77 (3) Bagus

Clinical Pharmacology & Therapeutics
Copyright 2005 American Society for Clinical Pharmacology and Therapeutics

Volume 77, Issue 3, Pages 113-233 (March 2005)

1.

Inside front cover CONTENTS LIST
Page CO2
2.

Table of contents CONTENTS LIST
Pages A1-A4
3.

Editorial board EDITORIAL BOARD
Pages A5-A6
4.

Information for authors MISCELLANEOUS
Pages A7-A10
5.

American Society for Clinical Pharmacology and Therapeutics Position
Statement on dietary supplement safety and regulation ARTICLE
Pages 113-122
Jason D. Morrow, Timi I. Edeki, Mohamed El Mouelhi, Raymond E.
Galinsky, Rose Kovelesky, Robert J. Noveck, Charles Preuss, American
Society for Clinical Pharmacology and Therapeutics Task Force on Dietary
Supplements
6.

Pharmacogenetics and response to -adrenergic receptor antagonists in
heart failure ARTICLE
Pages 123-126
Mordechai Muszkat and C. Michael Stein
7.

-adrenergic receptor polymorphisms and responses during titration of
metoprolol controlled release/extended release in heart failure ARTICLE
Pages 127-137
Steven G. Terra, Daniel F. Pauly, Craig R. Lee, J. Herbert Patterson,
Kirkwood F. Adams, Jr, Richard S. Schofield, Bernadette S. Belgado, Karen
K. Hamilton, Juan M. Aranda et al.

8.

Polymorphism screening in the cardiac K
+
channel gene KCNA5
ARTICLE
Pages 138-144
Chantale Simard, Benoit Drolet, Ping Yang, Richard B. Kim and Dan M.
Roden
9.

Implications of CYP2A6 genetic variation for smoking behaviors and
nicotine dependence ARTICLE
Pages 145-158
Viba Malaiyandi, Edward M. Sellers and Rachel F. Tyndale
10.

Influence of menstrual cycle on cytochrome P450 2A6 activity and
cardiovascular effects of nicotine ARTICLE
Pages 159-169
Janne Hukkanen, Steven G. Gourlay, Saraswati Kenkare and Neal L.
Benowitz
11.

Effect of grapefruit juice volume on the reduction of fexofenadine
bioavailability: Possible role of organic anion transporting polypeptides
ARTICLE
Pages 170-177
George K. Dresser, Richard B. Kim and David G. Bailey
12.

Inhibition of human intestinal wall metabolism by macrolide antibiotics:
Effect of clarithromycin on cytochrome P450 3A4/5 activity and
expression ARTICLE
Pages 178-188
Amar G. Pinto, Ying-Hong Wang, Naga Chalasani, Todd Skaar, Dhanashri
Kolwankar, J. Christopher Gorski, Suthat Liangpunsakul, Mitchell A.
Hamman, Million Arefayene and Stephen D. Hall
13.

Association between adherence measurements of metoprolol and health
care utilization in older patients with heart failure ARTICLE
Pages 189-201
Wanzhu Tu, Andrew B. Morris, Jingjin Li, Jingwei Wu, James Young, D.
Craig Brater and Michael D. Murray
14.

BIBN4096BS antagonizes human -calcitonin gene related peptide
induced headache and extracerebral artery dilatation ARTICLE
Pages 202-213
Kenneth A. Petersen, Lisbeth H. Lassen, Steffen Birk, Lynna Lesko and Jes
Olesen

15.

Peginterferon alfa-2a does not alter the pharmacokinetics of methadone
in patients with chronic hepatitis C undergoing methadone maintenance
therapy ARTICLE
Pages 214-224
Mark Sulkowski, Teresa Wright, Stephen Rossi, Sanjeev Arora, Matthew
Lamb, Ka Wang, Jean-Michel Gries and Sreeni Yalamanchili
16.

On the move MISCELLANEOUS
Page 224
17.

Disposition of cisapride appears to be influenced by P-glycoprotein in
the mouse CORRESPONDENCE
Pages 225-226
Edna F. Choo, Karthik Venkatakrishnan, Heather L. Hatch and Sandhya
Rahematpura
18.

Modulation of celecoxib pharmacokinetics by food in pediatric patients
CORRESPONDENCE
Pages 226-228
Diana Stempak, Janet Gammon, Jacqueline Halton, Martin Champagne,
Gideon Koren and Sylvain Baruchel
19.

Limited association of the 2988g>a single nucleotide polymorphism with
CYP2D6*41 in black subjects CORRESPONDENCE
Pages 228-230
Andrea Gaedigk, Liliane Ndjountch, J. Steven Leeder and L. DiAnne
Bradford
20.

Limited association of the 2988g>a single nucleotide polymorphism with
CYP2D6*41 in black subjects: Reply CORRESPONDENCE
Pages 230-231
Ulrich M. Zanger, Matthias Schwab, Claudia Toscano, Michel Eichelbaum
and Sebastian Raimundo
21.

A message from the ASCPT membership committee chairperson
ANNOUNCEMENT
Pages 232-233
John T. Sullivan
22.

Information for readers MISCELLANEOUS
Page A13
EDITOR
C. MICHAEL STEIN, MD Associate Professor of Medicine
and Pharmacology, Division of Clinical Pharmacology, Vander-
bilt University School of Medicine, Nashville, Tenn
EDITOR EMERITUS
MARCUS M. REIDENBERG, MD
FOUNDING EDITOR
WALTER MODELL, MD (Deceased)
ASSOCIATE EDITORS
ARTHUR J. ATKINSON, JR, MD Office of the Director,
Clinical Center, National Institutes of Health, Bethesda, Md
RICHARD B. KIM, MD Associate Professor of Medicine and
Pharmacology, Division of Clinical Pharmacology, Vanderbilt
University School of Medicine, Nashville, Tenn
DAVID W. NIERENBERG, MD Professor of Medicine and
Pharmacology/Toxicology, Chief, Division of Clinical Pharma-
cology, Dartmouth-Hitchcock Medical Center, Lebanon, NH
GRANT R. WILKINSON, PHD, DSC Professor of Pharma-
cology, Division of Clinical Pharmacology, Vanderbilt University
School of Medicine, Nashville, Tenn
SENIOR ISSUE MANAGER
KAY G. GOEHLER Elsevier Inc., St Louis, Mo
EDITORIAL BOARD
DARRELL R. ABERNETHY, MD, PHD Clinical Director,
National Institute on Aging, National Institutes of Health, Ger-
ontology Research Center, Baltimore, Md
ADEDAYO ADEDOYIN, PHD Department of Drug Dispo-
sition and Metabolism, Zeneca Pharmaceuticals, Wilmington, Del
MATTHEW M. AMES, PHD Professor of Pharmacology,
Mayo Clinic and Foundation, Rochester, Minn
FREDY. AOKI, MD Professor of Medicine, Medical Microbi-
ology, and Pharmacology and Therapeutics, University of Mani-
toba School of Medicine, Winnipeg, Manitoba, Canada
GLEN APSELOFF, MD Director of Clinical Pharmacology,
The Ohio State University College of Medicine, Columbus, Ohio
LARRY A. BAUER, PHARMD Professor, Departments of
Pharmacy and Laboratory Medicine, Schools of Pharmacy and
Medicine, University of Washington, Seattle, Wash
LEIF BERTILSSON, PHD Professor, Department of Clinical
Pharmacology, Karolinska Institute at Huddinge University Hos-
pital, Stockholm, Sweden
TERRENCE F. BLASCHKE, MD Chief, Division of Clinical
Pharmacology, Stanford University School of Medicine,
Stanford, Calif
R. A. BRANCH, MD Professor of Medicine and Pharmacol-
ogy, University of Pittsburgh School of Medicine, Pittsburgh, Pa
ERIC P. BRASS, MD, PHD Chair, Department of Medicine,
Harbor-UCLA Medical Center, Professor of Medicine, UCLA
School of Medicine, Torrance, Calif
D. CRAIGBRATER, MD Dean, Indiana University School of
Medicine, Indianapolis, Ind
ALASTAIR BRECKENRIDGE, MSC, MD Professor, Uni-
versity of Liverpool, Liverpool, United Kingdom
DOUWE D. BREIMER, PHD Professor Doctor, Rector Mag-
nificus of Leiden University, Leiden, The Netherlands
MICHAEL E. BRIER, PHD Associate Professor of Medicine,
Kidney Disease Program, University of Louisville, and the De-
partment of Veterans Affairs, Louisville, Ky
JAMES F. BURRIS, MD Deputy Chief Research and Devel-
opment Officer, Veterans Health Administration, and Clinical
Professor of Medicine and Pharmacology, Georgetown Univer-
sity School of Medicine, Washington, DC
S. GEORGE CARRUTHERS, MD, FRCPC Dean, United
Arab Emirates University, Al Ain, United Arab Emirates
DENNIS J. CHAPRON, RPH, MS Associate Professor of
Pharmacy, School of Pharmacy, University of Connecticut,
Storrs, Conn
DAVID M. COCCHETTO, PHD Group Director, Regula-
tory Affairs, GlaxoSmithKline, Research Triangle Park, NC, and
Adjunct Associate Professor, School of Pharmacy, University of
North Carolina, Chapel Hill, NC
C. LINDSAY DEVANE, PHARMD Professor of Psychiatry and
Pharmacy, Medical University of South Carolina, Charleston, SC
COLIN DOLLERY, MB, CHB, BSC Senior Consultant,
GlaxoSmithKline, Harlow, Essex, United Kingdom
MARGARET S. DORDAL, MD, PHD Tapp Holdings, Inc,
Deerfield, Ill
AKIO EBIHARA, MD Director, Kaminokawa Hospital,
Kawachi-gun, Tochigi-ken, Japan
DAVID J. EDWARDS, PHARMD Professor, Eugene Apple-
baum College of Pharmacy and Health Sciences, Wayne State
University, Detroit, Mich
MERRILL J. EGORIN, MD University of Pittsburgh Cancer
Institute, Pittsburgh, Pa
THOMAS EISSENBERG, PHD Assistant Professor of Psy-
chology and Institute for Drug and Alcohol Studies, Virginia
Commonwealth University, Richmond, Va
WILLIAM E. EVANS, PHARMD First Tennessee Professor of
Clinical Pharmacy and Pediatrics, University of Tennessee, and
Scientific Director and Deputy Director, St Jude Childrens Re-
search Hospital, Memphis, Tenn
ROSS D. FELDMAN, MD R. W. Gunton Professor of Ther-
apeutics, University of Western Ontario, London, Ontario, Canada
CHARLES FLEXNER, MD Division of Clinical Pharmacol-
ogy, The Johns Hopkins Hospital, Johns Hopkins University,
Baltimore, Md
MARILYNNC. FREDERIKSEN, MD Associate Professor of
Obstetrics and Gynecology, Clinical Pharmacology Center,
Northwestern University Medical School, Chicago, Ill
CLINICAL PHARMACOLOGY & THERAPEUTICS MARCH 2005 Page 5A
CLINICAL
PHARMACOLOGY
T
HERAPEUTICS &
KATHLEEN GIACOMINI, PHD Associate Professor, Uni-
versity of California at San Francisco Schools of Pharmacy and
Medicine, San Francisco, Calif
JEANGRAY, MD, FRCP(C) Professor of Medicine and Phar-
macology, Dalhousie University, Halifax, Nova Scotia, Canada
PERRY V. HALUSHKA, MD, PHD Professor of Pharmacol-
ogy and Medicine, Medical University of South Carolina,
Charleston, SC
THOMAS K. HENTHORN, MD Professor and Chair of An-
esthesia, University of Colorado Health Sciences Center, Denver,
Colo
LINDA A. HERSHEY, MD, PHD Professor of Neurology
and Pharmacology, State University of New York at Buffalo,
Buffalo, NY
PETER H. HINDERLING, MD Senior Advisor, Office of
Clinical Pharmacology and Biopharmaceutics, Center for Drug
Evaluation and Research, Food and Drug Administration, Rock-
ville, Md, and Adjunct Professor, Department of Medicine,
Georgetown University, Washington, DC
TAKASHI ISHIZAKI, MD Department of Pharmacotherapy,
Graduate School of Pharmacy, Kumamoto University,
Kumamoto, Japan
PATRICE JAILLON, MD Professor of Clinical Pharmacol-
ogy, Hopital Saint-Antoine, Paris 6 University, Paris, France
WILLIAM J. JUSKO, PHD Professor of Pharmaceutics, State
University of New York at Buffalo, Buffalo, NY
SHIGERUKAGEYAMA, MD, PHD Head, Clinical Pharma-
cology and Therapeutics Unit, The Jikei University School of
Medicine, Tokyo, Japan
HELEN KASTRISSIOS, PHD Assistant Professor, Colleges
of Pharmacy and Medicine, University of Illinois, Chicago, Ill
JOAN M. KORTH-BRADLEY, PHARMD, PHD Director,
Clinical Pharmacokinetics, Wyeth-Ayerst Research, Philadelphia,
Pa
JUANJ. L. LERTORA, MD, PHD Professor of Medicine and
Pharmacology, Tulane University School of Medicine,
New Orleans, La
LIONEL D. LEWIS, MB, BCH, MD Associate Professor of
Medicine and Pharmacology/Toxicology, Division of Clinical
Pharmacology, Dartmouth Hitchcock Medical Center, Lebanon,
NH
DAVIDT. LOWENTHAL, MD, PHD Professor of Medicine,
Pharmacology, and Exercise Science, University of Florida,
Gainesville, Fla
CARL V. MANION, MD Department of Medicine, Baptist
Hospital, Oklahoma City, Okla
MICHAEL C. ORME, MB, BS Professor of Pharmacology
and Therapeutics, University of Liverpool, Liverpool, United
Kingdom
RUSSELL K. PORTENOY, MD Chairman, Department of
Pain Medicine and Palliative Care, Beth Israel Medical Center,
New York, NY, and Professor of Neurology and Anesthesiology,
Albert Einstein College of Medicine, Bronx, NY
WILLIAM Z. POTTER, MD, PHD Lilly Research Fellow,
Lilly Research Laboratories, Indianapolis, Ind
MARKJ. RATAIN, MD Professor of Medicine and Chairman,
Committee on Clinical Pharmacology, The University of Chi-
cago, Chicago, Ill
MARY V. RELLING, PHARMD Associate Member, Pharma-
ceutical Department, St Jude Childrens Hospital, and Associate
Professor, University of Tennessee, Memphis, Tenn
JANICE B. SCHWARTZ, MD Director, Long-term Care Re-
search Center, Goldman Institute on Aging/Jewish Home, Uni-
versity of California, San Francisco, San Francisco, Calif
EDWARD M. SELLERS, MD, PHD, FRCP(C) Director,
Psychopharmacology and Dependence Research Unit, Womens
College Hospital, Toronto, Ontario, Canada
STEVEN L. SHAFER, MD Anesthesiology Service, Palo Alto
VA Health Care System, Department of Anesthesia, School of
Medicine, Stanford University, Stanford, Calif
ALEXANDER M. M. SHEPHERD, MD, PHD Professor of
Medicine and Pharmacology, Chief, Division of Clinical Pharma-
cology, The University of Texas Health Science Center, San An-
tonio, Tex
SANG-GOO SHIN, MD, PHD Professor of Clinical Pharma-
cology/Pharmacology, Director of Clinical Pharmacology Unit
and Clinical Trial Center/SNUH, Seoul National University
College of Medicine, Seoul, Korea
DANIEL S. SITAR, PHD Professor, Department of Medicine
and Pharmacology and Therapeutics, University of Manitoba,
Winnipeg, Manitoba, Canada
FOLKE SJO
QVIST, MD, PHD Professor of Clinical Pharma-

cology, Department of Clinical Pharmacology, Karolinska Insti-
tute at Huddinge University Hospital, Stockholm, Sweden
DONALD R. STANSKI, MD Vice President, Scientific and
Medical Program, Pharsight, Palo Alto, Calif
ARNOLDB. STERMAN, MD Executive Director, Aura Lab-
oratories, Hackensack, NJ
JOHN T. SULLIVAN, MB, CHB, FRACP Director, Early
Development, Medical Sciences, Amgen Inc, Thousand Oaks,
Calif
ROBERT TEMPLE, MD Director, Office of Drug Research
and Review, Food and Drug Administration, Rockville, Md
ANDRE TERZIC, MD, PHD Professor of Medicine and Phar-
macology, Division of Cardiovascular Diseases, Department of
Medicine and Molecular Pharmacology and Experimental Ther-
apeutics, Clinical Pharmacology Unit, Mayo Clinic, Mayo Foun-
dation, Rochester, Minn
TIMOTHY S. TRACY, PHD Professor, Experimental and
Clinical Pharmacology, College of Pharmacy, University of Min-
nesota, Minneapolis, Minn
ELLIOT S. VESELL, MD Professor of Pharmacology and
Medicine, the Milton S. Hershey Medical Center, Hershey, Pa
ROBERT E. VESTAL, MD Clinical Pharmacology Consult-
ing, Boise, Idaho
JOHN B. WARREN, MD, FRCP Department of Health,
Medicines Control Agency, London, United Kingdom
WILLIAMB. WHITE, MD Professor of Medicine, University
of Connecticut, Farmington, Conn
JOHN T. WILSON, MD Section on Clinical Pharmacology,
Departments of Pharmacology and Therapeutics, Pediatrics, and
Psychiatry, Louisiana State University Medical Center,
Shreveport, La
ALASTAIRJ. J. WOOD, MB, CHB Professor of Medicine and
Professor of Pharmacology, Vanderbilt University School of
Medicine, Nashville, Tenn
RAYMOND L. WOOSLEY, MD, PHD Vice President for
Health Sciences, University of Arizona, Tucson, Ariz
Page 6A MARCH 2005 CLINICAL PHARMACOLOGY & THERAPEUTICS
INFORMATION FOR AUTHORS
Send manuscripts and all other editorial communications to:
C. Michael Stein, MD, Editor
Clinical Pharmacology & Therapeutics
Division of Clinical Pharmacology
560 RRB
Vanderbilt University School of Medicine
Nashville, TN 37232-6602
Telephone: (615) 936-3420
E-mail: michael.stein@vanderbilt.edu
CONTENT OF THE JOURNAL
Clinical Pharmacology & Therapeutics is a monthly
journal devoted to the publication of articles of high quality
concerned with effects of drugs in human beings and in-
vites original articles in this eld. Hypothesis-driven studies
(as opposed to observational studies) are especially encour-
aged. Several broad categories are of interest to our readers:
Commentaries:
Brief, thoughtful essays on topics relevant to clinical phar-
macology (particularly public policy, drug regulation or
development, and education) may be submitted.
Perspectives in Clinical Pharmacology:
Articles provide state-of-the-art education about a topic of
general interest from a clinical pharmacology perspective.
These articles are solicited by the Editor; potential contrib-
utors should approach the Editor before submitting work.
Standard review articles about drugs, classes of drugs, or
treatment of a disease are normally not accepted.
ORIGINAL ARTICLES
Pharmacokinetics and Drug Disposition:
Papers should have wide generalizability or represent ad-
vances in technical or conceptual methodology and under-
standing of drug disposition. Articles that only describe the
kinetics or bioavailability of a drug in normal subjects or the
population pharmacokinetics of a drug whose pharmacoki-
netics are otherwise well known are not usually published.
Pharmacodynamics and Drug Action:
These papers describe the actions of drugs in humans and
study the mechanism of action of drugs.
Clinical Trials and Therapeutics:
Papers include controlled clinical trials and studies that
examine therapeutic effects of drugs.
Pharmacogenetics and Pharmacogenomics:
Studies of genetic factors inuencing drug disposition, drug
action, and adverse drug reactions are of interest, particu-
larly genotype-phenotype associations.
Pharmacoepidemiology and Drug Utilization Studies:
Studies of drugs that are widely used or are of particular
public health importance, or that seek to improve prescribing
are encouraged.
Clinical Pharmacology Grand Rounds:
These articles include reports of cases in which observations
are made that can be used to illustrate a principle of clinical
pharmacology that is applied in therapy. The discussion
must be limited and focused on the specic teaching point of
the case. Case reports without a clinical pharmacology
teaching point are not considered.
Letters to the Editor:
Letters to the Editor of fewer than 400 words, including
references, that comment on articles previously published in
the Journal will be considered.
Book Reviews:
Occasional reviews of books of particular interest to clinical
pharmacologists are published. Books reviewed may be text-
books, biography, history, or ction and are solicited and
published at the discretion of the Editor.
Clinical Pharmacology & Therapeutics welcomes papers
that do not t into a specic category but do contribute to the
knowledge base of the effects of drugs in human beings. In
vitro studies of human-derived materials that could not be
done with materials derived from other species and papers
that include animal experimental data in addition to human
data are also of interest, provided that the primary focus of
the paper is on drugs in human beings.
SUPPLEMENTS
Supplements to the Journal are occasionally published.
These are usually proceedings of meetings or guidelines
produced by an authoritative body on topics that are of
interest to clinical pharmacologists. Supplements that deal
with a single drug or the products of a single company or
that serve marketing or advertising functions are not pub-
lished. All contents of supplements are peer reviewed and
may be rejected by the Editor. The cost of publishing a
supplement is borne by its organizers.
Copyright. Articles are accepted on the condition that
they are contributed solely to Clinical Pharmacology &
Therapeutics. It is assumed that all of the observations
reported in a manuscript are original, have not been previ-
ously published, and are not under consideration by another
journal, and it is the corresponding authors responsibility to
indicate any data previously published in any form or in
press. A paper from a study that has generated multiple
publications should include references to all papers that have
presented any data that are included in the report and copies
of papers should be included and mentioned in the cover
CLINICAL
PHARMACOLOGY
T
HERAPEUTICS &
letter. Papers that present data that may be included in a
future report should mention such a possibility in a footnote
and the information must be included in the cover letter.
Commentaries and addresses submitted for publication may
contain, in part, material that has been previously published
but appropriate permission must be obtained by the author.
All material published in this Journal is copyrighted and may
be reproduced only with the permission of the publisher.
Statements and opinions expressed in articles and commu-
nications herein are those of the author(s) and not necessar-
ily those of the Editor, publisher, or American Society for
Clinical Pharmacology and Therapeutics, and the Editor,
publisher, and American Society for Clinical Pharmacology
and Therapeutics disclaim any responsibility or liability for
such material. Neither the Editor nor the publisher nor the
American Society for Clinical Pharmacology and Therapeu-
tics guarantees, warrants, or endorses any product or service
advertised in this publication, nor do they guarantee any
claim made by the manufacturer of such product or service.
COPYRIGHT TRANSFER
In accordance with the Copyright Act of 1976, which
became effective on January 1, 1978, copyright owner-
ship is to be transferred to the American Society for Clinical
Pharmacology and Therapeutics. A copyright transfer form
that must be signed by all authors will be sent to the
corresponding author if a manuscript is accepted or if a
revised manuscript is requested. The enclosure of a copy-
right transfer form in a request for a revised manuscript does
not imply that the revised manuscript will be accepted. The
copyright transfer form is also available electronically at
http://www.mosby.com/cpt. No changes should be made to
the wording of this form.
CONFLICT OF INTEREST AND ETHICS
Conict of interest and sources of funding. A potential
conict of interest exists when an author or the authors
institution has nancial or personal relationships that could
inuence or could be perceived to inuence the work. A
conict can be actual or potential, and therefore full disclosure
is best (see Uniform Requirements for Manuscripts Submitted
to Biomedical Journals available at http://www.icmje.org/).
All submissions must include disclosure of all relationships
that could be viewed as presenting a potential conict of
interest, and these will be published with the manuscript.
The Editor and Associate Editors will recuse themselves
from involvement in processing manuscripts where they
identify a conict of interest. A reviewer should not review
a manuscript if there is a conict of interest.
Ethical obligations. The Journal follows the guidelines
describing the ethical obligations of authors, reviewers, and
editorial staff in Uniform Requirements for Manuscripts
Submitted to Biomedical Journals (available at http://www.
icmje.org/). It is the authors responsibility to inform the
Editor if there is more than one manuscript published, in
process or planned, that will share data and to enclose copies
of those potentially competing manuscripts when a manu-
script is submitted. All persons designated as authors should
qualify for authorship, and all those who qualify should be
listed. Each author must have participated sufciently to take
public responsibility for appropriate portions of the content.
Authorship requires that all of the following conditions be
met: (1) substantial contributions to concept and design, or
acquisition of data, or analysis and interpretation of data; (2)
drafts of the article or critical revisions for important intel-
lectual content; and (3) nal approval of the version to be
published.
MANUSCRIPT SUBMISSION
General Instructions
Clinical Pharmacology & Therapeutics complies with
the Uniform Requirements for Manuscripts Submitted to
Biomedical Journals (Ann Intern Med 1997;126:36-47;
http://www.icmje.org/).
Manuscripts must be typed double-spaced with a 12-
point font, including references, gure legends, and tables, on
8 11 inch paper on one side of the page with liberal margins.
Number each page except the title page, including
gure legends, tables, and references. Cite each gure and
table in text in numeric order.
Assemble manuscripts in this order: (1) title page; (2)
abstract; (3) text, including Introduction, Methods, Results,
and Discussion; (4) acknowledgments and conict of inter-
est statement; (5) references; (6) gure legends; (7) tables;
and (8) gures. Staple only the copies of the manuscripts,
not the original. Cite each reference in text in numeric order.
Refer to drugs by accepted generic (public) or chemical
names only. Do not abbreviate these names; spell them in
full each time they are used. Code numbers alone are not
acceptable. USAN is the preferred name; if the USAN and
INN differ, the INN should be given in parentheses the rst
time the drug name is mentioned in the abstract and the rst
time it is mentioned in the text. Proprietary names should be
given in parentheses only at rst text mention. Insert IUPAC
(International Union of Pure and Applied Chemistry) name
if necessary. Specic stereoisomers, if used, or a racemic
mixture should be identied as such.
Hard-copy submission. 1. Enclose a cover letter that
includes the following statement of submission: All authors
have read the manuscript and conict of interest statement
and approved their submission for publication; the work is
original and has not been published and is not being con-
sidered for publication elsewhere, in whole or in part, in any
language, except as an abstract.
Indicate if you have potentially overlapping work that is
in preparation, has been previously submitted or published,
or is in press.
The cover letter can include the names and addresses of
4 potential reviewers, with e-mail addresses or telephone or
fax numbers, who are expert in the area of research and who
do not have a conict of interest. Names of individuals
whom the authors believe have a conict of interest and
should not be asked to review the work may also be listed.
The Editor may or may not decide to solicit reviews from the
persons listed.
2. Send 4 copies of the full manuscript, including one
original copy, and one full set of original gures on glossy-
nish paper. Manuscripts will not be returned. A stamped,
self-addressed large envelope should be sent if authors want
the gures from a rejected manuscript to be returned.
3. If applicable, enclose a copy of any potentially over-
lapping work that is in preparation, has been previously
submitted or published, or is in press.
Electronic submission. A cover letter and manuscript in
Word or WordPerfect can be e-mailed to the Editor at:
michael.stein@vanderbilt.edu. All electronic gures must be
accompanied by a high-quality hard copy on glossy paper.
When the article is accepted, electronic gures will have to
meet the publishers requirements (see Author Gateway at
http://authors.elsevier.com).
Revised manuscripts. Revised manuscripts should fol-
low the General Instructions, but the title page should in-
clude the manuscript number and R1 (for rst revision). One
copy of the revised manuscript should be marked with
word-processing tools such as track changes, highlight,
or underline to indicate precisely the changes that have
been made. Enclose a diskette in IBM PCcompatible for-
mat containing the nal version (Word or WordPerfect pre-
ferred). Label and date the diskette. The author accepts
responsibility for the submitted diskette exactly matching
the printout of the nal version of the manuscript. Revisions
not received within 3 months will be administratively with-
drawn. At the Editors discretion, extensions may be granted
for revisions. A copyright transfer agreement will be en-
closed in the request for revision. It must be signed by all
authors and accompany the revised manuscript.
Specic Instructions
1. Title page: Make the title specic and short (there is a
limit of 80 characters). Include the rst name and the highest
earned academic degree for each author. Include the name
and address of the institution from which the work origi-
nated and sources of all nancial support for the work. List
the name, e-mail address, complete mailing address, and
telephone and fax numbers of the corresponding author. List
the number of pages, number of gures, and number of
tables and suggest a short title for your paper (no more than
50 characters including spaces) and the Journal category
under which it should appear.
2. Abstract: Supply an abstract of no more than 150
words, which will appear underneath the title. A structured
abstract (see Ann Intern Med 1990;113:69-76 for details) of
up to 250 words may be used instead of the regular150-word
abstract. Do not use acronyms or abbreviations or cite ref-
erences. The abstract should include the rationale for the
study, a brief description of methods and presentation of
signicant results, and a succinct interpretation of the data.
3. Text (Introduction, Methods, Results, and Discussion):
Use only standard abbreviations for measurements and omit
periods; explain in tabular form all other abbreviations and
acronyms. Limit these other abbreviations and acronyms to
those in common use; do not invent abbreviations but rather
spell out the terms fully each time they are used.
The Journal endorses the principles embodied in the
Declaration of Helsinki and assumes that authors of papers
submitted to the Journal will have followed all ethical and
legal standards for performing experiments on human beings
or animals as set forth by their institutions and states. Manu-
scripts that describe studies on humans must indicate in the
Methods that the study was approved by an institutional review
committee and that the subjects gave informed consent.
4. Acknowledgments and conict of interest statement:
This should be placed at the end of the text and before the
reference list. The acknowledgments should include only the
names of those who have made a scientic contribution.
All authors must disclose any nancial and personal
relationships that could potentially be perceived as inuenc-
ing their described research or state that there are none.
Examples of nancial conicts include employment, consul-
tancies, stock ownership, honoraria, paid expert testimony,
patent applications, and research and travel grants within 3
years of beginning the work submitted. If there are no
conicts of interest, authors should state that there are none.
For example: Dr A has stock in company Y and has
received research funding from company X; Dr B is an
employee of company Y and holds stock options; Dr C has
no conict of interest, and Dr D has received honoraria from
companies X, Y and Z and consulted for X. Conict of
interest statements will be published at the end of the paper.
If individuals other than the authors contributed signi-
cantly to the writing or editing of the manuscript, their
names, roles, and conict of interest statements should be
provided, together with a signed statement from each indi-
vidual indicating that this information can be published.
5. References: Place the reference list immediately after
the acknowledgments. Type the list double-spaced according
to the Uniform Requirements for Manuscripts Submitted to
Biomedical Journals (Ann Intern Med 1997;126:36-47).
Journal names are to be abbreviated as they are in the
Cumulated Index Medicus. In press references may be
used only if the journal that has accepted the manuscript is
indicated. Personal communications and other unpublished
and nonarchival references should not be included in the
reference list. Instead, make these footnotes to the text; state
the source, and date the item.
Examples of references (if there are 6 or fewer authors,
list all of them; if there are 7 or more, list the rst 6 authors
and add et al):
Periodicals:
You CH, Lee KY, Chey RY, Menguy R. Electrogastro-
graphic study of patients with unexplained nausea, bloating
and vomiting. Gastroenterology 1980;79:311-4.
Books:
Eisen HN. Immunology: an introduction to molecular and
cellular principles of the immune response. 5th ed. New
York: Harper & Row; 1974. p. 406.
Chapters in books:
Weinstein L, Swartz MN. Pathogenic properties of invading
microorganisms. In: Sodeman WA Jr, Sodeman WA, edi-
tors. Pathologic physiology: mechanism of disease. Phila-
delphia: WB Saunders; 1974. p. 457-72.
6. Appendixes: Extensive or very detailed descriptions of
analytic methods of mathematic calculations should be in-
cluded as appendixes, following the references. A short
legend may be necessary.
7. Figure legends, tables, and gures: Type gure legends
double-spaced on a separate page marked Legends for
gures. If a gure has been previously published, the leg-
end must give full credit to the original source and the author
must obtain permission for reproduction.
Number tables in Roman numerals in order of mention in
the text; all tables must be mentioned. Type tables double-
spaced, including title, column heads, body, and footnotes. The
title must concisely describe the content of the table, and
column heads must do the same for the data in the columns. If
a table or any data therein have been previously published, a
footnote to the table must give full credit to the original source
and the author must obtain permission for reproduction.
Each copy of the manuscript should include a photocopy
of each of the gures. Illustrations will be scanned electron-
ically. Unless printed by a high-resolution printer (1200 dpi
or better), laser-printed illustrations should be furnished in a
larger size, preferably 6 inches or more wide. Photocopies
are not acceptable as original illustrations. Preferred fonts
for value labels and legends are Arial, Courier, Helvetica,
Symbol, or Times Roman. Bold or italic typefaces are not
acceptable unless used to emphasize particular words or
phrases. The ordinate should be labeled vertically. Labels
directly on the graph rather than in the legend are encour-
aged. Type size must be 14 point or more and line width 1
point or more for 6-inch graphics. The ll for bar graphs or
pie charts should be distinctive: avoid shading because dif-
ferent shades of gray (including black) tend to become
indistinguishable. If more than 2 bars are used in a graph,
please use cross-hatching for adequate contrast. Fills with
unnecessary patterns will be returned for revision. Three-
dimensional graphics may be used only when there are 3
coordinates (x, y, and z). Black and white graphs or draw-
ings must be legible and clearly printed in jet-black ink on
heavy paper; typewritten or freehand lettering is unacceptable.
Any special instructions regarding sizing or grouping should be
noted clearly. All lettering must be done professionally and
should be sized in proportion to the drawing, graph, or photo-
graph. Make sure that decimal points, dotted lines, and the like
are dark enough to reproduce plainly. Do not send original
artwork, x-ray lms, or electrocardiogram (ECG) strips.
In addition, submit ONE ORIGINAL of each photograph
or laser- or plotter-printed gure. If computer-generated
originals of gures are not available, one set of profession-
ally prepared glossy or matte nish prints must be submitted.
The photocopies are sent to the reviewers, and the set of
original illustrations is used for reproduction in the Journal.
Figures must be cited in the text and must be numbered in
order of mention. The rst authors last name, the gure
number, and an arrow to indicate the top edge should be
marked lightly on the back of each gure. A reasonable
number of black and white illustrations will be reproduced
free of charge to the author, but special arrangements must
be made for extra illustrations, color plates, or elaborate
tables. Illustrations will not be returned.
8. Permissions: Direct quotations, tables, or illustrations
that have appeared in copyrighted material must be accom-
panied by written permission for their use from the copyright
owner and original author along with complete information
as to source. Photographs of identiable persons must be
accompanied by signed releases that show informed consent.
9. Page charges: An article charge of $35 per page for the
rst 8 printed pages is made. For pages numbered 9 through
16 the charge per page is $50, and for additional pages the
charge per page is $100. Authors who do not have funds
available may request approval for waivers from the Editor.
Editorial consideration is in no way related to acceptance of
page charges. Waivers should not be sought until after the
author receives a bill.
REVIEW PROCESS
An acknowledgment that your manuscript has been re-
ceived will be sent to the corresponding author by e-mail.
New manuscripts are reviewed by the Editor and an Asso-
ciate Editor and if in their opinion the work is not suitable
for the Journal or is unlikely to have high enough priority for
publication, it will be rejected without review and the cor-
responding author will be notied. Manuscripts that undergo
full review are sent to at least 2 reviewers and are usually
processed within 8 weeks. The manuscript will be rejected or
the authors will have the opportunity to submit a revision. A
revised manuscript must be received within 3 months or it
will be administratively withdrawn. Manuscripts and illus-
trations will not be returned.
ANNOUNCEMENTS
The Journal will publish ofcial notices of afliated
societies only.
BOOKS RECEIVED
Receipt of books and monographs, domestic and inter-
national, will be acknowledged.
REPRINTS
For queries about author reprints, e-mail author.support@
elsevier.com. To order 100 or more reprints for educational,
commercial, or promotional use, contact the Commercial
Reprints Department, Elsevier Inc., 360 Park Ave South,
New York, NY 10010-1710. Fax 212-462-1935; e-mail
reprints@elsevier.com. Reprints of single articles available
online may be obtained by purchasing Pay-Per-View access
for $30 per article on the Journal Web site, http://www.mosby.
com/cpt.
CHECKLIST FOR AUTHORS
Cover letter
Original and 3 photocopies of typed manuscript,
with photocopies of illustrations attached to each
Title page (double-spaced)
Title of manuscript
Full name(s), highest earned academic degrees,
and afliations of author(s); institution(s) and
city(ies) from which work originated
Name, address, and business and home
telephone and fax numbers of corresponding
author
Sources of all nancial support
Number of pages, number of gures, and
number of tables; brief title
Abstract, 150 words or less (double-spaced)
Article proper (double-spaced)
Acknowledgment and conict of interest statement
for each author (double-spaced)
References (double-spaced), on a separate sheet
Figure legends (double-spaced), on a separate sheet
Tables (double-spaced), on separate sheets
Figures/Illustrations, properly labeled (one original
set and attach one set of photocopies to each of the
4 copies of manuscript)
Permissions to reproduce published material (if
applicable)
Informed consent for patient photographs (if
applicable)
With revision
Diskette with revised manuscript
Marked copy indicating changes with revised
manuscript
Copyright transfer agreement signed by all
authors with revised manuscript
CLINICAL
PHARMACOLOGY
&
T
HERAPEUTICS
VOLUME 77 NUMBER 3 MARCH 2005
ASCPT POSITION STATEMENT
American Society for Clinical Pharmacology
and Therapeutics Position Statement on
Dietary Supplement Safety and Regulation
Jason D. Morrow, MD, Chair, Timi I. Edeki, MD, PhD,
Mohamed El Mouelhi, MD, PhD, Raymond E. Galinsky, PharmD,
Rose Kovelesky, RPh, PhD, Robert J. Noveck, MD, PhD, and
Charles Preuss, RPh, PhD, American Society for Clinical Pharmacology and
Therapeutics Task Force on Dietary Supplements Nashville, Tenn, Waukegan, Ill, Mason, Ohio,
Indianapolis, Ind, Emeryville, Calif, New Orleans, La, and West Palm Beach, Fla
SUMMARY AND RECOMMENDATIONS
Dietary supplements are regularly used by a large
number of Americans with the belief that they improve
human health and well-being. Most consumers also
believe that dietary supplements are subject to signi-
cant governmental regulation and have undergone ex-
tensive review by the Food and Drug Administration
(FDA) to ensure their safety and efcacy. Dietary sup-
plements are regulated under the Dietary Supplement
Health and Education Act (DSHEA) of 1994. This act
assumes that supplements, unlike drugs and food addi-
tives, are inherently safe, so there is a relative lack of
regulation related to compound safety. DSHEA se-
verely limits the control that the FDA can exert over the
dietary supplement industry to ensure product safety.
This is a cause for concern because many dietary sup-
plements contain pharmacologically active ingredients
that are capable of causing serious adverse events
(AEs). Several government organizations have recently
sought to modify DSHEA to enhance oversight of
dietary supplement safety.
In response to concerns regarding the relative lack of
dietary supplement regulation, the American Society
for Clinical Pharmacology and Therapeutics (ASCPT)
established the ASCPT Task Force on Dietary Supple-
ments, under the purview of its Government Affairs
From Vanderbilt University, Nashville; Abbott Clinical Pharmacol-
ogy, Waukegan; Procter and Gamble Pharmaceuticals, Mason;
Purdue University, Indianapolis; Chiron Corporation, Emeryville;
MDS Pharma Services, New Orleans; and Palm Beach Atlantic
University, West Palm Beach.
The opinions, ndings, conclusions, or recommendations put forth in
this position paper are those of the authors and do not necessarily
reect the views of the organizations with which they are em-
ployed or engaged.
Reprint requests: Michele Boisse, American Society for Clinical
Pharmacology and Therapeutics, 528 N. Washington St, Alexan-
dria, VA 22314.
E-mail: michele@ascpt.org
Clin Pharmacol Ther 2005;77:113-22.
0009-9236/$30.00
Copyright 2005 by the American Society for Clinical Pharmacology
and Therapeutics.
doi:10.1016/j.clpt.2005.01.017
113
Committee, to examine this issue and provide recom-
mendations for changes in DSHEA that would increase
the safety of these agents.
The recommendations put forth by the Task Force
and detailed in this position statement were subse-
quently peer-reviewed and adopted by the ASCPT
Board of Directors on January 20, 2005.
It is the opinion of ASCPT that enhanced oversight
of dietary supplements is essential to increase the safety
of these products for the American consumer. ASCPT
proposes 5 recommendations that will provide mecha-
nisms by which oversight and safety can be enhanced.
These are as follows:
1. Provide legislation and funding to enable the FDA
to establish procedures to improve detection of AEs
associated with dietary supplements, obtain better
data as they relate to AE reporting (AER), improve
the assessment of AER data, and improve the dis-
semination of reports about AEs associated with
dietary supplements to the consumer.
2. Finalize and implement the FDAs proposed cur-
rent good manufacturing practices (cGMP) for di-
etary supplements because they promote and pro-
tect public health.
3. Implement the recommendations of the Department
of Health and Human Services Inspector General
regarding the template for labeling of dietary sup-
plements and provide additional contact informa-
tion on dietary supplement labels for consumers.
4. Provide enhanced educational opportunities on di-
etary supplements to health care professionals and
consumers.
5. Enhance research opportunities on the potential ad-
verse effects, as well as possible efcacy, of dietary
supplements.
Implementation of these recommendations will serve
to improve the safety of dietary supplements for the
American public and will allow consumers to become
better educated about the potential toxicities and ben-
ets associated with these agents.
INTRODUCTION
A dietary supplement is dened as a product (other
than tobacco) that is intended for use by humans to
supplement the diet by increasing total dietary intake
and that bears or contains 1 or more of the following
ingredients: a vitamin, a mineral, an herb or other
botanical, an amino acid, or a dietary substance.
1
Di-
etary supplements are widely used in the United States
today. It is estimated that at least 35% of adults regu-
larly use alternative and complementary medical ap-
proaches, including dietary supplements, and sales of
supplements in 2002 were in excess of $18 billion.
2-4
Supplements also are widely used by children and
adolescents, and the decision to use supplements in
these populations is often driven by parents. Use of
supplements has increased since the 1970s, suggesting
that Americans are becoming more receptive to alter-
native agents for treatment and prevention of disease.
2
Since 1994, there has been an explosion in the number
of dietary supplement products from 4000 to 29,000,
with over 1000 new products introduced annually. Re-
cent surveys suggest that consumers generally believe
that dietary supplements are subject to existing govern-
ment regulations, similar to those for over-the-counter
(OTC) medications sold without a prescription, and
have undergone extensive review by the Food and Drug
Administration (FDA).
2,5
Dietary supplements are regulated under the Dietary
Supplement Health and Education Act (DSHEA) of
1994.
1
The legislation was designed to protect the right
of access of the consumer to dietary supplements so as
to promote wellness. In the Act the assumption is that
dietary supplements are safe and, as such, dietary in-
gredients in dietary supplements are exempt from reg-
ulation as drugs or food additives. As a consequence,
premarketing approval (as for drugs) of a product is not
required nor does it need to obtain a generally recog-
nized as safe status (required of food additives). A
dietary supplement is considered unsafe if it presents a
signicant or unreasonable risk of illness or injury
under conditions of use recommended or suggested in
labeling. More importantly, however, the FDA bears
the burden of proof if it decides to assert that a supple-
ment is unsafe.
1
Furthermore, for new dietary ingredi-
ents introduced after 1994, manufacturers are required
to notify the FDA 75 days before marketing a dietary
supplement, and the agency may review safety claims,
although FDA approval is not required for marketing. It
also is a general assumption of DSHEA that dietary
supplements are efcacious as claimed. These regula-
tions severely limit the control that the FDA can exert
over the dietary supplement industry to ensure product
safety and efcacy.
Although many dietary supplements are likely safe,
there is little denitive scientic data to prove safety for
most agents.
6,7
Many supplements contain active ingre-
dients with the potential for serious adverse pharmaco-
logic effects if administered in excessive, or in some
cases recommended, dosages. A recent example of this
is Ephedra-containing products that were marketed
widely for weight loss and enhancement of physical
performance. Ingestion of these products can result in
CLINICAL PHARMACOLOGY & THERAPEUTICS
114 Morrow et al MARCH 2005
systemic concentrations of ephedrine (one of the active
herbs in Ephedra) that have led to a number of serious
adverse cardiovascular effects including hypertension,
stroke, seizures, and death.
7-9
Ephedra-containing prod-
ucts were eventually banned from the market but only
after a prolonged effort by the FDA in which the
agency had to prove the products unsafe. Other agents
that have been regarded as safe but which clearly have
potentially harmful effects include L-tryptophan, which
contained a contaminant that has caused a rare blood
disorder (the eosinophilia-myalgia syndrome) and
which was removed from the market.
10,11
More re-
cently, hepatotoxicitya state of toxic damage to the
liverassociated with kava (from the pepper Piper
methysticum) resulted in warnings to consumers about
the agent in several European markets.
12,13
Because of concern over the relative lack of regula-
tion of dietary supplements in comparison to drugs and
food additives, the FDA, various health care organiza-
tions, and the US Congress have considered changes to
DSHEA to bolster the general oversight of dietary
supplements and to address concerns related to safety.
Debate during the 108th Congress has highlighted a
need for enhanced regulation.
14
Furthermore, the FDA
has recently issued a strategy for enhanced regulation
of dietary supplements under DSHEA and has provided
guidance for the supplement industry regarding claims
made about supplements.
15,16
In addition, the FDA has
turned to the Institute of Medicine (IOM) and the
National Research Council (NRC) of The National
Academies to propose a framework for evaluating the
safety of dietary supplement ingredients marketed in
the United States. A committee of expertsthe Com-
mittee on the Framework for Evaluating the Safety of
Dietary Supplements (IOM/NRC Framework Commit-
tee)evaluating this issue presented its ndings in
2004, and they are available online.
2
The major con-
clusions of the IOM/NRC Framework Committee were
that current law may not adequately secure the Amer-
ican public from risks associated with dietary supple-
ments and that constraints imposed by DSHEA on the
FDA make it difcult for the American public to be
adequately protected. A number of recommendations
were made by the IOM/NRC Framework Committee
including the following: (1) Continue to maintain a
prospective monitoring system for AEs associated with
dietary supplement ingredients within the FDA. (2)
Congress should provide adequate resources to allow
the FDA to protect consumers under DSHEA. (3)
Amend DSHEA to require manufacturers and distribu-
tors of dietary supplements to report AEs to the FDA.
(4) The FDA should establish current good manufac-
turing practices (cGMP) for dietary supplements. (5)
Change dietary supplement labeling to list both the
manufacturer and distributor. (6) Encourage collabora-
tive efforts between the FDA and the National Institutes
of Health (NIH) as additional research becomes neces-
sary to address unresolved issues regarding the poten-
tial harm that certain ingredients may cause.
2
It is the position of ASCPT to support enhanced
regulation of dietary supplements to increase the safety
of these products. Following are the specic recom-
mendations of ASCPT that are designed to provide
mechanisms by which oversight of supplements can be
improved.
RECOMMENDATION 1
Recommendation 1 is to provide legislation and
funding to enable the FDA to establish procedures to
improve detection of AEs associated with dietary sup-
plements, obtain better data as they relate to AE report-
ing (AER), improve the assessment of AER data, and
improve the dissemination of reports of AEs associated
with dietary supplements to the consumer.
AEs caused by dietary supplements can be divided
into 3 broad categories: toxicity, interactions, and adul-
teration.
17
Toxicity can occur with dietary supplements
because of side effects that are inherent in the pharma-
cology of the ingredients contained in the dietary sup-
plement or are an idiosyncratic reaction. Interactions
occur when a dietary supplement is taken with a drug
(ie, prescription or OTC) or with another dietary sup-
plement. Adulteration of dietary supplements occurs
when they are contaminated with substances such as
heavy metals, pathogenic microbes, or allergens.
Unfortunately, the FDAs current AER system de-
tects only a small percentage of actual AEs associated
with dietary supplements.
17
The American Association
of Poison Control Centers, a network of hospitals and
academic health centers that respond to consumer calls
about problems with a variety of consumed products,
receives more AER than the FDA.
17
In 1999 the Amer-
ican Association of Poison Control Centers docu-
mented 13,000 reports attributed to dietary supplements
whereas the FDA documented 460.
17,18
Presumed safety, self-care, and perceived lack of
FDA involvement are thought to contribute to under-
reporting of dietary supplement AEs to the FDA.
17
Consumers believe that dietary supplements are derived
from natural ingredients and therefore must be safe,
unlike prescription drugs, which are most often syn-
thetically manufactured. Hence, consumers will substi-
tute self-care products including dietary supplements
for prescription or OTC drugs. These assumptions,
2005;77(3):113-22 Position paper on dietary supplements 115
combined with the perception of limited FDA involve-
ment regarding dietary supplements, lead to under-
reporting of AEs.
The FDAs AER system has difculty interpreting
data regarding potential problems with dietary supple-
ments because its reporting system receives limited
information from databases about dietary supplements.
The key information comes from medical profession-
als, product information, manufacturer information,
consumer contact information, and trend analysis.
17
In
this light, the FDA was unable to obtain the medical
records for 58% of the reports it was pursuing and it
was unable to determine the ingredients for 32% of the
dietary supplements that it was investigating.
17
The
FDA receives very few AE reports from the manufac-
turer. The FDA could not determine the identity of the
manufacturer for 32% of the dietary supplements in-
volved in the reports and was unable to follow up with
27% of the reports.
17
Therefore the FDA was unable to
analyze trends with statistical rigor because the data
were poor in quality and quantity.
17
To help resolve these gaps in obtaining AE data,
ASCPT recommends that legislation and funding be
provided to allow the FDA to quickly implement the
recommendations of the Department of Health and
Human Services (DHHS) Ofce of Inspector Generals
2001 report Adverse Event Reporting for Dietary
Supplements: An Inadequate Safety Valve.
17
What
ASCPT views as the most relevant issues related to
these important recommendations are as follows: (1)
improving detection of AEs, (2) obtaining better data
on AER, (3) improving the assessment of the AER
data, and (4) improving the reporting of AEs asso-
ciated with dietary supplements to the consumer.
Furthermore, ASCPT recommends that this AER
monitoring process, detailed below, be prospective
and ongoing so as to accumulate information on AEs
in real time. Importantly, the FDA should receive
adequate funds from Congress to carry out these
recommendations.
To improve the detection of AEs caused by dietary
supplements, the FDA should (1) require dietary sup-
plement manufacturers to report AEs, (2) obtain AER
from poison control centers, and (3) inform physicians,
pharmacists, and other health care providers about
AER.
In this regard ASCPT emphasizes that it should be a
primary responsibility of dietary supplement manufac-
turers and distributors to report AEs. In addition, better
utilization of MedWatch
19
the FDA safety informa-
tion and AER programby health care providers and
consumers would help improve AE information and
communication efforts.
To obtain better data on AER, the FDA should (1)
require better information on AER from the dietary
supplement industry, (2) require manufacturers to reg-
ister themselves and their products with the FDA, (3)
notify the manufacturer when an AE report is received,
and (4) enable consumers to submit an AE report in a
secure manner.
The FDA should work with the United States Phar-
macopeia (USP) to incorporate in its Dietary Supple-
ment Verication Program, or its other reporting pro-
grams, a system for reporting AEs with dietary
supplements by hospitals, poison control centers, and
emergency departments.
20
To improve the assessment of AER data, the FDA
should (1) require that safety and AE information be
provided when manufacturers submit the 75-day pre-
market notication, (2) develop a monograph system
for dietary supplements, (3) work more closely with the
NIH in setting rigorous research agendas for dietary
supplements to explore issues such as safety and ef-
cacy and work more closely with the USP in the veri-
cation of dietary supplements, and (4) develop better
software to analyze AER data.
17
Regulation of dietary supplement quality should be
similar to that for OTC drugs. The FDA should require
a supplement to comply with USP and new formulary
standards where a monograph exists unless the dietary
supplement specically states on its label that it is not
USP or new formulary quality.
To improve the reporting of dietary supplement AEs
to consumers, (1) the FDA should provide current and
relevant AER on its Web site that would allow the
consumer to be educated and make a better choice for
a safe dietary supplement, (2) physicians and pharma-
cists should document use of dietary supplements and
educate patients about dietary supplements, and (3) the
dietary supplement industry should inform the public of
the potential for dietary supplementdietary supple-
ment interactions or dietary supplementdrug interac-
tions, as well as dietary supplementfood product in-
teractions, through appropriate product labeling (see
Recommendation 3).
RECOMMENDATION 2
Recommendation 2 is to nalize and implement the
FDAs proposed cGMP for dietary supplements be-
cause they promote and protect public health.
Dietary supplements are currently subject to lower
safety standards than food additives or drugs.
21
Current
good manufacturing practices for dietary supplements
can be simply stated as regulations that ensure that
every dietary supplement on the market has the iden-
tity, purity, quality, strength, and composition that are
stated on the label. The key components of the pro-
posed FDA dietary supplements cGMP relate to per-
sonnel, physical plant, equipment, process controls,
holding, records, and consumer complaints.
22
Many of
these are similar to food cGMP. ASCPT supports the
institution of cGMP for the manufacturing of dietary
supplements through the belief that this will improve
the safety of these agents. The following highlights
selected aspects of the proposed cGMP that are of
particular importance to ASCPT.
There are 4 main provisions of the FDA report,
dealing with personnel, hygiene, qualications, and su-
pervision.
22
Hygienic practices of personnel should be
used to prevent the contamination of the dietary sup-
plements with pathogenic microbes, dirt, hair, and so
on. Only qualied personnel should be allowed to man-
ufacture, package, or hold dietary supplements to en-
sure production of an unadulterated product. The ap-
propriate training, education, and experience of
personnel are essential to competently manufacture,
package, and hold dietary supplements. Qualied su-
pervisory personnel should be responsible for all cGMP
requirements.
The main issue related to the physical plant is that it
be designed to sustain a contaminant-free environment
for dietary supplements during their manufacturing,
packaging, and holding.
22
The physical plant should be
required to be kept in a clean and sanitary condition to
prevent contamination from rodents, insects, patho-
genic microbes, dust, and so on. There should be writ-
ten procedures for sanitation, cleaning schedules, and
methods.
Equipment used in the manufacturing, packaging,
and holding of dietary supplements should have the
appropriate design, construction, and workmanship to
enable its suitable function and maintenance.
22
There
should be written procedures for adequate cleaning of
equipment. Furthermore, equipment should be
corrosion-resistant, made of nontoxic materials, and
designed to prevent contamination.
Process controls should be implemented in the man-
ufacturing, packaging, and holding of dietary supple-
ments.
22
Quality-control procedures should ensure that
the dietary supplements do not contain impurities. The
holding of dietary supplements should be done in a
manner that prevents contamination and deterioration.
They should be held under appropriate conditions of
temperature, humidity, and light to ensure that the
identity, purity, quality, strength, and composition are
not affected.
22
Finally, record-keeping is an extremely
important component of cGMP to ensure that a dietary
supplement company demonstrates compliance with
cGMP as evidenced by FDA inspections of the entire
manufacturing and holding process.
1
RECOMMENDATION 3
Recommendation 3 is to implement the recommen-
dations of the DHHS Inspector General regarding the
template for labeling of dietary supplements
23
and pro-
vide additional contact information on dietary supple-
ment labels for consumers.
The FDA is the federal agency primarily responsible
for regulating not only dietary supplements but also
their labels. The FDA oversees supplement labeling in
a postmarket system in that once a label has been
released to the public, the FDA has the burden of proof
to show that the label information is misleading or not
true before the product can be removed from the mar-
ketplace.
2
An important source of information for both
the professional and consumer is the product label.
Functions of the supplement label include facilitation of
product choice, description-of-use instructions, and
communication of safety information. Dietary supple-
ments are eligible for a variety of label statements and
claims, each of which has unique regulatory require-
ments. Two points discussed herein that form the basis
for recommendations of ASCPT are as follows: (1)
problems with current labeling practices for dietary
supplements under DSHEA and (2) previous recom-
mendations for dietary supplement labels from the
Commission on Dietary Supplement Labels, the FDA,
and the DHHS Inspector General.
A number of problems have been cited regarding
current labeling requirements for dietary supplements.
A central issue is that labels often fail to provide
sufcient information to guide the informed and appro-
priate use of supplements.
23
Second, ingredient infor-
mation is often difcult to interpret in that supplement
labels often fail to provide enough detail about supple-
ment composition to enable users to understand exactly
what they are taking. Third, statements of intended use
frequently provide limited information. Fourth, safety
information is often incomplete.
Because of these issues, there have been several
governmental reports related to dietary supplement la-
beling. In 1997 the Commission on Dietary Supplement
Labels recommended that the NIH Ofce of Dietary
Supplements serve as the principal advisor to the Sec-
retary of Health and provide advice to the directors of
the NIH and the Centers for Disease Control and Pre-
vention, as well as the commissioner of the FDA, on
issues related to labeling of dietary products.
24
The
commission had the view that this effort would allow
both the dietary supplement industry and consumers to
benet from an increased level of scientic input into
decisions regarding label statements for dietary supple-
ments. In addition, it was recommended that the dietary
supplement industry consider establishing an expert
advisory committee on dietary supplements to provide
scientic review of label statements and claims. Such a
committee could have been supported by 1 or more
industry trade associations or established as an inde-
pendent entity funded by extramural grants or fees for
services. These recommendations were not imple-
mented. That same year, the FDA proposed regulations
that require a specic labeling format for nutritional
supplements featuring a prominent box with the head-
ing Supplement Facts.
25
An accurate listing of sup-
plement ingredients was to be required on product
labels.
Because of ongoing labeling issues, in March 2003
the DHHS Inspector General proposed a template of
key elements for a dietary supplement label.
26
The label
content should include information related to ingredi-
ents; intended use; safety, including warnings about
potential interactions with drugs and other dietary sup-
plements; directions for use; and manufacturing infor-
mation. This template for label content is in line with
ndings of the 2004 Consumers Union Dietary Supple-
ments Survey
27
and is recommended by ASCPT.
In addition, ASCPT endorses the provision of en-
hanced contact information on the label for the con-
sumer regarding the manufacturer and distributor. Cur-
rent regulations related to the source of a product only
require the name and place of business of the manu-
facturer, packer, or distributor to be on the label. In
agreement with the IOM/NRC Framework Commit-
tee,
2
we urge the adoption of labeling changes such that
the label contains the name and place of business of
both the distributor and the manufacturer of the product
to facilitate tracing the source of the product if AEs or
other problems occur. Consideration also should be
given to providing the consumer with product contact
information. In particular, a telephone number for ques-
tions should be supplied by the manufacturer or distrib-
utor. Furthermore, ASCPT believes that dietary supple-
ments should be labeled with a toll-free FDA hotline
number for AER.
RECOMMENDATION 4
Recommendation 4 is to provide enhanced educa-
tional opportunities on dietary supplements to health
care professionals and consumers.
Health care professionals. A majority of health pro-
fessions schools (medical, pharmacy, and nursing)
teach students about selected aspects of dietary supple-
ments during their course work.
28-31
Because there are
a large number of patients who use dietary supple-
ments, some health care professionals feel compelled to
be educated on the subject even though they may not
agree with the usage of these products.
32,33
Often, their
concern about dietary supplements is generally a result
of the perception that there is a lack of high-quality,
peer-reviewed, critically evaluated and comprehensive
published literature and references.
29
The availability
of reliable, scientically sound analyses of these ther-
apies is, therefore, essential.
Many students in health professional schools today
are interested in dietary supplements, and this is often
the reason that natural products are included in the
curriculum. Programs that include dietary supplements
do so through required or elective courses, as part of a
required class such as pharmacology, therapeutics, and
OTC agents, or through required attendance at semi-
nars.
29,30,32,33
With the exception of pharmacy schools,
dietary supplement education is usually contained in
courses that cover different types of complementary
and alternative medicine including such therapies as
acupuncture and chiropractics. Courses at various
schools range from 2 to 3 credit hours and are offered
throughout the professional program.
29
Internships/
clinical rotations related to dietary supplements and
natural products are available at some schools including
those that do not teach natural products in the didactic
portion of their program.
29,33
The number of graduate
medical education programs with complementary and
alternative medicine curriculums available to medical
residents has remained the same over the past 4 years
(24%).
34
There is no clear consensus on how to incorporate
the topic of dietary supplements into the educational
process.
29,30 32,33
In an ideal education program, stu-
dents should focus on critical thinking regarding dietary
supplements that can serve as a springboard for teach-
ing students a skill that is crucial to their professions
how to critically evaluate published literature.
30
Also
included should be basic issues regarding quality and
safety, regulatory topics, and where to nd reliable
information.
29
For professionals already in practice, a
large amount of credible continuing education informa-
tion is available through professional organizations,
journals, seminars, and conferences, as well as online.
The FDA is beginning to put into place systems such
as one linking medical records through health care
providers and the Centers for Disease Control and
Prevention to identify potential AEs related to dietary
supplements more rapidly, using cutting-edge statistical
methods to assess the frequency and severity of AER.
Furthermore, it is partnering with the National Library
of Medicine, medication manufacturers and distribu-
tors, and health care information suppliers to set up
DailyMed, a database containing comprehensive med-
ication information that will be updated continually.
35
It is the view of ASCPT that education of health care
professionals on topics related to dietary supplements
can provide an important avenue to improve the safety
of consumers who use these products. Thus we favor
consideration of the following educational initiatives to
enhance training of health care professionals in dietary
supplements: (1) increasing the availability of reliable,
scientically sound analyses of these therapies; (2)
requiring courses that include dietary supplements as
described above in health care professions programs;
and (3) continuing to maintain reliable Web sites that
are updated regularly and that contain information on
supplements, such as the prospective DailyMed system
and the International Bibliographic Information on Di-
etary Supplements database.
36
Consumers. The goal of consumer education is to
educate the public so that informed choices about di-
etary supplements based on factual, scientically
driven and unbiased data can be made, preferably in
conjunction with health care professionals. A danger-
ous and mistaken belief held by much of the public is
that if a product is natural, there will be no risk of harm
and the product will have the advertised effect. In
addition, consumers generally are not aware that over-
doses and drug interactions can occur with dietary
supplements.
37
This can lead some patients not to dis-
close to their health care professional that they are
using dietary supplements.
37
Other alarming assump-
tions held by the public include that the government
must approve dietary supplements before they are mar-
keted (59%), that warning labels containing possible
side effects and toxicities are required (68%), and that
there is scientic support for any safety/efcacy claims
made (55%).
38
There also is an erroneous belief by
consumers that dietary supplements are held to the
same standard as OTC drugs in terms of manufacturing
and are free of any contaminants.
38
Clearly, there is a
dire need for consumer education.
Public opinion supports the adequate regulation of
dietary supplements generally to the same standard as
OTC medications. Consumers emphatically want sup-
plements to be proved safe and effective before their
introduction into the marketplace (91%) and manufac-
turers to be required to report any AEs experienced and
include this information on the label (96%).
39
A small
minority of adults (19%) believe that the dietary sup-
plement industry is adequately regulated.
39
Fundamental to educating the public about dietary
supplements is to use clear and concise language that is
easily understood. Basic, unbiased facts in laymans
terms about nutritional supplements should be made
readily available to the public. The sources of these
data would ideally come from health care professionals
and the federal government. Health care professionals
using their frequent interactions with patients, as well
as the publics perception that these professionals are
trustworthy, would provide an ideal place to start and
would require little capital investment. In particular,
health care professionals should be encouraged to be
proactive in initiating discussions with patients about
the topic of dietary supplements.
The FDAs MedWatch, a voluntary program through
which health professionals and consumers can report
serious adverse reactions and problems related to di-
etary supplements, as well as drugs, biologics, medical
devices, cosmetics, and infant formulas, should be pro-
moted throughout the medical community. MedWatch
also should be publicized among patients and consum-
ers, who should be encouraged to contact the FDA
directly or through their physicians to report AEs. In
addition, the federal government can be further in-
volved by (1) strongly promoting the National Center for
Complementary and Alternative Medicine (NCCAM) and
the Ofce of Dietary Supplements within the NIH,
along with the FDA, as reliable resources for informa-
tion about dietary supplements and (2) ensuring that
information about AEs for natural products is easily
accessible on the FDA Web site
The American Medical Association notes that the
law prohibiting promotion by dietary supplement man-
ufacturers and vendors that a cure or treatment of a
specic disease is afforded by their products is largely
ignored.
37
An observation by the Federal Trade Com-
mission (FTC) is that the health claims of natural prod-
ucts are commonly deceptive and unsubstantiated.
40
FTC or FDA review of advertisements (print, televi-
sion, or otherwise) before they are released to the
general public should therefore be considered.
ASCPT believes that a number of opportunities exist
to enhance the education of the American public re-
garding dietary supplements. As such, ASCPT supports
increased education of consumers through health care
professionals and through reliable updated sources of
information such as MedWatch. In addition, we suggest
that advance review of natural product advertisements
by the FTC or FDA should be considered.
RECOMMENDATION 5
Recommendation 5 is to enhance research opportu-
nities on the adverse effects and efcacy of dietary
supplements.
Expanding research on dietary supplements is critical
to advancing our understanding of the toxicities and
efcacies associated with these agents. In addition, it is
crucially important to gain a better understanding of the
effects of dietary supplements in certain populations,
such as children and the elderly. As discussed, dietary
supplements are not tested for adverse effects or ef-
cacy with the same stringent scientic rigor that is
required of OTC and prescription drugs,
2
nor are they
subject to the regulation and approval processes of the
FDA.
Although, by law, dietary supplements cannot be
marketed for the diagnosis, treatment, or cure of dis-
ease, DSHEA permits the use of these products for
promoting general well-being and health.
1
Close exam-
ination of some of the claims of dietary supplements,
however, may not support their efcacy for promoting
health.
2,6,40
In instances where there is not sufcient
scientic data to understand toxicities associated with
dietary supplements, to evaluate claims of efcacy, or
to study their effects in various populations, ASCPT
proposes that NCCAM be provided with sufcient
funds to support research in these areas. Because NIH
study sections may not consider traditional clinical
pharmacology studies of dietary supplements to be
particularly innovative, mechanisms other than the
investigator-initiated proposal (R01s), such as Requests
for Applications (RFAs) or Requests for Proposals
(RFPs), may be useful to fund these investigations.
Such steps would be in the best public interest, because
most dietary supplements cannot be patented and there-
fore usually lack sponsors for studies of efcacy and
toxicity.
41
Promoting such research initiatives will help to en-
sure that the consumer has adequate data from which to
make an informed decision about natural products. It is
the opinion of ASCPT that NCCAM and the FDA must
take leading roles in addressing research related to
dietary supplements. Ongoing and expanded interac-
tions between these agencies are critical to taking ad-
vantage of the opportunities to advance our understand-
ing of dietary supplements and, in so doing, to improve
the safety of these products for the American public and
assist the consumer in making informed choices with
respect to the use of dietary supplements.
CONCLUSION
ASCPT is the largest and oldest scientic and pro-
fessional organization devoted to serving the discipline
of clinical pharmacologythe study of drugs in hu-
mansthrough the continuum of discovery, develop-
ment, regulation, and use. As a learned organization,
our primary concerns are better understanding and use
of existing drug therapies and developing better and
safer treatments for the future. Our members comprise
physicians and other doctoral scientists, pharmacists,
nurses, research coordinators, fellows in training, and
other professionals committed to making an impact for
patients and society. They hold a unique combination
of scientic and clinical expertise making them espe-
cially qualied to understand the impact of disease on
patients, the compelling need for effective drug ther-
apy, and the efforts necessary to meet those needs.
ASCPT believes that enhanced oversight of dietary
supplements and research to better determine dietary
supplement safety and efcacy will help to identify
their potential for harm, as well as their benets. A
coordinated effort among dietary supplement manufac-
turers, government oversight and research agencies, the
health care community, and the public to discuss sup-
plement use, AEs, or suspected supplement interactions
with other supplements, drugs, or OTC products will
help to build and sustain a solid knowledge base and
advance overall health and wellness.
The reviewers of this position statement were as follows: Alan S.
Hollister, MD, PhD, Glaxo Smith-Kline, Philadelphia, Pa; Mitchell
A. H. Levine, MD, FRCPC, McMaster University, Hamilton, On-
tario, Canada; William M. Wardell, MD, PhD, Wardell Associates
International, LLC, Princeton, NJ; Roger L. Williams, MD, US
Pharmacopeia, Rockville, Md; and Raymond L. Woosley, MD, PhD,
University of Arizona, Tucson, Ariz.
This position paper was reviewed in draft form by ASCPT mem-
bers chosen for their diverse views and scientic expertise. Review-
ers provided candid and relevant comments and suggestions; how-
ever, they were not asked to endorse the recommendations or
conclusions, nor did they see the nal draft before its approval by the
ASCPT Board of Directors.
This position statement was also reviewed and approved by the
ASCPT Government Affairs Committee: Robert E. Vestal, MD,
Chair, Darrell R. Abernethy, MD, PhD, Matthew M. Ames, PhD,
Timi I. Edeki, MD, PhD, Raymond E. Galinsky, PharmD, Barry
Mangum, PharmD, Dennis J. McCarthy, PhD, Robert J. Noveck,
MD, PhD, FACEP, and Simi Vincent, MD, PhD.
Dr Morrow has consulted for Unilever Ltd (United Kingdom) and
has received research funding from Pharmacia-Upjohn Pharmaceuti-
cals; Dr Edeki has received consulting and lecture fees from Purdue
Pharma and Novartis Pharmaceuticals, has served as a consultant to
Janssen Pharmaceuticals, and is currently employed by Abbott Lab-
oratories; Dr El Mouelhi has no conict of interest; Dr Galinsky has
no conict of interest; Dr Kovelesky has no conict of interest; Dr
Noveck has no conict of interest; and Dr Pruess has no conict of
interest.
References
1. Dietary Supplement Health and Education Act
(DSHEA), Public Law 103-417, 1994.
2. National Research Council. 2005. Dietary supplements: a
framework for evaluating safety. Available from: URL:
http://www.nap.edu/books/0309091101/html/. Accessed
Jan 20, 2005.
3. Balluz LS, Kieszak SM, Philen RM, Mulinare J. Vitamin
and mineral supplement use in the United States: results
from the Third National Health and Nutrition Examina-
tion Survey. Arch Fam Med 2000;9:258-62.
4. Slesinski MJ, Subar AF, Kahle LL. Trends in use of
vitamin and mineral supplements in the United States: the
1987 and 1992 National Health Interview Surveys. J Am
Diet Assoc 1995;95:921-3.
5. Blendon RJ, Earaches CM, Benson JM, Brodie M,
Altman DE. Americans views on the use and regula-
tion of dietary supplements. Arch Intern Med 2001;
161:805-10.
6. De Smet PAGM. Health risks of herbal remedies: an
update. Clin Pharmacol Ther 2004;76:1-17.
7. Stein CM. Are herbal products dietary supplements or
drugs? An important question for public safety. Clin
Pharmacol Ther 2002;71:411-3.
8. Wooltorton E, Sibbald B. Ephedra/ephedrine: cardiovas-
cular and CNS effects. CMAJ 2002;166:633.
9. Haller CA, Benowitz NL. Adverse cardiovascular and
central nervous system events associated with dietary
supplements containing ephedra alkaloids. N Engl J Med
2000;343:1833-8.
10. Hertzman PA, Falk H, Kilbourne EM, Page S, Shulman
LE. The eosinophilia-myalgia syndrome: the Los Alamos
Conference. J Rheumatol 1991;18:867-73.
11. Center for Food Safety and Applied Nutrition, Ofce of
Nutritional Products, Labeling, and Dietary Supplements,
US Food and Drug Administration. Information paper on
L-tryptophan and 5-hydroxy-L-tryptophan. February
2001. Available from: URL: http://www.cfsan.fda.gov/
dms/ds-tryp1.html.Accessed Jan 20, 2005.
12. Wooltorton E. Herbal kava: reports of liver toxicity.
CMAJ 2002;166:777.
13. Russmann S, Lauterburg BH, Helbling A. Kava hepato-
toxicity. Ann Intern Med 2001;135:68-9.
14. Congressional Record, Proceedings and debates of the
108th Congress, second session [SenateJune 21].
2004;86:S7077-S7086.
15. 69 Federal Register 64957 (2004) [Docket No. 2004N-
0458]. Dietary supplements: strategy for further im-
plementation and enforcement of the Dietary Supple-
ment Health and Education Act of 1994; availability.
US Food and Drug Administration. Notice. Nov 9,
2004.
16. 69 Federal Register 64962 (2004) [Docket No. 2004D-
0466]. Draft guidance for industry: substantiation for
dietary supplement claims made under Section 403(r)(6)
of the Federal Food, Drug, and Cosmetic Act;
availability. Food and Drug Administration. Notice. Nov
9, 2004.
17. Ofce of the Inspector General. Adverse event reporting
for dietary supplements: an inadequate safety valve. OEI-
01-00-00180. Boston: Ofce of the Inspector General,
Department of Health and Human Services; 2001.
18. Litovitz TL, Klein-Schwartz W, White S, Cobaugh DJ,
Youniss J, Drab A, et al. 1999 Annual Report of the
American Association of Poison Control Centers Toxic
Exposure Surveillance System. Am J Emerg Med 2000;
18:517-74.
19. MedWatch: The FDA safety information and adverse
event reporting program. Food and Drug Administra-
tion (US). Available from: URL: http://www.fda.gov/
medwatch/. Accessed Nov 12, 2004.
20. US Pharmacopeia. Dietary supplements. Available from:
URL: http://www.usp.org/dietarysupplements/. Accessed
Nov 12, 2004.
21. Marcus DM, Grollman AP. Botanical medicinesthe
need for new regulations. N Engl J Med 2002;347:
2073-6.
22. 68 Federal Register 12158 (2003) (codied at 21 CFR
Parts 111 and 112). Current good manufacturing practice
in manufacturing, packing, or holding dietary ingredients
and dietary supplements. US Food and Drug Adminis-
tration. Proposed rule.
23. Ofce of the Inspector General. Dietary supplements
labels: an assessment. OEI-01-01-00121. Boston: Ofce
of the Inspector General, Department of Health and Hu-
man Services; 2003.
24. Report of the Commission on Dietary Supplement La-
bels. Presidents Commission on Dietary Supplement
Labels. Nesheim MC, Chair. 1997 Nov.
25. Moorman CH. Challenges of OTC labeling. Am Pharm
Outsourcing 2004;5:8-15.
26. Ofce of the Inspector General. Dietary supplements
labels: key elements. OEI-01-01-00120. Boston: Ofce
of the Inspector General, Department of Health and Hu-
man Services; 2003.
27. Consumers Union. Press release; 9 of 10 consumers want
dietary supplements to be proven safe before put on store
shelves. June 7, 2004. Available from: URL: http://www.
consumersunion.org/pub/core_product_safety/001171.html.
Accessed Jan 20, 2005.
28. Mackowiak ED. Natural product education. Am J Pharm
Ed 2003;67:15-7.
29. Shields KM, McQueen CE, Bryant PJ. Natural product
education in schools of pharmacy in the United States.
Am J Pharm Ed 2003;67:43-8.
30. Wetzel MS, Eisenberg DM, Kaptchuk TJ. Courses in-
volving complementary and alternative medicine at US
medical schools. JAMA 1998;280:784-7.
31. Sergeant MJ. Medical school courses in alternative med-
icine [letter]. JAMA 1999;281:609-10.
32. Park CM. Diversity, the individual, and proof of efcacy:
complementary and alternative medicine in medical ed-
ucation. Am J Public Health 2002;92:1568-72.
33. Wetzel MS, Kaptchuk TJ, Haramati A, Eisenberg DM.
Complementary and alternative medical therapies: impli-
cations for medical education. Ann Intern Med 2003;138:
191-6.
34. Brotherton SE, Rockey PH, Etzel SI. US graduate
medical education, 2003-2004. JAMA 2004;292:
1032-7.
35. DailyMed initiative enhancing patient safety through ac-
cessible medication information. Available from: URL:
http://www.fda.gov/cder/regulatory/ersr/2003_02_13_
dailymed/. Accessed Nov 12, 2004.
36. International Bibliographic Information on Dietary
Supplements (IBIDS) database. Available from: URL:
http://dietary-supplements. info. nih. gov/Health_
Information/IBIDS.aspx. Accessed November 12, 2004.
37. Landers SJ. Alphabet overload: which dietary supplements
are effective? American Medical News online. Sept 6, 2004.
Available from: URL: http://www.ama-assn.org/amednews/
2004/09/06/hlsa0906.htm. Accessed Nov 12, 2004.
38. ConsumerReports.org Web site. Dangerous supplements:
still at large. May 2004. Available from: URL: http://
www.consumerreports.org/main/content/display_report.
jsp?FOLDER%3C%3Efolder_id419337&ASSORTMENT
%3C%3East_id333141&bmUID1106339029120#.
Accessed Jan 12, 2005.
39. Consumers Union. Consumers Union dietary supple-
ments survey. June 4, 2004. Available from: URL:
http://www.consumersunion.org/pub/Supplements-press
%20version.pdf. Accessed Nov 12, 2004.
40. Degan FH. Federal regulation of dietary supplements:
eight years after DSHEA, where are we? J Am Nutra-
ceutical Assoc 2003;6:20-4.
41. OHara M, Kiefer D, Farrall K, Kemper K. A review of
12 commonly used medicinal herbs. Arch Fam Med
1998;7:523-36.
COMMENTARY
Pharmacogenetics and response to
-adrenergic receptor antagonists
in heart failure
Mordechai Muszkat, MD, and C. Michael Stein, MD Jerusalem, Israel, and Nashville, Tenn
Congestive heart failure (CHF) is a major cause of
morbidity and death that is becoming more prevalent as
the proportion of the population that is elderly increas-
es.
1
Several advances in the treatment of heart failure,
including therapy with -adrenergic receptor (-AR)
antagonists, have improved the prognosis of patients
with CHF. The -AR antagonists metoprolol, carve-
dilol, and bisoprolol improve quality-of-life and hemo-
dynamic measures and lower mortality and hospitaliza-
tion rates.
2-5
Treatment with a -blocker resulted in an
approximately 50% lower incidence of a combined end
point of hospital admission for worsening CHF and
cardiac death within a year.
5
However, despite the
proven efcacy of -blockers, the majority of elderly
patients with CHF do not receive them.
6
There are several reasons why so many patients with
CHF are not treated with -blockers, but concern about
the risks of adverse effects is prominent.
1
Adverse
events from -blocker therapy occur in 25% to 43% of
patients with CHF, and discontinuation of therapy is
frequent, ranging from 8% to 25%.
7,8
If -blockers are
withdrawn, this usually occurs during the rst 3 months
of treatment and is often a result of hypotension, bra-
dycardia, and worsening of heart failure. These adverse
effects may occur after very low doses and soon after
treatment is started.
8
For example, a test dose of 5 mg
metoprolol was not tolerated because of hypotension or
bradycardia in 4% of patients with idiopathic dilated
cardiomyopathy,
9
and some degree of hypotension oc-
curred after the rst dose of -blocker in more than half
of elderly patients with CHF.
10
Several clinical factors such as age and severity of
left ventricular dysfunction contribute to the ability of
an individual patient to tolerate a -blocker. However,
because genetic factors contribute to both the metabo-
lism of -blockers and the response to them, these may
be important. Several recent articles in the Journal have
addressed the contribution of genetic variants in cyto-
chrome P450 (CYP) 2D6 and the
1
-adrenergic recep-
tor (
1
AR) to interindividual variability in response to
-blocker therapy.
11-15
Metoprolol is a -blocker often used to treat patients
with heart failure. Genetic variants of CYP2D6 have a
marked effect on plasma concentrations of metoprolol.
The peak plasma concentration of metoprolol is 3-fold
higher and the area under the plasma concentration
time curve (AUC) is 6-fold higher in subjects with a
nonfunctional CYP2D6 enzyme (poor metabolizers
[PMs]) compared with extensive metabolizers (EMs).
16
In several studies these higher plasma concentrations of
metoprolol resulted in a larger immediate reduction in
resting heart rate and exercise-induced tachycardia in
CYP2D6 PMs compared with EMs.
16-18
CYP2D6 also
contributes to the metabolism of carvedilol, and plasma
concentrations are 2 to 3 times higher in PMs.
19
Despite evidence that CYP2D6 genotype affects
plasma concentrations of metoprolol and carvedilol, the
clinical implications have not been explored exten-
sively. In a study of 37 patients with hypertension who
discontinued treatment with metoprolol because of ad-
From the Unit of Clinical Pharmacology, Department of Medicine,
Hadassah University Hospital, Jerusalem, and Division of Clinical
Pharmacology, Vanderbilt University School of Medicine, Nash-
ville.
Supported by US Public Health Service grants HL 56251, HL 04012,
and GM 31304.
Received for publication Oct 8, 2004; accepted Oct 13, 2004.
Reprint requests: Mordechai Muszkat, MD, Unit of Clinical Pharma-
cology, Department of Medicine, Hadassah University Hospital,
PO Box 12000, Jerusalem, Israel, 91120.
E-mail: muszkatm@hadassah.org.il
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.007
123
verse events, the incidence of CYP2D6 PM phenotype
was similar in patients who discontinued treatment with
metoprolol and those who did not.
20
In contrast, a
recent retrospective study suggested that CYP2D6
polymorphism may be an important determinant of
adverse events and, thus, withdrawal of -blocker ther-
apy. Of 24 subjects who had pronounced adverse ef-
fects after treatment with metoprolol, 9 (37.5%) were
PMs; this frequency was approximately 5-fold higher
than that expected in the general population.
11
How-
ever, the study was small and retrospective and only
included 5 patients with CHF.
11
More recently, Zineh
et al
15
reported in the Journal that, despite a 29.6-fold
variability in metoprolol AUC in patients with hyper-
tension, side effects were not related to drug concen-
tration or CYP2D6 genotype.
In addition to genetic variants that affect plasma
concentrations of a drug, variants in the drug target, the
1
AR, could also alter clinical responses to -blocker
therapy. There are 2 common nonsynonymous poly-
morphisms in the coding region of the ADRB1 gene,
Gly49Ser, which is located in the extracellular termi-
nus, an area suggested to be important for receptor
internalization and trafcking, and Gly389Arg, which
is located in the intracellular carboxy terminus in the
putative G-protein binding site.
21,22
The prevalence of
the Gly49 allele is approximately 15%, with no differ-
ence between white subjects and black subjects,
whereas the Gly389 allele is more frequent in white
subjects (42%) than in black subjects (27%).
22
In vitro studies suggest that both polymorphisms are
functionally important. The Gly49 variant showed de-
creased receptor expression compared with Ser49 after
exposure to isoproterenol (INN, isoprenaline) for
24 hours,
23
and in the Arg389 variant, adenylyl cyclase
activity in response to agonist was 3 times greater than
that of the Gly389 variant.
24
Three studies have shown
that these variants affect responses to -blockers in
vivo.
12-14
In normotensive subjects the resting blood pressure
response to atenolol, a selective
1
AR antagonist, was
signicantly greater in Arg389 homozygous subjects
compared with Gly389 homozygous subjects.
12
A pro-
spective study of 40 hypertensive patients treated with
metoprolol conrmed this observation, showing a 2- to
3-fold greater reduction in diastolic blood pressure in
Arg389 homozygotes compared with Gly389 heterozy-
gotes.
13
However, in addition to the contribution of
variation at position 389, the effect of ADRB1 diplotype
was a signicant predictor of blood pressure response
to metoprolol. Patients with Ser49/Ser49 Arg389/
Arg389 had a decline of 14.7 mm Hg in diastolic blood
pressure as compared with 0.5 mm Hg in patients with
the Gly49/Ser49 Arg389/Gly389 diplotype.
13
In a third
study, there was a larger reduction in resting and exer-
cise heart rates and systolic blood pressure after meto-
prolol in healthy subjects homozygous for Arg389 than
in Gly389 homozygotes.
14
However, despite these
studies in healthy subjects and in hypertensive patients,
there is little information about the effect of ADRB1
genotype and response to -blockers in patients with
heart failure.
CHF results in increased sympathetic stimulation
and decreased -AR responsiveness
25
that is related to
decreased -AR density and reduced maximal
isoproterenol-stimulated adenylate cyclase stimula-
tion.
26
Treatment of patients with heart failure with a
1
AR antagonist was associated with an increase in
myocardial -AR density and improved hemodynamic
responses.
27
Thus variations in the ADRB1 gene that
alter
1
AR expression (the Ser49Gly polymorphism
23
)
or response to agonist (the Arg389Gly polymor-
phism
24
) could potentially be clinically relevant.
Studies in transgenic mice with cardiac overexpres-
sion of the ADRB1 variants found that hemodynamic
responses to -blockade were greater in mice overex-
pressing Arg389.
28
Similarly, in humans with heart
failure, preliminary evidence indicated that the pres-
ence of the Arg389 allele was associated with improve-
ment of left ventricular function during carvedilol treat-
ment.
28
There is, however, little information regarding
the effects of CYP2D6 and ADRB1 genotype on re-
sponse to -blocker therapy in patients with heart fail-
ure.
In this issue of the Journal, Terra et al
29
describe the
effect of variants in 2 candidate genes, CYP2D6 and
ADRB1, on response to -blocker therapy in 61 patients
with CHF. The primary end point was the inability to
tolerate -blocker therapy, dened as the need to dis-
continue metoprolol or the inability to reach the target
dose of 200 mg/d by the end of the titration period. A
secondary outcome was decompensated heart failure, a
composite measure that included any of the following
end points: death, cardiac transplantation, hospitaliza-
tion for worsening heart failure, increase in other heart
failure medications, or the need to discontinue meto-
prolol.
Although genotype did not affect the frequency of
the primary composite outcome measure, a signi-
cantly greater percentage of Gly389 carriers required an
increase in heart failure medications (48% compared
with 14%). Diplotype analysis of polymorphism at po-
sitions 49 and 389 revealed that 52% of patients car-
rying the combination of Ser49Ser/Arg389Gly, 42% of
124 Muszkat and Stein MARCH 2005
Ser49Gly/Arg389Gly patients, 23% of Ser49Ser/
Arg389Arg patients, and no Ser49Gly/Arg389Arg pa-
tients required an increase in concomitant medications
for heart failure. Interestingly, the 2 diplotypes that
included the Gly389 allele and were associated with an
increased need for additional heart failure treatment
were previously found to be associated with the small-
est reduction in diastolic blood pressure in response to
a -blocker.
13
Concordant with the notion that the
Gly389 allele is associated with decreased response to
-blocker was a trend toward a greater 6-minute walk
distance in Arg389 homozygous patients (357 m versus
293 m, P .06). However, ADRB1 variants did not
affect the nal doses achieved over the titration period.
As expected, metoprolol concentrations were signif-
icantly affected by CYP2D6 genotype. Concentrations
of S-metoprolol were 2 to 4 times higher in CYP2D6
PMs compared with EMs; however, as was observed in
patients with hypertension,
15
this pharmacokinetic dif-
ference was not associated with clinically detectable
differences in effect. The lack of a clinically detectable
effect of genetic variations that have marked effects on
metoprolol concentrations is counterintuitive and may
be related to the relatively small number of patients
studied, clinical heterogeneity, or a weak relationship
between plasma concentration and clinical response.
The ndings presented by Terra et al
29
are an impor-
tant preliminary step toward a better understanding of
interindividual variability in the response to
1
AR an-
tagonists in patients with heart failure during the dose-
escalation period. The number of patients studied was
small, and additional larger studies will be required to
better dene the clinical impact of ADRB1 and
CYP2D6 polymorphism in patients with heart failure
receiving -blockers.
The authors have identied no conict of interest.
References
1. Consensus recommendations for the management of
chronic heart failure. On behalf of the membership of the
advisory council to improve outcomes nationwide in
heart failure. Am J Cardiol 1999;83:1A-38A.
2. Effect of metoprolol CR/XL in chronic heart failure: Meto-
prolol CR/XL Randomised Intervention Trial in Congestive
Heart Failure (MERIT-HF). Lancet 1999;353:2001-7.
3. Krum H, Roecker EB, Mohacsi P, Rouleau JL, Tendera
M, Coats AJ, et al, Carvedilol Prospective Randomized
Cumulative Survival (COPERNICUS) Study Group. Ef-
fects of initiating carvedilol in patients with severe
chronic heart failure: results from the COPERNICUS
Study. JAMA 2003;289:712-8.
4. Leizorovicz A, Lechat P, Cucherat M, Bugnard F. Biso-
prolol for the treatment of chronic heart failure: a meta-
analysis on individual data of two placebo-controlled
studiesCIBIS and CIBIS II. Cardiac Insufciency
Bisoprolol Study. Am Heart J 2002;143:301-7.
5. Zugck C, Haunstetter A, Kruger C, Kell R, Schellberg D,
Kubler W, et al. Impact of beta-blocker treatment on the
prognostic value of currently used risk predictors in conges-
tive heart failure. J Am Coll Cardiol 2002;39:1615-22.
6. Sin DD, McAlister FA. The effects of beta-blockers on
morbidity and mortality in a population-based cohort of
11,942 elderly patients with heart failure. Am J Med 2002;
113:650-6.
7. Gottlieb SS, Fisher ML, Kjekshus J, Deedwania P,
Gullestad L, Vitovec J, et al. Tolerability of beta-blocker
initiation and titration in the Metoprolol CR/XL Ran-
domized Intervention Trial in Congestive Heart Failure
(MERIT-HF). Circulation 2002;105:1182-8.
8. Macdonald PS, Keogh AM, Aboyoun CL, Lund M,
Amor R, McCaffrey DJ. Tolerability and efcacy of
carvedilol in patients with New York Heart Association
class IV heart failure. J Am Coll Cardiol 1999;33:924-31.
9. Waagstein F, Bristow MR, Swedberg K, Camerini F,
Fowler MB, Silver MA, et al. Benecial effects of meto-
prolol in idiopathic dilated cardiomyopathy. Metoprolol
in Dilated Cardiomyopathy (MDC) Trial Study Group.
Lancet 1993;342:1441-6.
10. Baxter AJ, Spensley A, Hildreth A, Karimova G,
OConnell JE, Gray CS. Beta blockers in older persons
with heart failure: tolerability and impact on quality of
life. Heart 2002;88:611-4.
11. Wuttke H, Rau T, Heide R, Bergmann K, Bohm M, Weil
J, et al. Increased frequency of cytochrome P450 2D6
poor metabolizers among patients with metoprolol-
associated adverse effects. Clin Pharmacol Ther 2002;72:
429-37.
12. Sofowora GG, Dishy V, Muszkat M, Xie HG, Kim RB,
Harris PA, et al. A common beta1-adrenergic receptor
polymorphism (Arg389Gly) affects blood pressure re-
sponse to beta-blockade. Clin Pharmacol Ther 2003;73:
366-71.
13. Johnson JA, Zineh I, Puckett BJ, McGorray SP, Yarandi
HN, Pauly DF.
1
Adrenergic receptor polymorphisms
and antihypertensive response to metoprolol. Clin Phar-
macol Ther 2003;74:44-52.
14. Liu J, Liu ZQ, Tan ZR, Chen XP, Wang, LS, Zhou G,
et al. Gly389Arg polymorphism of beta1-adrenergic re-
ceptor is associated with the cardiovascular response to
metoprolol. Clin Pharmacol Ther 2003;74:372-9.
15. Zineh I, Beitelshees AL, Gaedigk A, Walker JR, Pauly
DF, Eberst K, et al. Pharmacokinetics and CYP2D6 ge-
notypes do not predict metoprolol adverse events or
efcacy in hypertension. Clin Pharmacol Ther 2004;76:
536-44.
16. Lennard MS, Silas JH, Freestone S, Ramsay LE, Tucker
GT, Woods HF. Oxidation phenotypea major determi-
nant of metoprolol metabolism and response. N Engl J Med
1982;307:1558-60.
2005;77(3):123-6 Pharmacogenetics and -blockers in heart failure 125
17. McGourty JC, Silas JH, Lennard MS, Tucker GT, Woods
HF. Metoprolol metabolism and debrisoquine oxidation
polymorphismpopulation and family studies. Br J Clin
Pharmacol 1985;20:555-66.
18. Hamelin BA, Bouayad A, Methot J, Jobin J, Desgagnes
P, Poirier P, et al. Signicant interaction between the
nonprescription antihistamine diphenhydramine and the
CYP2D6 substrate metoprolol in healthy men with high
or low CYP2D6 activity. Clin Pharmacol Ther 2000;67:
466-77.
19. Giessmann T, Modess C, Hecker U, Zschiesche M, Daz-
ert P, Kunert-Keil C, et al. CYP2D6 genotype and induc-
tion of intestinal drug transporters by rifampin predict
presystemic clearance of carvedilol in healthy subjects.
20. Clack DWJ, Morgan AKW, Waal-Manning H. Adverse
effects from metoprolol are not generally associated with
oxidation status [letter]. Br J Clin Pharmacol 1984;18:
965-7.
21. Maqbool A, Hall AS, Ball SG, Balmforth AJ. Common
polymorphisms of beta1-adrenoceptor: identication and
rapid screening assay [letter]. Lancet 1999;353:897.
22. Moore JD, Mason DA, Green SA, Hsu J, Liggett SB. Racial
differences in the frequencies of cardiac beta(1)-adrenergic
receptor polymorphisms: analysis of c145AG and
c1165GC. Hum Mutat 1999;14:271.
23. Rathz DA, Brown KM, Kramer LA, Liggett SB. Amino
acid 49 polymorphisms of the human beta1-adrenergic
receptor affect agonist-promoted trafcking. J Cardio-
vasc Pharmacol 2002;39:155-60.
24. Mason DA, Moore JD, Green SA, Liggett SB. A gain-of-
function polymorphism in a G-protein coupling domain of
the human beta1-adrenergic receptor. J Biol Chem 1999;
274:12670-4.
25. Ginsburg R, Bristow MR, Billingham ME, Stinson EB,
Schroeder JS, Harrison DC. Study of the normal and
failing isolated human heart: decreased response of fail-
ing heart to isoproterenol. Am Heart J 1983;106:535-40.
26. Bristow MR, Ginsburg R, Minobe W, Cubiccioti RS,
Segman WS, Lurie K, et al. Decreased catecholamine
sensitivity and beta-adrenergic-receptor density in failing
human hearts. N Engl J Med 1982;307:205-11.
27. Heilbrunn SM, Shah P, Bristow MR, Valantine HA, Gins-
burg R, Fowler MB. Increased beta-receptor density and
improved hemodynamic response to catecholamine stimu-
lation during long-term metoprolol therapy in heart failure
from dilated cardiomyopathy. Circulation 1989;79:483-90.
28. Perez JM, Rathz DA, Petrashevskaya NN, Hahn HS,
Wagoner LE, Schwartz A, et al. Beta1-Adrenergic recep-
tor polymorphisms confer differential function and pre-
disposition to heart failure. Nat Med 2003;9:1300-5.
29. Terra SG, Pauly DF, Lee CR, Patterson JH, Adams KF
Jr, Schoeld RS, et al. -Adrenergic receptor polymor-
phisms and responses during titration of metoprolol con-
trolled release/extended release in heart failure. Clin
126 Muszkat and Stein MARCH 2005
PHARMACOGENETICS AND
GENOMICS
-Adrenergic receptor polymorphisms and
responses during titration of metoprolol
controlled release/extended release in
heart failure
Objective: -Blockers require careful initiation and titration when used in patients with heart failure. Some
patients tolerate -blocker therapy initiation without difficulty, whereas in other patients this period presents
clinical challenges. We tested the hypothesis that polymorphisms at codons 389 (Arg389Gly) and 49
(Ser49Gly) of the
1
-adrenergic receptor would be associated with differences in initial tolerability of
-blocker therapy in patients with heart failure. We also tested whether polymorphisms in the
2
-adrenergic
receptor, G-protein
s
subunit (G
s
), and cytochrome P450 (CYP) 2D6 genes or S-metoprolol plasma
concentrations were associated with -blocker tolerability.
Methods: Sixty-one -blockernaive patients with systolic heart failure were prospectively enrolled. Patients
began taking 12.5 to 25 mg metoprolol controlled release/extended release with titration every 2 weeks (as
tolerated) to 200 mg/d or the maximum tolerated dose over a period of 8 to 10 weeks. Decompensation was
the composite of death, heart failure hospitalization, increase in other heart failure medications, or need to
discontinue metoprolol. End points were assessed during the titration period.
continued on next page
Steven G. Terra, PharmD, Daniel F. Pauly, MD, PhD, Craig R. Lee, PharmD,
J. Herbert Patterson, PharmD, Kirkwood F. Adams, Jr, MD,
Richard S. Schofield, MD, Bernadette S. Belgado, PharmD, Karen K. Hamilton, MD,
Juan M. Aranda, MD, James A. Hill, MD, MS, Hossein N. Yarandi, PhD,
Joseph R. Walker, PharmD, Michael S. Phillips, PhD, Craig A. Gelfand, PhD, and
Julie A. Johnson, PharmD Gainesville, Fla, Chapel Hill, NC, and Princeton, NJ
From the Department of Pharmacy Practice, College of Pharmacy,
Division of Cardiology, Department of Medicine, Department of
Veterans Affairs Medical Center, and College of Nursing and
Biostatistics Unit, University of Florida, Gainesville; Division of
Pharmacotherapy, University of North Carolina at Chapel Hill
School of Pharmacy, and Division of Cardiology, Department of
Medicine, University of North Carolina, Chapel Hill; and Orchid
Biosciences Inc, Princeton.
This work was supported by grants from Orchid Biosciences Inc, the
National Heart, Lung, and Blood Institute (HL68834), and by
General Clinical Research Centers program grants RR00082 (Uni-
versity of Florida) and RR00046 (University of North Carolina,
Chapel Hill), Division of Research Resources, National Institutes
of Health. Dr Terra was an American College of Clinical Pharmacy
Merck Cardiovascular Research Fellow and subsequently an
American Foundation for Pharmaceutical Education Clinical Phar-
macy Post-PharmD Fellow in Biomedical Research Sciences at the
time of this work.
Received for publication July 1, 2004; accepted Oct 1, 2004.
Reprint requests: Julie A. Johnson, PharmD, University of Florida
College of Pharmacy, Department of Pharmacy Practice, 1600 SW
Archer Rd, PO Box 100486, Gainesville, FL 32610.
E-mail: johnson@cop.u.edu
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.006
127
Results: The overall rate of decompensation was not different between the codon 49 or 389 genotypes.
However, a significantly greater percentage of patients with the Gly389 variant required increases in heart
failure medications as compared with Arg389 homozygotes (48% versus 14%, respectively; P .006).
Similarly, patients with the Ser49 homozygous genotype were significantly more likely to require increases in
concomitant heart failure therapy as compared with Gly49 carriers (41% versus 11%, respectively; P .03).
Neither CYP2D6 genotypes nor metoprolol pharmacokinetics differed between patients with and those
without a decompensation event. There was no association between the
2
-adrenergic receptor or G
s
polymorphisms with decompensated heart failure.

