RESEARCH ARTICLE OPEN ACCESS Associations of Disease-Modifying Therapies With COVID-19 Severity in Multiple Sclerosis Steve Simpson-Yap, PhD, MPH,* Edward De Brouwer, MSc,* Tomas Kalincik, MD, PhD, PGCertBiostat, Correspondence Nick Rijke, MA, Jan A. Hillert, MD, PhD, Clare Walton, PhD, Gilles Edan, MD, Yves Moreau, PhD, Dr. Liesbet Peeters Tim Spelman, PhD, Lotte Geys, PhD, Tina Parciak, MSc, Clement Gautrais, PhD, Nikola Lazovski, PDEng, liesbet.peeters@uhasselt.be Ashkan Pirmani, MSc, Amin Ardeshirdavanai, PhD, Lars Forsberg, MSc, Anna Glaser, PhD, Robert McBurney, PhD, Hollie Schmidt, MS, Arnfin B. Bergmann, MD, PhD, Stefan Braune, PhD, Alexander Stahmann, MSc, Rodden Middleton, MBA, Amber Salter, PhD, Robert J. Fox, MD, Anneke van der Walt, PhD, Helmut Butzkueven, PhD, Raed Alroughani, PhD, Serkan Ozakbas, PhD, Juan I. Rojas, MSc, Ingrid van der Mei, PhD, Nupur Nag, PhD, Rumen Ivanov, MSc, Guilherme Sciascia do Olival, PhD, Alice Estavo Dias, PhD, Melinda Magyari, PhD, Doralina Brum, PhD, Maria Fernanda Mendes, PhD, Ricardo N. Alonso, MD, MSc, Richard S. Nicholas, PhD, Johana Bauer, PhD, Anı́bal Sebastián Chertcoff, MD, Anna Zabalza, MD, Georgina Arrambide, MD, PhD, Alexander Fidao, MPH, Giancarlo Comi, MD,† and Liesbet Peeters, PhD† ® Neurology 2021;97:e1870-e1885. doi:10.1212/WNL.0000000000012753 Abstract RELATED ARTICLE Background and Objectives Editorial People with multiple sclerosis (MS) are a vulnerable group for severe coronavirus disease 2019 B-Cell Depletion and (COVID-19), particularly those taking immunosuppressive disease-modifying therapies COVID-19 Severity in Multiple Sclerosis: (DMTs). We examined the characteristics of COVID-19 severity in an international sample of Remaining Challenges people with MS. Page 885 Methods MORE ONLINE Data from 12 data sources in 28 countries were aggregated (sources could include patients from 1–12 countries). Demographic (age, sex), clinical (MS phenotype, disability), and DMT (un- Podcast treated, alemtuzumab, cladribine, dimethyl fumarate, glatiramer acetate, interferon, natalizumab, Npub.org/Podcast9719 ocrelizumab, rituximab, siponimod, other DMTs) covariates were queried, along with COVID-19 COVID-19 Resources severity outcomes, hospitalization, intensive care unit (ICU) admission, need for artificial ven- For the latest articles, tilation, and death. Characteristics of outcomes were assessed in patients with suspected/ invited commentaries, and confirmed COVID-19 using multilevel mixed-effects logistic regression adjusted for age, sex, MS blogs from physicians phenotype, and Expanded Disability Status Scale (EDSS) score. around the world NPub.org/COVID19 *These authors contributed equally to this work as first authors. †These authors contributed equally to this work as senior authors. From the CORe (S.S.-Y., T.K.), Department of Medicine, and Neuroepidemiology Unit (S.S.-Y., N.N., A.F.), Melbourne School of Population & Global Health, University of Melbourne, Parkville, Australia; Menzies Institute for Medical Research (S.S.-Y.), University of Tasmania, Hobart, Australia; ESAT-STADIUS (E.D.B., Y.M., A.P.), KU Leuven, Belgium; Department of Neurology (T.K.), Melbourne MS Centre, Royal Melbourne Hospital, Parkville, Australia; MS International Federation (N.R., C.W.), London, UK; Department of Clinical Neuroscience (J.A.H., T.S., L.F., A.G.), Swedish MS Registry, Stockholm, Sweden; Department of Neurology (G.E.), CHU Pontchaillou, Rennes, France; Karolinska Institutet (T.S.), Solna, Sweden; Biomedical Research Institute–Data Science Institute (L.G., A.P., L.P.), Hasselt University, Belgium; Department of Medical Informatics (T.P.), University Medical Center Göttingen, Germany; Department of Computer Science and AI (C.G.), KU Leuven, Belgium; QMENTA (N.L.), Barcelona, Spain; Medpace Reference Laboratories (A.A.), Molecular Unit, Leuven, Belgium; iConquerMS People-Powered Research Network (R. McBurney, H.S.), Accelerated Cure Project for MS, Waltham, MA; NeuroTransData Study Group (A.B.B., S.B.), Neuro- TransData, Neuburg, Germany; German MS-Register by the National MS Society (A. Stahmann), MS Forschungs- und Projektentwicklungs-gGmbH, Hannover, Germany; UK MS Register (R. Middleton, R.S.N.), Swansea University, UK; COViMS (A. Salter, R.J.F.), US; Division of Biostatistics (A. Salter), Washington University in St. Louis, St