE: a= None

(facing page) before making ratings. 1 = Minimal, may
MOVEMENT RATINGS: Rate highest be extreme normal
severity observed. Rate movements 2 = Mild
that occur upon activation one less 3 = Moderate
than those observed spontaneously. 4 = Severe

(Circle one l
FACIAL AND ORAL 1. Muscles of facial expression, a 1 2 3 4
MOVEMENTS e.g., movements of forehead,
eyebrows, periorbital area,
cheeks; include frowning,
blinking, smiling, grimacing.

2. Lips and perioral area, e.g.,

puckering, pouting, smacking. a 2 3 4

3. Jaw, e.g., biting, clenching, a 2 3 4

chewing, mouth opening,
lateral movements.

4. Tongue. Rate only increase in a 2 3 4

movement both in and out of
mouth, NOT inability to sustain

EXTREMITY 5. Upper (arms, wrists, hands, a 2 3 4

MOVEMENTS fingers). Include choreic
movements (Le., rapid, objec-
tively purposeless, irregular,
spontaneous), athetoid
movements (Le., slow, irreg-
ular, complex, serpentine).
Do NOT include tremor (Le.,
repetitive, regular, rhythmic).

6. Lower (legs, knees, ankles, a 2 3 4

toes), e.g., lateral knee move-
ment, foot tapping, heel
dropping, foot squirming, in-
version and eversion of foot.

TRUNK 7. Neck, shoulders, hips, e.g., a 2 3 4

MOVEMENTS rocking, twisting, squirming,
pelvic gyrations.

Figure 2. Abnormal involuntary movement scale (AIMS). Asterisk (*) denotes activated movements. From
Department of Health, Education, and Welfare, Public Health Service, Alcohol, Drug Abuse, and Mental
Health Administration, National Institute of Mental Health.
GLOBAL 8. Severity of abnormal movements. None, normal 0
Mild 2
Moderate 3
Severe 4

9. Incapacitation due to abnor- None, normal 0

mal movements. Minimal 1
Mild 2
Moderate 3
Severe 4

10. Patienfs awareness of abnor- No awareness 0

mal movements. Rate only Awareness, no distress 1
patient's report. Aware, mild distress 2
Aware, moderate distress 3
Aware, severe distress 4

DENTAL STATUS 11. Current problems with teeth No 0

and/or dentures. Yes 1

12. Does patient usually wear No 0

dentures? Yes 1

Either before or after completing the Examination Procedure, observe the patient unobtrusively, at
rest (e.g., in waiting room).

The chair to be used in this examination should be a hard, firm one without arms.

1. Ask patient whether there is anything in his/her mouth (i.e., gum, candy, etc.) and if there is,
to remove it.
2. Ask patient about the ~ condition of his/her teeth. Ask patient if he/she wears
dentures. Do teeth or dentures bother patient D.QW?
3. Ask patient whether he/she notices any movements in mouth, face, hands, or feet. If yes,
ask to describe and to what extent they ~ bother patient or interfere with his/her
4. Have patient sit in chair with hands on knees, legs slightly apart, and feet flat on floor. (Look
at entire body for movements while in this position.)
5. Ask patient to sit with hands hanging unsupported. If male, between legs; if female and
wearing a dress, hanging over knees. (Observe hands and other body areas.)
6. Ask patient to open mouth. (Observe tongue at rest within mouth.) Do this twice.
7. Ask patient to protrude tongue. (Observe abnormalities of tongue movement.) Do his twice.
*8. Ask patient to tap thumb, with each finger, as rapidly as possible for 10 to 15 seconds,
separately with right hand, then with left hand. (Observe facial and leg movements.)
9. Flex and extend patient's left and right arms (one at a time). (Note any rigidity.)
10. Ask patient to stand up. (Observe in profile. Observe all body areas again, hips included.)
*11. Ask patient to extend both arms outstretched in front with palms down. (Observe trunk,
legs, and mouth.)
*12. Have patient walk a few paces, tum, and walk back to chair. (Observe hands and gait.) Do
this twice.

TABLE 6. Tardive Dyskinesia: Differential Diagnosis

mental Family
Disorder disturbances history Laboratory tests

Dyskinesias on withdrawal from neuroleptic Yes No None

Schizophrenic stereotypies and mannerisms Yes No None
Spontaneous oral dyskinesias associated with aging Yes No None
(including Meige syndrome)"
Oral dyskinesias related to dental conditions or No No None
Torsion dystonia No Yes None
Idiopathic focal dystonia [oral mandibular dystonia, No No None
dystonia, blepharospasm, spasmodic "habit spasms"
Huntington's diseasec Yes Yes CAT scan
Wilson's disease Yes Yes Serum copper and
Magnesium and other heavy metals Yes No Specific for metals
Fahr's syndrome or other disorders with calcification No Yes Skull X-rays
of the basal ganglia
Extrapyramidal syndromes following anoxia or Yes No None
Rheumatic (Sydenham's) chorea ("Saint Vitus' dance") No No None
Drug intoxications-L-dopa, amphetamines, Yes No For specific agents
anticholinergics, antidepressants, lithium, phenytoin
CNS complications of systematic metabolic disorders Yes Possible For specific agents
(e.g., hepatic or renal failure, hyperthyroidism,
hypoparathyroidism, hypoglycemia, vasculitides)
Brain neoplasms (thalamic, basal ganglia) Yes No CAT scan, other
brain scans,
EEG, etc.

aMeige syndrome is a disorder of middle age characterized by progressive oral, lingual, and buccal dystonia together with
blepharospasm. The movements are indistinguishable from those of tardive dyskinesia, but patients with Meige syndrome
need not have had antedating antipsychotic drug exposure, and Meige syndrome is a progressive disorder.
"Be sure to ask patients about the state of their mouth and teeth and whether they have any gum in their mouth when you are
examining them for tardive dyskinesia.
<The abnormal movements of Huntington's disease are primarily chorea with little dystonia or athetosis. The movements are
generalized, producing a ''fidgety'' appearance in the early stages. Although movements of tardive dyskinesia are more
stereotyped and abnormal, the movements of Huntington's disease appear to be normal movements at an increased
frequency. Patients with tardive dyskinesia have dyskinesias other than chorea. In general, patients with Huntington's
disease have more trouble keeping their tongues out of their mouths, whereas tardive dyskinesia patients may have trouble
keeping their tongues in their mouths.

neurological examination should reveal only abnormal involuntary movements and

specifically those characteristic of tardive dyskinesia; other neurological systems, such
as sensory or pyramidal, should not be involved. A physical examination should search
for the presence of ancillary signs of other extrapyramidal disorders such as the Kay-
ser-Fleischer rings of Wilson's disease, Finally, laboratory testing can rule out other
disorders-e.g., copper and ceruloplasmin levels in Wilson's disease, radiological
evidence of caudate degeneration in Huntington's disease-if these conditions are

suspected. At times, the psychiatrist may wish to seek consultation from a neurological

Tips for the Diagnosis of Tardive Dyskinesia

1. History
a. Exposure to antipsychotic drug
b. Nonprogressive disorder
c. Rule out family history of movement disorders
2. Physical examination
a. Movements characteristic of tardive dyskinesia
b. Rule out neurological signs in other systems (e.g., cerebellar, sensory,
Not helpful
1. What makes movements better (e.g., sleep, relaxation) or worse (e.g., tension,
psychological "gain")
2. Whether the patient can voluntarily inhibit movements
3. Whether movements are worsened or "unmasked" by a decrease in the anti-
psychotic drug dose
4. Laboratory tests (Although laboratory tests are not helpful to rule in tardive
dyskinesia, they can be helpful in ruling out other disorders.)

If the diagnosis of tardive dyskinesia appears firm, what is the course of treat-
ment? Raising the dose of the antipsychotic drug might succeed in partially (or even
completely) suppressing the movements, but the clinician should avoid this strategy if
possible. Available evidence suggests that continuing neuroleptic therapy, although
unlikely to exacerbate the degree of dyskinesia, may enhance the probability that the
dyskinesia will become irreversible. 43 •44 Of course, if continuing an antipsychotic drug
is necessary to control the psychosis or prevent episodic relapses (as is often the case),
then the clinician must recommend this course, and it is up to the patient (or guardian)
to decide whether to consent.
If abnormal movements are detected early, and it is clinically practical to taper
and discontinue the antipsychotic medication, it is more likely that the dyskinesia will
ultimately improve (although it may be exacerbated temporarily). Although this is ideal
from the standpoint of the movement disorder, for most schizophrenic patients the
trade-off in exacerbation of psychosis will not be worth the price. Schizophrenic
psychosis usually is worse than tardive dyskinesia, and most often the clinician, patient,
and family will decide to continue drug therapy, always at the lowest effective dose.
If the movements continue and are troublesome (and irrespective of whether the
patient remains on antipsychotic drug therapy), the clinician will want to alleviate as
much of the patient's discomfort as possible. Unfortunately, there is no standard and
accepted treatment for tardive dyskinesia or dystonia. Therapy with a benzodiazepine
drug such as diazepam or clonazepam (Klonopin®) may help some patients, either by
its sedative or muscle-relaxing properties or possibly via its role in increasing GABA-
ergic tone.

Just as Parkinson's syndrome often responds to strategies that increase dopamine

neurotransmission, tardive dyskinesia may be alleviated by treatments that diminish
dopaminergic tone. Thus, a dopamine-depleting agent such as reserpine or tetra-
benazine may be useful. (As mentioned earlier, the use of dopamine blockers such as
antipsychotics themselves is generally avoided.)
Another treatment approach focuses on the cholinergic system. Much as Parkin-
son's is treated with anticholinergic drugs, so tardive dyskinesia has been approached
through the use of cholinomimetics. Deanol was earlier employed for this purpose, but
its status as a cholinergic agonist is in question, and its efficacy in tardive dyskinesia is
doubtful. Choline chloride has been employed in this fashion, but initial positive results
have not been sustained, and its many unwanted effects make it clinically troublesome.
Phosphatidylcholine, contained in the naturally occurring lipid lecithin, also has been
used for this purpose. Our own group has found lecithin to produce statistically signifi-
cant but clinically minimal benefits.45 If a patient is already taking an anticholinergic
drug, the abnormal movements of tardive dyskinesia may be diminished if the anti-
cholinergic is withdrawn; the "downside," however, is that parkinsonian signs may
emerge or become worse.
Another agent that has been used in tardive dyskinesia with varying degrees of
success is baclofen (Lioresal® and others), which may operate via a GABAergic
mechanism. Many other drugs (e.g., propranolol) have been administered experimen-
tally in patients with tardive dyskinesia, but unfortunately, no consensus has emerged.
Research in recent years suggests a possible role for vitamin E in diminishing the
abnormal and involuntary movements of tardive dyskinesia and theoretically even in
preventing them.46
As with tardive dyskinesia, treatments for tardive dystonia (Table 7) are unsatisfy-
ing. Some patients may improve with anticholinergic drug treatment, and others experi-
ence symptomatic relief from reserpine or tetrabenazine (not available in the United
States). Clozapine, which by itself may not engender new cases of tardive dyskinesia or
dystonia, actually might improve the abnormal movements. Injections of botulinum
toxin may be therapeutic for severe blepharospasm or other debilitating dystonias.

Steps in Diagnosis and Management of Tardive Dyskinesia and Dystonia

Preliminary procedure
1. Complete a baseline neurological examination.
2. Inform the patient and family about the risk of tardive dyskinesia.
Primary prevention
1. Avoid unnecessary exposure to antipsychotics.
2. Use lowest dose of antipsychotic for shortest time.
Secondary prevention
1. Routinely screen for dyskinetic movements [routine neurological examina-
tion or rating scale such as Abnormal Involuntary Movement Scale (AIMS)].
a. Administer at the beginning of treatment.
b. Administer every 6 to 12 months.

TABLE 7. Comparison of Tardive Dystonia and Tardive Dyskinesia

Tardive dystonia Tardive dyskinesia

Strikes younger Strikes older

Strikes sooner in the course of neuroleptic Strikes later in the course of neuroleptic treatment
Poor prognosis Variable prognosis
More males More females (?)
Patients with mood disorders may be more Patients with mood disorders may be more
susceptible susceptible
Anticholinergics may improve condition Anticholinergics usually worsen condition

2. Consider lowering dosage or discontinuing antipsychotic drug if early signs

1. Confirm diagnosis by reviewing symptoms, drug history, and neurological
2. Can antipsychotic be discontinued? If not, consider lowering dosage.
3. Treatment options:
a. Adjunctive benzodiazepines (e.g., diazepam, clonazepam).
b. Switch antipsychotic to clozapine-if an antipsychotic is required.
c. Botulinum toxin for severe dystonias
d. Vitamin E?
4. Refer patient to specialized center.

h. Sedation. It was initially believed that sedation was important for the effec-
tiveness of antipsychotic drugs, but this no longer appears true. Drugs can have a major
impact on the primary symptoms of psychosis such as hallucinations and thought
disorder without producing somnolence. As mentioned previously, the low-potency
antipsychotic drugs such as chlorpromazine, thioridazine, and chlorprothixene tend to
be more sedating. Clozapine is especially sedating. If around-the-clock sedation is
desired, these drugs can be administered several times daily during initial therapy. If
nighttime sedation is indicated, the drugs may be administered once daily at bedtime.
Tolerance to sedation tends to develop over a matter of days or several weeks.

c. Seizures. Antipsychotic drugs, particularly the low-potency agents, lower the

seizure threshold. For most patients, this is seldom a problem. However, an occasional
patient without a history of epilepsy will experience a seizure during treatment with
very high doses (or a particularly rapid increase in dose) of a low-potency agent. The
problem is more likely in a patient with marginally controlled seizures or in someone in
a state of heightened vulnerability, such as withdrawal from sedative-hypnotic drugs
or alcohol. Each case must be handled individually, but consideration should be given
to lowering the dose of the antipsychotic drug, changing to a high-potency agent,

and/or adding (or increasing) an anticonvulsant. Clozapine lowers the seizure threshold
in a dose-dependent fashion. If a patient is to be treated with very high doses of
clozapine, or if a seizure occurs during clozapine treatment but the clinician and patient
decide that the antipsychotic benefits of clozapine warrant its continuation, coad-
ministration of an anticonvulsant medication may be indicated. In such cases, di-
valproex may be the most logical choice. Knowledgeable clinicians in the United
States tend to avoid coadministering carbamazepine with clozapine for fear of increas-
ing the risk of bone marrow suppression.

d. Neuroleptic Malignant Syndrome. The neuroleptic malignant syndrome

(NMS) is a serious disorder consisting of fever, muscular rigidity, and stupor, which
develops in association with antipsychotic drug therapy.45 Other features include auto-
nomic dysfunction (e.g., increased pulse, respiration, sweating, and blood pressure
instability) and occasionally respiratory distress. Laboratory findings commonly indi-
cate leukocytosis and elevated serum creatine phosphokinase (CPK). This syndrome
typically develops explosively over a 24- to 72-hr period beginning anywhere from
hours to months after initial drug exposure. A patient may have received prior treat-
ment with antipsychotic drugs without showing this pattern.
The neuroleptic malignant syndrome has been associated with various antipsy-
chotic drugs but is more prevalent with high-potency agents, particularly when the dose
is being raised quickly. Haloperidol has been associated with a disproportionate share
of case reports. Both sexes may be affected at any age, but among reported cases,
young adult males predominate. The incidence of this disorder is probably around 0.1 %
of patients exposed to neuroleptics. 47 Mortality from NMS may be 10-20%, but when
the patient survives, sequelae are rare. 48
The mechanism of NMS has yet to be elucidated. Most theorizing has focused on
the basal ganglia and hypothalamus, where blockade of dopamine neurotransmission is
held responsible (withdrawal of dopamine agonists, as well as institution of dopamine
antagonists, has been associated with precipitation of NMS). An alternative site for the
pathology of NMS could be peripheral muscles, analogous to the malignant hyperther-
mia of anesthesia.
Levinson and Simpson have underscored the heterogeneity of cases of neurolep-
tic-induced extrapyramidal signs plus fever. 48 They emphasize that many such patients
have diverse and independently treatable medical or neurological conditions, which
require a careful differential diagnosis.
Since a patient may develop NMS on one exposure to a neuroleptic, yet not on
another exposure, it is likely that additional factors may playa role in the pathogenesis.
Dehydration may be one such factor, and clinicians would do well to attend to hydra-
tion status during antipsychotic therapy, particularly in warmer weather when patients
are exerting themselves, or if they have inadequate oral intake.
Appropriate management requires immediate discontinuation of antipsychotic
drugs and institution of supportive measures (e.g., lowering body temperature, hydra-
tion) along with ruling out infection and other possible metabolic abnormalities. Al-
though no contraactive treatment has been clearly proved safe and effective, most
attention in recent years has focused on dantrolene (Dantrium®) and bromocriptine.

