Manage obstetric cholestasis and liver disease in pregnancy

In one line

Liver disease in pregnancy divides into two groups. Some diseases are caused by the pregnancy: intrahepatic cholestasis, the pre-eclampsia and HELLP spectrum, and acute fatty liver. They exist only because she is pregnant, and they resolve once she delivers. The rest she brings to the pregnancy, which then destabilises them: viral hepatitis, gallstones, cirrhosis, a liver transplant. Almost every decision in this chapter follows from that division. The 2025 FIGO guideline on liver disease and pregnancy is the first to cover both groups for a global population.

The shared pre-eclampsia treatment (magnesium, blood-pressure control, fluid balance) is developed in severe early-onset pre-eclampsia and antihypertensives in pregnancy. This chapter builds on the Intermediate chapter medical complications in pregnancy, which carries the HELLP and medical-disorder groundwork you are assumed to know.

Intrahepatic cholestasis of pregnancy

Intrahepatic cholestasis is the commonest liver disease of pregnancy. The lesion is failed canalicular bile export. The hepatocyte normally pumps bile acids across its canalicular membrane through the bile salt export pump (BSEP), against a steep gradient and in one direction. In the third trimester, oestrogen and progesterone metabolites inhibit BSEP and its partner, the phospholipid flippase MDR3. Bile acids that should leave the cell accumulate in maternal serum instead. The placenta drives those hormones, so the disease appears in late pregnancy and clears within days of delivery. A minority of women carry inherited ABCB4 variants that impair MDR3; they present earlier, run more severely, overlap with progressive familial intrahepatic cholestasis, and warrant hepatology follow-up after delivery.

The danger lies with the fetus, not the mother. Her synthetic liver function is preserved, and her only symptom is pruritus. Bile acids cross the placenta and reach the fetal compartment, where they are directly arrhythmogenic to the fetal myocardium. At high concentrations they precipitate a sudden, fatal arrhythmia. The death is arrhythmic, not the slow placental failure of growth restriction, so it comes without warning. A reassuring CTG does not predict it, and antenatal surveillance does not prevent it. Timed delivery is the only effective intervention.

Delivery timing follows the peak maternal total bile acid, and the relationship is stepped, not linear. Below 40 µmol/L the stillbirth risk matches background. From 40 to 99 µmol/L it is modestly raised. At 100 µmol/L and above it climbs several-fold. The peak value sets the delivery date, not the severity of the pruritus and not the transaminases. The level fluctuates and runs higher after food, so the disease is graded on the peak non-fasting value tracked across the pregnancy, and a single low reading in a clearly itching woman does not exclude it.

FIGO makes the table operational. The diagnostic threshold is a non-fasting bile acid of 19 µmol/L, and once cholestasis is diagnosed, bile acids are measured weekly from 32 weeks, so a woman climbing toward 100 is caught before she crosses it. The EASL hepatology guideline (2023, by the same authors as FIGO) and South African practice (NDoH) follow the same framework. The US SMFM guidance diagnoses at any value above the laboratory upper limit, often around 10 µmol/L, and so labels a larger, milder group. That difference changes who is labelled, not who is delivered early, because every guideline times delivery on the same 100 µmol/L threshold.

Ursodeoxycholic acid (UDCA), 500 mg twice daily up to about 1.5 g/day, is first-line, but it has not been shown to reliably prevent stillbirth and must never be used to delay indicated delivery. FIGO offers it at a bile acid of 40 µmol/L or more, to reduce spontaneous preterm birth, a benefit with a number-needed-to-treat of 15, and it also relieves the pruritus. It does not replace timed delivery. Chlorphenamine aids sleep, emollients ease the itch, and vitamin K is given when the prothrombin time is prolonged or there is steatorrhoea, because fat-soluble-vitamin malabsorption raises maternal and neonatal bleeding risk. Rifampicin is the recognised second-line agent for refractory disease; it improves itch and biochemistry but does not change the delivery decision.

Acute fatty liver of pregnancy

In acute fatty liver the position is reversed. The mother is critically ill, and the only cure is to end the pregnancy. It is rare, about 1 in 10,000, but it is a true obstetric emergency.

Acute fatty liver is associated, in a subset of cases, with a fetal defect in mitochondrial fatty-acid oxidation, most often long-chain 3-hydroxyacyl-CoA dehydrogenase (LCHAD) deficiency. Where that association holds, the classic pattern is a homozygous affected fetus and a heterozygous carrier mother: the fetus and placenta cannot oxidise long-chain fatty acids, so the unoxidised 3-hydroxy intermediates spill into the maternal circulation, and the mother's own partially deficient liver is overwhelmed, fills with microvesicular fat, and fails. Most cases do not establish this specific genetics, so a firm genetic diagnosis and recurrence estimate should not be given on the strength of the acute fatty liver alone. Two consequences follow. The disease resolves only once the placenta, the source of the load, is delivered. And where a fetal disorder is present, the neonate carries the dangerous genotype, so the baby must be screened for fatty-acid-oxidation defects, which are treatable when caught and lethal when missed; overall recurrence in a future pregnancy is uncommon but real, generally under 10 percent.

The mother presents in the third trimester with malaise, nausea, vomiting and abdominal pain, then with the features of acute liver failure: hypoglycaemia, coagulopathy, hyperammonaemia, renal impairment and encephalopathy. The Swansea criteria support the diagnosis when six or more features are present without another cause:

The criteria overlap heavily with severe pre-eclampsia, so they support the diagnosis rather than confirm it. Liver biopsy is rarely justified, because the coagulopathy makes it dangerous and the result does not change management.

Management has two steps, in order: resuscitate, then deliver. Hypoglycaemia is corrected with a 10% dextrose infusion and frequent glucose checks; it is the most under-treated derangement and it drives the encephalopathy. The coagulopathy is corrected with fresh frozen plasma, cryoprecipitate and vitamin K. ICU, anaesthesia and hepatology are involved early, on the supportive principles of obstetric critical care. FIGO names the markers that warrant transfer to a transplant centre: a rising lactate, a MELD score (Model for End-stage Liver Disease) of 30 or more, or more than seven Swansea features. Delivery then follows at any gestation, by the quickest safe route the coagulopathy allows, because the liver recovers only after the pregnancy ends. Most survivors recover fully, often after a transient worsening in the first days, with no lasting liver damage. The disease recurs in future pregnancies, especially where a fetal LCHAD defect is confirmed.