Louis, MO; Mellen Center for Multiple Sclerosis (R.J.F.), Cleveland Clinic, Cleveland, OH; Department of Neuroscience (A.v.d.W., H.B.), Central Clinical School, Monash University, Melbourne, Australia; Al-Amiri Hospital (R.A.), Kuwait City, Kuwait; Dokuz Eylul University (S.O.), Izmir, Turkey; Neurology Department (J.I.R.), Hospital Universitario de CEMIC; RELACOEM (J.I.R., R.N.A.), Buenos Aires, Argentina; Australian MS Longitudinal Study (I.v.d.M.), Menzies Institute for Medical Research, University of Tasmania, Hobart, Australia; Bulgarian SmartMS COVID-19 Dataset (R.I.), Sofia, Bulgaria; ABEM–Brazilian MS Patients Association (G.S.d.O., A.E.D.), São Paulo, Brazil; Danish Multiple Sclerosis Registry (M.M.), Department of Neurology, University Hospital Rigshospitalet, Glostrup, Denmark; Universidade Estadual Paulista (D.B.), Unesp, Faculdade de Medicina, Botucatu, Brazil; REDONE.br–Brazilian Registry of Multiple Sclerosis and Neuromyelitis Optica Spectrum Disorders (D.B., M.F.M.); Irmandade da Santa Casa de Misericórdia de São Paulo (M.F.M.), Brazil; Multiple Sclerosis University Center (R.N.A.), Ramos Mejia Hospital–EMA, Buenos Aires, Argentina; Imperial College London (R.S.N.); Swansea University (R.S.N.), Swansea, UK; Mental Health Area (J.B.); MS and Demyelinating Diseases (A.S.C.)., Hospital Británico de Buenos Aires, EMA, Argentina; Servei de Neurologia-Neuroimmunologia (A.Z., G.A.), Centre d’Esclerosi Múltiple de Catalunya, (Cemcat); Vall d’Hebron Institut de Recerca (A.Z., G.A.), Vall d’Hebron Hospital Universitari; Universitat Autònoma de Barcelona (A.Z., G.A.), Spain; and Institute of Experimental Neurology (G.C.), Ospedale San Raffaele, Milan, Italy. 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 the authors. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. e1870 Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. Glossary aOR = adjusted OR; BMI = body mass index; CI = confidence interval; COVID-19 = coronavirus disease 2019; DMT = disease-modifying therapy; EDSS = Expanded Disability Status Scale; ICU = intensive care unit; MS = multiple sclerosis; MSDA = MS Data Alliance; OR = odds ratio; RRMS = relapsing-remitting MS; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2. Results Six hundred fifty-seven (28.1%) with suspected and 1,683 (61.9%) with confirmed COVID-19 were analyzed. Among suspected plus confirmed and confirmed-only COVID-19, 20.9% and 26.9% were hospitalized, 5.4% and 7.2% were admitted to ICU, 4.1% and 5.4% required artificial ventilation, and 3.2% and 3.9% died. Older age, progressive MS phenotype, and higher disability were associated with worse COVID-19 outcomes. Compared to dimethyl fumarate, ocrelizumab and rituximab were associated with hospitalization (adjusted odds ratio [aOR] 1.56, 95% confidence interval [CI] 1.01–2.41; aOR 2.43, 95% CI 1.48–4.02) and ICU admission (aOR 2.30, 95% CI 0.98–5.39; aOR 3.93, 95% CI 1.56–9.89), although only rituximab was associated with higher risk of artificial ventilation (aOR 4.00, 95% CI 1.54–10.39). Compared to pooled other DMTs, ocrelizumab and rituximab were associated with hospitalization (aOR 1.75, 95% CI 1.29–2.38; aOR 2.76, 95% CI 1.87–4.07) and ICU admission (aOR 2.55, 95% CI 1.49–4.36; aOR 4.32, 95% CI 2.27–8.23), but only rituximab was associated with artificial ventilation (aOR 6.15, 95% CI 3.09–12.27). Compared to natalizumab, ocrelizumab and rituximab were associated with hospitalization (aOR 1.86, 95% CI 1.13–3.07; aOR 2.88, 95% CI 1.68–4.92) and ICU admission (aOR 2.13, 95% CI 0.85–5.35; aOR 3.23, 95% CI 1.17–8.91), but only rituximab was associated with ventilation (aOR 5.52, 95% CI 1.71–17.84). Associations persisted on restriction to confirmed COVID-19 cases. No associations were observed between DMTs and death. Stratification by age, MS phenotype, and EDSS score found no indications that DMT associations with COVID-19 severity reflected differential DMT allocation by underlying COVID-19 severity. Discussion Using the largest cohort of people with MS and COVID-19 available, we demonstrated consistent associations of rituximab with increased risk of hospitalization, ICU admission, and need for artificial ventilation and of ocrelizumab with hospitalization and ICU admission. Despite the cross-sectional design of the study, the internal and external consistency of these results with prior