Ursodeoxycholic acid in intrahepatic cholestasis of pregnancy: an individual participant data meta-analysis

Caroline Ovadia, Jenna Sajous, Paul Townsend Seed, Kajol Patel, Nicholas J Williamson, George Attilakos, Francesco Azzaroli, Yannick Bacq, Linoy Batsry, Kelsey Broom, Romana Brun-Furrer, Laura Bull, Jenny Chambers, Yue Cui, Min Ding, Peter Dixon, Maria Estiu, Fergus Gardiner, Victoria Geenes, Monika GrymowiczBerrin Gunaydin, William Hague, Christian Haslinger, Yayi Hu, Ugo Indraccolo, Alexander Juusela, Stefan Kane, Ayse Kebapcilar, Levent Kebapcilar, Katherine Kohari, Jurate Kondrackiene, Maria Koster, Richard Lee, Xiaohua Liu, Anna Locatelli, Rocio Macias, Riza Madazli, Agata Majewska, Kasia Maksym, Jessica Marathe, Adam Morton, Martijn Oudijk, Deniz Oztekin, Michael Peek, Andrew Shennan, Rachel Tribe, Valeria Tripodi, Naciye Turk Ozterlemez, Tharni Vasavan, L.F. Audris Wong, Yoav Yinon, Qianweng Zhang, Keren Zloto, Hanns-Ulrich Marschall, Jim Thornton, Lucy Chappell, Catherine Williamson

Research output: Contribution to journalArticlepeer-review

44 Downloads (Pure)


Ursodeoxycholic acid (UDCA) is commonly used to treat intrahepatic cholestasis of pregnancy (ICP), yet its largest trial detected minimal benefit for a composite outcome (stillbirth, preterm birth, and neonatal unit admission). We aimed to determine whether UDCA affects specific adverse perinatal outcomes.
Following systematic review of the literature, authors of selected studies were invited to submit individual participant data for meta-analysis. Studies of all design without evidence of selection bias were eligible for inclusion, reported by January 2020. The primary outcome was stillbirth, although we anticipated insufficient data to achieve statistical power, and included a composite of stillbirth and preterm birth as a main secondary outcome. A mixed-effects meta-analysis was performed using multilevel modelling, adjusting for bile acid concentration, parity, and multifetal pregnancy. Analyses were performed in Stata version 16·0; the study was registered in PROSPERO: CRD42019131495.
Eighty-five studies were selected, from which data were provided for 6974 women from 34 studies, of whom 4726 (67·8%) took UDCA. With adjustment for confounders, the results including observational studies were similar to those of randomised controlled trials (RCT). In RCTs, UDCA treatment had no statistically significant effect on stillbirth (adjusted odds ratio (aOR) 0·29, 0·04 to 2·42, p=0·25), but was associated with a reduced composite outcome (aOR 0·60, 0·39 to 0·91, p=0·016), largely due to reduced total preterm birth (aOR 0·61, 0·40 to 0·92, p=0·019). This was likely due to reduced spontaneous preterm birth, which was reduced when comparing participants of all studies (aOR 0·55, 0·35 to 0·88, p=0·012).
IPD revealed no significant effect of UDCA on stillbirth, likely limited by the low overall event rate. However, rates in combination with preterm birth were reduced, providing evidence for the clinical benefit of antenatal UDCA treatment.
Tommy’s, Wellcome Trust, ICP Support, and the National Institute for Health Research.
Original languageEnglish
Pages (from-to)1-25
Number of pages25
JournalThe Lancet Gastroenterology & Hepatology
Publication statusAccepted/In press - 22 Feb 2021


Dive into the research topics of 'Ursodeoxycholic acid in intrahepatic cholestasis of pregnancy: an individual participant data meta-analysis'. Together they form a unique fingerprint.

Cite this