ARTICLE OPEN ACCESS CLASS OF EVIDENCE Five years of ocrelizumab in relapsing multiple sclerosis OPERA studies open-label extension Stephen L. Hauser, MD, Ludwig Kappos, MD, Douglas L. Arnold, MD, Amit Bar-Or, MD, Bruno Brochet, MD, Correspondence Robert T. Naismith, MD, Anthony Traboulsee, MD, Jerry S. Wolinsky, MD, Shibeshih Belachew, MD, PhD, Dr. Koendgen Harold Koendgen, MD, PhD, Victoria Levesque, PhD, Marianna Manfrini, MD, Fabian Model, PhD, harold.koendgen@ Stanislas Hubeaux, MSc, Lahar Mehta, MD, and Xavier Montalban, MD, PhD roche.com ® Neurology 2020;95:e1854-e1867. doi:10.1212/WNL.0000000000010376 Abstract MORE ONLINE Objective Class of Evidence To assess over 3 years of follow-up the effects of maintaining or switching to ocrelizumab (OCR) therapy on Criteria for rating clinical and MRI outcomes and safety measures in the open-label extension (OLE) phase of the pooled therapeutic and diagnostic OPERA: I/II studies in relapsing multiple sclerosis. studies NPub.org/coe Methods After 2 years of double-blind, controlled treatment, patients continued OCR (600 mg infusions every 24 weeks) or switched from interferon (IFN)-β-1a (44 μg 3 times weekly) to OCR when entering the OLE phase CME Course (3 years). Adjusted annualized relapse rate, time to onset of 24-week confirmed disability progression (CDP)/ NPub.org/cmelist improvement (CDP), brain MRI activity (gadolinium-enhanced and new/enlarging T2 lesions), and per- centage brain volume change were analyzed. Results Of patients entering the OLE phase, 88.6% completed year 5. The cumulative proportion with 24-week CDP was lower in patients who initiated OCR earlier vs patients initially receiving IFN-β-1a (16.1% vs 21.3% at year 5; p = 0.014). Patients continuing OCR maintained and those switching from IFN-β-1a to OCR attained near complete and sustained suppression of new brain MRI lesion activity from years 3–5. Over the OLE phase, patients continuing OCR exhibited less whole brain volume loss from double-blind study baseline vs those switching from IFN-β-1a (−1.87% vs −2.15% at year 5; p < 0.01). Adverse events were consistent with past reports and no new safety signals emerged with prolonged treatment. Conclusion Compared with patients switching from IFN-β-1a, earlier and continuous OCR treatment up to 5 years provided sustained benefit on clinical and MRI measures of disease progression. Classification of evidence This study provides Class III evidence that earlier and continuous treatment with OCR provided sustained benefit on clinical and MRI outcomes of disease activity and progression compared with patients switching from IFN- β-1a. The study is rated Class III because of the initial treatment randomization disclosure that occurred after inclusion in OLE. Clinical trial identifiers NCT01247324/NCT01412333. From the Department of Neurology (S.L.H.), University of California, San Francisco; Neurologic Clinic and Policlinic (L.K.), Departments of Medicine, Clinical Research, Biomedicine and Biomedical Engineering, University Hospital Basel, University of Basel, Switzerland; NeuroRx Research (D.L.A.); Departments of Neurology and Neurosurgery (D.L.A.), McGill University, Montreal, Canada; Department of Neurology and Center for Neuroinflammation and Experimental Therapeutics (A.B.-O.), University of Pennsylvania, Philadelphia; Department of Neurology (B.B.), CHU de Bordeaux, France; Department of Neurology (R.T.N.), Washington University School of Medicine, St. Louis, MO; Division of Neurology (A.T.), Department of Medicine, University of British Columbia, Vancouver, Canada; Department of Neurology (J.S.W.), McGovern Medical School, University of Texas Health Science Center at Houston (UTHealth); F. Hoffmann–La Roche Ltd (S.B., H.K., M.M., F.M., S.H.), Basel, Switzerland; Genentech, Inc. (V.L., L.M.), South San Francisco, CA; Division of Neurology (X.M.), University of Toronto, Canada; and Department of Neurology–Neuroimmunology (X.M.), Vall d’Hebron University Hospital, Barcelona, Spain. During completion of the work related to this article, S.B. and L.M. were employees of F. Hoffmann–La Roche Ltd; current affiliations are Biogen (S.B.), Cambridge, MA; and Alder Biopharmaceuticals Inc. (L.M.), Bothell, WA. Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article. The Article Processing Charge was funded by F. Hoffmann–La Roche Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal. e1854 Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. Glossary AE = adverse event; ARR = annualized relapse rate; CDI = confirmed disability improvement; CDP = confirmed disability progression; CI = confidence interval; DBP = double-blind controlled treatment phase; DMT = disease-modifying therapy; EDSS = Expanded Disability Status Scale; FSS = Functional Systems Score; Gd = gadolinium; GMV = gray matter volume; HR = hazard ratio; IFN = interferon; Ig = immunoglobulin; ITT = intention-to-treat; LLN = lower limit of normal; MMRM = mixed-effect model of repeated measures; MS = multiple sclerosis; NEDA = no evidence of disease activity; OCR = ocrelizumab; OLE = open-label extension; OPERA = A Study of Ocrelizumab in Comparison With Interferon Beta-1a (Rebif) in Participants With Relapsing Multiple Sclerosis; PML = progressive multifocal leukoencephalopathy; RMS = relapsing multiple sclerosis; UTI = urinary tract infection; WBV = whole brain volume; WMV = white matter volume. Ocrelizumab (OCR) is a humanized monoclonal antibody patients meeting specific criteria (data available from Dryad, that selectively targets CD20+ B cells while preserving the supplemental data, doi.org/10.5061/dryad.stqjq2bzw) were capacity for B-cell reconstitution and preexisting humoral eligible to enter an OLE phase to evaluate the long-term immunity.1,2 In phase III randomized, double-blind, double- safety, tolerability, and efficacy of OCR in RMS; patients who dummy trials with identical designs (OPERA [A Study of declined or who were not eligible (since there was an OLE Ocrelizumab in Comparison With Interferon Beta-1a (Rebif) screening phase) to participate in the OLE phase entered in Participants With Relapsing Multiple Sclerosis] I safety follow-up. Patients entering the OLE phase first entered [NCT01247324] and OPERA II [NCT01412333]) in re- the OLE screening phase, which lasted up to 4 weeks. At the lapsing multiple sclerosis (RMS), OCR demonstrated supe- start of the OLE phase, patients who received OCR in the rior efficacy for preventing relapses and disability worsening, DBP continued OCR and patients from the IFN-β-1a group and a higher rate of disability improvement, compared with were switched to OCR, given every 24 weeks; in order to subcutaneous interferon (IFN)-β-1a given 3 times weekly.3 maintain blinding and in accordance with the start of the DBP, Significant benefits of OCR on MRI outcomes, including the all patients received the first dose of OCR as 2 separate reduction of the number of gadolinium (Gd)-enhancing T1 300 mg infusions, 2 weeks apart. Patient treatment allocation lesions, new or newly enlarged hyperintense T2 lesions, new was unblinded