studies suggests that rituximab/ocrelizumab use may be a risk factor for more severe COVID-19. Disease-modifying therapies (DMTs) that act by of hospitalization, although no associations with intensive care immunomodulatory/immunosuppressive mechanisms are a unit (ICU) admission/ventilation or death were seen.2 mainstay of treatment of multiple sclerosis (MS) but can in- crease infection susceptibility.1 Cross-sectional2 and cohort3-5 Large and geographically inclusive cohorts are required to assess studies suggest that comorbid conditions, age, sex, progressive the risk of severe COVID-19 for specific DMTs. Accordingly, MS phenotype, and higher disability increase risk for developing we established a global data-sharing initiative6 to investigate severe coronavirus disease 2019 (COVID-19).3,5 Some studies characteristics of COVID-19 severity in people with MS. We have also identified associations of certain DMT classes with hypothesized that older age, progressive MS phenotype, and COVID-19 severity.3 Using the French Covisep registry, the higher disability were associated with more severe COVID-19, authors assessed 347 patients with MS with suspected/ while immunosuppressive DMTs (alemtuzumab/cladribine/ confirmed COVID-19, finding those treated with DMTs with fingolimod/ocrelizumab/rituximab) would be deleterious, but a greater risk of systemic infection (alemtuzumab/cladribine/ those with less infection risk (interferons/glatiramer acetate) fingolimod/ocrelizumab/rituximab) had >4 times higher pro- would be associated with a less severe COVID-19. portions with severe COVID-19 than DMTs with no infection risk (interferon beta/glatiramer acetate).3 Using the MuSC-19 retrospective cohort study, authors assessed COVID-19 severity among 844 patients with MS with suspected/confirmed Methods COVID-19, finding that those treated with anti-CD20 DMTs Standard Protocol Approvals, Registrations, (ocrelizumab/rituximab) had 2.4 times higher risk of severe and Patient Consents COVID-19 compared to those treated with dimethyl fumarate.5 This study received approval from an ethics standards The COVID in MS (COViMS) Registry assessed characteristics committee on human experimentation (institutional or re- of COVID-19 severity among 1,626 people with MS with gional) for any experiments using human participants (ethics suspected/confirmed COVID-19, finding that, compared to the committee of Hasselt University, CME2020/025). Other untreated patients, ocrelizumab-treated (odds ratio [OR] 1.63) ethics information from data custodians includes the and rituximab-treated (OR 4.56) patients had higher frequencies following. Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1871 MSBase data are provided with the consent of individual par- Indeed, given the high proportion of patients with suspected ticipants and principal investigators at each MSBase participating and confirmed COVID-19, it is likely that a minority of pa- center. The German MS-Register was first approved by ethics tients at each center participated. committee of Julius-Maximilians-University of Würzburg (vote 142/12). After a switch was made to the web-based documen- Data were entered for a given participant once, but information tation system, further positive votes, for example, by the ethics for that participant could be reentered, and these reentered data committee of the Thuringia State chamber of physicians, fol- replaced the original record. This made for serial iterations of the lowed by several ethics committees of different universities, were analysis dataset, which were analyzed over time as the dataset given, and all patients signed an informed consent. expanded, thus allowing for assessment of temporal consistency of observed associations between the versions of the dataset. In Research participant protection was sought from the Wash- this fashion, if associations were erratic in their appearance be- ington University in St. Louis Institutional Review Board for tween iterations, this might suggest them to be statistical artifact, housing COViMS Registry data, who determined it to be not whereas consistency would indicate their veracity. human participants research and therefore exempt from active Institutional Review Board oversight at Washington Univer- To improve the quality of the data continuously over time, we sity in St. Louis