How to treat subsequent psychotic episodes in a patient with a history of NMS is

often a dilemma. 49 Low-potency neuroleptics may be safer than those of higher poten-
cy, although even clozapine and risperidone have been associated with NMS. When a
patient has a history of this syndrome, the clinician should first verify the past diag-
nosis. If it is validated, the clinician should consider non-neuroleptic treatments such as
ECT or, in the case of a mood disorder, lithium (Eskalith® and others). When a
neuroleptic is thought to be required, one of lower potency should be considered, and
the dose raised gradually. It is of particular importance that all previous signs of NMS
should have ended completely at least 2 weeks before the reintroduction of an anti-

2. Anticholinergic
a. Peripheral. Antipsychotic drugs block a SUbtype of cholinergic receptor
known as muscarinic. This type of receptor, which responds to interneuronal release of
acetylcholine, is located on postganglionic neurons of the parasympathetic branch of
the autonomic nervous system (as well as autonomic ganglion cells and certain cortical
and subcortical neurons). Thus, the atropinelike action of antipsychotic drugs antago-
nizes the effects of the parasympathetic nervous system. The most strongly anti-
cholinergic among the traditional antipsychotic drugs is thioridazine, which is almost
as potent as the cyclic antidepressants. High-potency antipsychotic drugs (e.g., flu-
phenazine, haloperidol) are comparatively weak in their atropinelike action.
In the skin, anticholinergic drugs can produce warmth, flushing, and dryness; in
the eye, dilated pupils, difficulty with visual accommodation, and increased intraocular
pressure; in the mouth, dryness; in the lungs, drying of secretions; in the heart, in-
creased rate; in the stomach, decreased acid secretion; in the bowel, diminished mo-
tility with potential constipation; in the urinary tract, smooth muscle slowing, which
can lead to delayed urination; in the penis, delayed (or retrograde) ejaculation.
In general, the elderly are more sensitive to atropinelike effects than are younger
patients. Obviously, certain illnesses are exacerbated by anticholinergic actions of
drugs. An abrupt attack of narrow-angle glaucoma is a rare possibility that can occur in
a predisposed individual. Fortunately, the more common open-angle glaucoma, parti-
cularly when controlled by drugs, is less likely to be made worse by drug administra-
tion. Individuals prone to dental problems may suffer from diminished salivation.
Patients with respiratory problems could be adversely affected by diminished pulmo-
nary secretions. Similarly, those with cardiac disorders may be compromised by tachy-
cardia as well as by more direct cardiotoxic effects, which are discussed below. Indi-
viduals with gastrointestinal disturbances, including those who have recently had
abdominal surgery, may be adversely affected by the diminished bowel motility. Like-
wise, a man with prostatic enlargement could find urination more difficult during
treatment with an anticholinergic drug.
Some degree of tolerance develops to these effects over weeks and months (and
conversely, rebound can occur on drug withdrawal). When peripheral anticholinergic
activity of antipsychotic drugs is a problem, the clinician may wish to lower the dose of
the drug or switch to a drug with less atropinic activity (i.e., a higher-potency agent). At

times, symptomatic treatment may be offered, such as sugarless gum and lozenges for
dry mouth or a stool softener or mild laxative for constipation. In occasional patients,
the use of a peripherally active parasympathomimetic agent such as bethanechol
(Urecholine® and others) may be indicated.
In its anticholinergic activity, clozapine again is something of an anomaly. Its
binding potency at the muscarinic receptor is higher than that of older antipsychotic
drugs, even thioridazine. However, one of the common side effects of clozapine is
hypersalivation: patients typically awaken with a pillow soaked with saliva. Patients
have also reported experiencing enuresis in association with clozapine treatment. 50
Helpful antidotes were oxybutynin (Ditropan® and others) and desmopressin
(DDAVP®). Risperidone has minimal anticholinergic activity.

b. Central. Presumably through their action in blocking central muscanmc

cholinergic receptors, antipsychotic drugs can at times produce memory difficulties,
confusion, and, in the extreme, delirium.

c. Serious Toxicity. An antipsychotic agent alone, particularly a higher-potency

compound, is unlikely to produce serious atropine-type toxicity. However, the more
potent antimuscarinic antipsychotic drugs (especially thioridazine), or the combination
of an antipsychotic drug with an anticholinergic antiparkinson agent, tricyclic anti-
depressant, or other antimuscarinic compound, or a drug overdose can lead to serious
toxicity, especially in the elderly and others vulnerable to these conditions. The patient
suffering from an atropineiike delirium is typically confused, disoriented, and agitated.
Pupils are large (although responsive), mucous membranes are dry, and skin is hot and
flushed. Tachycardia and markedly diminished bowel sounds are common. (These
signs have been summarized as follows: Dry as a bone, Red as a beet, Blind as a bat,
Mad as a hatter.)
Intravenously administered physostigmine (Antilirium®), 1-2 mg in a single,
slow injection, can be both diagnostic and therapeutic, as it will reverse both the
peripheral and central signs of atropine-type toxicity. (If successful, physostigmine can
be repeated 1M every few hours until delirium clears.) Physostigmine is a centrally
active inhibitor of acetylcholinesterase, and, as such, it increases the amount of acetyl-
choline available to cholinergic receptors, thus circumventing the drug's blockade. The
infusion of physostigmine should be slow, with careful cardiac monitoring and readily
available means to support respiration, should this prove necessary. A physician who
wishes to administer physostigmine should be familiar with the pharmacology of this
drug and the hazards of cholinergic excess (which can be reversed by the administra-
tion of atropine). As a final note, physostigmine should be administered with the
greatest caution to a patient who is showing cardiovascular instability as a result of a
drug overdose or other medical problems51 (see Chapter 2).

3. Cardiovascular and Respiratory

a. Hypotension. Orthostatic hypotension during antipsychotic drug administra-

tion has been attributed to a combination of hypothalamic actions and peripheral
a-adrenergic blockade. Some degree of tolerance may develop. This reaction is more

common with the low-potency drugs (e.g., chlorpromazine) and may be more of a
problem with the elderly or with patients with preexisting postural hypotension and
vascular instability such as those undergoing a sedative-hypnotic withdrawal reaction.
It is likely that parenteral administration of an antipsychotic drug may provoke more
severe hypotension than oral ingestion of the same drug. Clozapine produces substan-
tial postural hypotension, but risperidone does not.
If postural hypotension does develop, it usually can be managed by keeping the
patient horizontal. The next step, if necessary, is the administration of intravenous
fluids to expand the vascular volume. If these two maneuvers are not effective, then a
pure a-adrenergic pressor agent should be administered. Metaraminol (Aramine® and
others) will usually suffice, although norepinephrine (Levophed® and others) also may
be considered. The use of a mixed a- and l3-stimulating drug such as epinephrine
(Adrenalin® and others) could lead to a paradoxical drop in blood pressure, since the a
receptors are blocked, and the unopposed 13 stimulation can promote further hypoten-
sion. Similarly, the use of a primarily l3-stimulating drug such as isoproterenol (Isup-
rel® and others) should be avoided.

h. Cardiac. Antipsychotic drugs have both antiarrhythmic and arrhythmogenic

effects on the heart. Antagonism of sympathetic activity in the hypothalamus and a
local anesthetic property that stabilizes the cardiac cell membrane [similar to the effects
of lidocaine (Xylocaine® and others)] may cause the antiarrhythmic effect. In addition,
antipsychotic drugs have direct quinidinelike effects on the myocardium and cardiac
conduction system and affect electrolyte balance. Explanations of the arrhythmogenic
action of phenothiazines and other antipsychotic drugs similarly include both central
activity and a direct effect on the myocardium. Pathologists have detected microscopic
and ultramicroscopic lesions in the cardiac tissue of patients who died suddenly during
the course of phenothiazine therapy, suggesting a direct toxic effect.
Whether antipsychotic drugs are protective or toxic to the heart probably depends
on the agent, the dose, and the underlying state of a patient's cardiac physiology. Low-
potency antipsychotic agents, especially thioridazine and clozapine, tend to be more
cardiotoxic. On the electrocardiogram (EKG), increased heart rate, prolongation of the
QT and PR intervals and T wave and depression of the ST segment are occasionally
observed, particularly with thioridazine, clozapine, and pimozide (Orap®). For most
patients, these effects are not clinically troublesome. Occasionally, however, a life-
threatening arrhythmia, such as torsade de pointes (polymorphic ventricular tachycar-
dia), can occur. Torsade de pointes also has occurred when high doses of haloperidol
have been administered IV. 52 Patients with preexisting cardiac disease should be moni-
tored carefully (e.g., clinical examinations, vital signs, EKGs), and low-potency drugs
(especially in excessive doses and particularly thioridazine, clozapine, and pimozide)
should be avoided.
Overdoses of antipsychotic drugs (especially thioridazine) can cause cardiac ar-
rhythmias. Because antipsychotic agents have quinidinelike actions, arrhythmias
caused by overdoses should not be treated with quinidine or related type 1 antiar-
rhythmic drugs [procainamide (Pronestyl® and others) and disopyramide (Norpace®
and others)]. In the treatment of cardiac arrhythmias that result from overdoses with

phenothiazines and related drugs, antiarrhythrnics such as lidocaine, phenytoin (Dilan-

tin® and others), and propranolol may have a role. The early use of transvenous pacing
also has been recommended.
Combinations of antipsychotic drugs and antidepressants can produce additive
cardiotoxic effects. In particular, combinations of thioridazine, clozapine, pimozide,
and tricyclic antidepressants, especially amitriptyline (Elavil® and others), are best
avoided. Occasional cases of sudden death reported in patients receiving antipsychotic
drugs are difficult to interpret and probably result from various causes. However,
episodes of potentially fatal ventricular arrhythmias are a possible factor.
Risperidone has vascular a-adrenergic antagonistic activity and causes such dose-
related effects as hypotension and reflex tachycardia. It and sertindole may also have
proarrhythrnic effects, as they can prolong the QT interval in some patients. The
severity of sertindole's cardiac effects will only be established with time and clinical
experience. For now, sertindole should be avoided in patients with preexisting cardiac

4. Ocular

As already noted, the anticholinergic effects of antipsychotic drugs can affect the
eye. Increased mydriasis can make a patient more light-sensitive. Interference with
visual accommodation may result in complaints of blurred vision. Because of the rare
possibility of precipitating an attack of angle-closure glaucoma, patients should be
asked prior to treatment with any anticholinergic agent whether they have a history of
visual symptoms-such as eye pain, blurring, and halos-that could suggest previous
narrow-angle episodes. Open-angle glaucoma is less likely to be a problem, but the
patient should be managed conjointly with an ophthalmologist.
Prolonged treatment with high doses of low-potency antipsychotic drugs has been
associated with the deposition of pigment in the lens, cornea, conjunctiva, and retina,
often together with skin pigmentation. Except in extreme cases, these are unlikely to
interfere with vision. However, if it is necessary to expose a patient to high doses of a
low-potency agent for long periods of time, periodic ophthalmologic examinations
would be worthwhile, perhaps annually.
Of greater clinical significance is the pigmentary retinopathy that may accompany
treatment with thioridazine or other phenothiazines with piperidine side chains. This
disorder, which can permanently impair vision, is unlikely if doses of thioridazine do
not exceed 800 mg per day, but even this is relative. Thus, thioridazine is the one
antipsychotic agent that should be considered to have an absolute "ceiling dose" -800
mg daily-which must not be exceeded for even brief periods of time. (If a patient
does require high doses of an antipsychotic drug, it is best to change to a different one.)
Although phenothiazines with aliphatic side chains, such as chlorpromazine, are much
less likely to cause ocular complications, there have been rare reports of chlor-
promazine-induced retinopathy, even at low doses. 53
Other effects of chlorpromazine on the eye are minimal. In patients who have been
receiving chlorpromazine at 500 mg/day or more for at least 3 years, however, granular
deposits have been noted on the corneal endothelium and the lens capsule. 54 These

chlorpromazine effects are probably due to the compound coming out of solution in the
aqueous humor and depositing on surfaces bathed in the aqueous.

5. Cutaneous

Virtually any drug is capable of producing an allergic rash in occasional patients,

typically between 2 and 10 weeks following initial exposure. This is usually macu-
lopapular, erythematous, and itchy, affecting the face, neck, trunk, and extremities
(often the palms and soles of the feet). Allergic reactions vary in distribution and
severity, the most extreme being exfoliative dermatitis, which can be life-threatening.
In most cases, discontinuation of the precipitating agent is followed by prompt remis-
sion of symptoms and signs. The itching and rash can be treated symptomatically (with
topical steroids). Subsequent antipsychotic therapy should be with a drug from a
different chemical group. Occasionally, contact dermatitis occurs in patients hypersen-
sitive to antipsychotic drugs, particularly to liquid preparations.
Patients receiving low-potency antipsychotic medication sometimes become very
sensitive to sunlight. The resulting reaction resembles severe sunburn and is managed
in the same way. Subsequent treatment should include switching to a higher-potency
antipsychotic drug or warning patients to protect themselves from sun, whether physi-
cally or through the use of a sunscreen preparation. Although dermatologic reactions to
clozapine are rare, there has been one report of photosensitivity with this atypical
antipsychotic. Patients who have developed sensitivity to sunlight during treatment
with other drugs should be monitored carefully if they are switched to clozapine.
Occasional patients treated with high doses of low-potency phenothiazines for
prolonged periods develop a blue-gray discoloration of the skin, specifically in skin
exposed to sunlight. This usually occurs in conjunction with pigmentary changes in the
eye. Aside from cosmetic concerns, it is unclear if this reaction has any clinical
As mentioned earlier, seborrheic dermatitis often occurs in conjunction with Par-
kinson's syndrome.

6. Hormonal, Sexual, and Hypothalamic Reactions

Previously, we discussed'how antipsychotic drugs increase prolactin release from

the anterior pituitary and consequently cause hyperprolactinemia, probably as a result
of dopamine-blocking activity. In females, this occasionally leads to galactorrhea (also,
rarely, in males) and decreased frequency or flow of menstruation, and in both sexes, to
a diminished libido. Although there is no evidence that chronic antipsychotic drug
treatment can result in increased incidence of pituitary adenomas, it remains at least a
theoretical possibility that lactotrophes in the pituitary will be stimulated. Therefore, in
the face of chronic hyperprolactinemia, patients should be examined occasionally for
evidence of pituitary enlargement. Persistent amenorrhea and galactorrhea suggest that
the clinician should lower the antipsychotic drug dose. At times, the dopamine agonist
bromocriptine may be considered, but this should be discussed with an endocrinologist
and also might exacerbate psychosis.

