Assessment of different treatment failure criteria in a cohort of relapsing-remitting multiple sclerosis patients treated with interferon beta: implications for clinical trials.

Clinical trials with interferons in relapsing-remitting multiple sclerosis have shown a modest effect on disability using fixed definitions of treatment failure to measure disease progression. However, in the course of the disease, treatment failure may be influenced by interrater variability and frequent remissions. Thus, the purpose of this study was to assess the clinical usefulness of different treatment failure criteria in a cohort of relapsing-remitting multiple sclerosis patients treated with interferon beta. We studied 252 patients with a follow-up of more than 2 years. We used four different criteria of treatment failure with increasing stringency (1 Expanded Disability Status Scale [EDSS] point increase confirmed at 3 months, 1 EDSS point increase confirmed at 6 months, 1.5 EDSS points increase confirmed at 3 months, and 1.5 EDSS points increase confirmed at 6 months). We divided treatment failure into permanent treatment failure and transient treatment failure. We considered permanent treatment failure when treatment failure was confirmed on the last two scheduled visits and transient treatment failure when treatment failure was not confirmed on these visits at different time points (9, 12, 18, and 24 months). We calculated the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the different criteria of treatment failure to identify patients who achieved a high degree of disability after 4 years of follow-up. Regardless of the stringency of treatment failure definitions, a variable proportion of patients with treatment failure had transient treatment failure depending on the criterion applied. Patients with transient treatment failure had a significantly lower EDSS at entry compared with those with permanent treatment failure or no treatment failure. The number of relapses in patients with transient treatment failure did not differ from that of patients with permanent treatment failure. The criterion of confirmed 1 EDSS point increase at 6 months showed the best sensitivity (76.5%), with satisfactory specificity (89%). Our study shows that a large proportion of patients treated with interferon experience transient treatment failure that may affect outcome interpretation in clinical trials. Using a more strict criterion, as extending time to confirmation of EDSS deterioration, and longer follow-up may reduce this proportion of patients with transient treatment failure and improve the validity of the results attained in clinical trials.