after the last data point from the last patient hypointense T1 lesions, and brain volume loss (baseline to from the DBP was received. The 96-week core studies were week 96) were also reported.3 We report the interim efficacy conducted between 2011 and 2015. The first patient com- and safety results of OCR over the 3-year follow-up in the pleting the DBP entered the OLE phase in August 2013; the open-label extension (OLE) phase of phase III trials in pa- last ongoing patient entered the OLE phase in February 2015. tients with RMS (5 years of total follow-up). The clinical cutoff date for inclusion of data in this analysis was February 5, 2018. The OLE phase is planned until De- cember 2020, with the collection of data for up to 8 years. Methods Classification of evidence Standard protocol approvals, registrations, This study provides Class III evidence that earlier and con- and patient consents tinuous treatment with OCR provided sustained benefit on The relevant institutional review boards/ethics committees ap- clinical and MRI outcomes of disease activity and progression proved the trial protocols (NCT01247324 and NCT01412333). compared with patients switching from IFN-β-1a. The study is All patients provided written informed consent. rated Class III because of the initial treatment randomization disclosure that occurred after inclusion in the OLE. Trial design and patients As previously detailed,3 OPERA I (NCT01247324) and Efficacy assessments OPERA II (NCT01412333) were phase III, multicenter, The following endpoints, which were also measured during randomized, double-blind, double-dummy, IFN-β-1a con- the 96-week DBP, are reported for the OLE phase, using the trolled trials with identical designs, of OCR in patients with pooled OPERA trial population. RMS (data available from Dryad, figure e-1, doi.org/10.5061/ dryad.stqjq2bzw). Key eligibility criteria included an age of Annualized relapse rate 18–55 years, diagnosis of multiple sclerosis (MS; 2010 revised The ARR was calculated as the total number of protocol- McDonald criteria),4 and screening Expanded Disability defined relapses (occurrence of new or worsening neurologic Status Scale (EDSS) score of 0–5.5. Consistency of baseline symptoms [consistent with an increase of at least half a step characteristics and treatment effects across both OPERA on the EDSS score, or 2 points on one of the appropriate studies met predetermined criteria for pooled efficacy analy- Functional Systems Scores (FSS), or 1 point on 2 or more of sis, including annualized relapse rate (ARR) and confirmed the appropriate FSS] that were attributable to MS) for all disability progression (CDP).3 Following completion of the patients in the treatment group divided by the total patient- double-blind controlled treatment phase (DBP) of both trials, years of exposure to that treatment. Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1855 Twenty-four week confirmed disability CDP/CDI at 24 weeks were assessed using Kaplan-Meier and progression/improvement Cox survival analysis in the ITT population and for CDI, in CDP was defined as an increase in the EDSS score from the patients with a baseline EDSS score of ≥2.0; hazard ratios (HRs) baseline of the DBP or of the OLE phase of at least 1.0 point were estimated by stratified Cox regression. Mean change from (increase of ≥0.5 points if baseline EDSS score >5.5); confirmed baseline in EDSS score was analyzed using the mixed-effect disability improvement (CDI) was defined as a reduction in model of repeated measures (MMRM). The number of new T1 EDSS score ≥1.0 point compared with baseline of the DBP or of Gd-enhancing lesions and the number of new or enlarging T2 the OLE phase (reduction of ≥0.5 point if baseline EDSS score lesions were analyzed using a negative binomial model; in >5.5). For analyses based on the OLE phase only, a rebaseline of previously reported analysis3 of lesion outcomes during the EDSS was performed for all patients, including those who had an DBP, results were adjusted for baseline T2 lesion volume, event of CDP/CDI during the DBP. All analyses from the DBP baseline EDSS score (≤2.5 vs >2.5), and geographic region baseline were based on the intention-to-treat (ITT) population, (United States vs rest of the world). However, as patients had except for CDI, which was performed in the prespecified sub- no new T1 Gd-enhancing lesions/new or enlarging T2 lesions group of the ITT population with EDSS ≥2.0. All analyses from at several time points, it was impossible to fit a statistical model, OLE baseline were based on the OLE population with treatment and unadjusted rates were adopted for the OLE instead. The group assignment based on the originally randomized treatment, change in WBV, GMV, and WMV was analyzed using the except for CDI, which was performed in the prespecified sub- MMRM, adjusted for baseline WBV or GMV or WMV and group of the OLE population with EDSS ≥2.0. For both CDP baseline EDSS score (<4.0 vs ≥4.0), study, and geographic re- and CDI, rebaselining at the start of OLE provides a represen- gion (United States vs rest of the world). NEDA was analyzed tation of efficacy during the OLE period unconfounded by any using the Cochran-Mantel-Haenszel test stratified by study, disease activity during the DBP, hence values for the proportion geographic region (United States vs rest of the world), and of patients with events during DBP and rebaselined OLE may not baseline EDSS score (<4.0 vs ≥4.0) in a modified ITT pop- correspond with those of the combined DBP and OLE period. ulation, which included all patients in the ITT population, but patients who discontinued treatment early for reasons other Mean change from baseline in EDSS score than lack of efficacy or death and had NEDA before early dis- Mean change from baseline in EDSS score was assessed at continuation were excluded. Safety outcomes are reported as scheduled visits during the DBP and the OLE phase. incidence rates (events per 100 patient-years of exposure) with Poisson distribution-based confidence intervals (CIs). Acute brain MRI lesion activity The total number of new or newly enlarged T2 lesions and Data availability total number of T1 Gd-enhancing lesions on brain MRI was Qualified researchers may request access to individual patient- assessed every 48 weeks in the OLE phase. level data through the clinical study data request platform (vivli.org). Further details on Roche’s criteria for eligible studies Brain volume change are available at vivli.org/members/ourmembers. For further Percentage change in whole brain volume (WBV) was assessed details on Roche’s Global Policy on the Sharing of Clinical using SIENA/X.5 Percent change in cortical gray matter volume Information and how to request access to related clinical (GMV) and white matter volume (WMV) was assessed using study documents, see roche.com/research_and_development/ paired Jacobian integration.6 All analyses from the DBP baseline who_we_are_how_we_work/clinical_trials/our_commitment_ were based on the ITT population. All analyses from the OLE