EASL Clinical Practice Guidelines

Cholestatic disorders in pregnancy

Intrahepatic cholestasis of pregnancy (ICP)

Intrahepatic cholestasis of pregnancy (ICP, also known as obstetric cholestasis) is a reversible form of cholestasis characterized by (i) intense pruritus in pregnancy (starting in the second or third trimester of pregnancy in most patients), (ii) elevated serum ALT activities and fasting serum bile acid levels, and (iii) spontaneous relief of signs and symptoms after delivery (within 4–6 weeks) [[188], [189]]. In Europe, about 0.4–2.0% of pregnancies are affected [[188], [190]]. The clinical importance of ICP lies in the potential fetal risks (spontaneous or iatrogenic prematurity, asphyxial events during delivery, intrauterine death), albeit perinatal mortality rates from recent studies (9/1000) are comparable to whole population figures, most likely due to improvements in obstetric and neonatal care [191]. Pruritus (typically worse at night) impairs the mother's quality of life. Only infrequently, ICP is associated with steatorrhea and postpartum haemorrhage due to vitamin K deficiency.

The pathogenesis of ICP is multifactorial, with genetic, hormonal and environmental factors playing important roles. During ICP, there is an increased flux of bile acids from the mother to the fetus, as indicated by elevated bile acid levels in amniotic fluid, cord blood and meconium [192]. The central role of hormonal factors is supported by the higher ICP incidence in twin pregnancies and the observation that high-dose oral contraceptives and progesterone can trigger ICP [188]. An increased ICP incidence in family members and ethnic differences point to genetic factors. Recent genetic studies have identified gene variants of hepatocanalicular transport proteins (ATP-binding cassette [ABC] transporter B4 = phosphatidylcholine floppase, ABC transporter B11 = bile salt export pump, ABC transporter C2 = conjugated organic anion transporter, ATP8B1 = FIC1) and their regulators (e.g., the bile acid sensor farnesoid X receptor, FXR) in some ICP patients [189]. Mild malfunction of these hepatocanalicular transporters could trigger cholestasis when their transport capacity for hormones or other substrates is exceeded during pregnancy. Currently, genetic tests are performed in research laboratories only and are not applicable for diagnosis or risk stratification. However, mutation analysis of ABCB4 might be considered in the future if cholestasis (with increased γGT levels) persists after delivery.

Diagnosis

The skin should be inspected to differentiate scratching lesions from other skin disorders such as eczema and pruritic eruption of pregnancy. Although pruritus can precede any abnormalities in liver function, serum liver tests (ALT, bilirubin, γGT, bile acids, prothrombin time) are to be performed in every pregnant woman who experiences itching and to be repeated if normal in case of persistent pruritus. The diagnosis of cholestasis of pregnancy is based on otherwise unexplained pruritus and elevated serum bile acid concentrations (⩾11 μmol/L) [192]. Isolated elevation of bile acids may occur but this is uncommon; in the majority of patients, ALT activities are elevated as well. Bile acids are the most sensitive indicator for cholestasis of pregnancy and may precede the abnormalities of other serum liver tests. Bile acid levels >40 μmol/L any time during pregnancy and early onset of ICP (<33 weeks of gestation) might be associated with significantly increased fetal complication rates [[190], [193], [194], [195]]. ICP patients with ABCB4 variants tend to display elevated γ GT levels, which are otherwise normal in ICP. Mild jaundice with serum levels of conjugated bilirubin only moderately elevated occurs in 10–15% of cases. Liver biopsy is generally not warranted.

Pre-eclampsia and acute fatty liver of pregnancy are pregnancy specific causes of abnormal serum liver tests that may form part of the differential diagnosis in atypical or early ICP cases (Table 6).

Table 6
Characteristics of ICP, HELLP Syndrome and acute fatty liver of pregnancy [196].


ICP HELLP AFLP
% Pregnancies 0.1–1.0 0.2–0.6 0.005–0.01
Trimester (2 or) 3 3 or postpartum 3 or postpartum
Family history Often No Occasionally
Presence of preeclampsia No Yes 50%
Typical clinical features Pruritus Elevated serum ALT/AST fasting bile acids Haemolysis Elevated serum liver tests Thrombocytopenia (often <50,000/μL) Liver failure with mild jaundice, coagulopathy, encephalopathy, hypoglycemia, DIC
ALT (above normal) Mild to 10–20-fold Mild to 10–20-fold 5–15-fold, variable
Bilirubin <5 mg/dL (<85 μmol/l) Mostly <5 mg/dL (<85 μmol/l) Often <5 mg/dL (<85 μmol/l)
Hepatic imaging Normal Hepatic infarcts, hematomas, hepatic rupture Fatty infiltration
Maternal mortality (%) 0 1–25 7–18
Fetal/perinatal mortality (%) 0.4–1.4 11 9–23
Recurrence in subsequent pregnancies (%) 45–70 4–19 20–70 (carriers of LCHAD mutations)
Rare (others)
LCHAD: α-subunit, long-chain 3-hydroxyacyl-CoA dehydrogenase.

Persistent abnormalities after delivery should prompt reconsideration of other chronic liver diseases like PBC, PSC, ABCB4 deficiency or chronic hepatitis C, which may be associated with development of pruritus during late pregnancy.

Therapy

Ursodeoxycholic acid (UDCA, 10–20 mg/kg per day) is regarded as the first-line treatment for ICP based on evidence obtained from randomized clinical trials [[193], [194], [197], [198], [199], [200]]. UDCA may improve pruritus and serum liver tests in 67–80% of ICP patients, but reduction of fetal complication rates is uncertain as fetal complication rates were low in recent trials both in UDCA and placebo-treated patients.