Antipsychotic agents appear to lower circulating levels of testosterone through

either a direct or an indirect mechanism. Together with the effects of elevated prolactin,
decreased testosterone may contribute to the diminished libido observed in some males
treated with these chemicals.
Effects on blood sugar, growth hormone, and thyroid hormone also have been
reported, although it is doubtful that these are of much clinical import. Chlorpromazine
impairs glucose tolerance and insulin release, which may be clinically noteworthy in
some prediabetic patients. Another sexual symptom, already mentioned under anti-
cholinergic effects, is interference with ejaculation. This is most frequently observed in
young males treated with thioridazine and is manifested by delayed ejaculation or
retrograde ejaculation. In the latter case, the patient reports orgasm without emission,
followed by urination that has a "foamy" appearance. Difficulty maintaining and sus-
taining erection sometimes improves with change to a high-potency drug.
Both females and males taking antipsychotic drugs sometimes complain of de-
layed, altered, or inadequate orgasms. The mechanism is unclear but may include
psychological, hormonal, or autonomic factors. Possible therapeutic strategies can
include lowering the dose of the antipsychotic agent, switching to a higher-potency
compound, or using a peripherally active cholinomimetic, such as bethanechol. The
antiserotonin antihistaminic drug cyproheptadine (Periactin® and others) has been re-
ported helpful in reversing orgasmic dysfunction attributed to antidepressant drugs;
whether it might have similar value in patients on neuroleptic therapy is untested.
In addition to effects on prolactin and the existence of the neuroleptic malignant
syndrome, antipsychotic drugs seem to have many other actions at the level of the
hypothalamus. These probably include cardiovascular effects as well as various hor-
monal and autonomic changes. Temperature regulation is also impaired by phe-
nothiazines and their relatives, particularly the low-potency ones, making patients more
vulnerable to hypo- or hyperthermia (depending on the ambient temperature).
Another unwanted effect, probably related to hypothalamic changes, is increased
appetite with resultant weight gain. In some patients, this can be most marked and
unpleasant. Appetite increase (which may be related to antihistaminic effects) appears
more common with low-potency drugs, including clozapine. It may be less of a prob-
lem with molindone. Management may include switching to a higher-potency agent
and dietary counseling.

7. Hepatic

The low-potency antipsychotic drugs occasionally produce a syndrome of choles-

tatic jaundice, probably a combination of a direct toxic effect and an allergic reaction.
Typically, within the first month of treatment, the patient develops fever, chills, nausea,
malaise, pruritus, and right upper quadrant abdominal pain, followed within days by
jaundice. Liver function tests reveal an obstructive pattern with increased alkaline
phosphatase and conjugated (direct) bilirubin. Transaminase enzymes also may be
elevated, but they do not reach the levels observed in hepatitis.
The recommended treatment for cholestatic jaundice is discontinuing the anti-
psychotic drug, although it is possible that patients will recover despite continued

therapy. In almost all cases, recovery occurs over a matter of weeks and is complete
and without sequelae. In predisposed patients, cholestatic jaundice may lead to chronic
biliary cirrhosis. Subsequent treatment probably should be with a different antipsycho-
tic drug, preferably a high-potency agent. The presence of preexisting liver disease
does not contraindicate the use of antipsychotic drugs. However, when the liver is
impaired, metabolism of antipsychotic drugs may be slowed, and other drugs that use
similar enzyme pathways could be affected by the addition of another liver-metabo-
lized agent. In addition, the central action of antipsychotic drugs may worsen a case of
hepatic encephalopathy.

8. Hematological

Low-potency antipsychotic agents probably are weakly toxic to some elements of

the bone marrow, particularly stem cells of the granulocyte series. Almost all patients
can compensate for this and show no more than a transient leUkopenia. However, a rare
patient develops agranulocytosis.
Perhaps I in 3000 or 4000 patients treated with chlorpromazine (and possibly
other low-potency drugs) will develop agranulocytosis, which occurs rarely with high-
potency drugs. The onset is typically within the first 2-3 months of drug therapy. (It
must be remembered, however, that switching to a different agent "restarts the clock.")
Many patients treated with antipsychotic drugs show a transient leukopenia, but resis-
tant stem cells probably allow compensation. However, in a patient vulnerable to
developing agranulocytosis, the granulocyte count drops precipitously over the course
of several days. The white blood cell count may fall below 1000, and practically all
those cells will be lymphocytes.
Routine blood counts are unlikely to detect the abrupt onset of agranulocytosis
unless they are performed 2-3 times each week for the first several months of treat-
ment, which is generally not recommended (except for clozapine). Instead, the best
approach is to maintain a high clinical index of suspicion. Thus, sore throat, fever,
malaise, or other symptoms or signs of infection in the first year of treatment should
prompt an immediate white blood cell count with a differential. If the count is low, the
antipsychotic drug should be discontinued immediately, and the patient put into reverse
isolation to prevent infection. If infection does not supervene, a normal blood count
returns within several weeks. If infection does occur, there is substantial mortality. The
mortality rate is considerably elevated if the antipsychotic drug is not discontinued.
Agranulocytosis occurs more often with clozapine. Among patients treated with
clozapine, the incidences of agranulocytosis at 12 and 18 months are 0.8% and 0.91 %,
respectively. Almost all cases occur during the first 18 months of clozapine therapy,
although in a case with multiple complicating factors, agranulocytosis occurred during
the third year of clozapine treatment. 55 Unlike the probable toxic mechanism underly-
ing agranulocytosis induced by chlorpromazine and other phenothiazines, agranulocy-
tosis engendered by clozapine is more likely caused by an immune mechanism. 56 This
suggests that reexposure of a patient to clozapine can result in an abrupt drop in white
blood cell counts, and this has been reported in several cases. 57
HLA tissue typing is currently being studied as a means to identify patients at risk

of developing this complication. There is evidence of genetic vulnerability to

clozapine-induced agranulocytosis, with patients of Ashkenazi Jewish background pos-
sibly at highest risk. 56 The risk is also greater for women and older patients. 58 Agranu-
locytosis is often preceded by a sharp upswing in white blood cell count. 58 The use of
granulocyte colony-stimulating factor (G-CSF) may shorten the duration of granu-
locytopenia from a mean of 16 to 8 days and reduce the morbidity of the disorder. 59
Eosinophilia is estimated to occur in 0.2-40% of clozapine-treated patients. It is
reversible and considered benign, although it has been associated with the development
of neutropenia, a possible precursor of agranulocytosis. 6o Clozapine-induced eosino-
philia appears to occur more often in women. 61
Antipsychotic drugs probably can affect platelet function, although whether this
may have therapeutic or toxic significance is unclear. A variety of immunologic and
coagulation changes have been described in patients taking chlorpromazine, sometimes
in association with hepatomegaly.
About half of a group of patients who had been treated chronically with chlor-
promazine had a positive antinuclear antibody test, and more than 75% showed in-
creased serum concentrations of immunoglobulin (IgM) associated with prolongation
of partial thromboplastin time. In addition, a number of autoantibodies have been found
in patients treated with chlorpromazine, often in association with splenomegaly.62
In a study contrasting the immunologic effects of chlorpromazine with those of
haloperidol, 6 of 29 patients taking chlorpromazine but none of 14 taking haloperidol
had progressive elevations of serum IgM at the end of 5 years. 63 After 5 years, 87% of
chlorpromazine-treated and 50% of haloperidol-treated patients had antinuclear anti-
bodies, but none had developed a lupuslike syndrome. Several of the chlorpromazine-
treated patients in this series, however, did develop immune syndromes (Wal-
denstrom's macroglobulinemia and immune thrombocytopenia), leading these authors
to recommend that patients who develop increased serum IgM while taking chlor-
promazine be switched to other antipsychotic drugs. Occasional reports associate antip-
sychotic drugs with other hematological syndromes, but no relationship has been
clearly established.
A promising atypical neuroleptic, remoxipride, was not released in the U.S. mar-
ket as, just before its likely introduction, cases of aplastic anemia developed in patients
taking it in Europe. To date, no atypical neuroleptics besides clozapine and remoxipride
have caused such life-threatening hematological complications, but physicians should
keep a weather eye out as new drugs are introduced.

9. Pregnancy and Lactation

All antipsychotic drugs can cross the placenta and enter the fetus. To date, no
consistent evidence has emerged from epidemiologic data suggesting that any anti-
psychotic drug is a teratogen. However, the possibility of a low-level association
cannot be ruled out. Moreover, antipsychotic drugs may affect the developing nervous
system, with possible long-lasting neurochemical and behavioral effects.64 In addition,
newborn babies have shown syndromes of neuroleptic withdrawal from the drugs along
with extrapyramidal and hypothalamic reactions. For these reasons, antipsychotic

drugs should be avoided during pregnancy, if at all possible. Similarly, antipsychotic

drugs can be detected in human milk at levels roughly similar to those found in blood.
Although this means that the nursling would receive relatively small amounts each day,
concern about the effects of psychotropic drugs on the developing nervous system,
referred to above, must be taken into consideration in weighing treatment decisions
(see Chapter 10).

10. Withdrawal Reactions

Habituation and addiction are not believed to occur with antipsychotic drugs.
Rarely, these agents have been used recreationally.
As noted earlier, tolerance does develop to some of the unwanted effects of these
agents. Partial to complete tolerance may develop to sedation, and some degree of
tolerance can occur for hypotension and anticholinergic actions. Conversely, when the
drug is stopped, rebound reactions may occur. These can include insomnia, nightmares,
and other disturbances of sleep, as well as cholinergic rebound such as increased
salivation, abdominal cramps, and diarrhea. To make patients more comfortable, dis-
continue these drugs gradually (e.g., 5-10% of dose per day).
Obviously, discontinuation of antipsychotic drugs may be followed by relapse of
psychosis. This is generally believed to be a recurrence of the underlying psychiatric
disorder. Although it has been suggested that in some patients psychosis actually is
created or exacerbated as a result of supersensitivity created by the drugs,12 this
remains controversial and inadequately proved.
As discussed in the section on tardive dyskinesia, discontinuation of antipsychotic
drugs is occasionally followed by transient withdrawal dyskinesias. In other cases,
discontinuing the drug unmasks persistent dyskinesia.
Because risperidone appears to have a less severe side-effect profile than
clozapine, is less expensive, does not require weekly venipunctures, and is also an
atypical antipsychotic, many patients have had clozapine abruptly discontinued so that
they could be switched to risperidone. Abrupt discontinuation of clozapine may be
more problematic than abrupt discontinuation of traditional neuroleptics: psychotic
relapse may occur early and be severe, and autonomic disturbances, including in-
creased cholinergic activity, have been reported. Because of its different neurophar-
macology, particularly muscarinic affinity, risperidone does not block clozapine with-
drawal symptoms, and acute dystonic reactions have been reported despite the
relatively low incidence of these symptoms with risperidone. Risperidone also may not
be efficacious for patients who have failed to respond to clozapine. If a trial is to be
considered, clozapine should be tapered gradually.

11. Overdose
Fortunately, antipsychotic drugs have a high therapeutic index-the ratio of toxic
to therapeutic dose. Ingestion of an antipsychotic agent alone in an overdose attempt
seldom results in death. (However, it should always be remembered that another agent
may have been ingested along with the antipsychotic drug). Possibly more dangerous
than other anti psychotics in overdose are thioridazine and loxapine.

When an antipsychotic drug is ingested in an overdose, the anticholinergic effects

will probably delay gastric emptying; passage of a nasogastric tube with the application
of suction should always be attempted, even many hours after the overdose. Instillation
of activated charcoal may further diminish absorption. Attempts at inducing emesis
may be unsuccessful because of the antiemetic action of these drugs, particularly the
low-potency agents. Because antipsychotic agents have a high lipophilicity coupled
with a strong adherence to tissue and proteins, attempts at dialysis are relatively
Antipsychotic drugs alone seldom produce lethal suppression of vital brainstem
centers. However, the low-potency agents are capable of inducing serious blood pres-
sure problems, which, as noted earlier, should be treated with pure a-stimulating drugs.
Low-potency drugs also can occasionally produce serious arrhythmias, which, with
thioridazine, are most likely to be dangerous. As noted earlier, quinidine and other type 1
antiarrhythmic agents (namely, procainamide and disopyramide) should be avoided in
the treatment of these arrhythmias, because of additive effects (see Section III.G.3).
We previously discussed how physostigmine may have a role in reversing some of
the central nervous system effects of atropinic toxicity caused by antipsychotic agents.
However, physostigmine can worsen some arrhythmias, lower the seizure threshold,
and produce respiratory arrest. For this reason, it should be used extremely cautiously,
if at all, in conditions of cardiovascular and respiratory instability.51

H. Drug Interactions and Combinations

In general, to minimize potential unwanted interactions, to simplify the therapeu-
tic regimen, to enhance compliance, and to assess specific contributions to a clinical
outcome more easily, a physician should prescribe the fewest possible drugs for a
patient. However, in a number of clinical situations, a physician will want to coad-
minister another drug with an antipsychotic agent. We have already mentioned the
treatment of extrapyramidal effects with antiparkinson drugs. In such circumstances,
several pharmacodynamic interactions are possible. Most antiparkinson drugs (with the
partial exception of amantadine) have anticholinergic actions that will add to those of
the antipsychotic agent. Many antiparkinson drugs also have antihistaminic effects,
which can enhance the sedation caused by antipsychotic drugs. In addition, phar-
macokinetic interactions also have been reported. It is possible that antiparkinson drugs
lower the blood levels of some antipsychotic agents; however, data about this point are
inconsistent. 65,66
Sometimes it may be useful to coadminister a benzodiazepine together with an
antipsychotic agent. As mentioned earlier, a drug like diazepam may effectively coun-
teract acute dystonic reactions or akathisia. There appears to be little potential for a
pharmacokinetic interaction between these two classes of drugs, but benzodiazepines
or any sedative-hypnotic drug (including alcohol) can cause added sedation. In the
extreme, depression of respiration or blood pressure may occur. There is some concern,
as yet unproven, that coadministering a benzodiazepine with clozapine can result in
potentially fatal respiratory suppression and arrest.
At times (e.g., in the treatment of psychotic depression), a physician may elect to

treat a patient with both an antipsychotic drug and an antidepressant. A pharmacokine-

tic interaction, probably at the level of hepatic drug-metabolizing enzymes, can raise
blood levels of both drugs higher than if either had been administered alone. In
addition, additive sedative, anticholinergic, and cardiovascular effects are likely.
Lithium and an antipsychotic drug are combined in the treatment of many patients
with mania. A syndrome of severe neurotoxicity-including neuromuscular signs,
cognitive changes, and hyperthermia-was described in four patients receiving treat-
ment with both lithium carbonate and haloperido1. 67 Similarly, EEG abnormalities
occurred in several patients who developed encephalopathy during therapy with a
combination of lithium and thioridazine. 68 The combination of lithium with clozapine
produced serious but reversible neurologic reactions in four patients. 69 On the other
hand, broad clinical experience and a number of published series have testified to the
safety of lithium/antipsychotic drug combinations for a vast majority of patients. It is
possible, though, that an occasional patient will experience enhanced neurotoxicity
from the combination, and any suggestion that this may be occurring should prompt a
rapid lowering of dosages or drug discontinuation. 70
A pharmacokinetic interaction has also been described between lithium and chlor-
promazine: mean plasma chlorpromazine levels were significantly lower during com-
bined therapy.71 Lithium may slow gastric emptying, which would enhance the gut
metabolism of chlorpromazine and diminish its absorption, or it may interfere with the
absorption of chlorpromazine by diminishing its transport across gut membranes.
Because clozapine causes a dose-related increase in the risk of seizures, it is
sometimes coadministered with an anticonvulsant, such as valproate. The combination
of clozapine and carbamazepine is generally avoided because carbamazepine has a
tendency to suppress blood cell production in rare patients, and this may increase the
risk of agranulocytosis. Both carbamazepine and valproate have been reported to
decrease serum concentrations of clozapine. 72- 74
Erythromycin, calcium channel blockers, quinidine, paroxetine, and fluoxetine all
moderately increase plasma sertindole concentrations. Ketoconazole, itraconazole, and
cimetidine are likely to cause larger increases in sertindole levels. Sertindole is de-
creased significantly by carbamazepine and phenytoin and modestly by cigarette
smoke. Plasma levels of olanzapine also are decreased by carbamazepine, phenytoin,
and cigarette smoke.
Antipsychotic agents can also interact with drugs used for the treatment of non-
psychiatric conditions. Phenothiazines (like cyclic antidepressants) can interfere with
the antihypertensive actions of guanethidine (Ismelin@ and others). Low-potency anti-
psychotic drugs may potentiate the postural hypotension observed with various anti-
hypertensive drugs. Haloperidol has been reported to cause an organic brain syndrome
when combined with methyldopa (Aldomet@ and others). By decreasing the metabo-
lism of phenytoin (and perhaps of other anticonvulsant drugs), antipsychotic drugs may
elevate anticonvulsant blood levels, potentially producing toxicity. Through their ten-
dency to produce extrapyramidal reactions, antipsychotic drugs can counteract the
antiparkinson effects of drugs used to treat naturally occurring Parkinson's syndrome.
Caffeine, contained in many beverages and over-the-counter medicines, also may
interact with antipsychotic drugs. At the dynamic level, it can counteract sedative

effects, possibly necessitating an increased dose of the antipsychotic drug. At a kinetic

level, some caffeine-containing beverages might diminish the absorption of anti-
psychotic drugs from the stomach, but this is controversial. 75 Nicotine has been report-
ed to decrease the plasma concentrations of antipsychotic drugs. The "chain smoking"
patient may have discovered that this lowered plasma concentration adds to the plea-
sure of the cigarette itself.
The rule of thumb is that when any drug is added to any other drug, the possibility
of interactions must be considered. For this reason, as stated at the beginning of the
section, the clinician should minimize the use of multiple drugs in the same patient at
the same time.