and did not require patient consent. set up a data quality assessment and enhancement pipeline. This pipeline consists of 2 major parts: unambiguously de- The patient data sent to analyses resulting in the Associations fining new variables that are used in downstream analysis (e.g., of DMT Therapies With COVID-19 Severity in Multiple defining COVID-19 suspected and confirmed cases, catego- Sclerosis study originated from a study approved by the ethics rizing continuous variables to allow aggregation of the counts) committee of the Faculdade de Medicina de Botucatu, Uni- and predefining pass/fail criteria for variables (e.g., negative versidade Estadual Paulista under the internal review board ages, unrealistically high numbers for height). Variables that number CAAE 31021220.2.0000.5411. All participants signed fail were flagged, and registry custodians were contacted to a written informed consent form before enrollment. repair failed variables in the next upload. Simultaneously, failed variables were cleaned and preprocessed so that records The Centre d’Esclerosi Múltiple de Catalunya cohort study could be incorporated into the downstream analysis. was approved by the ethics committee of the Vall d’Hebron University Hospital (XMG-INT-2014-01), and all patients Variables signed an informed consent. Definitions for all terms were provided to data partners and were available on the MSDA platform. Clinicians made all Data from a core questionnaire on COVID-19 and relevant judgments regardless of how data were entered. This data demographic/clinical information were reported by treating dictionary is available by contacting the corresponding author. clinicians. The methods underlying the MS Data Alliance (MSDA) COVID-19 collaboration and measures thereof have COVID-19 status was defined as confirmed, based on a positive been described previously.6 Briefly, treating clinicians entered diagnostic test, or suspected, based on clinician judgment. information on a range of demographic, lifestyle, and MS- specific and COVID-19–specific clinical characteristics. Here, Hospitalization was queried as admission to hospital because only age, sex, MS phenotype, disability, DMTs, smoking, body of COVID-19 (suspicious) infection. ICU admission was mass index (BMI), comorbid conditions, COVID-19 status, queried as stay in ICU because of COVID-19 (suspicious) hospitalization, ICU admission, artificial ventilation, and death infection. Requiring artificial ventilation was queried as ven- are described. Study participation was restricted to patients tilation needed during hospital stay. Death due to COVID-19 with MS who were ≥18 years of age; however, this article is was queried as “did the patient die because of the (suspected) limited to those with suspected/confirmed COVID-19. COVID-19 infection?” Data were entered in 3 fashions: (1) direct entry to central Patient age was categorized into 3 age groups: 18 to 49, 50 to platform; (2) patient-level data sharing via participating 69, and ≥70 years. MS phenotype was grouped into relapsing- registries/cohorts, whereby MS registries and cohorts are remitting MS (RRMS) and progressive MS (secondary and regularly invited to share and upload their COVID-19 core primary progressive MS). dataset into the central data platform; and (3) aggregated data sharing via participating registries/cohorts, whereby some Disability was assessed by the Expanded Disability Status registries do not share data from individual patients but share Scale (EDSS)7 or Neurostatus.8 Disability was dichotomized aggregated results from specific queries. into scores of 0 to 6.0 and >6.0. Comorbid conditions were queried, including cardiovascular disease, hypertension, di- Multidimensional contingency tables from 12 different data abetes, chronic liver disease, kidney disease, other neurologic/ sources were merged, and then a combined anonymized neuromuscular disorder, lung disease, or malignant neoplasia. dataset was reconstructed. Not all patients at each contrib- BMI was categorized as nonobese (BMI ≤30 kg/m2) and uting data source necessarily participated in this study. obese (BMI >30 kg/m2). Current smoker status was queried as e1872 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N yes or no. Current DMT use was queried, including alemtu- zumab, cladribine, dimethylfumarate, fingolimod, glatiramer