Journal of Neurology (2019) 266:1182–1193 https://doi.org/10.1007/s00415-019-09248-6 ORIGINAL COMMUNICATION Ocrelizumab efficacy in subgroups of patients with relapsing multiple sclerosis Benjamin Turner1 · Bruce A. C. Cree2 · Ludwig Kappos3 · Xavier Montalban4,5 · Caroline Papeix6 · Jerry S. Wolinsky7 · Regine Buffels8 · Damian Fiore9 · Hideki Garren9 · Jian Han9 · Stephen L. Hauser2 Received: 7 December 2018 / Revised: 11 February 2019 / Accepted: 14 February 2019 / Published online: 28 February 2019 © The Author(s) 2019 Abstract Objective The efficacy and safety of ocrelizumab, versus interferon (IFN) β-1a, for the treatment of relapsing multiple scle- rosis (RMS) from the identically designed OPERA I (NCT01247324) and OPERA II (NCT01412333) phase III studies has been reported; here we present subgroup analyses of efficacy endpoints from the pooled OPERA I and OPERA II populations. Methods Patients with RMS were randomized to either ocrelizumab 600 mg administered by intravenous infusion every 24 weeks or subcutaneous IFN β-1a 44 µg three times per week throughout the 96-week treatment period. Relapse, disability, and MRI outcomes were analyzed for predefined and post hoc subgroups based on demographic and disease characteristics along with prior treatment using appropriate statistical tests to determine the treatment effect in subgroups and treatment- by-subgroup interactions. Results The significant treatment benefit of ocrelizumab, versus IFN β-1a, observed in the overall OPERA I and OPERA II pooled populations was maintained across most subgroup strata for all endpoints, including annualized relapse rate, disability progression, and MRI outputs. Conclusions The treatment effect of ocrelizumab versus IFN β-1a, measured by clinical and MRI outcomes, was maintained across most of the subgroups and strata of interest, and the pattern of treatment benefit across all subgroups was consistent with that from the pooled OPERA studies. Keywords Relapsing · Multiple sclerosis · Ocrelizumab · Subgroup · Interferon β-1a · Phase 3 Introduction OPERA I and OPERA II studies [1]. The identical study designs, comparable baseline demographics and disease The efficacy and safety of ocrelizumab, a humanized C­ D20+ characteristics of patients, and the evidence of no interaction B-cell selective monoclonal antibody, versus high-dose, high- between treatment and trial on the primary endpoint (annual- frequency interferon (IFN) β-1a in patients with relapsing ized relapse rate [ARR]) between the OPERA I and OPERA multiple sclerosis (RMS), was demonstrated in the phase III II trials [1], permitted pooling of trial data. Compared with IFN β-1a, treatment with ocrelizumab Electronic supplementary material The online version of this resulted in a 47% lower ARR (primary endpoint; p < 0.001) article (https​://doi.org/10.1007/s0041​5-019-09248​-6) contains in the analysis of the pooled OPERA I and OPERA II supplementary material, which is available to authorized users. 4 * Benjamin Turner Vall d’Hebron University Hospital, Barcelona, Spain Benjamin.Turner@bartsandthelondon.nhs.uk 5 Division of Neurology, St Michael’s Hospital, University 1 of Toronto, Toronto, Canada Department of Neurology, The Royal London Hospital, 6 London E1 1BB, UK Pitié-Salpêtrière Hospital, Paris, France 2 7 Weill Institute for Neurosciences, Department of Neurology, McGovern Medical School, The University of Texas Health University of California, San Francisco, San Francisco, CA, Science Center at Houston (UTHealth), Houston, TX, USA USA 8 F. Hoffmann-La Roche Ltd, Basel, Switzerland 3 University Hospital Basel, University of Basel, Basel, 9 Genentech, Inc., South San Francisco, CA, USA Switzerland 13 Vol:.(1234567890) Journal of Neurology (2019) 266:1182–1193 1183 intent-to-treat (ITT) population [2]. Similarly, treatment Prespecified subgroups were: study (OPERA I versus with ocrelizumab was associated with relatively lower rates OPERA II), age (< 40 versus ≥ 40 years), sex, BMI (< 25 kg/m2 of disability progression confirmed at 12 weeks (12W-CDP; versus ≥ 25 kg/m2), region (USA versus rest of world), baseline 40%) and at 24 weeks (24W-CDP; 40%) in the prespeci- EDSS score (< 4 versus ≥ 4), baseline gadolinium-enhancing fied pooled analysis of both studies. These data were further T1 lesion status (0 versus ≥ 1), and pre-treated patients with supported by a significantly greater suppression of develop- active disease (pre-treated for ≥ 1 year with either ≥ 1 relapse ment of new areas of inflammation (assessed by MRI of the in the year prior to randomization or ≥ 1 baseline T1 gadolin- brain with the use of gadolinium enhancement) and new or ium-enhancing lesion) or highly active disease (≥ 1 relapse in newly enlarged plaque formation (as measured by lesions on the year prior to randomization and ≥ 9 T2 lesions or ≥ 1 T1 T2-weighted MRI) [1]. gadolinium-enhancing lesion at baseline). Additional post hoc The current analyses were undertaken to understand if subgroups, included because of their clinical relevance, con- the treatment effects of ocrelizumab are consistent across sisted of: baseline EDSS score (< 2.5 versus ≥ 2.5 [for further subgroups of patients with different baseline characteris- evaluation of patients with lower disability at baseline]), prior tics; to describe the efficacy of ocrelizumab in patient sub- DMT use within the 2 years prior to study inclusion (yes versus groups relating to disability and clinical and MRI disease no), prior relapse (≤ 1 versus ≥ 2), and baseline normalized activity; and to describe the efficacy of ocrelizumab in both brain volume (≥ 1500 cm3 versus < 1500 cm3). treatment-naïve patients and those previously treated with disease-modifying therapy (DMT). Statistical analyses of endpoint by subgroup Subgroup-level treatment comparisons were used to deter- Methods mine whether the treatment effects of ocrelizumab and IFN β-1a were the same within each subgroup level. Treatment- Trial design and patients by-subgroup interaction tests were used to determine whether the treatment effect of ocrelizumab versus IFN The methodology of the identically designed phase III β-1a was the same between the two levels of subgroup. OPERA I (NCT01247324) and OPERA II (NCT01412333) p values < 0.05 from the treatment-by-subgroup interac- studies has been reported previously [1]. Patients (aged tion test indicate that the treatment effect of ocrelizumab 18–55 years, McDonald criteria [2010] diagnosis of MS versus IFN β-1a was not the same between the two levels [3], baseline Expanded Disability Status Scale [EDSS] score of subgroup. 