I. Laboratory Tests and Monitoring

No routine laboratory tests are essential for all patients receiving antipsychotic
agents-except clozapine. Similarly, blood assays of antipsychotic levels have not yet
reached the stage of general applicability. However, pretreatment laboratory tests (sup-
plementing a history and physical examination) can be useful to assess a patient's
general health status prior to initiating antipsychotic drug therapy and for periodic
reassessments during treatment. These may include a complete blood count with a
differential,liver function tests, and, in men over 30 and women over 40, an electrocar-
diogram. (In the initial assessment of a psychotic patient, a differential diagnosis may
support assessment of other organ systems or metabolic and endocrine functions. In
addition, it would be wise to perform routine screening for syphilis, chest disease, and
other illness. 76) During treatment, the clinician should occasionally recheck a patient's
health status, although it makes little sense to repeat laboratory tests more frequently
than once per year in asymptomatic patients. To recapitulate earlier points, liver disease
does not automatically contraindicate the use of antipsychotic drugs, and maintaining a
high clinical index of suspicion for agranulocytosis and jaundice within the early
months of therapy is superior to routine laboratory testing-except for clozapine. As of
this writing, weekly blood counts are required for patients taking clozapine. Because
the risk of agranulocytosis decreases steadily over the first few years of therapy, it is
likely that requirements in the United States will become less stringent, following the
lead of many other countries. Routine EKGs may be prudent for patients maintained on
As mentioned in the section on tardive dyskinesia, monitoring of patients for early
signs of abnormal involuntary movements is wise before and during chronic anti-
psychotic drug therapy. Prior to treatment and at intervals of 6 to 12 months thereafter,
the performance of a basic examination (such as the AIMS) would be worthwhile (see
Fig. 2).
When patients are taking antipsychotic drugs, various laboratory tests may be
artifactually altered. As examples, invalid increases may be reported in serum alkaline
phosphatase, serum and urine bilirubin, cerebrospinal fluid protein, serum cholesterol,
urine diacetate, serum 17 -hydroxy steroids, urine porphyrins, serum transaminases, and
urine urobilinogen. In addition, false-positive results may be reported in urinary phe-

nylketonuria and pregnancy testing. Protein-bound iodine (PBI), radioactive iodine,

and serum and urine uric acid testing may be inaccurate as well.

J. Clinical Uses of the Antipsychotic Drugs

Based on a medical and psychological evaluation of the patient, the clinician
should determine the need for medication and its place within the treatment plan (see
Sections I and II in this chapter). The choice of a specific agent depends on the patient's
medication history and an understanding of pharmacokinetics, adverse effects, toxicity,
and drug interactions. The clinician should discuss the decision to medicate with the
patient and/or his or her family, guardian, etc. (For further discussion of informed
consent, competency, etc., see Chapter 14). Pretreatment laboratory tests might be
useful to assess a patient's general health prior to initiating drug therapy (see Section

1. Acute Treatment
In the early placebo-controlled trials of phenothiazines in the treatment of acute
schizophrenic psychosis, about one-fourth of patients who received placebo responded
reasonably well. This means that a minority of psychotic patients who meet behavioral
criteria for this undoubtedly heterogenous condition will improve without antipsycho-
tic therapy. However, there is no clear way to predict who these patients are.
Given the current state of scientific knowledge and clinical experience, there is
wisdom in allowing a day or two to pass before administering antipsychotic drugs to a
patient undergoing a first psychotic episode. In a supportive milieu, away from poten-
tially toxic substances and psychosocial stressors, some patients may show rapid remis-
sion, without medication or with only antianxiety agents. At the same time, staff can
conduct a comprehensive medical, neurologic, and psychiatric diagnosis, psychosocial
history, and screening for substances of abuse. With this approach, unnecessary use of
antipsychotic drugs with their attendant risks can be avoided.
On the other hand, psychotic episodes should not be allowed to continue for more
than a very brief period without trying to alleviate them through administration of
antipsychotics. Not only is each episode painful, disruptive, and potentially dangerous,
the growing evidence suggests that episodes may leave a "neurobiological trace" that
may aggravate the ultimate long-term prognosis and treatment response.
Some physicians used to employ extremely large doses of high-potency anti-
psychotic drugs early in treatment. This approach, known as rapid neuroleptization,
attempted to achieve rapid remission of psychosis and avoidance of hospitalization or
rapid discharge from the hospital. Unfortunately, double-blind studies were unable to
confirm any enhanced efficacy of this form of treatment contrasted with standard
dosage regimens. In other words, treating all psychotic patients with "megadoses" does
not control psychosis better or more rapidly, and it is associated with increased un-
toward effects. Thus, the clinician should treat most patients at the lower end of the
therapeutic range for acute psychosis, e.g., 5-10 mg/day of haloperidol or flu-
phenazine. Additional sedation can be achieved through the adjunctive administration

of a benzodiazepine [e.g., lorazepam (Ativan® and others)]. Only patients who in the
present or past have been unresponsive to such standard doses administered over a
period of weeks should be considered for higher doses.
For the highly excited or dangerous patient, the combination of a neuroleptic drug
and a benzodiazepine may provide rapid pharmacologic control in the context of a calm
but structured environment. A high-potency drug, such as haloperidol, can be adminis-
tered parenterally and is less likely to produce hypotension or seizures than lower-
potency antipsychotic agents,?7 A dose of 2.5-10 mg can be administered intra-
muscularly every 1-2 hr, alternating with intramuscular injections of lorazepam (the
only benzodiazepine to achieve reliable blood levels when administered through this
route), 1-2 mg, until the patient is quiet. Vital signs should be monitored at least
hourly. An alternative to haloperidol in a combined regimen would be mesoridazine,
25-50 mg, which is more sedating than haloperidol and less likely to produce extra-
pyramidal reactions but causes more hypotension.
Yet a third option for parenteral neuroleptic therapy is droperidol. Available in the
United States only for use in anesthesia, droperidol is a butyrophenone like haloperidol,
but more sedating and possibly faster acting. It has been described as very effective in
emergency management of acutely excited and violent patients, but because no oral
preparation is currently available, the patient must later be switched to another product.
Early in the treatment course, administer antipsychotic drugs in divided doses,
perhaps three or four times per day (more frequently for patients who are severely
excited). Once satisfactory levels have been achieved, and the patient has become
tolerant to unwanted effects (probably within several weeks), a once-daily dosage
usually will suffice. If a patient can be maintained out of the hospital, a once-daily
dosage, even from the beginning, may enhance compliance. Particularly when a more
sedating agent is employed, a bedtime dose may be optimal, as it can improve sleep,
and the peak anticholinergic and hypotensive effects will occur while the patient is
What if a patient does not appear to respond to an antipsychotic drug even after 6
to 8 weeks? First, the clinician must wonder whether the patient is actually receiving
the agent. An outpatient may not have filled a prescription or may be taking the drug
irregularly, if at all. An inpatient might be "cheeking" tablets or capsules. Questions to
be asked at this juncture include: Is anyone observing the patient taking the medication,
including an inspection of the patient's mouth? Does the patient have any noticeable
"side effects" from the drug, such as extrapyramidal reactions? If compliance is a
particular problem, consider changing the dosage regimen, administering a liquid prep-
aration, or using a parenteral agent. Always ask patients how they feel about a drug.
They might reveal covert effects such as sexual dysfunction or akathisia, which may be
improved by switching to a different agent, or one may learn of a patient's wish to
remain "crazy" or a patient's feeling that taking the drug is an indication that he or she
is mentally ill.
If compliance is not a problem, yet the patient has not improved, he or she may
require a higher dose of medication. Increase the dosage until improvement occurs or
limiting adverse effects develop. Although U.S. Food and Drug Administration (FDA)-
approved labeling (contained in the Physicians' Desk Reference and in the package

insert) provides useful guidelines about maximal dosage, the actual upper limit may
vary from patient to patient. The one exception to this rule is thioridazine, which must
not be administered in dosages exceeding 800 mg per day-even for a brief period of
time-because of the danger of pigmentary retinopathy with potential loss of vision.
Some patients metabolize antipsychotic drugs very rapidly in the gut and in the
first pass through the liver. A physician might suspect that this is the case in a patient
who is taking relatively large doses orally yet experiences neither clinical benefit nor
adverse effects. These patients may respond to higher oral doses or to a parenteral
Although all antipsychotic drugs are equally effective, some patients appear to do
better with one agent than another. Thus, failure to improve with one drug (after the
above factors have been considered) is a reason to change to a different agent. Clini-
cians should be wary, however, of a tendency to be "too quick on the prescription
trigger," switching agents every few days in a vain attempt to "do something." Al-
though side effects (including sedation) occur early, antipsychotic effects may take 3-6
weeks. Only after a drug has been given an adequate trial (at least 300-600 mg of
chlorpromazine-equivalent dose per day for 6-8 weeks) should the clinician switch
to a different agent. When it is time to change, it makes most sense to choose a drug in
a different chemical class.
One other consideration in the apparently refractory patient is psychosocial fac-
tors, which may be "fueling" the psychosis. Antipsychotic drugs raise the threshold at
which an individual predisposed to psychosis actually becomes psychotic in response
to interpersonal and intrapsychic stresses. A chronic schizophrenic, for example, might
remain quite stable and relatively intact (albeit still awkward, isolated, and apathetic)
when in a low-stress environment, even at a very low dose of antipsychotic. However,
given a precipitant (an argument with a friend, family member, or mental health worker
or being in a program that is too intense or overinvolved), the patient may decompen-
sate into a state of disorganized thinking and be unable to distinguish reality from
It takes more medication to "cap" a psychosis in the face of ongoing stresses and
threats than in more placid circumstances. In fact, if the pressure is sufficient, psycho-
pharmacology alone will not bring the psychosis under control. Hence, the clinician
must consider psychological and environmental factors when treating an acutely de-
compensated patient. If patients are in a stable environment and are unimproved
despite adequate chemotherapy, change the environment; e.g., arrange for hospital
admission. If patients are in the hospital, inquire about ongoing contacts both inside
and outside the hospital. Occasionally, friends and family will need to be counseled on
how to reduce the patient's stress, other patients may need to be controlled from
making sexual advances, or a naive staff member might be getting provocatively close.
In such circumstances, decreasing the intensity of the patient's conflict may allow a
hitherto ineffective drug to become strikingly efficacious.
And how effective is effective for an antipsychotic drug? Traditional drugs are
best at suppressing the most florid symptoms of psychosis such as hallucinations,
delusions, disorganized thinking, and excited, assaultive, and belligerent behavior. If
the patient has a primary disorder of mood, reconstitution should involve what Bleuler

called restitutio ad integrum-in other words, a return to an integrated and normal state
of personality and functioning. For a schizophrenic patient, however, recovery from
psychosis is likely to involve return to a somewhat strange, awkward, and isolated
Since the third edition of The Practitioner's Guide went to press several years ago,
the introduction of atypical neuroleptics has substantially altered the pharmacotherapy
of psychotic disorders. Those of us who conducted research on clozapine in the 1970s
were impressed that it was a highly efficacious antipsychotic with minimal (if any)
acute extrapyramidal reactions. What has emerged with subsequent study and clinical
experience, however, is that it is probably a superior antipsychotic, with particular
efficacy in patients previously unresponsive to traditional neuroleptics and against the
cluster of symptoms commonly termed "negative"-e.g., apathy, flattened affect, so-
cial awkwardness, and decreased motivation. However, this interpretation of the litera-
ture is under debate. Some clinicians would say that clozapine does not have direct
impact on negative symptoms, but, as is true with most antipsychotics, the negative
symptoms get better in conjunction with the positive symptoms. 78 ,79
Because of its range of serious and potentially lethal effects, clozapine is typically
reserved for patients unresponsive to or intolerant of traditional neuroleptics. However,
given the magnitude and unique nature of its efficacy against the spectrum of schizo-
phrenic and other psychotic symptoms (it may be more efficacious in psychotic mood
disorders 80,81), and the malignant (and often lethal) nature of chronic psychotic disor-
ders, one might consider whether, despite its risks, clozapine should be used as a first-
line agent for some patients (Table 8). In any case, its use should not be withheld for an
extended time, unless a patient's clinical response (in terms of both safety and efficacy)
is highly satisfactory on a traditional agent.
At the time of this writing, clozapine is expensive. The cost is partly due to the
manufacturer's pricing of the medicine itself and partly due to the currently required
weekly venopunctures and white blood cell (WBC) assays. The former is likely to
decrease with the introduction of new atypical neuroleptics, while the frequency of
blood count monitoring will also probably decrease within the next few years. Even
with the current high annual cost, however, clozapine affords many patients sufficient
clinical and functional benefits so that they are able to be treated in less costly set-
tings-thus providing an "offset" that often exceeds the cost of the medicine and
laboratory testing. 82 The benefits of clozapine treatment appear to accrue through a
second year of maintenance therapy and perhaps even beyond. This may be related to
ongoing neurobiological alterations brought about by the drug itself, or perhaps what-
ever normalizing effect it has on neuronal networks allows rehabilitative therapies and
ongoing social interactions to have a normal and salutary effect on the patient's think-
ing, feelings, and behavior.
Due to clozapine's cost, clinicians often wonder when to discontinue treatment for
lack of efficacy. The best available data suggest that some benefit should be evident
after 3 months of treatment, and if the patient is not clearly better after 6 months of
taking clozapine, the doctor can reasonably taper and stop it.
A proper role for risperidone, the second atypical neuroleptic introduced in the
United States, is less clear-cut at this time. There is no evidence that it is superior in

TABLE 8. Clozapine (Clozaril®)

Clinical indications
If a patient shows satisfactory response over 3-4 months of acute treatment for any of the indications below,
and if maintenance treatment is indicated, in most cases clozapine will be the maintenance treatment of
I. Treatment-resistant schizophreniaa
2. Schizophrenia or schizoaffective disorder-when typical neuroleptics have caused severe extra-
pyramidal reactions or tardive movement disorders.
3. Treatment-resistant or -intolerant mood disorders with psychotic features. b
4. Other psychiatric disorders for which an antipsychotic is indicated but in which typical neuroleptics
have been ineffective or poorly tolerated. c

Adverse effects
A. Serious: Agranulocytosis; cardiovascular: EKG changes and arrhythmias, orthostatic hypotension; res-
piratory arrest; seizures
B. Common: Constipation; dizziness; fever; hypertension, sedation, salivation; tachycardia; urinary incon-
tinence; weight gain

Abrupt withdrawal of clozapine may lead to a more rapid and severe return of psychotic symptoms than
usually observed with traditional antipsychotic agents. Very gradual withdrawal. therefore, is recommended,
except when serious toxicity (e.g., agranulocytosis) occurs.

The most serious consequences of clozapine overdose are depressed consciousness, cardiac arrythmias,
respiratory suppression, hypotension, and seizures. There are no specific antiodotes, and management is the
same as for other antipsychotic overdoses.