Table 1 Cohort Characteristics acetate, interferons, natalizumab, ocrelizumab, rituximab, Suspected and siponimod, teriflunomide, or other DMT, which was queried as confirmed Confirmed (n = 2,340), n (%) (n = 1,683), n (%) on another drug not listed. Due to patient numbers <20 among the suspected/confirmed COVID-19 cases, siponimod (n = Confirmed COVID-19 status 1,683 (71.9) 1,683 (100.0) 12) was aggregated with other DMT. Note that aggregation of Hospitalization 489 (20.9) 453 (26.9) siponimod with fingolimod did not materially change results (data not shown). ICU admission 127 (5.4) 122 (7.2) Ventilation 97 (4.1) 91 (5.4) Statistical Analysis Death 73 (3.2) 65 (3.9) Associations with hospitalization, ICU admission, ventilation, and death were assessed with multilevel mixed-effects logistic Female 1702 (72.7) 1,213 (72.1) regression, random effects grouped by data source, as uni- Age, y variable and adjusted for age, sex, MS phenotype, and disability. 18–<50 1,505 (64.3) 1,032 (61.3) Multilevel mixed-effects logistic regression is an appropriate methodology wherein multiple data sources with heterogeneity 50–<70 740 (31.6) 571 (33.9) in cohort characteristics or methods are included. This method ≥70 83 (3.5) 72 (4.3) applies fixed and random effects to the model, accounting for intraclass correlation for observations within data sources, RRMS phenotype 2,119 (80.4) 1,854 (80.5) providing aggregate statistics for the measures of association of EDSS score 0-6 1,837 (78.5) 1,275 (75.8) the independent model covariates. While all data sources used Has comorbid conditions 920 (40.7) 707 (43.2) the same data entry framework and core questionnaire, there are differences in the modes of clinical practice, for MS- and Obese BMI 437 (19.3) 376 (23.0) infection-related care, as well as clinical and demographic Current smoker 166 (7.3) 115 (7.0) characteristics, which need be accounted for and for which this statistical method is suited. DMT Untreated 284 (12.1) 231 (13.7) Subgroup analyses were also undertaken whereby data on Alemtuzumab 31 (1.3) 26 (1.5) comorbid condition, BMI, and smoking were available, allowing additional adjustment for these covariates. All anal- Cladribine 29 (1.2) 17 (1.0) yses were complete case. Dimethyl fumarate 275 (11.8) 199 (11.8) For DMTs, individual DMTs were first compared with di- Fingolimod 200 (8.5) 143 (8.5) methyl fumarate. Despite leading to lymphopenia in some Glatiramer acetate 87 (3.7) 70 (4.2) patients, dimethyl fumarate has not been associated with Interferon 124 (5.3) 85 (5.1) increased infection risk,9 and its biological mechanism of action is unlikely to interfere with anti–severe acute re- Natalizumab 221 (9.4) 164 (9.7) spiratory syndrome coronavirus 2 (SARS-CoV-2) immu- Ocrelizumab 471 (20.1) 365 (21.7) nologic response10 while being common in the sample. Of Rituximab 258 (11.0) 142 (8.4) note, interferons and glatiramer acetate were considered as potential comparators, but the lack of failures across all 4 Teriflunomide 97 (4.1) 75 (4.5) outcomes for these DMTs precluded their being the refer- Other DMT 70 (3.0) 47 (2.8) ence group. Next, ocrelizumab, rituximab, and untreated were compared against all other pooled DMTs. Finally, Abbreviations: BMI = body mass index; COVID-19 = coronavirus disease 2019; DMT = disease-modifying therapy, EDSS = Expanded Disability Status ocrelizumab and rituximab were evaluated vs natalizumab to Scale; ICU = intensive care unit; RRMS = relapsing-remitting multiple assess ascertainment bias because natalizumab-treated pa- sclerosis. tients present for infusions every 28 to 42 days compared to biannual infusions for anti-CD20 DMTs. Data analyses were carried out with STATA/SE 16.0 (Stata- Corp, College Station, TX). Adjustment for multiple comparisons was undertaken with the family-wise Holm step-down method, such that within Data Availability each hypothesis and within models 1 and 2, statistical tests Persons interested in acquiring the anonymized data un- were ranked by lowest p value and significance threshold derlying this analysis can inquire with a senior author (L.P.) to evaluated relative to the number of statistical tests within that make requests. In addition, supplementary tables not included family. Associations reaching significance after this adjust- in the main body text can be found in the preprint version of the ment are annotated as such in tables. article, accessible at doi.org/10.1101/2021.02.08.21251316. Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1873 Figure 1 Characteristics of Patients Treated With Ocrelizumab, Rituximab, or Other DMTs or Untreated, by Sex, Age, MS Phenotype, and Disability DMT = disease-modifying therapy; EDSS = Expanded Disability Status Scale; MS = multiple sclerosis. Results lower in sources 2, 4 through 7, 9, and 11; and death was higher in source 11 and lower in sources 2 and 4 through 9. The cohort comprised 2,340 patients, of whom 657 (28.1%) Results were comparable on restriction to confirmed-only had suspected COVID-19 and 1,683 (71.9%) had confirmed COVID-19 (data not shown). COVID-19. Among suspected/confirmed COVID-19 cases, which made up the primary analysis dataset, 20.9% were hos- Compared to dimethyl fumarate (76.1%), lower proportions of pitalized, 5.4% were admitted to ICU, 4.1% required artificial female patients were untreated (70.7%) or treated with in- ventilation, and 3.2% died. Proportions were slightly higher terferon (69.4%) or ocrelizumab (67.5%, eTable 1, links.lww. among confirmed COVID-19 cases (Table 1). com/WNL/B520). Larger proportions of those 50 to 69 and ≥70 years of age were untreated (42.8%, 13.4%) or treated with Cohort Characteristics ocrelizumab (34.2%, 1.5%), teriflunomide (49.5%, 3.1%), or Data sources were located in (1) Sweden (n = 290); (2) other DMTs (37.7%, 5.8%) than with dimethyl fumarate Australia, Belgium, Brazil, Kuwait, Romania, Saudi Arabia, and (29.9%, 0.8%). Greater proportions of patients with progressive Turkey (n = 97); (3) Argentina, Chile, Colombia, Ecuador, MS were untreated (44.9%) or treated with ocrelizumab Honduras, and Mexico (n = 159); (4) Bulgaria (n = 3); (5) (20.2%), rituximab (21.8%), or other DMTs (31.9%) than with Germany and Italy (n = 45); (6) Denmark (n = 56); (7) Brazil dimethyl fumarate (8.0%). Of patients with greater disability (n = 96); (8) Australia, Bahamas, Belgium, Czech Republic, (EDSS score >6), higher proportions were either untreated Finland, France, the Netherlands, New Zealand, Serbia, Spain, (35.9%) or treated with ocrelizumab (27.0%) or other DMTs the United Kingdom, and the United States (n = 114); (9) (34.8%) than with dimethyl fumarate (7.6%). Similar results Germany (n = 41); (10) the United States and Canada (n = were seen among confirmed-only COVID-19 (data not shown). 1,161); (11) the United Kingdom (n = 131); and (12) Spain (n = 147). Sources 2 and 10 had higher proportions with In evaluations of ocrelizumab and rituximab compared to confirmed COVID-19, and sources 4 and 5 had higher pro- pooled other DMTs and the untreated (Figure 1), anti-CD20 portions with nonsuspected COVID-19. Among suspected/ DMTs were comparable by sex and MS phenotype, but a confirmed COVID-19 cases, hospitalization was higher in greater proportion of ocrelizumab-treated patients were 50 to source 11 and lower in sources 2, 4, 5, and 7; ICU admission 69 years of age and had an EDSS score ≤6. Greater proportions was higher in sources 3, 9, and 10 and lower in sources 2, 4, 7, of anti-CD20–treated patients were of the progressive MS 9, 11, and 12; ventilation was higher in sources 3 and 8 and phenotype or had an EDSS score >6 than those treated with the e1874 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N Table 2 Characteristics of COVID-19 Severity Outcomes, DMTs vs Dimethyl Fumarate, Suspected Plus Confirmed COVID-19 Hospitalization ICU admission a n (%) OR (95% CI) aOR (95% CI) n (%) OR (95% CI) aOR (95% CI)a Untreated 97/266 (36.5) 3.34 (2.17, 5.16)b,c 1.79 (1.12, 2.87)c 17/261 (6.5) 2.62 (1.06, 6.48)c 1.36 (0.52, 3.55) Alemtuzumab 3/30 (10.0) 0.70 (0.20, 2.47) 0.86 (0.24, 3.07) 1/30 (3.3) 1.41 (0.16, 12.09) 1.93 (0.22, 16.77) Cladribine 2/27 (7.4) 0.60 (0.13, 2.72) 0.67 (0.14, 3.16) 0/27 (0.0) — — Dimethyl fumarate 38/268 (14.2) 1.00 (Ref) 1.00 (Ref) 7/259 (2.7) 1.00 (Ref) 1.00 (Ref) Fingolimod 16/195 (8.2) 0.60 (0.32, 1.11) 0.64 (0.34, 1.21) 4/193 (2.1) 0.80 (0.23, 2.79) 0.90 (0.25, 3.19) Glatiramer acetate 16/86 (18.6) 1.36 (0.71, 2.61) 1.06 (0.53, 2.10) 0/86 (0.0) — — Interferon 17/119 (14.3) 1.11 (0.59, 2.11) 0.89 (0.46, 1.70) 2/118 (1.7) 0.62 (0.13, 3.10) 0.54 (0.11, 2.73) Natalizumab 24/212 (11.3) 0.84 (0.48, 1.46) 0.82 (0.47, 1.45) 6/209 (2.9) 