0–5.5, ≥ 2 documented clinical relapses within the previous For ARR, both subgroup-level and treatment-by-subgroup 2 years or one clinical relapse within the year before screen- interactions testing were performed using a negative bino- ing) were randomized (1:1) to treatment with either ocreli- mial or quasi-Poisson model with the number of relapses zumab 600 mg administered by intravenous infusion every as the response variable and log-transformed exposure time 24 weeks, or subcutaneous IFN β-1a 44 µg three times per as the offset variable in both models. Factors included in week throughout the 96-week treatment period. All patients subgroup-level tests were treatment, study, region, and base- provided written informed consent. line EDSS score (< 4.0 versus ≥ 4.0); additional factors in treatment-by-subgroup interaction testing were subgroup Statistical analyses and treatment-by-subgroup interaction. Disability progression, with 12- or 24-week confirmation, Statistical approaches, including sample size calculations subgroup-level, and treatment-by-subgroup interactions test- and analyses of primary and secondary endpoints, have been ing were performed using Cox proportional hazard models described previously [1]. with time to onset of disability progression as the response Subgroup analyses, based on data from the 96-week variable and treatment (ocrelizumab versus IFN β-1a) as a double-blind treatment period, of the primary (ARR) and factor, and study, region and baseline EDSS score (< 4.0 secondary endpoints were assessed in the ITT population versus ≥ 4.0) as adjustments in both models; additional fac- (all randomized patients [ocrelizumab, n = 827; IFN β-1a, tors in the treatment-by-subgroup interaction testing were n = 829]); analyses of no evidence of disease activity (NEDA) subgroups and treatment-by-subgroup interaction. were conducted in the modified ITT (mITT) population For the MRI outcomes of T1 gadolinium-enhancing lesions (ocrelizumab, n = 761; IFN β-1a, n = 759), where patients and new/enlarging T2 lesions, subgroup-level and treatment- withdrawn from the trial for reasons other than efficacy fail- by-subgroup interactions testing were performed using a ure or death and who had no evidence of clinical disease negative binomial or quasi-Poisson model with the number of activity at the time of treatment discontinuation in the trial lesions as the response variable, the log-transformed number were excluded [1]. of MRI scans as the offset variable, and baseline lesion count, 13 1184 Journal of Neurology (2019) 266:1182–1193 treatment, study, region, and baseline EDSS score (< 4.0 ver- across the majority of subgroup-levels, including study, sus ≥ 4.0) as factors in both models; additional factors in the region, age, sex, baseline BMI, prior DMT use and prior treatment-by-subgroup interaction tests were subgroup and relapse, baseline EDSS, normalized brain volume, and gad- treatment-by-subgroup interaction. For change from baseline olinium-enhancing T1 lesions; a numerical trend favoring brain volume, subgroup and treatment-by-subgroup interaction ocrelizumab was observed for the specific subgroup strata testing used a mixed-effect model of repeated measures model of patients aged ≥ 40 years (Fig. 1). (unstructured covariance matrix) with percentage change in Significant treatment-by-subgroup interactions for ARR brain volume as the dependent variable and baseline brain were observed between patients aged < 40 versus ≥ 40 years volume, treatment, study, region, baseline EDSS score (< 4.0 (p = 0.006) and patients with or without baseline T1 gadolin- versus ≥ 4.0), week, baseline brain volume-by-week, and treat- ium-enhancing lesions (p < 0.001). Treatment-by-subgroup ment-by-week as factors in both models; additional factors in interaction p values for all endpoints and subgroups are pre- the treatment-by-subgroup interaction tests were subgroup and sented in Supplementary Table S2. treatment-by-week-by-subgroup. Subgroup-level testing of NEDA or NEDA 24–96 (NEDA Confirmed disability progression rebaselined at Week 24, which provides a representation of steady-state efficacy unconfounded by any initial disease The significant reductions in disease progression with activity carried over from baseline and recent pre-baseline 12-week confirmation (40% [p < 0.001]; Fig. 2) and 24-week disease state [4]) used the Cochran–Mantel–Haenszel test confirmation (40% [p = 0.003]; Supplementary Fig. S1), with treatment and NEDA status as the column/row factors relative to IFN β-1a, observed in the overall pooled anal- and study, region, and baseline EDSS score (< 4.0 versus ysis of the ITT population were maintained across most ≥ 4.0) as stratification factors. Treatment-by-subgroup inter- subgroup-levels. action used the Breslow–Day test with treatment/NEDA sta- For 12W-CDP, numerical trends favoring ocrelizumab tus as the column/row factors and subgroup as the stratifica- were observed in the subgroup strata of patients: from the tion factor. USA, who received prior DMT, with ≥ 2 prior relapses For subgroup-level analyses, key covariates (i.e., study, at baseline, with a baseline EDSS score < 2.5, or with a region, or baseline EDSS < 4.0 versus ≥ 4.0) were not baseline BMI ≥ 25 kg/m2 (Fig. 2). The pattern of numeri- included as a main effect if the key covariate was used as cal trends favoring ocrelizumab for 24W-CDP was the the subgroup. If the subgroup was EDSS < 2.5 versus ≥ 2.5, same as for 12W-CDP with the addition of trends for males then baseline EDSS < 4.0 versus ≥ 4.0 was not included as and for patients with a baseline normalized brain volume a main effect. < 1500 cm3 (Supplementary Fig. S1). Analyses of patients who were pre-treated and had active Significant treatment-by-subgroup interactions were or highly active disease were conducted in a similar way to observed between patients with a baseline BMI < 25 kg/ the subgroup-level analyses described above, with the excep- m2 versus ≥ 25 kg/m2 for both 12W-CDP (p = 0.026) and tion that no treatment-by-subgroup testing was conducted. 