Bezafibrate for primary biliary cholangitis: A number needed to treat analysis

Ellen Werner; Maria A van de Vrie; Maria C B van Hooff; Bettina E Hansen; Adriaan J van der Meer; Dutch PBC Study Group

doi:10.1016/j.jhepr.2026.101926

Bezafibrate for primary biliary cholangitis: A number needed to treat analysis

JHEP Rep. 2026 Jun 10;8(8):101926. doi: 10.1016/j.jhepr.2026.101926. Online ahead of print.

Authors

Ellen Werner¹, Maria A van de Vrie¹, Maria C B van Hooff¹, Bettina E Hansen², Adriaan J van der Meer³; Dutch PBC Study Group

Collaborators

Dutch PBC Study Group:
Sunje Abraham, Rob P R Adang, Huseyin Aktas, Yasser A Alderlieste, Maria N Aparicio-Pages, L Bert C Baak, Martine A M C Baven-Pronk, A Sander van der Beek, Frank C Bekkering, Jeroen D van Bergeijk, Ulrich Beuers, Koen Beukema, Menno Beukema, Tiki D Blom, Wink de Boer, Femke Boersma, Kirsten Boonstra, Frank Ter Borg, Martijn J Ter Borg, Pieter C J Ter Borg, Gijs J de Bruin, Paul J Bus, Djuna L Cahen, Marcel Cazemier, Frans J C Cuperus, Lisette J H van Dam, Maaike J Denters, Remco van Dijk, Joost P H Drenth, Ludger S M Epping, Nicole S Erler, Hajo J Flink, Philip W Friederich, Nicole F M van Gerven, Tom J G Gevers, Bettina E Hansen, Sven J van den Hazel, Bart van Hoek, Maria C van Hooff, Daphne M Hotho, Harry L A Janssen, Hendrik J M de Jonge, Matthias C Jurgens, J Netty van Kemenade, Marjo J Kerbert-Dreteler, Michael Klemt-Kropp, Ingrid C A W Konings, Sander de Kort, Matthijs Kramer, Edith M M Kuiper, Johan P H Kuyvenhoven, Adriaan J van der Meer, Suzanne van Meer, Susanne L Onderwater, Leendert H Oterdoom, Cyriel Y Ponsioen, Paul G van Putten, Janne E van Rooij, Robert Roomer, Johannes Schmidt-Böhmer, Stephan Schmittgens, Tim C M A Schreuder, Jerome Sint Nicolaas, Hanneke van Soest, Khalida Soufidi, Stephan H C van Stiphout, Hans H K Thio, Merel M Tielemans, Sigrid Vandebosch, Rozanne C de Veer, Bart J Veldt, Robert C Verdonk, J Marleen de Vree, Elsemieke de Vries, Anne Vrieze, Jan Maarten Vrolijk, Laurens A van der Waaij, Gemma X Weijsters, Ellen Werner, Ulrike de Wit, Frank H J Wolfhagen

Affiliations

¹ Department of Gastroenterology and Hepatology, Erasmus University Medical Center, Rotterdam, the Netherlands.
² IHPME, University of Toronto; Toronto, Canada; Toronto Center for Liver Disease & TGHRI, University Health Network, Toronto, Canada; Department of Epidemiology, Erasmus University Medical Center, Rotterdam, the Netherlands.
³ Department of Gastroenterology and Hepatology, Erasmus University Medical Center, Rotterdam, the Netherlands. Electronic address: a.vandermeer@erasmusmc.nl.

PMID: 42269833
DOI: 10.1016/j.jhepr.2026.101926

Abstract

Background & aims: The clinical benefit of bezafibrate (BZF) in primary biliary cholangitis (PBC) has not been extensively quantified using absolute measures. We assessed the number needed to treat (NNT) with BZF to prevent liver transplantation (LT) or death (including liver-related death) in patients with PBC.

Methods: The NNT was calculated using LT-free survival probabilities among ursodeoxycholic acid (UDCA)-treated patients with PBC from the Dutch PBC Cohort Study and the reported adjusted hazard ratios (aHRs) for the effect of BZF on LT or death (aHR 0.33) and liver-related death (aHR 0.27) from the Japanese cohort study (Tanaka et al., J Hepatol 2021). The NNT to prevent one LT, death, or liver-related death within 5 years (NNT_5y) or 10 years (NNT_10y) was estimated.

Results: In total, 2,271 (63.7%) patients with PBC had an alkaline phosphatase (ALP) level above the upper limit of normal (ULN) after 1 year of UDCA treatment. The median age was 57.1 years (IQR 50.1-67.0), 2,003 (88.2%) were female, and 187 (8.2%) had cirrhosis. Overall, the estimated NNT_5y and NNT_10y for BZF were 15 (95% CI 11-26) and 7 (95% CI 6-13), respectively, to prevent one LT or death, and 26 (95% CI 18-56) and 15 (95% CI 11-30), respectively, to prevent one LT or liver-related death. According to ALP category at one year of UDCA treatment (1.0-1.5 × ULN, 1.5-3.0 × ULN, and ≥3.0 × ULN), the NNT_5y to prevent one LT or death was 24 (95% CI 16-48), 12 (95% CI 9-23), and 7 (95% CI 5-15), respectively, while the corresponding NNT_10y values were 10 (95% CI 7-18), 7 (95% CI 5-13), and 4 (95% CI 3-8).

Conclusions: The projected efficacy of BZF, expressed as the NNT to prevent hard clinical endpoints, was substantial overall, including among patients with relatively favourable ALP levels (1.0-1.5 × ULN) after one year of UDCA treatment. These findings highlight the potential clinical benefit of off-label BZF in UDCA-treated patients without ALP normalisation.

Impact and implications: Although infrequently used in hepatology, the number needed to treat (NNT) can increase awareness among clinicians and patients about the potential impact of second-line therapy. Bezafibrate showed strong efficacy (NNT <20) for preventing liver transplantation or (liver-related) death in people with primary biliary cholangitis, even in patients with relatively favourable alkaline phosphatase levels (1.0-1.5 × the upper limit of normal) after 1 year of ursodeoxycholic acid. Findings support the potential clinical benefit of off-label bezafibrate for patients who do not achieve alkaline phosphatase normalisation on ursodeoxycholic acid alone. The NNT facilitates shared decision-making between patients and physicians on the use of second-line therapy in people with primary biliary cholangitis.

Keywords: Number Need to Treat; Primary Biliary Cholangitis; Second-line Therapy.