1.07 (0.35, 3.25) 1.11 (0.36, 3.43) Ocrelizumab 116/463 (25.1) 2.14 (1.41, 3.24)b,c 1.56 (1.01, 2.41)c 37/463 (8.0) 3.00 (1.30, 6.89)c 2.30 (0.98, 5.39) b,c b,c b,c Rituximab 70/252 (27.8) 3.12 (1.92, 5.07) 2.43 (1.48, 4.02) 24/251 (9.6) 4.63 (1.86, 11.49) 3.93 (1.56, 9.89)b,c Teriflunomide 14/93 (15.1) 1.21 (0.61, 2.38) 0.85 (0.42, 1.72) 4/93 (4.3) 1.59 (0.45, 5.59) 1.15 (0.32, 4.14) Other DMT 16/68 (23.5) 2.09 (1.07, 4.09)c 1.10 (0.54, 2.24) 4/67 (6.0) 2.34 (0.66, 8.33) 1.27 (0.34, 4.72) Ventilation Death a n (%) OR (95% CI) aOR (95% CI) n (%) OR (95% CI) aOR (95% CI)a Untreated 18/261 (6.9) 2.51 (1.00, 6.26)c 1.31 (0.49, 3.49) 27/261 (10.3) 5.69 (2.13, 15.19)b,c 1.64 (0.56, 4.80) Alemtuzumab 1/30 (3.3) 0.98 (0.11, 8.63) 1.27 (0.14, 11.34) 1/30 (3.3) 2.10 (0.23, 19.15) 2.11 (0.19, 23.29) Cladribine 0/27 (0.0) — — 0/27 (0.0) — — Dimethyl fumarate 7/259 (2.7) 1.00 (Ref) 1.00 (Ref) 5/259 (1.9) 1.00 (Ref) 1.00 (Ref) Fingolimod 5/193 (2.6) 0.76 (0.23, 2.50) 0.79 (0.24, 2.67) 0/193 (0.0) — — Glatiramer acetate 0/86 (0.0) — — 2/86 (2.3) 1.05 (0.20, 5.57) 0.56 (0.10, 3.25) Interferon 1/118 (0.8) 0.21 (0.02, 1.75) 0.18 (0.02, 1.52) 1/118 (0.8) 0.58 (0.07, 5.20) 0.31 (0.03, 2.82) Natalizumab 4/209 (1.9) 0.74 (0.21, 2.59) 0.72 (0.20, 2.55) 3/209 (1.4) 0.78 (0.18, 3.33) 0.72 (0.16, 3.30) Ocrelizumab 18/463 (3.9) 1.48 (0.60, 3.66) 1.04 (0.41, 2.64) 12/463 (2.6) 1.22 (0.42, 3.56) 0.49 (0.16, 1.52) Rituximab 24/251 (9.6) 5.11 (2.01, 12.99)b,c 4.00 (1.54, 10.39)c 7/251 (2.8) 2.61 (0.76, 8.96) 1.22 (0.35, 4.18) Teriflunomide 4/93 (4.3) 1.33 (0.37, 4.79) 0.83 (0.22, 3.10) 1/93 (1.1) 0.59 (0.07, 5.21) 0.27 (0.03, 2.53) Other DMT 2/67 (3.0) 1.12 (0.22, 5.71) 0.61 (0.12, 3.21) 4/67 (6.0) 3.82 (0.97, 15.00) 1.05 (0.25, 4.45) Abbreviations: aOR = adjusted OR; CI = confidence interval; COVID-19 = coronavirus disease 2019; DMT = disease-modifying therapy; EDSS = Expanded Disability Status Scale; ICU = intensive care unit; OR = odds ratio; Ref = referent. Analysis by multilevel mixed-effects logistic regression estimating OR (95% CI). a Multivariable models adjusted for age, sex, multiple sclerosis phenotype, and EDSS score. b Significant after family-wise Holm step-down multiple-comparisons adjustment. Other DMT was queried as on another drug not listed plus siponimod. c Statistical significance (p < 0.05). pooled other DMTs, while the untreated were typically older, associated with 61% and 92% greater risks of hospitalization had progressive MS phenotype, and had higher disability. and death, while older age was positively associated with hos- pitalization and death. Progressive MS phenotype was associ- COVID-19 Severity by Clinical/Demographic ated with 68% higher risk of hospitalization. Higher EDSS score Characteristics was associated with higher risks of all outcomes, including Demographic and clinical characteristics of hospitalization and 279% higher hospitalization, 211% higher ICU admission, ICU admission (eTable 1, links.lww.com/WNL/B520) and 281% higher ventilation, and 893% higher death frequencies. need for artificial ventilation and death (eTable 2, links.lww. Among confirmed-only COVID-19, all associations persisted com/WNL/B521) among suspected/confirmed COVID-19 except for age >70 years, which became nonsignificant (data cases were assessed. In multivariable models, male sex was not shown). Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1875 Figure 2 Risk of Hospitalization by DMT Among Suspected/Confirmed (Solid Line) and Confirmed-Only (Dashed Line) Cases Disease-modifying therapies (DMTs) compared to dimethyl fumarate adjusted for age, sex, multiple sclerosis phenotype, and Expanded Disability Status Scale score. Other DMTs also include siponimod. aOR = adjusted odds ratio; CI = confidence interval. *p < 0.05, **p < 0.001. In the subset of data sources with data (comorbid conditions rituximab consistently showed stronger associations with out- 84.2%, BMI 55.6%, smoking 79.3%), having one of the speci- comes than ocrelizumab for hospitalization and ICU admission fied comorbid conditions queried showed a positive trend with and was solely associated with requiring artificial ventilation increased risk of death (adjusted OR [aOR] 2.91), while obese (data not shown). BMI had a 152% higher risk of hospitalization, 312% higher risk of ICU admission, and 445% higher risk of requiring ventila- On