24W-CDP (p = 0.016; Fig. 2). Results MRI outcomes Patient disposition, demographic and disease characteris- tics, and safety findings from the individual OPERA I and The substantial reductions in both the average number of OPERA II studies were reported previously [1]. Baseline gadolinium-enhancing T1 lesions (94% [p < 0.001]; Fig. 3) demographic and disease characteristics between treatment and the average number of new or newly enlarged hyper- groups in the pooled ITT population were generally compa- intense T2 lesions per scan (80% [p < 0.001]; Fig. 4), of rable (Table 1), and characteristics within the mITT popu- ocrelizumab relative to IFN β-1a, observed in the overall lation were generally comparable to those within the ITT pooled analysis of the ITT population were maintained population (Supplementary Table S1). across all subgroup-levels (p < 0.001 for both endpoints and all comparisons). Subgroup analyses For T1 gadolinium-enhancing lesions, treatment-by-sub- group interactions were observed between patients aged < 40 Annualized relapse rate versus ≥ 40 years (p = 0.030) and patients with or without baseline T1 gadolinium-enhancing lesions (p = 0.001). With The significant reduction in ARR observed in the overall respect to new or newly enlarged hyperintense T2 lesions, pooled analysis of the ITT population with ocrelizumab, treatment-by-subgroup interactions were observed between relative to IFN β-1a (47% [p < 0.001]), was maintained patients with a baseline EDSS score < 4.0 versus ≥ 4.0 13 Journal of Neurology (2019) 266:1182–1193 1185 Table 1  Baseline demographic Characteristic IFN β-1a 44 µg Ocrelizumab and disease characteristics (N = 829) 600 mg of the pooled OPERA I and (N = 827) OPERA II intent-to-treat population Age Years, mean (SD) 37.2 (9.2) 37.1 (9.2) < 40 years, n (%) 484 (58.4) 496 (60.0) ≥ 40 years, n (%) 345 (41.6) 331 (40.0) Female, n (%) 552 (66.6) 541 (65.4) Body mass index kg/m2, mean (SD) 26.4 (6.2) 26.2 (5.8) < 25 kg/m2, n (%) 413 (50.2) 406 (49.6) ≥ 25 kg/m2, n (%) 409 (49.8) 412 (50.4) Time since MS symptom onset, years, mean (SD) 6.5 (6.1) 6.7 (6.2) Time since RMS diagnosis, years, mean (SD) 3.9 (4.9) 4.0 (4.9) No DMT in the 2 years before study inclusion, n (%) 606 (73.4)a,b 605 (73.3)a,b EDSS score Mean (SD) 2.8 (1.3)c 2.8 (1.3) < 2.5, n (%) 329 (39.7)c 310 (37.5) ≥ 2.5, n (%) 499 (60.3)c 517 (62.5) < 4.0, n (%) 627 (75.7)c 629 (76.1) ≥ 4.0, n (%) 201 (24.3)c 198 (23.9) Number of relapses, mean (SD) In the last year, mean (SD) 1.33 (0.69)d 1.32 (0.67)d In the last 2 years, mean (SD) 1.76 (0.92)d 1.79 (0.91)d ≤ 1 relapse, n (%) 584 (70.6)d 585 (70.8)d ≥ 2 relapses, n (%) 243 (29.4)d 241 (29.2)d MRI Patients with no T1 gadolinium-enhancing lesions, n (%) 495 (60.2)e 485 (59.3)e Patients with ≥ 1 T1 gadolinium-enhancing lesion, n (%) 327 (39.8)e 333 (40.7)e Number of T2 lesions, mean (SD) 51 (38)f 50 (39)f T2 lesion volume, ­cm3, mean (SD) 10.18 (11.8)f 10.79 (14.1)f Normalized brain volume, c­ m3, mean (SD) 1500 (89)g 1502 (88)g Normalized brain volume < 1500 cm3, n (%) 398 (48.7)g 402 (49.0)g Normalized brain volume ≥ 1500 cm3, n (%) 420 (51.3)g 418 (51.0)g DMT disease-modifying therapy, EDSS Expanded Disability Status Scale, IFN interferon, MRI magnetic resonance imaging, MS multiple sclerosis, OCR ocrelizumab, RMS relapsing multiple sclerosis, SD stand- ard deviation a Data include patients who were untreated with any DMT in the 2 years before screening b IFN β-1a n = 826, OCR n = 825 c IFN β-1a n = 828 d IFN β-1a n = 827, OCR n = 826 e IFN β-1a n = 822, OCR n = 818 f IFN β-1a n = 824, OCR n = 822 g IFN β-1a n = 818, OCR n = 820 (p = 0.007) and a baseline BMI < 25 kg/m2 versus ≥ 25 kg/m2 maintained across the majority of subgroup-levels; a (p = 0.043). numerical trend favoring ocrelizumab was observed in The benefit in change from baseline brain volume patients from the USA. A treatment-by-subgroup inter- observed in the overall pooled analysis of the ITT pop- action was observed between patients with a normal- ulation with ocrelizumab, relative to IFN β-1a (least ized baseline brain volume < 1500 cm 3 and ≥ 1500 cm3 square mean difference, 0.31% [p < 0.001]; Fig. 5), was (p = 0.002). 13 1186 Journal of Neurology (2019) 266:1182–1193 IFN β-1a (N=829) OCR (N=827) Group, Relapses, Patient Adjusted Group, Relapses, Patient Adjusted n n years ARR n n years ARR Favors OCR Favors IFN β-1a All patients 829 334 1339.16 0.29 827 194 1415.72 0.16 0.54 (95% CI: 0.44–0.66; p<0.001) Study OPERA I 411 166 678.12 0.29 410 96 706.26 0.16 0.54 (95% CI: 0.40–0.72; p<0.001) OPERA II 418 168 661.04 0.29 417 98 709.45 0.16 0.53 (95% CI: 0.40–0.71; p<0.001) Region ROW 610 254 1013.69 0.29 610 143 1058.03 0.15 0.52 (95% CI: 0.41–0.67; p<0.001) USA 219 80 325.47 0.29 217 51 357.69 0.17 0.58 (95% CI: 0.39–0.85; p=0.005) Age, yearsa <40 484 212 779.40 0.36 496 103 846.48 0.15 0.41 (95% CI: 0.31–0.55; p<0.001) ≥40 345 122 559.76 0.23 331 91 569.23 0.17 0.76 (95% CI: 0.56–1.03; p=0.073) Sex Female 552 233 887.84 0.30 541 130 920.62 0.16 0.53 (95% CI: 0.41–0.68; p<0.001) Male 277 101 451.33 0.25 286 64 495.09 0.14 0.56 (95% CI: 0.40–0.80; p=0.001) Baseline body mass index, kg/m2 <25 413 171 668.23 0.30 406 98 696.53 0.16 0.52 (95% CI: 0.39–0.70; p<0.001) ≥25 409 160 658.05 0.28 412 94 703.99 0.15 0.55 (95% CI: 0.41–0.73; p<0.001) Baseline normalized brain volume, cm3 <1500 398 168 645.88 0.28 402 103 688.08 0.16 0.58 (95% CI: 0.43–0.76; p<0.001) ≥1500 420 165 678.61 0.33 418 91 714.72 0.16 0.49 (95% CI: 0.36–0.67; p<0.001) Prior disease-modifying therapy use in the 2 years before study inclusion No 605 226 995.71 0.28 604 137 1034.43 0.16 0.57 (95% CI: 0.45–0.72; p<0.001) Yes 224 108 343.46 0.32 223 57 381.29 0.15 0.46 (95% CI: 0.31–0.69; p=0.001) Prior relapseb ≤1 584 203 942.02 0.24 585 122 1000.77 0.14 0.57 (95% CI: 0.45–0.71; p<0.001) ≥2 243 130 