Conclusions: Patients with the Gly389 variant and Ser49Ser genotype were significantly more likely to
require increases in heart failure medications during -blocker titration and thus may require more frequent
follow-up during titration. (Clin Pharmacol Ther 2005;77:127-37.)
The -blockers carvedilol, metoprolol succinate, and
bisoprolol have been shown to signicantly reduce
morbidity and mortality rates in patients with heart
failure and left ventricular dysfunction and, accord-
ingly, are recommended in consensus guidelines for the
treatment of systolic heart failure.
1
Despite the over-
whelming benets observed in clinical trials, -blocker
utilization has been less than optimal. A recent Euro-
pean survey demonstrated that only 34% of heart fail-
ure patients were receiving -blockers, and among pa-
tients with left ventricular dysfunction, -blockers were
used in only 20%.
2
Among the reasons for the rela-
tively low use of -blockers is the initial risk of cardiac
decompensation or worsening heart failure that can
occur on initiation of therapy in certain patients, par-
ticularly those with more advanced heart failure.
3
The
initiation of -blocker therapy also requires frequent
clinic visits for dose titration and diligent management
of uid balance and concomitant medications to de-
crease the risk of cardiac deterioration. Thus identi-
cation of factors that increase susceptibility to initial
decompensation might enhance the uptake of this ther-
apy in clinical practice.
The
1
-adrenergic receptor (
1
AR) mediates the del-
eterious effects of the heightened sympathetic nervous
system activity in heart failure, and it contains func-
tional genetic polymorphisms at codon 49 (Ser49Gly)
and 389 (Arg389Gly). For example, an in vitro func-
tional study of the codon 389 polymorphism showed
that the Arg389 form of the receptor exhibited greater
basal and agonist-stimulated adenylyl cyclase activity.
4
The results of studies of human tissues have either been
consistent with this nding
5
or shown no differences by
genotype.
6
The early human studies were somewhat
discouraging; 2 similar studies showed that the
1
AR
codon 389 genotype was not associated with the heart
rate response to exercise.
7,8
However, there is greater
consistency in the -blocker pharmacogenetic litera-
ture, with most studies to date showing greater
-blocker response in Arg389 homozygotes.
9-12
At
codon 49, the Ser49 variant has been shown in in vitro
studies to be resistant to agonist-mediated down-
regulation,
13,14
and in one study transfected cells with
the Gly49 variant produced markedly greater basal and
agonist-stimulated adenylyl cyclase activities com-
pared with Ser49.
14
Some studies in humans have also
suggested functional relevance of the codon 49 poly-
morphism.
15
Given the putative role of the
1
AR polymorphisms
on the response to -blockers, we hypothesized that
polymorphisms in the
1
AR gene would be associated
with differences in the initial tolerability or rates of
worsening heart failure from -blocker therapy in heart
failure patients. In addition, we studied polymorphisms
in the
2
2
AR) and G-protein
s
subunit (G
s
) genes because studies suggest that these
genetic variants are associated with response to
-agonists or -blockers.
16-18
Given that metoprolol is
metabolized primarily by cytochrome P450 (CYP)
2D6, we also determined whether the CYP2D6 geno-
typedetermined phenotype or S-metoprolol concentra-
tions were associated with either decompensation or
tolerability to metoprolol controlled release/extended
release (CR/XL) (Toprol XL; AstraZeneca LP, Wil-
mington, Del).
METHODS
Patients were recruited for this prospective study
from the University of Florida (Gainesville, Fla) and
University of North Carolina (Chapel Hill, NC) Heart
Failure Clinics. All clinicians involved in the clinical
care of these patients were blinded to the genotypic
data, and all investigators determining the genotypes
were blinded to the patients clinical course. The study
was approved by the institutional review board at each
institution, and each patient gave written informed con-
sent for participation in the study. The study population
included patients aged 18 years or greater with symp-
128 Terra et al MARCH 2005
toms of heart failure of an ischemic or nonischemic
etiology in New York Heart Association (NYHA) func-
tional class II or III and with an ejection fraction of
40% or lower. Patients were not eligible for enrollment
if they were receiving -blocker therapy. Other exclu-
sion criteria were NYHA functional class IV, systolic
blood pressure lower than 90 mm Hg, heart rate lower
than 55 beats/min, bronchospastic lung disease, greater
than rst-degree heart block, active myocarditis, hyper-
trophic obstructive cardiomyopathy, and treatment with
sotalol.
After informed consent was obtained, a 5-mL blood
sample for isolation of genomic deoxyribonucleic acid
(DNA) was collected. Baseline evaluations included
assessment of vital signs while the patient was sitting
and body weight, a 6-minute walk test to assess sub-
maximal exercise tolerance,
19
and the Minnesota Liv-
ing with Heart Failure (MLWHF) questionnaire.
20
Af-
ter completion of baseline studies, patients began
taking metoprolol CR/XL at a dose of 12.5 mg/d
(NYHA class III) or 25 mg/d (NYHA class II), depend-
ing on functional class. The dose was then doubled on
a biweekly basis up to 200 mg/d or maximum tolerated
dose. Thus the time for up-titration was a period of 8 to
10 weeks, depending on the starting dose of metoprolol
CR/XL. The dosing regimen of metoprolol CR/XL
could be adjusted by the clinician. Concomitant medi-
cation changes were allowed during the titration period
but were discouraged except in response to a change in
symptoms. The 6-minute walk test, functional class
assessment, MLWHF questionnaire, and vital sign and
body weight measurements were repeated after 2 weeks
of receiving each dose of metoprolol CR/XL. In a
subset of 45 patients who consented to undergo addi-
tional blood sampling, a pharmacokinetic blood sample
was obtained approximately 2 to 4 hours after dosing
following 2 weeks of receiving each metoprolol CR/XL
dose for determination of S-metoprolol (active isomer)
concentrations. S-metoprolol concentrations were de-
termined by chiral HPLC with uorescence detection.
21
Determination of genotypes
Genomic DNA was isolated from whole blood by
use of a commercially available kit (Puregene; Gentra
Systems, Minneapolis, Minn). Polymerase chain reac-
tion (PCR) was carried out as previously described.
7
Genotypes were determined by pyrosequencing tech-
nology by use of the Pyrosequencing PSQ HS 96 sys-
tem (Pyrosequencing AB, Uppsala, Sweden). Haplo-
types were inferred with double heterozygotes assigned
as Ser49Gly389 and Gly49Arg389. This assignment
was based on the absence of Gly49Gly389 alleles in
more than 1200 alleles in patients in whom
1
AR
diplotype could be assigned in a previous study, despite
being predicted in 52 alleles. In addition, computa-
tional assessment of
1
AR haplotype revealed no
Gly49Gly389 alleles in a large population.
22
GenomeLab SNPstream CYP2D6 genotyping
methods
Genotyping. For this study, 9 CYP2D6 assays were
developed on a Beckman GenomeLab SNPstream
high-throughput genotyping platform (Beckman
Coulter, Inc, Fullerton, Calif). In brief, this genotyping
platform uses single-base primer extension chemistry to
query the base present at the site of interrogation. This
method involves a solution-phase multiplex PCR strat-
egy followed by a cycled primer extension reaction
with labeled terminating nucleotides, which, in turn, is
followed by solid-phase sorting onto a DNA microarray
for readout.
23
Because there is an extremely high degree of homol-
ogy between CYP2D6 and the nearby pseudogenes
(CYP2D7 and CYP2D8P),
24
a long PCR strategy was
developed to specically amplify the correct regions of
the genome before the genotyping reactions were per-
formed. A multiplexed reaction tested known single
nucleotide polymorphism positions for the presence of
minor allele variants associated with the *2, *3, *4, *6,
*9, *10, *17, *29, and *41 alleles of the CYP2D6 gene.
Analysis was performed in both forward and reverse
directions for improved genotyping accuracy and ro-
bustness. Thus 2 independent genotyping calls were
generated at each locus, and their concordance was
compared against one another before a nal genotype
call was assigned.
To further demonstrate assay design and multiplex-
ing accuracy, assays were extensively tested on Centre
dEtude Polymorphisme Humain pedigrees and control
samples of known genotype before the samples from
this study were screened. One hundred percent concor-
dance was observed with samples of known genotype
with the use of this panel (data not shown).
The CYP2D6-predicted phenotype was assigned on
the basis of the activity of the combination of alleles
present (fully functional, partially functional, or
nonfunctional).
25-30
Extensive metabolizers (EMs)
were dened as having at least 1 fully functional allele.
Subjects with 2 partially functional alleles or a partially
functional allele and a nonfunctional allele were as-
signed to the intermediate metabolizer (IM) cohort, and
subjects with 2 nonfunctional alleles were assigned to
the poor metabolizer (PM) group.
2005;77(3):127-37 -Receptor polymorphisms and response to metoprolol 129
Outcome measures. The primary end point was a
composite of poor tolerability dened as the need to
discontinue metoprolol CR/XL or an inability to reach
the target dose of 200 mg/d by the end of the titration
period. Decompensated heart failure was a secondary
end point dened as the composite of death or cardiac
transplantation, hospitalization for worsening heart fail-
ure, increase in any heart failure medication (eg, diuret-
ics, angiotensin-converting enzyme inhibitors, angio-
tensin receptor blockers, digoxin, and spironolactone)
for worsening symptoms, or need to discontinue meto-
prolol CR/XL for any reason during the titration period.
The increase in heart failure medications was treated as
a dichotomous end point, and with regard to diuretics,
the increase was dened as a physician-initiated in-
crease as a result of volume overload. The incidence of
the decompensation end point was assessed during the
titration period (ie, 8-10 weeks) only. It was anticipated
that during this short time period there would be few, if
any, deaths or transplantations. Yet, if these events
occurred, they would clearly have represented decom-
pensation, and so, for completeness, these were in-
cluded in the composite end point. In addition to the
decompensation composite end point, we also assessed
the tolerability to metoprolol CR/XL on the basis of the
scores on the MLWHF questionnaire, distance covered
in the 6-minute walk test, and dose of metoprolol
CR/XL attained by the end of the titration period. These
end points were compared by genotype group for the
various polymorphisms in the
1
AR,
2
AR, and G
s
genes.
Statistical analysis
Power calculations were based on the assumption
that 15% of patients permanently discontinue -blocker
therapy and another 35% are unable to reach the target
dose (for a total of 50% with poor tolerability).
31
On the
basis of this and a 56% frequency of the Arg389 ho-
mozygous genotype, with 60 patients, we had 80%
power to detect a 2.4-fold relative risk of poor tolera-
bility among homozygous Arg389 patients as com-
pared with Gly389 carriers with a 2-tailed of .05. We
would similarly have 80% power to detect a 2.4-fold
relative risk of decompensation in Arg389 homozy-
gotes and Gly389 carriers. Patients were compared by
genotype relative to the decompensation endpoints by
use of a chi square or Fisher exact test, as appropriate.
Patients were also divided by genotype and compared
with regard to continuous parameters (eg, MLWHF
scores, 6-minute walk distance, dose of metoprolol
CR/XL at end of titration, systolic blood pressure, and
heart rate), as well as changes in these parameters.
Comparisons were made by unpaired t test, ANOVA,
or Kruskal-Wallis test, as appropriate. Discrete vari-
Table I. Baseline demographic characteristics stratied by
1
AR genotype
Variable
Arg389Arg
(n 28)
Gly389 carriers
(n 33)
Ser49Ser
(n 43)
Gly49 carriers
(n 17)
Age (y) 59 12 55 14 58 13 55 12
Men 18 (64%) 19 (57%) 25 (58%) 11 (64%)
White/black/Hispanic 21 (75%)/7 21 (63%)/11 32 (74%)/11 10 (59%)/6
(25%)/0 (0%) (34%)/1 (3%) (26%)/0 (0%) (35%)/1 (6%)
NYHA functional class II/III 17 (60%)/11 (40%) 22 (66%)/11 (34%) 25 (58%)/18 (42%) 13 (76%)/4 (23%)
Ischemic heart failure 7 (25%) 12 (36%) 15 (34%) 4 (23%)
Body mass index (kg/m
2
) 27 6 29 6 27 6 30 6
Left ventricular ejection fraction (%) 23 5 22 9 21 7 25 9
Serum sodium (mEq/L) 138 3 138 3 138 4 139 3
Concomitant medical history
Diabetes 7 (25%) 12 (36%) 14 (32%) 5 (29%)
Hypertension 14 (50%) 17 (51%) 21 (48%) 9 (52%)
Atrial brillation 6 (21%) 6 (18%) 11 (25%) 1 (5%)
Background therapy
ACE inhibitor/ARB 28 (100%) 33 (100%) 43 (100%) 17 (100%)
Furosemide 22 (78%) 31 (93%) 38 (88%) 14 (82%)
Digoxin 18 (64%) 28 (84%) 33 (76%) 13 (76%)
Spironolactone 9 (32%) 8 (24%) 12 (28%) 5 (29%)
Data are given as mean SD or number and percent. There were no signicant differences between genotype groups for any parameter.
1
AR,
1
-Adrenergic receptor; NYHA, New York Heart Association; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker.
ables are listed as frequency counts and percentages,
whereas continuous variables are described as mean
SD or median and interquartile range. Polymorphisms
that resulted in a homozygous variant genotype fre-
quency of less than 2% were combined with the het-
erozygotes for analysis.
All statistical analyses were performed with the use
of SAS (version 8; SAS Institute, Cary, NC). A P value
of less than .05 was considered statistically signicant.
RESULTS
Table I shows the clinical characteristics of the 61
patients enrolled in this study, with data stratied by the
1
AR genotypes at codons 49 and 389 because these
polymorphisms were the primary focus of the study.
There was 1 subject in whom determination of the
codon 49 genotype was unsuccessful. The Arg389 al-
lele frequency was 0.72; there were 32 heterozygotes
and 1 Gly389 homozygote. The Ser49 allele frequency
was 0.85; there were 17 heterozygotes and no Gly49
homozygotes. None of the baseline characteristics
listed in Table I was statistically signicant between
genotype groups. More Gly389 carriers were receiving
digoxin and furosemide at baseline compared with
Arg389 homozygotes, but the difference was not sta-
tistically signicant (P .07 and P .12, respec-
tively). Patients were predominantly middle-aged white
men; approximately 35% of patients had NYHA func-
tional class III heart failure at enrollment, and mean
ejection fraction was 22% 7%.
Impact of
1
AR polymorphisms on poor
tolerability to metoprolol CR/XL
A similar percentage of patients in each
1
AR ge-
notype was unable to reach the target dose of metopro-
lol CR/XL or required discontinuation of therapy dur-
ing the titration of therapy. Overall, 19 of 28 Arg389
homozygous patients (67%) had poor tolerability to
metoprolol CR/XL compared with 25 of 33 Gly389
carriers (75%) (P .62). For the codon 49 polymor-
phism, the frequency of poor tolerability was not dif-
ferent between genotype groups (72% and 76% for
Ser49Ser and Gly49 carriers, respectively).
Impact of
1
AR polymorphisms on risk of
decompensated heart failure
Table II lists the end points for decompensated heart
failure for the codon 49 and 389 genotype groups.
There was a trend for Gly389 carriers to have a higher
incidence of the composite end point of decompensated
heart failure compared with Arg389 homozygotes
within 8 to 10 weeks of initiation of metoprolol CR/XL.
This was largely the result of a signicantly greater
percentage of Gly389 carriers requiring increases in
other heart failure medications (mostly diuretics) for
symptoms of worsening heart failure during -blocker
titration (48% versus 14%, respectively; P .006).
Although not statistically signicant, patients with the
Ser49 homozygous genotype had a nominally higher
rate of decompensation compared with Gly49 carriers.
This nding was driven largely by a signicantly higher
proportion requiring increases in heart failure medica-
tions to treat worsening symptoms. The nal dose of
metoprolol CR/XL was not signicantly different be-
tween patients who required increases in heart failure
medications compared with those who did not reach
this end point (90 mg/d versus 118 mg/d, respectively;
P .16). There were also no signicant differences in
the baseline characteristics of age, serum sodium level,
furosemide dose, 6-minute walk distance, scores on the
Table II. Risk of decompensated heart failure stratied by
1
AR genotypes
Arg389Arg
(n 28)
Gly389 carriers
(n 33) P value
Ser49Ser
(n 43)
Gly49 carriers
(n 17) P value
Decompensated heart failure
(composite)
8 (28%) 16 (48%) .11 19 (44%) 5 (29%) .29
Death 0 0 0 0
Hospitalization for worsening
heart failure
1 (3.5%) 2 (6%) .99 3 (6%) 0
Increase of heart failure
medications
4 (14%) 16 (48%) .006 18 (41%) 2 (11%) .03
Discontinuation of metoprolol
CR/XL
3 (10%) 3 (9%) .99 3 (6%) 3 (17%) .33
Data are given as number and percent.
CR/XL, Controlled release/extended release.
MLWHF questionnaire, heart rate, or systolic blood
pressure between patients who required increased heart
failure medications and those who did not reach this
end point (data not shown).
1
AR diplotypes (based on codon 49 and 389
polymorphisms) were determined in 60 patients. The
observed diplotypes were SR/SG (Ser49Ser plus
Arg389Gly) (n 25), SR/SR (Ser49Ser plus Arg389Arg)
(n 17), SR/GR (Ser49Gly plus Arg389Arg) (n
10), SG/GR (Ser49Gly plus Arg389Gly) (n 7), and
SG/SG (Ser49Ser plus Gly389Gly) (n 1). Fig 1
shows that there was a signicant difference in the
percentage of patients who required increases in
heart failure medications when the response was
stratied by the 4 most commonly observed
1
AR
diplotypes (P .01). The presence of a Gly389
variant was associated with an increased likelihood
of requiring increases in concomitant heart failure
Fig 1. Percentage of patients requiring increases in concomitant heart failure medications during
metoprolol CR/XL titration according to
1
1
AR) diplotype.
Table III. Impact of
1
AR genotypes on 6-minute walk distance, quality-of-life scores, hemodynamics, and dose
of metoprolol CR/XL at end of titration
Arg389Arg
(n 28)
Gly389 carriers
(n 33)
Ser49Ser
(n 43)
Gly49 carriers
(n 17)
Baseline
End of
titration Baseline
End of
titration Baseline
End of
titration Baseline
End of
titration
Metoprolol CR/XL
dose (mg/d)
119 72 100 74 107 71 108 79
6-min walk
distance (m)
330 108 357 127 301 75 293 102* 298 89 306 111 346 88 359 131
MLWHF 28 21 21 22 36 27 30 20 31 25 25 20 37 25 31 25
Heart rate (beats/min) 80 12 64 12 79 10 70 12 80 11 68 12 79 12 65 12
Systolic blood
pressure (mm Hg)
120 18 110 20 119 17 111 18 118 18 110 18 121 17 111 21
Data are given as mean SD.
MLWHF, Minnesota Living with Heart Failure questionnaire.
*P .06, versus Arg389Arg at end of titration.
P .07, versus Ser49Ser at baseline.
P .09, versus Arg389Arg at end of titration.
therapy. In all, 52% of patients with the SR/SG
diplotype (Ser49Ser plus Arg389Gly) and 42% of
patients with the SG/GR diplotype (Ser49Gly plus
Arg389Gly) required increases in heart failure med-
ications during metoprolol CR/XL titration. The
composite decompensation end point was not signif-
icantly different between groups (52%, 29%, 30%,
and 42% for the SR/SG, SR/SR, SR/GR, and SG/GR
diplotypes, respectively; P .46).
Impact of
1
AR polymorphisms on 6-minute walk
distance, quality of life, and dose of metoprolol
CR/XL attained
We also assessed tolerability to metoprolol CR/XL
by measuring submaximal exercise tolerance (6-minute
walk distance) scores on the MLWHF questionnaire
and dose of metoprolol CR/XL achieved at the end of
the titration period. Table III lists these data for the
1
AR polymorphisms. At baseline, there were no sig-
nicant differences in any of these parameters between
the codon 49 and codon 389 genotypes. However,
Gly49 carriers tended to cover a greater distance in the
baseline 6-minute walk test compared with Ser49 ho-
mozygous patients. At the end of the titration period,
there was a trend for Arg389 homozygous patients to
have a greater exercise tolerance compared with
Gly389 carriers. The change from baseline in the
6-minute walk distance was not signicantly different
between the codon 389 genotype groups. Scores on the
MLWHF questionnaire and reduction in heart rate and
systolic blood pressure were similar between
1
AR
genotype groups (Table III). There were no differences
in metoprolol CR/XL dose at the end of the titration
period between genotype groups (Table III). In all, 67%
of patients with the Arg389Arg genotype attained
metoprolol CR/XL doses of 100 mg/d or greater com-
pared with 51% of Gly 389 carriers (P .30). For
codon 49, a similar percentage of patients reached a
nal dose of 100 mg/d or greater (58% in each codon
49 group).
Impact of
2
AR and G
s
polymorphisms on risk
of decompensation and tolerability to metoprolol
CR/XL
The
2
AR and G
s
polymorphisms were not associ-
ated with the dose of metoprolol CR/XL attained at the
end of the titration period. For example, among the 3
codon 16
2
AR genotypes, the dose of metoprolol was
essentially identical (107, 109, and 108 mg/d for
Arg16Arg, Arg16Gly, and Gly16Gly, respectively).
Other measures of tolerability such as 6-minute walk
distances or quality-of-life assessment were also not
different between genotype groups (data not shown).
The risk of development of decompensated heart failure
was also not different among the various genotypes
within these genes. For example, decompensated heart
failure occurred among 40% of Glu27 carriers com-
pared with 37% of patients who were Gln27 homozy-
gotes, and approximately 30% of patients with each
genotype required increased doses of heart failure med-
ication.
Metoprolol pharmacokinetics and CYP2D6
genotypedetermined phenotype
There were no pharmacokinetic differences between
patients with and those without a decompensation event
at any dose of metoprolol CR/XL. The medians and
interquartile ranges of S-metoprolol concentrations for
patients with decompensation were 4.8 ng/mL (4.0-7.7
ng/mL), 11.9 ng/mL (5.7-19.2 ng/mL), 26.8 ng/mL
(15.4-35.4 ng/mL), and 46.0 ng/mL (31.7-85 ng/mL)
for the 25-, 50-, 100-, and 200-mg doses, respectively.
Table IV. S-metoprolol plasma concentrations and decompensation rate stratied by CYP2D6 phenotype
EM IM PM P value*
Dose
25 mg 4.0 (2.8-7.3) (n27) 7.6 (4.0-10.4) (n8) 15.6 (12.9-20.2) (n4) .005
50 mg 11.9 (4.5-15.1) (n27) 6.4 (3.6-19.7) (n7) 27.0 (25.6-35.9) (n4) .01
100 mg 17.7 (11.9-28.5) (n21) 42.5 (13.3-42.9) (n5) 60.2 (49.2-83.2) (n3) .03
200 mg 50.3 (31.2-56.9) (n9) 69.5 (65.8-119.6) (n3) 169.6 (146.6-192.5) (n2) .02
Decompensation rate 32% 50% 25% .61
Data are given as median and interquartile range or percentage.
EM, Extensive metabolizer; IM, intermediate metabolizer; PM, poor metabolizer; n, sample size.
*P value between groups.
P .02, versus IM.
P .001, versus EM.
P .04, versus IM.
P .004, versus EM.
P .03, versus EM.
For patients with no decompensation, the medians and
interquartile ranges of concentrations were 5.7 ng/mL
(3.1-9.4 ng/mL), 12.1 ng/mL (4.8-20.1 ng/mL), 22.9
ng/mL (11.6-39.1 ng/mL), and 56.5 ng/mL (36.0-69.3
ng/mL) at these doses, respectively. Likewise, there
were no differences in S-metoprolol concentrations be-
tween the codon 389 genotype groups for all meto-
prolol CR/XL doses. For example, the median
S-metoprolol concentrations in Arg389 homozygotes
and Gly389 carriers at 100 mg/d were 22.8 ng/mL
(interquartile range, 13.3-35.3 ng/mL) and 22.6 ng/mL
(interquartile range, 12-41.6 ng/mL), respectively.
CYP2D6 genotype data were available in 51 pa-
tients, with genotype-determined phenotype distribu-
tions of 37 EMs, 10 IMs, and 4 PMs. As expected,
S-metoprolol plasma concentrations were signicantly
different between CYP2D6 phenotypes (Table IV). It
should be noted that plasma concentrations were mea-
sured in each patient at each dose that they received.
Therefore patients reaching the 200-mg dose would
have data included in each of the 4 dose steps. Consis-
tent with the plasma concentration data, the risk of
cardiac decompensation was not inuenced by
CYP2D6 genotypedetermined phenotype (32%, 50%,
and 25% for EMs, IMs, and PMs, respectively; P
.61). Similarly, the need for increases in background
heart failure medications was not different between the
3 CYP2D6 groups (data not shown). The nal daily
metoprolol CR/XL doses were 106 69, 100 75,
and 150 57 mg/d for EMs, IMs, and PMs (P .45).
There were no signicant differences in baseline or
end-of-study heart rate or systolic blood pressure values
when stratied by CYP2D6 genotypedetermined phe-
notype.
DISCUSSION
To our knowledge, this is the rst study that has
investigated the relationship between early tolerability
to a -blocker and genotype. The major nding of this
study was that the heart failure patients with the Gly389
variant and those with the Ser49 homozygous genotype
were signicantly more likely to require increases in
heart failure medications during -blocker titration than
patients with other
1
AR genotypes. Furthermore, the
combination of the codon 49 and 389 polymorphisms
(ie,
1
AR diplotypes) was also associated with a sig-
nicantly higher risk for increases in concomitant heart
failure medications, suggesting that these patients may
require more frequent follow-up during titration. In
support of the clinical end points, patients who were
Gly389 carriers tended to have depressed submaximal
exercise performance compared with Arg389 homozy-
gotes at the end of the titration period. Similarly, at
baseline, there was a trend for Gly49 carriers to have a
greater exercise performance compared with Ser49 ho-
mozygous patients. The composite end point of decom-
pensated heart failure tended to be higher among
Gly389 carriers as compared with Arg389 homozy-
gotes, but this study was slightly underpowered to
detect a difference in the magnitude observed. Specif-
ically, we had 80% power to detect a 2.4-fold differ-
ence in decompensation rates but observed a 1.7-fold
difference between Arg389 homozygotes and Gly389
carriers. Thus it is possible that there would be signif-
icant differences in the decompensation end point with
a larger sample.
The 6-minute walk distance is frequently used in
clinical trials as a measurement of submaximal exercise
performance, and exercise capacity is directly related to
survival rate among heart failure patients.
32
Our nd-
ings on submaximal exercise tolerance are consistent
with a previous study in which peak myocardial oxygen
consumption was signicantly lower among patients
with the Gly389 variant and Ser49 homozygotes.
33
The
nding of reduced submaximal and maximal exercise
tolerance among Gly389 carriers is also entirely con-
sistent with results of in vitro studies demonstrating that
the Gly389 variant is hypofunctional.
4
Moreover, in
one study heart failure patients with the Gly49 poly-
morphism had an improved survival rate as compared
with those who were Ser49 homozygotes.
15
It is unclear
whether the differences in exercise performance ac-
count for any of the apparent survival benet. The
1
AR polymorphisms at codons 49 and 389 are no
more common in heart failure patients than in a control
population.
15,34
We did not observe differences by genotype in the
heart rate or blood pressure response to metoprolol.
Among a number of studies looking at the association
of heart rate with codon 389 genotype, the vast majority
have shown no association. This is independent of
whether the study investigated the association with
resting heart rate,
7,8
the heart rate response to exer-
cise,
7,8
or the heart rate response to -blockade.
9,10
It is
unclear why heart rate or the heart rate response to
drugs is not associated with the codon 389 genotype
when other drug response phenotypes (eg, ejection
fraction change, blood pressure response to
-blockade) have been fairly consistently associated
with codon 389 genotype.
9-12
However, there is little
doubt that the processes controlling heart rate and blood
pressure, for example, are different, and thus it is not
unreasonable that genetic associations might be differ-
ent for different phenotype responses to an agonist
versus an antagonist. The blood pressure ndings are in
contrast to previous ndings from our laboratory
9
and
others regarding the genetic associations with
-blocker effect on blood pressure. However, this is not
surprising for a number of reasons, and these are ex-
plained relative to our previous ndings in hypertensive
patients.
The rst factor is the method of measuring blood
pressure. In this study blood pressure assessments were
based on duplicate values obtained in the clinic,
whereas in our previous study we used 24-hour ambu-
latory blood pressure. The latter approach signicantly
reduces variability and consequently increases study
power. The second factor is the study population. Pa-
tients with heart failure are known to have diminished
blood pressurelowering effects from -blockers and
other antihypertensive drugs. This is presumably a re-
sult of the heightened activity of the sympathetic and
renin-angiotensin system that occurs in patients with
heart failure and the interplay between vascular resis-
tance and stroke volume, which differs dramatically
between those with normal (eg, hypertensive) versus
abnormal left ventricular function. Third, in the hyper-
tension study metoprolol was titrated to a goal diastolic
blood pressure response and the mean doses achieved
were substantially higher than those in the population
of heart failure patients. Fourth, patients in this study
received a variety of medications known to affect blood
pressure. This is in contrast to the earlier study in
hypertensive patients, in which metoprolol was the only
prescription medication allowed. We believe these dif-
ferences limit any comparison to our previous study of
hypertensive patients
The 2 most commonly used -blockers in heart fail-
ure, metoprolol and carvedilol, are metabolized by the
polymorphic CYP2D6 enzyme. The CYP2D6 enzyme
accounts for 70% to 80% of the metabolism of meto-
prolol.
35
There have been suggestions that one potential
cause of initial poor tolerability might be excessively
high concentrations in those who are CYP2D6 poor
metabolizers,
36
although to our knowledge this has
never been documented. In this study there were no
pharmacokinetic differences between patients with de-
compensation versus those without decompensation
during the titration period. Thus, consistent with data
regarding hypertension from our laboratory,
37
we do
not have any evidence that S-metoprolol concentration
contributes importantly to the tolerability of metopro-
lol. This nding is perhaps surprising but suggests that
there are substantial differences in the sensitivity to
-blockade (ie, pharmacodynamics) that may be more
important in determining the level of -blockade at a
given concentration.
38
Polymorphisms within the coding region of
2
AR
have been shown to be associated with altered exercise
tolerance among heart failure patients.
39
The Glu27
allele in the
2
AR gene was also shown to be associ-
ated with improved left ventricular ejection fraction
after treatment with carvedilol.
18
However, in this
study we did not observe any differences in either
baseline or end-of-study 6-minute walk distance ac-
cording to the
2
AR genotypes. Similarly, the risk of
decompensated heart failure was similar for polymor-
phisms in
2
AR. There are several explanations for the
differences in ndings between our study and that of
carvedilol pharmacogenetics in heart failure. First,
carvedilol is a nonselective -blocker (and thus blocks
the
2
-receptor) and so would be expected to be more
likely to exhibit a signicant association with the
2
-
receptor polymorphisms than with the
1
-selective
metoprolol. In addition, the end points in the 2 studies
were quite different, with a focus on short-term toler-
ability in this study and a focus on long-term ventric-
ular changes in the carvedilol study. Thus it would
probably be inappropriate to view the ndings from
these 2 studies as inconsistent, because there were
important differences in the study designs.
Although initiation of -blocker therapy can cause
worsening heart failure, analyses from clinical trials
failed to demonstrate any increased risk of cardiac
deterioration as compared with placebo during the ti-
tration period.
3,40
These clinical trial data are likely
explained by those patients who had initial worsening
of heart failure symptoms with -blocker initiation
being counterbalanced by those deriving early benet
from the -blocker. Consensus guidelines recommend
that all patients with systolic heart failure and without
contraindications be treated with a -blocker, and most
can be maintained on some -blocker dose. It is rec-
ognized that there are certain challenges associated
with the initiation of -blocker therapy in heart failure,
and known clinical and demographic risk factors that
are associated with worsening heart failure
31,41
should
be taken into account and titration individualized on the
basis of the clinical condition of the patient. In the
future, genetic information may also be helpful in in-
dividualizing the titration plan.
The authors have no conicts of interest to disclose.
References
1. Hunt SA, Baker DW, Chin MH, Cinquegrani MP, Feld-
man AM, Grancis GS, et al. ACC/AHA guidelines for the
evaluation and management of chronic heart failure in the
adult: executive summary. A report of the American
College of Cardiology/American Heart Association Task
Force on Practice Guidelines (Committee to revise the
1995 Guidelines for the Evaluation and Management of
Heart Failure). J Am Coll Cardiol 2001;38:2101-13.
2. Cleland JGF, Cohen-Solal A, Aguilar JC, Dietz R, East-
augh J, Follath F, et al, for the IMPROVEMENT of Heart
Failure Programme Committees and Investigators. Man-
agement of heart failure in primary care (the IMPROVE-
MENT of Heart Failure Programme): an international
survey. Lancet 2002;360:1631-9.
3. Gottlieb SS, Fischer ML, Kjekshus J, Deedwania P,
Gullestad L, Vitovec J, et al. MERIT-HF Investigators.
Tolerability of beta-blocker initiation and titration in the
Metoprolol CR/XL Randomised Intervention Trial in
Congestive Heart Failure (MERIT-HF). Circulation
2002;105:1182-8.
4. Mason DA, Moore JD, Green SA, Liggett SB. A gain-
of-function polymorphism in a G-protein coupling do-
main of the human
1
-adrenergic receptor. J Biol Chem
1999;274:12670-4.
5. Sandilands AJ, OShaughnessy KM, Brown MJ. Greater
inotropic and cyclic AMP responses evoked by noradren-
aline through Arg389
1
-adrenoceptors versus Gly389
1
-adrenoceptors in isolated human atrial myocardium.
Br J Pharmacol 2003;138:386-92.
6. Molenaar P, Rabnott G, Yang I, Fong KM, Savarimuthu
SM, Li L, et al. Conservation of the cardiostimulant
effects of (-)-norepinephrine across Ser49Gly and
Gly389Arg beta
1
-adrenergic receptor polymorphisms in
human right atrium in vitro. J Am Coll Cardiol 2002;40:
1275-82.
7. Xie HG, Dishy V, Sofowora G, Kim, RB, Landau R,
Smiley RM, et al. Arg
389
Gly
1
-adrenoceptor polymor-
phism varies in frequency among different ethnic groups
but does not alter response in vivo. Pharmacogenetics
2001;11:191-7.
8. Buscher R, Belger H, Eilmes KJ, Tellkamp R, Radke J,
Dhein S, et al. In-vivo studies do not support a major
functional role for the Gly
389
Arg
1
-adrenoceptor poly-
morphism in humans. Pharmacogenetics 2001;11:199-
205.
9. Johnson JA, Zineh I, Puckett BJ, McGorray SP, Yarandi
HN, Pauly DF.
1
-Adrenergic receptor polymorphisms
and antihypertensive response to metoprolol. Clin Phar-
macol Ther 2003;74:44-52.
10. Sofowora GG, Dishy V, Muszkat M, Xie HG, Kim RB,
Harris PA, et al. A common
1
-adrenergic receptor poly-
morphism (Arg389Gly) affects blood pressure response
to -blockade. Clin Pharmacol Ther 2003;73:366-71.
11. Liu J, Liu ZQ, Tan ZR, Chen XP, Wang LS, Zhou G, et
al. Gly389Arg polymorphism of
1
-adrenergic receptor
is associated with the cardiovascular response to meto-
prolol. Clin Pharmacol Ther 2003;74:372-9.
12. Mialet Perez J, Rathz DA, Petrashevskaya NN, Hahn HS,
Wagoner LE, Schwartz A, et al. Beta 1-adrenergic recep-
tor polymorphisms confer differential function and pre-
disposition to heart failure. Nat Med 2003;9:1300-5.
13. Rathz DA, Brown KM, Kramer LA, Liggett SB. Amino
acid polymorphism of the human beta 1-adrenergic re-
ceptor affect agonist-promoted trafcking. J Cardiovasc
Pharmacol 2002;39:155-60.
14. Levin MC, Marullo S, Muntaner O, Andersson B, Mag-
nusson Y. The myocardium-protective Gly-49 variant of
the
1
-adrenergic receptor exhibits constitutive activity
and increased desensitization and down-regulation. J Biol
Chem 2002;277:30429-35.
15. Borjesson M, Magnusson Y, Hjalmarson Y, Andersson
B. A novel polymorphism in the gene coding for the
beta(1)-adrenergic receptor associated with survival in
patients with heart failure. Eur Heart J 2000;21:1853-8.
16. Dishy V, Sofowora GG, Xie HG, Kim RB, Byrne DW,
Stein CM, et al. The effect of common polymorphisms of
the beta-2-adrenergic receptor on agonist-mediated vas-
cular desensitization. N Engl J Med 2001;345:1030-5.
17. Jia H, Hingorani AD, Sharma P, Hopper R, Dickerson C,
Trutwein D, et al. Association of the G(s) alpha gene with
essential hypertension and response to beta-blockade.
Hypertension 1999;34:8-14.
18. Kaye DM, Smirk B, Williams C, Jennings G, Esler M,
Holst D. Beta-adrenoceptor genotype inuences the re-
sponse to carvedilol in patients with congestive heart
failure. Pharmacogenetics 2003;13:379-82.
19. Guyatt GH, Sullivan MJ, Thompson PJ, Fallen EL, Pugs-
ley SO, Taylor DW, et al. The 6-minute walk: a new
measure of exercise capacity in patients with chronic
heart failure. Can Med Assoc J 1985;132:919-23.
20. Rector TS, Cohn JN. Assessment of patient outcome with
the Minnesota Living with Heart Failure questionnaire:
reliability and validity during a randomized, double-
blind, placebo-controlled trial of pimobendan. Pimoben-
dan Multicenter Research Group. Am Heart J 1992;124:
1017-25.
21. Johnson JA, Burlew BS. Metoprolol metabolism via cy-
tochrome P4502D6 in ethnic populations. Drug Metab
Dispos 1996;24:350-5.
22. Terra SG, McGorray SP, Wallace MR, Picoult-Newberg
L, Pepine CJ, Johnson JA. Linkage disequilibrium of
common beta-1 adrenergic receptor polymorphisms [ab-
stract]. Clin Pharmacol Ther 2002;71:P70.
23. Bell PA, Chaturvedi S, Gelfand CA, Huang CY, Koch-
ersperger M, Kopla R, et al. SNPstream UHT: ultra-high
throughput SNP genotyping for pharmacogenomics and
drug discovery. Biotechniques 2002;Suppl:70-2, 74,
76-7.
24. Kimura S, Umeno M, Skoda RC, Meyer UA, Gonzalez
FJ. The human debrisoquine 4-hydroxylase (CYP2D)
locus: sequence and identication of the polymorphic
CYP2D6 gene, a related gene, and a pseudogene. Am J
Hum Genet 1989;45:889-904.
25. Gaedigk A, Gotschall RR, Forbes NS, Simon SD, Leeder
JS. Optimization of cytochrome P4502D6 (CYP2D6)
phenotype assignment using a genotyping algorithm
based on allele frequency data. Pharmacogenetics 1999;
9:669-82.
26. Griese EU, Zanger UM, Brudermanns U, Gaedigk A,
Mikus G, Morike K, et al. Assessment of the predictive
power of genotypes for the in-vivo catalytic function of
CYP2D6 in a German population. Pharmacogenetics
1998;8:15-26.
27. Marcucci KA, Pearce RE, Crespi C, Leeder JS, Gaedigk A.
Characterization of cytochrome P450 2D6.1 (CYP2D6.1),
CYP2D6.2 and CYP2D6.17 activities toward model
CYP2D6 substrates dextromethorphan, bufuralol, and de-
brisoquine. Drug Metab Dispos 2002;30:1-7.
28. Raimundo S, Fischer J, Eichelbaum M, Griese EU,
Schwab M, Zanger UM. Elucidation of the genetic basis
of the common intermediate metabolizer phenotype for
drug oxidation by CYP2D6. Pharmacogenetics 2000;10:
577-81.
29. Oscarson M, Hidestrand M, Johansson I, Ingelman-
Sundberg M. A combination of mutations in the
CYP2D6*17 (CYP2D6Z) allele causes alterations in en-
zyme function. Mol Pharmacol 1997;52:1034-40.
30. Sachse C, Brockmoller J, Bauer S, Roots I. Cytochrome
P450 2D6 variants in a Caucasian population: allele
frequencies and phenotypic consequences. Am J Hum
Genet 1997;60:265-71.
31. Anthonio RL, van Veldhuisen DJ, Breekland A, Crijns
HJGM, van Gilst WH. Beta-blocker titration failure is
independent of severity of heart failure. Am J Cardiol
2000;85:509-12.
32. Mancini DM, Eisen H, Kussmaul W, Mull R, Edmunds
LH Jr, Wilson JR. Value of peak exercise oxygen con-
sumption for optimal timing of cardiac transplantation in
ambulatory patients with heart failure. Circ Res 1991;83:
778-86.
33. Wagoner LE, Craft LL, Zengel P, McGuire N, Rathz DA,
Dorn GW II, et al. Polymorphisms of the
1
-adrenergic
receptor predict exercise capacity in heart failure. Am
Heart J 2002;144:840-6.
34. Tesson F, Charron P, Peuchmaurd M, Nicaud V, Cam-
bien F, Tiret L, et al. Characterization of a unique genetic
variant in the
1
-adrenoceptor gene and evaluation of its
role in idiopathic dilated cardiomyopathy. J Mol Cell
Cardiol 1999;31:1025-32.
35. Regardh CG, Johnsson G. Clinical pharmacokinetics of
metoprolol. Clin Pharmacokinet 1980;5:557-69.
36. Meadowcroft AM, Williamson KM, Patterson JH, Pieper
JA. Pharmacogenetics and heart failure: a convergence
with carvedilol. Pharmacotherapy 1997;17:637-9.
37. Beitelshees AL, Zineh I, Gaedigk A, Leeder JS, Walker
JR, Eberst K, et al. Effect of CYP2D6 phenotype and
metoprolol pharmacokinetics on adverse effects [ab-
38. Sowinski KM, Burlew BS, Johnson JA. Racial differ-
ences in sensitivity to the negative chronotropic effects of
propranolol in healthy males. Clin Pharmacol Ther 1995;
57:678-83.
39. Wagoner LE, Craft LL, Singh B, Suresh DP, Zengel PW,
McGuire N, et al. Polymorphisms of the
2
-adrenergic
receptor determine exercise capacity in patients with
heart failure. Circ Res 2000;86:834-40.
40. Krum H, Roecker EB, Mohacsi P, Rouleau JL, Tendera
M, Coats AJ, et al. Carvedilol Prospective Randomized
Cumulative Survival (COPERNICUS) Study Group. Ef-
fects of initiating carvedilol in patients with severe
chronic heart failure: results from the COPERNICUS
Study. JAMA 2003;289:712-8.
41. Schleman KA, Lindenfeld JA, Lowes BD, Bristow MR,
Ferguson D, Wolfel EE, et al. Predicting response to
carvedilol for the treatment of heart failure: a multivariate
retrospective analysis. J Card Fail 2001;7:4-12.
Polymorphism screening in the cardiac K
channel gene KCNA5

