Current guideline-based management of the most common pregnancy-specific liver disease with bile acid thresholds, risk of stillbirth, discussion of UDCA, and timing of delivery.
⚕️ Target audience: This clinical review is intended for physicians and healthcare professionals; It is a summary of current literature and guidelines for educational purposes. Clinical decisions should be made based on patient-specific evaluation and applicable institutional protocols.
Intrahepatic cholestasis of pregnancy (ICP, formerly known as obstetric cholestasis) is the most common pregnancy-specific liver disease. It is defined by the association of itching and elevated serum bile acid without a primary lesion on the skin. What makes the condition clinically critical is that it increases the risk of sudden and unpredictable stillbirth, even though it is generally benign in the mother; This risk has been redefined according to bile acid level with large-scale data conducted in the last decade.
ICP typically begins in the third trimester; Itching is mostly evident on the palms and soles of the feet, worsens at night, and resolves rapidly after birth. Biochemical abnormalities also return to normal within a few weeks during the puerperium. Incidence shows marked ethnic and geographic variability: approximately 0.6–0.7% in women of European descent, ~1.5% in Scandinavia, ~1.75% in mainland China, and ~4% in Andean South American populations (2024 review — PMC11384812). Women of Indian subcontinent origin in the United Kingdom and Indigenous communities in Australia are high-risk groups.
Risk factors include history of ICP (personal or familial), multiple pregnancy, in vitro fertilization, history of hepatitis C and cholelithiasis. Both incidence and severity increase in multiple pregnancy.
Diagnosis is based on demonstration of elevated serum bile acid with unexplained pruritus in normal-appearing skin; Other causes of pruritus and liver disease are excluded. In the current consensus, the diagnostic threshold is adopted as random (non-fasting) peak total bile acid ≥19 µmol/L (RCOG GTG 43, 2022; SOGC No. 452, 2024).
Sampling timing is critical: The bile acid sample should be collected in the non-fasting (postprandial) state, not fasting. The postprandial peak captures high-risk cases; a fasting sample alone may miss these patients. For reference, fasting levels are approximately 10 µmol/L and postprandial levels approximately 20 µmol/L.
Transaminases (ALT/AST) and bile acid are measured at diagnosis (SMFM Recommendation 1, GRADE 1B). However, it is bile acid that is decisive in risk stratification — this distinction stems from data showing that although ALT is used in diagnosis, it does not predict prognosis (below).
RCOG GTG 43 (2022) uses a tripartite classification based on peak total bile acid; SMFM Consult #53 (2021), on the other hand, aggregates management decisions into a binary threshold (<100 vs ≥100 µmol/L).
| Class | Peak total bile acid | Source / comment |
|---|---|---|
| Light | 19–39 µmol/L | RCOG GTG 43 (2022); risk of stillbirth is similar to that of the general population |
| Medium | 40–99 µmol/L | RCOG GTG 43 (2022) |
| Severe | ≥100 µmol/L | RCOG GTG 43 (2022); SMFM critical decision threshold—significant increase in risk of stillbirth |
ALT does not predict stillbirth: In the Ovadia 2019 meta-analysis with individual patient data (IPD), stillbirth in singleton pregnancies was strongly associated with maximum total bile acid (ROC AUC 0.83; 95% CI 0.74–0.92) but was not significantly associated with ALT (ROC AUC 0.46; 0.35–0.57) (PubMed; DOI: 10.1016/S0140-6736(18)31877-4). Risk is stratified only by bile acid, and also by the highest value measured during pregnancy; Transaminase elevation alone does not direct the timing of birth.
ICP results from impaired bile acid transport in genetically predisposed individuals under the high hormonal load of pregnancy. Estrogen, and especially its sulfated progesterone metabolites, inhibit hepatobiliary bile acid transporters—primarily BSEP (ABCB11) and MDR3 (ABCB4). Genetic predisposition overlapping with ABCB11/ABCB4 variants and benign recurrent intrahepatic cholestasis (BRIC) has been reported in early-onset, recurrent or treatment-resistant cases (Gonzalez 2025 — DOI: 10.1016/j.crwh.2025.e00714).
The mediator of pruritus is not the classic accumulated bile salts: lysophosphatidic acid (LPA) and the enzyme autotaxin and sulfated progesterone metabolites are the main candidates for pruritogens; it seems unlikely that total bile acid and bilirubin conjugates play a dominant role in pruritus (Beuers 2022, Nat Rev Gastroenterol Hepatol — DOI: 10.1038/s41575-022-00687-7).
