Intrahepatic cholestasis of pregnancy recurs in 4 in 10 pregnancies

Presenter Tatyana Kushner (Weill Cornell Medicine, New York, USA) told delegates at the EASL Congress 2026 in Barcelona, Spain: “This is the rate that I will now be using moving forward to counsel my patients in the US.”

She noted that although the rate of recurrence increased to 52% among women with more severe ICP whose total serum bile acid (SBA) peaked at 40 µmol/L or above in the index pregnancy, “peak bile acid levels alone did not predict ICP recurrence.”

Instead, they found that EASL’s 2023 clinical practice guidelines for stratifying women with suspected ICP provided a “clinically useful risk framework,” said the presenter, “with patients in Group B and Group C having highest risk of ICP recurrence,” relative to those in Group A.

According to the framework, women with pruritus with or without rash or suspected ICP are assigned to Group A if they have a normal total SBA and normal alanine aminotransferase/aspartate aminotransferase levels; to Group B if they have normal total SBA but raised liver enzymes/transaminases, or to Group C if they have raised total SBA and liver enzymes/transaminases.

Elevated total SBA according to the EASL diagnostic criteria is 19 µmol/L and above, whereas the US diagnostic criteria used in this study is 10 µmol/L and above.

Concordance in EASL-defined groups between pregnancies

The researchers first assessed 283 women (median age 32 years; 45% Hispanic; 35% White, non-Hispanic; 10% Asian, non-Hispanic; 7% Black, non-Hispanic, 3% other) within their healthcare system who had suspected ICP in an index pregnancy and had SBA assessed in both their index pregnancy and their subsequent one.

At their subsequent pregnancy, 163 of the women were classified as Group A, according to the EASL criteria, 71 as Group B, and 49 as Group C. As expected, women in group C were significantly more likely than those in Group B and Group A to have prior hepatobiliary disease, at 35% versus 18% and 12%, respectively, and a higher peak total SBA, at a median of 37 µmol/L versus 14 µmol/L and 9 µmol/L.

Kushner reported that the EASL ICP group classification between the index and subsequent pregnancy in these women was concordant in 52%, with only 12% of women transitioning to a higher risk group. This refutes the “traditional thinking that in a subsequent pregnancy ICP will become more severe,” she said.

Specifically, 81% of patients in Group A in their index pregnancy remained in Group A in their subsequent pregnancy, while 82% of patients in Group B remained in Group B or moved to Group A, and 66% of patients in Group C moved to Group B or Group A.

The team calculated that after taking into account a history of hepatobiliary disease, women were 4.6 times more likely to progress to Group C if they were in Group B rather than Group A in their index pregnancy, and 2.7 times more likely if they had a total SBA of 10 µmol/L or above in an index pregnancy.

EASL group classification predicts ICP recurrence

In all, 193 of the women had confirmed ICP in their index pregnancy, with a total SBA of 10 µmol/L or above (45% from Group A, 65% from Group B, and 90% from group C), 80 of whom experienced a recurrence in their subsequent pregnancy.

Among the 58 women with ICP who had total SBA levels of 40 µmol/L or above in their index pregnancy, 30 had a recurrence in their subsequent pregnancy.

Being in Group B or Group C versus Group A in the index pregnancy was a significant predictor of recurrence, at adjusted risk ratios of 2.95 and 2.62, respectively, as was being Hispanic, at an adjusted risk ratio of 1.99. However, total SBA levels of 40 µmol/L or above in the index pregnancy was not a significant predictor of recurrence.

Kusher noted that a recurrence of ICP did not increase the risk for adverse obstetric outcomes, and although ICP symptom onset before 28 weeks was more likely in patients who had recurrent ICP, she suggested that this might be due to “heightened awareness and earlier detection among patients with prior ICP.”

The researcher pointed out that the study did not involve any genetic data and called for “more prospective data with genetic data and standardized testing to learn more about recurrence risk and predictors.”