Low-dose methotrexate is ineffective in primary biliary cirrhosis: long-term results of a placebo-controlled trial.

Abstract

New treatments for primary biliary cirrhosis (PBC) need to be evaluated. We conducted a single-center double-blind, randomized trial of methotrexate, 7.5 mg/wk (n = 30), vs. placebo (n = 30) for up to 6 years in PBC. Methods included three monthly symptom assessment and liver function tests and liver biopsy and gastroscopy at baseline, after 2 years, and after 4-6 years. Patients randomized to methotrexate had, compared with patients randomized to placebo, (1) significantly lower on-treatment serum alkaline phosphatase, gamma-glutamyltransferase, immunoglobulin (Ig) M, IgG, and (after 24 months) aspartate aminotransferase and alanine aminotransferase levels (P < 0.02-0.001 by analysis of covariance to adjust for baseline differences); (2) a nonsignificant trend toward lower on-treatment pruritus scores; (3) similar on-treatment Knodell inflammatory scores but nonsignificant trends toward lower Knodell fibrosis score and less ductopenia; (4) a trend toward greater increases in serum bilirubin level and Mayo score with time (both significant after 5 years of follow-up); and (5) a 2.9-fold (95% confidence interval, 0.85-10.25-fold) increase the rate of death or liver transplantation as a result of liver disease during or after the trial (P = 0.09) in a Cox multivariate regression analysis compared with patients randomized to placebo. These results do not support the clinical use of low-dose methotrexate in PBC.