394.84 0.38 241 72 413.87 0.20 0.53 (95% CI: 0.40–0.71; p<0.001) Baseline Expanded Disability Status Scale scoreb <4 627 226 1023.33 0.22 629 118 1081.62 0.11 0.49 (95% CI: 0.40–0.62; p<0.001) ≥4 201 108 315.78 0.35 198 76 334.10 0.23 0.67 (95% CI: 0.50–0.89; p=0.006) <2.5 329 113 540.27 0.21 310 52 530.56 0.10 0.46 (95% CI: 0.31–0.69; p<0.001) ≥2.5 499 221 798.85 0.29 517 142 885.16 0.16 0.57 (95% CI: 0.45–0.73; p<0.001) Baseline T1 gadolinium-enhancing lesions, na 0 495 164 811.56 0.25 485 124 828.95 0.18 0.74 (95% CI: 0.56–0.96; p=0.025) ≥1 327 167 519.59 0.35 333 70 570.93 0.13 0.36 (95% CI: 0.26–0.50; p<0.001) 0.00 0.50 1.00 1.50 2.00 Rate ratio Fig. 1  Annualized relapse rate by subgroup in the pooled OPERA I subgroup interaction was observed for age (< 40 versus ≥ 40 years; and OPERA II intent-to-treat population. Subgroup-level and treat- p = 0.006) and for baseline T1 gadolinium-enhancing lesion status (0 ment-by-subgroup interactions were assessed by the negative bino- versus ≥ 1; p < 0.001); p values < 0.05 from the treatment-by-subgroup mial method, or quasi-Poisson model if appropriate (where indicated). interaction test indicate that the treatment effect of OCR versus IFN ARR​ annualized relapse rate, CI confidence interval, IFN interferon, is not the same at each subgroup-level. bEstimated by quasi-Poisson OCR ocrelizumab, ROW rest of world. aA significant treatment-by- model No evidence of disease activity population with ocrelizumab (72% [p < 0.001]), relative to IFN β-1a, was maintained across all subgroup-levels The increase in the proportion of patients with NEDA in ­( Supplementary Fig. S2). Treatment-by-subgroup inter- the overall pooled mITT population with ocrelizumab (75% actions were observed between patients with a baseline [p < 0.001]), relative to IFN β-1a, was maintained across EDSS score < 4.0 and ≥ 4.0 (p = 0.008), patients < 40 and most subgroup-levels (Fig. 6); a numerical trend favor- ≥ 40 years old (p < 0.001), and those with or without T1 ing ocrelizumab was observed in patients with a baseline gadolinium-enhancing lesions at baseline (p < 0.001). EDSS score ≥ 4.0. A treatment-by-subgroup interaction was observed between patients with a baseline EDSS score < 4.0 Efficacy in pre‑treated patients with active or highly and ≥ 4.0 (p = 0.022). active disease Similarly, when NEDA was rebaselined to Week 24, the improvement in the proportion of patients A total of 301 patients were defined as pre-treated with with NEDA 24–96 seen in the overall pooled mITT active disease (ocrelizumab, n = 153; IFN β-1a, n = 148) 13 Journal of Neurology (2019) 266:1182–1193 1187 IFN β-1a (N=829) OCR (N=827) Group, Events, Group, Events, n n n n Favors OCR Favors IFN β-1a All patients 829 113 827 75 0.60 (95% CI: 0.45–0.81; p<0.001) Study OPERA I 411 50 410 31 0.57 (95% CI: 0.37–0.90; p=0.015) OPERA II 418 63 417 44 0.63 (95% CI: 0.42–0.92; p=0.018) Region ROW 610 79 610 45 0.53 (95% CI: 0.36–0.76; p<0.001) USA 219 34 217 30 0.77 (95% CI: 0.47–1.26; p=0.3) Age, years <40 484 53 496 34 0.59 (95% CI: 0.39–0.92; p=0.019) ≥40 345 60 331 41 0.61 (95% CI: 0.41–0.92; p=0.018) Sex Female 552 70 541 48 0.64 (95% CI: 0.44–0.93; p=0.019) Male 277 43 286 27 0.56 (95% CI: 0.34–0.91; p=0.019) Baseline body mass index, kg/m2 a <25 413 54 406 23 0.39 (95% CI: 0.24–0.64; p<0.001) ≥25 409 59 412 51 0.81 (95% CI: 0.55–1.18; p=0.3) Baseline normalized brain volume, cm3 <1500 398 64 402 43 0.61 (95% CI: 0.41–0.90; p=0.013) ≥1500 420 47 418 32 0.63 (95% CI: 0.40–0.99; p=0.046) Prior disease-modifying therapy use in the 2 years before study inclusion No 605 82 604 53 0.60 (95% CI: 0.42–0.85; p=0.004) Yes 224 31 223 22 0.61 (95% CI: 0.35–1.06; p=0.080) Prior relapse ≤1 584 82 585 52 0.58 (95% CI: 0.41–0.82; p=0.002) ≥2 243 31 241 23 0.67 (95% CI: 0.39–1.16; p=0.2) Baseline Expanded Disability Status Scale score <4 627 87 629 64 0.66 (95% CI: 0.48–0.92; p=0.014) ≥4 201 26 198 11 0.39 (95% CI: 0.19–0.80; p=0.010) <2.5 329 44 310 37 0.82 (95% CI: 0.53–1.27; p=0.4) ≥2.5 499 69 517 38 0.48 (95% CI: 0.33–0.72; p<0.001) Baseline T1 gadolinium-enhancing lesions, n 0 495 65 485 43 0.60 (95% CI: 0.40–0.88; p=0.010) ≥1 327 47 333 32 0.60 (95% CI: 0.38–0.95; p=0.029) 0.00 0.50 1.00 1.50 2.00 Hazard ratio Fig. 2  Disease progression confirmed at Week 12 by subgroup in treatment-by-subgroup interaction was observed for BMI (< 25 kg/m2 the pooled OPERA I and OPERA II intent-to-treat population. Sub- versus ≥ 25 kg/m2; p = 0.026); p values < 0.05 from the treatment-by- group-level and treatment-by-subgroup interactions were assessed subgroup interaction test indicate that the treatment effect of OCR by the Cox proportional hazards method. CI confidence interval, versus IFN is not the same at each subgroup-level IFN interferon, OCR ocrelizumab, ROW rest of world. aA significant and 283 patients were defined as pre-treated with highly [p = 0.075] or a 50% relative reduction [p = 0.082], respec- active disease (ocrelizumab, n = 143; IFN β-1a, n = 140). tively) compared with a 40% reduction (p = 0.003) in the Consistent with the reduction in ARR in the overall ITT overall ITT pooled population; T1 gadolinium-enhancing pooled population (47%; p < 0.001), ocrelizumab, relative lesions (98% relative reduction [p < 0.001 in both sub- to IFN β-1a, reduced the ARR in patients who were pre- groups]) compared with a 94% reduction (p < 0.001) in treated with active disease by 65% (p < 0.001) and pre- the overall ITT pooled population; and new or enlarging treated with highly active disease by 68% (p < 0.001). T2 lesions (80% relative reduction [p < 0.001 in both sub- Similar benefits in patients who were pre-treated with groups]) compared with an 80% reduction (p < 0.001) in active disease or pre-treated with highly active disease were the overall ITT pooled population. observed for 12W-CDP (54% relative reduction [p = 0.032] For patients who were pre-treated with either active or a 53% relative reduction [p = 0.035], respectively) com- or highly active disease, the statistically significant treat- pared with a 40% reduction (p < 0.001) in the overall ITT ment benefits of ocrelizumab relative to IFN β-1a were pooled population; 