Background: Common deoxyribonucleic acid polymorphisms that modulate normal cardiac electrophysi-
ologic characteristics have previously been identified in long QT syndrome disease genes. In this study we
screened an additional gene encoding the cardiac potassium channel KCNA5 (underlying I
Kur
) in 3 ethnic
groups and evaluated the functional consequences of the variants identified.
Methods: The coding region was screened by single-stranded conformational polymorphism analysis and
direct sequencing, and nonsynonymous variants were studied by patch-clamping transfected Chinese hamster
ovary cells.
Results: Five synonymous and 6 nonsynonymous polymorphisms were found in KCNA5. None of these
polymorphisms was present in greater than 7% of alleles screened or in all 3 ethnic groups. Expression of the
nonsynonymous KCNA5 variants revealed normal gating. However, 2 variants (P532L and R578K, both in
the C-terminus) were resistant to block by the prototypical inhibitor quinidine; the concentration required to
block I
Kur
by 50% (IC
50
) was 8.4 mol/L for wild type versus 54 mol/L for R578K and 133 mol/L for
P532L (both P < .0001, versus wild type).
Conclusion: KCNA5 displays little variability in its coding region. C-terminal KCNA5 variants displayed
near-normal gating but striking resistance to drug block; thus these pharmacogenomic studies have identified
a heretofore-unappreciated role of this region as a modulator of channel sensitivity to drugs. Resistance to
I
Kur
blockers may be genetically determined. (Clin Pharmacol Ther 2005;77:138-44.)
Chantale Simard, PhD, Benoit Drolet, PhD, Ping Yang, PhD, Richard B. Kim, MD,
and Dan M. Roden, MD Nashville, Tenn
Deoxyribonucleic acid (DNA) variants are increas-
ingly well recognized as modulators of human physio-
logic characteristics and drug responses.
1-3
Very rare
variants, occurring in few kindreds, form the basis of
congenital disease, including familial arrhythmias as in
the long QT syndrome (LQTS).
4
At the other end of the
spectrum are polymorphisms, common DNA variants
that may nevertheless lead to variability in disease
phenotypes or in drug responses. One recent example is
the nding of the S1102Y variant (serine to tyrosine at
position 1102) in the cardiac sodium channel gene
SCN5A.
5
A tyrosine residue is present at this position in
about 10% of black persons and appears to confer
susceptibility to arrhythmias in a range of clinical set-
tings, including drug challenge. This variant is only
very rarely identied in other ethnic groups, emphasiz-
ing the need to consider ethnicity in efforts to screen
large numbers of subjects for DNA polymorphisms.
Other LQTS gene polymorphisms reported to modulate
arrhythmia presentations include H558R in SCN5A,
6
K897T in KCNH2 (or HERG),
7-9
D85N
10
and G38S in
KCNE1,
11
and T8A in KCNE2.
12
The ne balance between inward and outward cur-
rents that control the trajectory of repolarization in the
heart is determined not only by LQTS disease genes but
also other genes, including KCNA5, whose expression
results in I
Kur
. I
Kur
is an atrial-specic repolarizing
current that has been proposed as a target for atrial
brillation in humans.
13
Iwasa et al
14
screened the
coding and anking region of KCNA5 in Japanese
patients with LQTS and healthy Japanese control sub-
From the Division of Clinical Pharmacology, Departments of Med-
icine and Pharmacology, Vanderbilt University School of Medi-
cine.
Supported in part by grants from the US Public Health Service
(HL65962). Dr Drolet was the recipient of a postdoctoral fellow-
ship award from the Canadian Institutes of Health Research and the
Heart and Stroke Foundation of Canada. Dr Roden holds the
William Stokes Chair in Experimental Therapeutics, a gift from the
Dai-ichi Corporation.
Received for publication April 16, 2004; accepted Oct 4, 2004.
Reprint requests: Dan M. Roden, MD, Professor of Medicine and
Pharmacology, Director, Division of Clinical Pharmacology,
Vanderbilt University School of Medicine, 532 Medical Research
Bldg I, Nashville, TN 37232.
E-mail: dan.roden@vanderbilt.edu
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.008
138
jects and reported only 1 rare synonymous polymor-
phism (ie, it resulted in no amino acid change).
In this study we screened the coding region of this
gene in a large panel of ethnically dened but otherwise
anonymous individuals and assessed the function of
identied nonsynonymous variants. This approach not
only dened the frequency of polymorphisms in this
gene but also unexpectedly identied new modulators
of channel block by drug.
METHODS
Subjects. The genomic DNA of 95 unrelated white
subjects, 65 unrelated black subjects, and 30 unrelated
Asian subjects, maintained by the Coriell Institute for
Fig 1. Schematic representation of KCNA5 potassium channel with identication and localization
of polymorphisms found in this study by use of single-stranded conformational polymorphism
analysis.
Table I. Allelic frequency of KCNA5 polymorphisms
Gene
Nucleotide
change
Amino acid
change Region
Allele frequency (%)
White Black Asian
KCNA5 344CT A115V Exon 1 0 0.8 0
KCNA5 381CT S127S Exon 1 2.1 1.5 0
KCNA5 570CT N190N Exon 1 0.5 0 0
KCNA5 615GC L205L Exon 1 0 0.8 0
KCNA5 751AT A251T Exon 1 0.5 0 0
KCNA5 919CT P307S Exon 1 0.5 0 0
KCNA5 929TC P310L Exon 1 0.5 0 0
KCNA5 1149CT G383G Exon 1 0 6.9 0
KCNA5 1497GA L499L Exon 1 0 1.5 0
KCNA5 1595CT P532L Exon 1 0 0.8 0
KCNA5 1733GA R578K Exon 1 1.1 0 0
2005;77(3):138-44 Cardiac K
channel polymorphisms 139

DNA polymorphism discovery (Camden, NJ), was
screened.
15
Mutation analysis. The coding region of KCNA5
was amplied by use of polymerase chain reaction
(PCR) (Gene Amp PCR system 9700; PE Applied
Biosystems, Foster City, Calif) with newly designed
primer pairs. Eight primer pairs were used to screen
KCNA5 (1 exon, 1842 base pairs). PCR conditions were
optimized for each primer pair. Amplicons were then
subjected to cold single-stranded conformational poly-
morphism analysis (Novex Thermoow; Novex, Scar-
borough, Ontario, Canada). Silver staining was used to
identify abnormal conformers, each one of which was
then directly sequenced.
Construction and expression of recombinant
KCNA5 channels. Nonsynonymous variants were en-
gineered into the pBK-CMV Phagemid Vector (Strat-
agene, La Jolla, Calif) by use of the QuikChange Site-
Directed Mutagenesis Kit (Stratagene). The identity of
the inserts was reveried by sequencing. Vectors ex-
Fig 2. KCNA5 activating current at 50 mV and tail current at 30 mV under baseline conditions
and after 20-minute exposure to 10 mol/L of quinidine in wild-type (WT) (A), A115V (B), A251T
(C), P307S (D), and P310L channels (E).
Table II. Summary of electrophysiologic data with wild-type KCNA5 and variants (n 4 cells for each)
Act current
(pA/pF)
V
act
(mV)
act fast
(ms)
act slow
(ms)
deact fast
(ms)
deact slow
(ms)
Block by
10 -mol/L
quinidine (%)
Wild type 389 49 5.50 2.13 1.26 0.26 6.58 1.88 14.89 1.19 62.31 6.16 49.6 7.5
A115V 402 19 6.11 3.23 1.25 0.25 7.37 1.20 13.32 2.45 60.85 4.50 47.9 8.3
A251T 413 59 4.83 1.40 1.40 0.26 7.39 0.82 16.29 2.71 73.28 13.17 47.9 6.0
P307S 407 43 6.78 2.08 1.07 0.25 5.58 0.62 14.00 3.41 57.16 2.29 47.8 3.5
P310L 381 48 8.63 2.06 1.91 0.31 6.48 1.56 13.05 6.39 71.77 6.28 54.3 2.4
P532L 400 44 6.01 2.81 1.39 0.46 6.33 1.40 5.03 2.87* 42.95 11.93 7.3 2.1
R578K 406 56 3.59 1.20 1.22 0.28 7.15 1.10 12.60 1.53 47.48 3.88 16.1 2.0
act fast and slow were measured by use of a depolarizing pulse to 50 mV. deact fast and slow were measured from the tail current obtained at 30 mV, after
a depolarizing pulse to 50 mV. Act current, Steady-state current measured at end of depolarizing pulse to 50 mV; V
, voltage at which channels were 50% activated;

, time constant; act, activation; deact, deactivation.
*P .002.
P .00005.
P .003.
P .0001.
140 Simard et al MARCH 2005
pressing wild-type or mutant KCNA5 were transfected
into Chinese hamster ovary (CHO) cells for subsequent
electrophysiologic study; 2 g of KCNA5 complemen-
tary DNA and 6 L of the FuGENE 6 transfection
reagent (Roche Applied Science, Indianapolis, Ind)
were used. Green uorescent protein was coexpressed
to identify transfected cells.
Electrophysiologic studies. I
Kur
generated by ex-
pression of KCNA5 was measured in the whole-cell
conguration of the patch-clamp technique in cells
maintained at room temperature (22C-23C). Cells
were held at 80 mV and pulsed to 30 to 50 mV
for 250 ms, and tail currents were then measured at
30 mV. Pulses were delivered every 15 s throughout
the experiments, so use-dependent block was not ana-
lyzed. The composition of superfusion and internal
pipette solutions was described previously.
16
Statistical analysis. A biexponential function was t
to activation and deactivation time courses to derive
fast and slow time constants (
fast
and
slow
, respec-
tively). Time constants of activation (fast and slow)
were measured after a pulse to 50 mV. Time con-
and after 20-minute exposure to 10 mol/L of quinidine in WT (A), P532L (B), and R578K
channels (C) or exposure to 100 mol/L of quinidine in WT (D), P532L (E), and R578K channels
(F). G, KCNA5 activating current amplitude, measured at 50 mV (n 48 in total), normalized to
baseline, plotted as a function of quinidine concentration and tted to the Hill equation for WT
(circles), R578K (squares), and P532L (triangles). IC
50
, Concentration required to block I
Kur
by
50%; I
Drug
/I
Baseline
, relative current (current under drug compared with baseline current).
2005;77(3):138-44 Cardiac K

stants of deactivation (fast and slow) were measured
from the tail current obtained at 30 mV after a pulse
to 50 mV. The voltage at which channels were 50%
activated (V
) was obtained by tting the Boltzmann

function to current-voltage (I-V) relationships: I
I
max
/(1 exp[(V V
)/k]), where I
max
is the maximal
current and k represents the slope factor. Electrophysi-
ologic data are presented as means SEM. Differences
between wild-type and mutant currents were analyzed
with a 1-way ANOVA. A P value .05 was consid-
ered statistically signicant.
RESULTS
KCNA5. Five synonymous and 6 nonsynonymous
polymorphisms, all with minor allele frequencies lower
than 7%, were found in KCNA5 (Fig 1 and Table I). No
polymorphism was present in all 3 populations.
Electrophysiology. Table II and Fig 2 show that I
Kur
generated by the KCNA5 variants A115V, A251T,
P307S, and P310L were nearly indistinguishable from
wild type. In addition, 10 mol/L of the prototypical
blocker quinidine produced approximately 50% inhibi-
tion of wild-type current at 50 mV, as previously
reported,
17
and the extent of inhibition was similar with
these 4 variants.
The gating of the remaining 2 variants, P532L and
R578K (both in the C-terminus), displayed only minor
differences compared with wild type (Fig 3 and Table
II). P532L did display slightly faster deactivation ki-
netics, with a fast time constant () of 5.0 2.9 ms
versus 14.9 1.2 ms for wild type (P .002). R578K
displayed a 9-mV positive shift in the voltage of half-
activation (V
), as compared with the wild type (P

.003). However, despite the near identity of these cur-
rents with the wild type, both were strikingly resistant
to quinidine: 10 mol/L of the drug reduced R578K
I
Kur
by 16.1% 2.0% and P532L I
Kur
by only 7.3%
2.1% (Table II and Fig 3). The concentrations required
to block I
Kur
by 50% (IC
50
) calculated at 50 mV were
8.4 mol/L for wild type, 54 mol/L for R578K, and
133 mol/L for P532L (Fig 3, G). Fig 4 shows that the
pharmacologic response of wild-type, P532L, and
R578K I
Kur
to the more clinically used blocker (5
mol/L of propafenone)
18
was similar to that with 10
mol/L of quinidine.
DISCUSSION
In this study we screened panels across dened dif-
ferent ethnicities and identied 11 polymorphisms, all
of them uncommon, in the coding region of KCNA5.
Minor allele frequencies were low (7%). An align-
ment of KCNA5 among human, mouse, and rat shows
86% amino acid identity, and no monogenic human
arrhythmia syndrome has been linked to KCNA5.
The 4 KCNA5 variants located in the N-terminus
(A115V), S1 segment (A251T), and S1-S2 linker
(P307S, P310L) generated I
Kur
values whose gating and
drug sensitivity were essentially indistinguishable from
the wild type. The 2 C-terminus variants (P532L and
R578K) also show baseline I
Kur
almost identical to that
of the wild type. However, the sensitivity of these
channels was an order of magnitude (or more) less than
that of the wild type. The effect of propafenone was
similar, indicating that drug resistance was not unique
to quinidine.
A common theme in ion channel pharmacology is
that the S6 segment is a drug-binding/drug-blocking
site common to many channels, including cardiac and
neuronal sodium channels,
19,20
L-type calcium chan-
nels,
21
and KCNH2 (HERG)
22
and KCNQ1
23
channels.
Previous studies have similarly implicated this region
as a key determinant of KCNA5 channel block.
17
One
possibility is that the variant proteins (P532L and
R578K) generate a novel drug-binding site. However,
and after 20-minute exposure to 5 mol/L of propafenone in WT (A), P532L (B), and R578K
channels (C).
previous studies demonstrated that a 57amino acid
deletion of the C-terminus, which eliminates the R578
site, retains wild-type drug sensitivity.
17
R578 is also
included in a consensus target site for protein kinase A
(RRGS),
24,25
although the variant is also predicted to be
a protein kinase A site (RKGS). We considered that
CHO cells might constitutively phosphorylate KCNA5
channels at this site to alter function, but the S580A
mutation, which entirely disrupts the phosphorylation
site, showed I
Kur
gating and drug block similar to those
of the wild type (data not shown).
We have recently reported that the sensitivity of
another cardiac K
channel, KCNQ1, to drug block

was markedly reduced by phosphorylation at C- and
N-terminal sites of the channel protein.
26
Further ex-
periments supported a model in which drug access to its
S6 binding site was impeded by the addition of the
phosphate groups; further experiments will be required
to test the obvious hypothesis that similar structural
constraints underlie the altered drug sensitivity of
P532L and R578K KCNA5 channels.
The data we present have both clinical and mecha-
nistic implications. However, until follow-up studies
delineate these, some caution should be exercised in
interpretation. Our electrophysiology experiments were
conducted at room temperature, and it is possible that
gating changes that were absent at 22C would be-
come apparent at higher temperatures. Similarly,
posttranslational modication of R578K channels
might occur in myocytes even if they were absent in
CHO cells.
In summary, KCNA5 contains few polymorphisms in
its coding region. This gene seems to be well conserved
among different species, likely reecting an important
physiologic role. Expression of KCNA5 variants
showed nearly identical baseline current when com-
pared with that of the wild type. Moreover, they also
responded to quinidine in the same way, except for
C-terminal variants P532L and R578K, which showed
marked resistance to drug block. Thus these pharma-
cogenomic studies provide support for a model in
which regulated drug access to a binding site near the
pore contributes to variability in drug block. Further-
more, they indicate that a small percentage of patients
receiving quinidine and propafenone for atrial arrhyth-
mias
13,27
may have drug resistance as a result of these
genetic variants.
We thank Donna Choate for technical assistance.
The authors and their institution have no conicts of interest in
relation to this report.
References
1. McLeod HL, Evans WE. Pharmacogenomics: unlocking
the human genome for better drug therapy. Annu Rev
Pharmacol Toxicol 2001;41:101-21.
2. Small KM, McGraw DW, Liggett SB. Pharmacology and
physiology of human adrenergic receptor polymor-
phisms. Annu Rev Pharmacol Toxicol 2003;43:381-411.
3. Roden DM, George AL Jr. The genetic basis of variabil-
ity in drug responses. Nat Rev Drug Discov 2002;1:37-
44.
4. Keating MT, Sanguinetti MC. Molecular and cellular
mechanisms of cardiac arrhythmias. Cell 2001;104:569-
80.
5. Splawski I, Timothy KW, Tateyama M, Clancy CE,
Malhotra A, Beggs AH, et al. Variant of SCN5A sodium
channel implicated in risk of cardiac arrhythmia. Science
2002;297:1333-6.
6. Viswanathan PC, Benson DW, Balser JR. A common
SCN5A polymorphism modulates the biophysical effects
of an SCN5A mutation. J Clin Invest 2003;111:341-6.
7. Pietila E, Fodstad H, Niskasaari E, Laitinen PPJ, Swan H,
Savolainen M, et al. Association between HERG K897T
polymorphism and QT interval in middle-aged Finnish
women. J Am Coll Cardiol 2002;40:511-4.
8. Bezzina CR, Verkerk AO, Busjahn A, Jeron A, Erdmann
J, Koopmann TT, et al. A common polymorphism in
KCNH2 (HERG) hastens cardiac repolarization. Cardio-
vasc Res 2003;59:27-36.
9. Paavonen KJ, Chapman H, Laitinen PJ, Fodstad H,
Piippo K, Swan H, et al. Functional characterization of
the common amino acid 897 polymorphism of the cardiac
potassium channel KCNH2 (HERG). Cardiovasc Res
2003;59:603-11.
10. Wei J, Yang IC, Tapper AR. KCNE1 polymorphism
confers risk of drug-induced long QT syndrome by alter-
ing kinetic properties of IKs potassium channel [ab-
stract]. Circulation 1999;100:I-145.
11. Lai LP, Su MJ, Yeh HM, Lin JL, Chiang FT, Hwang JJ,
et al. Association of the human minK gene 38G allele
with atrial brillation: evidence of possible genetic con-
trol on the pathogenesis of atrial brillation. Am Heart J
2002;144:485-90.
12. Sesti F, Abbott GW, Wei J, Murray KT, Saksena S,
Schwartz PJ, et al. A common polymorphism associated
with antibiotic-induced cardiac arrhythmia. Proc Natl
Acad Sci U S A 2000;97:10613-8.
13. Nattel S. New ideas about atrial brillation 50 years on.
Nature 2002;415:219-26.
14. Iwasa H, Kurabayashi M, Nagai R, Nakamura Y, Tanaka
T. Twenty single-nucleotide polymorphisms in four
genes encoding cardiac ion channels. J Hum Genet 2002;
47:208-12.
15. Collins FS, Brooks LD, Chakravarti A. A DNA polymor-
phism discovery resource for research on human genetic
variation. Genome Res 1998;8:1229-31.
2005;77(3):138-44 Cardiac K

