Acute Fatty Liver of Pregnancy: An Experience in the

Diagnosis and Management of Fourteen Cases

Usta, Ihab M.; Barton, John R.; Amon, Erol A.; Gonzalez, Anthony; Sibai,
Baha M.

Abstract
OBJECTIVE: Our purpose was to investigate the diagnostic problems
and maternal-perinatal outcome in cases of acute fatty liver of
pregnancy.
STUDY DESIGN: Fourteen cases with acute fatty liver of pregnancy
managed during the past 8-years were studied with emphasis on
presenting symptoms, admitting diagnosis, laboratory findings, clinical
course, maternal complications, and neonatal outcome.
RESULTS:

The mean gestational age at onset was 34.5 weeks (range 28

to 39). Only seven patients had acute fatty liver of pregnancy as a
definite or suspected diagnosis on admission. Computed tomography of
the liver was performed on 10 patients, with only two positive results.
There were no maternal deaths; however, maternal morbidity was
frequent: four patients had hepatic encephalopathy, three pulmonary
edema, three ascites, four respiratory arrest, two diabetes insipidus,
and 10 had transfusion of blood or blood products to correct either
disseminated intravascular coagulation or excessive bleeding.
Coagulation abnormalities were common: hypofibrinogenemia (<300
mg/dl) in 13 patients (93%), prolonged prothrombin time in 12 (86%),
and prolonged partial thromboplastin time in 11 (79%). The corrected
perinatal mortality was 6.6%.
CONCLUSION: Acute fatty liver of pregnancy should be suspected in all
patients with symptoms of preeclampsia in the presence of
hypoglycemia, low fibrinogen, and prolonged prothrombin time,
particularly in the absence of severe abruptio placentae. Computed
tomography of the liver has a high false-negative rate in patients with
acute fatty liver of pregnancy. In spite of the literature's dismal
prognosis, our findings indicate that maternal and perinatal outcomes
appear favorable in well-managed patients. (AM J OBSTET GYNECOL
1994;171:1342-7.)
Key words: Acute fatty liver, pregnancy outcome, laboratory findings,
computed tomography

Acute fatty liver of pregnancy is a rare but potentially fatal complication

of the third trimester of pregnancy. Although described in the literature
for over a century, it was not recognized as a distinct clinical syndrome
until Sheehan's [1] report in 1940 of a series of six cases. Acute fatty
liver of pregnancy is characterized by jaundice, coagulopathy, central
nervous system disturbance, and microvesicular fatty infiltration of the
liver on liver biopsy specimens. As recently as 1980 fetal and maternal
mortality rates as high as 85% were reported [2]. With increased
awareness especially in less severe cases, the use of ultrasonography
and computed tomography (CT) of the liver for the diagnosis of this
condition, and more aggressive management the mortality from acute
fatty liver of pregnancy has been significantly reduced in the past
decade.
The objective of this study was to report our experience with the
diagnostic problems, the use of CT and ultrasonography of the liver for
diagnosis, and the maternal-perinatal outcome in 14 cases of acute
fatty liver of pregnancy managed during the past 8 years.

Material and methods

Medical records of patients with the discharge diagnosis of acute fatty
liver of pregnancy managed between January 1985 and September
1993 were reviewed. Thirteen patients were managed by one or more
of the authors. Records were reviewed for presenting symptoms,
admitting diagnosis, laboratory findings, maternal complications, clinical
course, and neonatal outcome. Management of these patients included
delivery; serial assessments of hematologic, renal, and hepatic
functions; continuous electrocardiographic monitoring; and careful
attention to fluid intake and output. Anemia and coagulopathy were
treated with transfusions. Ascites was treated with sodium restriction
and diuretics. Hepatic encephalopathy was managed with a highcarbohydrate, low-protein diet and oral lactulose. Disseminated
intravascular coagulation was defined as thrombocytopenia (<100 x
103/mm3), prolonged prothrombin time, prolonged partial
thromboplastin time, hypofibrinogenemia (<300 mg/dl), abnormal
peripheral blood smear, and elevated fibrin split products (>40
micrograms/dl). Preeclampsia was diagnosed as the development of
hypertension (diastolic blood pressure of >=90 mm Hg or a systolic
blood pressure >=140 mm Hg) with proteinuria or edema. The
diagnosis of hemolysis, elevated liver enzymes, and low platelets
(HELLP syndrome) was based on the presence of laboratory
abnormalities as described by Sibai [3].