24W-CDP (51% relative reduction 13 1188 Journal of Neurology (2019) 266:1182–1193 IFN β-1a (N=829) OCR (N=827) Group, Lesions, MRIs, Group, Lesions, MRIs, n n n n n n Favors OCR Favors IFN β-1a All patients 829 802 2081 827 42 2235 0.06 (95% CI: 0.04–0.08; p<0.001) Study OPERA I 411 337 1064 410 21 1118 0.06 (95% CI: 0.03–0.10; p<0.001) OPERA II 418 465 1017 417 21 1117 0.05 (95% CI: 0.03–0.09; p<0.001) Region ROW 610 474 1590 610 27 1686 0.06 (95% CI: 0.03–0.09; p<0.001) USA 219 328 491 217 15 549 0.05 (95% CI: 0.02–0.10; p<0.001) Age, yearsa,b <40 484 594 1208 496 24 1331 0.04 (95% CI: 0.03–0.06; p<0.001) ≥40 345 208 873 331 18 904 0.08 (95% CI: 0.05–0.13; p<0.001) Sex Female 552 558 1376 541 32 1445 0.07 (95% CI: 0.04–0.11; p<0.001) Male 277 244 705 286 10 790 0.04 (95% CI: 0.02–0.07; p<0.001) Baseline body mass index, kg/m2 <25 413 444 1044 406 23 1105 0.05 (95% CI: 0.03–0.09; p<0.001) ≥25 409 358 1016 412 19 1106 0.05 (95% CI: 0.03–0.08; p<0.001) Baseline normalized brain volume, cm3 <1500 398 388 1001 402 18 1090 0.04 (95% CI: 0.02–0.07; p<0.001) ≥1500 420 404 1059 418 24 1126 0.06 (95% CI: 0.03–0.10; p<0.001) Prior disease-modifying therapy use in the 2 years before study inclusionb No 605 544 1550 604 31 1632 0.05 (95% CI: 0.04–0.08; p<0.001) Yes 224 258 531 223 11 603 0.03 (95% CI: 0.02–0.06; p<0.001) Prior relapseb ≤1 584 508 1463 585 19 1575 0.04 (95% CI: 0.02–0.06; p<0.001) ≥2 243 293 615 241 23 659 0.06 (95% CI: 0.04–0.10; p<0.001) Baseline Expanded Disability Status Scale scoreb <4 627 673 1597 629 32 1707 0.04 (95% CI: 0.03–0.06; p<0.001) ≥4 201 129 484 198 10 528 0.08 (95% CI: 0.04–0.14; p<0.001) <2.5 329 315 843 310 20 841 0.07 (95% CI: 0.03–0.12; p<0.001) ≥2.5 499 487 1238 517 22 1394 0.05 (95% CI: 0.03–0.08; p<0.001) Baseline T1 gadolinium-enhancing lesionsa 0 495 173 1274 485 20 1310 0.11 (95% CI: 0.06–0.21; p<0.001) ≥1 327 626 801 333 22 905 0.03 (95% CI: 0.02–0.05; p<0.001) 0.00 0.50 1.00 1.50 2.00 Rate ratio Fig. 3  T1 gadolinium-enhancing lesions by subgroup in the pooled treatment-by-subgroup interaction was observed for age (< 40 versus OPERA I and OPERA II intent-to-treat population. Subgroup-level ≥ 40 years; p = 0.030) and for baseline T1 gadolinium-enhancing and treatment-by-subgroup interactions were assessed by the nega- lesion status (0 versus ≥ 1; p = 0.001); p values < 0.05 from the treat- tive binomial method, or quasi-Poisson model if appropriate (where ment-by-subgroup interaction test indicate that the treatment effect of indicated). CI confidence interval, IFN interferon, MRI magnetic res- OCR versus IFN is not the same at each subgroup-level. bEstimated onance imaging, OCR ocrelizumab, ROW rest of world. aA significant by quasi-Poisson model demonstrated for all endpoints except for 24W-CDP, which benefits of ocrelizumab, relative to IFN β-1a, are maintained comprised a relatively smaller number of patients. across most baseline-, clinically, and radiologically defined subgroups, although the magnitude of the treatment benefit of ocrelizumab over IFN β-1a may be greater in one of the two Discussion levels of subgroup. Treatment-by-subgroup interactions were observed for some endpoints over particular subgroups; how- In the OPERA I and OPERA II studies, ocrelizumab, rela- ever, all showed a benefit (significant or numerical) of ocre- tive to IFN β-1a, was associated with significantly better out- lizumab over IFN β-1a at each of the two levels of subgroup. comes in terms of ARR, disability progression, suppression OPERA I and OPERA II were not powered to detect of new inflammatory lesions in the brain (T1 gadolinium- statistically significant effects for all subgroup analyses, enhancing lesions), and new or newly enlarged plaque forma- including those predefined and with pre-planned pool- tion (new/enlarging T2 lesions) [1]. The subgroup analyses ing (12W-CDP or 24W-CDP), or to test for heterogene- presented here demonstrate that the significant treatment ity between subgroups. Additionally, post hoc analyses of 13 Journal of Neurology (2019) 266:1182–1193 1189 IFN β-1a (N=829) OCR (N=827) Group, Lesions, MRIs, Group, Lesions, MRIs, n n n n n n Favors OCR Favors IFN β-1a All patients 829 4019 2091 827 810 2246 0.20 (95% CI: 0.16–0.24; p<0.001) Study OPERA I 411 1916 1066 410 430 1123 0.23 (95% CI: 0.17–0.30; p<0.001) OPERA II 418 2103 1025 417 380 1123 0.17 (95% CI: 0.13–0.23; p<0.001) Region ROW 610 2821 1595 610 595 1690 0.20 (95% CI: 0.16–0.26; p<0.001) USA 219 1198 496 217 215 556 0.19 (95% CI: 0.13–0.27; p<0.001) Age, years <40 484 3064 1216 496 615 1340 0.18 (95% CI: 0.14–0.23; p<0.001) ≥40 345 955 875 331 195 906 0.23 (95% CI: 0.16–0.33; p<0.001) Sex Female 552 2281 1384 541 484 1457 0.20 (95% CI: 0.16–0.26; p<0.001) Male 277 1738 707 286 326 789 0.18 (95% CI: 0.13–0.25; p<0.001) Baseline body mass index, kg/m2 a <25 413 2181 1047 406 514 1107 0.22 (95% CI: 0.17–0.30; p<0.001) ≥25 409 1825 1023 412 290 1115 0.15 (95% CI: 0.12–0.20; p<0.001) Baseline normalized brain volume, cm3 <1500 398 1855 1004 402 317 1095 0.18 (95% CI: 0.14–0.24; p<0.001) ≥1500 420 2125 1066 418 492 1133 0.20 (95% CI: 0.15–0.25; p<0.001) Prior disease-modifying therapy use in the 2 years before study inclusion No 605 2709 1557 604 567 1641 0.21 (95% CI: 0.17–0.27; p<0.001) Yes 224 1310 534 223 243 605 0.16 (95% CI: 0.11–0.23; p<0.001) Prior relapse ≤1 584 2636 1469 585 532 1584 0.19 (95% CI: 0.15–0.25; p<0.001) ≥2 243 1381 619 241 278 661 0.20 (95% CI: 0.14–0.28; p<0.001) Baseline Expanded Disability Status Scale scorea <4 627 3225 1604 629 546 1719 0.17 (95% CI: 0.14–0.21; p<0.001) ≥4 201 794 487 198 264 527 0.32 (95% CI: 0.21–0.50; p<0.001) <2.5 329 1738 849 310 285 844 0.17 (95% CI: 0.12–0.22; p<0.001) ≥2.5 499 2281 1242 517 525 1402 0.22 (95% CI: 0.17–0.28; p<0.001) Baseline T1 gadolinium-enhancing lesions, n 0 495 945 1277 485 195 1313 0.21 (95% CI: 0.16–0.29; p<0.001) ≥1 327 3070 804 333 612 910 0.17 (95% CI: 0.14–0.22; p<0.001) 0.00 0.50 1.00 1.50 2.00 Rate ratio Fig. 4  New/enlarging T2 lesions by subgroup in the pooled OPERA I ROW rest of world. aA significant treatment-by-subgroup interaction and OPERA II intent-to-treat population. Subgroup-level and