16. Yang T, Snyders D, Roden DM. Drug block of I(kr):
model systems and relevance to human arrhythmias.
J Cardiovasc Pharmacol 2001;38:737-44.
17. Yeola SW, Rich TC, Uebele VN, Tamkun MM, Snyders
DJ. Molecular analysis of a binding site for quinidine in
a human cardiac delayed rectier K channel. Role of
S6 in antiarrhythmic drug binding. Circ Res 1996;78:
1105-14.
18. Franqueza L, Valenzuela C, Delpon E, Longobardo M,
Caballero R, Tamargo J. Effects of propafenone and
5-hydroxy-propafenone on hKv1.5 channels. Br J Phar-
macol 1998;125:969-78.
19. Ragsdale DS, McPhee JC, Scheuer T, Catterall WA.
Molecular determinants of state-dependent block of
Na channels by local anesthetics. Science 1994;265:
1724-8.
20. Weiser T, Qu Y, Catterall WA, Scheuer T. Differential
interaction of R-mexiletine with the local anesthetic re-
ceptor site on brain and heart sodium channel -subunits.
Mol Pharmacol 1999;56:1238-44.
21. Schuster A, Lacinova L, Klugbauer N, Ito H, Birnbaumer
L, Hofmann F. The IVS6 segment of the L-type calcium
channel is critical for the action of dihydropyridines and
phenylalkylamines. EMBO J 1996;15:2365-70.
22. Mitcheson JS, Chen J, Lin M, Culberson C, Sangui-
netti MC. A structural basis for drug-induced long QT
syndrome. Proc Natl Acad Sci U S A 2000;97:12329-
33.
23. Seebohm G, Chen J, Strutz N, Culberson C, Lerche C,
Sanguinetti MC. Molecular determinants of KCNQ1
channel block by a benzodiazepine. Mol Pharmacol
2003;64:70-7.
24. Uebele VN, Yeola SW, Snyders DJ, Tamkun MM. De-
letion of highly conserved C-terminal sequences in the
Kv1 K channel sub-family does not prevent expression
of currents with wild-type characteristics. FEBS Lett
1994;340:104-8.
25. Blom N, Gammeltoft S, Brunak S. Sequence and
structure-based prediction of eukaryotic protein phos-
phorylation sites. J Mol Biol 1999;294:1351-62.
26. Yang T, Kanki H, Roden DM. Phosphorylation of IKs
channel complex inhibits drug block. Novel mechanism
underlying variable antiarrhythmic drug actions. Circu-
lation 2003;108:132-4.
27. Dorian P. Antiarrhythmic drug therapy of atrial brilla-
tion: focus on new agents. J Cardiovasc Pharmacol Ther
2003;8(Suppl 1):S27-31.
PERSPECTIVES IN CLINICAL
PHARMACOLOGY
Implications of CYP2A6 genetic variation
for smoking behaviors and nicotine
dependence
Nicotine is the primary addictive compound in tobacco smoke. In this review we summarize nicotine
dependence and the genetics of smoking in brief before focusing on cytochrome P450 (CYP) 2A6. In humans
nicotine is mainly inactivated to cotinine and CYP2A6 mediates approximately 90% of this conversion. Some,
but not all, studies suggest that genetic variation in CYP2A6 may play a role in smoking. We review some of
the recent findings on the influence of CYP2A6 genetic polymorphisms on nicotine kinetics, smoking
behaviors, and how the gene appears to exert differential effects during various stages of smoking (eg,
initiation, conversion to dependence, amount smoked during dependence, and quitting). These new findings
will be put in the context of the discrepancies found in the literature. Implications of these recent findings on
current and novel treatment approaches for smoking cessation and tobacco-related lung cancer will also be
discussed. (Clin Pharmacol Ther 2005;77:145-58.)
Viba Malaiyandi, BSc, Edward M. Sellers, MD, PhD, and
Rachel F. Tyndale, PhD Toronto, Ontario, Canada
NICOTINE/TOBACCO DEPENDENCE
Nicotine is the major substance in tobacco that es-
tablishes and maintains dependence on cigarette smok-
ing.
1,2
Like other psychostimulants, such as cocaine
and amphetamines, nicotine modulates dopamine activ-
ity in the midbrain,
3-5
specically, the mesocorticolim-
bic system,
6,7
which contributes to the development
and maintenance of rewarding behaviors such as smok-
ing.
7,8
In addition, there are many factors that contrib-
ute to smoking, including hedonic, cognitive, social,
and contextual inuences.
9
Repeated exposure to nic-
otine causes neurobiologic changes, such as the up-
regulation of neuronal receptors,
10-12
producing toler-
ance, which is indicated by the appearance of
withdrawal symptoms on abrupt abstinence from ciga-
rettes or nicotine administration.
13,14
Nicotine inhaled from cigarette smoke enters the
lungs, is rapidly absorbed by the pulmonary capillaries,
and reaches the brain within 10 to 19 seconds after
inhalation.
15
Nicotine has a distribution half-life of 15
to 20 minutes and an elimination half-life of 1 to 2
hours.
16
Thus the drugs effects are experienced within
minutes of smoking, and smokers must smoke regularly
because of nicotines short half-life.
16,17
Cigarette
smoking presents an ideal drug delivery system for the
establishment of behavioral reinforcement leading to
dependence.
15-17
From the Departments of Pharmacology and Medicine and Psychia-
try, University of Toronto, and Centre for Addiction and Mental
Health.
This work was supported by Canadian Institutes of Health Research
MOP-53248 and by the Centre for Addiction and Mental Health.
Dr Tyndale is the recipient of a Canadian Research Chair in
Pharmacogenetics. V.M. receives funding from the Canadian In-
stitutes of Health Research Strategic Training Program in Tobacco
Research.
Received for publication Aug 3, 2004; accepted Oct 21, 2004.
Reprint requests: Rachel F. Tyndale, PhD, Department of Pharma-
cology, Medical Sciences Building, 1 Kings College Circle, Uni-
versity of Toronto, Toronto, Ontario, Canada M5S 1A8.
E-mail: r.tyndale@utoronto.ca
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.011
145
THE GENETICS OF SMOKING AND
NICOTINE DEPENDENCE
Smoking is a complex behavior inuenced by ge-
netic and environmental factors. This behavior can be
divided into various stages such as initiation, experi-
mentation/sporadic use, regular/daily use, dependence,
cessation, and relapse.
18
Biologic and psychosocial re-
inforcing effects experienced by individuals likely de-
termine the progression through these stages.
19
Individ-
uals with particular genetic predispositions and those
who are exposed to certain environmental inuences
may be at greater risk for dependence on nicotine.
19
A
review of data from twin, family, and adoption studies,
which assess the contribution of genes and environment
to a particular phenotype, estimates that the variance in
risk for initiating smoking is approximately 60% ge-
netic.
20,21
Once smoking is initiated, the variance in
liability for progressing to nicotine dependence is ap-
proximately 70% attributable to genetic sources.
20
In
addition, studies show a substantial genetic component
to the age of smoking onset, the amount smoked, and
smoking persistence.
22-24
It has also been shown that
smoking shares genetic risk with comorbid disorders.
For example, a study in male twins demonstrated
shared genetic risk between smoking experimentation
and low-frequency depressive symptoms.
25
Likewise, a
study in schizophrenic patients comparing smokers and
nonsmokers suggests an association between a genetic
polymorphism in the
7
nicotinic receptor subunit gene
and smoking status in this population.
26
A common
evoked auditory decit (P50) among schizophrenic pa-
tients,
27
which can be normalized by nicotine,
28
has
been linked to the
7
nicotinic receptor subunit gene
and may be one explanation for the association of this
gene variant and a high smoking prevalence in this
group.
29
Together, these studies provide convincing
evidence that smoking behavior and nicotine depen-
dence are strongly inuenced by genetics, and this has
prompted researchers to attempt to identify the specic
genes involved.
Many genes have been investigated, and these can be
thought of as clustering into the following 2 groups:
those involved in the general aspects of dependence
30
and those more specic to nicotine dependence.
21
Sim-
ilar to many behavioral disorders, nicotine dependence
is polygenic; genome scans suggest that genes on sev-
eral chromosomes are involved.
31-33
Although it is eas-
ier to think of candidate risk genes as belonging to
distinct groups, it is important to remember that they
are functionally overlapping; each contributes to a
small proportion of this complex disorder.
21
The focus
of a number of studies has been on neurotransmitter
systems involving dopamine, -aminobutyric acid, se-
rotonin, and glutamate because these modulate the re-
inforcing and hedonic effects experienced with several
addictive behaviors such as polydrug abuse and gam-
bling.
8,34,35
Genetic variation within these pathways
and their roles in smoking have been previously
reviewed.
19,21,31,36-38
The primary addictive substance in tobacco smoke is
nicotine. High doses of transdermal nicotine have been
shown to reduce the amount of nicotine extracted per
cigarette and the amount smoked.
39
Genes that can
transduce nicotines effects into physiologic experi-
ences (eg, acetylcholinergic receptors) and those that
can inactivate nicotine by removing it from the body
(eg, cytochrome P450 [CYP] enzymes) are logical can-
didates for altering smoking behaviors. Variation in
these genes may directly and indirectly affect nicotines
reinforcing and aversive properties and may alter an
individuals risk for nicotine dependence.
38
Nicotinic acetylcholine receptors present on dopami-
nergic neurons in the midbrain directly mediate nico-
tines reinforcing effects.
3-5,40
Nicotinic acetylcholine
receptors are pentameric channels consisting of combi-
nations of
3-7
and
2-4
subunits.
41
Differences in
smoking initiation, susceptibility to nicotine depen-
dence, and quitting may be a result of differential
sensitivity of nicotinic acetylcholine receptors to nico-
tine.
17,31,42
These differences may in part be attributed
to the combination of subunits that are present. For
example, knockout studies in mice indicate that nico-
tinic acetylcholine receptors containing
2
subunits me-
diate the reinforcing properties of nicotine
43
whereas
4
knockout mice demonstrate increased extracellular
levels of dopamine.
44
Other studies have suggested that
nicotinic acetylcholine receptors containing only
7
subunits are involved in nicotine-induced dopamine
release,
45
mediate effects of the nicotine withdrawal
syndrome,
46
and are associated with risk for smoking.
26
Genetic differences within subunits may contribute to
altered smoking risk, although these effects are still
unclear.
47,48
Smokers modulate their smoking to maintain brain
nicotine levels within a certain concentration range,
49
and factors that alter nicotine clearance affect smoking
behavior.
50
It has been established in a previous study
that craving for cigarettes increases as levels of plasma
nicotine and breath carbon monoxide, an indicator of
smoke exposure, decline.
51
In this study, participants
were asked to smoke as desired until they were satis-
ed. Individuals who eliminated nicotine rapidly were
less likely to achieve clearly low craving scores even
after smoking freely. Taken together, these ndings
146 Malaiyandi, Sellers, and Tyndale MARCH 2005
suggest that nicotine likely drives the urge to
smoke
51
and that this urge may be partly affected by
differing rates of nicotine metabolism. Genetic varia-
tion in CYP2A6, the main nicotine-inactivating en-
zyme, will be the focus of the remainder of this review.
INTERINDIVIDUAL AND INTERETHNIC
VARIATION IN NICOTINE METABOLISM
In humans approximately 70% to 80% of nicotine is
metabolized to cotinine
52
and roughly 90% of this
conversion is mediated by CYP2A6.
53,54
Cotinine is
subsequently oxidized, a step specically catalyzed by
CYP2A6 in vivo and in vitro, to form
3-hydroxycotinine.
55,56
CYP2A6 exhibits a narrow
substrate spectrum mediating the metabolism of cou-
marin, a few pharmaceutical compounds (eg, halo-
thane, valproic acid, disulram, and the antineoplastic
agent tegafur), and tobacco-specic nitrosamines.
57,58
Substantial variation in CYP2A6 activity and protein
levels exists. In vivo studies using coumarin
59-61
and
nicotine
52,55,62
as substrates indicate large interindi-
vidual variability in CYP2A6 activity. In addition, in
vitro kinetic studies using human liver microsomes
revealed interindividual variability in cotinine-to-
nicotine metabolic ratios.
53,54
Considerable interethnic
variability in coumarin
63
and nicotine
64
metabolism has
also been described. The presence of environmental or
therapeutic inducers and liver abnormality may contrib-
ute to these differences
58
; however, this variability has
been largely attributed to genetic polymorphisms in the
CYP2A6 gene.
65,66
Presently, there are 26 known variants of the
CYP2A6 gene that may affect enzyme activity or gene
regulation.
67
Table I summarizes the variants for which
functional consequences have been characterized in vitro or
in vivo. A number of alleles with single nucleotide polymor-
phisms in regulatory regions (CYP2A6*1B, CYP2A6*1C,
CYP2A6*1D, CYP2A6*1E, and CYP2A6*1H, CYP2A6*1J)
and coding regions (CYP2A6*1F, CYP2A6*1G, and
CYP2A6*13-*16) have also been identied; however,
little is known about their frequencies or effect on
function.
The frequencies of CYP2A6 alleles vary among
ethnicities (Table II). Generally, variant alleles re-
sulting in reduced enzyme activity are more common
among Asian populations (Chinese, Japanese, and
Korean) compared with Caucasian and African North
American populations (Fig 1).
64,68
In addition, cer-
Table I. Description of CYP2A6 alleles with known in vivo or in vitro functional consequences
CYP2A6
allele Nucleotide and structural changes Functional consequence Reference No.
*1A Normal activity
*8 6600GT (R485L) Normal in vivo activity 121, 122, 122a
*1x2 Reciprocal of *4A or *4D (gene duplication) Increased nicotine clearance in vivo 73, 74
*2 1799TA (L160H) No in vivo or in vitro activity 123, 124
*4A Unequal crossover (3-UTR) with CYP2A7
(gene deletion)
No in vivo activity 125-127
*4B Unequal crossover (distal 3-UTR) with
CYP2A7 (gene deletion)
No in vivo activity 128
*4D Unequal crossover (intron 8) with CYP2A7
(gene deletion)
No in vivo activity 74, 129
*5 6582GT (G479L) No in vivo or in vitro activity 129
*6 1703GA (R128Q) Reduced in vitro activity 130
*7 6558TC (I471T) Reduced in vivo and in vitro activity toward
nicotine but not coumarin
121, 122, 122a
*10 6558TC 6600GT (I471T R485L) Absent or reduced in vivo activity 122, 122a, 131
*11 3391TC (S224P) Reduced in vivo and in vitro activity toward
tegafur, an antineoplastic agent
132
*12 Hybrid gene CYP2A7: 5 regulatory to exon 2
CYP2A6: exon 3-3-UTR (10amino acid
substitution)
Reduced in vivo and in vitro activity toward
coumarin
133
*9 48TG (TATATAGA box) Decreased enzyme expression resulting in
reduced activity in vivo and in vitro
134-136
UTR, Untranslated region.
2005;77(3):145-58 CYP2A6 genetic variation and smoking 147
tain alleles vary substantially among Asian popula-
tions. For instance, the CYP2A6*4 allele frequency is
signicantly higher among Japanese subjects com-
pared with Chinese subjects.
68
Interethnic differ-
ences in CYP2A6-mediated cotinine formation
within genotype groups have also been described;
Korean subjects show substantially greater cotinine-
to-nicotine metabolic ratios compared with Japanese
subjects.
64
Although the frequency of variant alleles
is greater in Asian populations, smoking prevalence
is also much higher in this group.
69,70
However, even
among this population with a high smoking preva-
lence, the inuence of CYP2A6 on smoking is seen.
For example, Japanese poor metabolizers of CYP2A6
are less often found to be smokers, smoke less per
day if regular smokers, and have a reduced risk for
lung cancer.
71
However, this gene is only one part of
the genetic inuence, and in addition to genetic dif-
Fig 1. Frequencies reported as percentages by predicted activity. Calculations of genotype fre-
quencies are based on mean allelic frequencies and Hardy-Weinberg estimations of homozygotes
and heterozygotes. No CYP2A6 activity, Homozygous for CYP2A6*2, CYP2A6*4, and CYP2A6*5
inactive alleles; 50% CYP2A6 activity, heterozygous for inactive alleles and homozygous for any
combination of reduced activity alleles CYP2A6*6, CYP2A6*7, CYP2A6*9, CYP2A6*10, and
CYP2A6*12; 75% CYP2A6 activity, heterozygous for reduced activity alleles; Normal (100%)
CYP2A6 activity, remaining percentage of those not included in any other group. The African North
American ethnic group includes both African Americans and African Canadians.
Table II. Interethnic variation in CYP2A6 allele frequencies
CYP2A6
allele
Allele frequency
African North
American Caucasian Chinese Japanese Korean
*1x2 1.7%
74
0.5%
131
*2 0.3%-1.1%
68,137
1.1%-3.0%
68,74,124,135,138
0%-0.7%
68,127,138
0%
64,68
0%
64
*4 1.9%
68
0.5%-1.2%
68,74,127
6.7%-15.1%
68,127
20.1%-24.2%
64,68
11.0%
64
*5 0%
68
0%-0.1%
68,129
0%-0.5%
68,129
0%
64,68
0%-0.5%
64
*6 0%-0.4%
68,130,131
0%
131
*7 0%
68
0%-0.3%
68,122
2.2%-9.8%
68,122,122a
6.3%-12.5%
64,68,122a
3.6%
64
*8 0%
68
0%-0.1%
68,122
0%-3.6%
68,122,122a
0%-2.2%
64,68,122a
1.4%
64
*9 7.1%
68
5.2%-7.2%
68,135
15.6%-15.7%
68,135
20.3%-21.3%
68,136
22.3%
136
*10 0%
68
0%
68,122
0.4%-4.0%
68,122,122a
1.1%-3.2%
64,68,122a
0.5%
64
*12 0.4%
68
2.0%-2.2%
68,133
0%
68,133
0.8%
68
The African North American ethnic group includes African Americans and African Canadians.
ferences, environmental and cultural differences may
be inuential in driving smoking behavior. This may
be specic to populations such as the Chinese and
Japanese, in which smoking is socially and culturally
accepted and encouraged in men.
70
IMPACT OF CYP2A6 GENETIC VARIATION
ON NICOTINE DEPENDENCE AND SMOKING
BEHAVIORS
Susceptibility to nicotine dependence: Is slow me-
tabolism analogous to innate sensitivity? Several lines
of evidence indicate that smokers titrate their cigarette
consumption to maintain steady levels of nicotine in the
brain. It has been demonstrated that factors inuencing
nicotine plasma levels, by either intake (eg, high- ver-
sus low-yield cigarettes) or removal (eg, acidication
of urine to increase nicotine clearance), affect smoking
behaviors.
50,72
Variability in the rates of nicotine me-
tabolism, as a result of CYP2A6 genetic polymor-
phisms, alter nicotine plasma levels.
66,73,74
It was orig-
inally hypothesized by our laboratory that CYP2A6
genetic variation would affect smoking behaviors such
as the amount smoked and, secondarily, the risk for
nicotine dependence.
75
Specically, we hypothesized
that nicotine-dependent slow inactivators, individuals
with at least 1 inactive allele (CYP2A6*2 or
CYP2A6*4), would smoke fewer cigarettes and would
smoke less often (longer interval between cigarettes) to
avoid withdrawal symptoms compared with normal
inactivators.
76
In contrast, when learning to smoke,
slow nicotine inactivators might have prolonged or
elevated nicotine levels, which could alter the aversive
effects of nicotine. Consequently, we expected slow
inactivators to be at lower risk for dependence.
76
It is equally plausible that prolonged levels of brain
nicotine might increase the risk for dependence. The
sensitivity model proposes that those who have an
innate sensitivity to nicotine experience greater ef-
fects of nicotine, whether pleasurable or aversive, and
are more likely to progress to dependent smoking.
77,78
A retrospective study found that women who became
highly dependent smokers experienced greater pleasur-
able effects (rush, buzz, and relaxation) on initial
tobacco exposure compared with never-smokers.
79
These ndings are consistent with a prospective study
in youth showing that increased feelings of relaxation,
dizziness, and nausea after the rst nicotine exposure
may be risk factors for progression to nicotine depen-
dence.
80
Although individual pharmacokinetic effects
on nicotine levels and resulting nicotine sensitivity are
not well understood,
77
in this context one could think of
slow inactivators as being analogous to those who are
nicotine sensitive; they may experience greater nicotine
effects and thus may be more susceptible to nicotine
dependence.
CYP2A6 genetic variation and smoking: Retrospec-
tive and prospective studies. A number of studies in
adults have demonstrated that CYP2A6 genetic varia-
tion, causing reduced or absent enzyme activity, is
associated with a reduced risk for smoking, lower
amount smoked, altered smoking intensity, and in-
creased quitting
68,74,81-84
; however, not all studies are
in agreement with these ndings.
85-87
A meta-analysis
reviewing several studies on CYP2A6 and smoking
found no association between individuals with variant
alleles and smoking status or amount smoked.
88
As the
authors mention, the discordance in the literature may
be a result of various aspects of study design such as
broad denitions of smokers (ever-smoker, never-
smoker, former smoker, nonsmoker), population strat-
ication, genotype groups, and genotyping assays.
A recent study from our laboratory has addressed
some of these factors in an attempt to clarify the asso-
ciation between CYP2A6 genetic variation and smok-
ing.
68
Substantial differences in allele frequencies were
observed among ethnic groups. To minimize popula-
tion stratication, the smoking study population was
restricted to adults of Caucasian ancestry determined by
participants reports of having at least 3 grandparents of
Caucasian ethnicity. In this study the effect of the gene
on smoking maintenance, not initiation, was investi-
gated. To assess the effect of the gene (CYP2A6) on the
behavior (smoking), at least a single exposure to the
drug (nicotine) is necessary. Therefore in this study the
nonsmoker group was restricted to individuals who had
never become current smokers but had smoked 1 to 99
cigarettes in their lifetime. Smokers, those currently
smoking (100 cigarettes/lifetime), were stratied by
dependence. Those who met criteria for tobacco depen-
dence according to the American Psychiatric Associa-
tion Diagnostic and Statistical Manual of Mental Dis-
orders, Fourth Edition, were dened as dependent
smokers. Only variant allele products known to be
inactive or to have decreased activity with frequencies
greater than 1% in Caucasians were studied. Genotype
groups were segregated on the basis of combined the-
oretic and demonstrated CYP2A6 activity as slow in-
activators (50% activity), intermediate inactivators
(approximately 75% activity), and normal inactivators
(full activity).
Applying these specications to the study design and
data analysis, we found that several aspects of smoking
behavior were inuenced by CYP2A6 genotype. Slow
inactivators were signicantly more prevalent among
nonsmokers than smokers independent of tobacco de-
pendence (odds ratio, 0.52 [95% condence interval
(CI), 0.30-0.91]; P .027). This indicates that CYP2A6
slow inactivators of nicotine are at lower risk for being
current smokers regardless of whether they are
nicotine-dependent. CYP2A6 also affected the amount
smoked per day. Nicotine-dependent slow inactivators
smoked fewer cigarettes per day compared with normal
inactivators (21.3 cigarettes per day versus 28.2 ciga-
rettes per day, P .003), but this gene effect was not
found in individuals who were current nondependent
smokers. This highlights the importance of assessing
dependence because the gene effect on the amount
smoked may not be detected in groups in which there is
a greater proportion of nondependent smokers.
68
This
nding is consistent with our original hypothesis on the
amount smoked.
74,75
We expected that the titration of
plasma and brain nicotine levels by cigarette number
would occur only in dependent smokers.
49,76
Are slow nicotine-inactivator genotypes less preva-
lent among smokers because of a lower risk of becom-
ing a smoker or because these individuals quit smoking
sooner? In our adult study a trend toward a shorter
duration of smoking was observed among current
smokers who were slow inactivators. We found that the
proportion of slow inactivators was greatest among
nonsmokers, followed by short-duration smokers (10
years), and was lowest among long-duration smokers
(30 years), suggesting that smokers with slow inac-
tivator genotypes may be quitting sooner. This inter-
pretation is supported by studies assessing the effect of
CYP2A6 genotype on quitting. A smoking cessation
study conducted in African Americans showed that
slow inactivators had substantially increased cessation
rates (odds ratio, 3.8 [95% CI, 1.4-11.1]) compared
with normal inactivators.
89
In addition, Caucasians
having a CYP2A6*2 allele were 1.75 (95% CI, 1.17-
2.61) times more likely to quit compared with those
without the CYP2A6*2 allele.
90
These studies provide
support that smokers with a slow inactivator genotype
may be quitting sooner and, consequently, are under-
represented in smoker groups, particularly as the dura-
tion of smoking increases. However, there are still slow
inactivators who remain among long-duration smokers;
thus it will be important to determine whether these
slow inactivators continue smoking because they are
resistant to quitting or whether they have not yet made
serious attempts to stop smoking. A recent assessment
in a nicotine replacement clinical trial indicates that a
very low percentage of slow metabolizers enrolled (Ler-
man C, Malaiyandi V, Tyndale RF, unpublished data,
September 2004); this suggests that slow metabolizers
may be quitting before this time or may not be seeking
treatment. The apparent effect of CYP2A6 on increas-
ing quitting, particularly among shorter-duration smok-
ers, may contribute to the lack of agreement in pub-
lished studies between CYP2A6 slow inactivators and
decreased risk for smoking. A signicantly greater
prevalence of slow inactivator compared with normal
inactivator genotypes may not be apparent if the study
population is composed of smokers who have only been
smoking for a short duration (eg, 0-15 years).
To further investigate the role of CYP2A6 in smoking
and nicotine dependence, we examined the effect of
genetic variation on becoming a smoker. Retrospective
studies are adequate for examining the risk for current
or former smoking. However, they are weaker for as-
sessing the effect of CYP2A6 genetic variation on
smoking initiation and the risk for development of
tobacco dependence. Adults may have difculty accu-
rately recalling all aspects of their smoking histories,
especially during initiation, such as their experiences of
nicotines effects when rst learning to smoke. A pro-
spective study conducted in Caucasian adolescents fur-
ther elucidates the role of CYP2A6 genetic variation on
the risk for acquisition of nicotine dependence.
91
This
study followed an adolescent cohort to assess whether
CYP2A6 genetic variants altered the risk for becoming
nicotine-dependent from the point of rst inhalation.
The incidence of conversion to tobacco dependence
was approximately 3 times greater (hazard ratio, 2.8
[95% CI, 1.3-6.3]) among slow inactivators.
91
Although this nding is inconsistent with our origi-
nal hypothesis,
76
it is consistent with our ndings in
adults, in which a younger age of rst smoking was
reported among slow inactivators compared with nor-
mal inactivators (13 years versus 14.2 years, P
.03).
68
With poor recall accounted for, this result may
actually represent the age at which regular smoking was
established and appears to be consistent with a greater
occurrence of becoming tobacco-dependent among ad-
olescent slow inactivators.
91
Furthermore, in the ado-
lescent study it was observed that the prevalence of
smoking among parents of slow inactivators was lower
compared with parents of normal inactivators. The re-
duced frequency of smoking among parents of slow
inactivators, the presumed source of slow inactivator
alleles, is consistent with the reduced frequency of
smoking observed among adult slow inactivators. This
observation is also consistent with the reduced fre-
quency of slow inactivators among smokers as the
duration of smoking increases.
68
An increased incidence of conversion to tobacco
dependence among adolescent slow inactivators after
rst inhalation may also be consistent with the sensi-
tivity model. Slow rates of nicotine metabolism, pre-
dicted by CYP2A6 genotype, may cause more intense
and prolonged exposure of the brain to nicotine, which
may be analogous to increased nicotine sensitivity,
resulting in greater vulnerability to dependence. Ado-
lescent slow inactivators reported no differences in
symptoms of dizziness or nausea compared with nor-
mal inactivators.
91
It appears that when adolescents are
learning to smoke, slow nicotine inactivation, rather
than being protective against nicotine dependence by
enhancing the drugs aversive effects, may increase
susceptibility to dependence by greatly enhancing nic-
otines reinforcing effects. The sensitivity model sug-
gests that increased reinforcement, but not necessarily
increased aversive effects of nicotine, is predictive of
dependent smoking.
77-79
Although adolescent slow inactivators are becoming
dependent on nicotine sooner, cigarette consumption
among dependent adolescent smokers is markedly
lower in slow inactivators compared with normal inac-
tivators.
91
This nding is consistent with several studies
in adults
68,74,82
and supports the proposal that the gene
effect on the amount smoked is likely occurring only
among dependent smokers, whether in adults or ado-
lescents. In addition, the amount smoked by dependent
intermediate inactivators fell between the amounts of
slow and normal inactivators.
91
These adolescent data
parallel those of our adult study implicating a gene-
dose effect among dependent smokers and imply that
relatively small changes in nicotine metabolism may
affect the amount smoked.
68
Also, in support of the
gene effect on cigarette consumption, a kinetic study in
adults found that the metabolic ratio of
3-hydroxycotinine to cotinine, representing CYP2A6
activity, was signicantly correlated with the number of
cigarettes smoked per day, suggesting that the rate of
CYP2A6-mediated nicotine inactivation is a determi-
nant of the amount smoked.
92
Moreover, the adolescent
study suggests that among novice smokers slow nico-
tine inactivators are at greater risk for nicotine depen-
dence even though they are smoking less than normal
inactivators.
91
Temporal effect of CYP2A6 genetic variation:
Stages of smoking? Several studies in animals and
humans suggest that the adolescent brain is much more
sensitive to nicotine than the adult brain.
80,93-97
It has
been shown that adolescents show symptoms of nico-
tine dependence even if smoking 1 cigarette per
month.
96
Studies in the adolescent rat demonstrated that
even brief and low exposure to nicotine, similar to that
seen in adolescent smokers, results in the up-regulation
of nicotinic acetylcholine receptors in brain regions
associated with nicotine dependence, long-term reduc-
tions in cholinergic synaptic activity,
93
and increased
intravenous nicotine self-administration.
94
Taken to-
gether, these studies suggest that the adolescent brain
may be more susceptible to nicotine dependence as a
result of increased susceptibility to nicotines molecular
and behavioral effects. It is interesting that the potential
age-dependent effect of the CYP2A6 gene is consistent
with the period during which the brain appears to be
most sensitive to nicotines neuroplastic and behavioral
effects. It could be postulated that prolonged elevated
brain nicotine levels, as a result of CYP2A6 slow inac-
tivator alleles, may make adolescent slow inactivators
more vulnerable to nicotine dependence compared with
normal inactivators.
Collectively, these recent studies suggest that several
aspects of smoking are inuenced by CYP2A6 genetic
variation.
68,89,91
More specically, these effects appear
to vary with the point in smoking history and are
inuenced by factors such as status of nicotine depen-
dence and duration of smoking (Fig 2). Our studies
suggest that, during adolescence, having a slow inacti-
vator genotype increases the risk for nicotine depen-
dence while still reducing the amount smoked. As these
dependent novice smokers become adults and more
regular smoking occurs, it appears that slow nicotine
inactivators smoke less and for fewer years compared
with normal nicotine inactivators. Reduced cigarette
consumption has been associated with better cessation
outcomes in adults
98,99
; thus CYP2A6-mediated slow
nicotine inactivation may alter tolerance and with-
drawal such that these individuals are able to quit
sooner.
45
To further understand the temporal effect of
CYP2A6 genetic variation on smoking, we are continu-
ing to follow up this group of adolescents into adult-
hood. In addition, we are conducting a prospective
smoking study in college students in which we will be
able to assess the effect of CYP2A6 on the acquisition
of nicotine dependence and smoking behaviors in
young adults.
CYP2A6 GENETIC VARIATION:
IMPLICATIONS FOR SMOKING CESSATION
AND RISK FOR TOBACCO-RELATED
CANCERS
Cigarette smoking was the cause of 4.83 million
premature deaths globally in the year 2000.
100
Quitting
smoking has both immediate and long-term health ben-
ets such as reducing the risk for cardiovascular dis-
eases
101
and cancer.
102
Currently, nicotine replacement
therapy, including transdermal patch, nasal spray, gum
inhaler, and lozenges, is the most commonly used treat-
ment for smoking cessation.
103
Although these treat-
ments double the odds for quitting,
104
80% to 90% of
those who try these treatments have a relapse to their
former smoking patterns.
105
In addition to nicotine, 2 tobacco-specic nitro-
samines, 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanone (NNK) and N-nitrosonornicotine (NNN), are
metabolized by CYP2A6 to form active lung
carcinogens.
106-109
It has been shown that the risk for
cancer is related to the amount of smoke exposure,
110
and studies have found that the CYP2A6*4 inactive
allele is associated with a lower risk of lung can-
cer,
82,111,112
but not all studies are in agreement with
this nding.
86,113,114
Taken together, these results sug-
gest that smokers with CYP2A6 slow inactivator geno-
types may be at a lower risk for lung cancer as a result
of reduced bioactivation of these procarcinogens. In
addition, these individuals have lower cigarette con-
sumption and smoke for fewer years,
68
which will
reduce their overall exposure to these procarcino-
gens.
71,103
As mentioned, dependent smokers regulate their
smoking to maintain target nicotine concentrations in
the brain,
49,68
and CYP2A6 genetic variation affects
systemic levels of nicotine.
66,73
Inhibition of the
CYP2A6 enzyme would slow nicotine inactivation,
would prolong target brain nicotine levels, and would
be expected to reduce smoking behavior.
115
We have
done this phenocopying maneuver, using CYP2A6
inhibitors in normal metabolizers to mimic slow me-
tabolism, to demonstrate that the effect seen with ge-
netically reduced metabolism can also be reproduced
with inhibition of metabolism. In addition, CYP2A6
inhibition presents a novel treatment strategy. Methox-
salen is a potent CYP2A6 inhibitor currently used to
treat psoriasis.
116,117
Compared with placebo or nicoti-
nine alone, methoxsalen can signicantly increase
plasma nicotine levels. In addition, it reduced subjects
self-reported desire to smoke, decreased the amount
smoked, increased the latency to the next cigarette, and
decreased the total number of puffs taken.
118
In addi-
tion to altering smoking, inhibition of CYP2A6 is ex-
pected to decrease activation of procarcinogen sub-
strates. Methoxsalen, given to dependent smokers for 3
days during smoking as desired, also signicantly re-
duced smoke exposure and increased the formation of
the inactive metabolite of NNK, suggesting rerouting
from the procarcinogen activating pathway.
115,119
To-
gether, genetic and biochemical studies indicate that
inhibiting CYP2A6 in smokers may reduce the amount
smoked, reduce the activation of tobacco-specic pro-
carcinogens, and increase quitting.
115,120
Fig 2. CYP2A6 slow nicotine inactivators. Adolescent slow nicotine inactivators experimenting
with smoking may experience prolonged nicotine levels in the brain, which may increase their risk
for becoming nicotine-dependent, although their cigarette consumption, once nicotine-dependent, is
lower. As smoking continues into adulthood, the amount smoked daily remains lower and quitting
may increase for slow inactivators. The reduced cigarette consumption throughout a slow inacti-
vators smoking history may contribute to the lower risk for being a smoker, via increased quitting,
as the duration of smoking increases.
CONCLUSIONS
In summary, we conclude that slow nicotine inacti-
vation, predicted by CYP2A6 genetic variation, inu-
ences the risk for several aspects of smoking behavior
and nicotine dependence. However, the inuence of
CYP2A6 appears to vary along the smoking continuum
and is affected by factors such as status of nicotine
dependence and duration of smoking. The slow inacti-
vator genotype may increase the risk for nicotine de-
pendence when smoking is initiated during adolescence
but reduces the risk for smoking, lowers the amount
smoked, and decreases the duration of smoking among
adult dependent smokers. The apparent temporal effect
of CYP2A6 on smoking behavior and nicotine depen-
dence requires further investigation, but we postulate
that the continued effect of slow metabolism on reduc-
ing cigarette consumption, throughout the smoking his-
tory of slow inactivators, may affect tolerance and
withdrawal mechanisms, facilitating quitting among
these individuals. These ndings have implications for
the use of existing smoking cessation treatments in
adolescents and support CYP2A6 phenocopying as a
novel treatment for smoking reduction and cessation.
Moreover, the potential temporal effect of CYP2A6
genetic variation emphasizes the importance of un-
derstanding the various factors modulating vulnerabil-
ity to nicotine dependence and the need to identify
more specic smoking and disease prevention strate-
gies.
121,123,125,126,128,132,134
R.F.T. and E.M.S. hold shares in Nicogen Inc (Toronto, Ontario,
Canada) and are involved in the research program undertaken by
Nicogen Inc, a research pharmaceutical company that is focused on
the development of novel smoking cessation treatments based on the
inhibition of nicotine metabolism. V.M. is a student in the laboratory
of Dr Tyndale and has no involvement with Nicogen Inc.
References
1. Henningeld JE, Miyasato K, Jasinski DR. Abuse lia-
bility and pharmacodynamic characteristics of intrave-
nous and inhaled nicotine. J Pharmacol Exp Ther 1985;
234:1-12.
2. Stolerman IP, Jarvis MJ. The scientic case that nico-
tine is addictive. Psychopharmacology (Berl) 1995;117:
2-10, discussion 14-20.
3. Pidoplichko VI, DeBiasi M, Williams JT, Dani JA.
Nicotine activates and desensitizes midbrain dopamine
neurons. Nature 1997;390:401-4.
4. Pontieri FE, Tanda G, Orzi F, Di Chiara G. Effects of
nicotine on the nucleus accumbens and similarity to
those of addictive drugs. Nature 1996;382:255-7.
5. Salokangas RK, Vilkman H, Ilonen T, Taiminen T,
Bergman J, Haaparanta M, et al. High levels of dopa-
mine activity in the basal ganglia of cigarette smokers.
Am J Psychiatry 2000;157:632-4.
6. Laviolette SR, van der Kooy D. Blockade of mesolim-
bic dopamine transmission dramatically increases sen-
sitivity to the rewarding effects of nicotine in the ventral
tegmental area. Mol Psychiatry 2003;8:50-9.
7. Stein EA, Pankiewicz J, Harsch HH, Cho JK, Fuller SA,
Hoffmann RG, et al. Nicotine-induced limbic cortical
activation in the human brain: a functional MRI study.
Am J Psychiatry 1998;155:1009-15.
8. Koob GF. Neurobiology of addiction. Toward the de-
velopment of new therapies. Ann N Y Acad Sci 2000;
909:170-85.
9. Baker TB, Brandon TH, Chassin L. Motivational inu-
ences on cigarette smoking. Annu Rev Psychol 2004;
55:463-91.
10. Buisson B, Bertrand D. Nicotine addiction: the possible
role of functional upregulation. Trends Pharmacol Sci
2002;23:130-6.
11. Flores CM, Davila-Garcia MI, Ulrich YM, Kellar KJ.
Differential regulation of neuronal nicotinic receptor
binding sites following chronic nicotine administration.
J Neurochem 1997;69:2216-9.
12. Perry DC, Davila-Garcia MI, Stockmeier CA, Kellar
KJ. Increased nicotinic receptors in brains from smok-
ers: membrane binding and autoradiography studies.
J Pharmacol Exp Ther 1999;289:1545-52.
13. Epping-Jordan MP, Watkins SS, Koob GF, Markou A.
Dramatic decreases in brain reward function during
nicotine withdrawal. Nature 1998;393:76-9.
14. Teneggi V, Tiffany ST, Squassante L, Milleri S, Ziviani
L, Bye A. Smokers deprived of cigarettes for 72 h:
effect of nicotine patches on craving and withdrawal.
Psychopharmacology (Berl) 2002;164:177-87.
15. Benowitz NL. Pharmacology of nicotine: addiction and
therapeutics. Annu Rev Pharmacol Toxicol 1996;36:
597-613.
16. Jarvis MJ. Why people smoke. BMJ 2004;328:277-9.
17. Picciotto MR. Nicotine as a modulator of behavior:
beyond the inverted U. Trends Pharmacol Sci 2003;24:
493-9.
18. Mayhew KP, Flay BR, Mott JA. Stages in the develop-
ment of adolescent smoking. Drug Alcohol Depend
2000;59(Suppl 1):S61-81.
19. Lerman C, Berrettini W. Elucidating the role of genetic
factors in smoking behavior and nicotine dependence.
Am J Med Genet 2003;118B:48-54.
20. Sullivan PF, Kendler KS. The genetic epidemiology of
smoking. Nicotine Tob Res 1999;1(Suppl 2):S51-7, dis-
cussion S69-70.
21. Tyndale RF. Genetics of alcohol and tobacco use in
humans. Ann Med 2003;35:94-121.
22. Heath AC, Kirk KM, Meyer JM, Martin NG. Genetic
and social determinants of initiation and age at onset of
smoking in Australian twins. Behav Genet 1999;29:
395-407.
23. Koopmans JR, Slutske WS, Heath AC, Neale MC,
Boomsma DI. The genetics of smoking initiation and
quantity smoked in Dutch adolescent and young adult
twins. Behav Genet 1999;29:383-93.
24. Madden PA, Heath AC, Pedersen NL, Kaprio J, Kosk-
envuo MJ, Martin NG. The genetics of smoking persis-
tence in men and women: a multicultural study. Behav
Genet 1999;29:423-31.
25. Lessov CN, Benowitz NL, Swan GE. Evidence for
shared genetic risk between depressive symptoms and
smoking experimentation in men [abstract]. In Proceed-
ings of the Tenth Annual Meeting of the Society for
Research on Nicotine and Tobacco; 2004 Feb 18-21;
Scottsdale, Arizona. Scottsdale (AZ): The Society;
2004.
26. De Luca V, Wong AH, Muller DJ, Wong GW, Tyndale
RF, Kennedy JL. Evidence of association between
smoking and alpha7 nicotinic receptor subunit gene in
schizophrenia patients. Neuropsychopharmacology
2004;29:1522-6.
27. Freedman R, Adler LE, Bickford P, Byerley W, Coon
H, Cullum CM, et al. Schizophrenia and nicotinic re-
ceptors. Harv Rev Psychiatry 1994;2:179-92.
28. Adler LE, Hoffer LJ, Grifth J, Waldo MC, Freedman
R. Normalization by nicotine of decient auditory sen-
sory gating in the relatives of schizophrenics. Biol Psy-
chiatry 1992;32:607-16.
29. Freedman R, Coon H, Myles-Worsley M, Orr-Urtreger
A, Olincy A, Davis A, et al. Linkage of a neurophysi-
ological decit in schizophrenia to a chromosome 15
locus. Proc Natl Acad Sci U S A 1997;94:587-92.
30. Uhl GR, Liu QR, Walther D, Hess J, Naiman D. Poly-
substance abuse-vulnerability genes: genome scans for
association, using 1,004 subjects and 1,494 single-
nucleotide polymorphisms. Am J Hum Genet 2001;69:
1290-300.
31. Li MD, Ma JZ, Cheng R, Dupont RT, Williams NJ,
Crews KM, et al. A genome-wide scan to identify loci
for smoking rate in the Framingham Heart Study pop-
ulation. BMC Genet 2003;4(Suppl 1):S103-7.
32. Sullivan PF, Neale BM, van den Oord E, Miles MF,
Neale MC, Bulik CM, et al. Candidate genes for nico-
tine dependence via linkage, epistasis, and bioinformat-
ics. Am J Med Genet 2004;126B:23-36.
33. Wilhelmsen KC, Feiler HS, S-Cheng L, Lessov CN,
Ring HZ, Jack LM, et al. A genome-wide search for
quantitative trait loci associated with tobacco use phe-
notypes [abstract]. In Proceedings of the Tenth Annual
Meeting of the Society for Research on Nicotine and
Tobacco; 2004 Feb 18-21; Scottsdale, Arizona. Scotts-
dale (AZ): The Society; 2004.
34. Comings DE, Blum K. Reward deciency syndrome:
genetic aspects of behavioral disorders. Prog Brain Res
2000;126:325-41.
35. Goudriaan AE, Oosterlaan J, de Beurs E, Van den Brink
W. Pathological gambling: a comprehensive review of
biobehavioral ndings. Neurosci Biobehav Rev 2004;
28:123-41.
36. Arinami T, Ishiguro H, Onaivi ES. Polymorphisms in
genes involved in neurotransmission in relation to
smoking. Eur J Pharmacol 2000;410:215-26.
37. Batra V, Patkar AA, Berrettini WH, Weinstein SP,
Leone FT. The genetic determinants of smoking. Chest
2003;123:1730-9.
38. Lerman C, Niaura R. Applying genetic approaches to
the treatment of nicotine dependence. Oncogene 2002;
21:7412-20.
39. Benowitz NL, Zevin S, Jacob P III. Suppression of
nicotine intake during ad libitum cigarette smoking by
high-dose transdermal nicotine. J Pharmacol Exp Ther
1998;287:958-62.
40. Rose JE, Corrigall WA. Nicotine self-administration in
animals and humans: similarities and differences. Psy-
chopharmacology (Berl) 1997;130:28-40.
41. Klink R, de Kerchove dExaerde A, Zoli M, Changeux
JP. Molecular and physiological diversity of nicotinic
acetylcholine receptors in the midbrain dopaminergic
nuclei. J Neurosci 2001;21:1452-63.
42. Nguyen HN, Rasmussen BA, Perry DC. Subtype-
selective up-regulation by chronic nicotine of high-
afnity nicotinic receptors in rat brain demonstrated by
receptor autoradiography. J Pharmacol Exp Ther 2003;
307:1090-7.
43. Picciotto MR, Zoli M, Rimondini R, Lena C, Marubio
LM, Pich EM, et al. Acetylcholine receptors containing
the beta2 subunit are involved in the reinforcing prop-
erties of nicotine. Nature 1998;391:173-7.
44. Marubio LM, Gardier AM, Durier S, David D, Klink R,
Arroyo-Jimenez MM, et al. Effects of nicotine in the
dopaminergic system of mice lacking the alpha4 subunit
of neuronal nicotinic acetylcholine receptors. Eur
J Neurosci 2003;17:1329-37.
45. Schilstrom B, Svensson HM, Svensson TH, Nomikos
GG. Nicotine and food induced dopamine release in the
nucleus accumbens of the rat: putative role of alpha7
nicotinic receptors in the ventral tegmental area. Neu-
roscience 1998;85:1005-9.
46. Nomikos GG, Hildebrand BE, Panagis G, Svensson TH.
Nicotine withdrawal in the rat: role of alpha7 nicotinic
receptors in the ventral tegmental area. Neuroreport
1999;10:697-702.
47. Feng Y, Niu T, Xing H, Xu X, Chen C, Peng S, et al. A
common haplotype of the nicotine acetylcholine recep-
tor alpha 4 subunit gene is associated with vulnerability
to nicotine addiction in men. Am J Hum Genet 2004;
75:112-21.
48. Lueders KK, Hu S, McHugh L, Myakishev MV, Sirota
LA, Hamer DH. Genetic and functional analysis of
single nucleotide polymorphisms in the beta2-neuronal
nicotinic acetylcholine receptor gene (CHRNB2). Nic-
otine Tob Res 2002;4:115-25.
49. McMorrow MJ, Foxx RM. Nicotines role in smoking:
an analysis of nicotine regulation. Psychol Bull 1983;
93:302-27.
50. Benowitz NL, Jacob P III. Nicotine renal excretion rate
inuences nicotine intake during cigarette smoking.
51. Jarvik ME, Madsen DC, Olmstead RE, Iwamoto-
Schaap PN, Elins JL, Benowitz NL. Nicotine blood
levels and subjective craving for cigarettes. Pharmacol
Biochem Behav 2000;66:553-8.
52. Benowitz NL, Jacob P III. Metabolism of nicotine to
cotinine studied by a dual stable isotope method. Clin
53. Messina ES, Tyndale RF, Sellers EM. A major role for
CYP2A6 in nicotine C-oxidation by human liver micro-
somes. J Pharmacol Exp Ther 1997;282:1608-14.
54. Nakajima M, Yamamoto T, Nunoya K, Yokoi T, Na-
gashima K, Inoue K, et al. Role of human cytochrome
P4502A6 in C-oxidation of nicotine. Drug Metab Dis-
pos 1996;24:1212-7.
55. Dempsey D, Tutka P, Jacob P III, Allen F, Schoedel K,
Tyndale RF, et al. Nicotine metabolite ratio as an index
of cytochrome P450 2A6 metabolic activity. Clin Phar-
macol Ther 2004;76:64-72.
gashima K, Inoue K, et al. Characterization of CYP2A6
involved in 3-hydroxylation of cotinine in human liver
microsomes. J Pharmacol Exp Ther 1996;277:1010-5.
57. Pelkonen O, Rautio A, Raunio H, Pasanen M. CYP2A6:
a human coumarin 7-hydroxylase. Toxicology 2000;
144:139-47.
58. Xu C, Goodz S, Sellers EM, Tyndale RF. CYP2A6
genetic variation and potential consequences. Adv Drug
Deliv Rev 2002;54:1245-56.
59. Hadidi H, Irshaid Y, Vagbo CB, Brunsvik A, Cholerton
S, Zahlsen K, et al. Variability of coumarin 7- and
3-hydroxylation in a Jordanian population is suggestive
of a functional polymorphism in cytochrome P450
CYP2A6. Eur J Clin Pharmacol 1998;54:437-41.
60. Iscan M, Rostami H, Iscan M, Guray T, Pelkonen O,
Rautio A. Interindividual variability of coumarin
7-hydroxylation in a Turkish population. Eur J Clin
Pharmacol 1994;47:315-8.
61. Rautio A, Kraul H, Kojo A, Salmela E, Pelkonen O.
Interindividual variability of coumarin 7-hydroxylation
in healthy volunteers. Pharmacogenetics 1992;2:227-
33.
62. Benowitz NL, Jacob P III. Trans-3-hydroxycotinine:
disposition kinetics, effects and plasma levels during
cigarette smoking. Br J Clin Pharmacol 2001;51:53-9.
63. Shimada T, Yamazaki H, Guengerich FP. Ethnic-related
differences in coumarin 7-hydroxylation activities cat-
alyzed by cytochrome P4502A6 in liver microsomes of
Japanese and Caucasian populations. Xenobiotica 1996;
26:395-403.
64. Kwon JT, Nakajima M, Chai S, Yom YK, Kim HK,
Yamazaki H, et al. Nicotine metabolism and CYP2A6
allele frequencies in Koreans. Pharmacogenetics 2001;
11:317-23.
65. Nakajima M, Kwon JT, Tanaka N, Zenta T, Yamamoto
Y, Yamamoto H, et al. Relationship between interindi-
vidual differences in nicotine metabolism and CYP2A6
genetic polymorphism in humans. Clin Pharmacol Ther
2001;69:72-8.
66. Nakajima M, Yamagishi S, Yamamoto H, Yamamoto
T, Kuroiwa Y, Yokoi T. Decient cotinine formation
from nicotine is attributed to the whole deletion of the
CYP2A6 gene in humans. Clin Pharmacol Ther 2000;
67:57-69.
67. CYP2A6 allele nomenclature. Available from: URL:
http://www.imm.ki.se/CYPalleles/cyp2a6.htm. Ac-
cessed July 20, 2004.
68. Schoedel K, Hoffmann E, Rao Y, Sellers E, Tyndale
RF. Ethnic variation in CYP2A6 and association of
genetically slow nicotine metabolism and smoking in
adult Caucasians. Pharmacogenetics 2004;14:615-26.
69. Hsia FN, Spruijt-Metz D. The meanings of smoking
among Chinese American and Taiwanese American col-
lege students. Nicotine Tob Res 2003;5:837-50.
70. Spruijt-Metz D, Gallaher PE, Unger JB, Anderson-
Johnson C. Meanings of smoking and adolescent smok-
ing across ethnicities. J Adolesc Health 2004;35:197-
205.
71. Fujieda M, Yamazaki H, Saito T, Kiyotani K, Gyam
MA, Sakurai M, et al. Evaluation of CYP2A6 genetic
polymorphisms as determinants of smoking behavior
and tobacco-related lung cancer risk in male Japanese
smokers. Carcinogenesis 2004;25:2451-8.
72. Zacny JP, Stitzer ML. Cigarette brand-switching: ef-
fects on smoke exposure and smoking behavior. J Phar-
macol Exp Ther 1988;246:619-27.
73. Benowitz NL, Tyndale RF, Jacob P III, Swan GE.
CYP2A6 polymorphisms and nicotine metabolism [ab-
74. Rao Y, Hoffmann E, Zia M, Bodin L, Zeman M, Sellers
EM, et al. Duplications and defects in the CYP2A6
gene: identication, genotyping, and in vivo effects on
smoking. Mol Pharmacol 2000;58:747-55.
75. Pianezza ML, Sellers EM, Tyndale RF. Nicotine me-
tabolism defect reduces smoking. Nature 1998;393:750.
76. Tyndale RF, Sellers EM. Variable CYP2A6-mediated
nicotine metabolism alters smoking behavior and risk.
Drug Metab Dispos 2001;29:548-52.
77. Pomerleau OF. Individual differences in sensitivity to
nicotine: implications for genetic research on nicotine
dependence. Behav Genet 1995;25:161-77.
78. Pomerleau OF, Collins AC, Shiffman S, Pomerleau CS.
Why some people smoke and others do not: new per-
spectives. J Consult Clin Psychol 1993;61:723-31.
79. Pomerleau OF, Pomerleau CS, Namenek RJ. Early ex-
periences with tobacco among women smokers, ex-
smokers, and never-smokers. Addiction 1998;93:595-9.
80. DiFranza JR, Savageau JA, Fletcher K, Ockene JK,
Rigotti NA, McNeill AD, et al. Recollections and re-
percussions of the rst inhaled cigarette. Addict Behav
2004;29:261-72.
81. Ando M, Hamajima N, Ariyoshi N, Kamataki T, Mat-
suo K, Ohno Y. Association of CYP2A6 gene deletion
with cigarette smoking status in Japanese adults. J Epi-
demiol 2003;13:176-81.
82. Ariyoshi N, Miyamoto M, Umetsu Y, Kunitoh H,
Dosaka-Akita H, Sawamura Y, et al. Genetic polymor-
phism of CYP2A6 gene and tobacco-induced lung can-
cer risk in male smokers. Cancer Epidemiol Biomarkers
Prev 2002;11:890-4.
83. Iwahashi K, Waga C, Takimoto T. Whole deletion of
CYP2A6 gene (CYP2A6*4C) and smoking behavior.
Neuropsychobiology 2004;49:101-4.
84. Minematsu N, Nakamura H, Iwata M, Tateno H, Naka-
jima T, Takahashi S, et al. Association of CYP2A6
deletion polymorphism with smoking habit and devel-
opment of pulmonary emphysema. Thorax 2003;58:
623-8.
85. London SJ, Idle JR, Daly AK, Coetzee GA. Genetic
variation of CYP2A6, smoking, and risk of cancer.
Lancet 1999;353:898-9.
86. Loriot MA, Rebuissou S, Oscarson M, Cenee S, Miy-
amoto M, Ariyoshi N, et al. Genetic polymorphisms of
cytochrome P450 2A6 in a case-control study on lung
cancer in a French population. Pharmacogenetics 2001;
11:39-44.
87. Sabol SZ, Hamer DH. An improved assay shows no
association between the CYP2A6 gene and cigarette
smoking behavior. Behav Genet 1999;157:632-4.
88. Carter B, Long T, Cinciripini P. A meta-analytic review
of the CYP2A6 genotype and smoking behavior. Nico-
tine Tob Res 2004;6:221-7.
89. Goodz S, Alhuwalia J, Harris K, Xu C, Rao YS, Xu B,
et al. CYP2A6 genetic variants among African-
Americans. In Proceedings of the Eighth Annual Meet-
ing of the Society for Research on Nicotine and To-
bacco; 2002 Feb 20-23; Savannah, Georgia. Savannah
(GA): The Society; 2002.
90. Gu DF, Hinks LJ, Morton NE, Day IN. The use of long
PCR to conrm three common alleles at the CYP2A6
locus and the relationship between genotype and smok-
ing habit. Ann Hum Genet 2000;64:383-90.
91. OLaughlin J, Paradis G, Kim W, DiFranza JR,
Meshefedijan G, McMillan E, et al. Genetically de-
creased CYP2A6 and the risk of tobacco dependence: a
prospective study in novice smokers. Tob Control 2004;
13:422-8.
92. Benowitz NL, Pomerleau OF, Pomerleau CS, Jacob P
III. Nicotine metabolite ratio as a predictor of cigarette
consumption. Nicotine Tob Res 2003;5:621-4.
93. Abreu-Villaca Y, Seidler FJ, Qiao D, Tate CA, Cousins
MM, Thillai I, et al. Short-term adolescent nicotine
exposure has immediate and persistent effects on cho-
linergic systems: critical periods, patterns of exposure,
dose thresholds. Neuropsychopharmacology 2003;28:
1935-49.
94. Adriani W, Spijker S, Deroche-Gamonet V, Laviola G,
Le Moal M, Smit AB, et al. Evidence for enhanced
neurobehavioral vulnerability to nicotine during peri-
adolescence in rats. J Neurosci 2003;23:4712-6.
95. Belluzzi JD, Lee AG, Oliff HS, Leslie FM. Age-
dependent effects of nicotine on locomotor activity and
conditioned place preference in rats. Psychopharmacol-
ogy (Berl) 2004;174:389-95.
96. DiFranza JR, Rigotti NA, McNeill AD, Ockene JK,
Savageau JA, St Cyr D, et al. Initial symptoms of
nicotine dependence in adolescents. Tob Control 2000;
9:313-9.
97. Vastola BJ, Douglas LA, Varlinskaya EI, Spear LP.
Nicotine-induced conditioned place preference in ado-
lescent and adult rats. Physiol Behav 2002;77:107-14.
98. Breslau N, Johnson EO. Predicting smoking cessation
and major depression in nicotine-dependent smokers.
Am J Public Health 2000;90:1122-7.
99. Hymowitz N, Cummings KM, Hyland A, Lynn WR,
Pechacek TF, Hartwell TD. Predictors of smoking ces-
sation in a cohort of adult smokers followed for ve
years. Tob Control 1997;6(Suppl 2):S57-62.
100. Ezzati M, Lopez AD. Estimates of global mortality
attributable to smoking in 2000. Lancet 2003;362:847-
52.
101. McBride PE. The health consequences of smoking. Car-
diovascular diseases. Med Clin North Am 1992;76:333-
53.
102. Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R.
Smoking, smoking cessation, and lung cancer in the UK
since 1950: combination of national statistics with two
case-control studies. BMJ 2000;321:323-9.
103. Sellers EM, Tyndale RF, Fernandes LC. Decreasing
smoking behaviour and risk through CYP2A6 inhibi-
tion. Drug Discov Today 2003;8:487-93.
104. Silagy C, Mant D, Fowler G, Lancaster T. Nicotine
replacement therapy for smoking cessation. Cochrane
Database Syst Rev 2000;CD000146.
105. Fiore MC, Smith SS, Jorenby DE, Baker TB. The ef-
fectiveness of the nicotine patch for smoking cessation.
A meta-analysis. JAMA 1994;271:1940-7.
106. Patten CJ, Peterson LA, Murphy SE. Evidence for met-
abolic activation of N-nitrosonornicotine and
N-nitrosobenzylmethylamine by a rat nasal coumarin
hydroxylase. Drug Metab Dispos 1998;26:177-80.
107. Patten CJ, Smith TJ, Friesen MJ, Tynes RE, Yang CS,
Murphy SE. Evidence for cytochrome P450 2A6 and
3A4 as major catalysts for N-nitrosonornicotine alpha-
hydroxylation by human liver microsomes. Carcinogen-
esis 1997;18:1623-30.
108. Smith GB, Bend JR, Bedard LL, Reid KR, Petsikas D,
Massey TE. Biotransformation of 4-(methylnitros-
amino)-1-(3-pyridyl)-1-butanone (NNK) in peripheral
human lung microsomes. Drug Metab Dispos 2003;
31:1134-41.
109. Yamazaki H, Inui Y, Yun CH, Guengerich FP, Shimada
T. Cytochrome P450 2E1 and 2A6 enzymes as major
catalysts for metabolic activation of N-nitroso-
dialkylamines and tobacco-related nitrosamines in hu-
man liver microsomes. Carcinogenesis 1992;13:1789-
94.
110. Law MR, Morris JK, Watt HC, Wald NJ. The dose-
response relationship between cigarette consumption,
biochemical markers and risk of lung cancer. Br J
Cancer 1997;75:1690-3.
111. Kamataki T, Nunoya K, Sakai Y, Kushida H, Fujita K.
Genetic polymorphism of CYP2A6 in relation to can-
cer. Mutat Res 1999;428:125-30.
112. Miyamoto M, Umetsu Y, Dosaka-Akita H, Sawamura
Y, Yokota J, Kunitoh H, et al. CYP2A6 gene deletion
reduces susceptibility to lung cancer. Biochem Biophys
Res Commun 1999;261:658-60.
113. Wang H, Tan W, Hao B, Miao X, Zhou G, He F, et al.
Substantial reduction in risk of lung adenocarcinoma
associated with genetic polymorphism in CYP2A13, the
most active cytochrome P450 for the metabolic activa-
tion of tobacco-specic carcinogen NNK. Cancer Res
2003;63:8057-61.
114. Tan W, Chen GF, Xing DY, Song CY, Kadlubar FF,
Lin DX. Frequency of CYP2A6 gene deletion and its
relation to risk of lung and esophageal cancer in the
Chinese population. Int J Cancer 2001;95:96-101.
115. Sellers EM, Ramamoorthy Y, Zeman MV, Djordjevic
MV, Tyndale RF. The effect of methoxsalen on nicotine
and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone
(NNK) metabolism in vivo. Nicotine Tob Res 2003;5:
891-9.
116. Kharasch ED, Hankins DC, Taraday JK. Single-dose
methoxsalen effects on human cytochrome P-450 2A6
activity. Drug Metab Dispos 2000;28:28-33.
117. Koenigs LL, Peter RM, Thompson SJ, Rettie AE,
Trager WF. Mechanism-based inactivation of human
liver cytochrome P450 2A6 by 8-methoxypsoralen.
118. Sellers EM, Tyndale RF. Mimicking gene defects to
treat drug dependence. Ann N Y Acad Sci 2000;909:
233-46.
119. Takeuchi H, Saoo K, Yokohira M, Ikeda M, Maeta H,
Miyazaki M, et al. Pretreatment with 8-methoxy-
psoralen, a potent human CYP2A6 inhibitor, strongly
inhibits lung tumorigenesis induced by 4-(methylnitros-
amino)-1-(3-pyridyl)-1-butanone in female A/J mice.
Cancer Res 2003;63:7581-3.
120. Sellers EM, Kaplan HL, Tyndale RF. Inhibition of
cytochrome P450 2A6 increases nicotines oral bio-
availability and decreases smoking. Clin Pharmacol
Ther 2000;68:35-43.
121. Ariyoshi N, Sawamura Y, Kamataki T. A novel single
nucleotide polymorphism altering stability and activity
of CYP2a6. Biochem Biophys Res Commun 2001;281:
810-4.
122. Xu C, Rao YS, Xu B, Hoffmann E, Jones J, Sellers EM,
et al. An in vivo pilot study characterizing the new
CYP2A6*7, *8, and *10 alleles. Biochem Biophys Res
Commun 2002;290:318-24.
122a.Mwenifumbo JC, Myers MG, Wall T, Lin SK, Sellers
EM, Tyndale RF. Ethnic variation in CYP2A6*7,
CYP2A6*8 and CYP2A6*10 as assessed with a novel
haplotyping method. Pharmacogenetics. In press 2005.
123. Hadidi H, Zahlsen K, Idle JR, Cholerton S. A single
amino acid substitution (Leu160His) in cytochrome
P450 CYP2A6 causes switching from 7-hydroxylation
to 3-hydroxylation of coumarin. Food Chem Toxicol
1997;35:903-7.
124. Oscarson M, Gullsten H, Rautio A, Bernal ML, Sinues
B, Dahl ML, et al. Genotyping of human cytochrome
P450 2A6 (CYP2A6), a nicotine C-oxidase. FEBS Lett
1998;438:201-5.
125. Ariyoshi N, Takahashi Y, Miyamoto M, Umetsu Y,
Daigo S, Tateishi T, et al. Structural characterization of
a new variant of the CYP2A6 gene (CYP2A6*1B)
apparently diagnosed as heterozygotes of CYP2A6*1A
and CYP2A6*4C. Pharmacogenetics 2000;10:687-93.
126. Nunoya KI, Yokoi T, Kimura K, Kainuma T, Satoh K,
Kinoshita M, et al. A new CYP2A6 gene deletion re-
sponsible for the in vivo polymorphic metabolism of
()-cis-3,5-dimethyl-2-(3-pyridyl)thiazolidin-4-one hy-
drochloride in humans. J Pharmacol Exp Ther 1999;
289:437-42.
127. Oscarson M, McLellan RA, Gullsten H, Yue QY, Lang
MA, Bernal ML, et al. Characterisation and PCR-based
detection of a CYP2A6 gene deletion found at a high
frequency in a Chinese population. FEBS Lett 1999;
448:105-10.
128. Nunoya K, Yokoi T, Kimura K, Inoue K, Kodama T,
Funayama M, et al. A new deleted allele in the human
cytochrome P450 2A6 (CYP2A6) gene found in indi-
viduals showing poor metabolic capacity to coumarin
and ()-cis-3,5-dimethyl-2-(3-pyridyl)thiazolidin-4-
one hydrochloride (SM-12502). Pharmacogenetics
1998;8:239-49.
129. Oscarson M, McLellan RA, Gullsten H, Agundez JA,
Benitez J, Rautio A, et al. Identication and characteri-
sation of novel polymorphisms in the CYP2A locus:
implications for nicotine metabolism. FEBS Lett 1999;
460:321-7.
130. Kitagawa K, Kunugita N, Kitagawa M, Kawamoto T.
CYP2A6*6, a novel polymorphism in cytochrome p450
2A6, has a single amino acid substitution (R128Q) that
inactivates enzymatic activity. J Biol Chem 2001;276:
17830-5.
131. Yoshida R, Nakajima M, Watanabe Y, Kwon JT, Yokoi
T. Genetic polymorphisms in human CYP2A6 gene
causing impaired nicotine metabolism. Br J Clin Phar-
macol 2002;54:511-7.
132. Daigo S, Takahashi Y, Fujieda M, Ariyoshi N,
Yamazaki H, Koizumi W, et al. A novel mutant allele of
the CYP2A6 gene (CYP2A6*11) found in a cancer
patient who showed poor metabolic phenotype towards
tegafur. Pharmacogenetics 2002;12:299-306.
133. Oscarson M, McLellan RA, Asp V, Ledesma M, Ruiz
ML, Sinues B, et al. Characterization of a novel
CYP2A7/CYP2A6 hybrid allele (CYP2A6*12) that
causes reduced CYP2A6 activity. Hum Mutat 2002;20:
275-83.
134. Kiyotani K, Yamazaki H, Fujieda M, Iwano S, Mat-
sumura K, Satarug S, et al. Decreased coumarin
7-hydroxylase activities and CYP2A6 expression levels
in humans caused by genetic polymorphism in CYP2A6
promoter region (CYP2A6*9). Pharmacogenetics 2003;
13:689-95.
135. Pitarque M, von Richter O, Oke B, Berkkan H, Oscar-
son M, Ingelman-Sundberg M. Identication of a single
nucleotide polymorphism in the TATA box of the
CYP2A6 gene: impairment of its promoter activity.
Biochem Biophys Res Commun 2001;284:455-60.
136. Yoshida R, Nakajima M, Nishimura K, Tokudome S,
Kwon JT, Yokoi T. Effects of polymorphism in pro-
moter region of human CYP2A6 gene (CYP2A6*9) on
expression level of messenger ribonucleic acid and en-
zymatic activity in vivo and in vitro. Clin Pharmacol
Ther 2003;74:69-76.
137. Paschke T, Rieer M, Schuler-Metz A, Wolz L, Scherer
G, McBride CM, et al. Comparison of cytochrome P450
2A6 polymorphism frequencies in Caucasians and
African-Americans using a new one-step PCR-RFLP
genotyping method. Toxicology 2001;168:259-68.
138. Chen GF, Tang YM, Green B, Lin DX, Guengerich FP,
Daly AK, et al. Low frequency of CYP2A6 gene poly-
morphism as revealed by a one-step polymerase chain
reaction method. Pharmacogenetics 1999;9:327-32.
PHARMACOKINETICS AND
DRUG DISPOSITION
Influence of menstrual cycle on cytochrome
P450 2A6 activity and cardiovascular effects
of nicotine
Background and Objectives: The phase of the menstrual cycle has been reported to affect frequency of smok-
ing, withdrawal symptoms, and the likelihood of smoking cessation in women. Cytochrome P450 (CYP) 2A6
is primarily responsible for the metabolism of nicotine. Our objective was to evaluate the effect of the phase
of the menstrual cycle on the activity of CYP2A6 and the cardiovascular effects of nicotine.
Methods: Eleven healthy, nonsmoking women received a 30-minute combined infusion of deuterium-labeled
nicotine and cotinine (0.5 g kg
1
min
1
of each compound) during the midfollicular and midluteal
phases of the menstrual cycle. Nicotine and cotinine pharmacokinetic parameters and plasma adrenocortico-
tropic hormone (ACTH), epinephrine, and norepinephrine responses were measured over time.
Results: There were no biologically or statistically significant differences in the comparison of menstrual cycle
phases with regard to the pharmacokinetics of nicotine and cotinine. Nicotine clearance was 1000 315
mL/min and 1047 271 mL/min in the follicular and luteal phases, respectively (geometric mean ratio,
1.06; 90% confidence interval, 0.87-1.29). Cotinine clearance was 44 20 mL/min and 55 42 mL/min
in the follicular and luteal phases, respectively (geometric mean ratio, 1.13; 90% confidence interval, 0.90-
1.41). Nicotine infusion increased blood pressure, heart rate, and epinephrine concentrations. There were no
differences in catecholamine, ACTH, or hemodynamic responses to nicotine infusion between menstrual cycle
phases, although norepinephrine concentrations were constantly higher in the luteal phase compared with the
follicular phase.
Conclusions: CYP2A6 activity is not affected by menstrual cycle phase, and it is unlikely that menstrual
cyclerelated smoking habits of women are determined by changes in nicotine pharmacokinetics. The effects
of nicotine on plasma ACTH and catecholamine levels and hemodynamic parameters are not altered by
menstrual cycle phase in healthy, nonsmoking women. (Clin Pharmacol Ther 2005;77:159-69.)
Janne Hukkanen, MD, PhD, Steven G. Gourlay, MBBS, PhD,
Saraswati Kenkare, PhD, and Neal L. Benowitz, MD San Francisco and South San Francisco,
Calif
From the Division of Clinical Pharmacology and Experimental Thera-
peutics, Medical Service, San Francisco General Hospital Medical
Center, and Departments of Medicine, Psychiatry, and Biopharma-
ceutical Sciences, University of California, San Francisco, San Fran-
cisco, and Pharmacological Sciences, Genentech Inc, South San Fran-
cisco.
This study was presented in part at the annual meeting of the
American Society for Clinical Pharmacology and Therapeutics
(abstract MPI-107), Atlanta, Ga, March 24-27, 2002.
Supported by US Public Health Service grants DA02277 and
DA12393 from the National Institute on Drug Abuse, National
Institutes of Health, and grants from the Academy of Finland, the
Paavo Nurmi Foundation, and the Finnish Cultural Foundation to
J.H. Carried out in part at the General Clinical Research Center at
San Francisco General Hospital Medical Center with the support of
the Division of Research Resources, National Institutes of Health
(5-MO1-RR-00083-41).
Received for publication July 30, 2004; accepted Oct 24, 2004.
Reprint requests: Neal L. Benowitz, MD, Division of Clinical Phar-
macology and Experimental Therapeutics University of California,
San Francisco, Box 1220, San Francisco, CA 94143-1220.
E-mail: nbeno@itsa.ucsf.edu
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.012
159
The phase of the menstrual cycle inuences the smok-
ing habits of women; the frequency of smoking is slightly
increased during the late luteal phase.
1-3
Nicotine with-
drawal symptoms are also greater in the late luteal
phase,
4-8
and smoking cessation is reported to be harder
for female smokers in premenstruum compared with
women in midcycle and men.
9,10
The transdermal nicotine
patch has a greater effect in reducing craving during
smoking abstinence in the late luteal phase compared with
the follicular phase.
7
These observations suggest that hor-
monal changes during the menstrual cycle affect the
smoking behavior of women.
Changes in the metabolism of nicotine have been
shown to affect smoking behavior. Lower cytochrome
P450 (CYP) 2A6 activity caused by the presence of
defective alleles of CYP2A6, as well as pharmacologic
inhibition of CYP2A6, is associated with reduced cig-
arette consumption.
11,12
The main pathway of nicotine
metabolism is C-oxidation to cotinine, which accounts
for 70% to 80% of nicotine metabolism on average.
13,14
Cotinine is further metabolized to trans-3-hydroxy-
cotinine. Both of these reactions are catalyzed by he-
patic CYP2A6.
15,16
The inuence of sex hormones on nicotine metabo-
lism is suggested by the ndings that female gender,
pregnancy, and the use of oral contraceptives accelerate
the metabolism of nicotine and cotinine. Clearance is
increased by 60% and 140% for nicotine and cotinine,
respectively, in pregnancy compared with post partum
as a result of a rise in metabolic clearance.
17
A study
comparing women during pregnancy and again post
partum found that mean salivary cotinine concentration
per cigarette was higher when not pregnant (3.5 ng/mL
versus 9.9 ng/mL), consistent with higher cotinine
clearance during pregnancy.
18
Pregnant smokers had a
substantially lower level of serum nicotine than ex-
pected when standardized for their nicotine intake com-
pared with population-based values.
19
These results
suggest that CYP2A6 is induced during pregnancy,
possibly by sex hormones. Results from a recently
completed twin study with intravenous infusions of
both nicotine and cotinine show that nicotine and coti-
nine clearances are higher in women compared with
men and oral contraceptive use further accelerates nic-
otine and cotinine clearances in women.
20
Nicotine
clearance and cotinine clearance were 13% and 26%
higher, respectively, in women not taking oral contra-
ceptives compared with men. Oral contraceptive use
induced nicotine and cotinine clearances by 30% and
33%, respectively, compared with women not taking
oral contraceptives. The gender difference was also
detected in a recent study on smokers showing that the
ratio of nicotine to cotinine plus 3-hydroxycotinine in
24-hour urine collection is signicantly lower in
women, indicating faster metabolism in women com-
pared with men.
21
Thus there are reasons to suspect that the high estro-
gen and progesterone concentrations in the luteal phase
of the menstrual cycle might accelerate the metabolism
of nicotine. The effect of the menstrual cycle on
CYP2A6 activity has not been studied previously. The
aims of our study were to determine the effects of the
menstrual cycle on the disposition kinetics of nicotine
and to test the hypothesis that CYP2A6 activity is
induced in the luteal phase.
We also measured the effects of nicotine on cardio-
vascular parameters and hormonal responses in the
follicular and luteal phases of the menstrual cycle.
Nicotine is a sympathomimetic drug that causes cate-
cholamine release, increases heart rate and blood pres-
sure, and constricts some blood vessels. These hemo-
dynamic effects of nicotine may contribute to tobacco-
induced cardiovascular disease.
22
Nicotine-mediated
release of adrenocorticotropic hormone (ACTH) has
been speculated to contribute to the development of
addiction.
23
The menstrual cycle can also affect nor-
epinephrine concentrations and some hemodynamic re-
sponses to sympathomimetics.
24-26
No previous study
has examined the combined effect of menstrual cycle
phase and nicotine infusion on these parameters.
METHODS
Subjects. Healthy, premenopausal, nonsmoking
women were recruited for the study. The subjects used
no medications; they had not taken oral contraception
for the previous 3 months and had taken no other
prescription medication for the previous 28 days. Over-
the-counter medications had not been taken for the
previous 3 days. Exclusion criteria included an irregu-
lar menstrual cycle (dened as menstrual cycle duration
of 25 days or35 days or intermittent bleeding),
pregnancy, drug or alcohol abuse, and abnormal liver or
renal function (dened as elevated serum creatinine
level or abnormal liver function tests). Plasma nicotine
and cotinine concentrations were measured at screening
to conrm nonsmoking. The study was approved by the
Committee on Human Research at the University of
California, San Francisco (San Francisco, Calif). Writ-
ten, informed consent was obtained from each subject.
The subjects were nancially compensated for
participation.
The number of subjects (N11) was based on a power
analysis for paired Student t test of nicotine clearance to
160 Hukkanen et al MARCH 2005
detect an effect size of 25% assuming a coefcient of
variation of 25% with .05 and .2.
Experimental protocol. Subjects were admitted
twice to the General Clinical Research Center at San
Francisco General Hospital Medical Center (San Fran-
cisco, Calif)once during the follicular phase and
once during the luteal phase. Serum estradiol and pro-
gesterone concentrations were measured during each
admission to conrm the phase of menstrual cycle. Day
1 of the follicular phase was determined by the onset of
menstrual bleeding. The day of ovulation was deter-
mined by a positive luteinizing hormone surge detected
by use of a home diagnostic ovulation kit (OvuKit;
Quidel Corporation, San Diego, Calif) and was speci-
ed as day 1 of the luteal phase. The day of admission
was assigned by counting the days from the onset of
menstruation (day 5-7 for the follicular phase) and by
counting the days after ovulation (day 6-8 for the luteal
phase). Subjects were instructed to test the rst morn-
ing urine with the ovulation kit 4 days before the
predicted midcycle and to continue the use of the kit
every morning (at least 9 days) until a positive result
was detected. The sequence (ie, follicular or luteal
phase rst) of the 2 admissions was alternated to con-
trol for any inuence of the order of testing. Subjects
were disqualied if hormonal measurements or ovula-
tion kit results were inconsistent with a normal cycle.
(If the rst admission was planned to be in the luteal
phase and the subject did not ovulate, she was disqual-
ied. If the rst admission was in the follicular phase
and the subject did not ovulate 2 consecutive times, she
was disqualied.)
The study was conducted as an outpatient study. Sub-
jects were asked to come to the clinical research center at
7 AM after fasting overnight. Two intravenous catheters
were inserted, one in the nondominant forearm for blood
drawing and another in the dominant antecubital vein for
infusion, and samples for estradiol and progesterone mea-
surements were taken. Subjects received an intravenous
infusion of deuterium-labeled nicotine and cotinine (0.5
g kg
1
min
1
of nicotine-d
2
[3,3-dideuteronicotine]
and 0.5 g kg
1
min
1
of cotinine-d
4
[2,4,5,6-tetra-
deuterocotinine]) for 30 minutes. The dose for 1 subject
was calculated erroneously, resulting in a lower infusion
rate of 0.4 g kg
1
min
1
. The same dose was infused
during both menstrual phases for all patients. Blood pres-
sure, heart rate, and nger skin temperature were mea-
sured at 15, 10, 5, 0, 5, 10, 15, 20, 25, 30, 40, 45, 50,
60, and 90 minutes and then at 2, 3, and 4 hours after the
beginning of the infusion. Blood samples for nicotine and
cotinine measurements were drawn at 0, 15, 30, 45, 60,
and 90 minutes and then at 2, 3, 4, 6, 8, 24, 48, 72, and 96
hours after the start of the infusion. Blood samples for
measurements of norepinephrine and epinephrine concen-
trations were taken at 30, 15, 0, 15, and 30 minutes,
and blood samples for ACTHmeasurements were taken at
30, 15, 0, 15, 30, 45, 60, and 90 minutes. Urine was
collected for 8 hours after infusion. Subjects were permit-
ted to sit up or to stand, and a meal was served 1 hour after
the infusion was completed. Subjects were discharged 8
hours after the start of the infusion. Subjects came back to
the study center on the next 4 days to have blood samples
taken.
Analytic methods. Nicotine and metabolite concentra-
tions in plasma were determined by gas chromatography
mass spectrometry.
27
Nicotine and metabolite concentra-
tions in urine were determined by liquid chromatography
tandem mass spectrometry.
28
Glucuronide-conjugated
nicotine, cotinine, and trans-3-hydroxycotinine levels
were measured as the difference in the total concentration
before and after hydrolysis by incubation with 2000 U of
-glucuronidase from Escherichia coli (Sigma, St Louis,
Mo) per 1 mL of urine.
Plasma concentrations of epinephrine and norepi-
nephrine were determined by means of HPLC with a
reversed-phase column with electrochemical detection
as described previously.
29
Plasma ACTH measure-
ments were performed with a chemiluminescence im-
munometric assay with ACTH 100T Kit (Nichols In-
stitute Diagnostics, San Juan Capistrano, Calif). Serum
estradiol and progesterone measurements were per-
formed with automated immunoassay by Unilab Cor-
poration (Tarzana, Calif).
Data analysis. Pharmacokinetic parameters were es-
timated from the plasma concentrations by use of
model-independent methods described previously
13
with the WinNonlin pharmacokinetic analyses package
(Pharsight Corp, Mountain View, Calif). Total clear-
ance was :
CL
nic
Dose nicd
2
AUC nicd
2
CL
cot
Dose cotd
4
AUC cotd
4
where CL is clearance (in milliliters per minute), AUC
is the area under the plasma concentrationtime curve
extrapolated to innity, nic is nicotine, and cot is coti-
nine. Renal clearance (CL
renal
) was calculated as fol-
lows:
CL
renal
Urinary excretion of nicotine or cotinine

AUC
8h
based on urine collected and the area under the plasma
2005;77(3):159-69 CYP2A6 and menstrual cycle 161
concentrationtime curve for the 8-hour period after the
infusion (AUC
8h
). Nonrenal clearance was estimated as
total clearance minus renal clearance.
Fractional conversion of nicotine to cotinine (f) was
estimated by use of the plasma concentrations of coti-
nine formed as a metabolite during the nicotine infusion
and the clearance of cotinine estimated from its plasma
concentrations after the cotinine infusion, as follows:
f
AUC
cotd
2
Dose
nicd
2
CL
cotd
4
The metabolic clearance of nicotine via formation of
cotinine was computed as CL
nic
f.
Analysis of the urine metabolite data was based on
the urine collection over an 8-hour period after the
beginning of the nicotine and cotinine infusion. Urine
metabolite concentrations were expressed as a fraction
of the total dose of nicotine administered and as a
fraction of the total nicotine plus metabolites recovered.
The pharmacokinetic parameters were compared
across menstrual phases by the paired Student t test.
Pharmacodynamic parameters were compared across
menstrual phases by repeated-measures ANOVA. The
statistical signicance of changes in plasma concentra-
tions of ACTH, epinephrine, and norepinephrine and
blood pressure and heart rate responses from baseline to
the 30-minute time point was tested with the paired
Student t test. AUC from baseline to 30 minutes and
AUC from baseline to 60 minutes of the ACTH and
catecholamine concentration-time curves, as well as
AUCs of the cardiovascular parameter-time curves for
the follicular and luteal phases, were tested with the
paired Student t test or 1-sample t test. Three measure-
ments immediately before nicotine infusion were aver-
aged to establish baseline values (values at 15, 10,
and 5 minutes for blood pressure and heart rate and
values at 30, 15, and 0 minutes for ACTH and
catecholamine levels). Spearman rank order correla-
tions were used to analyze the correlation of estradiol
and progesterone with nicotine and cotinine clearances.
Geometric mean ratios and 90% condence intervals
for differences in means were calculated from the log-
transformed data.
RESULTS
The subjects were 11 healthy, premenopausal, non-
smoking women aged 19 to 43 years. The mean age
was 26 7 years, and the mean weight was 63 7 kg.
The ethnicity of the subjects was as follows: 5 His-
panic, 3 Asian, 1 black, and 2 Asian/Hispanic. One
subject consumed 20 g of alcohol weekly; the others
did not drink alcohol. Fifteen subjects were enrolled,
and there were 4 dropouts. The reasons for dropping
out were difcult venipuncture (1 subject) and anovu-
lation (2 subjects). One subject was lost to follow-up
for unknown reasons.
Disposition kinetics of nicotine and cotinine. The
mean dose of nicotine administered by intravenous
infusion was 0.92 mg (range, 0.79-1.08 mg). The ef-
fects of menstrual cycle phase on the disposition kinet-
ics of nicotine and cotinine are presented in Tables I
and II and Figs 1 and 2. There was no signicant
inuence of menstrual cycle on pharmacokinetic pa-
rameters of nicotine and cotinine. In 1 subject the
nicotine and cotinine clearance increased more than
Table I. Effect of menstrual cycle phases on disposition kinetics of nicotine-d
2
(N 11)
Follicular
(mean SD)
Luteal
(mean SD)
Geometric mean ratios
(luteal/follicular) and
90% condence interval
C
max
(ng/mL) 6.65 1.95 6.65 2.22 0.99 (0.84-1.16)
t
(min) 132.1 35.9 128.7 34.2 0.98 (0.78-1.23)

Clearance (mL/min) 1000 315 1047 271 1.06 (0.87-1.29)
Nonrenal clearance (mL/min) 971 308 1021 269 1.07 (0.88-1.30)
Renal clearance (mL/min) 28.9 20.6 25.1 13.1 1.03 (0.64-1.63)
Mean residence time (min) 207.3 140.5 162.5 37.0 0.86 (0.63-1.19)
V
ss
(L) 182.9 65.6 163.4 31.7 0.92 (0.80-1.05)
Fractional conversion of
nicotine to cotinine
0.81 0.20 0.76 0.16 0.95 (0.84-1.07)
Metabolic clearance by
cotinine pathway (mL/min)
837 385 828 370 1.01 (0.77-1.31)
Differences between follicular and luteal phases are not statistically signicant as tested with paired t test.
C
max
, Peak plasma concentration; t
, half-life; V
ss
, steady-state volume of distribution.
3-fold in the luteal phase compared with the follicular
phase. Exclusion of this subject did not change the
results. The sequence of the 2 treatments (follicular or
luteal phase rst) did not inuence the results. Mean
intraindividual coefcients of variation for clearance
were 15.3% and 17.5% for nicotine-d
2
and cotinine-d
4
,
respectively. Mean serum estradiol concentrations were
89.9 63.5 pg/mL and 194.0 52.1 pg/mL for the
follicular and luteal phases, respectively. Mean serum
progesterone concentrations were 0.4 0.2 ng/mL and
9.6 4.3 ng/mL, respectively. Estradiol and proges-
terone concentrations did not correlate with nicotine or
cotinine clearances.
Urine metabolite patterns. The values for percent-
age recovery of nicotine and its metabolites derived
from nicotine-d
2
are shown in Table III. No signicant
differences between phases were detected.
Table II. Effect of menstrual cycle phases on disposition kinetics of cotinine-d
4
(N 11)
Follicular
(mean SD)
Luteal
(mean SD)
Geometric mean ratios
(luteal/follicular) and
90% condence interval
C
max
(ng/mL) 25.2 8.4 22.3 7.3 0.89 (0.77-1.04)
t
(min) 1057 269 1149 468 1.04 (0.86-1.26)

Clearance (mL/min) 44.1 19.7 55.0 42.1 1.13 (0.90-1.41)
Nonrenal clearance (mL/min) 36.9 17.4 46.9 37.2 1.14 (0.91-1.42)
Renal clearance (mL/min) 7.1 4.1 8.1 5.4 1.14 (0.78-1.67)
Mean residence time (min) 1353 328 1410 508 1.01 (0.84-1.21)
V
ss
(L) 57.0 25.0 65.6 28.4 1.13 (0.98-1.31)
Differences between follicular and luteal phases are not statistically signicant as tested with paired t test.
Fig 1. Plasma clearance of nicotine-d
2
(A) and cotinine-d
4
(B) for individual subjects across menstrual cycle phases.
Data on mean and SD values for clearance in each menstrual
cycle phase are provided in Tables I and II.
Fig 2. Mean plasma epinephrine (A) and norepinephrine (B)
concentrations. Data are presented as mean SD.
Plasma epinephrine, norepinephrine, and ACTH
concentrations. Plasma epinephrine and norepineph-
rine concentrations are presented in Fig 2. Plasma nor-
epinephrine concentrations were consistently higher
during the luteal phase (29% higher on average in luteal
compared with follicular phase; ANOVA, P .13), but
there was no increase during infusion of nicotine. Epi-
nephrine and ACTH concentrations (data not shown)
did not differ between menstrual phases (ANOVA, P
.49 and P .21, respectively). Epinephrine level in-
creased 1.6-fold as a result of nicotine infusion (21.8-
34.3 pg/mL; Student t test, P .070 for both phases
combined; 1-sample t test, P .0098 for AUC from
baseline to 30 minutes for both phases combined).
Plasma ACTH concentrations did not change during
nicotine infusion. The change from baseline to the
30-minute time point and AUC from baseline to 30
minutes for ACTH and catecholamine levels did not
differ statistically signicantly between menstrual
phases.
Heart rate and blood pressure response. Heart rate
and blood pressure responses are presented in Fig 3.
Blood pressure was slightly reduced and heart rate
slightly increased in the luteal phase compared with the
follicular phase. These differences were not statistically
signicant. Nicotine infusion increased heart rate by
10.4 beats/min (63.3-73.7 beats/min; Student t test, P
.0006 for both phases combined), systolic blood pres-
sure by 5.4 mm Hg (105.3-110.7 mm Hg; Student t test,
P .0008 for both phases combined), and diastolic
blood pressure by 4.3 mm Hg (63.8-68.1 mm Hg;
Student t test, P .0024 for both phases combined).
There were no statistically signicant differences be-
tween phases in cardiovascular response to nicotine
infusion (analyzed as the change from baseline to the
30-minute time point, AUC from baseline to 30 min-
utes, and AUC from baseline to 60 minutes).
DISCUSSION
This study demonstrates that there are no signicant
differences in nicotine and cotinine pharmacokinetics
between the midfollicular and midluteal phases of the
menstrual cycle in healthy, nonsmoking women. Be-
cause both nicotine and cotinine are specic substrates
for CYP2A6, it can be concluded that CYP2A6 activity
Table III. Effect of menstrual cycle phases on
urinary recovery of nicotine and metabolites
(N 11)
Follicular (%)
(mean SD)
Luteal (%)
(mean SD)
Nicotine 28 27 24 22
Nicotine glucuronide 6 4 10 8
Cotinine 22 8 26 7
Cotinine glucuronide 6 4 6 4
3-Hydroxycotinine 35 17 30 20
3-Hydroxycotinine
glucuronide
4 4 3 3
Data are based on excretion of nicotine-d
2
and metabolites as percentage of
total recovery in 8-hour urine. Differences between follicular and luteal phases
are not statistically signicant as tested with paired t test.
Fig 3. Mean heart rate (A), systolic blood pressure (BP) (B),
and diastolic BP (C) during study. Data are presented as mean
SD. bpm, Beats per minute.
does not differ between the midfollicular and midluteal
phases. Thus pharmacokinetic factors cannot explain
the higher frequency of smoking and greater difculty
in smoking cessation observed for women in the luteal
phase compared with the follicular phase of the cycle.
Nonsmokers were studied to avoid the potential con-
founding effects of smoking on the metabolism and
cardiovascular effects of nicotine. Smoking is known to
inhibit nicotine metabolism.
30
Prior nicotine exposure
blunts the cardiovascular response to nicotine as a
result of the development of tolerance.
31
The changes in drug metabolism during the men-
strual cycle have received considerable attention over
the years. Most of the studies exploring the effects of
the menstrual cycle phases on CYP enzymes have been
negative or have shown only minor, clinically insignif-
icant effects.
32
However, many of the earlier studies
had methodologic problems such as not controlling for
anovulation, the use of a small sample size, and the use
of probes metabolized by several CYP enzymes. If
ovulation does not occur, concentrations of estrogen
and progesterone remain low throughout the cycle. In
our study subjects with anovulatory cycles were ex-
cluded. Although the sample size of this study was
relatively small, the use of a crossover design results in
statistical robustness. When interpreting the results of
this study, one must keep in mind that a lack of differ-
ence in nicotine pharmacokinetics between the midfol-
licular and midluteal phases does not exclude an effect
of other phases (ovulatory or menstrual) of the men-
strual cycle.
Most of the studies on CYP1A2 activity report no
difference in caffeine or theophylline pharmacokinetics
as a result of menstrual cycle phase,
33-37
but some
studies show slightly higher clearance in the follicular
phase
38-40
or menstrual phase.
41
One study that mea-
sured the saliva paraxanthine-to-caffeine ratio every
day throughout the menstrual cycle showed no trend for
menstrual cycle phase effects.
34
CYP2C9 activity as assessed by the plasma AUC
after oral phenytoin or steady-state phenytoin concen-
tration is not affected by menstrual cycle phase in the
general population or in patients with epilepsy.
42-45
However, a subgroup of patients with epilepsy with a
high frequency of seizures during the menstrual phase
of the menstrual cycle (catamenial epilepsy) had lower
phenytoin concentrations and higher oral clearance dur-
ing menses.
44-46
CYP2C19 activity in the midluteal and
midfollicular phases was studied recently by use of
omeprazole as a probe, and no differences were detect-
ed.
47
CYP2D6 activity in the menstrual cycle phases has
been investigated by use of dextromethorphan, meto-
prolol, ethylmorphine, and debrisoquin (INN, debriso-
quine) as substrates. Most of the studies have not de-
tected differences between cycle phases,
48-51
although
1 study showed a small decrease in debrisoquin meta-
bolic ratio (ie, higher activity) in the luteal phase.
52
CYP2D6 activity is higher during pregnancy.
53,54
CYP2D6 seems to be analogous to CYP2A6 in that the
activities of both are induced by pregnancy,
17
but nei-
ther is affected by menstrual cycle phase. The inducing
effect of sex hormones on CYP2A6 and CYP2D6 ac-
tivities during pregnancy is most likely both dose- and
time-dependent, requiring higher concentrations of hor-
mones and a longer time of exposure than are present in
the luteal phase of the normal menstrual cycle.
The inuence of menstrual cycle phase on the phar-
macokinetics of various CYP3A4 substrates has been
studied extensively. Studies with alprazolam,
55,56
mi-
dazolam,
57,58
triazolam,
59
alfentanil,
60
dextromethor-
phan,
48,61
and eletriptan
62
as substrates show no
changes in clearance during the menstrual cycle. Sim-
ilarly, the urinary ratio of 6-hydroxycortisol to corti-
sol remains constant during the menstrual cycle.
63,64
Only 1 study using oral midazolam as a probe detected
higher CYP3A4 activity in the follicular phase.
65
In-
terestingly, the messenger ribonucleic acid for
CYP3A7, a major fetal CYP enzyme, is expressed to a
much greater degree (approximately 10-fold) in the
endometrium in the proliferative (follicular) phase than
in the secretory (luteal) phase, whereas CYP3A4 levels
are constant.
66
However, CYP3A7 is not expected to
have a signicant effect on systemic clearance of drugs
in adults because of its low expression in adult livers.
67
It can be concluded that there are no differences
between menstrual cycle phases in the activity of drug-
metabolizing CYP enzymes studied with specic
probes thus far. However, CYP2C9 activity seems to be
higher in the menstrual phase of patients with catame-
nial epilepsy. Other major drug-metabolizing CYP en-
zymes such as CYP2B6, CYP2C8, and CYP2E1 have
not yet been studied in this regard. The numerous
studies on the effect of menstrual cycle phase on eth-
anol pharmacokinetics are not discussed here because
ethanol metabolism is mediated mainly by alcohol de-
hydrogenase and not CYP2E1.
In our study, consistently higher concentrations of
plasma norepinephrine were detected in the midluteal
phase compared with the midfollicular phase. Although
statistically nonsignicant, this nding is consistent
with earlier studies.
24-26
Baseline epinephrine and
ACTH concentrations were similar in the 2 phases of
the menstrual cycle, as shown in previous
studies.
24,25,68-70
Some older studies have detected a
higher ACTH level during ovulation
71,72
or in the fol-
licular phase.
73
Although statistically nonsignicant,
systolic and diastolic blood pressures were slightly
higher and heart rate slightly lower during the follicular
phase compared with the luteal phase. These small
differences have also been shown previously.
74,75
Plasma epinephrine concentrations increased mod-
estly during low-dose nicotine infusion. An increase in
epinephrine concentrations during nicotine infusion has
been reported previously.
29,76-79
The increase in plasma
epinephrine level was smaller in our study than in the
previous nicotine infusion studies,
29,76,77,79
including a
study in nonsmokers with the same low nicotine dose as
in our study.
79
This might be because of a different
gender distribution; the subjects in the previous studies
were almost exclusively male, whereas the subjects in
this study were exclusively female. Norepinephrine and
ACTH plasma concentrations remained unchanged
during nicotine infusion. Nicotine infusion has been
shown to result in a small increase (approximately 15%)
in the plasma norepinephrine concentrations in male sub-
jects.
29,79
Nicotine infusions have been shown to increase
ACTH concentrations.
80,81
The results from the study by
Newhouse et al
81
support our nding that plasma ACTH
level is not elevated by low-dose nicotine infusion. Nic-
otine infusion increased blood pressure and heart rate as
expected.
29,76-79
There were no signicant differences be-
tween menstrual cycle phases on hormonal and hemody-
namic response to nicotine infusion.
In conclusion, CYP2A6 activity is not affected by
menstrual cycle phase, and it is unlikely that menstrual
cyclerelated smoking habits of women are determined
by changes in nicotine pharmacokinetics. The effects of
nicotine on plasma ACTH and catecholamine levels, as
well as hemodynamic parameters, are not altered by
menstrual cycle phase in healthy, nonsmoking women.
We thank Dr Peyton Jacob III for supervision of analytic chem-
istry; Patricia Buley, Brenda Herrera, and Sandra Tinetti and the staff
of the General Clinical Research Center at San Francisco General
Hospital for assistance in conducting the clinical study; Lisa Yu for
analytic chemistry; Margaret Wilson for catecholamine analyses; and
Barbara Chang for ACTH analyses. We thank Gunnard Modin and
Faith Allen for help in statistical analyses and Dr Delia Dempsey for
assistance with medical supervision. We thank Pharsight Corp for
providing the WinNonlin package as an educational gift.
The authors have no conicts of interest.
References
1. Mello NK, Mendelson JH, Palmieri SL. Cigarette smok-
ing by women: interactions with alcohol use. Psycho-
pharmacology (Berl) 1987;93:8-15.
2. DeBon M, Klesges RC, Klesges LM. Symptomatology
across the menstrual cycle in smoking and nonsmoking
women. Addict Behav 1995;20:335-43.
3. Snively TA, Ahijevych KL, Bernhard LA, Wewers ME.
Smoking behavior, dysphoric states and the menstrual
cycle: results from single smoking sessions and the nat-
ural environment. Psychoneuroendocrinology 2000;
25:677-91.
4. OHara P, Portser SA, Anderson BP. The inuence of
menstrual cycle changes on the tobacco withdrawal syn-
drome in women. Addict Behav 1989;14:595-600.
5. Pomerleau CS, Garcia AW, Pomerleau OF, Cameron
OG. The effects of menstrual phase and nicotine absti-
nence on nicotine intake and on biochemical and subjec-
tive measures in women smokers: a preliminary report.
Psychoneuroendocrinology 1992;17:627-38.
6. Allen SS, Hatsukami D, Christianson D, Nelson D.
Symptomatology and energy intake during the menstrual
cycle in smoking women. J Subst Abuse 1996;8:303-19.
7. Allen SS, Hatsukami D, Christianson D, Brown S. Ef-
fects of transdermal nicotine on craving, withdrawal and
premenstrual symptomatology in short-term smoking ab-
stinence during different phases of the menstrual cycle.
Nicotine Tob Res 2000;2:231-41.
8. Perkins KA, Levine M, Marcus M, Shiffman S, DAmico
D, Miller A, et al. Tobacco withdrawal in women and
menstrual cycle phase. J Consult Clin Psychol 2000;68:
176-80.
9. Allen SS, Hatsukami DK, Christianson D, Nelson D.
Withdrawal and pre-menstrual symptomatology during
the menstrual cycle in short-term smoking abstinence:
effects of menstrual cycle on smoking abstinence. Nico-
tine Tob Res 1999;1:129-42.
10. Craig D, Parrott A, Coomber JA. Smoking cessation in
women: effects of the menstrual cycle. Int J Addict
1992;27:697-706.
11. Sellers EM, Kaplan HL, Tyndale RF. Inhibition of cyto-
chrome P450 2A6 increases nicotines oral bioavailabil-
ity and decreases smoking. Clin Pharmacol Ther 2000;
68:35-43.
12. Tyndale RF, Sellers EM. Genetic variation in CYP2A6-
mediated nicotine metabolism alters smoking behavior.
Ther Drug Monit 2002;24:163-71.
13. Benowitz NL, Jacob P III. Metabolism of nicotine to
cotinine studied by a dual stable isotope method. Clin
14. Benowitz NL, Jacob P III, Fong I, Gupta S. Nicotine
metabolic prole in man: comparison of cigarette smok-
ing and transdermal nicotine. J Pharmacol Exp Ther
1994;268:296-303.
gashima K, Inoue K, et al. Role of human cytochrome
P4502A6 in C-oxidation of nicotine. Drug Metab Dispos
1996;24:1212-7.
gashima K, Inoue K, et al. Characterization of CYP2A6
involved in 3-hydroxylation of cotinine in human liver
microsomes. J Pharmacol Exp Ther 1996;277:1010-5.
17. Dempsey D, Jacob P III, Benowitz NL. Accelerated
metabolism of nicotine and cotinine in pregnant smokers.
18. Rebagliato M, Bolumar F, Florey Cdu V, Jarvis MJ,
Perez-Hoyos S, Hernandez-Aguado I, et al. Variations in
cotinine levels in smokers during and after pregnancy.
Am J Obstet Gynecol 1998;178:568-71.
19. Selby P, Hackman R, Kapur B, Klein J, Koren G.
Heavily smoking women who cannot quit in pregnancy:
evidence of pharmacokinetic predisposition. Ther Drug
Monit 2001;23:189-91.
20. Benowitz NL, Swan GE, Lessov CN, Jacob P, III. Oral
contraceptives induce CYP2A6 activity and accelerate
nicotine metabolism [abstract]. Clin Pharmacol Ther
2004;75:P36.
21. Zeman MV, Hiraki L, Sellers EM. Gender differences in
tobacco smoking: higher relative exposure to smoke than
nicotine in women. J Womens Health Gend Based Med
2002;11:147-53.
22. Benowitz NL, Gourlay SG. Cardiovascular toxicity of
nicotine: implications for nicotine replacement therapy.
J Am Coll Cardiol 1997;29:1422-31.
23. Pickworth WB, Fant RV. Endocrine effects of nicotine
administration, tobacco and other drug withdrawal in
humans. Psychoneuroendocrinology 1998;23:131-41.
24. Goldstein DS, Levinson P, Keiser HR. Plasma and uri-
nary catecholamines during the human ovulatory cycle.
Am J Obstet Gynecol 1983;146:824-9.
25. Minson CT, Halliwill JR, Young TM, Joyner MJ. Inu-
ence of the menstrual cycle on sympathetic activity,
baroreex sensitivity, and vascular transduction in young
women. Circulation 2000;101:862-8.
26. Cooke WH, Ludwig DA, Hogg PS, Eckberg DL, Con-
vertino VA. Does the menstrual cycle inuence the sen-
sitivity of vagally mediated baroreexes? Clin Sci (Lond)
2002;102:639-44.
27. Jacob P, III, Yu L, Wilson M, Benowitz NL. Selected ion
monitoring method for determination of nicotine, cotin-
ine and deuterium-labeled analogs: absence of an isotope
effect in the clearance of (S)-nicotine-3,3-d2 in humans.
Biol Mass Spectrom 1991;20:247-52.
28. Jacob P, III, Yu L, Benowitz NL. Determination of
nicotine, its major metabolites, and deuterium-labeled
isotopomers in human urine using LC-MS/MS [abstract].
Drug Metab Rev 2002;34:160.
29. Gourlay SG, Benowitz NL. Arteriovenous differences in
plasma concentration of nicotine and catecholamines and
related cardiovascular effects after smoking, nicotine na-
sal spray, and intravenous nicotine. Clin Pharmacol Ther
1997;62:453-63.
30. Benowitz NL, Jacob P, III. Nicotine and cotinine elimi-
nation pharmacokinetics in smokers and nonsmokers.
31. Lee BL, Benowitz NL, Jacob P, III. Inuence of tobacco
abstinence on the disposition kinetics and effects of nic-
otine. Clin Pharmacol Ther 1987;41:474-9.
32. Kashuba AD, Nafziger AN. Physiological changes dur-
ing the menstrual cycle and their effects on the pharma-
cokinetics and pharmacodynamics of drugs. Clin Phar-
macokinet 1998;34:203-18.
33. McLean C, Graham TE. Effects of exercise and thermal
stress on caffeine pharmacokinetics in men and eumen-
orrheic women. J Appl Physiol 2002;93:1471-8.
34. Zaigler M, Rietbrock S, Szymanski J, Dericks-Tan JS,
Staib AH, Fuhr U. Variation of CYP1A2-dependent caf-
feine metabolism during menstrual cycle in healthy
women. Int J Clin Pharmacol Ther 2000;38:235-44.
35. Matsuki S, Kotegawa T, Tsutsumi K, Nakamura K, Na-
kano S. Pharmacokinetic changes of theophylline and
amikacin through the menstrual cycle in healthy women.
J Clin Pharmacol 1999;39:1256-62.
36. Kashuba AD, Bertino JS, Jr. Kearns GL, Leeder JS,
James AW, Gotschall R, et al. Quantitation of three-
month intraindividual variability and inuence of sex and
menstrual cycle phase on CYP1A2, N-acetyl-
transferase-2, and xanthine oxidase activity determined
with caffeine phenotyping. Clin Pharmacol Ther 1998;
63:540-51.
37. Kamimori GH, Joubert A, Otterstetter R, Santaromana
M, Eddington ND. The effect of the menstrual cycle on
the pharmacokinetics of caffeine in normal, healthy eu-
menorrheic females. Eur J Clin Pharmacol 1999;55:
445-9.
38. Lane JD, Steege JF, Rupp SL, Kuhn CM. Menstrual
cycle effects on caffeine elimination in the human fe-
male. Eur J Clin Pharmacol 1992;43:543-6.
39. Bruguerolle B, Toumi M, Faraj F, Vervloet D, Razzouk
H. Inuence of the menstrual cycle on theophylline phar-
macokinetics in asthmatics. Eur J Clin Pharmacol 1990;
39:59-61.
40. Balogh A, Irmisch E, Klinger G, Splinter FK, Hoffmann
A. Elimination of caffeine and metamizole in the men-
strual cycle of the fertile female [in German]. Zentralbl
Gynakol 1987;109:1135-42.
41. Nagata K, Ishitobi K, Yamamoto Y, Ikeda T, Hori S,
Matsumoto Y, et al. Increased theophylline metabolism
in the menstrual phase of healthy women. J Allergy Clin
Immunol 1997;100:39-43.
42. Meyer MC, Straughn AB, Mhatre RM, Shah VP, Chen
ML, Williams RL, et al. Variability in the bioavailability
of phenytoin capsules in males and females. Pharm Res
2001;18:394-7.
43. Backstrom T, Jorpes P. Serum phenytoin, phenobarbital,
carbamazepine, albumin; and plasma estradiol, proges-
terone concentrations during the menstrual cycle in
women with epilepsy. Acta Neurol Scand 1979;59:63-71.
44. Rosciszewska D, Buntner B, Guz I, Zawisza L. Ovarian
hormones, anticonvulsant drugs, and seizures during the
menstrual cycle in women with epilepsy. J Neurol
Neurosurg Psychiatry 1986;49:47-51.
45. Shavit G, Lerman P, Korczyn AD, Kivity S, Bechar M,
Gitter S. Phenytoin pharmacokinetics in catamenial epi-
lepsy. Neurology 1984;34:959-61.
46. Kumar N, Behari M, Ahuja GK, Jailkhani BL. Phenytoin
levels in catamenial epilepsy. Epilepsia 1988;29:155-8.
47. Kim MJ, Bertino JS, Jr. Gaedigk A, Zhang Y, Sellers
EM, Nafziger AN. Effect of sex and menstrual cycle
phase on cytochrome P450 2C19 activity with omepra-
zole used as a biomarker. Clin Pharmacol Ther 2002;72:
192-9.
48. McCune JS, Lindley C, Decker JL, Williamson KM,
Meadowcroft AM, Graff D, et al. Lack of gender differ-
ences and large intrasubject variability in cytochrome
P450 activity measured by phenotyping with dextro-
methorphan. J Clin Pharmacol 2001;41:723-31.
49. Labbe L, Sirois C, Pilote S, Arseneault M, Robitaille
NM, Turgeon J, et al. Effect of gender, sex hormones,
time variables and physiological urinary pH on apparent
CYP2D6 activity as assessed by metabolic ratios of
marker substrates. Pharmacogenetics 2000;10:425-38.
50. Kashuba AD, Nafziger AN, Kearns GL, Leeder JS,
Shirey CS, Gotschall R, et al. Quantication of intrain-
dividual variability and the inuence of menstrual cycle
phase on CYP2D6 activity as measured by dextrometho-
rphan phenotyping. Pharmacogenetics 1998;8:403-10.
51. Gerdin E, Rane A. N-demethylation of ethylmorphine in
pregnant and non-pregnant women and in men: an eval-
uation of the effects of sex steroids. Br J Clin Pharmacol
1992;34:250-5.
52. Llerena A, Cobaleda J, Martinez C, Benitez J. Interethnic
differences in drug metabolism: inuence of genetic and
environmental factors on debrisoquine hydroxylation
phenotype. Eur J Drug Metab Pharmacokinet 1996;21:
129-38.
53. Hogstedt S, Lindberg B, Peng DR, Regardh CG, Rane A.
Pregnancy-induced increase in metoprolol metabolism.
54. Wadelius M, Darj E, Frenne G, Rane A. Induction of
CYP2D6 in pregnancy. Clin Pharmacol Ther 1997;62:
400-7.
55. McAuley JW, Friedman CI. Inuence of endogenous
progesterone on alprazolam pharmacodynamics. J Clin
Psychopharmacol 1999;19:233-9.
56. Kirkwood C, Moore A, Hayes P, DeVane CL, Pelonero
A. Inuence of menstrual cycle and gender on alprazo-
lam pharmacokinetics. Clin Pharmacol Ther 1991;50:
404-9.
57. Kharasch ED, Mautz D, Senn T, Lentz G, Cox K. Men-
strual cycle variability in midazolam pharmacokinetics.
J Clin Pharmacol 1999;39:275-80.
58. Kashuba AD, Bertino JS, Jr. Rocci ML, Jr. Kulawy RW,
Beck DJ, Nafziger AN. Quantication of 3-month intra-
individual variability and the inuence of sex and men-
strual cycle phase on CYP3A activity as measured by
phenotyping with intravenous midazolam. Clin
59. Kamimori GH, Sirisuth N, Greenblatt DJ, Eddington ND.
The inuence of the menstrual cycle on triazolam and
indocyanine green pharmacokinetics. J Clin Pharmacol
2000;40:739-44.
60. Kharasch ED, Russell M, Garton K, Lentz G, Bowdle
TA, Cox K. Assessment of cytochrome P450 3A4 activ-
ity during the menstrual cycle using alfentanil as a non-
invasive probe. Anesthesiology 1997;87:26-35.
61. Kashuba AD, Nafziger AN, Kearns GL, Leeder JS,
Gotschall R, Gaedigk A, et al. Limitations of dextro-
methorphan N-demethylation as a measure of CYP3A
activity. Pharmacogenetics 1999;9:453-62.
62. Shah AK, Laboy-Goral L, Scott N, Morse T, Apseloff G.
Pharmacokinetics and safety of oral eletriptan during
different phases of the menstrual cycle in healthy volun-
teers. J Clin Pharmacol 2001;41:1339-44.
63. Lin Y, Anderson GD, Kantor E, Ojemann LM, Wilensky
AJ. Differences in the urinary excretion of 6-beta-hy-
droxycortisol/cortisol between Asian and Caucasian
women. J Clin Pharmacol 1999;39:578-82.
64. Burstein AH, Reiss WG, Kantor E, Anderson GD. Cy-
tochrome P450 3A4 activity in premenopausal and post-
menopausal women, based on 6-beta-hydroxycortisol:
cortisol ratios. Pharmacotherapy 1998;18:1271-6.
65. Zhu B, Liu ZQ, Chen GL, Chen XP, Ou-Yang DS, Wang
LS, et al. The distribution and gender difference of
CYP3A activity in Chinese subjects. Br J Clin Pharmacol
2003;55:264-9.
66. Sarkar MA, Vadlamuri V, Ghosh S, Glover DD. Expres-
sion and cyclic variability of CYP3A4 and CYP3A7
isoforms in human endometrium and cervix during the
menstrual cycle. Drug Metab Dispos 2003;31:1-6.
67. Burk O, Tegude H, Koch I, Hustert E, Wolbold R,
Glaeser H, et al. Molecular mechanisms of polymorphic
CYP3A7 expression in adult human liver and intestine.
J Biol Chem 2002;277:24280-8.
68. Altemus M, Roca C, Galliven E, Romanos C, Deuster P.
Increased vasopressin and adrenocorticotropin responses
to stress in the midluteal phase of the menstrual cycle.
J Clin Endocrinol Metab 2001;86:2525-30.
69. De Souza MJ, Maguire MS, Maresh CM, Kraemer WJ,
Rubin KR, Loucks AB. Adrenal activation and the pro-
lactin response to exercise in eumenorrheic and amenor-
rheic runners. J Appl Physiol 1991;70:2378-87.
70. Seki K. Variability of cortisol and adrenocorticotropic
hormone responses to metoclopramide during the men-
strual cycle. Gynecol Obstet Invest 1989;27:201-3.
71. Genazzani AR, Lemarchand-Beraud T, Aubert ML, Fel-
ber JP. Pattern of plasma ACTH, hGH, and cortisol
during menstrual cycle. J Clin Endocrinol Metab 1975;
41:431-7.
72. Burns JK. Proceedings: variation in plasma ACTH levels
during the human menstrual cycle. J Physiol 1975;249:
36P.
73. Stewart PM, Penn R, Holder R, Parton A, Ratcliffe JG,
London DR. The hypothalamo-pituitary-adrenal axis
across the normal menstrual cycle and in polycystic
ovary syndrome. Clin Endocrinol (Oxf) 1993;38:387-91.
74. Dunne FP, Barry DG, Ferriss JB, Grealy G, Murphy D.
Changes in blood pressure during the normal menstrual
cycle. Clin Sci (Lond) 1991;81:515-8.
75. Hassan AA, Carter G, Tooke JE. Postural vasoconstric-
tion in women during the normal menstrual cycle. Clin
Sci (Lond) 1990;78:39-47.
76. Zevin S, Jacob P, III, Benowitz NL. Nicotine-mecamyla-
mine interactions. Clin Pharmacol Ther 2000;68:58-66.
77. Fattinger K, Verotta D, Benowitz NL. Pharmacodynam-
ics of acute tolerance to multiple nicotinic effects in
humans. J Pharmacol Exp Ther 1997;281:1238-46.
78. Benowitz NL, Jacob P, III. Intravenous nicotine replace-
ment suppresses nicotine intake from cigarette smoking.
79. Andersson K, Arner P. Systemic nicotine stimulates hu-
man adipose tissue lipolysis through local cholinergic
and catecholaminergic receptors. Int J Obes Relat Metab
Disord 2001;25:1225-32.
80. Stalke J, Hader O, Bahr V, Hensen J, Scherer G, Oelkers
W. The role of vasopressin in the nicotine-induced stim-
ulation of ACTH and cortisol in men. Clin Investig
1992;70:218-23.
81. Newhouse PA, Sunderland T, Narang PK, Mellow AM,
Fertig JB, Lawlor BA, et al. Neuroendocrine, physio-
logic, and behavioral responses following intravenous
nicotine in nonsmoking healthy volunteers and in pa-
tients with Alzheimers disease. Psychoneuroendocrinol-
ogy 1990;15:471-84.
Effect of grapefruit juice volume on the
reduction of fexofenadine bioavailability:
Possible role of organic anion transporting
polypeptides
Objectives: The purpose of this study was to elucidate the potential clinical relevance and mechanism(s) of
action of 2 different volumes of grapefruit juice on the reduction of bioavailability of fexofenadine, a
substrate of organic anion transporting polypeptides.
Methods: Grapefruit juice or water at normal (300 mL) or high (1200 mL) volume was ingested concomi-
tantly with 120 mg fexofenadine by 12 healthy volunteers in a randomized 4-way crossover study, and
fexofenadine pharmacokinetics were determined over a period of 8 hours.
Results: The 300-mL volume of grapefruit juice decreased the mean area under the plasma drug concentra-
tiontime curve (AUC) and the peak plasma drug concentration of fexofenadine to 58% (P < .001) and 53%
(P < .001), respectively, of those with the corresponding volume of water, and 1200 mL grapefruit juice
reduced these parameters to 36% (P < .001) and 33% (P < .001), respectively, of those with the correspond-
ing volume of water. The 300-mL volume of grapefruit juice diminished the AUC of fexofenadine variably
among individuals. This decline correlated with baseline AUC of fexofenadine with water at equivalent
volume (r
2
0.97, P <.0001). The 1200-mL volume of grapefruit juice decreased the AUC of fexofenadine
more than the 300-mL volume of grapefruit juice compared with the corresponding volume of water in each
subject by a constant amount. Grapefruit juice, 300 mL and 1200 mL, reduced the coefficient of variation of
the AUC of fexofenadine by 2-fold compared with that with a matching volume of water.
Conclusions: Grapefruit juice at a commonly consumed volume diminished the oral bioavailability of fexo-
fenadine sufficiently to be pertinent clinically, likely by direct inhibition of uptake by intestinal organic anion
transporting polypeptide A (OATP-A; new nomenclature, OATP1A2). A much higher volume caused an
additional modest effect, possibly from reduced intestinal concentration and transit time of fexofenadine.
This food-drug interaction appears to be novel and may be relevant to other fruit juices and drugs. (Clin
Pharmacol Ther 2005;77:170-7.)
George K. Dresser, MD, PhD, Richard B. Kim, MD, and David G. Bailey, PhD London,
Ontario, Canada, and Nashville, Tenn
Clinical response to an identical dose of a medication
can uctuate substantially within individuals or differ
markedly among individuals, resulting in adverse events
ranging from lack of efcacy to appearance of toxicity.
Because the effect of a drug is most often linked to
concentration at the tissue site(s) of action, the outcome of
pharmacotherapy can be greatly dependent on the activity
of biologic processes that signicantly inuence the dis-
position of a drug. Moreover, variability in response
within and among individuals is frequently the result of
dissimilar activity of these processes, which are com-
From the Lawson Health Research Institute, London Health Sciences
Centre, and Departments of Medicine and Physiology and Phar-
macology, University of Western Ontario, London, and Division of
Clinical Pharmacology, Vanderbilt University School of Medicine,
Nashville.
Supported by a grant from the Internal Research Fund of London
Health Sciences Centre and US Public Health Service grants GM
31304 and GM 54724.
Presented in part at the Annual Meeting of the American Society for Clinical
Pharmacology and Therapeutics, Washington DC, April 2-5, 2003.
Received for publication May 18, 2004; accepted Oct 4, 2004.
Reprint requests: David G. Bailey, PhD, Department of Medicine,
London Health Sciences Centre, Victoria Campus, 375 South St,
London, Ontario, Canada N6A 4G5.
E-mail: david.bailey@lhsc.on.ca
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.005
170
monly regulated by environmental or genetic inuences.
These concepts are relevant and well established for drug-
metabolizing enzymes, such as cytochrome P450 (CYP)
3A4, which convert numerous medications to inactive or
less active derivatives, and for the efux transporter
P-glycoprotein, which limits intestinal absorption and tis-
sue distribution and facilitates the excretion of many
drugs.
1,2
Recently, molecular cloning and expression experi-
ments with members of the family of drug uptake
transporters known as organic anion transporting
polypeptides (OATPs) point to the likelihood that they
also play a crucial role in drug disposition.
3
In the small
intestine, OATPs appear to affect certain orally admin-
istered medications by enabling enteric absorption, a
key initial step for systemic action of a medication.
Consequently, modulation of the activity of intestinal
OATPs might be expected to alter the oral bioavailabil-
ity and subsequent clinical response of a drug.
The latest ndings have shown that constituents in
the diet reduced the activity of several OATPs.
4
In
vitro, substantially diluted concentrations of grapefruit,
orange, and apple juices and their specic ingredients
were potent inhibitors. In humans, a normal concentra-
tion of these juices at a high volume (1200 mL) pro-
foundly decreased the oral bioavailability of the anti-
histamine fexofenadine, a nonmetabolized substrate of
certain OATPs.
4-6
The purpose of this study was to reveal the potential
clinical relevance and mechanism(s) of this interaction
by use of grapefruit juice as a representative fruit juice
and by assessment of the effect of uid volume on the
single-dose oral pharmacokinetics of fexofenadine. A
randomized 4-way crossover study was conducted in
which 12 healthy volunteers received grapefruit juice or
water at normal (300 mL) or high (1200 mL) volume
with 120 mg fexofenadine. The 300-mL volume of
grapefruit juice was selected because it represents the
usual amount commercially available in individual
juice containers in North America and thus has practi-
cal clinical relevance, whereas the incorporation of the
1200-mL volume of grapefruit juice in the study served
as a reference point to our initial publication, which
used this large volume so that a possible effect would
not be missed. Results indicated that a routinely con-
sumed volume (300 mL) of grapefruit juice produced a
potentially clinically relevant reduction in the oral bio-
availability of fexofenadine, likely by a mechanism
involving direct inhibition of the inherent activity of
intestinal OATP-A (OATP1A2). A much higher vol-
ume (1200 mL) of juice caused a moderate additional
reduction, possibly from less exposure of fexofenadine
to this enteric transporter through diminished intestinal
drug concentration or transit time. This food-drug in-
teraction is likely to be original and applicable to other
fruit juices and drugs.
METHODS
Study population. We tested 12 subjects (7 men and
5 women; age range, 23-47 years). They were healthy
as determined by medical history, physical examina-
tion, and routine hematologic and serum chemical test-
ing. No subject had a signicant illness within the
preceding 2 weeks, had routinely used prescription or
over-the-counter medications, or had a history of car-
diac, renal, hepatic, or gastrointestinal disease or drug
or alcohol abuse. The Research Ethics Board for Health
Sciences Research Involving Human Subjects at the
University of Western Ontario (London, Ontario, Can-
ada) approved the study, and all subjects provided
written informed consent.
Experimental protocol. The study was a single-
dose, randomized, open, 4-way crossover, and con-
trolled clinical investigation. Subjects consumed 300
mL grapefruit juice (Everfresh, Windsor, Ontario, Can-
ada), 300 mL water, 1200 mL grapefruit juice (300 mL
grapefruit juice with drug followed by 150 mL every
0.5 hour until 3 hours after dosing), or 1200 mL water
(same schedule as for 1200 mL grapefruit juice) with
120 mg fexofenadine (Hoechst Marion Roussel Can-
ada, Inc, Laval, Quebec, Canada). Grapefruit juice was
purchased as a single lot in bottles containing 300 mL
juice at normal strength, which had been previously
reconstituted from concentrate.
Subjects fasted from 10 PM on the evening before
testing. Peripheral venous blood (5 mL) was sampled
into tubes that contained ethylenediaminetetraacetic
acid just before and at 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5,
4.0, 5.0, 6.0, and 8.0 hours after administration of
fexofenadine for measurement of plasma drug concen-
trations. A standardized lunch was provided at 4 hours
after dosing. All subjects were asked not to consume
citrus products or fruit juices for 1 week before and
throughout testing, with the exception of the grapefruit
juice provided in the study. They were required to
abstain from alcohol and not to take medications in-
cluding over-the-counter preparations for 48 hours be-
fore testing. Tobacco and caffeine-containing bever-
ages were not permitted during testing. The washout
period between testing was 1 week. Subjects were
asked to maintain a consistent diet throughout all
phases of testing.
Assay of plasma fexofenadine concentration. The
plasma concentration of fexofenadine was quantied as
2005;77(3):170-7 Grapefruit juice, fexofenadine, and OATP 171
outlined previously.
4
A 100-mg C18 preparatory solid-
phase extraction column (Sep-Pak Vac cartridge; Wa-
ters Corp, Mississauga, Ontario, Canada) was washed
with 1-mL volumes of isopropyl alcohol, methanol, and
water. A 500-L aliquot of aqueous 1% phosphoric
acid was added, followed by a 500-L aliquot of the
plasma sample for analysis. The mixture was then
passed through the column. The column was washed
once with a 1-mL volume of aqueous 1% phosphoric
acid water, 3 times with a 1-mL volume of water/
methanol/glacial acetic acid (88:10:2 [vol/vol]), and 3
times with a 1-mL volume of water/methanol/concen-
trated ammonium hydroxide (88:10:2 [vol/vol]). The
sample was eluted with 1 mL of methanol/triethylamine
(99.8:0.2 [vol/vol]), and the efuent was evaporated to
dryness at 40C under a gentle stream of nitrogen. The
residue was dissolved in a 100-L aliquot of HPLC
mobile phase that consisted of acetonitrile/water/trieth-
ylamine (33:67:1 [vol/vol] at pH 3.0) with phosphoric
acid. Recovery of fexofenadine from the solid-phase
extraction column was complete. The solution was l-
tered (0.45 m), and a sample (40 L) was injected
onto a Prodigy 5-m octadecylsilane (2) column (150
mm 3.2 mm; Phenomenex, Torrance, Calif) with a
mobile ow rate of 0.4 mL/min. Fluorescence detection
(excitation wavelength, 223 nm; emission wavelength,
290 nm) was used to monitor the efuent. The retention
time of fexofenadine was 10.9 minutes. The standard
curve of fexofenadine was linear over the range of
concentrations measured in the plasma samples from
the study (0-1000 ng/mL). The coefcient of variation
was 4.8% at 25 ng/mL (n 5), and the limit of
detection was 5 ng/mL.
Data analysis. Plasma fexofenadine concentrations
were analyzed by use of a noncompartmental model
method. The terminal elimination rate constant (k
e
) was
determined by log-linear regression of the nal data
points (at least 3). The apparent elimination half-life of
the log-linear phase (t
) was calculated as 0.693/k