Results
During the study period 13 patients (14 cases) had acute fatty liver of
pregnancy as the discharge diagnosis. One patient had acute fatty liver
of pregnancy documented in two consecutive pregnancies; details of
her recurrence have been previously reported [4]. Ten (77%) patients
were white and three were black. The mean +/-SD maternal age at
diagnosis was 27.9 +/-7.6 years (range 17 to 37). Eight patients (57%)
were nulliparous. Three patients had chronic hypertension, one had
previous preeclampsia, four had previous history of blood transfusions,
and three were remote from hepatitis A (these last seven patients had
liver biopsy --confirmed acute fatty liver of pregnancy). There was no
history of tetracycline use or acetaminophen overdose in any of these
patients. The mean +/-SD gestational age at delivery was 34.5 +/-3.5
weeks (range 28 to 39), with 10 (71%) delivered at <37 weeks'
gestation. There was one twin gestation.
Patients frequently complained of a viral-like illness on presentation,
noting a history that included fever, malaise, anorexia, chest pain,
dyspnea, or myalgias. The most common presenting symptom was
nausea and vomiting, noted in 10 patients (71%). Other presenting
complaints included: abdominal pain (seven), pruritus (two), mental
confusion (three), and pregnancy-related complaints in seven (three
patients had labor pains, one vaginal bleeding, two decreased fetal
movements, and one loss of fetal movements). Seven patients had
icterus on admission, five had abdominal tenderness, and two had
tremors or asterixis.
Table I summarizes the admitting diagnoses and admission-to-delivery
intervals. Only two patients were diagnosed with acute fatty liver of
pregnancy on admission; one because of acute fatty liver of pregnancy
in a previous pregnancy and one because of the typical presentation of
nausea and vomiting, icterus, abdominal tenderness, and laboratory
profile consistent with acute fatty liver of pregnancy. Preeclampsia was
the most commonly entertained diagnosis on initial presentation,
resulting in a significant delay in delivery for two patients.
One patient had acute fatty liver of pregnancy in the postpartum
period. This patient was admitted because of mild preeclampsia and
was delivered 12 days after admission. On postpartum day 1
hypoglycemia, high blood ammonia level, disseminated intravascular
coagulation, and respiratory arrest occurred. The diagnosis of acute
fatty liver of pregnancy was confirmed by CT of the liver.