treat- was observed for baseline BMI (< 25 versus ≥ 25 kg/m2; p = 0.043) and ment-by-subgroup interactions were assessed by the negative bino- baseline EDSS score (< 4.0 versus ≥ 4.0; p = 0.007); p values < 0.05 mial method, or quasi-Poisson model if appropriate (where indicated). from the treatment-by-subgroup interaction test indicate that the treat- CI confidence interval, EDSS Expanded Disability Status Scale, ment effect of OCR versus IFN is not the same at each subgroup-level IFN interferon, MRI magnetic resonance imaging, OCR ocrelizumab, subgroups can be difficult to interpret. Within this analysis For ARR, gadolinium-enhancing T1 lesions, and NEDA most outcomes and subgroups were predefined; only NEDA rebaselined at Week 24, significant subgroup interactions sug- 24–96 and brain volume change from baseline for outcomes, gested that patients who were younger and those with indica- and the subgroups baseline EDSS score < 2.5 versus ≥ 2.5, tors of active inflammatory disease appeared to gain a greater prior DMT use yes versus no, those with ≤ 1 or ≥ 2 prior treatment benefit with ocrelizumab, relative to IFN β-1a, than relapses, and those with a baseline normalized brain volume patients who were older and those without markers of active of < 1500 cm3 or ≥ 1500 cm3, were post hoc considerations. inflammatory disease. These findings are consistent with treat- Consequently, given the relatively smaller numbers of ment benefit of other DMTs, which has previously been sug- subjects in some of the subgroups, findings in these sub- gested to be higher in younger than in older patients on the groups should be interpreted with caution. Further studies basis of higher disease activity in younger populations with with an expanded data set with improved assessment tech- MS [5]. Subgroup analyses performed for other DMTs using niques would be required to address these points. either placebo or an active comparator, e.g., fingolimod [6, 7], cladribine [8], teriflunomide [9], dimethyl fumarate [10–12], 13 1190 Journal of Neurology (2019) 266:1182–1193 IFN β-1a (N=829) OCR (N=827) Change from Change from Group, Group, baseline, baseline, n n Favors IFN β-1a Favors OCR LS mean (%) LS mean (%) All patients 707 –1.29 827 –0.99 0.31 (95% CI: 0.20–0.42; p<0.001) Study OPERA I 353 –1.23 410 –0.94 0.29 (95% CI: 0.15–0.43; p<0.001) OPERA II 354 –1.35 417 –1.03 0.32 (95% CI: 0.16–0.49; p<0.001) Region ROW 529 –1.34 610 –0.98 0.36 (95% CI: 0.23–0.49; p<0.001) USA 178 –1.17 217 –1.02 0.15 (95% CI: –0.06–0.37; p=0.2) Age, years <40 413 –1.31 496 –1.09 0.22 (95% CI: 0.08–0.37; p=0.003) ≥40 294 –1.29 331 –0.85 0.44 (95% CI: 0.28–0.60; p<0.001) Sex Female 471 –1.34 541 –0.97 0.37 (95% CI: 0.23–0.50; p<0.001) Male 236 –1.21 286 –1.01 0.20 (95% CI: 0.01–0.39; p=0.044) Baseline body mass index, kg/m2 <25 360 –1.31 406 –1.01 0.30 (95% CI: 0.14–0.47; p<0.001) ≥25 340 –1.27 412 –0.95 0.32 (95% CI: 0.18–0.47; p<0.001) Baseline normalized brain volume, cm3 a <1500 338 –1.54 402 –1.06 0.48 (95% CI: 0.31–0.65; p<0.001) ≥1500 369 –1.08 418 –0.94 0.14 (95% CI: 0.01–0.28; p=0.042) Prior disease-modifying therapy use in the 2 years before study inclusion No 521 –1.33 604 –1.02 0.31 (95% CI: 0.18–0.43; p<0.001) Yes 186 –1.19 223 –0.90 0.29 (95% CI: 0.08–0.50; p=0.007) Prior relapse ≤1 494 –1.27 585 –0.97 0.30 (95% CI: 0.17–0.43; p<0.001) ≥2 212 –1.37 241 –1.04 0.33 (95% CI: 0.13–0.53; p=0.002) Baseline Expanded Disability Status Scale score <4 539 –1.24 629 –0.94 0.30 (95% CI: 0.18–0.42; p<0.001) ≥4 168 –1.43 198 –1.09 0.34 (95% CI: 0.07–0.61; p=0.013) <2.5 287 –1.11 310 –0.92 0.19 (95% CI: 0.04–0.34; p=0.014) ≥2.5 420 –1.39 517 –1.01 0.38 (95% CI: 0.23–0.53; p<0.001) Baseline T1 gadolinium-enhancing lesions, n 0 427 –1.06 485 –0.74 0.32 (95% CI: 0.21–0.44; p<0.001) ≥1 278 –1.64 333 –1.34 0.31 (95% CI: 0.12–0.50; p=0.002) –0.2 0.0 0.2 0.4 0.6 0.8 1.0 Least square mean difference, % Fig. 5  Change from baseline brain volume (%) by subgroup in the world. aA significant treatment-by-subgroup interaction was observed pooled OPERA I and OPERA II intent-to-treat population. Subgroup- for normalized baseline brain volume (< 1500 cm3 and ≥ 1500 cm3; level and treatment-by-subgroup interactions were assessed by the p = 0.002); p values < 0.05 from the treatment-by-subgroup interac- mixed-effect model of repeated measures model. CI confidence inter- tion test indicate that the treatment effect of OCR versus IFN is not val, IFN interferon, LS least square, OCR ocrelizumab, ROW rest of the same at each subgroup-level alemtuzumab [13], and natalizumab [14, 15], also found that ocrelizumab’s treatment benefit on brain atrophy was efficacy was highest in younger patients with more active greater in patients with lower baseline brain volume, disease. which may be related to differential rates of atrophy in Consistent effects across endpoints were seen for other the patient subgroups defined by brain volume thresh- subgroups of interest. There were no significant sex- olds. For NEDA (Weeks 0–96) and new or enlarging T2 based subgroup interactions and the same positive trends lesion endpoints, the treatment benefit of ocrelizumab was of treatment benefit with ocrelizumab were reported in greater in patients with lower measures of disability at female and male patients. The treatment benefit of ocreli- baseline. For both 12W-CDP and 24W-CDP, significant zumab, relative to IFN β-1a, was comparable across end- treatment-by-subgroup interactions for BMI indicate that points in patients with active or highly active disease who the magnitude of ocrelizumab treatment benefit is greater were pre-treated with DMT, while prior treatment per se in lighter patients versus heavier, consistent with lower did not impact the magnitude of the beneficial effect of ocrelizumab exposure in heavier patients (BMI > 25 kg/m2 ocrelizumab. Across other endpoints, the magnitude of versus < 25 kg/m 2). Analyses have shown that patients 13 Journal of Neurology (2019) 266:1182–1193 1191 IFN β-1a (N=759) OCR (N=761) Group, NEDA, Group, NEDA, n n (%) n n (%) Favors IFN β-1a Favors OCR All patients 759 206 (27.1) 761 363 (47.7) 1.75 (95% CI: 1.53–2.01; p<0.001) Study OPERA I 