Primary biliary cirrhosis is an uncommon chronic liver disease of unknown aetiology, affecting mostly women. It is characterised by progressive inflammation and destruction of the liver tissue, eventually progressing to liver cirrhosis and the need for liver transplantation. Methotrexate, a folic acid antagonist with immunosuppressive properties, has been used to treat patients with primary biliary cirrhosis. However, the evidence did not show a clear benefit of methotrexate on mortality or the need for liver transplantation in patients with primary biliary cirrhosis. This review is based on five randomised trials; four comparing methotrexate with placebo, and one comparing methotrexate with colchicine. Methotrexate, compared with placebo, has no significant beneficial effect on mortality and the need for liver transplantation is not significantly reduced. The effects of methotrexate on pruritus, fatigue, clinical complications, liver biochemistry levels, liver histology, and adverse events were not significantly different from placebo. There may be some beneficial effect on pruritus score (ie, an objective measure of subjective feeling of pruritus), but we cannot recommend methotrexate for this indication only, taken into account possible adverse events. In the small trial comparing methotrexate versus colchicine, methotrexate seemed to work superior to colchicine, but it is not clear if this stems from the fact that methotrexate exerts beneficial effects as colchicine exerts harmful effects. In comparison with both placebo and colchicine, methotrexate was associated with large risks of mortality and adverse events, but the increase did not reach statistical significance.
Methotrexate had no statistically significant effect on mortality in patients with primary biliary cirrhosis nor the need for liver transplantation. Although methotrexate may benefit other outcomes (pruritus score, serum alkaline phosphatase, immunoglobulin M levels), there is no sufficient evidence to support methotrexate for patients with primary biliary cirrhosis.
Methotrexate has been used to treat patients with primary biliary cirrhosis as it possesses immunosuppressive properties. The previously prepared version of this review from 2005 showed that methotrexate seemed to significantly increase mortality in patients with primary biliary cirrhosis. Since that last review version, follow-up data of the included trials have been published.
To assess the beneficial and harmful effects of methotrexate for patients with primary biliary cirrhosis.
Randomised clinical trials were identified by searching The Cochrane Hepato-Biliary Group Controlled Trials Register, The Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library, MEDLINE, and EMBASE (from their inception until September 2009). Reference lists were also read through. Authors of trials were contacted.
We searched to include randomised clinical trials comparing methotrexate with placebo, no intervention, or another drug irrespective of blinding, language, year of publication, or publication status.
Our primary outcomes were mortality, and mortality or liver transplantation combined. Dichotomous outcomes were reported as relative risks (RR) and hazard ratios (HR) if applicable. Continuous outcomes were reported as mean differences (MD).
Five trials were included. Four trials with 370 patients compared methotrexate with placebo or no intervention (three trials added an equal dose of ursodeoxycholic acid to the intervention groups). The bias risk of these trials was high. We did not find statistically significant effects of methotrexate on mortality (RR 1.32, 95% CI 0.66 to 2.64), mortality or liver transplantation combined, pruritus, fatigue, liver complications, liver biochemistry, liver histology, or adverse events. The pruritus score (MD - 0.17, 95% CI - 0.25 to - 0.09) was significantly lower in patients receiving methotrexate. The prothrombin time was significantly worsened in patients receiving methotrexate (MD 1.60 s, 95% CI 1.18 to 2.02). One trial with 85 patients compared methotrexate with colchicine. The trial had low risk of bias. Methotrexate, when compared to colchicine, did not significantly effect mortality, fatigue, liver biopsy, or adverse events. Methotrexate significantly benefited pruritus score (MD - 0.68, 95% CI - 1.11 to - 0.25), serum alkaline phosphatases (MD - 0.41 U/l, 95% CI - 0.70 to - 0.12), and plasma immunoglobulin M (MD - 0.47 mg/dl, 95% CI - 0.74 to - 0.20) compared with colchicine. Other outcomes showed no statistical difference.