Objectives To compare two modes of natalizumab cessation interventions: immediate versus tapered down, as measured by serial MRI and the occurrence of relapses during a 12-month period.

Background Weighing progressive multifocal encephalopathy risk associated with ≥24 months of natalizumab therapy against the benefits of disease control, we initiated a natalizumab discontinuation study.

Methods A phase IV, 12-month, single-blinded randomised (MRI) study. Fifty relapsing patients with multiple sclerosis (MS) who had been on natalizumab therapy ≥24 months and were contemplating natalizumab discontinuation were enrolled. Participants were randomised to either the immediate discontinuation group (IDG) or the tapered group (TG). IDG discontinued natalizumab at once and initiated another disease modifying therapy (DMT) following the last natalizumab infusion, while the TG received two more natalizumab infusions, at 6 and 8 weeks (14 weeks from study entry) before initiating another DMT. Standardised MRI was performed at baseline, 6 and 12 months from the last natalizumab infusion.

Results A higher rate of relapses in the IDG (n=28) compared to the TG (n=8) over 12 months from the last infusion (p=0.007) was observed, most relapses occurred within 3 months of discontinuation (20 vs 7 relapses, p=0.012). The IDG showed a higher number of new T2 lesions within 6–12 months of discontinuation (p=0.025), a higher mean absolute T2-LV change from 0 to 12 months (1.1 vs 0.1 mL, p=0.024) and a higher number of new T1-hypointense lesions over 0–12 months (p=0.005) as well as from baseline to 6 months (p=0.026) compared to the TG.

Conclusions Natalizumab discontinuation therapy was associated with development of new disease activity. Our tapered protocol showed benefits, as patients in the TG experienced less relapses and lower accumulation of MRI lesions compared to those in the IDG.