384 112 (29.2) 382 183 (47.9) 1.64 (95% CI: 1.36–1.98; p<0.001) OPERA II 375 94 (25.1) 379 180 (47.5) 1.89 (95% CI: 1.54–2.32; p<0.001) Region ROW 567 163 (28.7) 570 291 (51.1) 1.78 (95% CI: 1.53–2.07; p<0.001) USA 192 43 (22.4) 191 72 (37.7) 1.66 (95% CI: 1.20–2.30; p=0.002) Age, years <40 447 101 (22.6) 454 201 (44.3) 1.99 (95% CI: 1.63–2.42; p<0.001) ≥40 312 105 (33.7) 307 162 (52.8) 1.55 (95% CI: 1.29–1.87; p<0.001) Sex Female 507 150 (29.6) 493 245 (49.7) 1.66 (95% CI: 1.41–1.95; p<0.001) Male 252 56 (22.2) 268 118 (44.0) 2.05 (95% CI: 1.57–2.67; p<0.001) Baseline body mass index, kg/m2 <25 382 106 (27.7) 374 176 (47.1) 1.70 (95% CI: 1.41–2.06; p<0.001) ≥25 370 98 (26.5) 379 183 (48.3) 1.82 (95% CI: 1.49–2.23; p<0.001) Baseline normalized brain volume, cm3 <1500 364 92 (25.3) 371 172 (46.4) 1.84 (95% CI: 1.49–2.26; p<0.001) ≥1500 385 112 (29.1) 383 185 (48.3) 1.67 (95% CI: 1.39–2.01; p<0.001) Prior disease-modifying therapy use in the 2 years before study inclusion No 558 158 (28.3) 553 274 (49.5) 1.75 (95% CI: 1.50–2.05; p<0.001) Yes 201 48 (23.9) 208 89 (42.8) 1.79 (95% CI: 1.34–2.41; p<0.001) Prior relapse ≤1 531 155 (29.2) 537 264 (49.2) 1.68 (95% CI: 1.43–1.96; p<0.001) ≥2 226 51 (22.6) 224 99 (44.2) 2.00 (95% CI: 1.51–2.65; p<0.001) Baseline Expanded Disability Status Scale scorea <4 579 153 (26.4) 579 291 (50.3) 1.91 (95% CI: 1.63–2.24; p<0.001) ≥4 180 53 (29.4) 182 72 (39.6) 1.31 (95% CI: 0.98–1.75; p=0.064) <2.5 306 84 (27.5) 283 143 (50.5) 1.85 (95% CI: 1.49–2.29; p<0.001) ≥2.5 453 122 (26.9) 478 220 (46.0) 1.70 (95% CI: 1.42–2.04; p<0.001) Baseline T1 gadolinium-enhancing lesions, n 0 449 174 (38.8) 441 263 (59.6) 1.55 (95% CI: 1.35–1.78; p<0.001) ≥1 303 31 (10.2) 312 94 (30.1) 2.85 (95% CI: 1.96–4.13; p<0.001) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Rate ratio Fig. 6  Subgroup analyses of the proportion of patients with NEDA disease activity, OCR ocrelizumab, ROW rest of world. aA significant in the pooled OPERA I and OPERA II mITT population. Subgroup- treatment-by-subgroup interaction was observed for baseline EDSS level testing used the Cochran–Mantel–Haenszel test and treatment- score (< 4.0 versus ≥ 4.0; p = 0.022); p values < 0.05 from the treat- by-subgroup interactions were assessed by the Breslow–Day test. ment-by-subgroup interaction test indicate that the treatment effect of CI confidence interval, EDSS Expanded Disability Status Scale, OCR versus IFN is not the same at each subgroup-level IFN interferon, mITT modified intent-to-treat, NEDA no evidence of with a higher ocrelizumab exposure had a greater benefit interaction between age and disability progression, with on 12W-CDP and 24W-CDP. Therefore, it might be pos- lower levels of progression in younger patients; this was sible that in the BMI subgroups the ocrelizumab expo- not observed in the current study. sure had an influence on confirmed disability progression In conclusion, the treatment benefit of ocrelizumab ver- results. However, a contradictory significant treatment-by- sus IFN β-1a, measured by clinical and MRI outcomes, subgroup interaction between BMI strata was observed was maintained across almost all the subgroups and strata (p = 0.043) for the endpoint of new or enlarging T2 lesion of interest, and the pattern of treatment benefit across endpoints, indicating a higher magnitude of ocrelizumab all subgroups was consistent with that from the pooled treatment benefit in heavier versus lighter patients; how- OPERA studies. The results show that RMS patients with ever, this finding requires further study before clinical a wide range of demographic and clinical characteristics meaningfulness can be confirmed. Subgroup analyses benefit more from treatment with ocrelizumab than from of other studies with MS DMTs have demonstrated an IFN β-1a. 13 1192 Journal of Neurology (2019) 266:1182–1193 Data sharing statement Corporation. R. Buffels is an employee of F. Hoffmann-La Roche Ltd. H. Garren is an employee of Genentech, Inc., and a shareholder of F. Hoffmann-La Roche Ltd. D. Fiore is an employee of Genentech, Inc., Qualified researchers may request access to individual and a shareholder of F. Hoffmann-La Roche Ltd. J. Han is an employee patient-level data through the clinical study data request of Genentech, Inc. S.L. Hauser serves on the board of trustees for Neu- platform (http://www.clinic​ alstu​ dydat​ arequ​ est.com). Further rona and on scientific advisory boards for Annexon, Symbiotix, Bi- onure, Alector, and Molecular Stethoscope; he has also received travel details on Roche’s criteria for eligible studies are available reimbursement from F. Hoffmann-La Roche Ltd for CD20-related here: https​://clini​calst​udyda​tareq​uest.com/Study​-Spons​ors/ meetings and presentations. Study​-Spons​ors-Roche​.aspx. For further details on Roche’s Global Policy on the Sharing of Clinical Information and Ethical standards This trial was conducted in accordance with the International Conference on Harmonisation guidelines for Good Clini- how to request access to related clinical study documents, cal Practice and the Declaration of Helsinki. see: https​://www.roche​.com/resea​rch_and_devel​opmen​t/ who_we_are_how_we_work/clini ​ c al_trial ​ s /our_commi​ Informed consent All patients provided written informed consent. tment​_to_data_shari​ng.htm. Open Access This article is distributed under the terms of the Crea- Acknowledgements The authors thank all patients, their families, and tive Commons Attribution 4.0 International License (http://creat​iveco​ the investigators who participated in the OPERA I and OPERA II stud- mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribu- ies. We are also grateful to Gisèle von Büren (of F. Hoffmann-La Roche tion, and reproduction in any medium, provided you give appropriate Ltd) for additional critical review of this manuscript. F. Hoffmann-La credit to the original author(s) and the source, provide a link to the Roche Ltd provided financial support for these studies. Writing and Creative Commons license, and indicate if changes were made. editorial assistance was provided by Andrew Stead of Articulate Sci- ence Ltd. and was funded by F. Hoffmann-La