Data collection and analysis We used standard Cochrane methodological procedures. Main results The review included 10 trials (249 participants) using different treatment regimens. Seven of the 10 trials assessed single agents, and 3 assessed combined agents. Many of the studies did not present adequate data for the reporting of the primary outcome of the review, which was the percentage change in muscle strength score at six months. Pooled data from two trials of interferon beta-1a (n = 58) identified no important difference in normalised manual muscle strength sum scores from baseline to six months (mean difference (MD)-0.06, 95% CI-0.15 to 0.03) between IFN beta-1a and placebo (moderate-quality evidence). A single trial of methotrexate (MTX) (n = 44) provided moderate-quality evidence that MTX did not arrest or slow disease progression, based on reported percentage change in manual muscle strength sum scores at 12 months. None of the fully published trials were adequately powered to detect a treatment effect. We assessed six of the nine fully published trials as providing very low-quality evidence in relation to the primary outcome measure. Three trials (n = 78) compared intravenous immunoglobulin (combined in one trial with prednisone) to a placebo, but we were unable to perform meta-analysis because of variations in study analysis and presentation of trial data, with no access to the primary data for re-analysis. Other comparisons were also reported in single trials. An open trial of anti-T lymphocyte immunoglobulin (ATG) combined with MTX versus MTX provided very low-quality evidence in favour of the combined therapy, based on percentage change in quantitative muscle strength sum scores at 12 months (MD 12.50%, 95% CI 2.43 to 22.57). Data from trials of oxandrolone versus placebo, azathioprine (AZA) combined with MTX versus MTX, and arimoclomol versus placebo did not allow us to report either normalised or percentage change in muscle strength sum scores. A complete analysis of the effects of arimoclomol is pending data publication. Studies of simvastatin and bimagrumab (BYM338) are ongoing. All analysed trials reported adverse events. Only 1 of the 10 trials interpreted these for statistical significance. None of the trials included prespecified criteria for significant adverse events. Authors' conclusions Trials of interferon beta-1a and MTX provided moderate-quality evidence of having no effect on the progression of IBM. Overall trial design limitations including risk of bias, low numbers of participants, and short duration make it difficult to say whether or not any of the drug treatments included in this review were effective. An open trial of ATG combined with MTX versus MTX provided very lowquality evidence in favour of the combined therapy based on the percentage change data given. We were unable to draw conclusions from trials of IVIg, oxandrolone, and AZA plus MTX versus MTX. We need more randomised controlled trials that are larger, of longer duration, and that use fully validated, standardised, an...