Drug Holiday of Interferon Beta 1b in Multiple Sclerosis: A Pilot, Randomized, Single Blind Study of Non-inferiority

To compare a schedule with cyclic withdrawal (CW) of interferon beta (IFN-b) 1b, respect to the full regimen (FR), in relapsing-remitting MS (RR-MS). Participants were randomly assigned to CW or FR schedule and monthly monitored with brain MRI scans for 12 months (three of run-in and 9 of treatment). CW schedule included drug withdrawal for 1 month after two of treatment for a total of three quarters over the 9-month treatment period. The assessing neurologist and the expert neuroradiologists were blind. After the blind phase of the study all participants took their indicated disease modifying therapies in a prospectively planned, open-label extension phase (up to 120 months). Of 60 randomized subjects 56 (29 in FR and 27 in CW group) completed the single-blind phase: the two groups were comparable, except for a non-significant difference in the number of contrast-enhanced lesions (CEL) at the end of run-in. The two-sided 90% CI for the ratio between median number of cumulative CEL was 0.29-1.07, allowing to significantly reject the null hypothesis of a ratio ≥1.2 and to meet the primary end-point of non-inferiority (the threshold and the ratio between median were chosen according to the non-normal distribution of the data). The differences (CW vs. FR) were also non-significant for secondary end points: mean cumulative number of T2-weighted new and enlarging lesions (3.48 ± 5.34 vs. 3.86 ± 6.76); mean number and volume (cm) of black holes (1.24 ± 1.61 vs. 2.71 ± 4.56; 489.11 ± 1488.12 vs. 204.48 ± 396.98); number of patients with at least an active scan (21 vs. 22); mean relapse rate (0.52 ± 0.89 vs. 0.34 ± 0.66), relapse risk ratio adjusted for baseline variables (2.15 [0.64-7.18]), EDSS score (1.0 [1-1.56] vs. 1.5 [1-1.78]), proportion of patients with antibodies anti-IFN (5 [21%] vs. 8 [36%]). Fifty-four patients (27 for each study arm) completed the open-label phase. The annualized RR, EDSS, proportion of patients shifting to progressive disease and hazard ratio of shifting, adjusting for baseline covariates, were comparable between the two study groups. A calendar with CW was non-inferior than FR at the beginning of IFN-b therapy, and may not affect the long-term outcome. www.ClinicalTrials.gov, identifier: NCT00270816.