Laboratory testing and monitoring

A. As of this writing, patients being treated with clozapine must undergo a pretreatment white blood cell
(WBC) count and weekly counts thereafter for as long as they remain on clozapine treatment. Available
data, however, suggest that after a period of time, less frequent WBC counts are necessary, as the risk of
agranulocytosis diminishes. Presumably in time, the manufacturer and FDA will agree to more liber-
alized requirements.
B. Other laboratory tests, such as assessment of liver function, are not routine and are indicated by clinical
symptoms and signs. Pretreatment EKG and vital signs are recommended.
C. Close observation of patients undergoing clopazine therapy is useful during the initial weeks of treat-
ment. Most important to watch for are signs of infection, loss of consciousness, seizure activity,
dizziness, or difficulty breathing. If possible, vital signs should be checked several times a day during
dosage titration.

Clozapine costs $6,000-$10,000 annually. The high cost is due largely to the requirement of weekly
venipunctures and laboratory tests. As more novel antipsychotics are introduced over the next few years, the
manufacturer is likely to reduce the price of clozapine. If the requirement for WBC counts is liberalized, that
too will reduce the annual costs of treatment.

a As indicated in the text, one may argue about how many "typical" neuroleptics should be tried before c10zapine is
prescribed for a patient. Although c10zapine is unquestionably more toxic and hazardous that standard anti psychotics, its
superior efficacy suggests that in many cases it should be tried sooner rather than later in the course of a lifelong illness.
bEven though c10zapine does not carry FDA-approved labeling for this indication, there is compelling evidence that it is
efficacious and possibly even more efficacious for this indication than it is in the treatment of schizophrenia.
'Outside of FDA labeling.

efficacy, nor that it has special advantages for patients previously resistant to standard
neuroleptics. At lower doses-8 mg/day or less-it produces a much lower incidence
of acute extrapyramidal reactions than a high-potency neuroleptic such as haloperidol.
As noted earlier, it has produced cases of neuroleptic malignant syndrome (NMS), and
there is currently no basis on which to estimate its relative risk of causing tardive
movement disorders. A common practice at this time among clinicians treating psycho-
tic patients is to try one or two standard neuroleptics and, failing to see optimal
benefits, to then prescribe risperidone, followed by clozapine. There is certainly little
reason to believe that a patient who has failed to respond to clozapine will benefit from
risperidone; however, the reverse is likely to be true. Patients who have benefited from
clozapine and then been switched to risperidone often have not shown comparable
benefits. As mentioned earlier, a rapid discontinuation of clozapine and substitution of
risperidone has produced havoc in many patients, most likely due to a combined
withdrawal effect from clozapine and a lack of sufficient compensation by risperidone.
A gradual discontinuation from clozapine may ease the transition and reduce the risks
of psychotic relapse.
As of this writing, olanzapine has recently been introduced into clinical practice.
Preliminary results have been very favorable. Some clinicians believe it approaches
clozapine in efficacy-particularly in its effects on negative symptoms and in difficult
patients. Early clinical experience suggests that it is well tolerated. Sertindole may soon
be approved for marketing in the United States, but clinical experience beyond phase
III trials does not yet exist.
Other atypical neuroleptics are certain to be introduced over the next few years.
What benefit they will bring in the important safety and efficacy variables discussed
above remains to be seen as additional studies are conducted and clinicians gain
experience in the field with patients typically much more complex than those who
qualify for research protocols. The atypical agents currently being studied have varying
pharmacologic profiles. A common denominator is that virtually all have a compara-
tively low incidence of acute extrapyramidal reactions and uncertain long-term liability
of producing tardive syndromes. The differences among this group tend to be much
greater than differences among the typical neuroleptics, making generalizations impos-

2. Transition and Continuation Therapy

When psychotic symptoms have remitted successfully, we pass from the acute
phase of active symptom suppression to continuation therapy. Continuation therapy
entails the continued administration of medication in the hope of avoiding a relapse.
This is analogous to continuing antimicrobial therapy past the period of symptom relief
and fever defervescence. After the alleviation of an acute psychotic episode, continue
therapy for about 6 months.
It is best not to change the type and dosage of antipsychotic medication at the time
of an important transition, such as discharge from the hospital. If a patient has re-
sponded favorably to an antipsychotic drug, continue the same dose of that drug

through the return to home, job, and community. Antipsychotic agents presumably
provide a buffer for the stresses of the transitional period.
If a person resumes a normal life without evidence of recurrent psychosis, gradu-
ally lower the dose of the antipsychotic drug. If dosage tapering proceeds smoothly,
consider discontinuing the agent entirely within about 6 months. However, when a
psychotic illness has shown itself to be chronic, with frequent recurrences (a common
pattern among schizophrenic patients), maintenance therapy may be in order. Data
emerging from recent clinical studies suggest that at least in some patients diagnosed
with schizophrenia, clinicians may need to consider lifelong maintenance antipsychotic
therapy after only two episodes. 83 Factors that might support such a decision include
poor premorbid functioning, a failure to return to baseline, the presence of attenuated
symptoms, and family history of psychosis.

3. Maintenance Therapy

a. Definition. Maintenance therapy means long-term treatment. For some pa-

tients, such as those with chronic schizophrenia, discontinuing maintenance anti-
psychotic drug therapy (or even reducing the dose) leads to a rapid return of psychotic
symptoms. For these patients, the drugs play an ongoing, suppressive role.
For other patients on maintenance therapy, the drugs appear to playa "prophylac-
tic" role. In this group, drug discontinuation raises the probability of a symptomatic
relapse at an indeterminate time in the future. For them, most probably the drug raises
the threshold to psychosocial stressors that can precipitate a reexacerbation of the
illness. In the absence of this protective effect, a milder stress can lead to return of

h. Chronic Schizophrenia. If a group of schizophrenic patients in remission is

switched from active antipsychotic medication to placebo, roughly 65-80% will re-
lapse within about 1 year. In any given month, roughly 10% of the remaining unre-
lapsed medication-free patients may be expected to relapse. At the end of 1 to 2 years, a
small cohort of schizophrenic patients will not have relapsed despite the lack of
antipsychotic drug therapy; they have less need for maintenance treatment. Unfor-
tunately, we are unable to identify these patients in advance. How long a patient will
remain stable before relapsing is variable and unpredictable. Even with guaranteed
compliance, medication probably does not prevent relapse in some patients but only
delays the time to recurrence. Patients clinically stable on relatively low doses of
antipsychotic drugs may be more likely to maintain remission following complete
cessation of the agent. Also, patients in relatively low-stress environments are more
likely to remain stable without drugs, although the onset of stress can heighten their
As in the treatment of acute psychosis, growing evidence suggests that mainte-
nance therapy too may be approached with lower neuroleptic doses. For high-potency
agents such as haloperidol or fluphenazine, 0.5-1 mg daily may suffice to prevent
symptom exacerbation and relapse for many chronic patients. In maintenance therapy,
less may be more: lower doses can often control symptoms while diminishing the risks

of tardive dyskinesia as well as less troubling but more annoying side effects that might
undermine compliance.
Research has shown that lower antipsychotic drug doses do, on average, lead to
higher rates of relapse, especially with longer periods of follow-up.84 Nevertheless,
relapses in patients on low-dose antipsychotic drug therapy tend to be less severe than
those occurring in patients free of medication: they remit more easily when the drug
dose is raised and are less likely to result in social disruption and the need for hospital-
ization. Patients and families prefer low-dose treatment even with the increased risk for
Several groups of investigators have been examining an intermittent or "targeted
dose" strategy for chronic schizophrenic patients. This involves drug treatment only
during an acute episode, with drug discontinuation afterwards. During maintenance
therapy, patients are carefully monitored, and family members know how to reach a
clinician promptly to report prodromal symptoms. This approach relies heavily on
therapeutic rapport, and it also assumes that the clinician can identify prodromal
symptoms, which may include such nonspecific evidence of distress as anxiety and
insomnia. Unfortunately, a much higher rate of psychotic breakdowns with intermittent
than with continuous treatment makes this technique too risky for most patients. A
practical approach is to use low but continuous doses with intermittent increases when
symptoms increase. Patients can be maintained for prolonged periods on very low
doses of antipsychotic drugs, but any evidence of a "roughening" of their clinical
course should prompt a rapid raising of the dose until symptoms subside. Thus, a
patient receiving 0.25 mL of fluphenazine decanoate every 3-4 weeks might receive a
supplement of 1-5 mg of fluphenazine hydrochloride daily by mouth or an additional
0.25 mL of decanoate until acute symptoms have subsided.
The use of "drug holidays" -i.e., periodic, planned withdrawal of neuroleptic
therapy-has fallen out of favor. Initially designed as a means to lower a patient's
exposure to these agents, drug holidays have not appeared to lower the risk of tardive
dyskinesia (and may, in fact, increase it). By contrast, when patients are taking these
medicines chronically, a 5-out-of-7-day schedule (i.e., once-daily dosing Monday
through Friday, with the weekend "off") does not represent a neuropharmacologic
"holiday," since the brain will hardly notice the falloff in levels (because of saturation
of fat and tissue stores). This approach may make sense for patients in day treatment
programs, who can take the medication once each weekday in front of a nurse.
As of this writing, there are three long-acting antipsychotic drugs available by
prescription in the United States: the enanthate and decanoate esters of the phe-
nothiazine fluphenazine (Prolixin Enanthate® and Prolixin Decanoate®) and the de-
canoate ester of the butyrophenone haloperidol (Haldol Decanoate®). Fluphenazine
decanoate probably has a longer duration of action than its enanthate sister, which is
infrequently used.
Fluphenazine decanoate is usually administered intramuscularly (it may also be
given subcutaneously) in doses of 12.5 mg (0.5 mL) through 75 mg (3 mL) every 3-4
weeks. Although some patients have received injections greater than 125 mg (5 mL) at
a time (requiring the use of several injections), there is a growing recognition that
patients can be maintained on the lower end of the dosage scale, with some receiving as

little as 1.25 mg (0.05 mL) at each dose. Additionally, the dosing interval can be
progressively increased so that patients need only receive injections every 1-2 months,
based on their clinical responses. Symptomatic exacerbation is best treated through the
use of oral or 1M fluphenazine hydrochloride (short-acting), although the maintenance
dose also may be increased if need be. Since a decanoate cannot be removed once it is
injected, it is best to err on the lower side and supplement with shorter-acting prepara-
tions. Unfortunately, there is no established conversion factor to equate a dose of oral
fluphenazine hydrochloride to a dose of the decanoate, but each can be titrated within
the usual therapeutic ranges.
The Use of Fluphenazine Decanoate
1. Initially administer 0.25 mL of fluphenazine decanoate.
2. Supplement with oral or 1M fluphenazine hydrochloride as needed.
3. Give second dose of 0.25 mL of fluphenazine decanoate after 4 weeks.
4. Begin to taper oral fluphenazine.
5. If you cannot stop oral fluphenazine without reemergence of psychotic symp-
toms, increase decanoate dose to 0.5 mL at next 4-week administration.
6. Once patient is stable on fluphenazine decanoate alone, see if a 5- or 6-week
interval suffices. (Some patients may need decanoate only every 2 months or
less often!)
7. If patient begins to become symptomatic toward the end of a dosing interval,
shorten the interval. (Rare patients may need fluphenazine decanoate as often
as every 2 weeks.)
8. If a patient begins to become symptomatic around a time of stress, supple-
ment decanoate with oral or 1M fluphenazine hydrochloride as needed or
increase the decanoate dose for one to two dosing intervals until the stress and
symptoms pass.
Haloperidol decanoate is recommended for once-monthly administration, but
there are probably patients who can take it, too, at less frequent intervals. Concentra-
tions of haloperidol in the decanoate preparations are 50 mg/mL or 100 mg/mL. For
patients already stabilized on oral haloperidol, the recommended conversion formula is
to mUltiply the daily dose by 10-15 and then to administer that dose (in milligrams) of
haloperidol decanoate in a monthly injection. Again, the best approach is to "start low
and go slow," particularly when using a long-acting preparation.
Arguably the most dramatic effects observed with clozapine have been in the
long-term treatment of patients with chronic psychotic disorders. Many patients have
become less symptomatic and more functional on clozapine treatment than they had
been on standard neuroleptics, with benefit increasing steadily over the first several
years. Because of clozapine's unique side-effect profile, specific approaches to clinical
management are required for patients maintained on the drug. 85 As mentioned earlier,
patients who have recovered from clozapine-induced agranulocytosis should never
again be exposed to this drug. Those with seizures can, however, be reexposed, possi-
bly in conjunction with an anticonvulsant (usually divalproex), and certainly at a lower
initial dose. Sedation can be managed with a lower dose or a bedtime biasing of the
dosage schedule, and, as with other psychotropic medications, symptomatic hypoten-

sion may be treated with support stockings, increased dietary sodium, or the miner-
alocorticoid fludrocortisone (Florinef® and others). Hyperthermia is common with
clozapine, but usually benign and transient; infection and NMS must be ruled out.
Occasional patients who develop nausea and, less commonly, vomiting after weeks or
months of clozapine therapy have been treated successfully with metocloprarnide (Reg-
lan® and others).

c. Other Diagnostic Groups. In no other psychiatric disorder is the indication for

maintenance antipsychotic drug therapy as clearly established as it is in the treatment of
the schizophrenic syndrome. Nevertheless, clinical experience suggests that patients
with other diagnoses occasionally benefit from long-term treatment with antipsychotic
drugs. For example, an occasional patient with bipolar disorder may benefit from
maintenance therapy with an antipsychotic agent, either in addition to or as a substitute
for lithium. Similarly, a patient with recurrent episodes of psychotic depression might
be.assisted. We must remember, however, that the risk of developing tardive dyskinesia
or dystonia following neuroleptic exposure may be increased in patients with mood
disorders. Some patients with uncontrollable behavior caused by mental retardation or
organic brain syndrome will be more manageable on maintenance antipsychotic drug
therapy; however, it first should be demonstrated that nonpharmacologic approaches,
such as environmental alteration and behavioral therapy, do not alleviate the problem.
Patients with borderline personalities may benefit from antipsychotic drugs, usually in
relatively low doses, for long-term assistance with uncontrollable waves of emotions
and, at times, destructive impulses (see Chapter 13). The butyrophenone haloperidol
and the diphenylbutylpiperidine pimozide are often used for chronic symptom suppres-
sion in patients with Gilles de la Tourette syndrome. Pimozide might also be useful in
treating monosymptomatic hypochondriacal psychosis such as delusions of parasite
infestation or dysmorphophobia. In patients taking pimozide, the EKG should be
carefully monitored for QT prolongation.
If antipsychotic drugs are to be used for disorders other than schizophrenia,
particularly when they are used over a period of many months and even years, the
clinical record should reflect in detail the rationale behind the therapy and the care
taken in its administration. In light of the long-term risks of antipsychotic drug treat-
ment, a patient's chart should indicate why alternate treatments were not employed.
The record should also detail the continued need for antipsychotic agents, attempts
made to lower the dose or discontinue their use, and steps taken to monitor for
unwanted effects and toxicity, such as tardive dyskinesia. Finally, seek consultation,
even a quick "curbside consult" with a colleague, and note it in the record.