Ten of the 14 cases had the diagnosis confirmed by liver biopsy. One
additional case was diagnosed by CT of the liver (no biopsy), and the
remaining three cases were diagnosed on the basis of clinical
presentation and laboratory findings. There were no complications
related to the biopsy procedures. Seven patients underwent a liver
biopsy and also had CT performed within 48 hours of the biopsy, with
only one demonstrating a positive result. Six of these seven patients
had the CT performed after the liver biopsy, before the biopsy result
was available; one had the liver biopsy performed after a negative CT
result. Overall, 10 women had a CT evaluation of the liver, but only two
results were consistent with acute fatty liver of pregnancy. In addition,
one patient with biopsy-confirmed acute fatty liver of pregnancy
underwent a magnetic resonance imaging of the liver, but the
evaluation was nondiagnostic.
Ten patients had cesarean deliveries: five for fetal distress, one for
breech presentation, one for suspected macrosomia, and three for an
unfavorable cervical Bishop score. The mean +/-SD birth weight of the
15 infants (one set of twins) was 2187 +/-858 gm (range 980 to 3940
gm), with two (14%) being small for gestational age (<10th percentile
according to Brenner et al.) [5]. The median Apgar score of the
livebirths at 1 minute was 5 (range 1 to 9) and at 5 minutes was 7
(range 5 to 9). The mean +/-SD arterial cord pH and base deficit was
7.20 +/-0.14 and 6.6 +/-5.1, respectively (only eight infants had cord
blood sampled). Two infants were significantly acidotic at birth (pH 6.89
and 7.10), and four live infants (27%) had Apgar scores of <7 at 5
minutes. One patient had an intrauterine fetal death, and one neonatal
death occurred (an infant with trisomy 18), for a total perinatal
mortality of 13.3% and a corrected perinatal mortality of 6.6%.
Placental pathologic studies were available on seven patients: two were
normal, two showed focal fibrinosis, two revealed infarcts with necrosis,
and the last one was consistent with acute chorioamnionitis and
umbilical cord vasculitis.
There were no maternal deaths; however, maternal morbidity was
significant: 10 patients required transfusion with blood or blood
products to correct disseminated intravascular coagulation or severe
bleeding (uterine, gastrointestinal, subcutaneous hematoma). Four
patients had hepatic encephalopathy, four had respiratory arrest
requiring assisted ventilation, three had ascites, three had renal
insufficiency, two had diabetes insipidus, and one had adult respiratory
distress syndrome. Only one pregnancy was complicated by abruptio

placentae. The mean maternal hospital stay was 13.6 days (range 7 to
37 days).
Table II summarizes significant laboratory findings for the study
patients. It is important to note that 13 (93%) of the patients had a
fibrinogen level <300 mg/dl and 10 patients (71%) had a glucose level
<70 mg/dl in spite of the use of intravenous dextrose solutions. The
data were subsequently analyzed in the 10 patients with biopsy-proved
diagnosis. The frequency of abnormal tests remained the same.
Table III summarizes the serial changes in laboratory findings in the
postpartum period. Platelet count and serum glucose were consistently
delayed in recovery. All patients had evidence of improvement in
laboratory parameters by the seventh postpartum day.

Comment
The management of acute fatty liver of pregnancy remains maternal
stabilization followed by delivery and supportive care. Early diagnosis
and intervention can result in excellent maternal and fetal outcomes. As
with any severe complication of pregnancy, maternal assessment and
stabilization should be the initial management goal. Fetal assessment
with heart rate monitoring or a biophysical profile is essential because
of the increased risk of fetal distress or stillbirth in pregnancies
complicated by acute fatty liver of pregnancy [6,7,8].
Patients with acute fatty liver of pregnancy may present with a variety
of symptoms including nausea, vomiting, malaise, fatigue, anorexia,
abdominal pain (diffuse or localized), fever, headache, diarrhea,
backache, myalgias, and pruritus [2,9]. Nine of our 14 patients (64%)
had an initial diagnosis of preeclampsia or HELLP syndrome. A high
concurrence of acute fatty liver of pregnancy and preeclampsia has
been reported in the literature [10,11,12]. Some authors even suggest
that the spectrum of liver involvement in preeclampsia includes acute
fatty liver of pregnancy within its confines [10]. This hypothesis is linked
to the overlap between the two conditions, including occurrence in
nulliparity and twin gestations and its presentation in the third
trimester. Both conditions tend to improve with delivery and share
clinical manifestations and laboratory abnormalities [11,13].
Hypoglycemia (glucose <70 mg/dl) was present in 10 (71%) of our
patients in spite of the use of intravenous dextrose solutions.
Hypoglycemia has been reported in 25% of the cases of acute fatty
liver of pregnancy and may persist for days or weeks [14]. Our patients