e
.
The area under the plasma drug concentrationtime
curve from 0 to 8 hours [AUC(0-8)] was calculated by
use of the linear trapezoidal method. The AUC from 8
hours to innity [AUC(8-)] was determined by divid-
ing the nal plasma fexofenadine concentration by k
e
.
Peak plasma drug concentration (C
max
) and the time to
reach C
max
(t
max
) were obtained directly from the ex-
perimental data.
Statistical comparisons among the 4 treatment
groups initially used ANOVA for repeated measures.
For those analyses where P was less than .05, a priori
comparisons were performed between grapefruit juice
versus water treatments for 300-mL or 1200-mL vol-
umes and between 300-mL volumes versus 1200-mL
volumes for grapefruit juice or water treatments by use
of the paired t test. Statistical signicance was deter-
mined by the Bonferroni method by correcting for the
number of comparisons (ie, 0.05/4 0.0125). The
coefcient of variation was expressed as a percent of
SD/mean. Data are presented as mean values SEM.
RESULTS
Grapefruit juice, 300 mL or 1200 mL, produced
lower mean plasma concentrations, AUC, and C
max
of
fexofenadine compared with the corresponding volume
of water (Fig 1 and Table I). The 300-mL volume of
grapefruit juice reduced mean AUC and C
max
of fexo-
fenadine to 58% and 53%, respectively. The 1200-mL
volume of grapefruit juice lowered these parameters to
36% and 33%, respectively. Grapefruit juice or water at
300 mL or 1200 mL did not produce different t
max
or t
of fexofenadine. The 300-mL volume of grapefruit

juice had higher mean plasma concentrations, AUC,
and C
max
of fexofenadine than 1200 mL grapefruit
juice.
The 300-mL volume of grapefruit juice produced an
interindividual reduction in the AUC of fexofenadine
that was variable; a greater reduction in the AUC of
fexofenadine with 300 mL grapefruit juice occurred
with a higher baseline AUC of fexofenadine with an
equivalent volume of water (Fig 2). The 1200-mL
volume of grapefruit juice caused a moderately larger
reduction compared with 300 mL grapefruit juice rel-
ative to the corresponding volume of water; this addi-
Fig 1. Mean plasma drug concentrationtime curves of 120
mg fexofenadine for individuals (N 12) administered water
or grapefruit juice (GFJ), 300 mL or 1200 mL (300 mL with
drug followed by 150 mL every 0.5 hour until 3.0 hours).
Bars represent SEM.
172 Dresser, Kim, and Bailey MARCH 2005
tional interindividual decrease in the AUC of fexofe-
nadine was constant.
The 300-mL volume of water produced mean plasma
concentrations and pharmacokinetics of fexofenadine
that were not different from those with 1200 mL water.
However, 3 individuals with the highest AUC of fexo-
fenadine with 300 mL water had a decrease in this
parameter with 1200 mL water (Fig 3).
Both 300 mL and 1200 mL grapefruit juice markedly
reduced the variability of the AUC of fexofenadine
among individuals compared with the matching volume
of water (Fig 4). The 300-mL volume of grapefruit
juice had a coefcient of variation of 24% compared
with 49% as calculated for 300 mL water. The
1200-mL volume of grapefruit juice had a coefcient of
variation of 20% relative to 39% as calculated for 1200
mL water.
DISCUSSION
Because our previous pharmacokinetic interaction
study showed that grapefruit juice at a high volume
(1200 mL) markedly diminished the oral bioavailability
of fexofenadine to 35% of that seen with an equivalent
volume of water,
4
it was hypothesized that grapefruit
juice at a more commonly used volume (300 mL)
would also generate a potentially pertinent outcome.
Results showed that 300 mL grapefruit juice lowered
the mean oral bioavailability of fexofenadine to 58%
compare with 300 mL water. Thus concomitant grape-
fruit juice at a frequently consumed amount may cause
an unwanted decrease or duration of efcacy of fexo-
fenadine. Indeed, the reduction of the administered
dose of fexofenadine from 130 mg to 40 mg substan-
tially shortened the duration of clinical efcacy of
fexofenadine as measured by the single-dose change in
the wheal areatime prole.
7
Establishing the cause of this interaction necessitated
deliberation on major aspects that determine the oral
bioavailability of a drug, which requires that the chem-
ically unchanged form of a drug undergo a number of
essential sequential events. The drug must undergo
dissolution from the tablet matrix, transfer from the
stomach to the small intestine, passage through the
small intestinal wall into the portal circulation, and then
conveyance through the liver.
Because fexofenadine is a zwitterion, it carries an
ionic charge and has high aqueous solubility over a
Fig 2. Area under plasma drug concentrationtime curve
from 0 to 8 hours [AUC(0-8)] of 120 mg fexofenadine with
water, 300 mL or 1200 mL (300 mL with drug followed by
150 mL every 0.5 hour until 3.0 hours), plotted against
absolute change in this parameter with matching volume of
GFJ for each individual (N 12). Horizontal dashed line
indicates no change in fexofenadine AUC(0-8). The lines of
best t (solid lines) were determined by linear regression
analysis. Dashed lines with arrows indicate the variable and
constant components of the interaction.
Table I. Plasma pharmacokinetics of 120 mg fexofenadine administered with water or grapefruit juice, 300 mL or
1200 mL (300 mL with drug followed by 150 mL every 0.5 hour until 3.0 hours [total volume of 1200 mL])
Water Grapefruit juice
300 mL 1200 mL 300 mL 1200 mL
AUC(0-8) (ng h/mL) 1685 237 1379 155 980 68* 491 28
AUC(0-) (ng h/mL) 2167 283 1747 184 1379 122* 677 41
C
max
(ng/mL) 436 74 326 37 233 25* 109 8
t
max
(h) 2.0 0.4 2.1 0.3 3.3 0.6 2.9 0.4
t
(h) 3.0 0.5 3.2 0.2 3.1 0.2 3.5 0.2

Data are expressed as mean SEM.
AUC(0-8), Area under plasma drug concentrationtime curve from 0 to 8 hours; AUC(0-), area under plasma drug concentrationtime curve from 0 hours
extrapolated to innity; C
max
, peak plasma drug concentration; t
max
, time to reach peak plasma drug concentration; t
, elimination half-life.
*P .01 for 300 mL versus 1200 mL for water or grapefruit juice.
P .001 for water versus grapefruit juice for 300 mL or 1200 mL at equivalent volume.
broad pH range. Consequently, it seems unlikely that
the acidic pH of grapefruit juice would decrease the
dissolution of fexofenadine in the gastrointestinal tract.
Moreover, a greater volume of juice might be expected
to favor dissolution of fexofenadine. Given that 1200
mL grapefruit juice diminished the oral bioavailability
of fexofenadine more than 300 mL grapefruit juice, it
seemed unlikely that this juice acted primarily by ad-
versely affecting dissolution of this drug.
Fexofenadine is eliminated essentially unchanged,
and the t
max
values of this drug were similar among
treatments.
6
Thus it is doubtful that chemical instabil-
ity, changed drug metabolism, and altered transit to the
intestinal site of absorption were causes for the inter-
action.
The physicochemical properties of fexofenadine of a
high degree of ionization and large polar surface area
would predict negligible or low passive intestinal per-
meability.
8,9
Because this process is traditionally con-
sidered to be of primary importance for the passage of
drug from the small intestine into the portal circulation,
fexofenadine might be envisaged to have minimal or
minor intestinal absorption. Moreover, fexofenadine
encounters efux transport by P-glycoprotein in the
enterocytes.
5
Yet the absolute oral bioavailability of
fexofenadine is estimated to be 33% in humans.
10
Re-
cent ndings indicate that there is substantial expres-
sion of OATP-A, which enables uptake transport of
fexofenadine, on the luminal membrane of human in-
testinal enterocytes.
11
In addition, OATP-B (OATP2B1)
has been noted to be capable of mediating the cellular
uptake of fexofenadine and is expressed in the intes-
tine.
12,13
However, we have not been able to show that
cells overexpressing OATP-B cause signicant uptake
transport of fexofenadine (R.B.K., unpublished data,
September 2004). Although it is possible that both
OATP-A and OATP-B may play a role in fexofenadine
absorption from the intestine, OATP-A appears to have
far greater afnity for fexofenadine and is most likely
the major determinant of the extent of absorption of
fexofenadine into the portal circulation.
In the liver, OATPs are expressed on the sinusoidal
membrane of hepatocytes.
3
However, the OATP-A iso-
form has not been detected there. P-glycoprotein is
located on the bile canalicular membrane. Given that
biliary secretion is a major route for elimination of
fexofenadine, this might reect the cooperative action
of another OATP isoform plus P-glycoprotein. How-
ever, verapamil and ketoconazole, potent inhibitors of
P-glycoprotein, noticeably enhanced the systemic
availability of fexofenadine apparently by selectively
impairing biliary excretion.
14,15
Thus the biliary excre-
tion of fexofenadine might be mainly dependent on
efux transport mediated by P-glycoprotein.
In this investigation, grapefruit juice decreased the
AUC and C
max
but did not affect the t
max
or elimination
t
of fexofenadine. These data comprise the primary