Clozapine itself has represented little short of a revolution in the treatment of

patients with schizophrenia and other psychotic mood disorders. We are still exploring
the depth and breadth of its therapeutic possibilities. The quantum leap forward that
clozapine has provided in the treatment of these patients rivals that brought about by

the introduction of chlorpromazine in the 1950s. Clozapine unquestionably sets the

standards for the next generation of antipsychotic drugs currently being developed.
Efficacy enhanced beyond that of the original antipsychotics is possible. The challenge
is to push that efficacy even further while, at the same time, diminishing the many
serious toxicities associated with this breakthrough agent. Future approaches may
include novel agents not yet dreamed of, with various sources of therapeutic energy
with which we can enter the brain, the utilization of imaging technology to increase
specificity of therapeutics, brain grafting techniques, and genetic therapies to screen
and correct psychotic conditions before they have had an opportunity to manifest

I. May P. R. A.: Treatment of Schizophrenia; A Comparative Study of Five Treatment Methods. New York,
Science House, 1968.
2. Stanton A. H., Gunderson J. G., Knapp P. H., et al: Effects of psychotherapy in schizophrenia: I. Design
and implementation of a controlled study. Schizophrenia Bull 10:520-563, 1984.
3. Schooler N. R., Keith S. J.: The clinical research base for the treatment of schizophrenia. Psychophar-
macol Bull 29:431-446, 1993.
4. Leff J., Kuipers L., Berkowitz R., Eberlein-Vries R., Sturgeon D.: A controlled trial of social interven-
tion in the families of schizophrenic patients. Br J Psychiatry 141: 121-134, 1982.
5. Kottgen c., Sonnichsen I., Mollenhauer K., et al: The family relations of young schizophrenic patients:
Results of the Hamburg Camberwell Family Interview Study I. Int J Fam Psychiatry 5:61-70, 1984.
6. Gelenberg A J.: The catatonic syndrome. Lancet 1:1339-1341, 1976.
7. Gelenberg A 1.: ECT: Controversies and consensus. Mass Gen Hosp Bioi Ther Psychiatry Newsletter
8:29-32, 1985.
8. Bartlett 1., Bridges P., Kelly D.: Contemporary indications for psychosurgery. Br J Psychiatry 138:507-
511, 1981.
9. Bassuk E. L.: The homelessness problem. Sci Am 251 :40-45, 1984.
10. McEvoy 1. P.: The neuroleptic threshold as a marker of minimum effective neuroleptic dose. Compr
Psychiatry 27:327-335, 1986.
II. McEvoy 1. P., Stiller R. L., Farr R.: Plasma haloperidol levels drawn at neuroleptic threshold doses: A
pilot study. J Clin Psychopharmacol 6: 133-138, 1986.
12. Chouinard G., Jones B. D.: Neuroleptic-induced supersensitivity psychosis: Clinical and pharmacologic
characteristics. Am J Psychiatry 137:16-31, 1980.
13. Ross C. A: Buspirone in the treatment of tardive dyskinesia. Med Hypotheses 22:321-328, 1987.
14. Gardner E. L., Chen J., Paredes W.: Clozapine produces potent antidopaminergic effects anatomically
specific to the mesolimbic system. J Clin Psychiatry 55: 15-22, 1994.
15. Moghaddam B.: Preferential activation of cortical dopamine neurotransmission by clozapine: Functional
significance. J Clin Psychiatry 55:27-29, 1994.
16. Targeting schizophrenia: Pierre Sokoloff on dopamine receptors. Science Watch 5:3-4, 1994.
17. Cohen B. M., Lipinski J. E: In vivo potencies of antipsychotic drugs in blocking alpha-I noradrenergic
and dopamine D-2 receptors: Implications for drug mechanisms of action. Life Sci 39:2571-2580, 1987.
18. Ko G. N., Korpi E. R., Linnoila M.: On the clinical relevance and methods of quantification of plasma
concentrations of neuroleptics. J Clin Psychopharmacol 5:253-262, 1985.
19. Keepers G. A, Clappison V. J., Casey D. E.: Initial anticholinergic prophylaxis for neuroleptic-induced
extrapyramidal syndromes. Arch Gen Psychiatry 40: 1113-1117, 1983.
20. Van Putten T., Mutalipassi L. R., Malkin M. 0.: Phenothiazine-induced decompensation. Arch Gen
Psychiatry 30:102-106, 1974.
21. Ge1enberg A. J., Mandel M. R.: Catatonic reactions to high potency neuroleptic drugs. Arch Gen
Psychiatry 34:947-950, 1977.

22. Rifkin A .• Quitkin E, Klein D. E: Akinesia: A poorly recognized drug-induced extrapyramidal behavior
disorder. Arch Gen Psychiatry 32:642-674, 1975.
23. Wojcik J. D.: Antiparkinson drug use. Mass Gen Hosp Bioi Ther Psychiatry Newsletter 2:5-7, 1979.
24. Gelenberg A. J., Van Putten T., Lavori P., et al: Anticholinergic effects on memory: Benztropine versus
amantadine. J Clin PsychopharmacoI9:180-185, 1989.
25. Gelenberg A. 1.: Amantadine in the treatment of benztropine-refractory extrapyramidal disorders in-
duced by antipsychotic drugs. Curr Ther Res 23:375-380, 1978.
26. Smith J. M.: Abuse of the antiparkinson drugs: A review of the literature. J Clin Psychiatry 41:351-354,
27. Jeste D. V., Wyatt R. J.: Changing epidemiology of tardive dyskinesia: An overview. Am J Psychiatry
138:297-309, 1981.
28. Burke R. E., Fahn S., Jankovic J., et al: Tardive dystonia: Late-onset and persistent dystonia caused by
antipsychotic drugs. Neurology 32: 1335-1346, 1982.
29. Morgenstern H., Glazer W. H., Niedzwiecki D., et a1: The impact of neuroleptic medication on tardive
dyskinesia: A meta-analysis of published studies. Am J Public Health 77:717-724, 1987.
30. Tamminga C. A., Thaker G. K., Moran M., Kakigi T., Gao X. M.: C10zapine in tardive dyskinesia:
Observations from human and animal model studies. J Clin Psychiatry 55:102-106, 1994.
31. Gerlach 1., Peacock L.: Motor and mental side effects of c1ozapine. J Clin Psychiatry 55: 107-109, 1994.
32. Wojcik 1. D., Gelenberg A. 1., Labrie R. A., Mieske M.: Prevalence of tardive dyskinesia in an outpatient
population. Comp Psychiatry 21 :370-380, 1980.
33. Kane J. M., Woerner M., Borenstein M., et al: Integrating incidence and prevalence of tardive dyskin-
esia. IVth World Congress of Biological Psychiatry, Philadelphia, Sept. 8-13, 1985 (Abstract).
34. Glazer W. M., Morgenstern H., Doucette 1. T.: Predicting the long-term risk of tardive dyskinesia in
outpatients maintained on neuroleptic medications. J Clin Psychiatry 54: 133-139, 1993.
35. Gardos G., Cole J. 0., Tarsy D.: Withdrawal syndromes associated with antipsychotic drugs. Am J
Psychiatry 135:1321-1324, 1978.
36. Quitkin E, Rifkin A., Gochfeld L., et al: Tardive dyskinesia: Are first signs reversible? Am J Psychiatry
134:84-87, 1977.
37. Chakos M. H., Lieberman J. A., Bilder R. M., et al: Increase in caudate nuclei volumes of first-episode
schizophrenic patients taking antipsychotic drugs. Am J Psychiatry 151:1430-1431, 1994.
38. Keshavan M. S., Bagwell W. w., Haas G. L., Sweeney 1. A., Schooler N. R., Pettegrew J. w.: Changes in
caudate volume with neuroleptic treatment. Lancet 344:1434, 1994.
39. Doraiswamy P. M., Tupler L. A., Krishnan K. R. R.: Neuroleptic treatment and caudate plasticity. Lancet
345:734-735, 1995.
40. Chakos M. H., Lieberman 1. A., A1vir J., Bilder R., Ashtari M.: Caudate nuclei volumes in schizophrenic
patients treated with typical antipsychotics or cIozapine. Lancet 345:456-457, 1995.
41. Dalgalarrondo P., Gattaz W. E: Basal ganglia abnormalities in tardive dyskinesia: Possible relationship
with duration of neuroleptic treatment. Eur Arch Psychiatry Clin Neurosci 244:272-277, 1994.
42. Baldessarini R. 1., Tarsy D.: Tardive dyskinesia, in Lipton M. A., DiMascio A., Killam K. E (eds):
Psychopharmacology: A Generation of Progress. New York, Raven Press, 1978, pp. 993-1004.
43. Casey D. E., Povisen U. 1., Meidahl B., et al: Neuroleptic-induced tardive dyskinesia and parkinsonism:
Changes during several years of continuing treatment. Psychopharmacol Bull 8:250-253, 1986.
44. Yagi G., Itoh H.: A ten-year follow-up study of tardive dyskinesia-with special reference to the
influence of neuroleptic administration on the long-term prognosis. Keio J Med 34:211-219, 1985.
45. Caroff S. N.: The neuroleptic malignant syndrome. J Clin Psychiatry 41:79-83, 1980.
46. Dabiri L. M., Pasta D., Darby J. K., Mosbacher D.: Effectiveness of vitamin E for treatment of long-term
tardive dyskinesia. Am J Psychiatry 151 :925-926, 1994.
47. Shalev A., Munitz H.: The neuroleptic malignant syndrome: Agent and host interaction. Acta Psychiatr
Scand 73:337-347, 1986.
48. Levinson D. E, Simpson G. M.: Neuroleptic-induced extrapyramidal symptoms with fever: Hetero-
geneity of the "neuroleptic malignant syndrome". Arch Gen Psychiatry 43:839-848, 1986.
49. Gelenberg A. 1., Bellinghausen B., Wojcik J. D., Falk W. E., Farhadi A. M.: Patients with NMS histories:
What happens when they are rehospitalized? J Clin Psychiatry 50:178-180, 1989.

50. Frankenburg E R., Kando J. c., Centorrino E, Gilbert J. M.: Bladder dysfunction associated with
cIozapine therapy. J Clin Psychiatry 57:39-40, 1996.
51. Baldessarini R. J., Gelenberg A J.: Using physostigmine safely. Am J Psychiatry 136: 1608-1609, 1979.
52. Wilt 1. L., Minnema A M., Johnson R. E, Rosenblum A. M.: Torsade de pointes associated with the use
of intravenous haloperidol. Ann Intern Med 119:391-394, 1993.
53. Mitchell A C., Brown K. W: Chlorpromazine-induced retinopathy. Br J Psychiatry 166:822-823, 1995.
54. Potts A M.: Toxic responses of the eye, in Klaassen C. D., Amdur M. 0., Doull J. (eds): Casarett and
Doull's Toxicology; The Basic Science of Poisons, 5th ed.. New York, McGraw-Hili, 1996, pp. 583-
55. Frankenburg E R., Stormberg D., Gerson S. L.: Unsuccessful reexposure to cIozapine. J Clin Psycho-
pharmacol 14:428-429, 1994.
56. Lieberman 1. A, Johns C. A., Kane 1. M., et al: Clozapine-induced agranulocytosis: Non-cross-reactivity
with other psychotropic drugs. J Clin Psychiatry 49:271-277, 1988.
57. Safferman A Z., Lieberman J. A., Alvir J. M. J., Howard A.: Rechallenge in cIozapine-induced
agranulocytosis. Lancet 339: 1296-1297, 1992.
58. Alvir J. M. J., Lieberman 1. A: Agranulocytosis: Incidence and risk factors. J Clin Psychiatry 55:137-
59. Gerson S. L.: G-CSF and the management of cIozapine-induced agranulocytosis. J Clin Psychiatry
55:139-142, 1994.
60. Hummer M., Spemer-Unterweger B., Kemmler G., Falk M., Kurz M., Oberbauer H., Aeischhacker
W. W: Does eosinophilia predict cIozapine-induced neutropenia? Psychopharmacology 124:201-214,
1996. 1992 (Abstract).
61. Banov M. D., Tohen M., Friedberg 1.: High risk of eosinophilia in women treated with cIozapine. J Clin
Psychiatry 54:466-469, 1993.
62. Zarrabi M. H., Zucker S., Miller E, et al: Immunologic and coagulation disorders in chlorpromazine-
treated patients. Ann Intern Med 91:194-199, 1979.
63. Zucker S., Zarrabi H. M., Schuback W H., et al: Chlorpromazine-induced immunopathy: Progressive
increase in serum IgM. Medicine 69:92-100, 1990.
64. Gelenberg A. 1.: Psychotropic drugs during pregnancy and the perinatal period. Mass Gen Hosp Bioi
Ther Psychiatry Newsletter 2:41-42, 1979.
65. Rivera-Calimlim L., Castaneda L., Lasagna L.: Effect of management on plasma chlorpromazine in
psychiatric patients. Clin Pharmacol Ther 14:978-986, 1973.
66. Dysken M. W., Javaid J. I., Chang S. S., et al: Auphenazine pharmacokinetics and therapeutic response.
Psychopharmacology 73:205-210, 1981.
67. Cohen W J., Cohen N. H.: Lithium carbonate, haloperidol, and irreversible brain damage. JAMA
230:1283-1287, 1974.
68. Spring G. K.: Neurotoxicity with combined use of lithium and thioridazine. J Clin Psychiatry 40:135-
138, 1979.
69. Blake L. M., Marks R. C., Luchins D. J.: Reversible neurologic symptoms with cIozapine and lithium. J
Clin Psychopharmacol 12:297-299, 1992.
70. Gelenberg A. 1.: Is it safe to co-prescribe lithium with a neuroleptic? Mass Gen Hosp Bioi Ther
Psychiatry Newsletter 2:13, 1979.
71. Rivera-Calimlim L., Kerzner B., Karch F. E.: Effect of lithium on plasma chlorpromazine level. Clin
Pharmacol Ther 23:451-455, 1978.
72. Finley P., Warner D.: Potential impact of valproic acid therapy on cIozapine disposition. Bioi Psychiatry
36:487-488, 1994.
73. Tiihonen J., Vartiainen H., Hakola P.: Carbarnazepine-induced changes in plasma levels of neuroleptics.
Pharmacopsychiatry 28:26-28, 1995.
74. Longo L. P., Salzman c.: Valproic acid effects on serum concentrations of c10zapine and norcIozapine.
Am J Psychiatry 152:650, 1995.
75. Ge1enberg A J.: Coffee, tea, and antipsychotic drugs revisited. Mass Gen Hosp Bioi Ther Psychiatry
Newsletter 4:42-43, 1981.
76. Gelenberg A. 1.: Psychiatric emergencies: The psychotic patient. Drug Ther 25-36, 1981.

77. Granacher R. P., Jr.: Management of the acutely psychotic patient. Symposium on Antipsychotic Drug
Therapy: Current Concepts and Future Trends, Key Biscayne, Florida, March 28, 1986 (Abstract).
78. Meltzer H. Y.: Clozapine: Is another view valid? Am J Psychiatry 152:821-825, 1995.
79. Carpenter W. T. J., Conley R. R., Buchanan R. w., Breier A., Tamminga C. A.: Patient response and
resource management: Another view of clozapine treatment of schizophrenia. Am J Psychiatry 152:827-
832, 1995.
80. McElroy S. L., Dessain E. C., Pope H. G., et al: Clozapine in the treatment of psychotic mood disorders,
schizoaffective disorder, and schizophrenia. J Clin Psychiatry 52:411-414, 1991.
81. Suppes T., McElroy S. L., Gilbert J., Dessain E. c., Cole J. 0.: Clozapine in the treatment of dysphoric
mania. Bioi Psychiatry 32:270-280, 1992.
82. Meltzer H. Y., Cola P. A.: The pharmacoeconomics of clozapine: A review. J Clin Psychiatry 55:161-
165, 1994.
83. Kane 1. M.: Schizophrenia. N Engl J Med 334:34-41, 1996.
84. Marder S. R., Putten T., Mintz J., et al: Low- and conventional-dose maintenance therapy with flu-
phenazine decanoate. Arch Gen Psychiatry 44:518-521, 1987.
85. Lieberman J. A., Kane J. M., Johns C. A.: Clozapine: Guidelines for clinical management. J Clin
Psychiatry 50:329-338, 1989.
86. Marder S. R.: Atypical drugs. Psychiatric limes 2, 1994.
87. Gelenberg A. J.: Treating the outpatient schizophrenic. Postgrad Med 64:48-56, 1978.
88. Taylor W. J., Caviness M. H. D.: A Textbook/or the Clinical Application o/Therapeutic Drug Monitor-
ing, Irving, Texas, Abbott Laboratories, 1986, p. 467.


Anxiety is a common response to stressful situations, a consequence of nonpsychiatric

medical disorders, an effect induced by drugs or their discontinuation, a cardinal
feature of the anxiety disorders, and a manifestation of other psychiatric disorders. The
practitioner should establish a differential diagnosis, consider the established medica-
tion and nonmedication treatments, help patients understand their problem and the
treatment alternatives, and provide a full therapeutic trial.