The most accepted explanation for fetal mortality is the fetal cardiac hypothesis. High concentrations of bile acids—especially taurocholate—disrupt contraction rate and rhythm in fetal cardiomyocytes, destabilize intracellular calcium dynamics, and cause arrhythmia through a muscarinic M2 receptor-mediated mechanism; this may result in sudden intrauterine death (Williamson group, J Hepatol 2021 — PMC8062912; primary study: Clin Sci 2001, PMID 11256973). It has been experimentally shown that ursodeoxycholic acid can prevent this arrhythmia by reregulating T-type calcium current. The clinical relevance of the mechanism is decisive: stillbirth usually occurs suddenly, without any pre-asphyxia symptoms — so antenatal monitoring cannot reliably predict or prevent stillbirth.
The contemporary basis for ICP management is the Ovadia 2019 (Lancet) meta-analysis: synthesis of 23 studies total (5557 ICP cases, 165,136 controls) and 27 studies IPD (5269 cases). The risk of stillbirth in singleton pregnancies differs significantly according to the bile acid layer.
| Peak total bile acid | Prevalence of stillbirth | Hazard rate (HR) |
|---|---|---|
| <40 µmol/L | 3/2310 = 0.13% (0.02–0.38) | Reference (similar to general population) |
| 40–99 µmol/L | 4/1412 = 0.28% (0.08–0.72) | 2.35 (0.52–10.50); p=0.26 (insignificant) |
| ≥100 µmol/L | 18/524 = 3.44% (2.05–5.37) | 30.50 (8.83–105.30); p<0.0001 |
The clinical message is clear: increased risk of stillbirth occurs mainly above the threshold of ≥100 µmol/L (PubMed; DOI: 10.1016/S0140-6736(18)31877-4). The majority of ICP patients are below this threshold; When bile acid levels are monitored repeatedly throughout pregnancy in these patients, reassurance can be provided that the risk of stillbirth is similar to that of the general pregnant population. In risk calculation, the highest value measured so far is always taken as basis.
UDCA has been considered the first-line agent of ICP for decades; however, recent high-quality evidence has challenged this established consensus regarding perinatal benefit. The double-blind, placebo-controlled PITCHES study (Chappell 2019, Lancet; 605 women) failed to show superiority over placebo with UDCA 500 mg × 2/day (1–4 tablets/day as needed) on the primary composite outcome (perinatal death + <37 weeks preterm birth + ≥4 hours neonatal unit admission): UDCA 23% vs placebo 27%, aRR 0.85 (95% CI 0.62–1.15). There was also no clinically significant difference in pruritus (PubMed; DOI: 10.1016/S0140-6736(19)31270-X). Secondary analysis found no benefit in any subgroup, including high bile acid or high pruritus (Fleminger 2020, BJOG — DOI: 10.1111/1471-0528.16567). It should be noted that approximately 75% of the cohort was <40 µmol/L at baseline.
In contrast, the Ovadia 2021 IPD meta-analysis (Lancet Gastroenterol Hepatol; 34 studies, 6974 women) presented a more nuanced picture: UDCA alone had no effect on stillbirth (all studies aOR 1.04, 0.35–3.07; RCTs alone aOR 0.29, 0.04–2.42), but When only randomized trials were analyzed, there was a reduction in stillbirth + preterm birth composite: aOR 0.60 (95% CI 0.39–0.91; p=0.016) (PubMed; DOI: 10.1016/S2468-1253(21)00074-1). This is where the difference in interpretation arises: UDCA does not prevent stillbirth, but may provide clinical benefit by reducing preterm birth.
This divergence of evidence is directly reflected in the guideline recommendations:
In practice, UDCA remains a reasonable option for the control of maternal pruritus and biochemical disturbance; however, neither the RCOG nor the SOGC position it as a means to prevent stillbirth.
Rifampin (rifampicin) is added as an adjuvant in UDCA-resistant cases. As a PXR agonist, it induces bile acid detoxification and biotransformation enzymes. The typical dose range in the literature is 300–1200 mg/day, usually starting with 300 mg and increasing according to response. Although maternal symptoms and transaminases often resolve rapidly, fetal deterioration may still develop (Liu 2018, Obstet Gynecol — DOI: 10.1097/AOG.0000000000002794; Beuers 2022 — DOI: 10.1038/s41575-022-00687-7). Monitoring is required for reversible hepatotoxicity and rarely acute kidney injury/DIC; Since rifampin disrupts vitamin K metabolism, neonatal vitamin K is recommended at birth.