also had hyperuricemia, a combination of obstructive and hepatocellular

disturbance, and hypoalbuminemia. The plasma ammonia level was
elevated in 75% of our patients. Elevated ammonia level was implicated
as the cause of altered sensorium noted in 98% of patients with acute
fatty liver of pregnancy in one series [9]. Patients with altered
sensorium in the presence of a normal ammonia level should have
cranial CT to rule out other conditions such as cerebral ischemia,
cerebral edema, [15] or lipid emboli [16]. Evidence of encephalopathy
was present in four (29%) of our patients.
Hypofibrinogenemia was present in 93% of our patients, with fulminant
disseminated intravascular coagulation occurring in seven (50%) and
an altered coagulation status in three additional patients. Only one of
our patients had abruptio placentae, and it involved 20% of the
placental surface. Disseminated intravascular coagulation has been
commonly reported in association with acute fatty liver of pregnancy
[17]. Unlike in other obstetric accidents, disseminated intravascular
coagulation in acute fatty liver of pregnancy may continue long after
delivery and even worsen in the postpartum period, probably secondary
to the low antithrombin III levels, which may take several days to
recover [12,18]. Acute fatty liver of pregnancy should be suspected in all
patients with clinical signs and symptoms of preeclampsia in the
presence of laboratory evidence of disseminated intravascular
coagulation but in the absence of severe abruptio placentae.
Diabetes insipidus, which is the result of increased oxytocinase activity,
[14] complicated the course of two (14%) of our patients. In one patient
it was resistant to vasopressin administration. The diabetes insipidus
persisted for 7 days in one patient and 3 days in the other. In such
patients accurate determination of fluid intake and output with frequent
evaluation of serum electrolytes is recommended.
Thirteen of our patients had onset of the disease before delivery, and
50% had improvement in laboratory parameters by the second
postpartum day. Thus it seems that delivery promotes recovery of acute
fatty liver of pregnancy, and if initial conservative management of mild
cases is attempted it should be done with extreme caution [10]. Most
authors have recommended prompt delivery of patients with acute fatty
liver of pregnancy, irrespective of the severity of the condition, for
several reasons: the observed rapid improvement in maternal
laboratory parameters after delivery, the improvement in maternal and
fetal survival in recent years with concomitant use of immediate
delivery, the occurrence of the condition near term, the possibility of
sudden deterioration in the condition endangering the life of both

mother and fetus, and the observation that early delivery does not
worsen the low survival in fulminant hepatitis, a condition frequently
suspected in patients with acute fatty liver of pregnancy [2,19]. Whether
cesarean delivery or induction of labor is the best approach remains
controversial. Some authors have suggested that cesarean delivery is
hazardous for the mother and advise labor induction with close
monitoring [14]. Others have stated that early delivery by cesarean
section is advisable and may improve fetal prognosis, but whether this
benefits the mother is disputed [13,14].
In our series 10 patients had a liver biopsy performed without
complications. The definitive method of diagnosing acute fatty liver of
pregnancy remains a liver biopsy specimen demonstrating diffuse lowgrade centrilobular microvesicular fatty changes (steatosis) on light or
electron microscopic examination [20] clearly different from the classic
pathologic changes in the liver of patients with preeclampsia or HELLP
syndrome [21,22].
Of the 10 patients who underwent CT of the liver, only two (20%) had
abnormal results. One of our patients also had a magnetic resonance
imaging of the liver, which was nondiagnostic. Ultrasonography and CT
of the liver have been suggested as alternative diagnostic modalities
because they are noninvasive, safe with coagulation abnormalities,
widely available, able to assess the overall pattern of fatty liver changes
(thus avoiding sampling errors with nonuniform fat distribution), and
able to provide immediate results. Some investigators have reported a
20% false-negative rate using ultrasonography for diagnosis, whereas
others have reported a rate as high as 91% [23,24]. Similarly, the
accuracy of CT in detecting acute fatty liver of pregnancy is disputed
[7,14,25]. McKee et al [25] have reported that a liver density below the
normal range of 50 to 70 Hounsfield units on CT suggests acute fatty
liver of pregnancy; however, and as indicated from our work, CT has a
high false-negative rate and its use in diagnosis of acute fatty liver of
pregnancy needs further investigation. Improved timing of the
radiologic examination and serial examinations in patients suspected of
having acute fatty liver of pregnancy may reveal a clinical window with
greater sensitivity and specificity.
There were no maternal deaths in our study. Recent studies have
shown a maternal survival rate of 65% to 90% [9,10,13,19] for acute
fatty liver of pregnancy, compared with earlier reports where maternal
mortality reached 85% [2]. We attribute this improvement in maternal
mortality to the recognition of milder cases, use of maximal supportive
care, and expeditious delivery with reversal of the disease process.

