Letter on ‘Non‐Response to Obeticholic Acid Is Associated With Heightened Risks of Developing Clinical Events in Primary Biliary Cholangitis’

We read Abbas et al.'s study [1] in Aliment Pharmacol Ther with great interest. Its 4-year international multicenter real-world follow-up represents a valuable contribution—for the first time, it validates the long-term prognostic value of biochemical response to obeticholic acid (OCA) in a large primary biliary cholangitis (PBC) cohort, effectively addressing the gap in long-term real-world evidence for OCA. That said, several critical limitations warrant careful consideration when interpreting its implications. A key concern lies in the use of ‘per-protocol’ analysis to assess the link between biochemical non-response and clinical event risk, specifically when exploring associations between baseline covariates and clinical event occurrence. In practical terms, this means if a patient stopped OCA due to treatment failure and disease progression, they were censored from survival analyses immediately after discontinuation. Importantly, subsequent clinical events, such as liver transplantation or death, were excluded from risk assessments for the ‘OCA non-response group’. It substantially underestimates the true risk of adverse outcomes in non-responders and artificially weakens the association between biochemical non-response and clinical events, undermining the reliability of the study's core conclusions. Equally notable is the data on cirrhotic patients. Abbas et al. [1] reported that 56% of patients with cirrhosis discontinued OCA within 4 years, with the vast majority discontinuing due to progressive liver disease. During this period, 26 patients underwent liver transplantation and 16 died. These data indicate that OCA has limited efficacy in controlling disease progression in these cirrhotic patients and may even accelerate disease deterioration via drug-related mechanisms. However, in their multivariate analysis, the authors adjusted for ‘cirrhosis’ as a confounding factor and concluded that ‘biochemical non-response is an independent predictor of clinical events’. This approach overlooks the fundamental premise that OCA may be contraindicated or inappropriate for patients with cirrhosis. In summary, the study has notable methodological shortcomings. To mitigate non-random censoring bias, ‘intention-to-treat’ (ITT) analysis [2] should have been the analytical approach of choice. Furthermore, ‘cirrhosis’ ought to have been treated as a key effect modifier, with subgroup analyses to separately evaluate OCA's efficacy and safety in cirrhotic versus non-cirrhotic PBC patients. Only then could the findings provide more robust guidance for clinical decision-making in PBC management. Yi Pei: writing – original draft. Runze Jiang: writing – original draft. Shenglong Li: writing – original draft, writing – review and editing. The authors declare no conflicts of interest. This article is linked to Abbas et al. papers. To view these articles, visit https://doi.org/10.1111/apt.70378. The data that support the findings of this study are available from the corresponding author upon reasonable request.