In more severe/refractory cases, bezafibrate (PPAR agonist) appears promising; Biochemical normalization with bezafibrate was reported in a case resistant to UDCA, rifampin and cholestyramine, reaching 389 µmol/L (Gandelsman 2025, BMC Pregnancy Childbirth — DOI: 10.1186/s12884-025-08168-w). Cholestyramine can be used as a bile acid sequestrant, but the evidence is weak and it may worsen vitamin K deficiency by impairing fat-soluble vitamin absorption. Antihistamines and topical emollients only provide symptomatic relief.
Timing of birth is a matter of balancing the risk of stillbirth against the iatrogenic morbidity of preterm birth and is the topic where the most significant divergence is seen among the guidelines. While SMFM uses a binary threshold, SOGC 2024 offers the most granular four-layer scheme.
| Peak bile acid | SMFM #53 (2021) | SOGC No. 452 (2024) |
|---|---|---|
| 19/20–39 µmol/L | 36 0/7 – 39 0/7 weeks (<100) [GRADE 1C] | 39 0/7 – 39 6/7 weeks |
| 40–69 µmol/L | 38 0/7 – 38 6/7 weeks | |
| 70–99 µmol/L | (≥36 0/7 in the above range) | 36 0/7 – 37 6/7 weeks |
| ≥100 µmol/L | 36 0/7 weeks [GRADE 1B] | Up to 36 weeks |
SMFM also recommends against preterm birth before 37 weeks for a sole clinical diagnosis of ICP without laboratory confirmation [GRADE 1B] and recommends administering antenatal corticosteroids if delivery is planned before 37 weeks [GRADE 1A]. Rather than imposing a fixed scheme, RCOG GTG 43 (2022) individualizes the timing of birth according to the bile acid level and additional risk factors (multiple pregnancy, gestational diabetes, preeclampsia); At ≥100 µmol/L, earlier delivery is considered. ACOG also supports the approach of 36 weeks for ≥100 µmol/L and 36–39 weeks for lower levels, based on Committee Opinion 831. In severe cases with early onset, when bile acid levels rise well above 100 µmol/L, an individualized, earlier delivery decision can be made.
The value of antenatal fetal surveillance is limited in ICP and the guidelines are clear on this issue. SMFM #53 states that surveillance may be initiated at the gestational age when an abnormal test result would lead to delivery (or at diagnosis if the diagnosis is made late), but this is a weak recommendation [GRADE 2C]. SOGC 2024 is even clearer: antenatal fetal monitoring has not been shown to improve perinatal outcomes. In practical practice, SOGC recommends monitoring every 1–2 weeks at 40–99 µmol/L, weekly or twice weekly at ≥100 µmol/L, and at least weekly bile acid monitoring after the 32nd week.
Limit of monitoring: Because stillbirth is sudden and unpredictable, cardiotocography and biophysical profile cannot reliably predict or prevent stillbirth. Monitoring can give false reassurance; The main preventive intervention is timely delivery, determined by the bile acid level. Repeated measurement of bile acid is a more valuable indicator of risk than surveillance tests.
ICP recurs at a high rate in subsequent pregnancies. In the historical literature, the recurrence rate is given as 60–90% (~92% in familial cases, ~40% in sporadic cases; UK population 90%, Chinese cohort 55%); SOGC 2024 adopts the 70–90% range. A recent single-center cohort (2024, n=104) reported a lower rate (44.2%; 95% CI 34.6–53.8). In this study, peak total bile acid level >40 µmol/L in the index pregnancy was a significant risk factor for recurrence (30.4% vs 6.9% in the recurrence group; p<0.01); Interestingly, an interpregnancy interval of less than 18 months was found to be protective (OR 0.39, 0.16–0.91) (PMC12571205).
In the long term, a history of ICP has been associated with a slightly increased risk of hepatobiliary disease, hepatitis C, cardiovascular and metabolic disease; SOGC and RCOG therefore recommend postpartum liver function control and long-term lifestyle counseling in appropriate cases. In case of persistent biochemical abnormality, underlying chronic liver disease should be investigated.
Last updated: . Guide numbers and publication years have been verified against current catalogues.