Our perinatal mortality rate was 13.3%, which is similar to recent

reports of 14% to 18% [12,17] and much improved from the 60% to
80% perinatal mortality reported earlier [2,8,11]. The only stillbirth in
our study was in a patient with nausea and vomiting who was given
antiemetics, only to return 1 week later with an intrauterine fetal death,
at which time the diagnosis of acute fatty liver of pregnancy was
confirmed. The reasons for the poor fetal outcome in patients with
acute fatty liver of pregnancy remains an enigma. Fetal acidosis
secondary to maternal acidosis is a possible explanation. Moise and
Shah [8] have proposed that fetal distress results from decreased
placental perfusion secondary to fibrin deposition in the chorionic villi in
low-grade coagulopathy exacerbated by the replacement therapy in an
effort to correct disseminated intravascular coagulation.
There was one case of recurrent acute fatty liver of pregnancy in our
series [4]. We are unaware of any subsequent pregnancies complicated
by acute fatty liver of pregnancy for the other patients. It is generally
believed that acute fatty liver of pregnancy does not recur. The paucity
of previous reports of recurrence may be related to medical counseling
against further pregnancy, infrequent use of liver biopsy in some cases,
desire against further pregnancies, or a low incidence of recurrence. In
a review of the literature Watson and Seeds [13] could not find any
recurrence in 21 patients with 25 subsequent pregnancies. We
recommend close follow-up of subsequent pregnancies in patients with
previous acute fatty liver of pregnancy, especially in the third trimester.
Acute fatty liver of pregnancy is a rare but potentially fatal complication
of the third trimester of pregnancy. The morbidity and mortality of this
disorder has decreased significantly in recent years because of
increased awareness and aggressive management. Recommended
treatment continues to be expeditious delivery with maximal supportive
care.

REFERENCES
1. Sheehan HL. The pathology of acute yellow atrophy and delayed chloroform
poisoning. J Obstet Gynaecol Br Empire 1940;47:49-61. [Context Link]
2. Kaplan MM. Acute fatty liver of pregnancy. N Engl J Med 1985;313:367-70.
Bibliographic Links [Context Link]

3. Sibai BM. The HELLP syndrome (hemolysis, elevated liver enzymes, and low
platelets): much ado about nothing? AM J OBSTET GYNECOL 1990;162:311-6.
Bibliographic Links [Context Link]

4. Barton JR, Sibai BM, Mabie WC, Shanklin DR. Recurrent acute fatty liver of
pregnancy. AM J OBSTET GYNECOL 1990;163:534-8. Bibliographic Links
[Context Link]

5. Brenner WE, Edelman DA, Hendricks CH. A standard of fetal growth for the
United States of America. AM J OBSTET GYNECOL 1976;126:555-64.
Bibliographic Links [Context Link]

6. Varner M, Rinderknecht NK. Acute fatty metamorphosis of pregnancy: a

maternal mortality and literature review. J Reprod Med 1980;24:177-80.
Bibliographic Links [Context Link]

7. Mabie WC, Dacus JV, Sibai BM, et al. Computed tomography in acute fatty
liver of pregnancy. AM J OBSTET GYNECOL 1988;158:142-5. Bibliographic
Links [Context Link]