A. Symptoms
Anxiety consists of cognitive, physiological, and behavioral symptoms. Cognitive
symptoms include feelings of apprehension or fear, a sense of uncertainty, helpless-
ness, or danger, and frightening or worrisome thoughts. Physiological symptoms in-
clude respiratory symptoms such as shortness of breath, a smothering sensation, hyper-
ventilation, and numb or tingling paresthesias; cardiac symptoms such as chest
tightness, sensations of heart pounding or racing, and increased heart rate; gastrointesti-
nal symptoms such as nausea, abdominal distress, diarrhea, a lump in the throat, and
swallowing problems; neurological symptoms such as light-headedness and tremulous-
ness; neuromuscular symptoms such as tension and headaches; and other autonomic
sensations such as hot flashes or chills, sweating, dry mouth, and urinary urgency.
Behavioral symptoms include avoidance of, escaping from, or freezing in the anxiety-
provoking situation, restlessness, fidgeting, nail-biting, and impairments in concentra-
tion and performance.

Eric M. Reiman, M.D. • Department of Psychiatry, University of Arizona; Samaritan PET Center, Good
Samaritan Regional Medical Center, Phoenix, Arizona 85006.
The Practitioner's Guide to Psychoactive Drugs (Fourth Edition), Gelenberg and Bassuk, eds.
Plenum Publishing Corporation, New York, 1998


Anxiety is considered a "normal" (indeed, adaptive) response to stressful situa-

tions when its onset, intensity, and duration are tolerable and situationally appropriate.
In such situations, anxiety warns us about possible dangers, prepares us for "fight or
flight," focuses our attention on the source of danger, and mobilizes us to respond
effectively. Anxiety is considered "pathological" when its onset, frequency, intensity,
or duration is excessive or situationally inappropriate and when it is associated with
intolerable distress, interferes with our ability to cope with the anxiety-provoking
situation, impairs our educational, occupational, or social activities, or leads to mal-
adaptive coping responses such as avoidance, compUlsions, alcohol abuse, or suicidal
gestures. Pathological anxiety can be viewed as a false alarm: it focuses our attention
on and mobilizes us to respond to normally innocuous cognitive or sensory stimuli,
causes us unnecessary distress, distracts us from everyday challenges, and interferes
with our ability to cope with these challenges.

B. Substrates
1. Neuroanatomical Substrates
The brain regions most commonly implicated in the production of anxiety and fear
include limbic areas (the amygdala, hippocampal formation, and septum), which are
anatomically well situated to attach emotional significance to different kinds of sensory
stimuli; paralimbic areas (orbitofrontal cortex, insular cortex, the temporal poles, the
parahippocampal gyrus, and the cingulate complex), which are the main inputs and
outputs of the limbic areas; the hypothalamus, which receives information from limbic
areas and participates in autonomic and endocrine expressions of anxiety; and certain
brainstem nuclei (such as the locus coeruleus, raphe nuclei, barabrachial nuclei, and
medullary chemoreceptors).
Investigating the effects of selective brain lesions in laboratory rats, LeDoux and
his colleagues l have begun to characterize the neuroanatomical structures and path-
ways that are involved in the dissectable components of anxiety. They demonstrated
that the amygdala participates in the evaluation procedure which labels simple sensory
stimuli (such as a pure tone or flashing light) with emotional significance, that a
projection from the amygdala to the hypothalamus participates in certain autonomic
expressions of anxiety, and that a projection from the amygdala to or through the
midbrain participates in certain behavioral expressions of anxiety. Interestingly, lesions
of the relevant primary sensory areas (e.g., lesions of auditory cortex in rats condi-
tioned to fear a tone) had no effects on the animal's fearful response to the sensory
stimulus, perhaps accounting for the observation that individuals with a phobic disor-
der have difficulty appealing to conscious, rational thought to overcome their fear.
Comparing the effects of antianxiety medications and selective brain lesions on a
laboratory rat's hypervigilant response to novel or unpleasant stimuli, Gray 2 postulated
that the septohippocampal system, its incoming noradrenergic projections from the
locus coeruleus, and its serotonergic projections from the raphe nuclei are involved in
forms of anxiety other than conditioned fear. He postulated that antianxiety medica-
5. ANXIETY 215

tions like the benzodiazepines and barbiturates exert their therapeutic effect by inhibi-
ting the noradrenergic and serotonergic projections to the septohippocampal system.
Using depth electrodes in the presurgical evaluation of awake patients with intrac-
table epilepsy, neurosurgeons found that seizure discharge and electrical stimulation of
the amygdala, hippocampal formation, parahippocampal gyrus, and temporal poles are
most commonly associated with the subjective experience of fear.
Positron-emission tomography (PET) studies find that patients with obsessive-
compulsive disorder have abnormally increased activity in the vicinity of the caudate
and orbitofrontal cortex, that these structures, the thalamus, and the anterior cingulate
gyrus are activated by the elicitation of obsessive-compulsive anxiety, and that antiob-
sessional medication and nonmedication treatments both correct the caudate abnor-
mality. These and other studies led investigators to postulate that the caudate is failing
to filter out adventitious, worrisome thoughts in these patients, that this abnormality
leads to an abnormally active circuit involving the caudate, thalamus, and orbitofrontal
cortex, and that medication, psychological, and neurosurgical treatments exert their
therapeutic effect by interrupting this self-perpetuating circuit.
Other PET studies suggest that patients with panic disorder have a functional
abnormality in the vicinity of the right parahippocampal gyrus, that other paralimbic
regions (such as insular cortex) are activated during experimentally generated panic
attacks (and normal anticipatory anxiety), and that anti panic medication does not
correct the parahippocampal abnormality. According to one theory, the regional brain
abnormality is involved in the genetic predisposition to panic attacks; the abnormal
region responds to a normally innocuous triggering event (such as a normal increase in
locus coeruleus activity) by sending a message to paralimbic alarm centers which are
involved in both pathological and normal forms of anxiety; antipanic medications
interfere with the initiation of panic through its effect on the triggering event (e.g., by
reducing noradrenergic activity); and cognitive-behavioral therapy works downstream,
resetting the alarm centers, and interfering with the elaboration of panic.

2. Neurochemical Substrates

The chemical processes implicated in the development and treatment of anxiety

include, but are not limited to, norepinephrine and its receptors, serotonin and its
receptors, 'Y-aminobutyric acid (GABA) and its receptors, and the benzodiazepine
More than 70% of noradrenergic neurons (i.e., those that synthesize, store, and
release the neurotransmitter norepinephrine) arise in a pontine nucleus called the locus
coeruleus and project to widespread regions of the brain. Stimulation of these neurons
appears to be associated with increased arousal in laboratory rats, increased expres-
sions of anxiety in monkeys, an increased frequency of panic attacks in patients with
panic disorder, and an increased frequency of panic attacks and flashbacks in patients
with posttraumatic stress disorder. Some, but not all, of the medications used in the
treatment of excessive anxiety, panic disorder, and posttraumatic stress disorder de-
crease noradrenergic neurotransmission.

Most serotonergic neurons (i.e., those that synthesize, store, and release serotonin)
arise in the raphe nuclei in the midbrain and project to widespread regions of the brain.
Some studies relate anxiety and some of its disorders to increases in serotonergic
activity, others to decreases. This system may have an especially important role in the
treatment of obsessive-compulsive disorder.
GAB A, the major inhibitory neurotransmitter, is located in intemeurons through-
out the central nervous system. Antianxiety medications, such as the benzodiazepines
and barbiturates, appear to reduce anxiety by facilitating GABA neurotransmission.
When GAB A binds to the postsynaptic GABA A receptor, a chloride channel opens,
causing chloride ions to enter, hyperpolarize, and, thus, inhibit the postsynaptic neuron.
Benzodiazepines, barbiturates, and ethanol all change the conformation of the GABA A
receptor, increasing its affinity for GABA. Benzodiazepines do so by binding to and
activating a postsynaptic benzodiazepine receptor, which exists in a supramolecular
receptor complex with the GABA A receptor and a chloride channel; barbiturates do so
by binding to a separate postsynaptic receptor; and eth.anol appears to do so by altering
the membrane around the GABA A receptor. While GABA inhibits neurons throughout
the brain, some researchers postulate that GABA exerts its antianxiety effects by
inhibiting noradrenergic neurons arising in the locus coeruleus, serotonergic neurons
arising in the raphe nuclei, or both.
Additional chemical processes are likely to have a role in anxiety and its disor-
ders. For instance, investigators recently postulated that noradrenergic, opiate, cor-
ticotropin-releasing factor, and other neurochemical systems are involved in the patho-
physiology and potential treatment of posttraumatic stress disorder.
While researchers continue to uncover the chemical processes that are involved in
the development and treatment of anxiety and its disorders, several caveats are in order.
First, medications act on multiple neurotransmitter systems and neuroreceptor sub-
types, many of which remain to be char~cterized. Second, additional neurotransmitter,
neuropeptide, neuroreceptor, intraneuronal, and endocrine processes are likely to have
a role in different aspects of anxiety, anxiety disorders, and their treatment. Third,
researchers are just beginning to characterize the interactions among different neuro-
transmitter systems. Finally, it is difficult, if not impossible, to make valid inferences
about the pathophysiology of psychiatric disorders on the basis of a medication's
therapeutic effects, since pharmacological agents could work by compensating for
rather than correcting the underlying abnormality.

3. Psychological Substrates

Learning theorists postulate that classical conditioning plays an important role in

the development of anxiety and phobic disorders. When an innocuous situation (the
conditioned stimulus) is paired with an anxiety-provoking situation (the unconditioned
stimulus), the individual learns to respond to the previously innocuous situation with
anxiety. For instance, when an emotionally neutral auditory or visual stimulus is paired
with an electric shock, laboratory rats are conditioned to respond to the sensory stimu-
lus with autonomic and behavioral expressions of fear. Similarly, behavioral therapists
postulate that patients with panic disorder learn to respond to normally innocuous
5. ANXIETY 217

sensory stimuli, such as palpitations or shortness of breath (conditioned stimuli), with

the symptoms of panic. Cognitive-behavioral therapists postulate that panic disorder is
sustained by the habit of responding to panic-eliciting situations and sensory stimuli
with unrealistic, catastrophic thoughts.
Some learning theorists postulate that extinction plays an important role in the
treatment of anxiety and phobic disorders. When an individual is repeatedly exposed to
the conditioned stimulus in the absence of another aversive stimulus, conditioned fear
is eventually extinguished. Extinction may be related to erasure of the fear-generating
association, or to the formation of new associations that mask the fear-generating
Psychoanalysts postulate that anxiety serves as an intrapsychic alarm system,
which warns the individual about forbidden impulses, thoughts, and feelings that are
threatening to emerge from the unconscious and mobilizes mechanisms of defense to
help repress them. Some theorists postulate that phobic disorders reflect a failure to
repress an unconscious aggressive or sexual impulse, and the threat associated with this
impulse is displaced onto a symbolically related external object or situation to disguise
the internal threat and help bolster repression. While these theories are thought-provok-
ing, they remain to be empirically tested.

C. Integrating Biological and Psychosocial Contributions to the Problem

Evidence supports biological contributions to the development and treatment of
anxiety disorders. Consider the evidence for panic disorder. First, approximately 25%
of patients' immediate family members and almost 50% of their daughters develop
probable panic disorder; family and twin studies support the possibility that the disor-
der is transmitted by autosomal dominant inheritance with variable penetrance. Second,
several medications have been demonstrated to block panic attacks in patients with
panic disorder. Third, several biological challenges can precipitate a panic attack more
in patients with panic disorder, but rarely do so in other individuals. Fourth, patients
can have panic attacks at totally unexpected times, including non-REM sleep. Finally,
PET studies suggest that patients have a regional brain abnormality in the nonpanic
Other evidence supports psychosocial contributions to the development and treat-
ment of anxiety disorders. Again, consider the evidence for panic disorder. First, many
patients develop panic attacks that occur in stressful or agoraphobic situations. Second,
the likelihood of having a biologically induced panic attack is affected by the patients'
sense of control. Third, cognitive-behavioral therapy appears to block panic attacks in
many patients.
It is helpful to review this evidence with patients and their families and help them
reconcile biological and psychosocial contributions to the problem. For instance, clini-
cians could provide the following information to those affected by panic disorder: "As
we've discussed, biological-indeed, genetic-factors appear to be involved in the
predisposition to panic attacks. In biologically predisposed individuals, additional bio-
logical, psychological, and social factors appear to determine when a panic attack is
generated. Like other medical problems, panic disorder is not your fault; excessive self-

TABLE 1. Differential Diagnosis of Anxiety

Responses to everyday stressors

Nonpathological reaction to stress
Adjustment disorder
Physiological effect of nonpsychiatric medical disorders
Anxiety disorder due to a generalized medical disorder, such as:
a dementia
a neurological disorder (e.g., temporal lobe epilepsy, vestibulopathies, basal ganglia diseases)
an endocrine disorder (e.g., thyroid disorders, hypoparathyroidism, pheochromocytoma)
a cardiovasular disorder (e.g., paroxysmal atrial tachycardia, ischemic heart disease)
a respiratory disorder (e.g., asthma)
other medical problems (e.g., porphyria)
Physiological effect of substances or their discontinuation
substance-induced anxiety disorder
caffeine, cannabis, cocaine, psychostimulants, thyroxine, and other medications, toxins, or
recreational drugs
Substance withdrawal
ethanol, sedative-hypnotics, or other drugs
Anxiety disorders
Panic disorder with or without agoraphobia
Agoraphobia without a history of panic disorder
Social phobia, circumscribed or generalized type
Specific phobias
Obsessive-compulsive disorder
Acute or posttraumatic stress disorder
Generalized anxiety disorder
Other psychiatric disorders
Disorders first diagnosed in infancy or childhood (e.g., autism, attention-deficit/hyperactivity disorder,
separation anxiety disorder)
Mood disorders (e.g., major depressive disorder, bipolar disorder)
Psychotic disorders (e.g., schizophrenia)
Somatoform disorders (e.g., somatization disorder, hypochondriasis)
Personality disorders (e.g., avoidant, borderline, or histrionic personality disorders)

blame is neither warranted nor helpful. At the same time, we need to work together to
address the problem in the most constructive way possible. There are medication and
nonmedication treatments that have been established to block panic attacks. Once the
panic attacks are blocked, I will encourage you to confront and learn to overcome those
situations which you have come to fear and avoid. Let's review the options and decide
the best way to proceed at this time .... "

D. Differential Diagnosis
The differential diagnosis of anxiety is summarized in Table 1 and discussed
5. ANXIETY 219

1. The Anxious Response to Everyday Stressors

Nonpathological reactions to stress are characterized by anxiety that is situa-
tionally appropriate in its intensity and duration, tolerable, does not interfere with one's
ability to manage the stressful situation, and does not interfere with one's social,
educational, or occupational activities. In contrast, an adjustment disorder (with
anxiety, with mixed anxiety and depressed mood, or with mixed disturbance of
emotions and conduct) is characterized by anxiety that is situationally excessive,
difficult to tolerate, interferes with one's ability to manage the stressful situation, or
interferes with one's social, educational, or occupational activities. By definition, ad-
justment disorders begin within three months after the onset of the stressor and last less
than six months after the stressor is terminated. These disorders are common, appar-
ently affecting 5-20% of patients in mental health outpatient clinics and many of those
who are exposed to nonpsychiatric medical stressors.
Patience, listening, understanding, and empathy (a nonjudgmental effort to under-
stand what it is like to be in the patient's situation) can often reduce patients' feelings of
helplessness and social isolation, increase their confidence and self-esteem, and help
them consider constructive coping alternatives. A benzodiazepine can have a beneficial
effect if it is used as a temporary, adjunctive treatment, which supports rather than
substitutes for other coping efforts. A benzodiazepine can have a detrimental effect if it
is used as a substitute for the practitioner's time or concern, distracts the clinician and
patient from other constructive coping efforts, is used in patients who are searching for
"the magic pill," or is used in patients with a history of an alcohol, sedative-hypnotic,
or other psychoactive substance use disorder.

