BackgroundInjectable disease-modifying therapies (iDMTs) are often used as first-line treatments for relapsing multiple sclerosis. Fingolimod is frequently used following treatment with iDMTs. Whether prior iDMT treatment impacts the effectiveness of subsequent fingolimod therapy is unclear. Here, we assessed switching from iDMTs to fingolimod, and the impact of treatment history on fingolimod escalation using data from the 12-month ‘Prospective, Randomized, active-controlled, open-label study to Evaluate patient retention on Fingolimod versus approved first-line disease-modifying thErapies in adults with Relapsing–remitting Multiple Sclerosis’ (PREFERMS). The study design and results at the end of randomized treatment (EoRT) in PREFERMS have been published. MethodsBoth treatment-naïve patients and those who had previously received an iDMT were eligible for enrolment in PREFERMS, and one treatment switch was permitted on study. Pre-specified exploratory analyses compared outcomes in those randomized to fingolimod or to an iDMT at end of study (EoS), which included time spent on randomized and on switch treatment. Post hoc exploratory analyses (unadjusted for multiplicity owing to the large number of comparisons) among patients randomized to an iDMT who switched to fingolimod, compared outcomes longitudinally before (EoRT) and after (EoS) switching, and compared outcomes at EoRT and EoS among subgroups stratified by iDMT-treatment history. Outcomes included brain volume, various measures of gadolinium-enhancing [Gd+] lesion counts, annualized relapse rate (ARR), Symbol Digit Modalities Test (SDMT) score, patient-reported treatment satisfaction using the Medication Satisfaction Questionnaire (MSQ) and adverse event (AE) rates. ResultsAt EoS, 255 of 439 patients randomized to an iDMT had switched to fingolimod and 27 of 436 patients randomized to fingolimod had switched to an iDMT. By EoS, 44.2% of total treatment exposure in the iDMT group was to fingolimod and the mean time spent on fingolimod in this group was 220 days (approximately 7 months). Outcomes in the fingolimod group at EoS (brain volume, changes in Gd+ lesion counts, ARR, oral SDMT score and MSQ score) were similar to those seen at EoRT, but in the iDMT group these outcomes were more favorable at EoS than at EoRT and were similar to rates seen in the fingolimod group. Among patients who switched from iDMT to fingolimod, there were longitudinal improvements in ARR (EoRT, 0.3 [95% confidence interval (CI), 0.2–0.4]; EoS, 0.2 [0.1–0.3]; odds ratio, 0.5 [0.3–0.9]) and in treatment satisfaction (proportion of patients with MSQ > 5; EoRT, 67.4%; EoS, 90.4%; odds ratio, 5.7 [95% CI, 3.4–9.4]) after fingolimod treatment, and changes in brain volume, Gd+ lesion count, and AEs or AEs causing discontinuation were also more favorable at EoS than at EoRT. In all patient groups stratified by iDMT-treatment history, differences in outcomes narrowed or disappeared after fingolimod treatment. ConclusionThese analyses indicate that patients in PREFERMS had improved outcomes within months of switching to fingolimod from an iDMT and that improvements occurred irrespective of the number of iDMTs previously administered. These data provide a unique opportunity to explore clinical, radiological and safety outcomes associated with a range of clinically relevant treatment pathways.