8. Moise KJ Jr, Shah DM. Acute fatty liver of pregnancy: etiology of fetal
distress and fetal wastage. Obstet Gynecol 1987;69:482-5. Bibliographic
Links [Context Link]

9. Pircon R, Freeman R. Acute fatty liver in pregnancy--it's manageable.

Contemp Obstet Gynecol 1988;4:55-64. [Context Link]
10. Riely CA, Latham PS, Romero R, Duffy TP. Acute fatty liver of pregnancy. A
reassessment based on observations in nine patients. Ann Intern Med
1987;106:703-6. Bibliographic Links [Context Link]
11. Burroughs AK, Seong NG, Dojcinov DM, Scheuer PJ, Sherlock SV.
Idiopathic acute fatty liver of pregnancy in 12 patients. Q J Med
1982;204:481-97. Bibliographic Links [Context Link]
12. Pokros PJ, Peters RT, Reynolds TB. Idiopathic fatty liver of pregnancy:
findings in ten cases. Medicine 1983;63:449-52. [Context Link]
13. Watson WJ, Seeds JW. Acute fatty liver of pregnancy. Obstet Gynecol Surv
1990;45:585-93. Bibliographic Links [Context Link]
14. Purdie JM, Waters BNJ. Acute fatty liver of pregnancy. Clinical features and
diagnosis. Aust N Z J Obstet Gynaecol 1988;28:62-7. Bibliographic Links
[Context Link]

15. Sherlock S. Acute fatty liver of pregnancy and the microvesicular fat
disease. Gut 1983;24:265-9. Bibliographic Links [Context Link]
16. MacKenna J, Pupkin M, Crenshaw C, McLeod M, Parker RT. Acute fatty
metamorphosis of the liver. A report of two patients who survived. AM J
OBSTET GYNECOL 1977;127:400-4. Bibliographic Links [Context Link]

17. Rolfes DB, Ishak KG. Acute fatty liver of pregnancy: a clinicopathologic
study of 35 cases. Hepatology 1985;5:1149-58. [Context Link]
18. Liebman HA, McGhee WG, Patch MJ, Feinstein DI. Severe depression of
antithrombin III associated with disseminated intravascular coagulation in
women with fatty liver of pregnancy. Ann Intern Med 1983;98:330-3.
Bibliographic Links [Context Link]

19. Hou SH, Levin S, Ahola S, et al. Acute fatty liver of pregnancy. Survival
with early cesarean section. Dig Dis Sci 1984;29:449-52. Bibliographic Links
[Context Link]

20. Riely CA. Acute fatty liver of pregnancy. Semin Liver Dis 1987;7:47-54.
Bibliographic Links [Context Link]

21. Arias F, Mancilla-Jimenez R. Hepatic fibrinogen deposits in preeclampsia. N

Engl J Med 1976;295:578-82. [Context Link]
22. Barton JR, Riely CA, Adamec TA, et al. Hepatic histopathologic condition
does not correlate with laboratory abnormalities in HELLP syndrome
(hemolysis, elevated liver enzymes, and low platelet count). AM J OBSTET
GYNECOL 1992;167:1538-43. Bibliographic Links [Context Link]
23. Campillo B, Bernuau J, Wite MO, et al. Ultrasonography in acute fatty liver
of pregnancy. Ann Intern Med 1986;105:383-4. Bibliographic Links [Context
Link]

24. Van Le L, Podrasky A. Computed tomography and ultrasonographic

findings in women with acute fatty liver of pregnancy. J Reprod Med
1990;35:815-7. Bibliographic Links [Context Link]
25. McKee CW, Weir PE, Foster JH, et al. Acute fatty liver of pregnancy and
diagnosis by computed tomography. BMJ 1986;292:291-2. [Context Link]

10

Table I. Admitting diagnosis and admission-to-delivery interval in

patients with acute fatty liver of pregnancy

Table II. Laboratory findings in acute fatty liver of pregnancy

Table III. Serial changes in laboratory findings post partum

11