2. Anxiety Disorders

a. The Physiological Effect of a Nonpsychiatric Medical Condition. Anxiety

disorders due to a generalized medical condition are characterized by the presence of
generalized anxiety, panic, or (less commonly) obsessive-compulsive symptoms that
appear to reflect the physiological response to a nonpsychiatric medical condition.
These nonpsychiatric medical conditions include cognitive disorders (e.g., the demen-
tias), other neurologic disorders (e.g., temporal lobe epilepsy, vestibulopathies, and
basal ganglia diseases), endocrine disorders (e.g., hyper- and hypothyroidism, hypo-
parathyroidism, pheochromocytoma, hyperadrenocorticism, and the hypoglycemic re-
sponse to exogenous insulin), metabolic disorders (e.g., porphyria, vitamin B12 defi-
ciency), cardiovascular disorders (e.g., paroxysmal atrial tachycardia, coronary artery
disease), respiratory diseases (e.g., asthma), and other medical problems.
The identification of a nonpsychiatric medical problem does not exclude the
possibility that the patient has a coexisting anxiety disorder. For instance, the presence
of mitral valve prolapse does not exclude the coexisting diagnosis of panic disorder.
Although patients with panic disorder have an increased risk of mitral valve prolapse,
this usually benign medical condition does not appear to affect the patients' natural
history of panic attacks, family history of panic disorder, or response to antipanic

Still, it is important to consider the possibility that the patient's anxiety symptoms
are the direct physiological consequence of a medical condition. Thus, it is helpful to
consider a temporal association between the onset, exacerbation, or remission of non-
psychiatric medical and anxiety symptoms, especially in patients with atypical features
of an anxiety disorder, such as atypical age of onset, symptoms, or clinical course.
Laboratory tests are considered on the basis of the patients' psychiatric and medical
histories, review of symptoms, and relevant physical findings. For instance, findings of
weight loss, heat intolerance, and ocular, dermatological, or hair changes should lead
the practitioner to consider the possibility that the patient has hyperthyroidism.

b. The Physiological Effect of a Substance or Its Discontinuation. Substance-

induced anxiety disorders are characterized by anxiety symptoms that are a direct
physiological effect of substance intoxication (e.g., caffeine, cannabis, cocaine, psy-
chostimulants, thyroxine, other prescription medications, other drugs, or other toxins)
or substance withdrawal (e.g., ethanol, sedative-hypnotics, or other drugs). It is impor-
tant to consider the temporal association between substance use, or substance with-
drawal, and the onset, exacerbation, or remission of anxiety symptoms.

c. Panic Disorder. Panic disorder is characterized by recurrent panic attacks, at

least some of which occur at unexpected times. A panic attack is a sudden, discrete
episode of intense fear or discomfort that is accompanied by at least four of the
following symptoms: shortness of breath or smothering sensations; choking sensations;
numb or tingling paresthesias; palpitations or increased heart rate; chest pain or dis-
comfort; dizziness or light-headedness; trembling or shaking; sweating; hot flashes or
chills; nausea or abdominal distress; feelings of unreality or depersonalization; fear of
losing control or going crazy; and fear of dying. (Episodes which are accompanied by
only two or three symptoms are called limited-symptom attacks.)
The best way to recognize a panic attack is by characterizing its temporal features.
By definition, a panic attack is abrupt, reaching maximal intensity within 10 minutes.
In most cases, a panic attack is also brief, usually lasting 2-30 minutes and rarely
lasting longer than one hour. (Since a panic attack is brief, health care providers rarely
have the chance to observe an episode directly; practitioners often underestimate the
impact of this disorder on patients and their families.)
While it is important to document a history of at least some panic attacks that
begin at unexpected times, many patients have attacks that occur in times of stress or in
situations which they have come to fear. Many patients will visit the emergency room,
cardiologists, neurologists, and other nonpsychiatric physicians, urgently seeking a
medical explanation for the problem.
Panic disorder is a common, distressing, and potentially disabling problem. It
afflicts 1.5-3.5% of the population, affects females twice as commonly as males and
typically begins in the late teens to early thirties. As a consequence of recurrent panic
attacks, patients can develop anxiety and concern about having additional attacks,
situations in which they might have these attacks, and the consequences of these
attacks, such as dying, going crazy, or losing control. If the disorder is not properly
recognized or treated, patients can feel discouraged, demoralized, ashamed, and social-
5. ANXIETY 221

ly isolated. Approximately one-third to one-half of patients with panic disorder develop

agoraphobia, a fear and avoidance of situations from which it would be difficult or
embarrassing to escape in the event of panic symptoms. Other complications include
major depressive disorder, marital problems, suicide attempts, and excess mortality
from suicide and cardiovascular causes. In a 20-year follow-up study of patients with
recurrent panic attacks conducted prior to the advent of established treatment, approx-
imately five out of six patients were still symptomatic and about half were at least
mildly disabled. Established treatments include certain antidepressants, monoamine
oxidase inhibitors, high-potency benzodiazepines, and cognitive-behavioral therapy.

d. Agoraphobia. Agoraphobia is a fear or avoidance of situations in which it is

difficult or embarrassing to escape in the event of a panic attack or panic-like symp-
toms. In clinical settings, almost all patients who present with agoraphobia have a
history of recurrent panic attacks. Thus, over 95% of these patients have panic disor-
der with agoraphobia. Although the Epidemiological Catchment Area study raised
the possibility that there were many other individuals with agoraphobia without a
history of panic disorder who did not seek treatment, a recent investigation suggests
that the majority of these individuals had a specific phobia. Still, it is conceivable that
some patients fear and avoid situations that might be associated with panic-like symp-
toms, such as incapacitating dizziness, or other symptoms, such as an embarrassing loss
of bladder control. This diagnosis could also be given to patients with medical prob-
lems, such as Crohn's disease, which are associated with realistic concerns, such as
diarrhea, but only if the fear or avoidance appears excessive for that condition.
Left untreated, agoraphobia is a chronic and disabling condition. Agoraphobic
situations include driving, airplanes, elevators, theaters, restaurants, or, in its most
extreme form, any place outside the home. Some patients become dependent on their
companions or friends to help them confront these situations or compensate for their
Although patients can respond extremely well to treatment, it seems that the
longer they have lived with agoraphobic fear and avoidance, the harder it is to treat and
the longer it takes to improve. Like other phobic disorders, the best-established treat-
ment is exposure therapy, which consists of frequent and prolonged exposure to in-
creasingly intense phobic situations. In patients with panic disorder with agoraphobia,
it is crucial to look for and treat their recurrent panic attacks.

e. Social Phobia. Sometimes called social anxiety disorder, social phobia is

characterized by an excessive, persistent fear of being scrutinized and embarrassed in
performance or social situations. While many people have stage fright, performance
anxiety, or shyness in social situations, the diagnosis of social phobia is reserved for
those individuals whose distress and disability are clinically significant. These individ-
uals recognize that the fear is unreasonable, excessive, and problematic; their avoid-
ance, anticipatory anxiety, or distress in the feared situation interferes with their educa-
tional, occupational, or social activities.
Individuals with social phobia, circumscribed (or discrete) type, have a persis-
tent, excessive fear of certain performance situations, such as a fear of public speaking

(the most common fear), a fear of other kinds of public performances (such as playing a
musical instrument), or fears of eating, drinking, writing, or using rest rooms in the
presence of others. Individuals with social phobia, generalized type, have a persistent,
excessive fear of most performance and social situations; in addition to several of the
performance fears noted above, they fear social situations such as speaking in small
groups, dating, calling strangers on the telephone, returning goods to a store, and
attending parties.
It is also important to make the distinction between social phobia and
agoraphobia. In contrast to social phobia, panic disorder with agoraphobia is charac-
terized by at least some panic attacks that occur at unexpected times; panic disorder
with agoraphobia and agoraphobia without a history of panic disorder are not always
limited to those situations involving possible scrutiny by others (e.g., driving, eleva-
In the feared situation, there often appears to be a vicious cycle of anticipatory
fear of failure or embarrassment, somatic anxiety symptoms, perceived less of control,
actual or perceived performance problems, embarrassment, anticipatory fear, and so
forth. The fear of public speaking is often exacerbated by a fear of heart racing,
trembling hands, and a trembling voice; the fear of eating or drinking in public is often
related to a fear of hand trembling (as the eating utensil reaches the mouth) or choking
on one's food (perhaps related to anxiety-related esophageal motility problems); the
fear of writing checks in public is characteristically related to a fear of hand trembling;
and the fear of using public rest rooms is often related to a fear of urinary hesitancy.
Social phobia appears to affect up to 13% of the population, depending on the
number of social phobic stressors investigated and the threshold used to characterize
clinically significant distress and disability, and is twice as common in females as in
males. The majority of individuals with this disorder have a fear of more than one kind
of performance or social situation. Social phobia typically begins in the mid-teenage
years, lasts for many years, and waxes and wanes in accordance with demands related
to the individual's social, academic, and occupational advancement, such as those
involving dating, oral reports and oral examinations in school, and work-related oral
presentations. The generalized type of social phobia is often characterized by a child-
hood history of social inhibition and shyness and enduring problems related to poor
self-esteem, hypersensitivity to rejection, and difficulty with assertiveness. Indeed,
some individuals with the generalized type of social phobia also satisfy criteria for
avoidant personality disorder. Family, twin, and developmental studies support the
possibility of a genetic contribution to the problem.

f. Specific Phobia. Previously known as a simple phobia, a specific phobia is a

persistent, excessive fear of specific objects or situations, such as small animals,
insects, heights, blood, and air travel. Individuals with a specific phobia recognize that
their fear is excessive and unreasonable; they avoid the phobic situation or endure it
with intense anxiety. Although phobias are extremely common, (affecting about lO-
II % of the population), they usually do not result in sufficient distress or disability to
satisfy the criteria for a diagnosis. They are more common in females than in males.
The fourth edition of the Diagnostic and Statistical Manual of Mental Disorders
5. ANXIETY 223

(DSM-IV) distinguishes several types of specific phobia by their natural history, psy-
chophysiological characteristics, and familial transmission. Specific phobia, animal
type, is characterized by an excessive fear of certain animals or insects and is more
common in an individual's first-degree relatives. Specific phobia, natural environment
type, is characterized by fear of environmental situations, such as storms, water, and
heights. Specific phobia, blood-injection-injury type, is characterized by a fear of
seeing blood, injuries, injections, or other invasive medical procedures. In contrast to
the other phobias, it is usually characterized by a strong vasovagal response, with
decreased heart rate, decreased blood pressure, and occasional syncope. Specific pho-
bia, situational type, is characterized by a fear of specific situations, such as driving,
tunnels, bridges, elevators, enclosed places, and flying and is more common in an
individual's first-degree relatives. In contrast to agoraphobia, this and other types of
specific phobia are not characterized by a history of unexpected panic attacks; the focus
of fear is the situation itself rather than a panic attack or its consequences. Specific
phobia, other type, is characterized by a fear of other stimuli, such as costumed
characters or loud noises (which sometimes occurs in children), choking, vomiting, or
acquiring an illness. Individuals with one phobia appear more likely to have additional
phobias within the same SUbtype.

g. Obsessive-Compulsive Disorder. Obsessive-compulsive disorder (OCD) is

characterized by obsessions or compulsions that are distressing, disabling, or time-
consuming (at least one hour each day). Obsessions are intrusive, unwanted thoughts,
impulses, or images that insistently enter the mind; they cause significant distress and
elicit efforts to suppress or neutralize them. Most individuals with OCD recognize that
their obsessions are excessive and unreasonable; nevertheless, they cannot suppress the
obsessions or shake the possibility that their obsessional fears might be realized.
Almost half of individuals with OCD have obsessions related to the fear of dirt,
disease, or contamination. In one study of OCD patients, 45% had contamination
obsessions (e.g., intrusive thoughts of becoming contaminated when shaking some-
one's hands); 42% had pathological doubts (e.g., recurrent doubts about whether or not
a door was left unlocked); 36% had somatic obsessions (which were distinguished from
hypochondriasis and a specific phobia by the presence of associated rituals);· 31 % had
the obsessional need to have things symmetrical or in a certain order; 28% had aggres-
sive or horrific obsessions (e.g., intensely distressing impulses to hurt one's child or
shout an obscenity in church); 26% had sexual obsessions (e.g., unwanted, intrusive,
distressing pornographic images involving the individual's parents); and 13% had other
kinds of obsessions. Sixty percent of these patients had mUltiple obsessions. It is
important to distinguish obsessions from excessive worries about real-life problems.
Compulsions are repetitive, ritualistic behaviors that individuals feel driven to
perform in order to neutralize their obsessions or obsession-elicited anxiety. In the
study noted above, 63% of OCD patients had checking rituals (e.g., checking the door a
large, sometimes "magical," number of times in order to reduce distressing doubts
about whether or not it is locked); 50% had washing or cleaning rituals (e.g., repeatedly
washing one's hands, sometimes until they are raw, in order to reduce contamination
fears); 36% had counting rituals; 31% had asking, demanding, or confessing rituals;

28% had rituals involving ordering, symmetry, and precision; and 18% had hoarding
rituals. About half of these patients had multiple compulsions.
OCD is not the same thing as obsessive-compulsive personality disorder, a condi-
tion which is characterized by a disabling preoccupation with details and rules, dis-
abling perfectionism, excessive devotion to work and productivity, inflexible insistence
on certain ethical values, an inability to discard worthless objects, a reluctance to
delegate tasks to others, a miserly spending style, and stubborn insistence that others
conform to one's way of doing things. In one study, only 14% of OCD patients had
obsessive-compulsive personality features.
Less than one generation ago, OCD was considered an extremely rare disorder,
affecting only 1 in 2000 individuals. Individuals with OCD were unlikely to volunteer
these embarrassing, seemingly "crazy," presumably untreatable symptoms. Clinicians
were unlikely to ask about obsessions or compulsions. Since then, studies have consis-
tently demonstrated that OCD is a common disorder, affecting 2.5% of the population.
If the disorder is recognized and properly treated, symptoms can be significantly
reduced. Thus, it is important to suspect the possibility of OCD. Consider using a
checklist (such as the Yale-Brown Obsessive Compulsive Scale checklist) to identify
symptoms that patients have difficulty volunteering.
Although OCD is equally common in males and females, it tends to develop in the
early teens in males and the early twenties in females. About 10% of cases begin prior
to the age of 10, and about 15% begin after the age of 35. In most cases, OCD is a
chronic, waxing and waning, mildly to moderately disabling condition; in about 5% of
cases, it is episodic; and in about 15% of cases, it is a progressively severe, disabling
disorder. About 10% of individuals have OCD with poor insight, in which the individu-
al does not recognize that the obsessions or compulsions are excessive or unreasonable.
This form of the disorder appears to be more progressive, more disabling, and less
responsive to antiobsessive treatments.
Patients with OCD often have additional problems, such as depressed mood,
major depressive disorder, social phobia, eating disorders, alcohol and sedative-hyp-
notic abuse, panic disorder, Tourette's disorder, and personality disorders. OCD pa-
tients can spend much of their day engaged in the battle between anxiety-eliciting
obsessions and anxiety-relieving but disabling compulsions. Others can avoid an in-
creasing number of situations (e.g., some individuals can avoid handshaking, rest
rooms, and all sorts of situations that are perceived to be associated with possible
contamination). Like those with other obsessions, individuals with intrusive, distress-
ing aggressive or sexual urges do not appear to have an increased risk of suicide
attempts, antisocial behaviors, or acting out on their obsessions.
The differential diagnosis of OCD includes basal ganglia disorders, which are
sometimes associated with obsessive-compulsive symptoms; hypochondriasis, which
is associated with recurrent, distressing thoughts about physical symptoms and ill-
nesses; and "compulsive" eating, gambling, sexual activity, and substance use and
other behaviors which are intrinsically pleasurable but which lead to deleterious conse-
quences. (The role of serotonin reuptake inhibitors in the treatment of these disorders is
now being investigated.) In those individuals who have OCD with poor insight, it is
sometimes difficult to distinguish the disorder from schizophrenia.

Based on their compulsive features, familial transmission, or response to serotonin

reuptake inhibitors, some investigators have proposed that there may be a spectrum of
OCD-related disorders. For instance, trichotillomania is characterized by compUlsive
hair pulling, a response to the serotonin reuptake inhibitors clomipramine and fluox-
etine (though tolerance may develop), and an increased risk of OCD in first-degree
relatives. Tourette's disorder is characte