rationale for concluding that the interaction resulted
from limiting oral drug bioavailability. In vitro, grape-
fruit juice at only 0.5% normal strength decreased
OATP-Amediated cellular uptake transport of fexofe-
nadine by half.
4
In contrast, grapefruit juice at 10-fold
higher concentration had no effect on cellular mono-
layer P-glycoproteinmediated efux transport of
digoxin. This highly potent and selective effect by
grapefruit juice on the activity of OATP-A might be the
key to understanding the mechanism of this interaction
clinically. Given that inhibition of hepatic OATP-
mediated fexofenadine uptake would be expected to
enhance plasma concentrations of this drug, the most
compelling explanation for the mechanism of this in-
teraction is primarily reduction in the amount of fexo-
fenadine undergoing active uptake transport mediated
by intestinal OATP-A.
The quantity of fexofenadine that is normally intes-
tinally absorbed might depend on the inherent activity
of OATP-A. Given that variability is an intrinsic prop-
erty of biologic systems, it might be expected that the
activity of intestinal OATP-A transporter would range
substantially among individuals. Low innate activity
might be expected to produce a small AUC for fexo-
fenadine. Conversely, high innate activity might be
Fig 3. AUC(0-8) of 120 mg fexofenadine with 300 mL water
plotted against absolute change in this parameter with 1200
mL water (300 mL with drug followed by 150 mL every 0.5
hour until 3.0 hours) for each individual (N 12). Horizontal
dashed line indicates no change in fexofenadine AUC(0-8).
The line of best t (solid line) was determined by linear
regression analysis.
envisaged to cause an opposite outcome. Moreover,
inhibition of low activity would be predicted to produce
less effect than inhibition of high activity. Given that
the 4-fold variation of baseline AUC of fexofenadine
with water strongly inversely correlated with the dim-
inution of this parameter with grapefruit juice among
individuals, the interaction is consistent with an effect
on innate intestinal OATP-A. In addition, specic con-
stituents in grapefruit juice including the furanocouma-
rins bergamottin and 6,7-dihydroxybergamottin and
the bioavonoid naringin are normally present in this
juice at concentrations that are 5-, 8-, and 150-fold
higher, respectively, than those shown to cause at least
50% inhibition of in vitro activity of oatp3, the rat
intestinal ortholog of human OATP-A.
4,16
Thus data
from 2 sources point to 1 or more specic ingredients in
grapefruit juice directly inhibiting the inherent activity
of intestinal OATP-A as one basis for the pharmacoki-
netic interaction.
Recent ndings show that 300 mL grapefruit juice
can lower the AUC and C
max
of fexofenadine even
when juice is consumed 2 hours before administration
of this drug.
17
The magnitude of reduction was essen-
tially the same as that with concomitant ingestion of
juice. Given that simultaneous consumption of grape-
fruit juice produced a t
max
of fexofenadine of 2.2 0.2
hours (mean SEM), it appears unlikely that grape-
fruit juice would be left in the stomach at this time.
Furthermore, ingestion of juice 2 hours earlier would
most likely result in a negligible amount remaining in
the small intestine at the time of peak absorption of
fexofenadine. Data from other sources reported that the
clearance of saline solution and 25% dextrose from the
stomach ranged from 30 to 90 minutes.
18-20
Thus re-
sults indicate that specic ingredients in grapefruit juice
might have residual inhibitory action on the activity of
enteric OATP-A.
Grapefruit juice is known to cause prolonged inhibi-
tion of intestinal drug metabolism mediated by
CYP3A4.
21-24
Furanocoumarins in grapefruit juice are
considered to be metabolized by CYP3A4 in vivo to
reactive intermediates that then bond covalently and
irreversibly inactivate this enzyme, a process termed
suicide or mechanism-based inhibition.
25-34
Moreover,
these same furanocoumarins can inhibit in vitro activ-
ities of oatp3.
4
Consequently, reactive intermediates
may inactivate both CYP3A4 and OATP-A in entero-
cytes. Altered trafcking of enteric OATP-A from the
membrane into the cytosol might also occur. Either
mechanism might explain the residual effect of grape-
fruit juice on the oral absorption of fexofenadine and
require testing.
Two substantially different volumes of grapefruit
juice were tested, and dissimilar quantitative responses
were observed. The 300-mL and 1200-mL volumes of
grapefruit juice diminished the mean oral bioavailabil-
ity of fexofenadine to 58% and 36% of those observed
with equivalent volumes of water, respectively. Given
that a 4-fold higher volume of grapefruit juice reduced
the AUC of fexofenadine only moderately more, 300
mL grapefruit juice may be near the upper portion of
the volume-response curve. Consequently, a volume of
grapefruit juice of less than 300 mL may also bring
about a clinically relevant reduction in the oral bio-
availability of fexofenadine.
The 300-mL volume of grapefruit juice created an
interindividual decrease in the AUC of fexofenadine
that was variable. The 1200-mL volume of grapefruit
juice resulted in an additional reduction that was con-
stant. These divergent responses may reect dissimilar
mechanisms. The smaller volume of grapefruit juice
brought about the majority of the effect, probably by
near-maximum direct inhibition of the intrinsic activity
of intestinal OATP-A. The substantially higher volume
of juice added a supplementary smaller effect. More-
over, certain individuals who initially absorbed fexofe-
nadine well with 300 mL water had a decline in the oral
bioavailability of fexofenadine with 1200 mL water.
Thus the 1200-mL volume containing more solutes
(nonabsorbed carbohydrates or salts) and osmotic ef-
fects or just more uid might decrease intestinal con-
Fig 4. AUC(0-8) of 120 mg fexofenadine with water, 300
mL or 1200 mL (300 mL with drug followed by 150 mL
every 0.5 hour until 3.0 hours), plotted against same param-
eter with matching volume of GFJ for each individual (N
12). The solid lines with end brackets indicate the observed
range of values for each treatment. The dashed line is the line
of identity.
centration or transit time of drug. Under linear transport
kinetics, a diminution of either of these parameters
would be calculated to cause a proportional decline in
the amount of fexofenadine undergoing intestinal up-
take transport. Thus it seems reasonable to speculate
that markedly additional uid in the gastrointestinal
tract, whether through more solutes and enhanced os-
motic effect or simply greater intake of total volume,
might diminish the concentration or transit time of
fexofenadine sufciently to affect the intestinal absorp-
tion of this drug adversely.
Grapefruit juice halved the coefcient of variation in
the AUC of fexofenadine among individuals compared
with the equivalent volume of water. Despite the un-
derstandable disadvantage of an observed reduction in
the oral bioavailability of a drug, the better reliability of
this parameter with grapefruit juice may improve the
dose-response relationship of fexofenadine among and
within individuals. In other words, it may make the
effect of a drug more predictable. This approach may
have some application for medications that have vari-
able oral pharmacokinetics and a narrow range of
plasma concentrations between therapeutic efcacy and
toxicity. For example, grapefruit juice has been shown
to produce a relevant decrease in the plasma concen-
trations of the chemotherapeutic agent etoposide.
35
If
the mechanism of this interaction with etoposide were
found to be inhibition of intestinal OATP-A, this ap-
proach might have some practical value. However, a
reduction in the coefcient of variation of the AUC of
fexofenadine by grapefruit juice may not be a consis-
tent nding.
4
Consequently, additional research would
need to be conducted.
In conclusion, a commonly consumed volume (300
mL) of grapefruit juice produced a diminution of oral
bioavailability of the drug probe fexofenadine of suf-
cient magnitude to be pertinent clinically, potentially
resulting in reduced benet of drug. The mechanism
appeared to be unique. The most convincing explana-
tion was that this volume of grapefruit juice contained
sufcient specic active ingredients to produce a direct
and prolonged inhibitory effect on inherent activity of
intestinal OATP-A. Because it caused a near-maximum
reduction in the amount of fexofenadine undergoing
active uptake transport, a marked interaction may be
apparent also at a lower volume of this juice. Ingestion
of a markedly higher quantity (1200 mL) of grapefruit
juice brought about a modest additional effect, possibly
as a result of a mechanism involving lessened exposure
of fexofenadine to intestinal OATP-A through reduced
concentration and transit time of drug in the gastroin-
testinal tract. Despite decreased systemic availability,
grapefruit juice created a more standard relationship
between dose and oral bioavailability of fexofenadine
in this study, supporting the importance of enteric
OATP-A as a determinant of the variability of oral
bioavailability of fexofenadine among subjects and
suggesting an approach to obtain a more predictable
clinical effect for drugs possessing a narrow therapeutic
index.
None of the authors has any nancial or other relationships that
could lead to conicts of interest.
References
1. Dresser GK, Spence JD, Bailey DG. Pharmacokinetic-
pharmacodynamic consequences and clinical relevance
of cytochrome P450 3A4 inhibition. Clin Pharmacokinet
2000;38:41-57.
2. Lin JH, Yamazaki M. Role of P-glycoprotein in pharma-
cokinetics: clinical implications. Clin Pharmacokinet
2003;42:59-98.
3. Kim RB. Transporters and drug disposition. Curr Opin
Drug Discov Deliv 2000;3:94-101.
4. Dresser GK, Bailey DG, Leake BF, Schwarz UI, Dawson
PA, Freeman D, et al. Fruit juices inhibit organic anion
transporting polypeptide-mediated drug uptake to de-
crease the oral availability of fexofenadine. Clin Pharma-
col Ther 2002;71:11-20.
5. Cvetkovic M, Leak B, Fromm MF, Wilkinson GR, Kim
RB. OATP and P-glycoprotein transporters mediate the
cellular uptake and excretion of fexofenadine. Drug
Metab Dispos 1999;27:866-71.
6. Lippert C, Ling J, Brown P, Burmaster S, Eller M, Cheng
L, et al. Mass balance and pharmacokinetics of MDL
16455A in healthy male volunteers [abstract]. Pharm Res
1995;12:S-390.
7. Russell T, Stoltz M, Weir S. Pharmacokinetics, pharma-
codynamics, and tolerance of single- and multiple-dose
fexofenadine hydrochloride in healthy male volunteers.
8. Winiwarter S, Bonham NM, Ax F, Hallberg A, Lenner-
nas H, Karlen A. Correlation of jejunal permeability (in
vivo) of drugs with experimentally and theoretically de-
rived parameters. A multivariate data analysis approach.
J Med Chem 1998;41:4939-49.
9. Kelder J, Grootenhuis PD, Bayada DM, Delbressine LP,
Ploemen JP. Polar molecular surface as a dominating
determinant for oral absorption and brain penetration of
drugs. Pharm Res 1999;16:1514-9.
10. Fexofenadine [product information]. Laval, Quebec
(Canada): Hoechst Marion Roussel Canada, Inc.
11. Smith LH, Lee W, Kim RB. Differential expression of
OATP drug uptake transporters in human liver, intestine,
and kidney [abstract]. Drug Metab Rev 2003;35:73.
12. Nozawa T, Imai K, Nezu J, Tsuji A, Tamai I. Functional
characterization of pH-sensitive organic anion transport-
ing polypeptide OATP-B in humans. J Pharmacol Exp
Ther 2004;308:438-45.
13. Kullak-Ublick GA, Ismair MG, Stieger B, Landmann L,
Huber R, Pizzagalli F, et al. Organic anion-transporting
polypeptide B (OATP-B) and its functional comparison
with three other OATPs of human liver. Gastroenterol-
ogy 2001;120:525-33.
14. Tannergren C, Petri N, Knutson L, Hedeland M, Bond-
esson U, Lennernas H. Multiple transport mechanisms
involved in the intestinal absorption and rst-pass extrac-
tion of fexofenadine. Clin Pharmacol Ther 2003;74:423-
36.
15. Tannergren C, Knutson T, Knutson L, Lennernas H. The
effect of ketoconazole on the in vivo intestinal perme-
ability of fexofenadine using a regional perfusion tech-
nique. Br J Clin Pharmacol 2003;55:182-90.
16. Walters HC, Craddock AL, Fusegawa H, Willingham
MC, Dawson PA. Expression, transport properties, and
chromosomal location of organic anion transporter sub-
type 3. Am J Physiol Gastrointest Liver Physiol 2000;
279:G1188-200.
17. Dresser GK, Kim RB, Bailey DG. Duration of grapefruit
juice effect on fexofenadine interaction [abstract]. Clin
Pharmacol Ther 2003;73:P48.
18. Collins PJ, Houghton LA, Read NW, Horowitz M, Chat-
terton BE, Heddle R, et al. Role of the proximal and
distal stomach in mixed solid and liquid meal emptying.
Gut 1991;32:615-9.
19. Husebye E. The patterns of small bowel motility: phys-
iology and implications in organic disease and functional
disorders. Neurogastroenterol Motil 1999;11:141-61.
20. Ouyang A, Sunshine AG, Reynolds. Caloric content of a
meal affects duration but not contractile pattern of duo-
denal motility in man. Dig Dis Sci 1989;34:528-36.
21. Lundahl J, Regardh CG, Edgar B, Johnsson G. Relation-
ship between time of intake of grapefruit juice and its
effect on pharmacokinetics and pharmacodynamics of
felodipine in healthy subjects. Eur J Clin Pharmacol
1995;49:61-7.
22. Takanaga H, Ohnishi A, Murakami H, Matsuo H, Higu-
chi S, Urae A, et al. Relationship between time after
intake of grapefruit juice and the effect on the pharma-
cokinetics and pharmacodynamics of nisoldipine in
healthy subjects. Clin Pharmacol Ther 2000;67:201-14.
23. Lilja JJ, Kivisto KT, Neuvonen PJ. Duration of effect of
grapefruit juice on the pharmacokinetics of the CYP3A4
substrate simvastatin. Clin Pharmacol Ther 2000;68:384-
90.
24. Greenblatt DJ, von Moltke LL, Harmatz JS, Chen G,
Weemhoff JL, Jen C, et al. Time course of recovery of
cytochrome P450 3A function after single doses of
grapefruit juice. Clin Pharmacol Ther 2003;74:121-9.
25. Bailey DG, Dresser GK, Kreeft JH, Munoz C, Freeman
DJ, Bend JR. Grapefruit-felodipine interaction: effect of
unprocessed fruit and probable active ingredients. Clin
26. Lown KS, Bailey DG, Fontana RJ, Janardan SK, Adair
CH, Fortlage LA, et al. Grapefruit juice increases felo-
dipine oral availability in humans by decreasing intestinal
CYP3A protein expression. J Clin Invest 1997;99:2545-
53.
27. Edwards DJ, Bellevue FH, Woster PM. Identication of
6,7-dihydroxybergamottin, a cytochrome P450 inhibi-
tor, in grapefruit juice. Drug Metab Dispos 1996;24:
1287-90.
28. Schmiedlin-Ren P, Edwards DJ, Fitzsimmons ME, He K,
Lown KS, Woster PM, et al. Mechanisms of enhanced
oral availability of CYP3A4 substrates by grapefruit con-
stituents. Decreased enterocyte CYP3A4 concentration
and mechanism-based inactivation by furanocoumarins.
29. Fukuda K, Ohta T, Oshima Y, Ohashi N, Yoshikawa M,
Yamazoe Y. Specic CYP3A4 inhibitors in grapefruit
juice: furocoumarin dimers as components of drug inter-
action. Pharmacogenetics 1997;7:391-6.
30. He K, Iyer KR, Hayes RN, Sinz MW, Woolf TF, Hol-
lenberg PF. Inactivation of cytochrome P450 3A4 by
bergamottin, a component of grapefruit juice. Chem Res
Toxicol 1998;11:252-9.
31. Ohnishi A, Matsuo H, Yamada S, Takanaga H, Morimoto
S, Shoyama Y, et al. Effect of furanocoumarin deriva-
tives in grapefruit juice on the uptake of vinblastine by
Caco-2 cells and on the activity of cytochrome P450
3A4. Br J Pharmacol 2000;130:1369-77.
32. Tassaneeyakul W, Guo LQ, Fukuda K, Ohta T, Yamazoe
Y. Inhibition selectivity of grapefruit juice components
on human cytochromes P450. Arch Biochem Biophys
2000;378:356-63.
33. Guo LQ, Fukuda K, Ohta T, Yamazoe Y. Role of furano-
coumarin derivatives on grapefruit juice-mediated inhi-
bition of human CYP3A activity. Drug Metab Dispos
2000;28:766-71.
34. Ohta T, Nagahashi M, Hosoi S, Tsukamoto S. Dihy-
droxybergamottin caproate as a potent and stable
CYP3A4 inhibitor. Bioorg Med Chem 2002;10:969-73.
35. Reif S, Nicolson MC, Bisset D, Reid M, Kloft C, Jaehde
U, et al. Effect of grapefruit juice intake on etoposide
bioavailability. Eur J Clin Pharmacol 2002;58:491-4.
Inhibition of human intestinal wall
metabolism by macrolide antibiotics: Effect
of clarithromycin on cytochrome P450
3A4/5 activity and expression
Background: Clarithromycin increases both hepatic and intestinal availability of the selective cytochrome P450
(CYP) 3A probe midazolam. This study was designed to identify determinants of variability in the extent of
intestinal wall CYP3Ainhibition by clarithromycin, such as CYP3A5 genotype, and the mechanismof inhibition.
Methods: Ten healthy volunteers received 500 mg oral clarithromycin twice a day for 7 days. Before and after
administration of clarithromycin, small-bowel mucosal biopsy specimens were obtained endoscopically.
Intestinal CYP3A activity was determined from the rate of 1-hydroxymidazolam and 4-hydroxymidazolam
formation by incubation of small-bowel homogenate with midazolam (25 mol/L) and NADPH for 5
minutes. Intestinal CYP3A4 and CYP3A5 messenger ribonucleic acid was quantified by real-time reverse
transcriptasepolymerase chain reaction. Intestinal CYP3A4 and CYP3A5 protein concentrations were de-
termined by immunoblotting. Serum and homogenate concentrations of midazolam, clarithromycin, and
metabolites were determined by liquid chromatographymass spectrometry. CYP3A5 genotype was deter-
mined by real-time polymerase chain reaction.
Results: The formation of 1-hydroxymidazolam (1.36 0.46 pmol min
1
mg
1
at baseline versus 0.35
0.16 pmol min
1
mg
1
after administration) and 4-hydroxymidazolam (0.39 0.12
pmol min
1
mg
1
at baseline versus 0.12 0.05 pmol min
1
mg
1
after administration) was
significantly (P <.001) reduced after clarithromycin administration. Clarithromycin administration did not
result in a significant change in intestinal CYP3A4 and CYP3A5 messenger ribonucleic acid and protein
expression. All subjects had detectable serum clarithromycin concentrations after 7 days of clarithromycin
(3.71 2.43 mol/L). The mean concentration of clarithromycin in the intestinal biopsy homogenate was
1.2 0.7 nmol/L (range, 0.42-2.39 nmol/L). Compared with CYP3A5 nonexpressers, subjects with at least
1 CYP3A5*1 allele (CYP3A5 expressers) had greater inhibition of intestinal CYP3A activity after treatment
with clarithromycin. There was a strong linear relationship between the decrease in intestinal CYP3A activity
and baseline catalytic activity (R
2
0.9).
Conclusion: Baseline intestinal activity of CYP3A4 was a key determinant of variability of the inhibitory effect
of clarithromycin among individuals. CYP3A5*1 alleles were associated with greater baseline intestinal
CYP3A activity and, therefore, greater extent of inhibition. The primary in vivo mechanism was not rapidly
reversible competitive or irreversible inhibition but was likely formation of metabolic intermediate com-
plexes. (Clin Pharmacol Ther 2005;77:178-88.)
Amar G. Pinto, MD, Ying-Hong Wang, PhD, Naga Chalasani, MD, Todd Skaar, PhD,
Dhanashri Kolwankar, PhD, J. Christopher Gorski, PhD, Suthat Liangpunsakul, MD,
Mitchell A. Hamman, MS, Million Arefayene, MS, and Stephen D. Hall, PhD Indianapolis, Ind
Clarithromycin is a macrolide antibiotic that has
been shown to exhibit a broad in vitro antibacterial
spectrum that includes staphylococci, streptococci, le-
gionella, Haemophilus inuenzae, Neisseria gonor-
From the Department of Medicine, Indiana University School of Med-
icine.
This study was supported by National Institutes of Health grants
T32GM08425, AG 13718, and MO1-RR00750 and Food and Drug
Administration Cooperative Agreement FDT-001756.
Received for publication June 4, 2004; accepted Oct 5, 2004.
Reprint requests: Stephen D. Hall, PhD, Indiana University School of
Medicine, Department of Medicine, Division of Clinical Pharma-
cology, Wishard Memorial Hospital, Indianapolis, IN 46202.
E-mail: sdhall@iupui.edu
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.002
178
rhoeae, chlamydia, and many anaerobes.
1
It has pri-
marily been used in the treatment of respiratory tract
infections, such as pharyngitis, sinusitis, acute bronchi-
tis, and community-acquired pneumonia. In combina-
tion with amoxicillin (INN, amoxicilline) and a proton
pump inhibitor such as omeprazole, clarithromycin is
indicated for the treatment of Helicobacter pylori
associated duodenal and gastric ulcers.
2
The cyto-
chrome P450 (CYP) 3A enzymes are abundantly ex-
pressed in the liver and small-intestinal epithelia of
adults
3
and are responsible for the metabolism of a
large proportion of the drugs in current clinical use,
including macrolide antibiotics, benzodiazepines, cal-
cium channel blockers, immunosuppressants, and anti-
convulsants.
4
In humans clarithromycin is a potent in-
hibitor of both hepatic and intestinal CYP3A activity.
5
Examples of drugs that are metabolized by human
CYP3A enzymes and whose elimination is inhibited by
clarithromycin include cyclosporine (INN, ciclos-
porin),
6
midazolam, omeprazole,
7
and tacrolimus.
8
The mechanism underlying clinically important in-
hibition of CYP3A catalytic activity in vivo is usually
assumed to be both reversible and competitive. How-
ever, several clinically important CYP3A4 inhibitors,
such as clarithromycin, diltiazem, and verapamil, are
weak competitive inhibitors in vitro (inhibition con-
stant, 10 mol/L) and would not be predicted to
cause signicant drug interactions in vivo. These inhib-
itors have the capability to efciently form metabolic
intermediate (MI) complexes in human liver micro-
somes in vitro, and consequently, it has been suggested
that this phenomenon may be responsible for the inhi-
bition of CYP3A activity in vivo.
5,9,10
MI complex
formation is important because it causes slowly revers-
ible inhibition of CYP3A4, resulting in prolonged re-
duction of activity of this enzyme under physiologic
conditions and may explain the unexpected potency of
many of the CYP3A inhibitors and the delayed onset
and offset of the inhibitory effect.
4,5,9
However, there is
no evidence for MI complexes with human CYP3A in
the liver and intestine in vivo. Therefore we designed
this study to test the hypothesis that clarithromycin
inhibits intestinal wall CYP3A activity by a non
rapidly reversible mechanism. We reasoned that 2
hours after the last dose the concentration of clarithro-
mycin in the intestinal biopsy homogenate (50 dilu-
tion) would be negligible and any observed reduction in
catalytic activity would reect a nonrapidly reversible
mechanism of inhibition. The potential contribution of
decreased intestinal wall CYP3A messenger ribonu-
cleic acid (mRNA) and protein concentration was also
examined to assess whether the mechanism might in-
volve irreversible inhibition. The extent of inhibition of
CYP3A activity in the liver and gut wall is greater in
individuals who have a higher baseline activity.
5
Thus
a secondary goal of our study was to determine whether
the capability of clarithromycin to inhibit intestinal
CYP3A was dependent on baseline CYP3A activity, as
well as to quantify the impact of the genetically poly-
morphic expression of CYP3A5.
10-14
METHODS
Subjects. The study was performed in 10 healthy
subjects (7 men and 3 women). All subjects were aged
18 years or older and received no prescription or over-
the-counter medications for 2 weeks before the study.
Individuals with intolerance to macrolide antibiotics or
benzodiazepines or a signicant medical history, smok-
ing, or alcohol intake were excluded from the study.
One week before and during the study, subjects ab-
stained from consuming grapefruit or juice, apple juice,
citrus products, or vegetables from the mustard green
family. The Clarian and Indiana University Purdue
University Indianapolis (IUPUI) Institutional Review
Board approved this study. All subjects provided writ-
ten informed consent.
Experimental design. After an overnight fast, all
subjects underwent upper intestinal endoscopy and 4
mucosal biopsy specimens were obtained from the sec-
ond portion of the duodenum close to the ampulla
before and after administration of 500 mg clarithromy-
cin for 7 days. Subjects received intravenous midazo-
lam (Roche Pharmaceuticals, Nutley, NJ) to achieve
conscious sedation for esophagogastroduodenoscopy,
and the dose of midazolam used to achieve conscious
sedation varied from subject to subject (range, 2-13
mg). Blood samples were drawn at 1-hour intervals for
3 hours for the determination of serum midazolam and
1-hydroxymidazolam concentrations. Beginning on
the next day, all subjects received 500 mg clarithromy-
cin twice daily by mouth for 7 days. After the last dose
of clarithromycin (day 8), subjects returned to the en-
doscopy suite and were sedated as before and proximal
small bowel biopsy and blood collection were repeated.
All biopsy specimens were immediately snap frozen in
liquid nitrogen and stored at 80C until the time of
analysis. An additional blood sample was obtained just
before esophagogastroduodenoscopy after clarithromy-
cin administration for 7 days to measure serum concen-
trations of clarithromycin and its metabolites. Compli-
ance with clarithromycin dosing and diet, alcohol, and
drug restrictions was assessed by pill count and patient
questioning.
2005;77(3):178-88 Clarithromycin and intestinal CYP3A4/5 activity 179
Duodenal homogenates. Duodenal biopsy speci-
mens (approximately 20 g wet tissue weight) were
homogenized with 20 strokes by use of a handheld
glass homogenizer in an ice jacket with 1 mL ice
cold buffer consisting of 50-mmol/L Tris(hydroxym-
ethyl)aminomethane, 2-mmol/L ethylenediaminetet-
raacetic acid, 1-mmol/L phenylmethylsulfonyl uoride,
1-mmol/L benzamidine, 50 g aprotinin, and 20%
glycerol (pH 7.4). The resultant homogenate was then
snap frozen in liquid nitrogen and stored at 80C until
assayed. Homogenate protein concentrations were de-
termined by the microassay of Lowry et al
15
by use of
bovine serum albumin.
Midazolam hydroxylation activities. An aliquot of
0.5 mL of duodenal homogenate was diluted with 0.5
mL of 100-mmol/L sodium phosphate buffer (pH 7.4)
and incubated at 37C with 25 mol/L midazolam, and
the reaction was started with the addition of 1 mol
NADPH. The reaction was quenched after 10 minutes
with 1 mL of ethyl acetate/hexane (50:50 [vol/vol]) and
then 1 mL of 1N sodium hydroxide/1-mol/L glycine
(pH 11.3) buffer on ice. N-desmethyldiazepam was
then added as an internal standard. The mixture was
then extracted with an additional 3 mL of ethyl acetate/
hexane (50:50 [vol/vol]). The organic phase was dried
under vacuum. N-desmethyldiazepam and 1- and 4-
hydroxymidazolam were analyzed by HPLCmass
spectrometry (Navigator; Finnigan, San Jose, Calif).
The samples were reconstituted in 100 L of mobile
phase (20-mmol/L ammonium acetate [pH 7.4]/aceto-
nitrile/methanol [40:40:20, vol/vol/vol]). The separa-
tion was achieved with an isocratic ow of 1 mL/min to
a Phenomenex Luna C18 column (5 m, 150
4.6mm internal diameter; Phenomenex, Torrance,
Calif). The eluent was analyzed by atmospheric pressure
chemical ionization (APCI) in the positive mode with
cone voltage of 25 V and source and probe temperatures
of 200C and 550C, respectively. N-desmethyldiazepam
and 1- and 4-hydroxymidazolam were detected with
selected ion monitoring at mass-to-charge ratios of 271
and 342, respectively. The limit of quantication of the
hydroxylated metabolite of midazolam was 0.2 ng, and
the precision at 2 ng was 3.2%.
CYP3A4 and CYP3A5 mRNA. Total ribonucleic
acid (RNA) from the samples was prepared by use of
Trizol reagent (Invitrogen Corp, Carlsbad, Calif) ac-
cording to the manufacturers instructions. RNA yield
was determined by spectrophotometry (Beckman DU
640; Beckman Coulter, Inc, Fullerton, Calif), and qual-
ity was assessed by the calculated 260/280-nm ratio.
After RNA isolation, all samples were stored at 80C
until complementary deoxyribonucleic acid (cDNA)
preparation. In a 20-L reaction containing random
hexamers, 2 g of total RNA was reverse-transcribed
into cDNA by use of the Promega Reverse Transcrip-
tion System (Promega Corp, Madison, Wis) according
to the manufacturers instructions.
Real-time polymerase chain reaction analysis. A
real-time polymerase chain reaction (PCR) method was
used to measure the amount of CYP3A4 mRNA in
intestinal biopsy specimens by use of a previously
described method
16
with minor modications. Primers
specic to CYP3A4 transcripts were purchased from
Integrated DNA Technologies (Coralville, Iowa). The
CYP3A4 forward primer sequence was 5-CAT TCC
TCA TCC CAA TTC TTG AAG T-3, and the reverse
primer sequence was 5-CCA CTC GGT GCT TTT
GTG TAT CT-3. As a control, each sample was as-
sayed for the expression of the housekeeping gene
glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
by use of the forward primer 5-GAA GGT GAA GGT
CGG AGT C-3 and the reverse primer 5-GAA GAT
GGT GAT GGG ATT TC-3. The cDNA samples were
assayed for the expression of CYP3A4 and GAPDH by
use of SYBR Green Core Reagents (Applied Biosys-
tems, Foster City, Calif) on a BioRad iCycler Thermal
Cycle (Applied Biosystems). The running conditions
include 3 steps as follows: 95C for 10 minutes (acti-
vation of Taq), 95C for 15 seconds (denaturation), and
58C for 1 minute (annealing step and extension). Steps
2 and 3 are repeated for 40 cycles. Melt-curve analysis
is used to ensure that the expected PCR products are
being generated continuously and reproducibly in each
real-time PCR reaction. The presence of a single in-
ection point on the melt curve, at the previously
established temperature, indicates that a single PCR
product is generated. CYP3A5 mRNA expression was
determined by use of a previously described method.
17
A dilution series (200 femtograms to 0.2 attograms)
containing the cDNA for CYP3A4 and GAPDH was
used as a positive control and to generate a calibrator
series. These controls are required to determine the
PCR efciencies and were included with every assay. A
line was obtained by plotting cycle threshold (C
T
) val-
ues as a function of starting cDNA. The slope of this
curve was used for efciency calculation by use of the
following formula: E 10
(1/Slope)
1. The relative
expression of CYP3A4 and CYP3A5 genes, expressed
as fold variation over control, was calculated after
determination of the difference between C
T
of the
CYP3A4 and CYP3A5 gene and that of the calibrator
gene GAPDH in the intestinal biopsy samples before
and after treatment with clarithromycin.
180 Pinto et al MARCH 2005
Immunoblotting. An aliquot (45 g for CYP3A5
and 20 g for CYP3A4) of small-bowel homogenate
was mixed with Laemmli sample buffer containing
sodium dodecyl sulfate (SDS) and 2-mercaptoethanol
and heated to 95C for 4 minutes. The samples under-
went electrophoresis in 9% polyacrylamide/0.1% SDS
gels until the bromphenol blue dye front had run off the
gel. The samples were transferred onto Hybond-P hy-
drophobic polyvinylidene diuoride membrane (Amer-
sham Pharmacia Biotech, Piscataway, NJ) with a 25-
mmol/L Tris and 192-mmol/L glycine buffer at 100 V
for 20 minutes. The blots were blocked for 1 hour in
phosphate-buffered saline solution containing 0.1%
Tween 20 and 5% nonfat dry milk. The membrane was
then probed for 1 hour with polyclonal antibody to
CYP3A4 (Gentest, Woburn, Mass), anti-CYP3A5
polyclonal antibody (gift from Steve Leeder, Kansas
City, Mo), and a mouse monoclonal antibody to chick
villin that is known to cross-react with human villin
(Chemicon, Temecula, Calif). After extensive washing
with phosphate-buffered saline solution containing
0.1% Tween 20, the membranes were probed with a
horseradish proxidase secondary antibody and devel-
oped with an enhanced chemiuorescence kit (Amer-
sham Pharmacia Biotech). The membrane was exposed
to BioMax lm (Kodak, Rochester, NY), and optical
densities of the bands on the lm were converted to
quantitative numbers by use of Kodak Electrophoresis
Documentation and Analysis System 290 (EDAS 290)
and Kodak 1D image analysis software. The levels of
villin, CYP3A4, and CYP3A5 in the biopsy specimens
were obtained by comparison with villin standard and
serial dilutions of recombinant human CYP3A4 and
CYP3A5, respectively. The standards were run on the
same gel as the biopsy samples. The relative levels of
CYP3A4 and CYP3A5 in the enterocytes were ex-
pressed as a ratio with the villin level of the same
sample.
Correction for interbiopsy variation of enterocyte
content. CYP3A4 and CYP3A5 proteins are expressed
exclusively in mature enterocytes. Enterocytes represent
only a small portion of tissue obtained in the pinch biopsy
of the intestine.
18
It is well known that there is signicant
interbiopsy variation in the content of enterocytes among
biopsy specimens obtained from single individuals, and
differences in percent enterocyte content in intestinal bi-
opsy specimens will alter the amount of immunoreactive
protein observed on a blot. Villin, an enterocyte-specic
protein, is able to control for the variation in biopsy
content of mature enterocytes.
18
Therefore biopsy levels
of CYP3A4 were expressed as a ratio with the villin level
in the same sample. These villin-corrected values provide
a relative measure of enterocyte CYP3A4/CYP3A5
concentration.
Midazolam and clarithromycin assay. Midazolam
and 1-hydroxymidazolam were quantied in human se-
rumby use of desmethyldiazepamas the internal standard,
as previously described in detail.
19
Compounds of interest
were separated by use of a Luna C18 column (5 m, 4.6
150mm internal diameter; Phenomenex) equipped
with a Brownlee RP-18 guard column (PerkinElmer,
Shelton, Conn). The mobile phase (acetonitrile/20-
mmol/L ammonium acetate [pH 7.4]/methanol [40:40:20,
vol/vol/vol]) was delivered isocratically at 1 mL/min.
Desmethyldiazepam, midazolam, and 1-hydroxy-
midazolam were quantied by use of a mass spectrometer
(Navigator; Finnigan) equipped with an APCI interface.
The limit of quantication in serum was 0.25 ng/mL for
midazolam and 1-hydroxymidazolam; the respective cor-
responding coefcient of variation and relative error for
midazolam and 1-hydroxymidazolam at 1.4, 10, and 40
ng/mL were less than 7% and 10%.
Clarithromycin, 14-hydroxyclarithromycin, and
N-desmethylclarithromycin serum concentrations were
determined by use of troleandomycin as the internal
standard, as previously described in detail.
20
In brief,
the compounds of interest were separated by use of a
C8 column (5 m, 100 4.6mm internal diameter;
Brownlee) and a mobile phase of 20-mmol/L ammo-
nium acetate (pH 7)/methanol (20:80 [vol/vol]). The
mobile phase was delivered isocratically at 1 mL/min.
Detection was achieved by use of a mass spectrometer
with an APCI inlet (Navigator; Finnigan). The limit of
quantitation in serum and homogenate was 2.5 ng/mL
for clarithromycin and its metabolites. The respective
corresponding coefcients of variation and relative er-
ror for clarithromycin and its metabolites at 10 ng/mL
were less than 8% and 9%.
CYP3A5 genotype. Genomic deoxyribonucleic acid
was extracted from whole blood by use of the Qiagen
Midi Kit according to the manufacturers instructions
(Qiagen, Valencia, Calif). CYP3A5*3, CYP3A5*6, and
CYP3A5*7 genotyping was carried out by use of a
method described previously
21,22
with PCR conditions
as follows: 5 minutes at 95C; 35 cycles of 0.5 minute
at 94C, 0.5 minute at 55C, and 1 minute at 72C; and,
nally, 10 minutes at 72C. As appropriate, the PCR
product (5 L) was then digested with DdeI restriction
enzyme (Roche Pharmaceuticals) in a 15-L reaction
for 8 hours at 37C and subsequently analyzed by
capillary electrophoresis by use of an Agilent Bioana-
lyzer (Agilent Technologies).
Statistical analysis. Data are reported as mean and
SD. The effect of clarithromycin was determined by
paired Student t test, and signicance was associated
with P .05.
RESULTS
Effect of clarithromycin on intestinal CYP3A cata-
lytic activity. This xed-order study was completed by
10 healthy adult volunteers (7 men and 3 women)
without endoscopy- or sedation-related side effects.
The rate of formation of 1-hydroxymidazolam and
4-hydroxymidazolam in intestinal biopsy homogenate
after incubation with 25 mol/L midazolam for 10
minutes is shown in Fig 1. There was a signicant
reduction (P .001) in the rate of formation of 1-
hydroxymidazolam after 7 days of treatment with clar-
ithromycin (0.35 0.16 pmol min
1
mg
1
) compared
with that at baseline (1.36 0.46 pmol min
1
mg
1
)
(Fig 1). Similarly, there was a signicant reduction (P
.001) in the rate of formation of 4-hydroxymidazolam
after 7 days of treatment with clarithromycin (0.12 0.05
pmol min
1
mg
1
) compared with that at baseline
(0.39 0.12 pmol min
1
mg
1
) (Fig 1). These
differences corresponded to 74% and 62% reductions
in the rate of formation of 1-hydroxymidazolam and
4-hydroxymidazolam, respectively, after treatment
with clarithromycin compared with baseline. There
was a strong correlation (R
2
0.89) between base-
line intestinal CYP3A activity as measured by the
rate of formation of 1-hydroxymidazolam and
change (Baseline Treatment) in CYP3A activity
after treatment with clarithromycin (Fig 2). The
mean concentration of clarithromycin in the intesti-
nal biopsy homogenate was 1.2 0.7 nmol/L (range,
0.42-2.39 nmol/L). One subject had an undetectable
clarithromycin concentration, and another subject had
insufcient biopsy homogenate for analysis. Of 10 sub-
jects, 3 had detectable 14-hydroxyclarithromycin (range,
34.0-70.7 nmol/L) and N-desmethylclarithromycin
(range, 46.4-96.3 nmol/L) concentrations in the intestinal
homogenate.
Effect of clarithromycin on intestinal CYP3A4 ex-
pression. Representative immunoblots of intestinal
CYP3A4 and villin proteins before and after treatment
with clarithromycin for 7 days are shown in Fig 3, A.
The concentrations of CYP3A4 protein in the intestinal
biopsy homogenates expressed as the ratio of integrated
density values of CYP3A4 to villin proteins at baseline
and after treatment are shown in Fig 3, B. There was no
signicant change (P .18) in intestinal CYP3A4
protein concentration ratio at baseline (20.05 14.03)
when compared with the concentration after 7 days of
treatment with clarithromycin (27.4 9.37). There was
no change (P .87) in the ratio of intestinal CYP3A4
to GAPDH mRNA concentration before (0.18 0.20)
and after (0.19 0.18) treatment with clarithromycin
(Fig 4).
Effect of clarithromycin on hepatic CYP3A activ-
ity. A single-point dose-normalized midazolam con-
centration and the ratio of serum 1-hydroxymidazolam
to midazolam have been shown to predict hepatic
CYP3A activity when intravenous midazolam is used
as a probe of hepatic CYP3A in vivo.
23,24
There was a
signicant increase in dose-normalized serum midazo-
lam concentration after 7 days of treatment with clar-
ithromycin at 120 minutes (7.6 2.9 ng/mL, P .001)
and 180 minutes (6.9 2.5 ng/mL, P .001) after
midazolam dosing when compared with corresponding
concentrations at baseline (4.53 1.3 ng/mL and 3.0
0.6 ng/mL, respectively) (Fig 5). There was a signi-
cant decrease (P .001) in the ratio of serum 1-
hydroxymidazolam to midazolam after treatment with
clarithromycin at 120 minutes (0.07 0.05) and 180
minutes (0.05 0.03) compared with corresponding
ratios at baseline (0.21 0.09 and 0.24 0.11, re-
spectively) (Fig 5). The 2-fold increase in dose-
normalized serum midazolam concentration after intra-
venous administration is consistent with results of our
previous study, in which intravenous midazolam area
under the curve was increased by 2.7-fold after admin-
istration of clarithromycin for 7 days.
5
This serves as an
internal control to verify that the systemic effects of
clarithromycin are comparable between this study and
our previous study, which demonstrated increased gut
wall availability of midazolam after the same clarithro-
mycin dosing regimen.
Effect of CYP3A5 genotype on intestinal CYP3A
activity and expression. Of 10 subjects, 3 had at least 1
CYP3A5*1 allele (CYP3A5 expressers) and 6 had 2
CYP3A5*3 alleles (CYP3A5 nonexpressers) (Table I).
One subject was characterized as having CYP3A5*1/*3
and CYP3A5*1/*7 and was presumed to be CYP3A5*3/*7
on the basis of the absence of CYP3A5 mRNA and
protein expression in the intestinal biopsy homogenate,
and consequently, this subject was included in the
CYP3A5 nonexpresser group (Table I). All subjects in the
CYP3A5 expresser group were of African American de-
scent. One of the subjects with the CYP3A5*1/*1 geno-
type possessed a CYP3A5*6 variant allele characterized as
CYP3A5*1/*6. After clarithromycin administration for 7
days, CYP3A5 expressers had a signicantly greater
change (P .01) in the rate of formation of 1-
hydroxymidazolam (1.53 0.25 pmol min
1
mg
1
)
compared with CYP3A5 nonexpressers (0.83 0.4
pmol min
1
mg
1
) (Fig 1). There was no signicant
difference in the observed serum concentrations of clar-
ithromycin (5.9 3.6 mol/L versus 3.0 1.8 mol/L,
P .18), 14-hydroxyclarithromycin (1.1 0.6 mol/L
versus 0.7 0.3 mol/L, P .13), or N-desmethyl-
clarithromycin (0.7 0.4 mol/L versus 0.6 0.3
mol/L, P .44) between the CYP3A5 expresser and
CYP3A5 nonexpresser groups (Table I).
There was no difference in the intestinal CYP3A4
protein concentration ratio between CYP3A5 express-
ers and CYP3A5 nonexpressers at baseline (22.11
17.9 versus 19.16 13.6, P .77) or after treatment
with clarithromycin (22.0 14.1 versus 29.8 6.5, P
.24). One subject with both CYP3A5*3 and
CYP3A5*7 alleles failed to express CYP3A5 protein in
the intestine. One of the subjects with the
CYP3A5*1/*1 genotype also failed to express CYP3A5
protein in the intestine despite expressing wild-type
CYP3A5 mRNA. This discrepancy in the expression of
intestinal CYP3A5 protein in a subject with
CYP3A5*1/*1 genotype may be a result of secondary
degradation of CYP3A5 protein.
There was no signicant difference (P .64) in the
ratio of CYP3A4 to GAPDH mRNA expression at
baseline between CYP3A5 expressers (0.13 0.15)
and CYP3A5 nonexpressers (0.20 0.22). All
CYP3A5 expressers had detectable intestinal CYP3A5
wild-type mRNA expression compared with only 3 of 7
CYP3A5 nonexpressers. There was no signicant dif-
ference (P .94) in the ratio of intestinal CYP3A5 to
GAPDH mRNA before (0.17 0.20) and after (0.16
0.21) treatment with clarithromycin. The ratio of intes-
tinal CYP3A5 to GAPDH mRNA was 6-fold higher in
the CYP3A5 expresser group (0.25 0.24) when com-
pared with the CYP3A5 nonexpresser group (0.04
0.05), but this difference was not statistically different
(P .22).
DISCUSSION
We found a signicant (P .001) reduction in
intestinal CYP3A activity after treatment with 500 mg
clarithromycin for 7 days (Fig 1). There was a signif-
icant (P .001) 74% reduction in the rate of formation
of 1-hydroxymidazolam after treatment with clarithro-
mycin compared with baseline. This reduction in intes-
tinal CYP3A activity represents a reduction in intrinsic
clearance of gut wall CYP3A (CL
int
). On the basis of
our previous study, administration of clarithromycin for
7 days increased gut wall bioavailability of midazolam
from 0.4 (baseline) to 0.8 (after treatment).
5
With the
assumption that the absorption rate constant of the gut
wall (A
G
) remains unchanged, a 75% reduction in in-
testinal CYP3A intrinsic clearance (CL
int,G
) of mida-
zolam accurately predicts doubling of gut wall bioavail-
ability (F
G
) of midazolam by use of the following
equation
10
: F
G
A
G
/(A
G
CL
int,G
). Thus the site of
intestinal mucosal sampling used in this study is rep-
resentative of the overall effect of clarithromycin on
intestinal wall availability of midazolam.
The loss of CYP3A activity in the intestinal wall
after clarithromycin treatment could reect altered ex-
pression of enzymes and inhibition of activity by clar-
ithromycin and its metabolites. We were able to rule
out altered enzyme expression (see later), but the mech-
anism of intestinal CYP3A inhibition by clarithromycin
could involve (1) reversible, competitive inhibition; (2)
covalent modication of CYP apoprotein and irrevers-
ible inhibition; or (3) formation of a slowly reversible
Fig 1. Intestinal CYP3A activity before and after treatment with clarithromycin (500 mg twice daily
for 7 days). CYP3A activity was determined by incubation of intestinal biopsy homogenate with 25
mol/L midazolam (MDZ) for 10 minutes and rate of hydroxymidazolam formation.
Fig 2. Relationship between intestinal CYP3A activity as measured by rate of formation of
1-hydroxymidazolam at baseline and decrease (Baseline Treatment) in CYP3A activity after 7
days of treatment with clarithromycin. Solid squares represent CYP3A5 expressers, and open
squares represent CYP3A5 nonexpressers.
Fig 3. A, Representative immunoblots of intestinal CYP3A4 and villin proteins before (B) and after
(A) 500 mg clarithromycin twice daily for 7 days. Immunoblots used 50 g of homogenate protein
prepared from intestinal biopsy specimens. B, Intestinal CYP3A4 protein concentration before and
after treatment with 500 mg clarithromycin twice daily for 7 days. Concentrations are expressed as
ratio of integrated density values of intestinal CYP3A4 to villin protein.
complex with CYP heme (MI complex).
4
Competitive
inhibition cannot explain our data because the non
enzyme-bound inhibitor is essentially absent from the
intestinal tissue and our catalytic activities were deter-
mined in tissue homogenates that further dilute the
inhibitor by approximately 50-fold. The mean concen-
tration of clarithromycin in the intestinal homogenate
was 1.2 0.7 nmol/L, which is approximately 10,000-
fold lower than the reported inhibition constant (10
mol/L) for reversible competitive inhibition of
CYP3A4 by clarithromycin in vitro.
25
Similarly, the
concentrations of clarithromycin metabolites were in
the low nanomolar range, and this is not consistent with
signicant reversible inhibition. Thus the inhibition of
intestinal wall CYP3A by clarithromycin in vivo must
involve either irreversible modication of the apopro-
tein or MI complex formation, and these 2 possibilities
can be distinguished in vitro. We have previously dem-
onstrated that when human liver microsomes are pre-
incubated with clarithromycin a spectrophotometrically
determined MI complex is formed with CYP3A, result-
ing in a loss of catalytic activity.
9
Indeed, the formation
of the MI complex accounts for all of the loss of
CYP3A activity in vitro, and consequently, this phe-
nomenon is the most likely mechanism underlying the
intestinal CYP3A inhibition in vivo. However, the
small quantity of biopsy tissue available precluded a
direct demonstration of MI complex formation in vivo.
Our conclusion that clarithromycin inhibits intestinal
CYP3A in vivo by MI complex formation is supported
by the ndings that clarithromycin formed MI complex
with hepatic CYP3A when administered alone and to
dexamethasone-pretreated rats.
26,27
In humans a related
macrolide antibiotic, troleandomycin, formed MI com-
plex with hepatic CYP3A in vivo, but the amount of
CYP3A protein in the liver was increased by 5-fold and
erythromycin demethylation activity was reduced com-
pared with that in controls.
28
Similarly, 500 mg eryth-
romycin 3 times a day for 7 days formed MI complex
with hepatic CYP3A in vivo and increased CYP3A
protein by 2-fold in the livers of patients undergoing
elective surgery.
29
Clarithromycin had no effect on intestinal CYP3A pro-
tein expression after administration for 7 days compared
with baseline (Fig 3, B). Several studies have demon-
strated that erythromycin and troleandomycin administra-
tion results in increases in hepatic CYP3A content in both
humans and rodents by protecting the enzyme from deg-
radation (ie, substrate stabilization).
28-31
In rodents clar-
ithromycin administration for 7 days increased hepatic
CYP3A protein content by 3-fold compared with controls.
However, the effect of clarithromycin or any other mac-
rolide on intestinal CYP3A protein expression has not
been studied in humans or animals. Loss of intestinal
CYP3A can result from a diminished protein concentra-
tion in small-intestinal mucosa biopsy specimens, as noted
for grapefruit juice. In one study administration of the
intestinal CYP3A inhibitor grapefruit juice (8 oz 3 times a
day for 6 days) to healthy human subjects reduced
CYP3A4 concentration by 62% in the intestinal homog-
enate, and the authors
32
subsequently demonstrated that
components of grapefruit juice modify CYP3A4 through
suicide inhibition, resulting in accelerated degradation of
CYP3A4 enzyme. However, MI complex formation be-
tween clarithromycin and CYP3A protein does not result
in stabilization or accelerated degradation of CYP3A pro-
tein in the human intestine.
In rodents treatment with troleandomycin for 5 days
increased the expression of hepatic CYP3A mRNA.
30
In contrast to the up-regulation of CYP3A mRNA in
the rat liver after troleandomycin administration, our
study did not demonstrate a signicant change in intes-
tinal CYP3A4 or CYP3A5 mRNA expression after
treatment with clarithromycin compared with baseline
(Fig 4). Thus long-term administration of clarithromy-
cin does not result in a change in the rate of transcrip-
tion of intestinal CYP3A4/CYP3A5 mRNA compared
with baseline. This difference in ndings may reect
differences in the way CYP3A mRNA is regulated in
humans and rodents, as well as differences in the ef-
fects of individual macrolides on CYP3A mRNA ex-
pression within a given species. In addition, there is
evidence to suggest that intestinal CYP3A4 and hepatic
CYP3A4 are regulated differentially by the tissue-
selective expression of nuclear factors.
33
Fig 4. Ratio of intestinal CYP3A4 to glyceraldehyde
3-phosphate dehydrogenase (GAPDH) messenger ribonucleic
acid (mRNA) before and after treatment with 500 mg clar-
ithromycin twice daily for 7 days.
CYP3A activity in the intestine and liver reects the
combined effects of CYP3A5 and CYP3A4, and
CYP3A5 is polymorphically expressed in the human
liver and intestine. There are several known variants in
the human CYP3A5 gene that affect CYP3A5 enzyme
expression and activity in the liver and intestine. The
CYP3A5*3 and CYP3A5*6 variant alleles result in im-
properly spliced CYP3A5 mRNA that fails to produce
a functional CYP3A5 protein.
13
The CYP3A5*7 variant
allele causes single nucleotide insertion, results in
frame-shift mutation, and produces a nonfunctional
CYP3A5 protein.
14
In a study of 500 white subjects
who donated blood, the respective frequencies of
CYP3A5*3, CYP3A5*6, and CYP3A5*7 alleles were
91%, 0.1%, and 0.0%.
22
The respective frequencies of
CYP3A5*3, CYP3A5*6, and CYP3A5*7 alleles among
black subjects are 27%, 15%, and 10%.
13,22,34
The
studies correlating CYP3A5 genotype and pharmaco-
kinetic parameters have yielded mixed results. In a
study limited to CYP3A5*3/*3 and CYP3A5*1/*3 in a
Chinese population, midazolam area under the curve
was not different in CYP3A5*3/*3 and CYP3A5*1/*3
subjects after oral midazolam administration.
35
In con-
trast, CYP3A5*1 genotype positively correlated with
Fig 5. Dose-normalized midazolam (MDZ) concentration and 1-hydroxymidazolam (1-OH
MDZ)/MDZ concentration ratio at 120 minutes and 180 minutes after intravenous midazolam
administration in 10 healthy volunteers before (open bars) and after (solid bars) administration of
500 mg clarithromycin twice daily for 7 days. Asterisk, P .001 compared with baseline.
Table I. Volunteer demographics and serum midazolam, clarithromycin, 14-hydroxyclarithromycin, and
N-desmethylclarithromycin concentrations, with all subjects having received 500 mg clarithromycin twice daily for
7 days
Subject
No. Race
Body
mass
index
(kg/m
2
)
Midazolam intravenous
dose (mg)
Serum
clarithromycin
(mol/L)
Serum
14-hydroxy-
clarithromycin
(mol/L)
Serum
N-desmethyl-
clarithromycin
(mol/L)
CYP3A5
genotype Baseline
After
clarithromycin
1 Black 37.5 13 12 3.1 0.77 0.61 *3/*7
2 Black 35.1 10 12 6.9 1.1 0.92 *1/*3
3 White 26.1 10 12 2.6 0.75 0.57 *3/*3
4 Black 28.0 10 12 1.5 0.75 0.61 *3/*3
5 White 21.5 6 5 5.9 1.0 0.99 *3/*3
6 White 30.0 2 7 4.9 0.71 0.64 *3/*3
7 Black 27.5 10 11 1.1 0.4 0.28 *3/*3
8 Black 31.4 6 7 8.1 1.9 0.98 *1/*1
9 White 25.1 5 4 1.7 0.2 0.15 *3/*3
10 Black 30.0 11 8 1.3 0.44 0.26 *1/*6
Genotyped by use of restriction fragment length polymorphism (Liu et al
21
and van Schaik et al
22
).
Subject was *1/*3 and *1/*7.
hepatic CYP3A activity and tacrolimus dose require-
ment.
36
The inuence of CYP3A5 genotype on the
extent of inhibition by CYP3A inhibitors such as clar-
ithromycin in humans is unknown. In our study
CYP3A5 expressers displayed a signicantly greater
change (1.53 0.24 pmol min
1
mg
1
) in CYP3A
activity after treatment with clarithromycin for 7 days
compared with the CYP3A5 nonexpresser group (0.83
0.4 pmol min
1
mg
1
) (P .01) (Fig 2). The
likely explanation for this phenomenon is that close-to-
complete inhibition by a strong CYP3A inhibitor such
as clarithromycin causes a greater absolute change in
activity in individuals with high baseline CYP3A ac-
tivity and CYP3A expressers fall into the high activity
range. This phenomenon is consistent with ndings in
our previous study, in which a signicant correlation
between initial midazolam intestinal bioavailability and
increase in intestinal bioavailability was observed after
7 days of treatment with clarithromycin.
5
Subjects with
higher baseline CYP3A activity such as CYP3A5 ex-
pressers may be more vulnerable to severe drug inter-
actions compared with CYP3A5 nonexpressers when a
strong CYP3A inhibitor such as clarithromycin is ad-
ministered. However, further studies are needed to fully
characterize the inuence of CYP3A5 genotype on the
degree of intestinal CYP3A activity inhibition by cla-
rithromycin.
In conclusion, long-term administration of clarithro-
mycin causes slowly reversible inhibition of intestinal
CYP3A activity, most likely as a result of formation of
an MI complex. Subjects with a CYP3A5*1 allele tend
to have higher baseline intestinal CYP3A activity and a
signicantly greater inhibition of intestinal CYP3A ac-
tivity after clarithromycin administration.
None of the authors reports any personal or nancial conicts of
interest relevant to this report.
References
1. Sturgill MG, Rapp RP. Clarithromycin: review of a new
macrolide antibiotic with improved microbiologic spec-
trum and favorable pharmacokinetic and adverse effect
proles. Ann Pharmacother 1992;26:1099-108.
2. Markham A, McTavish D. Clarithromycin and omepra-
zole as helicobacter pylori eradication therapy in patients
with H. pylori-associated gastric disorders. Drugs 1996;
51:161-78.
3. Watkins PB. Drug metabolism by cytochromes P450 in
the liver and small bowel. Gastroenterol Clin North Am
1992;21:511-26.
4. Thummel KE, Wilkinson GR. In vitro and in vivo drug
interactions involving human CYP3A. Annu Rev Phar-
macol Toxicol 1998;38:389-430.
5. Gorski JC, Jones DR, Haehner-Daniels BD, Hamman
MA, OMara EM Jr, Hall SD. The contribution of intes-
tinal and hepatic CYP3A to the interaction between mi-
dazolam and clarithromycin. Clin Pharmacol Ther 1998;
64:133-43.
6. Sadaba B, Lopez dO, Azanza JR, Quiroga J, Cienfuegos
JA. Concurrent clarithromycin and cyclosporin A treat-
ment. J Antimicrob Chemother 1998;42:393-5.
7. Gustavson LE, Kaiser JF, Edmonds AL, Locke CS, De-
Bartolo ML, Schneck DW. Effect of omeprazole on
concentrations of clarithromycin in plasma and gastric
tissue at steady state. Antimicrob Agents Chemother
1995;39:2078-83.
8. Wolter K, Wagner K, Philipp T, Fritschka E. Interaction
between FK 506 and clarithromycin in a renal transplant
patient. Eur J Clin Pharmacol 1994;47:207-8.
9. Mayhew BS, Jones DR, Hall SD. An in vitro model for
predicting in vivo inhibition of cytochrome P450 3A4 by
metabolic intermediate complex formation. Drug Metab
Dispos 2000;28:1031-7.
10. Wang YH, Jones DR, Hall SD. Prediction of cytochrome
P450 3A inhibition by verapamil enantiomers and their
metabolites. Drug Metab Dispos 2004;32:259-66.
11. Thummel KE, Shen DD, Podoll TD, Kunze KL, Trager
WF, Bacchi CE, et al. Use of midazolam as a human
cytochrome P450 3A probe. II. Characterization of inter-
and intraindividual hepatic CYP3A variability after liver
transplantation. J Pharmacol Exp Ther 1994;271:557-66.
12. Paine MF, Khalighi M, Fisher JM, Shen DD, Kunze KL,
Marsh CL, et al. Characterization of interintestinal and
intraintestinal variations in human CYP3A-dependent
metabolism. J Pharmacol Exp Ther 1997;283:1552-62.
13. Kuehl P, Zhang J, Lin Y, Lamba J, Assem M, Schuetz J,
et al. Sequence diversity in CYP3A promoters and char-
acterization of the genetic basis of polymorphic CYP3A5
expression. Nat Genet 2001;27:383-91.
14. Hustert E, Haberl M, Burk O, Wolbold R, He YQ, Klein
K, et al. The genetic determinants of the CYP3A5 poly-
morphism. Pharmacogenetics 2001;11:773-9.
15. Lowry OH, Rosenbrough NH, Farr AL, Randall FJ. Pro-
tein measurement with Folin phenol reagent. J Biol Chem
1951;193:265-75.
16. Westlind A, Malmebo S, Johansson I, Otter C, Andersson
TB, Ingelman-Sundberg M, et al. Cloning and tissue
distribution of a novel human cytochrome p450 of the
CYP3A subfamily, CYP3A43. Biochem Biophys Res
Commun 2001;281:1349-55.
17. Lin YS, Dowling AL, Quigley SD, Farin FM, Zhang J,
Lamba J, et al. Co-regulation of CYP3A4 and CYP3A5
and contribution to hepatic and intestinal midazolam
metabolism. Mol Pharmacol 2002;62:162-72.
18. Lown KS, Kolars JC, Thummel KE, Barnett JL, Kunze
KL, Wrighton SA, et al. Interpatient heterogeneity in
expression of CYP3A4 and CYP3A5 in small bowel.
Lack of prediction by the erythromycin breath test. Drug
Metab Dispos 1994;22:947-55.
19. Belle DJ, Callaghan JT, Gorski JC, Maya JF, Mousa O,
Wrighton SA, et al. The effects of an oral contraceptive
containing ethinyloestradiol and norgestrel on CYP3A
activity. Br J Clin Pharmacol 2002;53:67-74.
20. Chu SY, Wilson DS, Guay DR, Craft C. Clarithromycin
pharmacokinetics in healthy young and elderly volun-
teers. J Clin Pharmacol 1992;32:1045-9.
21. Liu TC, Lin SF, Chen TP, Chang JG. Polymorphism
analysis of CYP3A5 in myeloid leukemia. Oncol Rep
2002;9:327-9.
22. van Schaik RH, van der Heiden IP, van den Anker JN,
Lindemans J. CYP3A5 variant allele frequencies in
Dutch Caucasians. Clin Chem 2002;48:1668-71.
23. Thummel KE, Shen DD, Podoll TD, Kunze KL, Trager
WF, Hartwell PS, et al. Use of midazolam as a human
cytochrome P450 3A probe: I. In vitro-in vivo correla-
tions in liver transplant patients. J Pharmacol Exp Ther
1994;271:549-56.
24. Lin YS, Lockwood GF, Graham MA, Brian WR, Loi CM,
Dobrinska MR, et al. In-vivo phenotyping for CYP3A by a
single-point determination of midazolam plasma concentra-
tion. Pharmacogenetics 2001;11:781-91.
25. Jurima-Romet M, Crawford K, Cyr T, Inaba T. Terfena-
dine metabolism in human liver. In vitro inhibition by
macrolide antibiotics and azole antifungals. Drug Metab
Dispos 1994;22:849-57.
26. Tinel M, Descatoire V, Larrey D, Loeper J, Labbe G,
Letteron P, et al. Effects of clarithromycin on cytochrome
P-450. Comparison with other macrolides. J Pharmacol
Exp Ther 1989;250:746-51.
27. Ohmori S, Ishii I, Kuriya S, Taniguchi T, Rikihisa T,
Hirose S, et al. Effects of clarithromycin and its metab-
olites on the mixed function oxidase system in hepatic
microsomes of rats. Drug Metab Dispos 1993;21:358-63.
28. Watkins PB, Wrighton SA, Maurel P, Schuetz EG,
Mendez-Picon G, Parker GA, et al. Identication of an
inducible form of cytochrome P-450 in human liver. Proc
Natl Acad Sci U S A 1985;82:6310-4.
29. Larrey D, Funck-Brentano C, Breil P, Vitaux J, Theodore
C, Babany G, et al. Effects of erythromycin on hepatic
drug-metabolizing enzymes in humans. Biochem Phar-
macol 1983;32:1063-8.
30. Hostetler KA, Wrighton SA, Molowa DT, Thomas PE,
Levin W, Guzelian PS. Coinduction of multiple hepatic
cytochrome P-450 proteins and their mRNAs in rats
treated with imidazole antimycotic agents. Mol Pharma-
col 1989;35:279-85.
31. Wrighton SA, Campanile C, Thomas PE, Maines SL,
Watkins PB, Parker G, et al. Identication of a human
liver cytochrome P-450 homologous to the major
isosafrole-inducible cytochrome P-450 in the rat. Mol
Pharmacol 1986;29:405-10.
32. Lown KS, Bailey DG, Fontana RJ, Janardan SK, Adair
CH, Fortlage LA, et al. Grapefruit juice increases felo-
dipine oral availability in humans by decreasing intestinal
CYP3A protein expression. J Clin Invest 1997;99:2545-
53.
33. von Richter O, Burk O, Fromm MF, Thon KP, Eichel-
baum M, Kivisto KT. Cytochrome P450 3A4 and
P-glycoprotein expression in human small intestinal en-
terocytes and hepatocytes: a comparative analysis in
paired tissue specimens. Clin Pharmacol Ther 2004;75:
172-83.
34. Blanco JG, Edick MJ, Hancock ML, Winick NJ, Der-
vieux T, Amylon MD, et al. Genetic polymorphisms in
CYP3A5, CYP3A4 and NQO1 in children who devel-
oped therapy-related myeloid malignancies. Pharmaco-
genetics 2002;12:605-11.
35. Shih PS, Huang JD. Pharmacokinetics of midazolam and
1-hydroxymidazolam in Chinese with different
CYP3A5 genotypes. Drug Metab Dispos 2002;30:
1491-6.
36. Macphee IA, Fredericks S, Tai T, Syrris P, Carter ND,
Johnston A, et al. Tacrolimus pharmacogenetics: poly-
morphisms associated with expression of cytochrome
p4503A5 and P-glycoprotein correlate with dose require-
ment. Transplantation 2002;74:1486-9.
PHARMACODYNAMICS AND
DRUG ACTION
Association between adherence
measurements of metoprolol and health
care utilization in older patients with heart
failure
Objective: Data from electronic dosing monitors and published pharmacokinetic parameters were used to
derive medication adherence measures for immediate-release metoprolol and examine their association with
health care utilization of outpatients aged 50 years or older with heart failure.
Methods: We used a 1-compartment model and published population pharmacokinetic parameters to estimate
mean plasma metoprolol concentrations for patients treated for 6 to 12 months. In the absence of directly
measured plasma concentrations, we calculated the intended mean plasma concentration (Cp
ave
) under the
assumption of perfect adherence to the prescribed dose and frequency of administration. Projected mean
plasma concentrations (Cp
ave
) were estimated by use of data from recorded dosing times. In addition to
taking adherence (percentage of dose taken) and scheduling adherence (percentage of doses taken on sched-
ule), we calculated the deviation from the intended exposure (Cp
ave
Cp
ave
Cp
ave
) and the proportion
of intended exposure achieved by the patient (Cp
ave
/Cp
ave
). We assessed the association between the
adherence measures and the numbers of emergency department visits and hospital admissions experienced by
the patients.
Results: Patients (N 80) were aged 62 8 years. Mean Cp
ave
and Cp
ave
/Cp
ave
were 7.9 ng/mL (SD,
10.7) and 0.6 (SD, 0.3), respectively. Log-linear models adjusted for patient functional status indicated that
greater deviation from the intended metoprolol exposure (Cp
ave
) was associated with increased numbers of
emergency department visits (P < .0001) and hospital admissions (P < .0001). A higher proportion of
intended exposure (Cp
ave
/Cp
ave
) corresponded to a reduced number of emergency department visits (P
.0204) and hospital admissions (P .0093). Taking adherence was univariately associated with both
emergency department visits and hospital visits (P < .0001 and P .0010, respectively). Scheduling
continued on next page
Wanzhu Tu, PhD, Andrew B. Morris, PharmD, Jingjin Li, PhD, Jingwei Wu, MS,
James Young, PharmD, D. Craig Brater, MD, and Michael D. Murray, PharmD, MPH
Indianapolis, Ind, and Chapel Hill, NC
From the Department of Medicine, Indiana University School of
Medicine, Indianapolis; Indiana University Center for Aging Re-
search, Indianapolis; Regenstrief Institute, Inc, Indianapolis;
Wishard Health Services, Indianapolis; Department of Pharmacy
Practice, Purdue University School of Pharmacy, Indianapolis; and
Pharmaceutical Policy and Evaluative Sciences Division, Univer-
sity of North Carolina School of Pharmacy, Chapel Hill.
This research was supported by grants AG19105, AG07631,
HL69399, and HD042404 from the National Institutes of Health.
Received for publication April 9, 2004; accepted Oct 14, 2004.
Reprint requests: Michael D. Murray, PharmD, MPH, Mescal S.
Ferguson Distinguished Professor, Pharmaceutical Policy and
Evaluative Sciences, University of North Carolina at Chapel Hill,
CB 7360 Beard Hall, Chapel Hill, NC 27599-7360.
E-mail: mick@unc.edu
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.004
189
adherence was associated with the number of emergency department visits (P .0181) but not with the
number of hospital admissions (P .1602). Model selection procedures consistently chose the proposed
measures over taking adherence and scheduling adherence.
Conclusion: Deviation from the intended exposure and proportion of intended exposure achieved by the
patient are valid adherence measures for immediate-release metoprolol and are associated with health care
utilization. The potential utility of these measures for other -adrenergic antagonists and perhaps other
cardiovascular drugs should be investigated. (Clin Pharmacol Ther 2005;77:189-201.)
Patients with chronic diseases must take medications
regularly to prevent deteriorating illness requiring
costly use of health care services such as emergency
department (ED) visits and hospitalization. However,
approximately half of all patients who are prescribed
medications for chronic diseases do not adhere to their
prescriptions.
1-3
Medication adherence is particularly
important to patients with chronic heart failure (CHF)
because of the propensity for such patients to require
expensive health care services. Adherence to the pre-
scribed regimen by patients with heart failure varies
from 10% to 85%.
4,5
Indiscretions in medication ad-
herence or diet may result in decompensation in 44% of
patients with CHF and other chronic cardiovascular
diseases.
6
Adherence measurement has been greatly facilitated
by electronic dosing monitors that record the temporal
pattern of medication container openings.
7
Assuming
that an appropriate number of tablets are taken each
time a lid is opened, aggregate adherence measures
may be computed to summarize medication event data
generated by electronic monitors.
8-10
Choo et al
11
ex-
plored the interrelationships among these measure-
ments and elucidated their mathematic foundation. Be-
cause of the relationship between dosing and plasma
concentration,
12,13
we believed that it would be appro-
priate to explore this relationship and assess its effect
on relevant patient outcomes. As such, we derived 2
adherence measures that couple electronic dosing
records with the pharmacokinetic properties of the
monitored drug. In doing so, the resulting adherence
measures consider the frequency, dose administration
time, dose strength, and population pharmacokinetic
parameters of the medication. We hypothesized that the
proposed measures would be associated with disease
exacerbation. We tested this hypothesis in 80 older
patients with CHF treated with immediate-release
metoprolol and then assessed the associations between
the new measures and the numbers of ED visits and
hospital admissions experienced by these patients.
In the past decade, several investigators have ex-
plored the use of dosing history data to improve the
estimation of drug concentration curves by measuring
plasma drug concentrations.
12,14,15
Their work has
highlighted the importance of dosing history data in
pharmacokinetic and pharmacodynamic studies. We
took a different approach. Instead of relying on directly
measured drug concentrations, we estimated plasma
concentrations by using an individual patients dosing
records and published population pharmacokinetic pa-
rameters. One advantage of using published population
parameters is that the resulting measures could be ap-
plicable to nonpharmacokinetic clinical trials in which
drug concentrations were unavailable.
METHODS
Setting and participants. Data were obtained from a
health services research, randomized, controlled trial
being conducted in Indianapolis, Ind. The primary ob-
jective of the trial was to test the effectiveness of a
pharmacy-based intervention program aimed at im-
proving medication adherence in older adults with
CHF.
16
The study protocol was approved by the Indi-
ana UniversityPurdue University Institutional Review
Board (Indianapolis, Ind).
Participants were recruited from Wishard Health Ser-
vices, which serves the inner-city residents of Indianap-
olis, Ind. The study population may be generally char-
acterized as having limited resources and being
ethnically diverse. Eligible participants were English-
speaking, were aged 50 years or older, had a diagnosis
of CHF, and were currently taking at least 1 cardiovas-
cular medication, including angiotensin-converting en-
zyme (ACE) inhibitors, angiotensin II receptor antag-
onists, -adrenergic receptor antagonists, digoxin, an
aldosterone antagonist, and loop or nonloop diuretics.
On enrollment, participants were provided with elec-
tronic monitor lids (see below) on their prescription
containers for each of the aforementioned cardiovascu-
lar medications. Other prescription medication contain-
ers did not have electronic monitors afxed. Medica-
tion use was monitored for 6 to 12 months.
To test the operating characteristics of the proposed
adherence measures, we used data from the trials con-
trol arm for 80 subjects who had been prescribed
immediate-release metoprolol, a selective
1
-
190 Tu et al MARCH 2005
adrenergic antagonist with well-established pharmaco-
kinetics. The rationale for using only control arm pa-
tients was to avoid any effects of the intervention,
which was aimed at improving medication adherence
among participants assigned to the intervention arm.
16
Electronic dosing records. We used MEMS Track
V prescription container lids as electronic monitoring
devices (AARDEX Ltd, Zug, Switzerland).
17
Data re-
corded by the MEMS monitors were extracted into a
desktop computer that in turn displayed the event pat-
terns graphically and calculated medication taking and
scheduling adherence.
7
MEMS adherence logs were
validated by pharmacists and data analysts via compar-
ison of metoprolol MEMS dosing patterns with patient
self-reported adherence and with prescription rell ad-
herence.
18
We also veried the phase start and end
dates to ensure that the MEMS lids were correctly set.
Veried MEMS data were then exported from MEMS
software to SAS (SAS Institute, Cary, NC) for further
analyses.
Pharmacokinetic adherence measures based on es-
timated plasma concentrations. To accommodate the
pharmacokinetic properties of metoprolol, we con-
structed 2 adherence measures. The rst measure esti-
mated the magnitude of the patients deviation from the
intended metoprolol exposure, whereas the second
measure estimated adherence as the proportion of in-
tended metoprolol exposure achieved by the patients
dosing pattern. Both measures were calculated by use
of published metoprolol pharmacokinetic parameters
and dosing history data recorded by MEMS. Intended
drug exposure was calculated as the average plasma
concentration under perfect adherence to the prescribed
regimen.
Plasma concentrations (Cp) of metoprolol and most
other cardiovascular medications were not measured.
We, therefore, estimated the Cp levels by using estab-
lished pharmacokinetic parameters of metoprolol. Con-
sidering the standard 1-compartment model for the Cp
level in the ith subject, we estimated the Cp levels at
preselected time points t
j
, where j 0, 1, 2, . . . J
i
.
Herein, we chose the time points at hourly intervals and
at all of the recorded metoprolol dosing times. For the
ith patient, we used the rst recorded dosing time as the
starting point (t
0
0) and the last time point in the
observation period as t
Ji
, where t
Ji
was allowed to vary
from subject to subject. We then estimated the plasma
concentration of metoprolol in the ith subject at time t
j
,
Cp[i,t], for all of the preselected time points t
j
, where j
0, 1, 2, . . . J
i
. Using the multidose 1-compartment
model,
19
we considered the iterative algorithm below.
An illustration for the use of the algorithm is provided
in the Appendix.
Algorithm. Let d[i,j] be the dose amount of the med-
ication that the ith subject took at time t
j
. If the patient
did not take his or her medication at the time, d[i,j]
0. At time t
0
0, the cumulative dose D[i,0] is the rst
dose taken (d[i,0]), and the plasma concentration is as
follows:
Cp[i,0]
F d[i,0] k
a
V[i](k
a
k
e
)
[exp(k
e
t
0
) exp(k
a
t
0
)] 0
(1)
where F is the population-based bioavailability of the
medication, V[i] is the volume of distribution adjusted
for the ith patients weight, and k
a
and k
e
are the
rst-order absorption and elimination rate constants,
respectively.
At all later time points t
j
, where j 1, 2, . . . J
i
, we
rst calculated the cumulative dose as follows:
D[i,j] D[i,j 1]exp(k
a
[t
j
t
j1
]) di,j (2)
Then we wrote the plasma concentration at time t
j
as
follows:
Cp[i,j] Cp[i,j 1]exp(k
e
[t
j
t
j1
])
F D[i,j 1] k
a
V[i](k
a
k
e
)
(exp[k
e
(t
j
t
j1
)]
exp[k
a
(t
j
t
j1
)]) (3)
We used the above-described iterative algorithm to
estimate the hourly plasma concentration levels based
on the dosing events recorded by the MEMS records.
Specically, we calculated the values of Cp[i,j] in
hourly increments after the rst dosing event for the
entire duration of the study period covered by the
MEMS logs. We then averaged the estimated hourly
Cp[i,j] in each study subject for the mean hourly
plasma concentration for the study period. We denoted
this average as Cp
ave
[i] for the ith subject as follows:
Cp
ave
J
i j1
J
i
Cp
i,j
, where J
i
was the num-
ber of hours contained in the subjects MEMS log.
Therefore Cp
ave
[i] can be viewed as an approximation
of the mean plasma concentration of metoprolol during
the study period.
In a similar fashion, we calculated the intended
plasma concentration level under the assumption of
perfect adherence to the prescribed dose and frequency
of administration. We denoted the average of hourly Cp
values under the assumption of perfect adherence as
Cp
ave
[i]. It should be noted that Cp
ave
[i] can also be
obtained by use of a steady-state equation.
19
The dif-
ference between Cp
ave
[i] and Cp
ave
[i], therefore, re-
2005;77(3):189-201 Adherence measures for electronic dosing data 191
ected the deviation of the projected Cp from the
intended (or prescribed) level. The greater this devia-
tion, the poorer the patients adherence to the pre-
scribed metoprolol regimen. Hence we propose to use
Cp
ave
[i] Cp
ave
[i] Cp
ave
[i] to measure the med-
ication adherence of the ith patient. Unlike most of the
existing adherence measures, Cp
ave
[i] also accommo-
dates the pharmacokinetic properties of the drug. For a
drug with a longer half-life, the plasma concentration
has a slower decrement and Cp
ave
[i] is less sensitive
to a delayed administration. Fig 1 depicts the estimated
Cp curves of metoprolol for an abbreviated time period
in 4 representative participants.
An alternative adherence measure was the ratio
Cp
ave
[i]/Cp
ave
[i], which quantied the proportion of
intended plasma concentration achieved by the ith pa-
tient. A larger value for this ratio indicated better pa-
tient adherence to the prescribed regimen.
Metoprolol. At therapeutic doses, metoprolol acts by
selectively antagonizing the
1
receptor in the myocar-
dium, producing negative chronotropism and inotro-
pism. Long-term use of metoprolol in heart failure
patients has been shown to be benecial in reducing
hospitalization and mortality rates.
20
Abrupt discontin-
uation and erratic administration of the drug, on the
other hand, are known risk factors of rebound tachy-
cardia, hypertension, and subsequent cardiac isch-
emia.
21
Patients in this study took immediate-release
metoprolol at doses ranging from 12.5 mg to 100 mg at
a frequency of 12 hours.
We used a 1-compartment model with published
pharmacokinetic parameters to estimate plasma con-
Fig 1. Estimated and intended metoprolol plasma concentration curves for 4 study subjects during
abbreviated time interval for exemplication. The dashed curve depicts the time course of the
estimated plasma concentration under perfect adherence and thus represents the intended dosing
pattern. The continuous curve depicts the MEMS-recorded dosing pattern. The upper horizontal line
represents the overall average value of intended metoprolol exposure (Cp
ave
[i]), and the lower
horizontal line represents the hourly average metoprolol concentration (Cp
ave
[i]) estimated from the
MEMS-recorded dosing pattern.
centrations of metoprolol, although it more appropri-
ately follows a 2-compartment kinetic model. Two-
compartment models could be used when kinetic data
were available to t those models. Clearance based on
differences in metabolism varies between 50 L/h and
409 L/h. In the absence of individual data, we used the
population-based value of the elimination rate constant
(k
e
) and the 1-compartment model to estimate the
plasma concentration.
The pharmacokinetic parameters of metoprolol are
well established. The drug has an elimination half-life
of approximately 3.5 hours, bioavailability (F) of 0.5,
k
e
equal to 0.198/h, and time to peak Cp (t
max
) approx-
imately equal to 1.5.
22,23
On the basis of the following
equation, we solved for an estimate of the absorption
parameter (k
a
) equal to 1.585/h:
t
max
1
(k
a
k
e
)
ln
k
a
k
e
By use of these parameters, we computed the values

of Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i] for each of the study
patients based on the dosing events recorded by
MEMS.
Statistical analysis. To assess the performance of
the proposed pharmacokinetic-based adherence mea-
sures, we rst examined their correlations with 2 com-
monly used MEMS adherence measures, namely, tak-
ing adherence and scheduling adherence. Taking
adherence was dened as the percentage of dose taken
in the observation period, whereas scheduling adher-
ence was dened as the percentage of dose taken on
schedule (ie, 2 hours of the desired dosing time). We
then evaluated their associations with 2 important clin-
ical outcomesnumbers of hospital admissions and
numbers of ED visits, which are regarded as surrogate
markers of disease exacerbation. Specically, by use of
an electronic medical record system,
24,25
we deter-
mined the numbers of hospital admissions and ED
visits experienced by each study participant during the
time period of observation. We then tted log-linear
models by using the numbers of hospital admissions
and ED visits as outcome variables. Important explan-
atory variables considered for the model included de-
mographic characteristics (age, race, and gender), New
York Heart Association (NYHA) classication, num-
bers and indicators of cardiovascular drugs, and taking
and scheduling adherence and the proposed adherence
measure of metoprolol. Race, gender, and NYHA clas-
sication were treated as categoric variables. Each level
of a categoric variable was compared with a reference
level.
The effects of variables were evaluated by use of the
Wald test.
26
Because lengths of observation differed
among subjects, we introduced the length of follow-up
period into the model as an offset parameter.
27
The 4
metoprolol adherence measures (taking and scheduling
adherence and the 2 new pharmacokinetic-based mea-
sures) were used as independent variables for the ex-
amination of their associations with the numbers of ED
visits and hospital admissions. Our general model se-
lection strategy was to rst screen all of the indepen-
dent variables for their univariate associations with
outcome variables. Then, for each outcome, we used
both forward and stepwise selection procedures to
choose the nal models. Because 2 new adherence
measures were proposed in this research, to assess the
performance of each one independently, we conducted
2 parallel selection processes for each outcome, one for
Cp
ave
[i] and all other covariates (excluding Cp
ave
[i]/
Cp
ave
[i]) and the other for Cp
ave
[i]/Cp
ave
[i] and all of
the covariates (excluding Cp
ave
[i]). This resulted in
the 4 nal models.
Although the primary objective of this research was
to develop new pharmacokinetic-based adherence mea-
sures and use them to assess the effect of metoprolol
adherence on health care utilization, it was important to
present the results within the context of other concom-
itantly used cardiovascular drugs. In our study sample,
87% and 83% of the patients, respectively, took diuret-
ics and ACE inhibitors concurrently with metoprolol.
Fewer patients were taking angiotensin II receptor an-
tagonists, digoxin, and spironolactone. To assess the
adherence effects of concomitant drugs, we rst eval-
uated the taking and scheduling effects for each of these
drugs univariately via log-linear models. Taking adher-
ence and scheduling adherence of these other cardio-
vascular drugs were used in the model selection pro-
cess, but none of them were selected into the nal
models. We re-examined the adherence effects of di-
uretics and ACE inhibitors by including them in the
nal models. We did not enter diuretics and ACE
inhibitors into the nal models simultaneously because
the stability of the models was undermined by strong
multicollinearity from multiple factors and the much
reduced sample size (only 53 patients were prescribed
metoprolol, diuretics, and ACE inhibitors concomi-
tantly). We do not present data on angiotensin II recep-
tor antagonists, digoxin, and spironolactone in multi-
variate models because of Journal space limitation.
RESULTS
Demographic and clinical characteristics of the study
patients are listed in Table I. The mean age (SD) of
Table I. Sample statistics and univariate associations between patient characteristics and outcomes
Variable
Sample statistics
(N 80)
Univariate model for
No. of hospitalizations
Univariate model for
No. of ED visits
Regression
coefcient
and SE P value
Regression
coefcient
and SE P value
Age (y) 62 (8) 0.0131 (0.023) .5704 0.0095 (0.013) .4791
Gender .3482 .5890
Female 55 (69%) 0.4320 (0.461) 0.1291 (0.239)
Male 23 (31%) 0.0 0.0
Race .1645 .0093
Nonwhite 41 (51%) 0.5625 (0.405) 0.5940 (0.228)
White 39 (47%) 0.0 0.0
NYHA classication* .0001 .0001
Class I 13 (16%) 1.9974 (0.570) 1.4786 (0.312)
Class II 41 (51%) 2.2215 (0.465) 1.9248 (0.278)
Class III 23 (29%) 2.7832 (0.612) 2.6879 (0.375)
Class IV 3 (4%) 0.0 0.0
No. of cardiovascular drugs 4.3 (SD, 1.76) 0.1236 (0.099) .2108 0.0787 (0.057) .1662
Metoprolol
Dose at baseline (mg) 48.15 (SD, 22.8) 0.0131 (0.009) .1522 0.0024 (0.005) .6078
Intended concentration (Cp
ave
) (ng/mL) 20.76 (SD, 13.7) 0.0270 (0.020) .1792 0.0164 (0.014) .2313
Measure of Cp
ave
(ng/mL) 7.9 (SD, 11.0) 0.0412 (0.009) .0001 0.0270 (0.006) .0001
Measure of Cp
ave
/Cp
ave
(%) 0.6 (SD, 0.3) 2.7598 (0.647) .0001 1.7548 (0.334) .0001
Taking adherence (%) 63 (SD, 34.1) 1.9199 (0.583) .0010 1.3313 (0.320) .0001
Scheduling adherence (%) 32.7 (SD, 31.0) 0.9487 (0.676) .1602 0.8877 (0.376) .0181
Diuretics
Diuretic use (yes 1, no 0) 69 (87%) 0.8182 (0.418) .0505 0.1385 (0.283) .6248
Taking adherence (%) 56.4 (SD, 32.3) 2.2319 (0.756) .0031 1.1252 (0.365) .0020
ACE inhibitors
ACE inhibitor use (yes 1, no 0) 66 (83%) 0.5218 (0.611) .3931 0.5863 (0.352) .0954
ARBs
ARB use (yes 1, no 0) 12 (15%) 0.2370 (0.461) .6068 0.3226 (0.310) .2986
Digoxin
Digoxin use (yes 1, no 0) 21 (26%) 0.1287 (0.405) .7504 0.5356 (0.215) .0128
Spironolactone
Spironolactone use (yes 1, no 0) 12 (15%) 0.6160 (0.436) .1581 1.1373 (0.221) .0001
No. of hospitalizations 0.4 (SD, 1.1)
No. of ED visits 1.3 (SD, 2.5)
ED, Emergency department; Cpc
ave
, intended level of hourly average metoprolol concentration in plasma; Cp
ave
, deviation from intended level of hourly average
metoprolol concentration in plasma; Cp
ave
/Cp
ave
, ratio of estimated and intended levels of hourly average metoprolol concentration in plasma; NYHA, New York Heart
Association; ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker.
*Functional performance is best for class I and worst for class IV.
the study participants was 62 8 years, 69% were
women, and 51% were nonwhite. Approximately half
(51%) of the participants had an NYHA functional
class of II. Participants were prescribed a mean of 4.3
2 cardiovascular drugs, including diuretics (87%),
ACE inhibitors (83%), angiotensin II receptor antago-
nists (15%), digoxin (26%), and spironolactone (15%).
Mean taking adherence of metoprolol was 63%, mean
scheduling adherence was 33%, average Cp
ave
was
7.87 ng/mL, and average Cp
ave
[i]/Cp
ave
[i] was 60%.
For metoprolol, correlation between Cp
ave
[i] and tak-
ing adherence was 0.4718 (P .0001) and correla-
tion between Cp
ave
[i] and scheduling adherence was
0.3746 (P .0005). Cp
ave
[i]/Cp
ave
[i] correlated
strongly with both taking adherence ( 0.8109, P
.0001) and scheduling adherence ( 0.5918, P
.0001). The signicant correlations among these 4 mea-
sures were not unexpected, because the information on
the drug consumption used to calculate taking and
scheduling adherence was incorporated in the compu-
tation of Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i]. Graphic pre-
sentations of these relationships are shown in Fig 2 with
histograms (Fig 2, f and i) depicting the distributions of
Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i]. We noted that
Fig 2. Scatter plot of Cp
ave
(estimated level of hourly average metoprolol concentration in plasma)
versus Cp
ave
(intended level of hourly average metoprolol concentration in plasma) (a), Cp
ave
[i]
(patients deviation from intended level of hourly average metoprolol concentration in plasma)
versus prescribed dose (b), scatter plot of Cp
ave
/Cp
ave
(ratio of estimated and intended levels of
hourly average metoprolol concentration in plasma) versus prescribed dose (c), scatter plot of
Cp
ave
[i] versus taking adherence (d), scatter plot of Cp
ave
[i] versus scheduling adherence (e),
frequency distribution of Cp
ave
[i] (f), scatter plot of Cp
ave
/Cp
ave
versus taking adherence (g),
scatter plot of Cp
ave
/Cp
ave
versus scheduling adherence (h), and frequency distribution of Cp
ave
/
Cp
ave
(i). r, Correlation coefcient.
Cp
ave
[i] was skewed (skewness 2.5) probably be-
cause of 2 large values in the right tail. The rest of the
distribution was relatively symmetric (skewness
0.58). Fig 2, b and c, showed that Cp
ave
[i] was more
sensitive to the dose input than was Cp
ave
[i]/Cp
ave
[i].
Univariate analysis suggested that all 4 adherence
measures for metoprolol were independently associated
with hospitalizations and ED visits (better adherence
corresponded to decreased utilization), with the excep-
tion of the nonsignicant association between schedul-
ing adherence and hospitalization (P .1602). Wald
statistics (Coefcient/SE) corresponding to each of the
4 adherence measures showed that the values of Wald
statistics for Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i] were
larger than those for taking adherence and scheduling
adherence, suggesting that stronger associations existed
between the 2 pharmacokinetic-based adherence mea-
sures and the outcomes. Better adherence to diuretics,
ACE inhibitors, digoxin, and spironolactone was also
independently associated with decreased utilization
(Table I).
Factors signicantly associated with hospital admis-
sions and ED visits, as identied by the nal models,
were tabulated in part 1 of Tables II and III, respec-
tively. For each of the outcome variables, we reported
2 nal models, as follows: (1) a model selected from a
pool of candidate variables including Cp
ave
[i] and (2)
a model selected from a pool of candidate variables
including Cp
ave
[i]/Cp
ave
[i]. The reported models were
chosen through a formal stepwise model selection pro-
cedure, and the signicant variables meeting the selec-
tion criteria (P .05) were included in the nal mod-
els. The forward selection process yielded the same
models.
Without exception, a patients functional status and
pharmacokinetic-based measures of metoprolol adher-
ence were selected into the nal models, suggesting
that both were strong correlates of health care utiliza-
tion in these heart failure patients. Specically, Cp
ave
was signicantly associated with the number of hospital
admissions after we controlled for the effect of NYHA
classication. Exponentiating the coefcient estimate
of Cp
ave
, we found the relative risk of hospitalization
associated with a 1-unit increase in Cp
ave
to be 1.05.
In other words, for each 1-ng/mL deviation from the
intended exposure, there was a 5.0% increase in the risk
of hospital admission (P .0001). Similarly, increased
Cp
ave
[i]/Cp
ave
[i] (a greater proportional metoprolol ex-
posure) was associated with a decreased number of
hospital admissions (P .0093). Not surprisingly,
NYHA functional classication was also found to be
signicantly associated with the number of hospital
admissions; compared with patients with NYHA class
IV, patients with all other classes had fewer numbers of
hospital admissions.
For ED visits, Cp
ave
and Cp
ave
[i]/Cp
ave
[i] were
again identied by model selection procedures as sig-
nicant correlates of the number of ED visits, along
with NYHA classications. The nal models show that
greater metoprolol adherence (measured by Cp
ave
and
Cp
ave
[i]/Cp
ave
[i]) was associated with a reduction in
the number of ED visits. Again, taking and scheduling
adherence measures did not reach the level of signi-
cance to be selected into the nal models.
In addition to the assessment of the effect of meto-
prolol adherence, we also examined the adherence ef-
fects of other cardiovascular drugs. Whereas adherence
effects of diuretics, ACE inhibitors, digoxin, and spi-
ronolactone were signicant in univariate analyses (Ta-
ble I), none were strong enough to be selected into nal
models. Furthermore, adherence to these other drugs
was not signicant when we forced them into the nal
models (parts 2 and 3 in Tables II and III).
DISCUSSION
The pharmacokinetic-based adherence measures
Cp
ave
and Cp
ave
[i]/Cp
ave
[i] account for adherence to
the regimen and pharmacokinetic properties of meto-
prolol. From the estimated plasma concentration
curves, it appeared that the proposed measures were
sensitive to missing doses, as well as erratic drug ad-
ministration. Signicant associations between the pro-
posed measure and ED and hospital admissions indicate
their practical utility and validity.
An important feature of the proposed measures is
that they can be used in nonpharmacokinetic studies
where drug levels are not measured. They also offer the
exibility to incorporate dose changes because equa-
tions 2 and 3 allow the dose d[i,j] to vary across the
time points. This is a clinically important feature be-
cause -blockers such as metoprolol are usually titrated
to patient response.
Unlike the taking and scheduling adherence com-
puted by MEMS software, the pharmacokinetic mea-
sures focus on drug exposure as opposed to the medi-
cation use behaviors only. In doing so, they account for
metoprolols kinetic characteristics, as well as the pa-
tients adherence behaviors. Furthermore, certain med-
ications are less forgiving than others in terms of clin-
ical outcomes from failure to take them.
13
For this
reason, we contend that the average plasma concentra-
tion of some cardiovascular medications may prove to
be a more sensitive and objective assessment of true
therapeutic exposure received by patients.
Table II. Factors associated with number of hospitalization admissions*
Variable Coefcient and SE P value
Part 1: Patients prescribed metoprolol
Model 1: Selected model containing Cp
ave
NYHA classication .0001
Class I 2.2313 (0.574)
Class II 2.7142 (0.518)
Class III 2.7312 (0.613)
Class IV 0.0
Cp
ave
for metoprolol 0.0485 (0.010) .0001
ave
/Cp
ave
Class I 1.1650 (0.639)
Class II 1.5052 (0.521)
Class III 1.6770 (0.717)
Class IV 0.0
Cp
ave
/Cp
ave
for metoprolol 1.9032 (0.732) .0093
Part 2: Patients prescribed metoprolol and a diuretic
ave
Class I 2.2257 (0.926)
Class II 3.3555 (0.846)
Class III 2.5646 (0.884)
Class IV 0.0
Cp
ave
for metoprolol 0.0465 (0.026) .0762
Diuretic taking adherence 0.0386 (1.084) .9716
ave
/Cp
ave
Class I 1.3101 (0.766)
Class II 2.4764 (0.743)
Class III 1.5748 (0.801)
Class IV 0.0
Cp
ave
/Cp
ave
for metoprolol 2.9271 (1.193) .0141
Part 3: Patients prescribed metoprolol and an ACE inhibitor
ave
Class I 2.4980 (0.635)
Class II 3.2846 (0.690)
Class III 2.7990 (0.678)
Class IV 0.0
Cp
ave
for metoprolol 0.0550 (0.012) .0001
ACE inhibitor taking adherence 0.6179 (0.637) .3319
ave
/Cp
ave
Class I 1.5535 (0.646)
Class II 1.8602 (0.588)
Class III 1.9969 (0.718)
Class IV 0.0
Cp
ave
/Cp
ave
for metoprolol 2.3532 (1.059) .0262
*Because of missing data, 75 subjects were used for the model selection process and the tting of the nal models.
Reference category.
Table III. Factors associated with number of ED admissions*
Variable Coefcient and SE P value
Part 1: Patients prescribed metoprolol (N 80)
ave
Class I 1.6112 (0.315)
Class II 2.2153 (0.297)
Class III 2.6227 (0.376)
Class IV 0.0
Cp
ave
for metoprolol 0.0315 (0.007) .0001
ave
/Cp
ave
Class I 1.0710 (0.356)
Class II 1.5661 (0.314)
Class III 2.1430 (0.439)
Class IV 0.0
Cp
ave
/Cp
ave
for metoprolol 0.8887 (0.383) .0204
Part 2: Patients prescribed metoprolol and a diuretic (n 69)
ave
Class I 1.9023 (0.466)
Class II 2.5968 (0.413)
Class III 3.1401 (0.506)
Class IV 0.0
Cp
ave
for metoprolol 0.0332 (0.013) .0126
ave
/Cp
ave
Class I 1.2457 (0.409)
Class II 2.0234 (0.383)
Class III 2.5065 (0.509)
Class IV 0.0
Cp
ave
/Cp
ave
for metoprolol 1.2287 (0.502) .0143
Part 3: Patients prescribed metoprolol and an ACE inhibitor (n 66)
ave
Class I 1.9710 (0.348)
Class II 2.6699(0.361)
Class III 3.1329 (0.436)
Class IV 0.0
Cp
ave
for metoprolol 0.0314 (0.007) .0001
ave
/Cp
ave
Class I 1.4599 (0.352)
Class II 2.0296 (0.343)
Class III 2.6683 (0.458)
Class IV 0.0
Cp
ave
/Cp
ave
for metoprolol 1.2461 (0.542) .0214
*Because of missing data, 75 subjects were used for the model selection process and the tting of the nal models.
Reference category.
Both Cp
ave
and Cp
ave
[i]/Cp
ave
[i] for metoprolol in
patients with CHF were associated with hospital admis-
sions and ED visits. Empiric evidence suggests that
associations with utilization were stronger than those
for taking or scheduling adherence. In 2 ways,
Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i] measure the estimated
drug exposure by comparing it with the intended expo-
sure. When there is a large deviation from the intended
level (ie, larger Cp
ave
[i] or smaller Cp
ave
[i]/Cp
ave
[i]),
disease exacerbation occurs and may result in an ED
visit or hospital admission. From a clinical perspective,
poor adherence in the long run and occasional drug
holidays are harmful behaviors.
20,28,29
Because both
Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i] are calculated for an
extended period of time, they are likely to be more
sensitive to long-term poor adherence than to occa-
sional drug holidays. Both measures can be modied
for the examination of adherence in shorter time win-
dows. Short-term adherence may then be linked to
outcomes such as heart rate or blood pressure to explore
the health effects of drug holidays. Unfortunately, we
did not have frequently measured heart rates or blood
pressure.
Our data show that the proposed adherence measure
Cp
ave
was associated with the hospital and ED visits
whereas intended metoprolol concentration (Cp) was
not. Intended drug concentration does not always re-
ect the patients drug-taking behavior. A patients
intended metoprolol concentration assumes 100% ad-
herence. However, perfect adherence often does not
occur in the clinical setting. In our study sample of 80
patients, the mean taking adherence of metoprolol was
63% and the scheduling adherence was 33%, suggest-
ing a considerable lack of adherence. When a patient
fails to take the prescribed medication appropriately,
the intended drug concentration (Cp) remains simply a
target therapeutic level under perfect adherence to the
physicians prescription. It often does not represent the
patients true drug exposure and, as such, will not likely
be associated with the patients health outcomes.
In contrast, the deviation from the intended exposure
(Cp
ave
) incorporates the patients medication adher-
ence. Greater values of Cp
ave
indicate poorer adher-
ence. Our data suggested that Cp
ave
correlated well
with other established adherence measures such as tak-
ing and scheduling adherence. As an adherence mea-
sure, Cp
ave
for metoprolol is a strong correlate of
health care utilization in older heart failure patients. Yet
it would be reasonable to expect an association between
Cp and health outcomes in situations where there is a
clear dose-response relationship and adherence is good.
Therefore the effect of Cp is likely to vary depending
on the specic drug, the outcome considered, and the
level of adherence.
As adherence measures, Cp
ave
[i] and Cp
ave
[i]/
Cp
ave
[i] differed. Cp
ave
[i] was an estimate of the
magnitude of a patients deviation from the intended
drug exposure, whereas Cp
ave
[i]/Cp
ave
[i] estimated the
relative level of the intended drug exposure that had
been achieved by patients. The correlation between
Cp
ave
[i]/Cp
ave
[i] and taking adherence was much
stronger than that between Cp
ave
[i] and taking adher-
ence. In addition, the range of Cp
ave
[i]/Cp
ave
[i] re-
mained relatively stable at all dose levels, whereas the
range of Cp
ave
[i] tended to vary across the doses,
although not in a monotonic fashion (Fig 2, b and c).
Therefore we believe that Cp
ave
[i] might be more
appropriate when dose variation is more limited. Ulti-
mately, the choice of an appropriate adherence measure
is likely to depend on the objective of the study and the
drug under investigation.
Although it is not the focus of the current report, our
data clearly show independent adherence effects of
concomitant diuretics, ACE inhibitors, digoxin, and
spironolactone among patients who were taking meto-
prolol. The lack of inclusion of these concomitant drugs
in the nal models was likely a result of their weaker
effects compared with those of the metoprolol ad-
herence measures. Correlations between the 2
pharmacokinetic-based metoprolol adherence measures
and the taking and scheduling adherence of other con-
comitant cardiovascular drugs were relatively strong
(correlation coefcients between 0.33 and 0.51); thus,
when Cp
ave
or Cp
ave
[i]/Cp
ave
[i] was in the model, the
other drug effects were unlikely to be signicant. We
also noted that our sample represented patients who
were prescribed metoprolol. The effects of other car-
diovascular drugs in this group may not be directly
comparable to the effects of the same drugs in a general
patient population, in which patients may not be pre-
scribed metoprolol.
The proposed pharmacokinetic-based measures have
several limitations. First, as with other MEMS adher-
ence measures, they assume that the correct dose of
medication was consumed when the MEMS lid was
opened. However, unlike MEMS taking and scheduling
adherence, they account for the dosage and thus pro-
vide more exibility in handling dosage changes. Sec-
ond, the proposed measures rely on several additional
assumptions, including the correctness of the pharma-
cokinetic model and the appropriateness of the phar-
macokinetic parameters for a specic patient popu-
lation. Although we did not have individual
pharmacokinetic data in this study, it would be inter-
esting to assess the performance of this method by use
of individual pharmacokinetic models. Given the recent
work by Vrijens and Goetghebeur
14,15
showing the
importance of dosing history data in the estimation of
pharmacokinetic and pharmacodynamic models, we be-
lieve that the explanatory power of the proposed mea-
sures would increase signicantly where plasma con-
centrations are available. Third, we recognize that both
Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i] are pharmacodynami-
cally naive measures. Published ndings on metoprolol
suggested that after the drug is stopped a rebound
increase in cardiac sensitivity to endogenous cat-
echolamines was likely to ensue in the following 2 to 7
days.
29
Short of a full pharmacodynamic analysis, we
are investigating the relationship between the proposed
measures in shorter time windows (in which drug ex-
posure could be low) and clinical outcomes in the
ensuing 7-day periods. Finally, the iterative computa-
tion of the proposed measures makes them more suit-
able for clinical investigations involving a small num-
ber of drugs as opposed to a general-purpose adherence
measure for clinical practice.
As proposed measures, many important aspects of
Cp
ave
[i] and Cp
ave
[i]/Cp
ave
[i] remain to be studied.
As mentioned previously, subject-specic pharmaco-
kinetic models with individual dosing logs could be
used to further improve the predictive power of
pharmacokinetic-based adherence measures. Given
the recent development in pharmacokinetic and phar-
macodynamic models based on dosing history da-
ta,
14,15
it would be of interest to compare the predic-
tive powers of population-based and individual-
based pharmacokinetic adherence measures. Such a
comparative study is also likely to provide some
practical guidance on the balance between the mea-
sures predictive power and their general applicabil-
ity. Another important aspect is the measures utility
with regard to other cardiovascular medications, es-
pecially those drugs with longer half-lives. We are
currently studying this approach in patients pre-
scribed carvedilol, a 3-compartment drug with a
much longer half-life than metoprolol.
Notwithstanding these limitations, we conclude
that by incorporating adherence and pharmacokinetic
parameters the described adherence measures could
contribute to our understanding of the relationships
between medication adherence of metoprolol and
health care utilization in heart failure patients.
We thank the Editor and 4 reviewers for their many constructive
comments. We also thank Dr Song Mu for useful discussion during
the preparation of this report.
None of the authors has professional or nancial conicts of
interest.
References
1. Eraker SA, Kirscht JP, Becker MH. Understanding and
improving patient compliance. Ann Intern Med 1984;
100:258-68.
2. Finnerty FA Jr, Shaw LW, Himmelsbach CK. Hyperten-
sion in the inner city. II. Detection and follow-up. Cir-
culation 1973;47:76-8.
3. Sackett DL, Haynes RB, Gibson ES, Hackett BC, Taylor
DW, Roberts RS, et al. Randomised clinical trial of
strategies for improving medication compliance in pri-
mary hypertension. Lancet 1975;1:1205-7.
4. Eisen SA, Miller DK, Woodward RS, Spitznagel E, Pr-
zybeck TR. The effect of prescribed daily dose frequency
on patient medication compliance. Arch Intern Med
1990;150:1881-4.
5. Monane M, Bohn RL, Gurwitz JH, Glynn RJ, Avorn J.
Noncompliance with congestive heart failure therapy in
the elderly. Arch Intern Med 1994;154:433-7.
6. Ghali JK, Kadakia S, Cooper R, Ferlinz J. Precipitating
factors leading to decompensation of heart failure. Traits
among urban blacks. Arch Intern Med 1988;148:2013-6.
7. Urquhart J. The electronic medication event monitor.
Lessons for pharmacotherapy. Clin Pharmacokinet 1997;
32:345-56.
8. Eisen SA, Woodward RS, Miller D, Spitznagel E,
Windham CA. The effect of medication compliance on
the control of hypertension. J Gen Intern Med 1987;2:
298-305.
9. Kruse W, Nikolaus T, Rampmaier J, Weber E, Schlierf
G. Actual versus prescribed timing of lovastatin doses
assessed by electronic compliance monitoring. Eur J Clin
Pharmacol 1993;45:211-5.
10. Rudd P, Ahmed S, Zachary V, Barton C, Bonduelle D.
Improved compliance measures: applications in an am-
bulatory hypertensive drug trial. Clin Pharmacol Ther
1990;48:676-85.
11. Choo PW, Rand CS, Inui TS, Lee MT, Canning C, Platt
R. Derivation of adherence metrics from electronic dos-
ing records. J Clin Epidemiol 2001;54:619-26.
12. Rubio A, Cox C, Weintraub M. Prediction of diltiazem
plasma concentration curves from limited measurements
using compliance data. Clin Pharmacokinet 1992;22:238-
46.
13. Urquhart J. Patient non-compliance with drug regimens:
measurement, clinical correlates, economic impact. Eur
Heart J 1996;17(Suppl A):8-15.
14. Vrijens B, Goetghebeur E. The impact of compliance in
PK studies. Stat Methods Med Res 1999;8:247-62.
15. Vrijens B, Goetghebeur E. Electronic monitoring of vari-
ation in drug intakes can reduce bias and improve preci-
sion in pharmacokinetic/pharmacodynamic population
studies. Stat Med 2004;23:531-44.
16. Murray MD, Young JM, Morrow DG, Weiner M, Tu W,
Hoke SC, et al. Methodology of an ongoing randomized,
controlled trial to improve drug use for elderly patients
with chronic heart failure. Am J Geriatr Pharmacother
2004;2:53-65.
17. AARDEX Ltd Web site. Available from: URL: http://
www.aardex.net/ or http://www.aardex.ch/. Accessed on
March 30, 2004.
18. Choo PW, Rand CS, Inui TS, Lee ML, Cain E, Cordeiro-
Breault M, et al. Validation of patient reports, automated
pharmacy records, and pill counts with electronic moni-
toring of adherence to antihypertensive therapy. Med
Care 1999;37:846-57.
19. Gibaldi M, Perrier D. Pharmacokinetics. 2nd ed. New
York: Marcel Dekker; 1982.
20. Shibata MC, Flather MD, Wang D. Systematic review of
the impact of beta blockers on mortality and hospital
admissions in heart failure. Eur J Heart Fail 2001;3:
351-7.
21. Psaty BM, Koepsell TD, Wagner EH, LoGerfo JP, Inui
TS. The relative risk of incident coronary heart disease
associated with recently stopping the use of beta-
blockers. JAMA 1990;263:1653-7.
22. AHFS drug information. Bethesda (MD): American So-
ciety of Health-System Pharmacists, Inc; 2001.
23. Cerqueira PM, Cesarino EJ, Mateus FH, Mere Y Jr,
Santos SR, Lanchote VL. Enantioselectivity in the
steady-state pharmacokinetics of metoprolol in hyperten-
sive patients. Chirality 1999;11:591-7.
24. McDonald CJ, Overhage JM, Tierney WM, Dexter PR,
Martin DK, Suico JG, et al. The Regenstrief Medical
Record System: a quarter century experience. Int J Med
Inf 1999;54:225-53.
25. Murray MD, Smith FE, Fox J, Teal EY, Kesterson JG,
Stifer TA, et al. Structure, functions, and activities of a
research support informatics section. J Am Med Inform
Assoc 2003;10:389-98.
26. Le CT. Introductory biostatistics. New York: John Wiley
& Sons; 2003.
27. SAS/STAT users guide. 8th ed. Cary (NC): SAS Insti-
tute, Inc; 2004.
28. Rangno RE, Langlois S, Lutterodt A. Metoprolol with-
drawal phenomena: mechanism and prevention. Clin
29. Rangno RE, Langlois S. Comparison of withdrawal phe-
nomena after propranolol, metoprolol and pindolol. Br J
Clin Pharmacol 1982;13(Suppl 2):345S51S.
APPENDIX
To demonstrate the use of the iterative computational
algorithm, we derived the metoprolol plasma concen-
tration formulas at times t
0
0, t
1
1.0, t
2
1.2, and
t
3
2.0 hours in an iterative fashion under the assump-
tion that a patient took the medication at time t
0
0
and t
2
1.2 at a constant dose amount d.
At time t
0
0, the patient took the medication at
dose d, so d[i,0] d and the cumulative dose D[i,0]
d[i,0] d. From equation 1, we have Cp[i,0] 0.
At time t
1
1.0, the patient did not take the
medication, so d[i,1] 0. From equation 2, we
derived the cumulative dose D[i,1] d exp(k
a
) at
this time, and from equation 3, we calculated the
plasma concentration as follows:
Cp[i,1]
F d k
a
V[i](k
a
k
e
)
[exp(k
e
) exp(k
a
)]
which followed the standard 1-compartment model ex-
actly.
Similarly, at time t
2
1.2, because the subject took
another dose of the medication, d[i,2] d. The cumu-
lative dose at this time point was calculated from equa-
tion 2 as D[i,2] d exp(1.2k
a
) d. From equation
3, we calculated the plasma concentration as follows:
Cp[i,2]
F d k
a
V[i](k
a
k
e
)
[exp(k
e
) exp(k
a
)] exp(0.2k
e
)
F d exp(k
a
) k
a
V[i](k
a
k
e
)
[exp(0.2k
e
) exp(0.2k
a
)]
F d k
a
V[i](k
a
k
e
)
[exp(1.2k
e
) exp(1.2k
a
)]
which again followed the standard 1-compartment
model.
At time t
3
2.0, we had d[i,3] 0 because the
patient did not take the medication at this time point;
the cumulative dose was as follows: D[i,3]
d [exp(2k
a
) exp(0.8k
a
)]. From equation 3, we
calculated the plasma concentration at this time point as
follows:
Cp[i,3]
F d k
a
V[i](k
a
k
e
)
[exp(1.2k
e
) exp(1.2k
a
)]
exp(0.8k
e
)
F d (exp(1.2k
a
) 1) k
a
V[i](k
a
k
e
)
[exp(0.8k
e
)
exp(0.8k
a
)]
F d k
a
V[i](k
a
k
e
)
[exp(2.0k
e
)
exp(2.0k
a
)]
F d k
a
V[i](k
a
k
e
)
[exp(0.8k
e
)
exp(0.8k
a
)]
where the rst term in the above-described expression
represented the portion of plasma concentration con-
tributed by the rst dosing event at time t
0
0 and the
second term represented the portion contributed by the
second dosing event at t
2
1.2. This showed how the
iterative algorithm was able to accommodate the cu-
mulative drug effect in multiple dose regimens.
BIBN4096BS antagonizes human
-calcitonin gene related peptideinduced
headache and extracerebral artery dilatation
Background and Objective: Calcitoningenerelatedpeptide (CGRP) plays a pivotal role inmigraine pathogenesis.
BIBN4096BS is the first CGRP receptor antagonist available for human studies, and its efficacy in the acute
treatment of migraine has been demonstrated. We investigated the ability of BIBN4096BS to inhibit human
CGRP (h-CGRP)induced headache and cerebral hemodynamic changes in healthy volunteers.
Methods: Ten healthy volunteers completed this double-blind, placebo-controlled crossover study with
2.5 mg BIBN4096BS and placebo as pretreatments before a 20-minute intravenous infusion of h-CGRP
(1.5 g/min). Transcranial Doppler ultrasonography was used to measure blood flow velocity in the middle
cerebral artery (MCA); regional and global cerebral blood flow(CBF) was measured by xenon 133 inhalation
single-photon emission computed tomography. The temporal and radial artery diameter was measured by
high-frequency ultrasound. Systemic hemodynamics, end-tidal partial pressure of carbon dioxide (PETCO
2
),
and headache were monitored.
Results: Of the 10 volunteers, 6 had a CGRP-induced headache during the in-hospital phase after placebo
pretreatment but none after BIBN4096BS (P .031). BIBN4096BS did not affect changes in the diameter
of the MCA or changes in CBF induced by h-CGRP. Vasodilatation of the extracranial arteries was,
however, significantly inhibited (P < .001 for temporal artery and P .001 for radial artery).
Conclusions: These results show that BIBN4096BS effectively prevents CGRP-induced headache and extra-
cerebral vasodilatation but does not significantly affect the induced cerebral hemodynamic changes. (Clin
Kenneth A. Petersen, MD, Lisbeth H. Lassen, PhD, Steffen Birk, PhD,
Lynna Lesko, PhD, and Jes Olesen, DMSc Glostrup, Denmark, and Ridgefield, Conn
Calcitonin generelated peptide (CGRP) is a neu-
ropeptide found in the perivascular nerve terminals
surrounding arteries.
1
A measurable concentration of
CGRP is circulating in the blood at rest,
2
and CGRP
receptors are localized throughout the body.
3
Cerebral
and other cephalic arteries have a particularly rich
innervation of CGRP-containing afferent trigeminal
nerve bers, and these arteries, as studied in tissue
baths, are particularly sensitive to CGRP.
4
CGRP is
found in an increased concentration in external jugular
venous blood but not in blood from the cubital vein
during a migraine attack.
5
After infusion in patients
with migraine, CGRP caused a migraine-like headache
and in some a genuine migraine attack with associated
symptoms that were indistinguishable from the pa-
tients normal migraine attacks.
6
Peptide antagonists of
CGRP receptors have been available for experimental
studies for several years, but previously available com-
pounds have not been tested for safety and are, there-
fore, not suitable for human clinical studies. One CGRP
receptor antagonist, BIBN4096BS, has been developed
with the purpose of treating acute migraine, and a phase
II study has provided proof of efcacy.
7
Although
BIBN4096BS potently interacts with the human CGRP
From the Danish Headache Center, University of Copenhagen, and
Department of Neurology, Glostrup University Hospital, Glostrup,
and Boehringer Ingelheim Pharmaceuticals Inc, Ridgeeld.
Boehringer Ingelheim sponsored the study and provided
BIBN4096BS. The authors were independently responsible for the
study design, data analysis, and manuscript. The technical equip-
ment used was partly sponsored by the Villum Kann Rasmussen
Foundation, Toyota Foundation, and Simon Fougner Hartmann
Foundation. The Lundbeck Foundation funds the research of the
Danish Headache Center.
Received for publication June 29, 2004; accepted Oct 6, 2004.
Reprint requests: Kenneth A. Petersen, MD, Danish Headache Center,
University of Copenhagen and Department of Neurology, Glostrup
University Hospital, KAS Glostrup, DK-2600 Glostrup, Denmark.
E-mail: kapetersen@dadlnet.dk
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.10.001
202
receptor in vitro,
8,9
no information is available on its
ability to inhibit CGRP changes in human volunteers at
increased levels of the peptide as seen during a mi-
graine attack.
5
We, therefore, decided to conduct a placebo-
controlled, double-blind crossover experiment, in
which BIBN4096BS and placebo in random order were
used as pretreatment followed by infusion of human
CGRP (h-CGRP). The aims of this study were to
validate the ability of h-CGRP to provoke headache
and to describe its effect on cerebral, extracerebral, and
systemic circulatory parameters. Furthermore, we ana-
lyzed whether BIBN4096BS could partly or fully block
the changes evoked by h-CGRP. It was our hope that
the study could thus contribute to a better understand-
ing of the role of CGRP in neurovascular headache and
the site of action of this new antimigraine compound.
METHODS
Design and participants. This was a placebo-
controlled, double-blind crossover study that included
11 healthy subjects (7 men and 4 women). One female
participant had claustrophobia that developed during
baseline single-photon emission computed tomography
(SPECT) scanning and was excluded before any trial
medication was given. Ten participants completed both
treatment days. The participants were aged 24 to 31
years (mean, 26.5 years) and weighed 68 to 89.4 kg
(mean, 77.4 kg). The participants had no current or
previous cardiovascular, cerebrovascular, endocrine, or
neurologic disorder, including no migraine, hypoten-
sion, or hypertension. A frequency of tension-type
headache of 4 d/mo or lower was accepted. On the day
of enrollment, physical and neurologic examination,
electrocardiography, and blood sampling were done.
The healthy volunteers were randomized to receive
either 2.5 mg BIBN4096BS or placebo (xylitol 5%) as
an intravenously administered pretreatment of 10 min-
utes duration. After a free interval of 10 minutes,
1.5 g/min h-CGRP was administered continuously
for 20 minutes on both trial days. The 2 trial days were
separated by at least 1 week.
Boehringer Ingelheim GmbH supplied BIBN4096BS
and performed the randomization and blinding, which
was balanced (ClinPro, version 6; Clinical Systems,
Inc, Garden City, NY). The dose effective in the treat-
ment of acute migraine attacks was used.
7
Human-CGRP was purchased from Clinalfa AG,
Lufelngen, Switzerland. In a study performed previ-
ously, we used a dose of 2 g/min.
6
This dose, how-
ever, induced pronounced hypotension that in 2 patients
necessitated premature termination of the infusion. In
the current study we, therefore, used the lower dose of
1.5 g/min.
All participants gave written informed consent be-
fore randomization. The Ethical Committee of Copen-
hagen (KA00079gs) and the Danish Medicines Agency
(2612-1376) approved the study, which was conducted
in accordance with the Helsinki II Declaration and the
Guidelines for Good Clinical Practice.
10
Recording of adverse events. Every 15th minute
from time (T) zero (T
0
) (baseline) to T
240
(end of study
period), the volunteers were questioned regarding the
presence of adverse events (AEs) and rated headache.
Between questionings, the participants self-reported
any changes that they might have. The intensity of the
AEs was graded as mild, moderate, or severe, and their
relationship to study medication was classied as re-
lated or not related by the investigator. Headache in-
tensity was scored on an 11-point verbal rating scale
with 0 indicating no headache; 1 indicating a feeling of
the occurrence of something unusual inside the head,
not necessarily actual pain; 5 indicating headache of
medium severity; and 10 indicating worst imaginable
headache. Accompanying symptoms were recorded ac-
cording to the International Headache Classication.
11
During the study period, the investigator recorded the
AEs. After discharge, the volunteers made an hourly
recording of AEs up to 24 hours after the infusion of
placebo or BIBN4096BS.
Cerebral blood ow measurements. Global and re-
gional cerebral blood ow (CBF) was measured with
xenon 133 inhalation and SPECT with a brain-
dedicated camera (Ceraspect; DSI, Waltham, Mass).
The apparatus consisted of a stationary annular sodium
iodide crystal and a fast-rotating collimator system.
Each rotation took 10 seconds, thereby acquiring 1
frame in a 30-frame dynamic protocol of
133
Xe inha-
lation, with 3 background, 9 wash-in, and 18 wash-out
frames by use of the Kanno-Lassen algorithm.
12
A
photoelectric window of 70 to 100 keV was used.
Thirty-two slices were reconstructed in a 64 64
matrix with each pixel measuring 0.33 0.33 cm by
use of a Butterworth one-dimensional lter (cutoff, 1.5;
order, 6). The 32 slices were reduced to sets of 8
transaxial slices generated by adding 4 slices together
to a total slice thickness of 1.32 cm. A correction by use
of the Chang algorithm (m 0.05 cm) and nose
artifact was performed. The output for each pixel was
the inhibition constant (K
i
) value, and ow values were
estimated from these by use of the partition coefcient
() of 0.85 (gray matter).
A Datex Normocap 200 (Dameca, Roedovre, Den-
mark) was used for end-tidal partial pressure of carbon
2005;77(3):202-13 CGRP receptor antagonism and vascular headache 203
dioxide (PETCO
2
) measurements during the CBF acqui-
sitions. A Ceratronic XAS SM 32C (Randers, Den-
mark) was used for the
133
Xe administration. Each
measurement lasted 5 minutes.
Calculations of ow in the perfusion territories of the
major cerebral arteries were performed by tting of
standard vascular regions of interest on the 5 rostral
slices at 3.6, 5.0, 6.3, 7.6, and 9 cm above the orbito-
meatal line. Flow in the territory of the middle cerebral
artery (MCA) (rCBF
MCA
) was calculated as a mean of
the left and right side.
Transcranial Doppler and C-scan. Transcranial
Doppler (TCD) ultrasonography (2 MHz) (Multi-Dop
X; DWL, Sipplingen, Germany) was used for the mea-
surement of blood ow velocity. The recordings were
done simultaneously and bilaterally as previously de-
scribed but with handheld probes.
13
Along the MCA, a
xed point was found for the measurement. The xed
point was chosen as close as possible to the bifurcation
of the anterior cerebral artery and MCA. The same x
point was used for each individual and for each record-
ing, for which the signal was optimized. On the basis of
the envelope curve (the spectral TCD curve), a time
averaged mean (V
mean
) over approximately 4 cardiac
cycles or 4 seconds was calculated by the built-in
software (version 7.40x of MDX TCD-7 software for
Multi-Dop X hardware, DWL). The nal measure used
for each time point was an average of 4 cycles (V
MCA
).
Simultaneously with the TCD recording, a mask cov-
ering the subjects mouth and nose region was placed
for the measurement of PETCO
2
(Datex Normocap 200;
Dameca).
A high-resolution ultrasound scanner, C-scan
(Dermascan C, 20 MHz; bandwidth, 15 MHz) (Cortex
Technology, Hadsund, Denmark),
14
was used to mea-
sure the diameter of the left temporal and left radial
artery. The diameter of the former was measured at the
front branch of the supercial temporal artery and the
latter at the wrist. To ensure that the repeated measure-
ments with TCD and C-scan were performed in the
same place, marks were drawn on the skin. After the
last recording on the rst trial day, the coordinates of
the marks were kept for reuse on the following trial
day.
Pharmacokinetics. Plasma concentrations of
BIBN4096BS were sampled at the following time
points: T
10
(baseline), T
9.5
, T
30
, T
60
, and T
180
on each
trial day in Vacutainer blood-collecting tubes with eth-
ylenediaminetetraacetic acid (K3 10-mL glasses;
Becton Dickinson, Rutherford, NJ). Samples were
stored on ice for a maximum of 30 minutes before
centrifuged for 10 minutes (2000 rpm) at 4C. The
plasma was stored at 20C until analyzed at Boehr-
inger Ingelheim Pharma GmbH & Co KG (Biberach an
der Riis, Germany). The plasma concentration of
CGRP was determined twice, at baseline (T
10
) and at
the end of the h-CGRP infusion (T
40
).
BIBN4096BS antibodies. BIBN4096BS was modi-
ed with succinic acid anhydride. This hapten was
covalently coupled to human serum albumin. Poly-
clonal antibodies were produced by immunization of
3-month-old female New Zealand rabbits with the im-
munogen in complete Freunds adjuvant. After several
booster immunizations, the antibodies were puried
from rabbit serum by use of protein ASepharose
(Sepharose is a registered trademark of Amersham
Biosciences).
BIBN4096BS analytic methods. The procedures
were conducted in accordance with current interna-
tional guidelines.
15
In this competitive enzyme-linked
immunosorbent assay, the biotinylated anti-
BIBN4096BS antibodies (immunoglobulin G fraction)
were bound to microtiter plates that were adsorptive-
coated with avidin. BIBN4096BS in the plasma sample
competed with added horseradish peroxidaselabeled
BIBN4096BS reagent for binding sites on the solid-
phase antibodies. After incubation, unbound
BIBN4096BS and plasma components were removed
by washing. Antibody-bound enzyme activity was de-
tected with a chromogenic substrate. The amount of
colored product formed was measured photometrically
and decreased with the increasing concentration of
BIBN4096BS in the plasma sample. The BIBN4096BS
concentration corresponding to the measured optical
absorbance was calculated via data tting of the non-
linear standard curve.
To compensate for slight variations in immuno-
chemical reaction parameters (such as temperature and
antibody binding capacity) between microplates, a stan-
dard curve was included on each plate. All steps of the
enzyme-linked immunosorbent assay were performed
at 22C 1C, which corresponded to room tempera-
ture of the air-conditioned laboratory.
Assay precision as assessed from 886 triplicate de-
terminations by construction of a precision prole was
9.1% coefcient of variance (CV) at the lower limit of
quantication, 2.7% CV at the upper limit of quanti-
cation, and 1.6% CV in the middle of the working range
(0.5 ng/mL).
Human-CGRP analysis. The analysis of
h-CGRP plasma concentrations was performed at the
Department of Clinical Physiology and Nuclear Medi-
cine, Glostrup Hospital (Glostrup, Denmark). The
method of analysis has been described in detail else-
204 Petersen et al MARCH 2005
where.
2
In this study only 100 L of serum was used.
The normal values were 85 35.4 pmol/L for women
and 88 36.2 pmol/L for men (Schifter S, oral com-
munication, November 2002).
Trial procedure. The healthy volunteers began the
study at 8 AM, headache-free. For the preceding 8 hours,
they had abstained from drinking coffee, tea, and
caffeine-containing beverages and smoking tobacco
and they had not taken any medication, except oral
contraception. They rested in the supine position
throughout the study period (T
20
to T
180
) Two intra-
venous catheters (Optiva*2 [18 gauge]; Johnson &
Johnson, Ethicon SpA, Pomezia, Italy) were inserted
into the cubital veins, one for the administration of
human CGRP and BIBN4096BS and the other for
blood sampling. The volunteers rested for at least 30
minutes before baseline values of CBF, V
MCA
, tempo-
ral and radial diameter, blood pressure (BP), heart rate
(HR), and electrocardiogram were recorded. The start
of infusion of 2.5 mg BIBN4096BS or placebo was
designated as time zero (T
0
). The infusion lasted 10
minutes. At T
20
, a 20-minute infusion of h-CGRP (1.5
g/min) was initiated. Infusions were administered by
a time- and volume-controlled infusion pump (Braun
perfusor; B. Braun Melsungen AG, Melsungen,
Germany).
All measurements, except the CBF measurements,
were recorded quarterly for 3 hours (study period), and
BP, HR, electrocardiogram (Cardiofax; Nihon Kohden
Corporation, Tokyo, Japan), AEs, and headache were
recorded for an additional hour. BP and HR were
measured every 5 minutes for the rst hour and there-
after every 15th minute with an automatically inating
cuff (Omega 1400, In Vivo Research Laboratories Inc,
Copiague, NY). In the observation period from T
180
to
T
240
, the participants were allowed to sit upright. Three
SPECT scans were done as follows: at baseline, at T
60
,
and at T
90
. V
MCA
was measured immediately after each
SPECT scan.
The estimated perfusion (rCBF
x
) in the area of a
given artery (x) is dependent on the mean blood ow
velocity [V
mean(x)
] and the cross-sectional area (r
2
)
of the artery. The following equation is valid for the
regional CBF:
rCBF
(x)
Vmean
(x)
r
2
Hence,
Diameter
rCBF
2(x)
V
mean
2(x)
V
mean
1(x)
rCBF
1(x)
1
100
Diameter is the relative percentage change in diam-
eter, V
mean1(x)
is the mean blood velocity before infu-
sion of drugs, and V
mean2(x)
is the velocity at a relevant
time point after the infusion; the same designation is
applied for rCBF.
16,17
Statistics. Baseline was calculated as a mean of the
measurements at time points T
20
and T
10
in the
analysis. Values are presented as means SD. P .05
was considered signicant. All analyses were per-
formed by use of SPSS statistical software, version 10.0
(SPSS Inc, Chicago, Ill).
For changes over time on each trial day, V
MCA
,
global CBF, rCBF
MCA
, diameter of the temporal and
radial artery, BP, and PETCO
2
were analyzed by a uni-
variate ANOVA for the factors time and subject. If a
signicant change was found, a post hoc analysis
(Dunnett multiple comparisons test) was performed to
localize the change. To eliminate the risk of mass
signicance of measurements with numerous repeated
measurements, 4 points of interest were chosen as
follows: baseline, 45 minutes, 105 minutes, and 165
minutes. Absolute values were used for the statistical
analysis. For the comparison between BIBN4096BS
and placebo, a paired t test was performed for the
following measurements: V
MCA
, global CBF,
rCBF
MCA
, diameter of the temporal and radial arteries,
BP, and PETCO
2
. The summary measure for the t test
was the area under the curve (AUC) calculated on
percentage changes from baseline.
Immediate headache was dened as any headache
during the rst 60 minutes after the start of the
h-CGRP infusion. Any headache occurring thereafter
was referred to as delayed headache. Peak values and
area under the curve for headache (AUC
headache
) were
compared between the 2 trial days by use of the
Wilcoxon signed rank test. The occurrence of headache
and AEs on the 2 trial days was compared by use of the
McNemar test.
RESULTS
Baseline values. All baseline measurements of the
hemodynamic responses are summarized in Table I.
Only the baseline PETCO
2
measured simultaneously
with the TCD recordings showed a signicant differ-
ence between study days (P .03). The nding was
interpreted as incidental and was not taken into account
in the processing of data.
Effect of BIBN4096BS on CGRP-induced head-
ache and other AEs. On placebo days, 5 participants
had an immediate headache and 3 had a delayed head-
ache; the maximum immediate headache score was 2
and the maximum delayed headache score was 1. No
participants had an immediate headache but 1 had a
delayed headache after BIBN4096BS pretreatment.
The delayed headache occurred 6 hours after the infu-
sion of BIBN4096BS, lasted 3 hours, and was scored 1.
The effect of BIBN4096BS in preventing immediate
headache was signicant (P .034 for peak headache
and P .04 for AUC
headache
) and in preventing the
occurrence of any headache during the in-hospital
phase (P .031, McNemar test).
After placebo pretreatment, h-CGRP caused ush-
ing in all participants and all but 1 had bilateral con-
junctival injection. Eight experienced a sensation of
heat. Five reported palpitations. None of these CGRP-
induced changes were seen on days when
BIBN4096BS was administered as pretreatment (Table
II). AEs that could possibly be assigned to the CGRP
receptor antagonist were located to the infusion site.
CBF. Global CBF increased signicantly after
h-CGRP on both study days (P .007 after placebo
and P .009 after BIBN4096BS pretreatment). The
increase was measured 20 minutes after the h-CGRP
infusion was stopped. No difference was found be-
tween the 2 days (P .42).
After h-CGRP, rCBF
MCA
increased signicantly on
both trial days (P .003 and P .01), again 20
minutes after the h-CGRP infusion. No signicant
difference was observed between the 2 study days (P
.38). Data were not corrected for PETCO
2
, because no
signicant changes were found on either day (P .2
and P .6) or between treatment days (P .1).
TCD. V
MCA
did not vary signicantly over time
(P .3 for placebo and P .7 for BIBN4096BS), and
between the 2 trial days, no difference was seen (P
.74). On the basis of the rCBF
MCA
and V
MCA
measure-
ments, the effect on the relative percentage diameter
change of MCA can be estimated.
16,18
As seen in Table
III, a dilation of the MCA was found on both study
days. Compared with baseline, the dilation occurring on
placebo days reached signicance at T
60
(P .005).
This corresponded to a diameter increase of 9.3%
8.1%. When BIBN4096BS was given as pretreatment,
Table II. Effect of BIBN4096BS pretreatment on
h-CGRPinduced symptoms
Symptom
Placebo
plus
h-CGRP
(No.)
BIBN4096BS
(2.5 mg) plus
h-CGRP
(No.)
P value
(McNemar
test)
Flushing 10 0 P .002
Heat sensation 8 0 P .008
Palpitations 5 0 P .063
Conjunctival
injection
9 0 P .004
Headache 6 0 P .031
The ushing and conjunctival injection was based on the investigators
observations. Heat sensation and palpitation were reported and headache was
systematically scored. The data shown are from the entire in-hospital study
period.
Table I. Baseline values of measured variables
Measured variable
Placebo plus
h-CGRP
BIBN4096BS
(2.5 mg) plus
h-CGRP
P
value
Global CBF (mL 100 g brain tissue
1
min
1
) 46.7 10.8 45.6 10.5 .5
rCBF
MCA
(mL 100 g brain tissue
1
min
1
) 45.9 10.5 44.7 10.6 .3
PETCO
2
(mm Hg)
CBF 39 3.5 39 4.1 .9
TCD 41 3.0 39 3.6 .03*
V
MCA
(cm/s) 78 17.0 74 15.3 .1
C-scan
Temporal (mm) 1.26 0.4 1.29 0.3 .8
Radial (mm) 2.76 0.5 2.63 0.4 .2
Systolic blood pressure (mm Hg) 113 8 113 6 .9
Diastolic blood pressure (mm Hg) 64 8 64 5 .8
Mean arterial blood pressure (mm Hg) 79 6 80 7 .5
Heart rate (beats/min) 54 8 53 4 .7
Plasma CGRP (pmol/L) 89 20.3 91 20.4 .5
h-CGRP, Human -calcitonin generelated peptide; CBF, cerebral blood ow; rCBF
MCA
, cerebral blood ow in territory of middle cerebral artery; PETCO
2
, end-tidal
partial pressure of carbon dioxide; TCD, transcranial Doppler; V
MCA
, middle cerebral artery blood ow velocity (average of 4 cycles); CGRP, calcitonin generelated
peptide.
*Signicant difference between baseline on placebo and BIBN4096BS pretreatment days (paired t test).
an increase was seen as well, but it did not reach
statistical signicance (P .2). There was no differ-
ence between placebo and BIBN4096BS pretreatment
(P .17). Data were not corrected for PETCO
2
, because
no signicant changes were found on the trial days
(P .4 on both days) or between treatment days (P
.45).
C-scan. A signicant change in temporal artery di-
ameter over time on placebo days was observed (P
.001). This signicance was seen at all time points after
the infusion of h-CGRP.
A signicant diameter change was not seen when
BIBN4096BS was infused as pretreatment (P .7).
The increase in temporal artery diameter was signi-
cantly inhibited by BIBN4096BS (P .001). The
radial artery revealed a similar nding, with a signi-
cant change on placebo days (P .01) and a nonsig-
nicant response after BIBN4096BS pretreatment
(P .07). The difference between placebo and
BIBN4096BS was signicant (P .001) (Fig 1).
Peripheral hemodynamics. Table IV summarizes
data on systolic BP, diastolic BP, mean arterial BP, and
HR. No signicant time-dependent changes were seen
in systolic, diastolic, or arterial mean BP on either trial
day or between days. The HR increased signicantly on
placebo days (P .001) at all time points compared
with baseline. This increase was not seen on
BIBN4096BS pretreatment days. The increase in HR
was signicantly inhibited by BIBN4096BS (P .003).
Pharmacokinetics of BIBN4096BSand h-CGRP. The
highest measured plasma concentration of
BIBN4096BS occurred just before the end of infusion
at 9.5 minutes after the start of the infusion, with a
mean of 170.4 25.4 ng/mL. The lowest plasma
concentration was measured 180 minutes after the start
of the infusion, with a mean of 8.7 5.7 ng/mL. No
BIBN4096BS was detected on placebo days (Fig 2).
The pharmacokinetic parameters for BIBN4096BS
have previously been published and are summarized.
On the basis of administration of 5 and 10 mg intrave-
nously for 10 minutes, BIBN4096BS had a total plasma
clearance of 12 L/h and a terminal half-life of 2.5 hours.
Approximately 15% of the dose was excreted un-
changed in urine, with a mean renal clearance of 2 L/h.
The geometric mean maximum plasma concentration
after 2.5 mg of BIBN4096BS was 210 ng/mL.
19
At the end of the CGRP administration, a signicant
difference between treatment days was found (P
.001, paired t test); the CGRP concentration was 342
68 pmol/L when placebo was administered as pretreat-
ment and 442 70 pmol/L when BIBN4096BS was
given (Fig 3). To elucidate whether this difference was
caused by a cross-reaction between BIBN4096BS and
our CGRP analytic kit, blood samples from 3 healthy
volunteers with the following concentrations of
BIBN4096BS were analyzed for CGRP: sample 1 and
2, 0 ng/mL (control); sample 3, 1000 ng/mL; sample 4,
500 ng/mL; sample 5, 250 ng/mL; and sample 6, 125
ng/mL. The analysis did not reveal any correlation
between the concentration of BIBN4096BS and CGRP.
Thus BIBN4096BS did not cross-react with CGRP
detection in our analytic kit.
DISCUSSION
This study has demonstrated that a specic CGRP
antagonist, BIBN4096BS, can prevent CGRP-induced
symptoms such as headache, ushing, heat sensation,
and palpitations. It furthermore prevents a CGRP-
induced increase in HR and dilatation of the supercial
temporal and radial artery, whereas it has no signicant
Table III. Effect of BIBN4096BS pretreatment on h-CGRPinduced cerebral hemodynamic changes
Time point
V
MCA
(cm/s)
Diameter
MCA
(%) and
Area
MCA
(%)
rCBF
MCA
(mL 100 g
1
min
1
)
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Baseline 78 17.0 74 15.3 0 (0) 0 (0) 45.9 10.5 44.7 10.6
60 min 75 15.5 75 12.9 9.3 8.1*
(19.95 17.9)
3.8 5.4
(7.96 11.3)
52.4 9.4 49.1 12.1
90 min 77 14.3 74 15.5 3.4 7.5
(7.4 15.1)
3.0 7.9
(6.8 16.3)
49.3 12.7 47.9 10.1
Values are given as mean SD. The percentage change in the mean diameter (left and right sides) of the middle cerebral artery (MCA) was estimated according to
Dahl et al.
16
V
MCA
and rCBF on left and right sides analyzed separately showed similar results.
*Difference from baseline on 2 trial days (ANOVA, Dunnett post hoc): P .003.
Difference from baseline on 2 trial days (ANOVA, Dunnett post hoc): P .002.
effect on CGRP-induced increase in CBF and dilatation
of the MCA.
Localization, function, and role of CGRP in mi-
graine. CGRP is one of the most potent vasodilators
known.
4
In the brain, immunohistochemical studies
have located the peptide to perivascular sensory
Table IV. Effect of BIBN4096BS pretreatment on h-CGRPinduced systemic hemodynamic changes
Time
point
Systolic blood pressure
(mm Hg)
Diastolic blood pressure
(mm Hg)
Mean blood pressure
(mm Hg)
Heart rate
(beats/min)
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Placebo
plus CGRP
BIBN4096BS
plus CGRP
Baseline 113 8 113 6 64 8 64 5 79 6 80 7 54 8 53 4
45 min 115 9 111 9 59 5 67 7 80 6 81 7 72 8* 57 10
105 min 112 11 110 10 62 6 66 7 83 9 80 7 61 8* 54 5
165 min 115 12 115 7 66 9 67 9 81 10 80 7 60 8 57 6
All values are given as mean SD. Only time points selected for the statistical analysis (ANOVA) are shown.
*Signicant difference between placebo and BIBN4096BS pretreatment (P .005, paired t test).
Fig 1. Diameter (in millimeters) of supercial temporal artery and radial artery. Asterisk denotes
signicant difference (P .05) between placebo and BIBN4096BS pretreatment on human
-calcitonin generelated peptide (h-CGRP)induced vasodilatation (paired t test). Open squares,
Temporal artery, placebo pretreatment; open circles, radial artery, placebo pretreatment; closed
squares, temporal artery, 2.5-mg BIBN4096BS pretreatment; solid circles, radial artery, 2.5-mg
BIBN4096BS pretreatment (mean SD).
C-bers surrounding cerebral and extracerebral arteries
and to the trigeminal ganglion cell bodies.
20
CGRP is present in plasma from healthy volun-
teers at rest.
2
We have previously shown that infu-
sion of the selective CGRP receptor antagonist
BIBN4096BS did not alter CBF or the diameter of
cerebral and peripheral arteries.
21
Hence circulating
levels of CGRP do not seem to exert a tonic dilator
action in these vascular beds. During a migraine
attack
5
and between attacks, CGRP is increased in
venous blood,
22
implicating a role of CGRP in mi-
graine pathogenesis. During an attack, the plasma
levels were increased (2-2.5 times) compared with
normal controls.
5
The infusion of 1.5 g/min of
h-CGRP in the current study increased the plasma
concentration approximately 3 to 4 times.
In healthy volunteers the infusion of CGRP has pre-
viously been shown to induce a sensation of fullness in
the head or mild headache,
23,24
corresponding to a
headache score of 1 in the current study. Because the
Fig 2. Plasma concentration of BIBN4096BS after 10-minute infusion of 2.5 mg. Values are given
as mean SD.
Fig 3. Plasma concentration of calcitonin generelated peptide (CGRP) on 2 different trial days.
Two asterisks denote a signicant difference (P .001) between BIBN4096BS and placebo
pretreatments (mean SD) (paired t test).
study was performed in healthy volunteers, a genuine
migraine attack was not expected. Why a migraine
attack is induced only in migraine patients and not in
healthy volunteers is most likely a threshold phenom-
enon. Through use of the human experimental head-
ache model, previously performed studies have shown
that headache response and MCA changes in healthy
volunteers compose a valid model for investigations of
migraine pathogenesis.
25-27
Methods of pharmacologic studies of cephalic he-
modynamics relevant to migraine. We have developed
the combination of methods applied in this study to
provide the most extensive and precise description of
pharmacologic effects on cephalic hemodynamics as
related to migraine. For CBF measurement, positron
emission tomography (PET) and magnetic resonance
(MR) techniques provide better spatial solution than
133
Xe SPECT, but this is of minor importance in phar-
macologic studies because no major variation has been
described in the response of cortical CBF to drugs.
28-30
Our pharmacologic experiments last several hours, and
the costs of blocking MR and especially PET equip-
ment for half a day or more on at least 30 experimental
days in 1 drug study are prohibitive in most centers.
Furthermore, studies of migraine patients in whom nau-
sea and vomiting may develop are very difcult inside
a magnet. The day-to-day CV of
133
Xe SPECT mea-
surements is 8%.
29
These gures are roughly equal to
those of PET and slightly better than MR-determined
CBF.
31
TCD measurements of velocity (V
mean
) as
stand-alone measurements are difcult to interpret be-
cause they depend both on the diameter of the MCA
and on CBF in the MCA territory (rCBF
MCA
), but they
are useful in combination with quantitative rCBF. TCD
measurements are most precise in the MCA, with a CV
of 26% between subjects, 20% between sides, 16%
between days, and 7% 5 minutes apart.
13
In studies of
headache, xed probes and continuous measurements
cannot be used because the probes cause local discom-
fort and headache.
32
Because of the relatively high
intersubject and interobserver variability of TCD mea-
surements, the current study was designed as a cross-
over study and the same trained investigator performed
all measurements. Power calculations based on an SD
of TCD measurements of 7.8 cm/s and allowing a 5%
type I and a 20% type II error showed that 7 subjects
were needed in a crossover design to detect a difference
of 15 cm/s (change from baseline of approximately
25%).
13
An alternative to these measurements could be MR
angiography, but MR angiography is difcult to time
with CBF measurements and poses problems in relation
to migraine induction. Its precision is not yet suf-
ciently documented and may not be high enough to
detect modest pharmacologic effects. In summary, the
combination of methods applied in this study remains
preferable to other alternatives for a broad character-
ization of the effects of messenger molecules and drugs
in the cephalic circulation.
Hemodynamic effects of CGRP. CGRP increases
CBF in dogs
33
and striatal blood ow in rats.
34
In
guinea pigs a pretreatment with urea was necessary
before CGRP-induced CBF changes could be in-
duced.
35
The peptide is partly responsible for cerebral
vasodilatation in response to hypotension,
36
postocclu-
sive hyperemia,
37
and cortical spreading depression.
38
In humans h-CGRP (0.6 g/min for 3 hours)
caused a signicant decrease of approximately
6 mL 100 g brain tissue
1
min
1
in the left hemi-
sphere CBF and a nonsignicant decrease of 5 mL
100 g brain tissue
1
min
1
in the right hemisphere in
one study.
39
The decrease was seen 30 minutes after the
initiation of the CGRP infusion but not after 2.5 hours
and was attributed to a decrease in PETCO
2
. Design, area
of measurement, and method of administration of
CGRP were different from those in this study, which
showed a slight but signicant increase in global and
regional CBF without changes in PETCO
2
.
In healthy volunteers (n 5) long-term infusion of
CGRP with a maximum dosage of 1.15 g/min caused
a nonsignicant decrease in the blood ow velocity in
the MCA (60 cm/s to 54 cm/s); CBF was not mea-
sured.
40
A similar decrease was seen with 0.6 g/
min.
39
In patients studied after subarachnoidal hemor-
rhage, V
MCA
decreased only on the side of the bleeding.
It was suggested that preconstriction of the artery or a
disruption of the blood-brain barrier was necessary for
CGRP to dilate MCA.
41,42
In the current study V
MCA
did not change after CGRP. We found, however, that
global CBF, as well as rCBF
MCA
, increased after CGRP
administration and calculated that CGRP increased the
diameter of the MCA
16
(Table III).
BIBN4096BS pretreatment did not prevent the
CGRP-induced increase in global CBF or rCBF
MCA
.
Furthermore, it did not prevent the CGRP-induced in-
crease in MCA diameter. However, the study was not
powered to rule out effects on MCA diameter, which is
calculated from 2 observed values and, therefore, is
more variable than the parameters alone.
In healthy volunteers and patients with subarachnoi-
dal hemorrhage, h-CGRP did not affect systolic or
diastolic BP. HR was increased signicantly.
24,41
In
healthy male volunteers 1.18 g/min induced a 30%
increase in HR.
43,44
In the current study we found
similar effects of CGRP.
CGRP receptors and CGRP receptor blockade. The
CGRP receptor can be subdivided into 2 subtypes,
CGRP
1
and CGRP
2
; however, some discussion of the
validity of this classication exists.
45
The CGRP
1
re-
ceptor predominates in cerebral arteries, trigeminal
ganglion cell bodies, and perivascular nerve bers.
4,46
BIBN4096BS is the rst CGRP receptor antagonist
available for human clinical trials. BIBN4096BS has
high specicity and afnity for the human CGRP re-
ceptor; in human SK-N-MC (neuroblastoma cell line of
human origin) cells, it displayed an afnity of K
i
of
14.4 6.3 pmol/L, and for rat spleen, it demonstrated
a K
i
of 3.4 0.5 nmol/L.
47
The antagonist is 10 times
more potent in blocking rat or h-CGRP and h-CGRP
responses compared with the peptide antagonist
CGRP
8-37
. Furthermore, it was 10 times more potent in
inhibiting responses in rat atrium than in rat vas defer-
ens.
48
These results, together with the ndings that
SK-N-MC cell lines only express the CGRP
1
recep-
tor,
49
indicate that BIBN4096BS preferably binds to
this receptor subtype. BIBN4096BS inhibits both
CGRP-induced BP changes in rats and the neurogenic-
induced increase in facial blood ow in marmosets.
47,50
In isolated human middle cerebral, meningial, and tem-
poral arteries, BIBN4096BS blocks the dilatatory re-
sponse to CGRP.
9,51
In our study, the CGRP plasma concentration was
found to be signicantly higher on BIBN4096BS pre-
treatment days than on placebo pretreatment days. An
explanation would be a cross-reaction between
BIBN4096BS and CGRP, but this was ruled out. The
nding can be explained by a complete blockade of
CGRP receptors by the antagonist, resulting in a higher
concentration of circulating CGRP. This explanation
remains hypothetic until conrmed by further studies.
Our study showed that BIBN4096BS completely in-
hibits the effects of CGRP on the supercial temporal
and radial artery, as well as its effect on HR. In con-
trast, it had no signicant effect on the CGRP-mediated
CBF and V
MCA
increase. These results are important
with regard to the understanding of the action of CGRP
blockade in the treatment of acute migraine. In a phase
II proof-of-concept study in migraine patients,
BIBN4096BS showed a signicant effect in aborting
migraine attacks.
7
In agreement with this clinical result,
BIBN4096BS completely prevented CGRP-induced
headache in normal volunteers in our study.
In summary, our study has demonstrated the potency
of the CGRP receptor antagonist BIBN4096BS in hu-
mans. Thus it prevented all symptoms and signs elicited
by h-CGRP and caused no symptoms on its own.
More importantly, h-CGRPelicited headache was
completely prevented. Given that the compound inhib-
ited all effects of h-CGRP on extracerebral arteries
but had no effect on the induced increase in CBF or on
the calculated diameter of the MCA, it is suggested that
BIBN4096BS prevents or treats headache predomi-
nantly in an extracerebral manner. Whether the effect
takes place in the dura mater or in extracranial arteries
and whether areas of the brain stem or hypothalamus
devoid of a blood-brain barrier also play a role remain
to be determined.
We thank Lene Elkjr and Kirsten Bruunsgaard for excellent
technical support and Dr Kirsten Kasse for making everything work
perfectly.
None of the authors at the Danish Headache Center received
payment or honoraria. Lynna Lesko is an employee of Boehringer
Ingelheim Pharmaceuticals Inc. However, she did not receive special
honoraria from the company and has no equity or other ownership
interest.
References
1. Uddman R, Edvinsson L, Ekblad E, Hakanson R, Sundler
F. Calcitonin gene-related peptide (CGRP): perivascular
distribution and vasodilatory effects. Regul Pept 1986;
15:1-23.
2. Schifter S. Circulating concentrations of calcitonin gene-
related peptide (CGRP) in normal man determined with a
new, highly sensitive radioimmunoassay. Peptides 1991;
12:365-9.
3. Brain SD, Williams TJ, Tippins JR, Morris HR, MacIn-
tyre I. Calcitonin gene-related peptide is a potent vaso-
dilator. Nature 1985;313:54-6.
4. Jansen I, Mortensen A, Edvinsson L. Characterization of
calcitonin gene-related peptide receptors in human cere-
bral vessels. Vasomotor responses and cAMP accumula-
tion. Ann N Y Acad Sci 1992;657:435-40.
5. Goadsby PJ, Edvinsson L, Ekman R. Vasoactive peptide
release in the extracerebral circulation of humans during
migraine headache. Ann Neurol 1990;28:183-7.
6. Lassen LH, Haderslev PA, Jacobsen VB, Iversen HK,
Sperling B, Olesen J. CGRP may play a causative role in
migraine. Cephalalgia 2002;22:54-61.
7. Olesen J, Diener HC, Husstedt IW, Goadsby PJ, Hall D,
Meier U, Pollentier S, et al. Calcitonin gene-related pep-
tide receptor antagonist BIBN 4096 BS for the acute
treatment of migraine. N Engl J Med 2004;350:1104-10.
8. Schindler M, Doods HN. Binding properties of the novel,
non-peptide CGRP receptor antagonist radioligand
[(3)H]BIBN4096BS. Eur J Pharmacol 2002;442:187-93.
9. Verheggen R, Bumann K, Kaumann AJ. BIBN4096BS is
a potent competitive antagonist of the relaxant effects of
alpha-CGRP on human temporal artery: comparison with
CGRP(8-37). Br J Pharmacol 2002;136:120-6.
10. Good Clinical Practice: consolidated guideline (E6) step 4
adopted 1996. Available from: URL: http://www.ich.org.
11. Classication and diagnostic criteria for headache disor-
ders, cranial neuralgias and facial pain. Headache Clas-
sication Committee of the International Headache So-
ciety. Cephalalgia 1988;8:1-96.
12. Kanno I, Lassen NA. Two methods for calculating re-
gional cerebral blood ow from emission computed to-
mography of inert gas concentrations. J Comput Assist
Tomogr 1979;3:71-6.
13. Thomsen LL, Iversen HK. Experimental and biological
variation of three-dimensional transcranial Doppler mea-
surements. J Appl Physiol 1993;75:2805-10.
14. Nielsen TH, Iversen HK, Tfelt-Hansen P. Determination
of the luminal diameter of the radial artery in man by
high frequency ultrasound: a methodological study. Ul-
trasound Med Biol 1990;16:787-91.
15. Shah VP, Midha KK, Dighe S, McGilveray IJ, Skelly JP,
Yacobi A, et al. Analytical methods validation: bioavail-
ability, bioequivalence and pharmacokinetic studies.
Conference report. Eur J Drug Metab Pharmacokinet
1991;16:249-55.
16. Dahl A, Russell D, Nyberg-Hansen R, Rootwelt K. Effect
of nitroglycerin on cerebral circulation measured by
transcranial Doppler and SPECT. Stroke 1989;20:
1733-6.
17. Sorteberg W. Cerebral artery velocity and cerebral blood
ow. In: Newell DW, Aaslid R, editors. Transcranial
Doppler. New York: Raven Press; 1992. p. 57-66.
18. Dahl A, Russell D, Nyberg-Hansen R, Rootwelt K. A
comparison of regional cerebral blood ow and middle
cerebral artery blood ow velocities: simultaneous mea-
surements in healthy subjects. J Cereb Blood Flow Metab
1992;12:1049-54.
19. Iovino M, Feifel U, Yong CL, Wolters JM, Wallenstein
G. Safety, tolerability and pharmacokinetics of BIBN
4096 BS, the rst selective small molecule calcitonin
gene-related peptide receptor antagonist, following single
intravenous administration in healthy volunteers. Ceph-
alalgia 2004;24:645-56.
20. Edvinsson L, Ekman R, Jansen I, Ottosson A, Uddman R.
Peptide-containing nerve bers in human cerebral arter-
ies: immunocytochemistry, radioimmunoassay, and in
vitro pharmacology. Ann Neurol 1987;21:431-7.
21. Petersen K, Lassen LH, Birk S, Olesen J. The novel
CGRP-antagonist BIBN4096BS does not affect the cere-
bral hemodynamics in healthy volunteers [abstract].
Cephalalgia 2003;23:729.
22. Ashina M, Bendtsen L, Jensen R, Schifter S, Olesen J.
Evidence for increased plasma levels of calcitonin gene-
related peptide in migraine outside of attacks. Pain 2000;
86:133-8.
23. Struthers AD, Brown MJ, Macdonald DW, Beacham JL,
Stevenson JC, Morris HR, et al. Human calcitonin gene
related peptide: a potent endogenous vasodilator in man.
Clin Sci 1986;70:389-93.
24. Howden CW, Logue C, Gavin K, Collie L, Rubin PC.
Haemodynamic effects of intravenous human calcitonin-
gene-related peptide in man. Clin Sci (Lond) 1988;74:
413-8.
25. Kruuse C, Thomsen LL, Jacobsen TB, Olesen J. The
phosphodiesterase 5 inhibitor sildenal has no effect on
cerebral blood ow or blood velocity, but nevertheless
induces headache in healthy subjects. J Cereb Blood
Flow Metab 2002;22:1124-31.
26. Kruuse C, Thomsen LL, Birk S, Olesen J. Migraine can
be induced by sildenal without changes in middle cere-
bral artery diameter. Brain 2003;126:241-7.
27. Thomsen L. Investigations into the role of nitric oxide
and the large intracranial arteries in migraine headache.
Cephalalgia 1997;17:873-95.
28. Edvinsson L, Krause DN. Cerebral blood ow and me-
tabolism. 2nd ed. Philadelphia: Lippincott Williams &
Wilkins; 2002.
29. Vorstrup S. Tomographic cerebral blood ow measure-
ments in patients with ischemic cerebrovascular disease
and evaluation of the vasodilator capacity by the acet-
azolamide test. Acta Neurol Scand Suppl 1988;114:1-48.
30. Vorstrup S, Henriksen L, Paulson OB. Effect of acetazol-
amide on cerebral blood ow and cerebral metabolic rate
for oxygen. J Clin Invest 1984;74:1634-9.
31. Carroll TJ, Teneggi V, Jobin M, Squassante L, Treyer V,
Hany TF, et al. Absolute quantication of cerebral blood
ow with magnetic resonance, reproducibility of the
method, and comparison with H2(15)O positron emission
tomography. J Cereb Blood Flow Metab 2002;22:1149-56.
32. Thomsen LL, Ahlgren M, Iversen HK, Olesen J. Middle
cerebral artery blood velocity increases during experi-
mental pressure pain in the head. In: Olesen J, Moskowitz
MA, editors. Experimental headache models. New York:
Raven Press; 1995. p. 337-9.
33. Baskaya MK, Suzuki Y, Anzai M, Seki Y, Saito K,
Takayasu M, et al. Effects of adrenomedullin, calcitonin
gene-related peptide, and amylin on cerebral circulation
in dogs. J Cereb Blood Flow Metab 1995;15:827-34.
34. Suzuki Y, Satoh S, Ikegaki I, Okada T, Shibuya M,
Sugita K, et al. Effects of neuropeptide Y and calcitonin
gene-related peptide on local cerebral blood ow in rat
striatum. J Cereb Blood Flow Metab 1989;9:268-70.
35. Beattie DT, McNeil DK, Connor HE. The inuence of
neurokinins and calcitonin gene-related peptide on cere-
bral blood ow in anaesthetized guinea-pigs. Neuropep-
tides 1993;24:343-9.
36. Hong KW, Pyo KM, Lee WS, Yu SS, Rhim BY. Phar-
macological evidence that calcitonin gene-related peptide
is implicated in cerebral autoregulation. Am J Physiol
1994;266:H11-6.
37. Macfarlane R, Moskowitz MA, Tasdemiroglu E, Wei EP,
Kontos HA. Postischemic cerebral blood ow and neu-
roeffector mechanisms. Blood Vessels 1991;28:46-51.
38. Wahl M, Schilling L, Parsons AA, Kaumann A. Involve-
ment of calcitonin gene-related peptide (CGRP) and nitric
oxide (NO) in the pial artery dilatation elicited by cortical
spreading depression. Brain Res 1994;637:204-10.
39. Stanley J, Martin J, Richards H, Lovick M, Barron M,
Pickard J. Effects of calcitonin gene related peptide on
cerebral blood ow in healthy volunteers [abstract].
J Cereb Blood Flow Metab 1991;11(Suppl 2):s273.
40. Naylor AR, Miller JD, Edwards CR, Merrick MV, Sellar
RJ, OShaughnessy D, et al. Effect of calcitonin gene-
related peptide on middle cerebral artery velocities [ab-
stract]. J Cereb Blood Flow Metab 1989;9.
41. Juul R, Aakhus S, Bjornstad K, Gisvold SE, Brubakk AO,
Edvinsson L. Calcitonin gene-related peptide (human alpha-
CGRP) counteracts vasoconstriction in human subarach-
noid haemorrhage. Neurosci Lett 1994;170:67-70.
42. Naylor AR, Robertson IJ, Edwards CR, Merrick MV,
Sellar RJ, OShaughnessy D, et al. Cerebral vasospasm
following subarachnoid hemorrhage: effect of calcitonin
gene-related peptide on middle cerebral artery velocities
using transcranial Doppler sonography. Surg Neurol
1991;36:278-80.
43. Johnston FG, Bell BA, Robertson IJ, Miller JD, Haliburn
C, OShaughnessy D, et al. Effect of calcitonin-gene-
related peptide on postoperative neurological decits af-
ter subarachnoid haemorrhage. Lancet 1990;335:869-72.
44. Salmon P, Fitzgerald D, Lambe R, Darragh A,
OShaughnessy D, Riddell A, et al. A single rising intrave-
nous dose tolerance and pharmacodynamic study of human
calcitonin gene related peptide (CGRP) in healthy male
volunteers [abstract]. Clin Pharmacol Ther 1989;45:170.
45. Poyner DR, Sexton PM, Marshall I, Smith DM, Quirion
R, Born W, et al. International Union of Pharmacology.
XXXII. The mammalian calcitonin gene-related pep-
tides, adrenomedullin, amylin, and calcitonin receptors.
Pharmacol Rev 2002;54:233-46.
46. Edvinsson L, Cantera L, Jansen-Olesen I, Uddman R.
Expression of calcitonin gene-related peptide1 receptor
mRNA in human trigeminal ganglia and cerebral arteries.
Neurosci Lett 1997;229:209-11.
47. Doods H, Hallermayer G, Wu D, Entzeroth M, Rudolf K,
Engel W, et al. Pharmacological prole of BIBN
4096BS, the rst selective small molecule CGRP antag-
onist. Br J Pharmacol 2000;129:420-3.
48. Wu D, Eberlein W, Rudolf K, Engel W, Hallermayer G,
Doods H. Characterisation of calcitonin gene-related
peptide receptors in rat atrium and vas deferens: evidence
for a [Cys(Et)(2, 7)]hCGRP-preferring receptor. Eur
J Pharmacol 2000;400:313-9.
49. McLatchie LM, Fraser NJ, Main MJ, Wise A, Brown J,
Thompson N, et al. RAMPs regulate the transport and
ligand specicity of the calcitonin-receptor-like receptor.
Nature 1998;393:333-9.
50. Doods H, Rudolf K, Hallermayer G, Eberlein W. CGRP
antagonistic properties of BIBN 4096BS in rats. Cepha-
lalgia 2000;20:414-20.
51. Edvinsson L, Alm R, Shaw D, Rutledge RZ, Koblan KS,
Longmore J, et al. Effect of the CGRP receptor antago-
nist BIBN4096BS in human cerebral, coronary and
omental arteries and in SK-N-MC cells. Eur J Pharmacol
2002;434:49-53.
CLINICAL TRIALS
Peginterferon alfa-2a does not alter the
pharmacokinetics of methadone in patients
with chronic hepatitis C undergoing
methadone maintenance therapy
Objective: Our objective was to quantify the pharmacokinetics of methadone and the pharmacokinetics and
pharmacodynamics of peginterferon alfa-2a (40 kd) in patients with chronic hepatitis C undergoing meth-
adone maintenance therapy.
Methods: Adults with chronic hepatitis C who had been receiving a consistent methadone maintenance
regimen for at least 3 months were eligible for this open-label, multicenter, nonrandomized drug interaction
study. All patients received 180 g subcutaneous peginterferon alfa-2a once weekly for 4 weeks and contin-
ued their methadone regimen. Serial blood samples were collected at baseline and immediately before and for
up to 168 hours after study drug administration for the purposes of quantifying methadone and peginter-
feron alfa-2a serum concentrations, measuring serum 2,5-oligoadenylate synthetase activity, and determin-
ing hepatitis C virus ribonucleic acid levels.
Results: Twenty-four patients were enrolled. Methadone exposure, as measured by maximum serum concen-
tration (C
max
) and area under the concentration-time curve (AUC) normalized to a 100-mg/d dose, after 4
doses of peginterferon alfa-2a increased by 10% to 15% when compared with baseline. The week 4/baseline
ratio of the mean C
max
was 1.11 (90% confidence interval [CI], 1.02-1.22), and for AUC from time 0 to 24
hours, the week 4/baseline ratio was 1.15 (90% CI, 1.08-1.23). The mean accumulation ratios (week 4/first
dose) for C
max
and AUC from time 0 to 168 hours of peginterferon alfa-2a were 2.1 and 2.3, respectively.
Conclusions: Peginterferon alfa-2a does not appreciably alter the pharmacokinetics of methadone. (Clin
Mark Sulkowski, MD, Teresa Wright, MD, Stephen Rossi, PharmD,
Sanjeev Arora, MD, Matthew Lamb, PharmD, Ka Wang, PhD,
Jean-Michel Gries, PhD, and Sreeni Yalamanchili, PharmD Baltimore, Md, San Francisco and
Thousand Oaks, Calif, Albuquerque, NM, and Nutley, NJ
Injection drug use is the most common mode of
acquisition of new hepatitis C virus (HCV) infections in
the United States.
1,2
Injection drug users (IDUs) are at
high risk of HCV infection because of the efciency
with which the virus is parenterally transmitted. Indeed,
within 1 year of needle use, approximately 77% of
intravenous drug users become infected with HCV
3
; in
those who have used injectable drugs for 10 years or
more, the prevalence of HCV exposure is as high as
94%.
4,5
Comprehensive strategies for the management
From the Viral Hepatitis Center, Johns Hopkins University School
of Medicine, Baltimore; Gastrointestinal Section, Veterans
Affairs Medical Center, University of California, San Fran-
cisco, San Francisco; Science Center, University of New Mex-
ico, Albuquerque; Roche, Nutley; and Amgen, Inc, Thousand
Oaks.
The study was sponsored by Roche, Nutley, NJ.
Received for publication June 9, 2004; accepted Sept 16, 2004.
Reprint requests: Mark Sulkowski, MD, Viral Hepatitis Center, Johns
Hopkins Medical Institutions, 1830 E Monument St, Room 448,
Baltimore, MD 21205.
E-mail: msulkows@jhmi.edu
0009-9236/$30.00
and Therapeutics.
doi:10.1016/j.clpt.2004.09.008
214
of HCV infection must encompass IDUs. Methadone
maintenance programs are effective in diminishing opi-
oid drug use and decreasing other high-risk behaviors;
accordingly, such drug treatment programs may serve
as an important link to provide IDUs with access to
treatment for chronic hepatitis C.
6
The combination of pegylated interferon plus ribavi-
rin, the treatment of choice in patients with chronic
hepatitis C,
7
produces sustained virologic responses in
52% to 63% of patients with HCV monoinfection and
40% of patients with HCV/human immunodeciency
virus coinfection.
8-12
Until recently, methadone treat-
ment has been a strict exclusion criterion in clinical
trials, and as a result, patients with hepatitis C who are
receiving methadone maintenance therapy have been
identied as an understudied population.
13
Thus this
study was designed to address this critical problem of
chronic hepatitis C in methadone recipients.
Because of the magnitude of the epidemic of hepa-
titis C in IDUs, it is important to understand the clinical
implications of the concomitant use of opiates (ie,
methadone) and anti-HCV therapies. Moreover, poten-
tial pharmacokinetic interactions between therapies for
hepatitis C and opiate agonists need to be investigated
to establish the safety of concurrent HCV and metha-
done treatment.
6
Methadone is a racemic mixture of which the
R-enantiomer is the active moiety.
14-17
There is marked
interindividual variability in the distribution and clear-
ance of the drug that is derived from variability in the
individual serum concentrations of
1
-acid glycopro-
tein and cytochrome P450 (CYP) 3A4 activity.
14,16,17
The stereoselective metabolism of methadone is medi-
ated by CYP3A4, CYP2C8, and CYP2D6, and data
obtained in vitro and in vivo demonstrate that there is
signicant potential for drug interactions with inhibi-
tors of these CYP isozymes.
18,19
Conventional interferon alfa has been shown to in-
hibit various CYP isozymes including CYP3A4 and
CYP2D6.
20-23
Moreover, some patients with chronic
hepatitis C have antibodies directed at CYP2D6,
24,25
and a study conducted during the rst month of treat-
ment with interferon alfa revealed an increase in the
activity of CYP3A4 and CYP2D6 in patients with
chronic hepatitis C.
26
It is, therefore, important to as-
sess the potential for clinically signicant CYP-
mediated interactions in drugs that may be used by
patients with chronic hepatitis C who are candidates for
interferon-based therapy.
Methadone has been shown to have anti-interferon
properties in mice.
27
Whether the reduction in serum
levels was related to decreased production or increased
clearance of endogenous interferon is another unre-
solved but relevant question with potential implications
for the treatment of chronic hepatitis C in patients
undergoing methadone maintenance therapy.
The objective of this study was to quantitatively
evaluate the effects of peginterferon alfa-2a on the
pharmacokinetics of methadone in patients with
chronic hepatitis C who were receiving a stable meth-
adone maintenance regimen. The pharmacokinetics and
pharmacodynamics of peginterferon alfa-2a were also
quantied in these individuals.
METHODS
Subjects. Eligible subjects were aged 18 years or
greater; had evidence of HCV infection, dened as a
positive anti-HCV antibody test result and either a
positive radioimmunoblot assay for HCV or quanti-
able HCV ribonucleic acid (RNA) in serum by poly-
merase chain reaction (COBAS AMPLICOR HCV
MONITOR Test, v2.0; Roche Diagnostics, Basel, Swit-
zerland) (limit of detection, 2000 IU/mL); and had
compensated liver disease. Women of childbearing po-
tential were required to have a negative urine preg-
nancy test result documented within 24 hours before
administration of the rst dose of peginterferon alfa-2a.
All participants had to have been receiving a consis-
tent methadone maintenance regimen, in terms of the
formulation, dose, and frequency of administration, for
at least 3 months before study entry. No specic dose of
methadone was selected for the trial. Because alcohol
inuences the disposition of methadone,
28
patients
were not eligible if the dosage form of methadone
included alcohol (ie, use of methadone tincture or
United States Pharmacopeia oral solution containing
8% alcohol was prohibited). Methadone diskettes, tab-
lets, alcohol-free liquid concentrate, and powder were
acceptable dosage forms.
Patients were excluded if they had a positive urine
screening result for drugs, except methadone, or evi-
dence of excessive alcohol or drug use within 6 months
of study entry. Drug use was dened as the use of drugs
without a documented indication, including amphet-
amines, barbiturates, benzodiazepines, cannabis, co-
caine, or opiates other than methadone. A history of
severe psychiatric disease, including depression, was
also grounds for exclusion.
Patients coinfected with hepatitis A or B virus or
human immunodeciency virus were also excluded, as
were patients with evidence of severe or advanced liver
disease of any etiology or other severe or chronic
systemic disease.
2005;77(3):214-24 Peginterferon alfa-2a and methadone PK 215
Prescription medications were allowed for the treat-
ment of concurrent chronic conditions, provided that
the patient had been receiving a consistent regimen for
at least 2 months and adjustments to the regimen were
not anticipated during the study and that the medication
was not a known inhibitor of CYP3A4 (ie, clarithro-
mycin, erythromycin, uconazole, uoxetine, uvox-
amine, itraconazole, ketoconazole, nefazodone, and
omeprazole). Because methadone is a substrate of
CYP3A4, consumption of grapefruit juice was prohib-
ited during the study. In addition to the above-
mentioned medications, the use of theophylline was
prohibited. With the exception of acetaminophen (INN,
paracetamol), which could be used at a maximum dose
of 4 g/d, the use of nonprescription medications was
prohibited for 2 weeks before and during the study.
Female participants were required to use contracep-
tion throughout the study.
Study design. Patients enrolled in this open-label,
multicenter, nonrandomized drug interaction study con-
tinued their ongoing daily methadone maintenance
therapy regimen and received 180 g subcutaneous
peginterferon alfa-2a (40 kd) (Pegasys; Roche, Nutley,
NJ) once weekly for 4 weeks. Patients fasted for 12
hours before peginterferon alfa-2a administration dur-
ing weeks 1 and 4. Caffeinated drinks were not allowed
for 48 hours before or on study drug administration
days. Patients remained in the study clinic for 24 hours
after peginterferon alfa-2a administration for blood
sampling. The date and time of methadone administra-
tion were recorded at baseline, on day 1 (week 1), and
on day 22 (week 4). The date and time of administra-
tion of all peginterferon alfa-2a doses and collection of
all blood samples were recorded.
Serial blood samples of 5 mL were collected at
baseline (day 7) and during week 4 (day 22) at 0 hours
and at 0.5, 1, 2, 3, 4, 6, 8, 12, and 24 hours after
methadone administration for the purpose of quantitat-
ing methadone serum levels.
Serial blood samples of 5 mL were collected during
week 1 (day 1) and week 4 (day 22) before adminis-
tration of peginterferon alfa-2a (0 hours) and at 3, 8, 24,
48, 72, 84, 96, 120, 144, and 168 hours after adminis-
tration for the purpose of quantitating peginterferon
alfa-2a serum levels.
Blood samples (2 mL) were collected before meth-
adone administration at baseline (day 7); before pegin-
terferon alfa-2a administration during weeks 1, 2, 3,
and 4; and before dosing and at 6, 12, 24, 36, 48, 72, 96,
and 168 hours after administration of peginterferon
alfa-2a during week 1 for the purpose of quantifying
serum HCV RNA levels. Serial blood samples were
also collected before dosing and at 3, 8, 24, 48, 72, 84,
96, 120, 144, and 168 hours after administration of
peginterferon alfa-2a at baseline (day 7) and during
weeks 1 and 4 for the purpose of measuring serum
2,5-oligoadenylate synthetase (2,5-OAS) activity, a
widely used marker for interferon-induced antiviral ac-
tivity.
Serum concentrations of methadone were quantied
by a validated gas chromatographymass spectroscopy
assay at PPD Development (Richmond, Va). Samples
were kept frozen until analyzed, at which time they
were thawed at room temperature in a water bath. One
milliliter of sample and 100 L of internal standard (50
ng/mL methadone-d
9
) were placed in a silylated 16
125mm culture tube. After the mixture was allowed to
stand for 20 minutes, 4 mL of water and 2 mL of
100-mmol/L phosphate buffer/methanol (80:20) were
added; the tube was blended in a vortex mixer and then
centrifuged at 3000 rpm for 5 minutes. The supernatant
was decanted into a second silylated 16 125mm
culture tube. The sample was then placed on a solid
phase extraction cartridge (Bond Elut Certify SPE col-
umns, 130 mg, LRC [large reservoir cartridge], Varian
No. 1211-3050; Varian, Inc, Palo Alto, Calif ). The
cartridge was then washed with 2 mL of water, 2 mL of
100-mmol/L acetic acid, and then 3 mL 100% metha-
nol. The columns were dried under vacuum for 5 min-
utes before being eluted into silylated, conical centri-
fuge tubes with 3 mL of dichloromethane/isopropyl
alcohol/ammonium hydroxide (78:20:2). The eluate
was then evaporated to dryness at 40C or greater under
a gentle stream of nitrogen. Samples were reconstituted
with 25 L of 10% N-methyl-N-trimethylsilyltriu-
oroacetamide in toluene. Reconstituted samples were
stored at 20C until analysis. A 1.0-L volume was
then injected onto a DB-1 30-m 0.32-mm inner
diameter column (J&W Scientic, Folsom, Calif)
mounted in a Varian 3400 gas chromatograph (Varian,
Inc) coupled with a mass spectrometer. The carrier gas
was helium, the injector temperature was 300C, and
the instrument was programmed to increase the column
temperature from 165C to 300C at a rate of 20C/
min. The analytic range of the assay was 0.1 to 50
ng/mL. Samples with concentrations greater than 50
ng/mL were diluted and reanalyzed. The coefcient of
variation for the assay ranged from 2.5% to 35.1%.
Serum concentrations of peginterferon alfa-2a were
quantied by a validated enzyme-linked immunosor-
bent assay at MDS Pharma Services (Montreal, Que-
bec, Canada). In a 1-step immunoreaction, the pegin-
terferon alfa-2a (40 kd) contained in the sample is
bound by the peroxidase-conjugated antibody. The as-
216 Sulkowski et al MARCH 2005
say is a quantitative sandwich enzyme immunoassay
which uses 2 monoclonal antibodies that recognize
different epitopes. It is not known whether the antibod-
ies bind to receptor-bound peginterferon alfa-2a (40
kd). One hundred microliters of substrate buffer and 50
L of assay dilution buffer were added to all wells in
microtiter plates. A 50-L aliquot of the standard or
study sample was then added, followed by 50 L of
peroxidase-conjugated solution. This complex binds
via the captureantipeginterferon alfa-2a (40 kd) an-
tibody to the multiprotein layercoated surface of the
microtiter plate. The plate is sealed and incubated at
room temperature for 30 minutes with shaking at 150
rpm before incubation for 16 to 24 hours at 37C
without shaking. After incubation, the plate was shaken
at room temperature at 150 rpm for 60 minutes. The
plate was then washed 5 times with 300 L of wash
buffer. Next, 200 L of substrate solution (0.8 mL of
0.2% tetramethylbenzidine in acetone/ethanol/hydro-
gen peroxide [10:89:1]/22 mL substrate buffer, pH 4.1)
was added to each well. This solution is converted by
the peroxidase to a colored product that can be deter-
mined photometrically. The plates were sealed and
incubated in the dark at room temperature for approx-
imately 15 to 20 minutes with shaking at 150 rpm. Fifty
microliters of 2N sulfuric acid was then added to stop
the reaction, and the plate was read within 15 minutes
by use of a SpectraMax 340 PC microplate reader
(Molecular Devices Corporation, Sunnyvale, Calif) set
to 450 to 650 nm. The concentration of the samples was
calculated by comparison with a standard curve. The
analytic range of the assay was 350 to 3000 pg/mL.
Samples with concentrations greater than 3000 pg/mL
were diluted and reanalyzed. The coefcient of varia-
tion for the assay ranged from 9.1% to 13.5%.
HCV RNA levels in serum were determined with the
COBAS AMPLICOR HCV MONITOR Test, v2.0
(Roche Diagnostics, Basel, Switzerland). Serum 2,5-
OAS activity was determined with a commercially
available radioimmunoassay kit.
After completion of the study, patients were allowed
to participate in an open-label continuation study with
peginterferon alfa-2a and ribavirin combination ther-
apy. Patients who declined combination therapy were
followed up for 4 weeks after administration of the last
dose of peginterferon alfa-2a.
The institutional review boards at each center ap-
proved the protocol. All patients provided informed
written consent before enrollment. The study was con-
ducted in conformance with the principles of the Dec-
laration of Helsinki, the principles outlined in the
Guideline for Good Clinical Practice, and the laws and
regulations of the United States.
Pharmacokinetic parameters. The following phar-
macokinetic parameters were determined for peginter-
feron alfa-2a and methadone by use of noncompart-
mental methods: observed maximum serum drug
concentration (C
max
), time to C
max
(t
max
), time of last
observed serum concentration (t
last
), area under the
serum concentrationtime curve (AUC) (AUC between
time 0 and the time of the last observed serum concen-
tration [AUC
0-last
], AUC from time 0 extrapolated to
innity [AUC
0-
], and AUC from time 0 to 24 hours
[AUC
0-24
] for methadone and AUC from time 0 to 168
hours [AUC
0-168
] for peginterferon alfa-2a). The total
apparent clearance (CL
ss
/F) and apparent terminal
phase volume of distribution (V
z
/F) were calculated for
methadone, and the accumulation ratio for week 4:1 for
C
max
and AUC
0-168
was calculated for peginterferon
alfa-2a.
Pharmacokinetic characteristics of peginterferon
alfa-2a and methadone were evaluated by model-
independent analysis by use of WinNonlin Pro (version
4.0; Pharsight Corporation, Mountain View, Calif). Ac-
tual sampling times rather than scheduled sampling
times were used in all analyses. For individual pegin-
terferon alfa-2a proles with measurable baseline con-
centrations, all data for the subject were excluded if the
baseline concentration exceeded 1000 pg/mL. If the
baseline value was less than 1000 pg/mL, subsequent
values were adjusted for the baseline concentration by
subtracting the baseline value from the measured value.
After adjustment, if the value was below the limit of
quantitation, it was excluded from the analysis.
The slope of the terminal log-linear portion of each
peginterferon alfa-2a and methadone concentration-
time prole was determined by least squares linear
regression. The log-linear trapezoidal method with ex-
trapolation beyond the last experimental concentration
value was used to determine AUC
0-
.
The total apparent clearance of methadone (CL
ss
/F)
was calculated as Dose/AUC
0-24
; the apparent volume
of distribution of methadone (V
z
/F) was estimated as
Dose/(Terminal slope AUC
0-
).
Dose-normalized methadone pharmacokinetic pa-
rameters were used to compare the C
max
and AUC of
methadone at baseline and during week 4; the param-
eters were normalized to a 100-mg dose.
Pharmacodynamic parameters. 2,5-OAS is an im-
portant effector of interferon-induced antiviral effects,
and the serum activity of this enzyme is widely used as
an indicator of interferon activity. Calculations of the
following parameters were based on week-1 measure-
ments of 2,5-OAS data: observed maximum activity
(OAS
max
), time to OAS
max
(t
max
), observed last activity
(OAS
last
), time to OAS
last
(t
last
), and area beneath the
effect curve at the last time point (ABEC
last
).
Serum 2,5-OAS activitytime proles were ana-
lyzed with noncompartmental methods by use of Win-
Nonlin Pro (version 4.0). The log-linear trapezoidal
method was used to determine ABEC
last
. For individual
proles with measurable baseline 2,5-OAS activity,
subsequent values were adjusted for the baseline con-
centration by subtracting the baseline value from the
measured value. After adjustment, if the value was
below the limit of quantitation, it was excluded from
the analysis.
Statistical analysis. No formal sample size calcula-
tion was made. A planned enrollment of 24 patients
was established, with the expectation that 18 individu-
als would complete the study.
A 90% condence interval (CI) for the true ratio of
week 4 to baseline for the C
max
and AUC
0-24
of meth-
adone and a 90% CI for the true difference between
means for the t
max
of methadone were calculated. The
residual mean square error from an ANOVA was used
as the estimate of variance for the CI.
RESULTS
A total of 24 patients were enrolled in the study, 22
of whom received 4 doses of peginterferon alfa-2a and
completed the pharmacokinetic portion of the study.
Data from 2 patients who received only 3 doses of
peginterferon alfa-2a were excluded from the pharma-
cokinetic analyses. In addition, pharmacokinetic data
for peginterferon alfa-2a from 2 further patients were
excluded because of analytic artifacts. All patients were
included in the safety analysis. The baseline character-
istics of the patients are presented in Table I. The mean
methadone dose at baseline was 88 mg/d.
Pharmacokinetic results
Inuence of peginterferon alfa-2a on pharmacoki-
netics of methadone. Methadone exposure after 4
doses of peginterferon alfa-2a increased by 10% to 15%
when compared with baseline (Table II). However, the
90% CIs for the ratio of the mean C
max
and AUC
0-24
at
baseline and during week 4 were 1.02 to 1.22 and 1.08
to 1.23, respectively, which fell within the generally
accepted equivalence interval of 0.80 to 1.25, indicat-
ing no signicant effect of peginterferon alfa-2a on the
pharmacokinetics of methadone (Table II).
Paired values for C
max
and AUC
0-24
for methadone
are presented for each patient in Fig 1, A and B,
respectively. Intersubject variabilities at baseline and
during week 4 were 36% and 34%, respectively, for
C
max
and 28% and 33%, respectively, for AUC
0-24
. Of
the patients, 8 of 22 (36%) and 16 of 21 (76%) had
higher individual C
max
and AUC
0-24
values, respec-
tively, during week 4 than at baseline.
The mean 24-hour serum concentration prole for
methadone (normalized to a 100-mg dose) at baseline
and during week 4 is presented in Fig 2. Methadone
Fig 1. Individual dose-normalized maximum serum concen-
tration (C
max
) (A) and area under concentration-time curve
from time 0 to 24 hours (AUC
0-24
) (B) for methadone at
baseline and during week 4.
Table I. Baseline characteristics of 24 patients
enrolled in study
Characteristic No.
Male/female 15/9
Race (white/black/other) 13/6/5
Age (y) (median and range) 49.5 (33-61)
Weight (kg) (median and range) 78.5 (52-129)
Height (cm) (median and range) 170 (137-185)
Body mass index (kg/m
2
) (median
and range)
26.9 (19.1-48.9)
Methadone dose (mg/d) (median
and range)
95 (30-150)
concentrations were higher at each time point during
the 24-hour sampling period.
All patients continued to receive stable doses of
methadone throughout the study.
Inuence of methadone on pharmacokinetics of
peginterferon alfa-2a. Three patients had detectable
predose concentrations of peginterferon alfa-2a, and
subsequent determinations were adjusted before analy-
sis. The intersubject variability of C
max
was 38% and
that of AUC
0-168
was 39% after the rst dose of pegin-
terferon alfa-2a, and it was 42% for both parameters
during week 4.
Single- and multiple-dose pharmacokinetic data for
peginterferon alfa-2a in patients receiving methadone
maintenance therapy are presented in Table III. The
mean accumulation ratios (week 4/rst dose) for C
max
and AUC
0-168
were 2.1 (90% CI, 1.62-2.58) and 2.3
(90% CI, 1.82-2.78), respectively. Values for peginter-
feron alfa-2a during week 1 and week 4 reveal a general
increase during the study (Fig 3).
Pharmacodynamic results
Mean 2,5-OAS activity at baseline and after the
rst (week 1) and last doses of peginterferon alfa-2a
(week 4) are presented in Fig 4. A summary of 2,5-
OAS activity parameters is presented in Table IV. The
induction of 2,5-OAS activity in methadone recipi-
ents was similar to that previously reported in subjects
not receiving methadone maintenance therapy.
Of 24 patients, 12 (50%) had either a greater than
2-log
10
decrease in serum HCV RNA levels or unde-
tectable serum HCV RNA levels after 4 weeks of
treatment with peginterferon alfa-2a. Of these 12
Table II. Methadone noncompartmental pharmacokinetic parameters
Parameter Baseline (n 22) Week 4 (n 22)
t
max
(h) 2.9 1.7 (2.2-3.6) 2.6 1.0 (2.2-3.0)
C
max
(ng/mL)* 701 253 (595-807) 774 266 (663-885)
AUC
last
(ng h/mL)* 10,671 2974 (9428-11,914) 12,511 4300 (10,714-14,308)
AUC
0-24
(ng h/mL)* 10,576 3012 (9288-11,864) 12,576 4153 (10,841-14,311)
CL
ss
/F (L/h) 11 5 (8.9-12.2) 9 4 (7.3-10.7)
V
z
/F (L) 634 541 (397-871) 616 574 (376-856)
Geometric least squares means
C
max
(ng/mL) 656.7 731.3
Week 4/baseline (mean and 90% CI) 1.11 (1.02-1.22)
AUC
0-24
(ng h/mL) 10,063.9 11,602.3
Week 4/baseline (mean and 90% CI) 1.15 (1.08-1.23)
Values are given as mean SD with 95% CI, unless otherwise indicated.
t
max
, Time to maximum serum concentration; C
max
, maximum serum concentration; AUC
last
, AUC between time 0 and the time of the last observed serum
concentration; AUC
0-24
, area under concentration-time curve from time 0 to 24 hours; CL
ss
/F, total apparent clearance; V
z
/F, apparent volume of distribution; CI,
condence interval.
*Individual parameter estimates were normalized to a 100-mg dose.
n 21.
n 20.
Table III. Single- and multiple-dose peginterferon alfa-2a noncompartmental pharmacokinetic parameters in
patients receiving stable methadone maintenance therapy
Parameter First dose (n 22) Week 4 (n 20)
t
max
(h) 84 34 (70-98) 78 24 (67-89)
C
max
(ng/mL) 14.4 5.4 (12.1-16.7) 26.0 10.9 (21.2-30.8)
t
last
(h) 167 5 (164.9-169.1) 168 0.5 (167.8-168.2)
C
last
(ng/mL) 9.9 4.5 (8.0-11.8) 15.9 6.3 (13.1-18.7)
AUC
last
(ng h/mL) 1769 669 (1489-2049) 3459 1451 (2823-4095)
AUC
0-168
(ng h/mL) 1788 700 (1495-2081) 3515 1463 (2874-4156)
Accumulation ratio: C
max
NA 2.1 1.3
Accumulation ratio: AUC
0-168
NA 2.3 1.3
Values are given as mean SD with 95% CI.
t
last
, Time of last observed serum concentration; C
last
, serum concentration from time 0 to t
last
; AUC
0-168
, area under concentration-time curve from time 0 to 168 hours;
NA, not applicable.
week-4 responders, 8 had HCV RNA levels below the
limit of quantitation of the assay by week 4.
Safety
A total of 123 adverse events were reported by 20
patients (83%). The most common adverse events were
those typically associated with conventional interferon
or pegylated interferon administration (ie, ulike symp-
toms). Of the patients, 4 (17%) had a neutrophil count
of less than 0.75 10
9
/L at some point during the trial
but no patient had a neutrophil count of less than 0.50
10
9
/L. Of the patients, 2 (8%) had a platelet count of
Fig 2. Mean dose-normalized methadone serum concentration (Conc) (SD) at baseline and during
week 4 after 4 180-g doses of peginterferon alfa-2a.
Fig 3. Mean serum concentration of peginterferon alfa-2a (SD) during the rst and last (week 4)
dose intervals.
less than 50 10
9
/L at some point during the trial but
no patient had a platelet count of less than 20 10
9
/L.
There were no serious adverse events, and no patients
withdrew prematurely because of adverse events or
laboratory abnormalities. In addition, no subject had
clinical signs or symptoms of intoxication or with-
drawal related to methadone.
DISCUSSION
The results of our study demonstrate that peginter-
feron alfa-2a does not inuence the pharmacokinetics
of methadone to a clinically signicant extent in pa-
tients receiving ongoing methadone maintenance ther-
apy. Baseline pharmacokinetic and pharmacodynamic
data for peginterferon alfa-2a in the absence of meth-
adone administration were not available for the patients
enrolled in our study; however, the pharmacokinetic
prole of peginterferon alfa-2a in these individuals was
generally similar to that observed previously in healthy
volunteers and patients with chronic hepatitis C.
29-32
The pharmacokinetic values for racemic methadone
reported in our study (normalized to 100 mg) compare
well with those reported previously in opiate users
(normalized to 10 mg
14
or 70 mg
16
). It must be noted
that we did not evaluate the pharmacokinetic properties
of the active R-isomer of methadone, so it is not pos-
sible to comment on the impact of peginterferon alfa-2a
on the pharmacokinetics of this entity. We also did not
monitor
1
-acid glycoprotein levels, which have an
inuence on the plasma pharmacokinetics of the drug.
15
As anticipated, the mean serum concentrations of
peginterferon alfa-2a were higher during week 4 than
during week 1. This agent has a long elimination half-
life and, as a result, does not reach steady-state levels
until 6 to 8 weeks after the initiation of weekly admin-
istration. The accumulation ratios (steady-state concen-
Table IV. Summary of baseline-adjusted
2,5-oligoadenylate synthetase activity at week 1 (N
24)
Parameter Mean
t
max
(h) 98 49 (78-118)
2,5-OAS
max
(nmol/L h
1
) 2.9 1.5 (2.3-3.5)
t
last
(h) 162 13 (157-167)
2,5-OAS
last
(nmol/L h
1
) 2.1 1.2 (1.6-2.6)
ABEC
last
(nmol/L) 253 139 (197-309)
Values are given as mean SD with 95% CI.
2,5-OAS
max
, Observed maximum 2,5-oligoadenylate synthetase activity;
2,5-OAS
last
, observed last 2,5-oligoadenylate synthetase activity; ABEC
last
,
area beneath effect curve at last time point.
Fig 4. Mean serum 2,5-oligoadenylate synthetase (OAS) activity-time proles (SD) at baseline,
during week 1, and during week 4.
tration/maximum concentration after 1 dose) reported
previously have been approximately 2 to 3.
29,32
Thus
our nding of an accumulation ratio of approximately 2
after 4 weeks of weekly dosing is in line with our
expectations.
The biologic response to peginterferon alfa-2a, as
assessed by 2 5-OAS activity, was similar in patients
receiving ongoing methadone maintenance therapy in
our study and in healthy subjects enrolled in previous
studies.
31,33
The decrease in HCV RNA levels during
the rst 4 weeks of treatment with peginterferon alfa-2a
was typical of that seen in chronic hepatitis C patients
who are not receiving methadone maintenance therapy,
and one half of the patients had a substantial reduction
in HCV RNA after only 4 doses of peginterferon alfa-
2a.
34
Peginterferon alfa-2a monotherapy was well toler-
ated during methadone maintenance therapy in patients
with chronic hepatitis C, and no serious adverse events
were observed. The spectrum and frequency of adverse
events were similar to those reported in patients with
chronic hepatitis C who were not receiving methadone
in a phase III trial,
35
although it must be acknowledged
that the patients in our trial were monitored for only 4
weeks.
There are several limitations to our study. As noted,
we evaluated patients during an abbreviated course of
treatment with peginterferon alfa-2a monotherapy
rather than for the full 48-week duration recommended
for the treatment of chronic hepatitis C at the time the
study was planned. However, the 4-week treatment
period was sufcient to conrm whether peginterferon
alfa-2a interfered with the pharmacokinetics of metha-
done and to determine whether the pharmacodynamics
of peginterferon alfa-2a was impaired by methadone.
Thus, although the most important therapeutic end
point in patients with chronic hepatitis C is the eradi-
cation of HCV infection, our study was not designed
with sufcient statistical power to examine sustained
virologic response rates. In addition, it must be noted
that therapy for chronic hepatitis C has continued to
evolve since our study was conceived, and the current
treatment of choice is the combination of pegylated
interferon plus ribavirin.
7
Thus it remains to be deter-
mined whether ribavirin has any inuence on the phar-
macokinetics of methadone.
These limitations notwithstanding, our data provide
important information to clinicians treating the rela-
tively large population of HCV-infected persons receiv-
ing methadone maintenance therapy worldwide. De-
spite the size of the population, patients with chronic
hepatitis C who are receiving methadone maintenance
therapy have been understudied, a problem that was
recognized by the US National Institutes of Health
Consensus Panel in their 2002 recommendations for the
management of chronic hepatitis C.
7
However, there is
increasing interest in the treatment of hepatitis C in
methadone recipients, and recent studies have sug-
gested that sustained viral response rates in patients
receiving methadone maintenance therapy are similar
to those observed in other HCV-infected patient popu-
lations. For example, among patients included in a
large, retrospective, matched cohort study, Van Thiel et
al
36
reported that sustained virologic response rates
were similar in patients receiving methadone mainte-
nance therapy and in patients without a history of
intravenous drug use after treatment with daily conven-
tional interferon (33% and 37%, respectively). Interest-
ingly, the researchers observed that the methadone dose
was increased in some patients during treatment with
conventional interferon alfa. However, consistent with
our ndings, the authors reported that these increases in
methadone dose generally represented the use of meth-
adone to treat side effects of conventional interferon
rather than a response to clinical evidence of a phar-
macokinetic interaction. Similarly, in an interim anal-
ysis of 50 methadone maintenance patients treated with
conventional interferon plus ribavirin, Sylvestre
37
re-
ported that the end-of-treatment response rate was sim-
ilar to that of patients without a history of injection drug
use. It must be noted that we did not observe a similar
increase in methadone doses during our study. A
smaller, prospective, open-label German study has also
reported success in treating intravenous drug users en-
rolled in a detoxication program with conventional
interferon-based therapy.
38
Although preliminary, these
encouraging results suggest that patients with a recent
history of injection drug use, particularly those under-
going stable methadone maintenance therapy, can be
successfully treated for hepatitis C with interferon-
based regimens.
In conclusion, our data indicate that peginterferon
alfa-2a does not appreciably alter the pharmacokinetics
of methadone. Thus concurrent treatment with metha-
done is not a contraindication to therapy with peginter-
feron alfa-2a for chronic hepatitis C. Although the
decision to treat a patient who is using illicit drugs
should be made by the patient after consulting with his
or her physician, this decision need not be inuenced
by the use of methadone maintenance therapy, and a
priori adjustment of either the methadone or the pegin-
terferon alfa-2a dose is not recommended in patients
receiving stable methadone maintenance therapy who
undergo treatment for chronic hepatitis C with pegin-
terferon alfa-2acontaining regimens.
Drs Sulkowski and Arora have received research grants from
Roche. Dr Wright has no potential conicts of interest. Dr Yalaman-
chili is a former employee of Roche. Drs Rossi, Lamb, Wang, and
Gries are currently employed by Roche.
References
1. Recommendations for prevention and control of hepatitis
C virus (HCV) infection and HCV-related chronic dis-
ease. Centers for Disease Control and Prevention.
MMWR Recomm Rep 1998;47:1-39.
2. Alter MJ. Prevention of spread of hepatitis C. Hepatology
2002;36:S93-8.
3. Garfein RS, Vlahov D, Galai N, Doherty MC, Nelson
KE. Viral infections in short-term injection drug users:
the prevalence of the hepatitis C, hepatitis B, human
immunodeciency, and human T-lymphotropic viruses.
Am J Public Health 1996;86:655-61.
4. Thomas DL. Correlates of hepatitis C virus infections
among injection drug users. Medicine (Baltimore) 1995;
74:212-20.
5. Novick DM. The impact of hepatitis C virus infection on
methadone maintenance treatment. Mt Sinai J Med 2000;
67:437-43.
6. Edlin BR. Prevention and treatment of hepatitis C in
injection drug users. Hepatology 2002;36:S210-9.
7. National Institutes of Health Consensus Development
Conference Statement: Management of hepatitis C:
2002June 10-12, 2002. Hepatology 2002;36:S3-20.
8. Manns MP, McHutchison JG, Gordon SC, Rustgi VK,
Shiffman M, Reindollar R, et al. Peginterferon alfa-2b
plus ribavirin compared with interferon alfa-2b plus riba-
virin for initial treatment of chronic hepatitis C: a ran-
domised trial. Lancet 2001;358:958-65.
9. Fried MW, Shiffman ML, Reddy KR, Smith C, Marinos
G, Goncales FL Jr, et al. Peginterferon alfa-2a plus
ribavirin for chronic hepatitis C virus infection. N Engl
J Med 2002;347:975-82.
10. Hadziyannis SJ, Sette H Jr, Morgan TR, Balan V, Diago
M, Marcellin P, et al. Peginterferon alfa-2a and ribavirin
combination therapy in chronic hepatitis C: randomized
study of treatment duration and ribavirin dose. Ann In-
tern Med 2004;140:346-55.
11. Zeuzem S, Diago M, Gane E, Reddy KR, Pockros P, Prati
D, et al. Peginterferon alfa-2a (40 kilodaltons) and ribavirin
in patients with chronic hepatitis C and normal aminotrans-
ferase levels. Gastroenterology 2004;127:1724-32.
12. Torriani FJ, Rodriguez-Torres M, Rockstroh JK, Lissen
E, Gonzalez-Garcia J, Lazzarin A, et al. Peginterferon
alfa-2a plus ribavirin for chronic hepatitis C virus infec-
tion in HIV-infected patients. N Engl J Med 2004;351:
438-50.
13. Strader DB. Understudied populations with hepatitis C.
Hepatology 2002;36:S226-36.
14. Wolff K, Rostami-Hodjegan A, Shires S, Hay AW, Feely
M, Calvert R, et al. The pharmacokinetics of methadone
in healthy subjects and opiate users. Br J Clin Pharmacol
1997;44:325-34.
15. Foster DJ, Somogyi AA, Dyer KR, White JM, Bochner
F. Steady-state pharmacokinetics of (R)- and (S)-
methadone in methadone maintenance patients. Br J Clin
Pharmacol 2000;50:427-40.
16. Foster DJ, Somogyi AA, White JM, Bochner F. Popula-
tion pharmacokinetics of (R)-, (S)- and rac-methadone in
methadone maintenance patients. Br J Clin Pharmacol
2004;57:742-55.
17. Boulton DW, Arnaud P, DeVane CL. Pharmacokinetics
and pharmacodynamics of methadone enantiomers after a
single oral dose of racemate. Clin Pharmacol Ther 2001;
70:48-57.
18. Wang JS, DeVane CL. Involvement of CYP3A4, CYP2C8,
and CYP2D6 in the metabolism of (R)- and (S)-methadone
in vitro. Drug Metab Dispos 2003;31:742-7.
19. Begre S, von Bardeleben U, Ladewig D, Jaquet-Rochat
S, Cosendai-Savary L, Golay KP, et al. Paroxetine in-
creases steady-state concentrations of (R)-methadone in
CYP2D6 extensive but not poor metabolizers. J Clin
Psychopharmacol 2002;22:211-5.
20. Israel BC, Blouin RA, McIntyre W, Shedlofsky SI. Ef-
fects of interferon-alpha monotherapy on hepatic drug
metabolism in cancer patients. Br J Clin Pharmacol 1993;
36:229-35.
21. Horsmans Y, Brenard R, Geubel AP. Short report:
interferon-alpha decreases 14C-aminopyrine breath test
values in patients with chronic hepatitis C. Aliment Phar-
macol Ther 1994;8:353-5.
22. Pageaux GP, le Bricquir Y, Berthou F, Bressot N, Picot
MC, Blanc F, et al. Effects of interferon-alpha on cyto-
chrome P-450 isoforms 1A2 and 3A activities in patients
with chronic hepatitis C. Eur J Gastroenterol Hepatol
1998;10:491-5.
23. Islam M, Frye RF, Richards TJ, Sbeitan I, Donnelly SS,
Glue P, et al. Differential effect of IFNalpha-2b on the
cytochrome P450 enzyme system: a potential basis of
IFN toxicity and its modulation by other drugs. Clin
Cancer Res 2002;8:2480-7.
24. Duclos-Vallee JC, Nishioka M, Hosomi N, Arima K,
Leclercq A, Bach JF, et al. Interferon therapy in LKM-1
positive patients with chronic hepatitis C: follow-up by a
quantitative radioligand assay for CYP2D6 antibody de-
tection. J Hepatol 1998;28:965-70.
25. Dalekos GN, Wedemeyer H, Obermayer-Straub P, Kay-
ser A, Barut A, Frank H, et al. Epitope mapping of
cytochrome P4502D6 autoantigen in patients with
chronic hepatitis C during alpha-interferon treatment.
J Hepatol 1999;30:366-75.
26. Becquemont L, Chazouilleres O, Serfaty L, Poirier JM,
Broly F, Jaillon P, et al. Effect of interferon alpha-
ribavirin bitherapy on cytochrome P450 1A2 and 2D6
and N-acetyltransferase-2 activities in patients with
chronic active hepatitis C. Clin Pharmacol Ther
2002;71:488-95.
27. Geber WF, Lefkowitz SS, Hung CY. Effect of morphine,
hydromorphone, methadone, mescaline, trypan blue, vi-
tamin A, sodium salicylate, and caffeine on the serum
interferon level in response to viral infection. Arch Int
Pharmacodyn Ther 1975;214:322-7.
28. Quinn DI, Wodak A, Day RO. Pharmacokinetic and
pharmacodynamic principles of illicit drug use and treat-
ment of illicit drug users. Clin Pharmacokinet 1997;33:
344-400.
29. Modi MW, Fried M, Reindollar RW, Rustgi VR, Kenny
R, Wright TL, et al. The pharmacokinetic behavior of
pegylated (40KDA) interferon alfa-2a (Pegasys) in
chronic hepatitis C patients after multiple dosing [ab-
stract]. Hepatology 2000;32:394A.
30. Martin NE, Modi MW, Reddy R. Characterization of
pegylated (40KDA) interferon alfa-2a (Pegasys) in the
elderly [abstract]. Hepatology 2000;32:348A.
31. Xu Z-X, Hoffman J, Patel I, Joubert P. Single-dose
safety/tolerability and pharmacokinetic/pharmacodynam-
ics (PK/PD) following administration of ascending sub-
cutaneous doses of pegylated-interferon (PEG-IFN) and
interferon alfa-2a (IFN alfa-2a) to healthy subjects [ab-
stract]. Hepatology 1998;28:702A.
32. Perry CM, Jarvis B. Peginterferon-alpha-2a (40 kD): a
review of its use in the management of chronic hepatitis
C. Drugs 2001;61:2263-88.
33. Yan S. Clinical study reportProtocol NP16569: A
comparability study of peginterferon (40kDa) alfa-2a of
clinical trial material and the commercial product admin-
istered subcutaneously via needle injection in healthy
subjects. Nutley (NJ): Roche; 2002 Mar 29. Research
report 1007463.
34. Zeuzem S, Herrmann E, Lee JH, Fricke J, Neumann AU,
Modi M, et al. Viral kinetics in patients with chronic
hepatitis C treated with standard or peginterferon
alpha2a. Gastroenterology 2001;120:1438-47.
35. Zeuzem S, Feinman V, Rasenack J, Heathcote EJ, Lai
MY, Gane E, et al. Peginterferon alfa-2a in patients with
chronic hepatitis C. N Engl J Med 2000;343:1666-72.
36. Van Thiel DH, Anantharaju A, Creech S. Response to treat-
ment of hepatitis C in individuals with a recent history of
intravenous drug abuse. Am J Gastroenterol 2003;98:2281-8.
37. Sylvestre DL. Treating hepatitis C in methadone main-
tenance patients: an interim analysis. Drug Alcohol De-
pend 2002;67:117-23.
38. Backmund M, Meyer K, Von Zielonka M, Eichenlaub D.
Treatment of hepatitis C infection in injection drug users.
Hepatology 2001;34:188-93.
LETTERS TO THE EDITOR
Disposition of cisapride appears to be inuenced
by P-glycoprotein in the mouse
To the Editor:
Recently, Lowry et al
1
evaluated cisapride as a potential
model substrate to assess cytochrome P450 (CYP) 3A4 ac-
tivity in vivo. As part of their evaluation, they conducted
in vitro studies to determine whether cisapride is a substrate
of P-glycoprotein (P-gp). These studies were conducted with
the use of LLC-PK1 cells and the derivative cell line
L-MDR1. Results from these studies suggest that cisapride is
not a substrate of P-gp. However, cisapride markedly in-
creased the accumulation of vinblastine in L-MDR1 cells,
with an apparent inhibition constant of approximately 16.1
mol/L. Cisapride (20 mol/L) also inhibited vinblastine
transport from the basal to the apical side by approximately
30% from baseline. These in vitro data suggest that cisapride
inhibits P-gp.
We evaluated the disposition of cisapride (Janssen Phar-
maceutica NV, Beerse, Belgium) in mdr1a/1b (/) mice
after oral administration of a 5-mg/kg dose and determined
the plasma and brain (whole-brain homogenates) concen-
trations of cisapride at 0.5, 1, 2, and 5 hours. Previous
studies have shown that the bioavailability and the brain
penetration of drugs that are substrates of P-gp are en-
hanced in mice lacking P-gp expression,
2
because P-gp is
highly expressed in the capillary endothelial cells of the
blood-brain barrier and the epithelial cells of the intes-
tine.
3,4
The area under the concentration versus time curve
from 0 to 5 hours [AUC(0-5)] of cisapride in the brains of
mdr1a/1b (/) was 5-fold higher in the mdr1a/1b (/)
mice compared with the mdr1a/1b (/) mice (316
ng h/mL versus 62.9 ng h/mL) (Fig 1). In contrast, the
AUC(0-5) of cisapride in the plasma of mdr1a/1b (/)
was 1.3-fold higher in the mdr1a/1b (/) mice compared
with the mdr1a/1b (/) mice (121 ng h/mL versus
90.2 ng h/mL) (Fig 1). With normalization for plasma
concentrations of cisapride, the brain-to-plasma area under
the concentration versus time curve (AUC) ratio of cisa-
pride in mdr1a/1b (/) mice was 2.6 compared with 0.7
in mdr1a/1b (/) mice; a similar observation was made
with maximum plasma concentration (C
max
) (Fig 2).
Our ndings suggest that, in vivo, cisapride appears to be
a moderate substrate of P-gp and that the brain penetration of
cisapride is enhanced in mice lacking P-gp expression. How-
ever, P-gp expression in the intestine appears to have minimal
inuence on the disposition of cisapride, possibly as a result
of both the inhibition and the saturation of P-gp at the level of
the intestine by cisapride. Although cisapride is no longer
marketed in the United States because of cardiac safety con-
cerns (arrhythmias),
5
it continues to be used as an orally
administered prokinetic agent for the treatment of gastro-
esophageal reux disease in many European and Asian coun-
tries. These ndings suggest not only that CYP3A4-based
interactions should be considered when drugs are coadminis-
tered with cisapride but also that caution should be exercised
when drugs that may be substrates or inhibitors of P-gp are
coadministered.
Edna F. Choo, PhD
Karthik Venkatakrishnan, PhD
Fig 1. Plasma (squares) and brain (triangles) levels of cisa-
pride in mdr1a/1b (/) (solid symbols) and mdr1a/1b
(/) (open symbols) mice. Data represent mean (SD)
values of 4 or 5 determinations at each time point.
Fig 2. Mean brain-to-plasma area under the concentration
versus time curve (AUC) and maximum plasma concentration
(C
max
) (SD) ratios in mdr1a/1b (/) (solid bars) and
mdr1a/1b (/) (open bars) mice.
CLINICAL PHARMACOLOGY & THERAPEUTICS MARCH 2005 225
Heather L. Hatch, MA
Sandhya Rahematpura, BSc
Pharmacokinetics, Dynamics and Metabolism
Pzer Inc
Groton, Conn
No conict of interest has been identied.
References
1. Lowry JA, Kearns GL, Abdel-Rahman SM, Nafziger AN, Khan
IS, Kashuba ADM, et al. Cisapride: a potential model substrate to
assess cytochrome P4503A4 activity in vivo. Clin Pharmacol Ther
2003;73:209-22.
2. Kim RB, Fromm MF, Wandel C, Leake B, Wood AJJ, Roden DM,
et al. The drug transporter P-glycoprotein limits oral absorption
and brain entry of HIV-1 protease inhibitors. J Clin Invest 1998;
101:289-94.
3. Gottesman MM, Pastan I. Biochemistry of multidrug resistance
mediated by the multidrug transporter. Annu Rev Biochem 1993;
62:385-427.
4. Cordon-Cardo C, OBrien JP, Casals D, Rittman-Grauer L, Bied-
ler JL, Melamed MR, et al. Multidrug-resistance gene (P-
glycoprotein) is expressed by endothelial cells at blood-brain
barrier sites. Proc Natl Acad Sci U S A 1989;86:695-8.
5. Michalets EL, Williams CR. Drug interactions with cisapride:
clinical implications. Clin Pharmacokinet 2000;39:49-75.
doi:10.1016/j.clpt.2004.10.009
Modulation of celecoxib pharmacokinetics by food
in pediatric patients
To the Editor:
In the November 2002 issue of the Journal, we presented
preliminary pharmacokinetic results of a clinical trial evalu-
ating the use of celecoxib in conjunction with metronomic
dosing of chemotherapy for antiangiogenic treatment in chil-
dren with recurrent solid tumors.
1
The purpose of this report
is 2-fold: (1) to present new pharmacokinetic data for cele-
coxib in a larger group of children taking celecoxib at a dose
of 250 mg/m
2
twice daily and (2) to present the effect of
high-fat food on celecoxib kinetics in a subset of children.
Our data suggest a trend toward increased absorption and
increased overall systemic exposure when celecoxib is taken
with a high-fat meal or snack. In addition, it appears that food
high in fat tends to modulate trough plasma levels such that
patients spend a smaller proportion of the dosing interval
below the targeted antiangiogenic concentration of 500 g/L
(Dr Jaime L. Masferrer, oral communication, 2002).
Data presented in our original report and data that have been
accrued since that publication indicate that the trough concen-
trations of celecoxib in our patients were below the desired
concentration that possesses antiangiogenic properties (500
g/L) (Dr Jaime L. Masferrer, oral communication, 2002). For
safety and regulatory reasons, we opted to try to increase the
systemic exposure to celecoxib by coadministering the drug with
a high-fat meal or snack rather than increasing the dose. Early
clinical studies investigating the pharmacokinetics of celecoxib
in healthy adults showed that the maximum plasma concentra-
tion (C
max
) and area under the plasma concentrationtime curve
(AUC) were approximately dose-proportional across the dose
range of 100 to 200 mg. However, at higher doses (400 mg,
which is approximately equivalent to our study dose of 250
mg/m
2
), there appears to be a less than proportional increase in
C
max
and AUC. This is thought to be a result of the low
solubility of celecoxib in aqueous media.
2
Therefore it was
recommended that higher doses of celecoxib be coadministered
with food to try to improve absorption.
3
Our study was amended
Table I. Single-dose and steady-state pharmacokinetics of celecoxib administered with and without food in
pediatric cancer patients
No.
Dose
(mg)
t
max
(h)
C
max
(g/L)
C
min
(g/L)
AUC*
(g/L h)
V
d
/F
(L/kg)
Cl/F
(L h
1
kg
1
) t
(h)
Predicted
time of
500 g/
L (h)
Time
spent at
500 g/
L (h)
Single dose
Without food 20 350
(median)
3
(median)
1257 869 213 114 7,810 3,484 8.2 6.7 1.4 0.9 4.0 1.5
With food 7 300
(median)
3
(median)
2284 998 297 155 12,498 3,871 4.0 1.5 0.8 0.2 3.6 1.4
% Change 82% 39% 60% 51% 43% 10%
P value .016 .138 .006 .115 .033 .599
Steady state
Without food 16 300
(median)
3
(median)
1440 756 296 162 8,886 4,521 6.8 4.9 0.9 0.4 5.0 2.4 8.1 3.6 4.1 3.2
With food 7 300
(median)
3
(median)
2867 1581 394 271 15,574 6,774 2.9 2.1 0.7 0.2 3.4 0.9 10.5 3.0 2.2 1.4
% Change 99% 33% 75% 57% 22% 32% 30% 46%
P value .007 .289 .011 .063 .103 .105 .147 .154
Data are expressed as mean SD. Comparisons within the single-dose and steady-state data were made by use of the Student t test with the signicance level set
at P .05.
t
max
, Time to maximum plasma concentration; C
max
, maximum plasma concentration; C
min
, trough concentration; AUC, area under plasma concentrationtime curve;
t
, elimination half-life; Vd/F, apparent volume of distribution; Cl/F, apparent clearance.

