Should women with intrahepatic cholestasis of pregnancy be delivered early?
Yes. The only known way to reduce risk of stillbirth is with early delivery.
Intrahepatic cholestasis of pregnancy (ICP) is associated with an increased risk of stillbirth, and the only known way to reduce this risk is early delivery. The rationale is 3-fold: 1) incidence of stillbirth in ICP is higher than in the general pregnant population; 2) ICP-associated stillbirths cluster toward the end of the third trimester; and 3) early delivery reduces the ICP-related perinatal mortality rate.
ICP increases risk of stillbirth
Compared to patients without ICP, those affected by ICP have a higher stillbirth rate. The stillbirth rate at 37 weeks’ gestation and beyond for the entire population in the United States is approximately 0.1% to 0.3% (1–3 per 1000).1,2 It is lower than the term stillbirth rate in ICP that was published by Henderson et al. Their study reported a 1.2% (4/331) term stillbirth rate attributed to ICP with expectant management.3 This was calculated after excluding cases of ICP with comorbidities such as diabetes, preeclampsia, abruption, intrauterine growth restriction, and congenital fetal anomalies. Puljic et al. studied more than 1.6 million singleton pregnancies between 34 and 40 weeks’ gestation and compared the stillbirth rate in those with and without ICP.4 They found a significantly higher risk of stillbirth in those with ICP compared to the unaffected population. It persisted for pregnancies between 32 and 40 weeks’ gestation. In a prospective cohort study evaluating patients affected by ICP with total bile acid concentrations ³ 40 μmol/L, Geenes et al., after adjusting for confounders, found a higher incidence of stillbirth in the population with ICP compared to the unaffected controls.5
Stillbirths in ICP cluster toward the end of the third trimester. More than 20 years ago, a striking case series of stillbirths attributed to ICP described obstetric outcomes of 8 women with 13 pregnancies affected with ICP. Twelve of the pregnancies were managed expectantly. Eight of the pregnancies resulted in stillbirth, and the gestational ages at death were: 37, 39, 32, 37 and 37 weeks, with 3 more listed as “at term.”6 Williamson et al. reported their findings in 227 women with 352 total pregnancies affected by ICP. Of these pregnancies, 5.7% (20/352) had an intrauterine death in a singleton pregnancy. The median gestational age when death occurred was 38 weeks (IQR 2.5) with 10% (2/20) occurring before 37 weeks’ gestation.7 In the aforementioned prospective study by Geenes et al., which had a 1.5% (10/664) incidence of stillbirth with ICP, the median gestational age at delivery for the cases with ICP and stillbirth was 36 weeks ± 2 days (IQR 35 ± 4 to 38 ± 1 days). Six of the 10 stillbirths occurred before 37 weeks’ gestation. Of these 10 stillbirths, 2 were to mothers with preeclampsia, 2 had gestational diabetes, and 2 had nonspecified complications. The authors clarified that no stillborn fetus was small for gestational age. Three stillbirths were large for gestational age; none of these were from mothers with gestational diabetes.5 Kawakita et al. had 4 stillbirths in 26 patients with ICP and a total bile acid concentration ³ 100 μmol/L. These stillbirths occurred at 37 1/7, 35 3/7, 24 1/7, and 35 5/7 weeks’ gestation. None of the stillbirths were affected by congenital abnormalities. The one stillbirth at 24 1/7 weeks’ gestation had a normal karyotype and occurred in a mother who had elevated transaminases prior to pregnancy.8 Alsulyman et al. reported 2 stillbirths in 79 patients with ICP who were managed expectantly. The mean gestational age at delivery was 38.5 ± 1.9 weeks for the cohort, and the 2 stillbirths occurred at 36 to 37 weeks’ gestation. Both were appropriately grown with no gross abnormalities.9
The case for early delivery
Early delivery is the only known intervention that reduces the risk of stillbirth. This practice was recommended more than 40 years ago by Reid et al., who reported 5 stillbirths in 56 pregnancies affected by ICP.10 Subsequently, they adopted a practice of inducing labor at term that led to a decrease in their hospital’s ICP-related perinatal mortality rate from 107 per 1000 to 35 per 1000.11 Several other studies soon followed that demonstrated the effect of early delivery on lowering the perinatal mortality rate in ICP. Rioseco et al. reviewed 320 pregnancies with ICP during a time when their practice was to deliver at 38 weeks’ gestation. They found this led to similar perinatal mortality rates between the ICP affected and unaffected populations. In the cohort of patients with ICP, there were 4 stillbirths (1.3% [4/320]), which occurred between 33 and 38 weeks’ gestation. None of the mothers had obstetrical complications nor did the fetuses have growth abnormalities.12 Roncaglia et al. studied 206 pregnancies with ICP that occurred during a time when they induced patients with ICP at 37 weeks’ gestation; there were no stillbirths. Those managed with early delivery had a significantly lower fetal death rate compared to their historical cohort of patients with ICP who were managed expectantly.13 Turunen et al. studied 687 pregnancies with ICP and compared them to 1374 unaffected controls.14 A proxy for early delivery, the labor induction rate was higher in the cohort with ICP compared to controls. There was no statistical difference in the incidence of stillbirth between the 2 groups. Kenyon et al. reported no stillbirths in 70 pregnancies with ICP that were managed with a protocol that offered elective delivery at 37–38 weeks’ gestation.15 Of these patients, 76% had labor induced. Rook et al., whose clinical practice was to deliver at approximately 37 weeks’ gestation, reported no stillbirths in 101 women with ICP. The average gestational age at delivery was 37 ± 1.2 weeks, and 87% were induced.16
We evaluated 122 patients with ICP when our practice was to deliver at 37 weeks’ gestation. The average gestational age at delivery for our cohort was 36.7 ± 2.1 weeks. Elective delivery occurred in 86.9% and spontaneous labor occurred in 13.1%. One stillbirth occurred at 30 weeks’ gestation.17
Lo et al. meticulously calculated the optimal gestational age for delivery in women with ICP. After accounting for neonatal mortality and morbidities associated with early delivery and the risk of stillbirth with ICP, they demonstrated the optimal time to deliver patients with ICP is at 36 weeks’ gestation.18