Dexamethasone (12 mg/day for 7 days) promotes fetal lung maturity, but is ineffective in reducing pruritus and ALT levels in patients with ICP [197]. Thus, this drug is not an adequate treatment of ICP [191].

S-Adenosyl-l-methionine is less effective than UDCA [200], but may have an additive effect [199]. If pruritus does not adequately respond to UDCA standard therapy for several days, the dose may be increased up to 25 mg/kg/day [201], or alternatively, treatment with S-adenosylmethione (combined with UDCA) or rifampicin might be considered on an individual basis (see Section 4.1). Topical emollients are safe but their efficacy is unknown.

Active obstetric management (including amnioscopy and generous induction of labour) has been reported to reduce perinatal mortality but increases intervention and complication rates [[194], [202], [203]]. The practice of considering delivery at (36 to) 38 weeks of gestation appears to prevent stillbirth beyond that gestation, but is not evidence-based [191].

Diagnosis and treatment of obstructive cholestasis during pregnancy

Although up to 10% of patients develop stones or sludge over the course of one pregnancy, symptomatic gallstones occur in only 1.2% of these pregnancies [204]. The diagnosis is based on clinical symptoms, elevated serum liver tests (ALT, bilirubin, γGT, AP) and abdominal (or endoscopic) ultrasound. Obstructive cholestasis due to an impacted common bile duct stone or worsening gallstone pancreatitis are indications to proceed to endoscopic retrograde cholangiography (ERCP), sphincterotomy and stone extraction under antibiotic coverage. Several series have demonstrated the safety of ERCP in pregnancy [[205], [206]]. An experienced physician should perform the intervention. Ultrasound-guidance might be helpful to minimize ionising radiation of the fetus (uterus dose 24 mSv/min). For deep sedation, consultation of an anesthesiologist and obstetrician is recommended. Meperidine, propofol, fentanyl and midazolam may be used at low doses [207]. Ampicillin is the preferred antibiotic and is compatible with breastfeeding [207] (Table 7).

Table 7
Medical treatment options in cholestatic disorders during pregnancy [[207], [208]].


Indication/drug Fetal risk (FDA category) Use and safety
Immune-mediated disorders
UDCA B Low risk
Prednisolone C Low risk: increased risk of cleft palate [211], adrenal insufficiency [212]
Azathioprine D Low risk
Bacterial cholangitis
Ampicillin B Low risk
Sedation and analgesia
Fentanyl C Use in low doses
Meperidine B Use in low doses
Midazolam D Use in low doses
Propofol B Avoid in first (and second) trimester
Fetal risk categories (FDA): A – no risk; B – risk in animal studies, but not in humans; C – human risk cannot be excluded; D – risk; X – absolute contraindication.

Drugs for cholestatic conditions during pregnancy

Women with cholestatic liver diseases may be of childbearing age, and an uncomplicated pregnancy with no disease flare is expected in those with mild or inactive disease. The course of autoimmune hepatitis or overlap syndrome in pregnancy is highly variable, and a flare of activity may occur during pregnancy or, more likely, in the post partum period. Table 7 summarizes the safety of drugs for cholestatic liver diseases [208].

UDCA. Although UDCA is not approved, but likely to be compatible, for use during early pregnancy, UDCA can be administered in cholestatic liver disease, when the pregnant woman is symptomatic during the second or third trimesters [209]. No adverse effects in mothers or newborns have been observed [210] including recent RCT, using UDCA for up to 8 weeks [[189], [197], [198], [199]]. UDCA is not approved during breastfeeding, but likely to be safe for the baby, since significant amounts of UDCA cannot be found in milk during lactation.

Corticosteroids. The use of prednisolone is considered safe during pregnancy and lactation, but is associated with an increased risk of cleft palate in children to women using the drug in the first trimester [211]. An increased risk of premature rupture of membranes and adrenal insufficiency was reported in the transplant setting [212].

Azathioprine. Azathioprine appears to be a safe drug during pregnancy, although it is teratogenic in animals. Steadily increasing experience is being reported in women with autoimmune hepatitis, rheumatoid arthritis, inflammatory bowel diseases, and after organ transplantation [[208], [213]]. The benefits and risks of therapy should be discussed in detail with the patient. Although very little azathioprine is excreted in breast milk, breastfeeding should be discussed on an individual basis.

Recommendations

  1. Diagnosis of ICP is based on (i) pruritus in pregnancy, (ii) elevated serum ALT activities and fasting bile acid levels, and (iii) exclusion of other causes of liver dysfunction or itching (II-2/C2). ICP is confirmed when serum liver tests completely normalize after delivery.
  2. Women with ICP should be advised that the incidence of premature birth is increased, both spontaneous and iatrogenic (II-2/B1). No specific fetal monitoring can be recommended (III/C2). UDCA ameliorates pruritus and improves serum liver tests (I/B1), but there are insufficient data concerning protection against fetal complications (II-1/C2). Vitamin K should be supplemented when prothrombin time is prolonged (III/C2). Timing of delivery should be discussed on an individual basis (II-2/C2).
  3. UDCA can be administered to pregnant women with cholestatic liver diseases during the second or third trimesters, when the patients are symptomatic (I/B1). Prednisolone ± azathioprine for treatment of autoimmune hepatitis should be continued during pregnancy to prevent disease flares, which might be more deleterious to pregnancy outcome than any potential risk of the medication (III/C2).
  4. Symptomatic bile duct stones in pregnancy are treated by endoscopic sphincterotomy and stone extraction (II-3/B1). X-ray is not absolutely contraindicated even in the first trimester (III/C2). Patients with simultaneous gallbladder and bile duct stones who are asymptomatic after clearance of the bile duct should undergo cholecystectomy post partum (III/C2).