to_data_sharing.htm. baseline were based on the OLE population with treatment group assignment based on the originally randomized treatment. No evidence of disease activity (NEDA) Results NEDA status was defined as the combined absence of Patient disposition, demographics, and protocol-defined relapses, 24-week CDP, new or enlarging T2 disease characteristics lesions, and T1 Gd-enhancing lesions. Patient disposition and reasons for discontinuation are shown in figure 1 by treatment group. More than 94% of patients in Safety both groups who completed the DBP entered the OLE phase. All patients who received any study treatment were included Of 827 patients treated with OCR during the DBP, 702 in the safety population. All data collected during the DBP (84.9%) entered the OLE phase and 623 (75.3%) remained at and OLE and the safety follow-up were included in the safety year 5. Of the 829 patients treated with IFN-β-1a in the DBP, analyses. Safety outcomes are reported for the OLE phase 623 (75.2%) entered the OLE phase and 551 (66.5%) using the pooled OPERA I and OPERA II ITT population. remained at year 5. Overall, 1,174 of 1,325 patients (89%) who entered the OLE phase of the OPERA studies completed Statistical analyses year 5, representing 71% of the originally enrolled population ARR was analyzed with the use of a generalized estimating (n = 1,656). Patient demographics and disease characteristics equation Poisson regression model in the ITT population. for the pooled studies at baseline of the DBP and at the start of e1856 Neurology | Volume 95, Number 13 | September 29, 2020 Neurology.org/N Figure 1 Patient disposition in the pooled OPERA I and OPERA II intent-to-treat (ITT) population Percentages in parentheses are of the ITT population. aClinical cutoff date: February 5, 2018; patients entering the open-label extension (OLE) phase who completed the double-blind period (DBP): interferon (IFN)-β-1a 623/660 (94.4%) and ocrelizumab (OCR) 702/726 (96.7%). b88.4%, and c88.7%, of patients who entered the OLE completed year 3. SFU = safety follow-up. the OLE phase were well balanced between treatment arms and (10.1% vs 15.6%; p = 0.002), at the end of year 4 (13.9% vs between the ITT population and those patients with a baseline 18.1%; p = 0.03), and at the end of year 5 (16.1% vs 21.3%; p EDSS score of ≥2.0 (table 1). The number of OCR doses = 0.014; figure 3A). The HR for patients receiving contin- received (mean ± SD; median [range]) in the OLE phase uous OCR compared with those switching from IFN-β-1a to population was 7.3 ± 2.0, 8.0 (1–10) for continuous OCR OCR for time to first 24-week CDP during the DBP was 0.60 patients and 7.4 ± 1.9; 8.0 (1–10) in patients switching from (95% CI 0.43–0.84; p = 0.003); after rebaselining at the end IFN-β-1a to OCR. of the DBP (i.e., start of the OLE), the HR during the OLE phase was 1.06 (95% CI 0.8–1.41; p = 0.7) (data Clinical efficacy assessments available from Dryad, figure e-2A, doi.org/10.5061/dryad. Patients receiving continuous OCR maintained a low ARR stqjq2bzw). through year 1 to year 5 (0.14, 0.13, 0.10, 0.08, and 0.07; figure 2). Switching from IFN-β-1a to OCR at the start of the OLE The proportion of patients with 24-week CDI over the du- phase was associated with a significant reduction in ARR (0.20 ration of the DBP and OLE periods (figure 3B) in patients in year 2 to 0.10 in year 3; adjusted rate ratio [95% CI] 2.069 with a baseline EDSS score of ≥2.0 was numerically higher in [1.547–2.766], p < 0.001; 52% relative reduction), which was patients receiving continuous OCR compared with those maintained through years 4 and 5 (0.08 and 0.07; figure 2). In switching from IFN-β-1a to OCR at the end of the DBP (year the OLE phase, there was no difference in ARR at years 3, 4, 2: 16.8% vs 13.3%; p = 0.099), at the end of year 3 (20.6% vs and 5 in patients receiving continuous OCR compared with 16.6%; p = 0.089), at the end of year 4 (23.7% vs 18.9%; p = those switching from IFN-β-1a (p > 0.5, all comparisons; 0.057), and at the end of year 5 (25.8% vs 20.6%; p = 0.046). figure 2). The HR for improvement (time to first 24-week CDI) from baseline to the end of the DBP in patients receiving contin- Over the duration of the DBP and OLE periods, time point uous OCR compared with those switching from IFN-β-1a to analysis demonstrated that the proportion of patients with OCR was 1.31 (95% CI 0.96–1.78; p = 0.06; figure 3B); after 24-week CDP from baseline remained significantly lower in rebaselining at the end of the DBP (i.e., start of the OLE), the patients receiving continuous OCR compared with those HR during the OLE phase was 0.89 (95% CI 0.61–1.31; p = switching from IFN-β-1a to OCR at the end of the DBP 0.6) (data available from Dryad, figure e-2B, doi.org/10.5061/ (year 2: 7.7% vs 12.0%; p = 0.005), at the end of year 3 dryad.stqjq2bzw). Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1857 e1858 Neurology | Volume 95, Number 13 | September 29, 2020 Table 1 Baseline demographics and disease characteristics for the pooled OPERA I and OPERA II populations at the start of the double-blind period (DBP) and open-label extension (OLE) DBP baseline OLE baseline DBP baseline EDSS score of ≥2.0 OLE baseline EDSS score of ≥2.0 IFN-β-1a 44 μg/OCR OCR 600 mg/OCR IFN-β-1a 44 μg/OCR OCR 600 mg/OCR IFN-β-1a 44 μg/OCR OCR 600 mg/OCR IFN-β-1a 44 μg/OCR OCR 600 mg/OCR 600 mg (n = 829) 600 mg (n = 827) 600 mg (n = 623) 600 mg (n = 702) 600 mg (n = 614) 600 mg (n = 628) 600 mg (n = 439) 600 mg (n = 500) Age, y, mean (SD) 37.2 (9.2) 37.1 (9.2) 39.3 (9.2) 39.2 (9.1) 38.6 (9.0) 38.3 (9.3) 41.0 (9.0) 40.4 (9.3) Female, n (%) 552 (66.6) 541 (65.4) 408 (65.5) 454 (64.7) 406 (66.1) 408 (65.0) 297 (67.7) 326 (65.2) EDSS, mean (SD) 2.8 (1.3) 2.8 (1.3) 2.7 (1.5) 2.6 (1.3) 3.3 (1.1) 3.3 (1.0) 3.4 (1.3) 3.1 (1.0) d e e g g d d Patients with T1 Gd- 327 (39.8) 333 (40.7) 106 (17.3) 5 (0.7) 233 (38.3) 256 (41.2) 77 (17.8) 4 (0.8)l enhancing lesions, n (%)a Number of T1 Gd-enhancing 1.9 (5.0)d 1.8 (4.6)e 0.5 (2.1)e 0.02 (0.2)g 2.0 (5.3)g 1.8 (4.8)d 0.4 (1.4)d 0.02 (0.3)l lesions, mean (SD)a Number of T1-hypointense 32.9 (35.1)d 32.4 (35.2)e 35.6 (36.8)h 33.9 (36.1)h 36.9 (36.8)g 36.5 (37.4)d 39.8 (38.0)h 37.8 (38.4) lesions, mean (SD)b T2-hyperintense lesion 10.2 (11.8)f 10.8 (14.1)f 9.4 (11.5)i 10.1 (13.8)j 11.3 (12.4)k 12.2 (15.0)f 10.6 (12.0)m 11.6 (15.1)i volume, cm3, mean (SD)c Number of T2 lesions, mean 51.0 (37.8)f 50.1 (38.8)f 55.5 (41.3) 50.9 (39.4) 53.9 (39.0)k 53.2 (39.8)f 59.3 (42.9) 53.2 (39.8) (SD)b Abbreviations: EDSS = Expanded Disability Status Scale; Gd = gadolinium; IFN = interferon; OCR = ocrelizumab. Demographics and disease characteristics at week 96 of the DBP (clinical cutoff dates: OPERA I, April 2, 2015; OPERA II, May 12, 2015) are considered baseline for the OLE phase. For MRI measurements: a OLE baseline is the assessment at week 96. b OLE baseline is the sum of lesion counts at baseline, week 24, week 48, and week 96. c OLE baseline is the last assessment prior to or at the start of OLE treatment. Patients missing and excluded for calculating percentages: d n = 7. e n = 9. f n = 5. g n = 6. h n = 1. i n = 33. j n = 44. k n = 4. l n = 3. m n = 28. Neurology.org/N Figure 2 Annualized relapse rate (ARR) in double-blind controlled treatment phase (DBP) study years 1 and 2 and open- label extension (OLE) years 1–3 Estimates are from analysis based on generalized estimating equation Poisson regression model with repeated measurements using unstructured co- variance matrix, adjusted by randomized treatment, study, baseline Expanded Disability Status Scale (<4.0 vs ≥4.0), geographic region (United States vs rest of the world), year, and treatment-by-year interaction. Log-transformed exposure time is included as an offset variable. aThe total number of relapses for all patients in the treatment group divided by the total patient-years of exposure to that treatment. bDBP year 1 and DBP year 2 data include the intention-to- treat population (number of patients available); year 3 (OLE year 1), year 4 (OLE year 2), and year 5 (OLE year 3) data include the OLE ITT population (number of patients available). Clinical cutoff date: February 5, 2018. IFN = interferon; OCR = ocrelizumab. Patients continuing OCR had a lower mean change from At 5 years (OLE year 3), patients treated continuously with baseline in EDSS score compared with patients switching OCR compared with those switching from IFN-β-1a to OCR from IFN-β-1a to OCR, at the end of the DBP (year 2: −0.146 experienced significantly lower brain atrophy measured by vs 0.032; p < 0.001), during year 3 (−0.118 vs 0.010; p = change from DBP baseline in WBV, cortical GMV, and 0.008), during year 4 (−0.057 vs 0.041; p = 0.06), and during WMV (adjusted rate: WBV –1.87% vs −2.15%, cortical year 5 (−0.043 vs 0.058; p = 0.07) (data available from Dryad, GMV −2.02% vs −2.25%, and WMV –1.33% vs −1.62%; p < figure e-3, doi.org/10.5061/dryad.stqjq2bzw). 0.01 for all comparisons; figure 5, A1, B1, and C1). Similarly, significant differences between patients treated continuously MRI efficacy assessments with OCR compared with patients switching from IFN-β-1a Patients receiving continuous OCR maintained the near to OCR were seen at years 2, 3, and 4 for change from complete suppression of MRI disease activity seen in the DBP baseline in WBV, cortical GMV, and WMV (adjusted rates: p through to year 5 (figure 4, A and B): the unadjusted rate of ≤ 0.01 for all comparisons; figure 5, A1, B1, and C1), except total T1 Gd-enhancing lesions was 0.017 at the end of the DBP for cortical GMV at year 3, where no difference was found (for year 2 only), 0.005 at year 3, 0.017 at year 4, and 0.006 at (adjusted rate −1.47% vs −1.56%; p = 0.155; figure 5B1). year 5; and for new or newly enlarged T2 lesions, the un- When expressed as annualized percent change in volume adjusted rate was 0.063 over year 2, 0.091 over year 3, 0.080 occurring during the preceding 48 weeks, the adjusted rates over year 4, and 0.031 over year 5. Patients switching from IFN- were generally stable during the OLE phase for patients β-1a to OCR had almost complete and sustained suppression switching from IFN-β-1a to OCR and those continuing of MRI lesion disease activity from year 3 to year 5 (figure 4, A OCR (figure 5, A2, B2, and C2). and B): the unadjusted rate of total T1 Gd-enhancing lesions was 0.491 at the end of the DBP (year 2), decreasing to 0.007 at NEDA in the OLE year 3, 0.004 at year 4, and 0.004 at year 5; and for new or In patients treated with OCR compared with IFN-β-1a, the enlarging T2 lesions, the unadjusted rate was 2.583 over the relative proportion of patients with NEDA was increased by duration of year 2, decreasing to 0.371 over year 3, 0.063 over 74% in the DBP (OCR 48.5% [n = 369/761] vs IFN-β-1a year 4, and 0.038 over year 5. In the OLE phase, there was no 27.8% [n = 210/756]; p < 0.001) and by 88% over the du- difference in MRI lesion counts at years 3, 4, and 5 in patients ration of the DBP and OLE periods (OCR 35.7% [n = 255/ receiving continuous OCR compared with those switching 715] vs IFN-β-1a/OCR 19.0% [n = 140/738]; p < 0.001). from IFN-β-1a (p > 0.5, all comparisons; figure 4), except for During the OLE period alone, the proportion of patients with new or enlarging T2 lesions at year 3, where the number was NEDA was 65.4% (n = 409/625) in patients treated contin- lower in patients receiving continuous OCR compared with uously with OCR compared with 55.1% (n = 310/563) in those switching from IFN-β-1a (0.091 vs 0.371; p < 0.001; those switching from IFN-β-1a to OCR, a relative difference of figure 4B). 19% (p < 0.001). Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1859 Figure 3 Time to onset of confirmed disability progression (CDP) and confirmed disability improvement (CDI) for at least 24 weeks during double-blind controlled treatment phase (DBP) and open-label extension (OLE) periods Time to onset of (A) CDP and (B) CDI. Curves show Kaplan-Meier estimates of the proportion of patients with disability progression/improvement events relative to the original double-blind treatment period baseline throughout double-blind and OLE periods. Without imputation. Intention-to-treat (ITT) population. Pooled OPERA I and OPERA II population; DBP clinical cutoff dates: April 2, 2015, and May 12, 2015, respectively; OLE clinical cutoff date: February 5, 2018. Data shown up to week 240, the last visit all ongoing patients completed. HR = hazard ratio; IFN = interferon; OCR = ocrelizumab. Safety in the OLE the primary DBP analysis cutoff date (May 2015) for each Table 2 summarizes the safety analyses in all patients treated group (events per 100 patient-years [95% CI], IFN-β-1a: 296 with OCR over a period of 5 years, which includes patients [287–305]; OCR 290 [281–299]). The most common AEs treated continuously with OCR (5 years) and patients switched included infusion-related reactions, urinary tract infections from IFN-β-1a (3 years). As of February 2018, 1,448 patients (UTIs), and upper respiratory tract infections. Overall, the rate with MS received OCR during the DBP and associated OLE of serious AEs was 6.1 (95% CI 5.5–6.8) events per 100 patient- periods of the OPERA trials. The rate of adverse events (AEs) years, consistent with the rate observed at the primary analysis in the overall OCR exposure population was 220 (95% CI cutoff date for each group (IFN-β-1a: 6.29 [5.05–7.75]; OCR: 217–224) per 100 patient-years, lower than the rate observed at 5.39 [4.26–6.72]). The rate of infections per 100 patient-years e1860 Neurology | Volume 95, Number 13 | September 29, 2020 Neurology.org/N Figure 4 Total number of T1 gadolinium (Gd)-enhancing lesions and new or enlarging T2 lesions in the double-blind controlled treatment phase (DBP) and open-label extension (OLE) (A) T1 Gd-enhancing lesions and (B) new or enlarging T2 lesions in the DBP and OLE. aDBP week 24, DBP year 1, and DBP year 2 data include the intention-to- treat (ITT) population; year 3 (OLE year 1), year 4 (OLE year 2), and year 5 (OLE year 3) data include the OLE ITT population; clinical cutoff date: February 5, 2018. b Unadjusted rate. IFN = interferon; OCR = ocrelizumab. was 75.3 (95% CI 73.1–77.6), consistent with the rate observed incidence rate for IFN-β-1a was 0.14 (0.02–0.52) and for at the primary analysis cutoff date for each group (IFN-β-1a: OCR was 0.28 (0.08–0.71). A list of malignancies from the 67.8 [95% CI 63.5–72.2]; OCR: 84.5 [95% CI 79.9–89.4]). DBP and OLE phase of the pooled OPERA studies as of February 2018 is provided in the supplemental data (available The most common serious AEs were classified as infections, from Dryad, supplemental data, doi.org/10.5061/dryad. which occurred at a rate of 1.5 per 100 patient-years (95% CI stqjq2bzw). 