Patients were retrospectively analyzed using distinct HDA criteria: 1. High relapse

activity (HRA): ≥2 relapses in the year before study entry whether on disease modifying drugs (DMD) or not; 2. HRA+ disease activity on treatment (DAT): ≥1 relapse AND ≥1 T1 gadolinium-enhancing or ≥9 T2 lesions in the year before study entry while on other DMDs. The CT3.5 group were patients exposed to CT3.5 in Years 1 and 2, while the CT7.0 group received further courses of CT in Years 3 and 4. Adverse Event of Special Interest (AESI) lymphopenia was analyzed across CLARITY and CLARITY EXT and was determined by absolute lymphocyte count (ALC).

In CT3.5 at Year 2 (CT3.5-Y2), AESIs/100 patient-years were similar in HRA (n=196) and HRA+DAT (n=219) subgroups (11.50; 13.09) and similar between corresponding non-HDA groups (14.08 [n=489]; 13.46 [n=466]). In CT7.0 at Year 4 (CT7.0-Y4) rates approximately doubled from CT3.5-Y2. Median ALC nadir (x10/L) for CT3.5-Y2 HRA and HRA+DAT were 0.71 and 0.70; median nadirs were 0.65 in corresponding non-HDA subgroups. CT7.0-Y4 had lower nadirs than CT3.5-Y2. CT3.5-Y2 incidence of Grade 3 lymphopenia, was similar between HDA and non-HDA subgroups whereas, CT7.0-Y4 was lower. Overall, Grade 3 lymphopenia incidence was lower in CT3.5-Y2 than CT7.0-Y4. Grade 4 lymphopenia incidence was low in all subgroups.