*AUC data are for the interval of 0 hours to innity for single-dose data and 0 to 12 hours for steady-state data.
226 Letters to the Editor MARCH 2005
to include fatty food coadministration. Examples of appropriate
foods included peanut butter on toast or a buttered mufn (Phar-
macia, oral communication, 2002). In the children in our study,
high-fat food has resulted in signicant differences in the C
max
and AUC values both after a single dose and at steady state
(Table I and Fig 1). We were particularly interested in the trough
concentration at steady state because we were aiming to main-
tain plasma concentrations above the antiangiogenic concentra-
tion of 500 g/L (Dr Jaime L. Masferrer, oral communication,
2002). Although the difference observed for the trough concen-
tration (C
min
) was not statistically different, the trend we were
expecting has been demonstrated and may reach signicance
with an appropriate sample size. Without increasing the cele-
coxib dose, this seems to be a reasonable way to increase the
plasma concentrations of celecoxib in an effort to achieve anti-
angiogenic therapeutic levels (Table I and Fig 1).
An additional valuable piece of information that arose
from this study concerns the safety of long-term use of
celecoxib in children at a twice-daily dose of 250 mg/m
2
. The
median duration of treatment was 10 weeks (range, 3 weeks
to 18 months). In fact, 1 patient was treated for 16 months and
another for 18 months, and none showed any signs of central
nervous system or cardiovascular toxicity.
There are several limitations of this study resulting from
factors that were beyond our control given the vulnerability of
the children involved. First of all, our study did not have a
blinded, randomized crossover design. The data we present
are opportunistic in nature and were collected as part of a
phase I pilot study whose main objective was to determine the
safety of a particular drug combination. In addition, we rec-
ognize the unbalanced sample sizes of our 2 groups (with
versus without high-fat food), but this is a result of the
original study design. Future studies of celecoxib at this dose
should include the recommendation for high-fat food coad-
ministration. Finally, although high-fat food intake was not
recommended for the original group of children studied,
fasting was not recommended either. However, experience
with pediatric cancer patients eligible for phase I studies
suggests that their eating habits generally include light meals,
and because the pharmacokinetic studies were started early in
the morning, in most cases the children had not yet eaten
breakfast. For the patients who did take celecoxib with food,
the high-fat meal or snack was not standardized because,
again, this was beyond the scope of this study. Instead, we
suggested several examples of foods that were considered to
be high in fat, and the patients were asked to keep a food diary
for the pharmacokinetic study days. Despite these limitations,
the evidence for increased systemic exposure to celecoxib
with high-fat food in children is quite convincing (Table I).
Diana Stempak, MSc
New Agents and Innovative Therapy Program
Divisions of Hematology/Oncology and Clinical
Pharmacology and Toxicology
Hospital for Sick Children
Toronto, Ontario, Canada
Fig 1. Plasma concentrationtime curve (mean SD) for 1 dose interval after single dose and at
steady state in pediatric cancer patients who were administered 250 mg/m
2
of celecoxib orally with
and without food.
2005;77(3):225-31 Letters to the Editor 227
Janet Gammon, BScN, RN
Division of Hematology/Oncology
Jacqueline Halton, MD
Childrens Hospital of Eastern Ontario
Ottawa, Ontario, Canada
Martin Champagne, MD
Hpital Sainte-Justine
Montreal, Quebec, Canada
Gideon Koren, MD
Division of Clinical Pharmacology and Toxicology
Sylvain Baruchel, MD
Division of Hematology/Oncology
E-mail: sylvain.baruchel@sickkids.ca
This work was supported by a research grant from Phar-
macia/Pzer Canada. D.S. is a recipient of an Ontario Grad-
uate Scholarship and a Hospital for Sick Children Research
Training Centre award. G.K. is a senior scientist of the
Canadian Institutes for Health Research and the Ivey Chair
holder in Molecular Toxicology at the University of Western
Ontario.
The authors have no conicts of interest to disclose.
References
1. Stempak D, Gammon J, Klein J, Koren G, Baruchel S. Single-dose
and steady-state pharmacokinetics of celecoxib in children. Clin
2. Celecoxib (SC-58635): a cyclooxygenase-2-specic inhibitor of
inammatory prostaglandin production [investigational brochure].
6th ed. Skokie (IL): Searle; 1999.
3. Celebrex (celecoxib capsules) [prescribing information]. Skokie
(IL): Searle, Pzer; 2001.
doi:10.1016/j.clpt.2004.10.015
Limited association of the 2988G>A single
nucleotide polymorphism with CYP2D6*41 in
black subjects
To the Editor:
In the August 2004 issue of the Journal, Raimundo et al
1
described a novel intronic single nucleotide polymorphism
(SNP), 2988GA, that is highly predictive for intermediate
metabolizer (IM) status as assessed with the probe drug
sparteine. In their white population, 2988GA was present in
52 of 56 CYP2D6*41 alleles. Because all 2988A/*null sub-
jects were IMs and this SNP was not found on any other
allele, they recommended identifying CYP2D6*41 as
[1584C, 2988A] and using 2988A as a single assay marker.
Currently, CYP2D6*2 and *41 are dened by identical coding
regions but discriminated by 1584CG (*41 C and *2
G).
Investigating phenotype-genotypediscordant cases, we
have independently discovered this SNP in North American
white subjects. Genotyping of 203 subjects revealed that all
31 CYP2D6*41 alleles carried 2988A (frequency 0.076).
2,3
Within this cohort, the only CYP2D6*41x2 allele observed
was negative for the 2988GA SNP (conrmed by pedigree
analysis), similar to that for subject Tu 140 described by
Raimundo et al
1
(Fig 1). Whereas the 2988GA genotyping
results are in agreement with their ndings, the functional
consequences with the use of dextromethorphan as a pheno-
typing probe are less conclusive. Of 7 subjects in our data set
with a *41[2988A]/*null genotype, 3 were extensive metabo-
lizers (EMs) (Fig 2). Hence, although 2988A appears to be
associated with an IM sparteine phenotype, this does not
necessarily seem to be the case for assessments with dextro-
methorphan.
In contrast to the strong linkage of 2988A with
CYP2D6*41 in white subjects, only 10 of 48 CYP2D6*41
alleles among 251 black subjects were positive for this SNP
(phenotype available for 8) (Fig 2).
3
Allele frequencies were
0.076 and 0.021 for CYP2D6*41[2988G] and *41[2988A],
respectively. As a sole marker, 2988GA would correctly
exclude a poor metabolizer (PM) phenotype but would not
detect about 80% of CYP2D6*41 alleles in this population
sample.
Among other explanations, Raimundo et al
1
speculated
that 2988GA may cause low expression as a result of
erroneously spliced messenger ribonucleic acid. Information
theorybased splice site analysis, however, did not reveal any
cryptic splice sites that may be activated by 2988A but did
reveal the potential loss of an SF2/ASF splice-enhancer bind-
ing site.
4
Furthermore, resequencing 9 kilobases of a
CYP2D6*41[2988A] allele did not reveal any other unique
SNP that may explain lower activity when compared with
CYP2D6*1 (AY545216) and CYP2D6*2 (NG_003180) ref-
erence sequences.
In summary, cytochrome P450 (CYP) 2D6 phenotype pre-
diction from genotype data remains a challenging task, espe-
cially in nonwhite and ethnically diverse or admixed popula-
tions. Our previously reported approach using 1584CG as
a PM exclusion criterion
5
may be compromised by the oc-
currence, albeit infrequently, of nonfunctional 1584G-
positive alleles (1 CYP2D6*4 and *8, respectively, reported
by Raimundo et al
1
) and may rather be used as a screening
tool than a denite assessment. Using 2988GA as a marker
may be helpful to identify a small proportion of African
Americans and other subjects of black African descent as
non-PMs, but it is not as powerful a predictor in these
populations as it may be in white subjects. Furthermore, our
genotype-to-phenotype correlation data imply that the pres-
ence of 2988A does not necessarily predict intermediate
metabolism for all CYP2D6 substrates alike. To further assess
the usefulness and associations of these SNPs, 1584CG
and 2988GA, both are included in ongoing investigations.
Fig 1. Simultaneous polymerase chain reaction (PCR)restriction fragment length polymorphism
(RFLP)based genotype analysis of 2850CT and 2988GA. A 225base pair (bp) fragment was
generated from a 6.6-kilobase CYP2D6-specic amplicon
5
with primers 5-TGC TCT CGG CCC
TGC TC and 5-TTC ATG GGC CCC CGC Cga aAC CCT T (partial XmnI restriction site [bold
type]). PCR fragments were cut with FspI (2850C) and XmnI (2988G). Depending on the absence
or presence of 2850CT and 2988GA, the following patterns were generated: *1, 62 142
21 bp; *2 and *41[2988G], 204 21 bp; and *41[2988A], 225 bp. Presence of 2850CT and
2988GA and genotypes are as indicated for each subject. F, Father; M, mother; D, daughter
(segregation of CYP2D6*41x2[2988G] in a 2-generation family). *41/*41 with diamond symbol
denotes a black subject.
Fig 2. Correlation between CYP2D6*41[2988A] alleles and CYP2D6 activity. Metabolic urinary
metabolic ratios of dextromethorphan (DM) to dextrorphan (DX) lower than 0.3, greater than 0.3,
and greater than 0.03 indicate poor metabolizer (PM), intermediate metabolizer (IM), and extensive
metabolizer (EM) phenotypes, respectively; *null, *red, and *fct denote nonfunctional, reduced, and
fully functional alleles, respectively. The sequenced CYP2D6*5/*41[2988A] subject is indicated by
asterisk within triangle. The number of subjects within each genotype group is as shown.
Andrea Gaedigk
Liliane Ndjountch
J. Steven Leeder
Division of Clinical Pharmacology and
Experimental Therapeutics
Childrens Mercy Hospital and Clinics
Kansas City, Mo
L. DiAnne Bradford
Morehouse School of Medicine
Atlanta, Ga
E-mail: agaedigk@cmh.edu
The authors do not have any conicts of interest
References
1. Raimundo S, Toscano C, Klein K, Fischer J, Griese E-U, Eichel-
baum M, et al. A novel intronic mutation, 2988GA, with high
predictivity for impaired function of cytochrome P450 2D6 in
white subjects. Clin Pharmacol Ther 2004;76:128-38.
2. Gaedigk A, Gotschall RR, Forbes NS, Simon SD, Leeder JS.
Optimization of cytochrome P4502D6 (CYP2D6) phenotype as-
signment using a genotyping algorithm based on allele frequency
data. Pharmacogenetics 1999;9:669-82.
3. Gaedigk A, Bradford LD, Marcucci KA, Leeder JS. Unique
CYP2D6 activity distribution and genotype-phenotype discor-
dance in African Americans. Clin Pharmacol Ther 2002;72:76-89.
4. Nalla R, Rogan PK. Automated splice site analyses. Available
from: URL: https://splice.cmh.edu. Accessed Aug 25, 2004.
5. Gaedigk A, Ryder DL, Bradford LD, Leeder JS. CYP2D6 poor
metabolizer status can be ruled out by a single genotyping assay
testing for the 1584G promoter polymorphism. Clin Chem
2003;49:1008-11.
doi:10.1016/j.clpt.2004.10.014
Reply
To the Editor:
The data obtained by Gaedigk et al in white subjects
entirely conrm our conclusion that 2988GA is a specic
marker for CYP2D6*41 and that no other specic marker for
this allele can be found within coding, intronic, and extended
upstream regions of the gene. The authors question, however,
the usefulness of genotyping for *41 to predict the interme-
diate metabolizer (IM) phenotype for substrates other than
sparteine and in individuals with ethnic origin other than
Caucasian.
It should be remembered that the IM phenotype was clearly
dened only for sparteine oxidation in white subjects, in
which it constitutes a separate subgroup within a trimodal
distribution comprising about 10% to 15% of the population.
More overlapping phenotypic modi, especially with respect to
the IM and extensive metabolizer phenotypes, have generally
been observed with other model substrates including dextro-
methorphan, resulting in part from their lower specicity for
cytochrome P450 (CYP) 2D6. It should also be kept in mind
that phenotyping is not as reliable an indicator for metabolic
activity as is often assumed, as a result of environmental
interactions and analytic or compliance problems.
1
Most im-
portantly, however, our major interest is not to genetically
predict the metabolism of CYP2D6 probe drugs with higher
and higher precision but rather to establish the basis for
meaningful application of genotyping in the clinical setting.
Identication of CYP2D6*41 as the second most frequent
functionally relevant allele in white subjects was an important
advance in this direction. Previously, carriers of 1 functional
and 1 nonfunctional allele (classical heterozygotes) have been
regarded as IMs. This misconception has now been corrected
by showing that the distinction between fully functional al-
leles (*1, *2, *35) and functionally decient alleles (*9, *10,
*41) dramatically improves genotype-phenotype relation-
ships. There is growing evidence to support our expectation
that this will be of clinical relevance. In addition to the
long-term metoprolol study
2
referenced in our article,
3
Stei-
mer et al
4
recently showed that amitriptyline and nortriptyline
plasma levels are signicantly higher in genotype-predicted
IMs (carriers of only 1 functional allele [*41 or *10]) but not
in carriers of 1 normal allele (*1, *2, *35), in striking agree-
ment with our ndings. Furthermore, preliminary results ob-
tained by genotyping individuals undergoing long-term clo-
mipramine therapy revealed that, in both poor metabolizers
and IMs, equally elevated plasma levels developed over time,
whereas carriers of 1 or 2 fully functional alleles (*1, *2, or
*35) did not differ from each other.
5
Other clinical studies
showing an impact of the IM status include studies of dihy-
drocodeine,
6
propafenone,
7
perhexiline,
8
and desipramine.
9
Of course, prospective clinical studies should now focus not
only on pharmacokinetics but also on outcome and on several
candidate genes rather than only on metabolic enzymes.
Careful reading of our article would reveal that we are very
well aware of the more limited role *41 may play in Asian or
African populations, in which the major low-activity alleles
have been identied as *10 and *17, respectively.
3
Neverthe-
less, data obtained in Ethiopians also revealed lower activity
of *41 (-related) alleles, which were present at even higher
frequencies than in white subjects.
10
However, in contrast to
our comprehensive analysis of *41, no complete sequence
analysis has yet been carried out in nonwhite subjects. With-
out this information and a careful phenotype-genotype corre-
lation analysis, genotyping for 1584CG or 2988GA in
nonwhite individuals may well lead to confusing results as
presented by Gaedigk et al.
Ulrich M. Zanger
Matthias Schwab
Claudia Toscano
Michel Eichelbaum
Dr Margarete Fischer-Bosch Institute of
Clinical Pharmacology
Stuttgart, Germany
Sebastian Raimundo
Institute of Cell Biology
University of Tbingen
Tbingen, Germany
E-mail: uli.zanger@ikp-stuttgart.de
This work was supported by the Robert Bosch Foundation,
Stuttgart, Germany.
The authors declare that they have no conicts of interest.
References
1. Chou WH, Yan FX, Robbins-Weilert DK, Ryder TB, Liu WW,
Perbost C, et al. Comparison of two CYP2D6 genotyping meth-
ods and assessment of genotype-phenotype relationships. Clin
Chem 2003;49:542-51.
2. Rau T, Heide R, Bergmann K, Wuttke H, Werner U, Feifel N, et al.
Effect of the CYP2D6 genotype on metoprolol metabolism persists
during long-term treatment. Pharmacogenetics 2002;12:465-72.
3. Raimundo S, Toscano C, Klein K, Fischer J, Griese E-U, Eich-
elbaum M, et al. A novel intronic mutation, 2988GA, with high
predictivity for impaired function of cytochrome P450 2D6 in
white subjects. Clin Pharmacol Ther 2004;76:128-38.
4. Steimer W, Zopf K, von Amelunxen S, Pfeiffer H, Bachofer J, Popp
J, et al. Allele-specic change of concentration and functional gene
dose for the prediction of steady-state serum concentrations of
amitriptyline and nortriptyline in CYP2C19 and CYP2D6 extensive
and intermediate metabolizers. Clin Chem 2004;50:1623-33.
5. Zanger UM, Furuno T, Schaeffeler E, Klein K, Schwab M, Gram
LF, et al. Poor metabolism of clomipramine in depressed patients
with CYP2D6 intermediate metabolizer genotype: impact of
long-term treatment [abstract]. Clin Exp Pharmacol Physiol
2004;31(Suppl 1):A134.
6. Platten HP, Schweizer E, Dilger K, Mikus G, Klotz U. Pharma-
cokinetics and the pharmacodynamic action of midazolam in
young and elderly patients undergoing tooth extraction. Clin
7. Cai WM, Zhang YD, Chen B, Cai MH, Luo JP, Ling SS.
Simultaneous modeling of pharmacokinetics and pharmacody-
namics of propafenone in healthy subjects. Acta Pharmacol Sin
2001;22:956-60.
8. Barclay ML, Sawyers SM, Begg EJ, Zhang M, Roberts RL,
Kennedy MA, et al. Correlation of CYP2D6 genotype with
perhexiline phenotypic metabolizer status. Pharmacogenetics
2003;13:627-32.
9. Furman KD, Grimm DR, Mueller T, Holley-Shanks RR, Bertz
RJ, Williams LA, et al. Impact of CYP2D6 intermediate metabo-
lizer alleles on single-dose desipramine pharmacokinetics. Phar-
macogenetics 2004;14:279-84.
10. Aklillu E, Herrlin K, Gustafsson LL, Bertilsson L, Ingelman-
Sundberg M. Evidence for environmental inuence on CYP2D6-
catalysed debrisoquine hydroxylation as demonstrated by pheno-
typing and genotyping of Ethiopians living in Ethiopia or in
Sweden. Pharmacogenetics 2002;12:375-83.
doi:10.1016/j.clpt.2004.10.014
Notes of the American Society for Clinical
Pharmacology and Therapeutics
A MESSAGE FROM THE ASCPT MEMBERSHIP COMMITTEE
CHAIRPERSON
On behalf of the Membership Committee, I am
pleased to welcome the new members accepted into the
Society in November 2004. Please take a moment to
review their names and welcome those in your geo-
graphic area. Thank you also to the current ASCPT
members who have sponsored these applicants.
Please continue your recruitment efforts on behalf of
ASCPT and take an active role in ensuring new mem-
bers can both contribute to and benet from the Soci-
ety. Again, welcome to our new members!
John T. Sullivan, MD
Chairperson, Membership Committee
WELCOME NEW ASCPT MEMBERS
FULL MEMBERS
Nancy J. Braden, MD
Sunny Chapel, PhD
Pzer, Inc
Julia Finkel, MD
Childrens National Medical Center
Michael Garvin, PharmD
Pharmaceutical Research and Manufacturers of America
(PhRMA)
Igor Grachev, MD, PhD
Schering-Plough Research Institute
Kathryn Kersey, MSc
Ronald Li, PhD
Sano Aventis
Judith Racoosin, MD, MPH
Center for Drug Evaluation and Research / FDA
Asia Rehman, MBBS, MSc
Universiti Sains Malaysia
Ricky Reinhardt, MD
Novo Nordisk Pharma
Nathan Teuscher, PhD
NPS Pharmaceuticals
ASSOCIATE NOT IN TRAINING MEMBERS
Daniel Selness, RN, MBA
Takeda
Andrea Yoder Graber, BA
University of Chicago
ASSOCIATE IN TRAINING MEMBERS
Chaline Brown, PharmD
University of California, San Francisco
Ying-Jun Cao, MD
The Johns Hopkins Hospital
Robert Freimuth, PhD
Washington University
Yan Gong, PhD
University of Florida
Elizabeth Shang, PhD
University of Buffalo-Pzer
Nathalie Zgheib, MD
University of Pittsburgh
STUDENT MEMBERS
Mindy Cohen
University of Colorado School of Medicine
Tanya Gore, PharmD
Childrens Mercy Hospitals and Clinics
Pejvak Soltany, PhD
Louisiana State University
FULL MEMBER UPGRADES
Ademola Abiose, MD
Creighton University Medical Center
232 MARCH 2005 CLINICAL PHARMACOLOGY & THERAPEUTICS
Michael Blake, MD, PhD
Apurva Desai, MD
The University of Chicago
Mehul Desai, MD
FDA
Tomoe Fujita, MD, PhD
Kitasato University School of Medicine
Aliceson King, MD, MPH
PAREXEL
Peter Kuebler, PharmD
Genetech, Inc
Mindy Magee, PharmD
GlaxoSmithKline
Donald Mager, PharmD, PhD
University at Buffalo, SUNY
Brian Overholser, PharmD
Purdue University
Robin Pearce, PhD
Kyung-Soo Kim, MD
Catholic University Medical Center
2005;77(3):232-3 ASCPT New Members 233
INFORMATION FOR READERS
COMMUNICATIONS
Send manuscripts and all other editorial communications to C.
Michael Stein, MD, Editor, Clinical Pharmacology & Therapeutics,
Division of Clinical Pharmacology, 560 RRB, Vanderbilt Univer-
sity School of Medicine, Nashville, TN 37232-6602.
Information for authors appears in each issue. Authors should
consult these instructions before submitting manuscripts to this
Journal.
Customer Service (orders, claims, online, change of address):
Elsevier Periodicals Customer Service, 6277 Sea Harbor Drive,
Orlando, FL 32887-4800. Tel: (800) 654-2452 (U.S. and Canada);
(407) 345-4000 (outside U.S. and Canada). Fax: (800) 225-6030
(U.S. and Canada): (407) 363-9661 (outside U.S. and Canada).
E-mail: elspcs@elsevier.com.
Address changes must be submitted four weeks in advance.
Yearly Subscription Rates: United States and possessions: Indi-
vidual $216; Institution $456; Student/Resident: $108. All other
countries (prices include airspeed delivery): Individual $258; Insti-
tution $500; Student/Resident $129. Single Issues $42. To receive
student/resident rate, orders must be accompanied by name of
afliated institution, date of term and the signature of program/
residency coordinator on institution letterhead. Orders will be billed
at the individual rate until proof of status is received. Current prices
are in effect for back volumes and back issues.
Further information on this journal is available from the Publisher or
from this journals website (http://www.mosby.com/cpt). Informa-
tion on other Elsevier products is available through Elseviers web-
site (http://www.elsevier.com).
Advertising information. Advertising orders and inquiries can be
sent to: USA, Canada, and South America, Michael Targowski,
Elsevier, 360 Park Ave South, New York, NY 10010; phone (212)
633-3693. Classied advertising orders and inquiries can be sent to
Simone Imbert, Elsevier, 360 Park Ave South, New York, NY
10010; phone (212) 462-1908. Europe and the rest of the world,
Julie Toop; phone 44 (0) 1865 843016; fax 44 (0) 1865 843976;
e-mail media@elsevier.com.
Author inquiries
For inquiries relating to the submission of articles (including elec-
tronic submission where available), please visit Elseviers Author
Gateway at http://authors.elsevier.com. The Author Gateway also
provides the facility to track accepted articles and set up e-mail
alerts to inform you of when an articles status has changed, as well
as detailed artwork guidelines, copyright information, frequently
asked questions, and more. Please see Information for Authors for
individual journal. Contact details for questions arising after accep-
tance of an article, especially those relating to proofs, are provided
after registration of an article for publication.
The paper used in this publication meets the requirements of

ANSI/NISO Z39.48-1992 (Permanence of Paper).
Reprints. For queries about author offprints, e-mail
authorsupport@elsevier.com. To order 100 or more reprints for
educational, commercial, or promotional use, contact the Commer-
cial Reprints Department, Elsevier Inc., 360 Park Avenue South,
New York, NY 10010-1710. Fax: (212) 462-1935; e-mail
reprints@elsevier.com. Reprints of single articles available online
may be obtained by purchasing Pay-Per-View access for $30 per
article on the journal website www.mosby.com/cpt.
Available electronically: on-line from Ovid Technologies, Inc, 333
7th Ave, New York, NY 10001; telephone 800-950-2035.
Microform edition available from ProQuest Information and
Learning, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346.
Copyright 2005 by the American Society for Clinical Phar-
macology and Therapeutics. This Journal and the individual con-
tributions contained in it are protected under copyright by the
American Society for Clinical Pharmacology and Therapeutics, and
the following terms and conditions apply to their use:
Photocopying. Single photocopies of single articles may be made for
personal use as allowed by national copyright laws. Permission of the
Publisher and payment of a fee is required for all other photocopying,
including multiple or systematic copying, copying for advertising or
promotional purposes, resale, and all forms of document delivery.
Special rates are available for educational institutions that wish to make
photocopies for non-prot educational classroom use.
Permissions may be sought directly from Elseviers Rights Depart-
ment in Philadelphia, PA, USA: phone (1) 215 238 7869, fax (1)
215 238 2239, e-mail healthpermissions@elsevier.com. Requests
may also be completed online via the Elsevier homepage (http://
www.elsevier.com/locate/permissions).
In the USA, users may clear permissions and make payments through
the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, USA; phone: (1) (978) 7508400, fax: (1) (978)
7504744, and in the UK through the Copyright Licensing Agency
Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, Lon-
don W1P 0LP, UK; phone: (44) 20 7631 5555; fax: (44) 20 7631
5500. Other countries may have a local reprographic rights agency for
payments.
Derivative Works. Subscribers may reproduce tables of contents or
prepare lists of articles including abstracts for internal circulation
within their institutions. Permission of the Publisher is required for
resale or distribution outside the institution.
Permission of the Publisher is required for all other derivative
works, including compilations and translations.
Electronic Storage or Usage. Permission of the Publisher is re-
quired to store or use electronically any material contained in this
journal, including any article or part of an article.
Except as outlined above, no part of this publication may be repro-
duced, stored in a retrieval system or transmitted in any form or by
any means, electronic, mechanical, photocopying, recording or
otherwise, without prior written permission of the Publisher.
Address permissions requests to: Elsevier Rights Department, at the
fax and e-mail addresses noted above.
Notice. No responsibility is assumed by the Publisher for any injury
and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods,
products, instructions or ideas contained in the material herein. Because
of rapid advances in the medical sciences, in particular, independent
verication of diagnoses and drug dosages should be made.
Although all advertising material is expected to conform to ethical
(medical) standards, inclusion in this publication does not constitute
a guarantee or endorsement of the quality or value of such product
or of the claims made of it by its manufacturer.
CLINICAL
PHARMACOLOGY
T
HERAPEUTICS &

77 (3) Bagus

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

77 (3) Bagus

Enviado por

Direitos autorais:

Formatos disponíveis

Clinical Pharmacology & Therapeutics

Copyright 2005 American Society for Clinical Pharmacology and Therapeutics

QVIST, MD, PHD Professor of Clinical Pharma-

polymorphisms with decompensated heart failure.

channel gene KCNA5

channel polymorphisms 139

, voltage at which channels were 50% activated;

channel polymorphisms 141

) was obtained by tting the Boltzmann

), as compared with the wild type (P

channel, KCNQ1, to drug block

channel polymorphisms 143

Urinary excretion of nicotine or cotinine

(min) 132.1 35.9 128.7 34.2 0.98 (0.78-1.23)

(min) 1057 269 1149 468 1.04 (0.86-1.26)

) was calculated as 0.693/k

of fexofenadine. The 300-mL volume of grapefruit

(h) 3.0 0.5 3.2 0.2 3.1 0.2 3.5 0.2

of fexofenadine. These data comprise the primary

By use of these parameters, we computed the values

, elimination half-life; Vd/F, apparent volume of distribution; Cl/F, apparent clearance.

The paper used in this publication meets the requirements of

Você também pode gostar