1.2–1.8) over the 5-year period, and this was similar to the rate observed at the primary analysis cutoff date (IFN-β-1a: 1.79 The rate per 100 patient-years of AEs leading to treatment [1.16–2.64]; OCR: 0.83 [0.43–1.45]). The most common withdrawals in the overall OCR exposure population (year 5: serious infections were UTIs and pneumonia. Two potential 1.36 [95% CI 1.08–1.70]) did not increase over time (rate serious opportunistic infections were reported in the OLE observed at the primary analysis cutoff date: IFN-β-1a: 3.93 period; both patients recovered with treatment by standard [2.96–5.12]; OCR: 2.35 [1.63–3.28]). therapies (systemic Pasteurella infection in a patient with RMS following a cat bite; enterovirus-induced fulminant Over 5 study years in the pooled OPERA study population, a hepatitis in a diabetic patient with RMS, resulting in liver reduction in serum immunoglobulin (Ig) levels was observed. transplant). As of end of April 2019, no cases of progressive At baseline, the number (%) of patients with Ig concentra- multifocal leukoencephalopathy (PML) were identified in the tions below the lower limit of normal (LLN) were 7 (0.5%) overall OCR exposure population for the pooled OPERA for IgG, 17 (1.2%) for IgA, and 7 (0.5%) for IgM. Over 5 study studies. The crude incidence rate of all malignancies per 100 years, for the majority of patients, Ig levels remained above the patient-years in the overall OCR exposure population was LLN (data available from Dryad, table e-1, doi.org/10.5061/ 0.40 (0.26–0.61), and at the primary analysis cutoff date the dryad.stqjq2bzw); the number (%) of patients with a decrease Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1861 Figure 5 Percentage change from baseline and annualized, in whole brain volume (WBV), cortical gray matter volume (CGMV), and white matter volume (WMV) in the double-blind controlled treatment phase (DBP) and open-label extension (OLE) Percentage change (A.a, B.a, C.a) from baseline and (A.b, B.b, C.b) annualized, in (A) WBV, (B) CGMV, and (C) WMV in the DBP and OLE mixed-effect model of repeated measures (MMRM) plot, intention-to-treat (ITT) population. Pooled OPERA [Rebif] I and OPERA II (clinical cutoff date: February 5, 2018); p values shown for difference in adjusted means. Graph includes patients with assessment at baseline and at least 1 postbaseline value. Estimates are from analysis based on MMRM using unstructured covariance matrix: percentage change = baseline brain volume + geographic region (United States vs rest of the world) + baseline Expanded Disability Status Scale (<4.0 vs ≥4.0) + study + week + treatment + treatment × week (repeated values over week) + baseline brain volume × week. aThe bars represent the annualized change in WBV, CGMV, or WMV occurring during the time period delineated by the axis label and the preceding 24 or 48 weeks. BL = baseline; CI = confidence interval; IFN = interferon; OCR = ocrelizumab. e1862 Neurology | Volume 95, Number 13 | September 29, 2020 Neurology.org/N Table 2 Adverse events, serious adverse events, and adverse events leading to discontinuation during the double-blind period (DBP) and open-label extension (OLE) phase of the pooled OPERA I and OPERA II populations OPERA (pooled) controlled treatment perioda IFN-β-1a rate per 100 OCR rate per 100 OPERA DBP/OLE,b OCR rate Event patient-years (95% CI)c patient-years (95% CI)c per 100 patient-years (95% CI)c Any adverse eventd 296 (287–305) 290 (281–299) 220 (217–224) Adverse events leading to study 3.93 (2.96–5.12) 2.35 (1.63–3.28) 1.36 (1.08–1.70) treatment discontinuation Infections and infestationsd 67.8 (63.5–72.2) 84.5 (79.9–89.4) 75.3 (73.1–77.6) Urinary tract infection 9.7 (8.1–11.4) 11.6 (9.9–13.5) 11.7 (10.8–12.6) Nasopharyngitis 8.3 (6.9–9.9) 13.0 (11.2–15.0) 11.0 (10.2–11.9) Upper respiratory tract infection 9.4 (7.8–11.1) 13.3 (11.5–15.3) 12.6 (11.7–13.6) Bronchitis 2.2 (1.5–3.1) 3.5 (2.6–4.6) 3.7 (3.2–4.2) Influenza 3.3 (2.4–4.4) 3.1 (2.3–4.2) 2.7 (2.3–3.1) Injury, poisoning and procedural 17.1 (15.0–19.4) 45.9 (42.4–49.5) 28.2 (26.8–29.6) complicationsd Infusion-related reactions 7.9 (6.5–9.5) 34.9 (31.9–38.1) 18.2 (17.1–19.3) Nervous system disordersd 34.8 (31.8–38.0) 31.6 (28.8–34.7) 20.7 (19.5–21.9) Headache 12.4 (10.6–14.4) 9.5 (8.0–11.3) 4.7 (4.2–5.3) Musculoskeletal and connective 25.0 (22.5–27.8) 24.3 (21.8–27.0) 18.8 (17.6–20.0) tissue disordersd Back pain 3.1 (2.2–4.1) 4.1 (3.1–5.3) 3.0 (2.6–3.5) Arthralgia 3.9 (3.0–5.1) 3.5 (2.6–4.6) 2.8 (2.3–3.2) Pain in extremity 2.9 (2.1–4.0) 3.7 (2.7–4.8) 2.3 (1.9–2.8) General disorders and administration 51.3 (47.6–55.2) 17.3 (15.2–19.5) 10.6 (9.8–11.5) site conditionsd Fatigue 5.7 (4.5–7.1) 5.4 (4.3–6.7) 3.5 (3.1–4.0) d Psychiatric disorders 14.2 (12.3–16.3) 14.4 (12.5–16.5) 8.5 (7.7–9.2) Depression 4.2 (3.2–5.4) 4.9 (3.8–6.2) 2.9 (2.5–3.4) Malignanciesd,e,f 0.14 (0.02–0.52) 0.28 (0.08–0.71) 0.40 (0.26–0.61) d Serious adverse events 6.29 (5.05–7.75) 5.39 (4.26–6.72) 6.1 (5.5–6.8) g Serious infections 1.79 (1.16–2.64) 0.83 (0.43–1.45) 1.5 (1.2–1.8) Number of potential serious OIsh 0 0 2 Fatalitiesi 0.14 (0.02–0.52) 0.07 (0–0.38) 0.04 (0–0.13) Abbreviations: CCOD = clinical cutoff date; CI = confidence interval; IFN = interferon; MedDRA = Medical Dictionary for Regulatory Activities; OCR = ocreli- zumab; OI = opportunistic infection. a Includes patients who received IFN-β-1a during the controlled treatment period of the OPERA I (CCOD April 2015) and OPERA II (CCOD May 2015) studies. b Includes patients who received any dose of OCR during the controlled treatment and associated OLE phase of the OPERA I and OPERA II studies; data from patients who were originally randomized to comparator (IFN-β-1a) are included after the switch to open-label OCR treatment. CCOD for inclusion of data in this analysis was February 5, 2018. c Multiple occurrences of the same adverse event (except for malignancies) in one patient are counted multiple times. d Includes adverse events falling into the MedDRA versions 18.0 (OPERA [pooled] controlled treatment period) and 20.1 (CCOD: February 2018). e Malignancies are identified using adverse events falling into the standard MedDRA query “Malignant tumours (narrow).” f Reported as incidence rate per 100 patient-years of first malignancy. g Serious infections are defined using adverse events falling into the MedDRA system organ class Infections and Infestations, and using “Is the event nonserious or serious?” from the adverse event case report form. h Potential serious OIs were medically reviewed. i A causal association between OCR and fatal events could not be established. Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1863 below the LLN at year 5 were 33 (5.4%) for IgG, 31 (5.1%) switched from IFN-β-1a (16.1% vs 21.3%). Similarly, by the for IgA, and 164 (29.5%) for IgM. end of the OLE phase (year 5), a significantly higher pro- portion of patients initiating OCR 2 years earlier achieved improvements in disability compared with patients switching from IFN-β-1a (25.8% vs 20.6%). These results add to the body Discussion of evidence that early high-efficacy treatment is more favorable These interim results from the OPERA extension study provide than the traditional escalation approach. A recent observational several important insights on the efficacy and safety of long- study demonstrated that patients commencing high-efficacy term (up to 5 years) treatment with OCR in patients with DMTs, including OCR, early after disease onset accumulate RMS. The low levels of disease activity observed in patients less long-term disability compared with those exposed later in treated with OCR during the 2-year controlled period3 were their disease.8–11 sustained from years 3–5, suggesting a benefit of earlier treat- ment and persistence of the effect with maintained therapy. In OCR also slowed the rate of brain volume loss. At 5 years from patients who switched from IFN-β-1a, OCR consistently led to DBP baseline, patients on early and continuous OCR experi- a rapid decrease in relapse activity and an almost complete enced lower WBV loss during the DBP compared with IFN- suppression of MRI activity. Disability accrual occurred at β-1a, and the difference between the 2 groups was maintained similarly low rates in the 2 groups during the OLE phase but, at throughout the OLE phase. The beneficial effect was also ob- the end of year 5, the proportion of patients with CDP from served in both the white matter and cortical gray matter baseline of the DBP was lower in those originally treated with compartments, with the exception of cortical GMV at year 3 OCR. The safety profile observed in years 3–5 was generally where the absence of a difference might be due to the reversal consistent with that observed during the controlled period, of pseudoatrophy affecting gray matter in patients switching with no additional safety findings that suggest potential dele- from IFN-β-1a to OCR.12 In patients with MS, a parenchymal terious effects of long-term exposure to OCR. loss of ≥0.4% per year (≥2.0% over a 5-year period) has been proposed as a pathologic atrophy rate.13 The 5-year cumulative In patients treated early and continuously with OCR, consistent WBV loss from DBP baseline in patients treated early and and persistent effects were evident on clinical and MRI mea- continuously with OCR (−1.87%) was lower than in those that sures of inflammatory disease activity. ARR remained low over 5 switched from IFN-β-1a (−2.15%). Taken together, our find- years. Consistent with the effects on relapses, the near complete ings converge in demonstrating that patients with RMS who suppression of subclinical disease activity, as measured by MRI, start later with high-efficacy treatment do not catch up on observed in the DBP was maintained throughout the OLE clinical and MRI outcomes reflecting residual functional loss phase, with few patients exhibiting either T1 Gd-enhancing and tissue damage. lesions (0.006 lesions per scan) or new/enlarging T2 lesions (0.031 lesions per scan) at the end of year 5. Patients who The durable efficacy of OCR across multiple endpoints, with an started OCR after switching from IFN-β-1a experienced sig- early impact not only on clinical and MRI measures of in- nificant decreases in levels of disease activity, with relapse rates flammation but also on disease progression, was accompanied and lesion counts becoming similar to those observed in pa- by a safety profile consistent with previous reports.3 This was tients on early and continuous OCR at year 3 and consistently reflected in the low on-study attrition rates, where almost 90% maintained through to year 5. The effect of OCR in reducing of all patients who entered the OLE phase completed year 5, the relapse rate is of particular relevance to the long-term indicating that OCR was generally well-tolerated. The reported prognosis, as natural history and long-term follow-up studies of rates of AEs per 100 patient-years with OCR during the OLE MS disease-modifying therapies (DMTs) show that more fre- out to year 5 continue to be generally consistent with those quent relapses early in the disease course, and relapses occurring seen during the DBP in patients with RMS. Consistent with over the 2- to 3-year interval of a randomized clinical trial, are results of previously reported trials,3,14 the incidence rates of associated with adverse long-term disability outcomes.7 malignancies in patients treated with OCR remain within the range of placebo data from clinical trials in MS and epidemi- The benefits of earlier and sustained OCR treatment for ologic data for this patient population.14,15 A small increase in measures of disease progression were evidenced by lower levels serious infections has been previously reported during OLE of clinically evaluated CDP, with only 16.1% of patients who observations in patients with primary progressive MS treated continued OCR experiencing 24-week CDP at the end of the with OCR.16 No cases of PML occurred in the controlled OLE phase (year 5). In patients who switched from IFN-β-1a, a treatment period of the OCR MS clinical trials (pivotal phase treatment effect of OCR was observed starting in the first year III studies and the phase II study) and to date, no cases have of the OLE, leading to rates of disability progression similar to been observed in the ongoing OLE of these studies or in the those seen in patients continuing OCR in the OLE phase. other ongoing MS clinical trials. Outside of clinical trials, as of However, a higher proportion of IFN-β-1a–treated patients October 2019, there have been 8 confirmed cases of PML in experienced 24-week CDP by the end of the OLE phase (year patients with MS treated with OCR: 7 cases of probable car- 5). Those patients who initiated OCR 2 years earlier accrued ryover PML (6 switching from natalizumab, 1 from fingoli- significantly less disability progression compared with those mod); in one case, the patient had not been treated with other e1864 Neurology | Volume 95, Number 13 | September 29, 2020 Neurology.org/N DMTs but had contributing risk factors for PML, including old The authors had full editorial control of the manuscript and age (78 years) and preexisting grade 1 lymphopenia (which provided their final approval of all content. worsened to grade 2 during treatment). No other PML cases with OCR have been reported to date (based on drug product Disclosure sales as of October 2019, where an estimated 130,000 patients S.L. Hauser serves on the board of trustees for Neurona and on with RMS and