Roche Ltd. Compliance with ethical standards References Conflicts of interest B. Turner has received honoraria, travel grants 1. Hauser SL, Bar-Or A, Comi G et al (2017) Ocrelizumab versus and been a member of advisory boards for Biogen, Merck Serono, interferon beta-1a in relapsing multiple sclerosis. N Engl J Med Novartis, Sanofi Genzyme, and Roche. C. Papeix has received speak- 376:221–234 ing honoraria and travel expense reimbursement for participation in 2. Papeix C, Cree B, Turner B et al (2017) Subgroup analyses of scientific meetings, and has participated in advisory boards in the past annualised relapse rates in patients with relapsing multiple sclero- years with Biogen, MedDay, Merck Serono, Novartis, Roche, and Sa- sis who received ocrelizumab or interferon beta-1a in the phase III nofi Genzyme. B. Cree has received personal compensation for con- OPERA I and OPERA II studies. In: 7th Joint European Commit- sulting from AbbVie, Akili, Biogen, EMD Serono, GeNeuro, Novartis, tee for Treatment and Research in Multiple Sclerosis (ECTRIMS) and Sanofi Genzyme. L. Kappos’s institution, the University Hospital and Americas Committee for Treatment and Research in Multi- Basel, has received research support and payments that were used ex- ple Sclerosis (ACTRIMS) Meeting; 25–28 October 2017; Paris, clusively for research support for Prof. Kappos’s activities as principal France. Poster P687 investigator and member or chair of planning and steering committees 3. Polman CH, Reingold SC, Banwell B et al (2011) Diagnostic or advisory boards in trials sponsored by Actelion, Addex, Almirall, criteria for multiple sclerosis: 2010 revisions to the McDonald Bayer HealthCare Pharmaceuticals, CSL Behring, Genentech, Inc., criteria. Ann Neurol 69:292–302 GeNeuro SA, Genzyme, Merck Serono, Mitsubishi Pharma, Novartis, 4. Giovannoni G, Turner B, Gnanapavan S, Offiah C, Schmierer K, Octapharma, Ono Pharma, Pfizer, Receptos, F. Hoffmann-La Roche Marta M (2015) Is it time to target no evident disease activity Ltd, Sanofi, Santhera, Siemens, Teva, UCB, and XenoPort; license (NEDA) in multiple sclerosis? Mult Scler Relat Disord 4:329–333 fees for Neurostatus products; and research grants from the Swiss MS 5. Tremlett H, Paty D, Devonshire V (2006) Disability progression Society, the Swiss National Research Foundation, the European Un- in multiple sclerosis is slower than previously reported. Neurology ion, the Gianni Rubatto Foundation, the Novartis Research Founda- 66:172–177 tion and the Roche Research Foundation. X. Montalban has received 6. Devonshire V, Havrdova E, Radue EW, O’Connor P, Zhang- speaking honoraria and travel expense reimbursement for participa- Auberson L, Agoropoulou C, Häring DA, Francis G, Kappos L, tion in scientific meetings, has been a steering committee member of FREEDOMS Study Group (2012) Relapse and disability out- clinical trials or participated in advisory boards of clinical trials in the comes in patients with multiple sclerosis treated with fingolimod: past years with Actelion, Almirall, Bayer, Biogen, Genzyme, Merck, subgroup analyses of the double-blind, randomised, placebo-con- Novartis, Octapharma, Receptos, F. Hoffmann-La Roche Ltd, Sanofi, trolled FREEDOMS study. Lancet Neurol 11:420–428 Teva, and Trophos. J.S. Wolinsky has served on advisory boards, 7. Derfuss T, Ontaneda D, Nicholas J, Meng X, Hawker K (2016) data monitoring or steering committees, has consulting agreements, Relapse rates in patients with multiple sclerosis treated with fin- or received speaker honoraria from the following entities: AbbVie, golimod: subgroup analyses of pooled data from three phase 3 AcademicCME, Actelion, ACTRIMS, Alkermes, Bayer HealthCare, trials. Mult Scler Relat Disord 8:124–130 Biogen, Bionest, Celgene, Clene Nanomedicine, CMSC, ECTRIMS, 8. Giovannoni G, Cook S, Rammohan K, Rieckmann P, Sørensen EMS Serono, Forward Pharma A/S, France Foundation, GeNeuro, PS, Vermersch P, Hamlett A, Viglietta V, Greenberg S, CLAR- MA’s Communications, Masters MS, MedDay Pharmaceuticals, Mc- ITY Study Group (2011) Sustained disease-activity-free status in Donnell Boehnen Hulbert & Berghoff, Novartis Pharmaceuticals, Ot- patients with relapsing–remitting multiple sclerosis treated with suka, PlatformQ Health Education, PRIME, PTC Therapeutics, Roche cladribine tablets in the CLARITY study: a post-hoc and subgroup Genentech, Sanofi Genzyme, Strategic Consultants International, analysis. Lancet Neurol 10:329–337 Takeda, Teva Pharmaceuticals, and WebMD; royalties are received for out-licensed monoclonal antibodies through UTHealth from Millipore 13 Journal of Neurology (2019) 266:1182–1193 1193 9. Miller AE, O’Connor P, Wolinsky JS et al (2012) Pre-speci- 13. Coles AJ, Twyman CL, Arnold DL et al (2012) Alemtuzumab fied subgroup analyses of a placebo-controlled phase III trial for patients with relapsing multiple sclerosis after disease-mod- (TEMSO) of oral teriflunomide in relapsing multiple sclerosis. ifying therapy: a randomised controlled phase 3 trial. Lancet Mult Scler 18:1625–1632 380:1829–1839 10. Bar-Or A, Gold R, Kappos L, Arnold DL, Giovannoni G, Selmaj 14. Hutchinson M, Kappos L, Calabresi PA et al (2009) The effi- K, O’Gorman J, Stephan M, Dawson KT (2013) Clinical efficacy cacy of natalizumab in patients with relapsing multiple sclero- of BG-12 (dimethyl fumarate) in patients with relapsing-remitting sis: subgroup analyses of AFFIRM and SENTINEL. J Neurol multiple sclerosis: subgroup analyses of the DEFINE study. J Neu- 256:405–415 rol 260:2297–2305 15. van Pesch V, Bartholomé E, Bissay V et al (2014) Safety and 11. Hutchinson M, Fox RJ, Miller DH et al (2013) Clinical efficacy of efficacy of natalizumab in Belgian multiple sclerosis patients: sub- BG-12 (dimethyl fumarate) in patients with relapsing–remitting group analysis of the natalizumab observational program. Acta multiple sclerosis: subgroup analyses of the CONFIRM study. J Neurol Belg 114:167–178 Neurol 260:2286–2296 12. Kappos L, Gold R, Miller DH et al (2012) Effect of BG-12 on contrast-enhanced lesions in patients with relapsing–remitting multiple sclerosis: subgroup analyses from the phase 2b study. Mult Scler 18:314–321 13