MedWire News: A large "real-world" study fails to show reduced disability progression in patients with relapsing-remitting multiple sclerosis (MS) given interferon beta.
For the study, which appears in JAMA, Helen Tremlett (University of British Columbia, Vancouver, Canada) and colleagues analyzed prospectively collected data for nearly 3000 patients with up to 10 years of follow up.
In an accompanying editorial, Tobias Derfuss and Ludwig Kappos (University Hospital Basel, Switzerland) stress: "Lacking evidence of treatment effect is not proof of lacking effect."
They observe that "although methodologically sound, this study cannot avoid the inherent challenges of data analysis and interpretation in nonrandomized observational studies."
Foremost of these is selection bias, and this, they argue, is visible in the team's findings.
Tremlett et al assessed the time to disability progression, defined as a sustained score of 6 on the Expanded Disability Status Scale (EDSS). The progression rates were 10.8% among 868 patients treated with interferon beta (median 5.1 years follow up) and 5.3% among 829 untreated contemporary controls (median 4.0 years follow up).
After accounting for confounders including age, disease duration, and baseline EDSS score, treated patients were a nonsignificant 1.3-fold more likely to progress than the untreated group. But compared with a group of 959 historical patients (10.8 years follow up), who had MS before interferon beta became available, treated patients were a nonsignificant 23% less likely to progress.
The difference in the direction of these associations is "compatible with the assumption that the contemporary cohort included a high number of patients who did not qualify for treatment because their disease was relatively benign, thus introducing a bias against interferon treatment," say Derfuss and Kappos.
However, the results were unaffected by further adjustment for propensity score (for receiving interferon beta).
Both the researchers and the editorialists note that the study could have been underpowered to detect an effect of interferon beta.
Derfuss and Kappos say the findings "reinforce the conclusion that the associations between use of interferons and long-term disability, although plausible, remain unproven."
They add that "more effective treatment options and better criteria that lead to more accurate selection of patients who might best respond to these treatments are needed."
MedWire (www.medwire-news.md) is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2012