There is currently no pharmacological therapy proven to alter the natural course of primary sclerosing cholangitis (PSC). The PRIMIS trial evaluated the efficacy and harms of the farnesoid X-activated receptor agonist cilofexor in participants with non-cirrhotic PSC.

In this phase 3, double-blind, placebo-controlled, multicentre trial (205 sites across 16 countries), adults aged 18–75 years with non-cirrhotic (F0–F3 [Ludwig classification]) large-duct PSC were randomly assigned (2:1) via an interactive web response system to receive cilofexor 100 mg or placebo (identical in appearance) orally once daily for 96 weeks. Randomisation was stratified by ursodeoxycholic acid use (yes or no) and the presence of bridging fibrosis (F3 vs F0–F2), with a block size of six within each stratum. Participants, personnel directly involved in the conduct of the study, and outcome assessors were masked to treatment assignment. The primary endpoint was the proportion of participants with histological progression of liver fibrosis (a stage increase of one or more [Ludwig classification]) at week 96. After study termination, the primary endpoint analysis set was amended to include all participants in the harms analysis set (all who received at least one dose of the study drug) who had biopsy data at baseline and week 96. This trial is complete (ClincalTrials.gov, NCT03890120).

Between June 13, 2019, and July 22, 2021, 419 participants were randomly assigned, and 416 were included in the full and harms analysis sets (cilofexor: n=277; placebo: n=139); 257 (62%) men and 159 (38%) women. The study was terminated early on Sept 26, 2022, after a planned interim futility analysis after 160 patients had reached 96 weeks of follow-up indicated a 6·8% probability of detecting a significant difference between cilofexor over placebo (futility boundary ≤10%). In the final analysis of the primary endpoint, for which 133 patients in the cilofexor group and 64 in the placebo group had liver biopsy results available, fibrosis progression occurred in 41 (31%) participants in the cilofexor group and 21 (33%) in the placebo group at week 96 (treatment difference −1·4% [95% CI –15·2 to 12·3]; p=0·42). The most common adverse events were pruritus (cilofexor: 136 [49%] of 277 patients; placebo: 50 [36%] of 139 patients; grade 3 or higher in 11 [4%] patients in the cilofexor group and one [1%] patient in the placebo group), COVID-19 (cilofexor: 65 [23%]; placebo: 26 [19%]), and upper abdominal pain (cilofexor: 40 [14%]; placebo 20 [14%]). The proportion of serious adverse events was similar between groups (cilofexor: 53 [19%]; placebo: 26 [19%]). There were no treatment-related deaths.

Cilofexor did not significantly reduce the rate of fibrosis progression (vs placebo) in participants with non-cirrhotic PSC. A greater percentage of cilofexor-treated participants had pruritus than placebo-treated participants; this study provides valuable harms data for cilofexor and other drugs in its class.