PPMS have received OCR treatment for ap- scientific advisory boards for Alector, Annexon, Bionure, Mo- proximately 140,000 patient-years).16 The observed decrease lecular Stethoscope, and Symbiotix and has received travel re- in Ig levels and potential implications for patient safety, such as imbursement and writing assistance from F. Hoffmann-La serious infections, need to be considered in the context of Roche Ltd for CD20-related meetings and presentations. L. longer follow-up data. Kappos’ institution, the University Hospital Basel, has received research support and payments that were used exclusively for As with other extension studies of DMTs, definitive con- research support for L. Kappos’ activities as principal in- clusions related to efficacy outcomes are limited by the ab- vestigator and member or chair of planning and steering sence of a control arm, and the influence of regression to the committees or advisory boards for trials sponsored by Actelion, mean, which has been described in MS studies, cannot be Addex, Almirall, Bayer HealthCare Pharmaceuticals, CLC excluded.17 As blinding was not maintained in the OLE Behring, F. Hoffmann-La Roche Ltd, Genentech, Inc, GeN- phase, some rater assessment bias could have been in- euro SA, Genzyme, Merck Serono, Mitsubishi Pharma, troduced, although EDSS raters continue to be blinded to all Novartis, Octapharma, Ono Pharmaceutical, Pfizer, Receptos, other procedures, and MRI interpretation remains masked Sanofi, Santhera, Siemens, Teva, UCB, and XenoPort; has re- in the OLE phase. There is a possibility of selection bias ceived license fees for Neurostatus products; and has received given that participation in the extension was voluntary. research grants from the European Union, Novartis Research However, this is unlikely to be significant as over 94% of Foundation, Roche Research Foundation, Swiss Multiple patients in both groups who completed the DBP entered the Sclerosis Society, Swiss National Research Foundation, and OLE, with 88% of these patients completing year 5. It should Innosuisse. D.L. Arnold has received personal fees for consul- also be noted that the disease duration at baseline for the ting from Acorda, Albert Charitable Trust, Biogen, Celgene, F. majority of the patients enrolled in the OPERA I/II studies Hoffmann-La Roche Ltd, Frequency Therapeutics, MedI- was longer than the 5 years defined in one study as the mmune, MedDay, Merck Serono, Novartis, Sanofi-Aventis, and threshold for early MS,9 but the duration was typical of other Wave Life Sciences; grants from Biogen, Immunotec, and clinical trials in the higher-efficacy DMT era. The long-term Novartis; and an equity interest in NeuroRx Research. A. Bar- outcome of ocrelizumab therapy in patient populations with Or has served on scientific advisory boards for Biogen, F. earlier MS disease course requires further study. Hoffmann-La Roche Ltd, Genentech, Inc, GlaxoSmithKline, Guthy-Jackson/GGF, MedImmune, Merck/EMD Serono, Studies that demonstrate the maintenance of long-term efficacy Mitsubishi Tanabe, Ono, Receptos, and Sanofi-Genzyme and and safety of DMTs are valuable since patients with MS need to has received research support from Novartis and Sanofi- receive DMTs over the long term. The results of this analysis of Genzyme. B. Brochet has received research support from 5-year data from the DBP and OLE of the pooled OPERA Bayer, Merck, Genzyme, Biogen, Novartis, Roche, Actelion, studies underscore persistence of the OCR treatment effect and MedDay, and Teva and has consulting agreements with suggest that early and continuous treatment provides sustained Novartis and Biogen. R.T. Naismith reports financial relation- long-term benefit. These and other data support the early use of ships with Alkermes, Biogen, Celgene, EMD Serono, Gen- highly effective therapies that impact clinical and MRI measures entech, Inc, Genzyme, Novartis, and TG Therapeutics. A. of disease activity and progression in RMS to optimize short- Traboulsee has received research support from Sanofi Gen- and long-term patient outcomes.18 zyme, Roche, Chugai, Novartis, and Biogen; has received consulting fees from Sanofi Genzyme, Roche, Teva Neurosci- Acknowledgment ence, Novartis, Biogen, and EMD Serono; and has received The authors thank the patients, their families, and the honoraria for his involvement in speaker bureau activities for investigators who participated in this trial (including the Sanofi Genzyme and Teva Neuroscience. J.S. Wolinsky has OPERA Study Steering Committee, which provided study received personal compensation for consulting, serving on a oversight); the independent data monitoring committee for scientific advisory board, speaking, or other activities with performing data analysis and safety monitoring; and Gisèle von AbbVie, Acorda Therapeutics, Alkermes, Biogen, Bionest, Büren (of F. Hoffmann-La Roche Ltd) for additional critical Celgene, Clene Nanomedicine, EMD Serono, Forward review of this manuscript and technical advice. Pharma A/S, GeNeuro, GW Pharma, MedDay Pharmaceuti- cals, Novartis, Otsuka, PTC Therapeutics, Roche/Genentech, Study funding and Sanofi Genzyme; royalties are received for out-licensed This research was funded by F. Hoffmann-La Roche Ltd, monoclonal antibodies through UTHealth from Millipore Basel, Switzerland. Terence Smith of Articulate Science, UK, Corporation. S. Belachew was an employee of F. Hoffmann-La wrote the first draft of the manuscript based on input from the Roche Ltd during the completion of the work related to this authors; his work was funded by F. Hoffmann-La Roche Ltd. manuscript. S. Belachew is now an employee of Biogen Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1865 (Cambridge, MA), which was not in any way associated with this study. H. Koendgen is an employee and shareholder of F. Appendix (continued) Hoffmann-La Roche Ltd. V. Levesque is an employee of Name Location Contribution Genentech, Inc, and shareholder of F. Hoffmann-La Roche Ltd. M. Manfrini is an employee and shareholder of F. Harold F. Hoffmann-La Roche Ltd, Reviewed the data, drafted Koendgen, Basel and revised the manuscript Hoffmann-La Roche Ltd. F. Model is an employee and MD, PhD content shareholder of F. Hoffmann-La Roche Ltd. S. Hubeaux is an Victoria Genentech, Inc., South San Drafted and revised the employee and shareholder of F. Hoffmann-La Roche Ltd. L. Levesque, Francisco, California manuscript content Mehta was an employee of Genentech, Inc., and a shareholder PhD of F. Hoffmann-La Roche Ltd. during the completion of the Marianna F. Hoffmann-La Roche Ltd, Drafted and revised the work related to this manuscript. L. Mehta is now an employee Manfrini, Basel manuscript content MD of Alder Biopharmaceuticals Inc (Bothell, WA), which was not in any way associated with this study. X. Montalban has re- Fabian F. Hoffmann-La Roche Ltd, Analyzed the data, drafted Model, PhD Basel and revised the manuscript ceived speaker honoraria and travel expense reimbursement for content participation in scientific meetings, been a steering committee member of clinical trials, or served on advisory boards of clinical Stanislas F. Hoffmann-La Roche Ltd, Analyzed the data, drafted Hubeaux, Basel and revised the manuscript trials for Actelion, Almirall, Bayer, Biogen, F. Hoffmann-La MSc content Roche Ltd, Genzyme, Merck, Novartis, Octapharma, Receptos, Lahar Genentech, Inc., South San Reviewed the data, drafted Sanofi, Teva, and Trophos. Go to Neurology.org/N for full Mehta, MD Francisco, California and revised the manuscript disclosures. content Xavier University of Toronto, Designed and Publication history Montalban, Canada; Vall d’Hebron conceptualized study, MD, PhD University Hospital, reviewed the data, drafted Received by Neurology July 10, 2019. Accepted in final form Barcelona, Spain and revised the manuscript April 10, 2020. content References Appendix Authors 1. DiLillo DJ, Hamaguchi Y, Ueda Y, et al. Maintenance of long-lived plasma cells and serological memory despite mature and memory B cell depletion during CD20 im- Name Location Contribution munotherapy in mice. J Immunol 2008;180:361–371. 2. Klein C, Lammens A, Schäfer W, et al. Epitope interactions of monoclonal anti- Stephen University of California, San Designed and bodies targeting CD20 and their relationship to functional properties. MAbs 2013; Hauser, MD Francisco conceptualized study, 5:22–33. reviewed the data, drafted 3. Hauser SL, Bar-Or A, Comi G, et al. Ocrelizumab versus interferon beta-1a in re- and revised the manuscript lapsing multiple sclerosis. N Engl J Med 2017;376:221–234. content 4. Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69: Ludwig University Hospital Basel, Designed and 292–302. Kappos, MD University of Basel conceptualized study, 5. Smith SM, Zhang Y, Jenkinson M, et al. Accurate, robust, and automated longitudinal reviewed the data, drafted and cross-sectional brain change analysis. Neuroimage 2002;17:479–489. and revised the manuscript 6. Nakamura K, Guizard N, Fonov VS, Narayanan S, Collins DL, Arnold DL. Jacobian content integration method increases the statistical power to measure gray matter atrophy in multiple sclerosis. Neuroimage Clin 2014;4:10–17. Douglas McGill University; NeuroRx Reviewed the data, drafted 7. Goodin DS, Reder AT, Bermel RA, et al. Relapses in multiple sclerosis: relationship to Arnold, MD Research, Montreal, and revised the manuscript disability. Mult Scler Relat Disord 2016;6:10–20. Canada content 8. Harding K, Williams O, Willis M, et al. Clinical outcomes of escalation vs early intensive disease-modifying therapy in patients with multiple sclerosis. JAMA Neurol Amit Bar- University of Pennsylvania, Drafted and revised the 2019;76:536–541. Or, MD Philadelphia manuscript content 9. Brown JWL, Coles A, Horakova D, et al. Association of initial disease-modifying therapy with later conversion to secondary progressive multiple sclerosis. JAMA 2019; Bruno CHU de Bordeaux, France Drafted and revised the 321:175–187. Brochet, manuscript content 10. Naegelin Y, Naegelin P, von Felten S, et al. Association of rituximab treatment with MD disability progression among patients with secondary progressive multiple sclerosis. JAMA Neurol 2019;76:274–281. Robert Washington University Drafted and revised the 11. He A, Merkel B, Zhovits L, et al. Early start of high-efficacy therapies improves Naismith, School of Medicine, St. manuscript content disability outcomes over 10 years. Presented at the 34th Congress of the European MD Louis, MO Committee for Treatment and Research in Multiple Sclerosis (ECTRIMS); October 10–12, 2018; Berlin. Poster P919. Anthony University of British Drafted and revised the 12. Battaglini M, Giorgio A, Luchetti L, et al. Dynamics of pseudo-atrophy in relapsing- Traboulsee, Columbia, Vancouver, manuscript content remitting MS patients treated with interferon beta-1a as assessed by monthly brain MD Canada MRI. Presented at the 34th Congress of the European Committee for Treatment and Research in Multiple Sclerosis (ECTRIMS); October 10–12, 2018; Berlin. Poster Jerry McGovern Medical School, Designed and conceptualized P798. Wolinsky, UTHealth, Texas study, reviewed the data, 13. De Stefano N, Stromillo ML, Giorgio A, et al. Establishing pathological cut-offs of MD drafted and revised the brain atrophy rates in multiple sclerosis. J Neurol Neurosurg Psychiatry 2016;87: manuscript content 93–99. 14. Hauser SL, Kappos L, Montalban X, et al. Safety of ocrelizumab in multiple sclerosis: Shibeshih F. Hoffmann-La Roche Ltd, Reviewed the data, drafted updated analysis in patients with relapsing and primary progressive multiple sclerosis. Belachew, Basel and revised the manuscript Presented at the 34th Congress of the European Committee for Treatment and MD, PhD content Research in Multiple Sclerosis (ECTRIMS); October 10–12, 2018; Berlin, Germany. Poster P1229. e1866 Neurology | Volume 95, Number 13 | September 29, 2020 Neurology.org/N 15. Nørgaard M, Veres K, Didden EM, Wormser D, Magyari M. Multiple sclerosis and 17. Martı́nez-Yélamos S, Martı́nez-Yélamos A, Martı́n Ozaeta G, Casado V, Carmona cancer incidence: a Danish nationwide cohort study. Mult Scler Relat Disord 2019;28: O, Arbizu T. Regression to the mean in multiple sclerosis. Mult Scler 2006;12: 81–85. 826–829. 16. Hauser SL, Kappos L, Montalban X, et al. Safety of ocrelizumab in multiple sclerosis: updated 18. Ziemssen T, De Stefano N, Sormani MP, Van Wijmeersch B, Wiendl H, Kieseier BC. analysis in patients with relapsing and primary progressive multiple sclerosis. Presented at the Optimizing therapy early in multiple sclerosis: an evidence-based view. Mult Scler 14th World Congress of Neurology Meeting; October 27–31, 2019; Dubai, UAE. Relat Disord 2015;4:460–469. Neurology.org/N Neurology | Volume 95, Number 13 | September 29, 2020 e1867 Five years of ocrelizumab in relapsing multiple sclerosis: OPERA studies open-label extension Stephen L. Hauser, Ludwig Kappos, Douglas L. Arnold, et al. Neurology 2020;95;e1854-e1867 Published Online before print July 20, 2020 DOI 10.1212/WNL.0000000000010376 This information is current as of July 20, 2020 Updated Information & including high resolution figures, can be found at: Services http://n.neurology.org/content/95/13/e1854.full References This article cites 14 articles, 2 of which you can access for free at: http://n.neurology.org/content/95/13/e1854.full#ref-list-1 Citations This article has been cited by 1 HighWire-hosted articles: http://n.neurology.org/content/95/13/e1854.full##otherarticles Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): All Clinical Neurology http://n.neurology.org/cgi/collection/all_clinical_neurology All Clinical trials http://n.neurology.org/cgi/collection/all_clinical_trials Multiple sclerosis http://n.neurology.org/cgi/collection/multiple_sclerosis Permissions & Licensing Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/about/about_the_journal#permissions Reprints Information about ordering reprints can be found online: http://n.neurology.org/subscribers/advertise Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology.. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.