adjustment for multiple comparisons, among suspected tion. Smoking was not associated with any outcomes. plus confirmed cases, rituximab associations with hospitaliza- tion and ICU admission remained significant, although that for Adjustment for multiple comparisons found most associations ventilation did not. However, among confirmed-only cases, the of sex, age, MS phenotype, and EDSS score with COVID-19 rituximab associations with hospitalization, ICU admission, outcomes persisted, as did BMI and comorbid conditions. and need for artificial ventilation remained significant after multiple-comparisons adjustment. The association of ocreli- COVID-19 Severity by DMT zumab with increased hospitalization did not remain significant Compared to dimethyl fumarate, rituximab use was associated in any analysis. with greater risks of hospitalization (aOR 2.43, 95% confi- dence interval [CI] 1.48, 4.02), ICU admission (aOR 3.93, COVID-19 Severity: Anti-CD20 DMTs vs Pooled 95% CI 1.56, 9.89), and artificial ventilation (aOR 4.00, 95% Other DMTs CI 1.54, 10.39, Table 2). Ocrelizumab showed similar asso- Compared to all other DMTs, those using rituximab had ciations for hospitalization (aOR 1.56, 95% CI 1.01, 2.41) and higher risks of hospitalization (aOR 2.76, 95% CI 1.87, 4.07, ICU admission (aOR 2.30, 95% CI 0.98, 5.39) but not arti- Table 3), ICU admission (aOR 4.32, 95% CI 2.27, 8.23), and ficial ventilation (aOR 1.04, 95% CI 0.41, 2.64). artificial ventilation (aOR 6.15, 95% CI 3.09, 12.27). Ocreli- zumab showed similar trends for hospitalization (aOR 1.75, No DMTs were associated with death. Untreated patients had 95% CI 1.29, 2.38) and ICU admission (aOR 2.55, 95% CI increased risk of hospitalization (aOR 1.79, 95% CI 1.12, 2.87), 1.49, 4.36) but not ventilation (aOR 1.60, 95% CI 0.82–3.14). but no independent associations with other outcomes were Neither rituximab (aOR 1.72, 95% CI 0.58, 5.10) or ocreli- seen. These associations persisted among confirmed-only zumab (aOR = 0.73, 95% CI = 0.32, 1.70) was associated with COVID-19 (Figures 2–4). Moreover, even on comparison to risk of death. Untreated patients had increased risks of hos- DMTs other than dimethyl fumarate as a post hoc analysis, pitalization (aOR 2.05, 95% CI 1.43, 2.94), ventilation (aOR e1876 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N Figure 3 Risk of ICU Admission by DMT Among Suspected/Confirmed (Solid Line) and Confirmed-Only (Dashed Line) Cases Disease-modifying therapies (DMTs) were compared to dimethyl fumarate, adjusted for age, sex, multiple sclerosis phenotype, and Expanded Disability Status Scale score. Other DMTs also include siponimod. Null set denotes analyses that could not be undertaken due to no events occurring in the exposed group. aOR = adjusted odds ratio; CI = confidence interval. *p < 0.05. 2.07, 95% CI 1.01.4.22), and death (aOR 2.53, 95% CI On adjustment for multiple comparisons, rituximab associa- 1.24.5.15). These results persisted among confirmed-only tions with hospitalization, ICU admission, and need for artifi- COVID-19 cases (data not shown). cial ventilation all persisted, while that for ocrelizumab persisted only for hospitalization. On restriction to confirmed- After adjustment for multiple comparisons, all rituximab only cases, this was also seen. and ocrelizumab associations persisted, while that for the untreated patients persisted for hospitalization and death. Stratification Sensitivity Analyses Among confirmed-only cases, rituximab associations To assess whether the associations seen for ocrelizumab and remained significant after multiple-comparisons adjust- rituximab were genuinely a function of the DMTs rather ment, while that for ocrelizumab persisted only for hos- than the characteristics of patients commonly treated with pitalization and for untreated for hospitalization and these medications (older, progressive MS, higher disabil- ventilation. ity), we undertook stratified analyses evaluating associations between DMTs and hospitalization, ICU admission, need COVID-19 Severity, Anti-CD20 DMTs vs for artificial ventilation, and death among persons >70 vs Natalizumab ≤70 years of age, among those with RRMS vs progressive, Compared to natalizumab, rituximab was associated with and among patients with EDSS scores ≤6 vs >6. By age, higher risks of hospitalization (aOR 2.88, 95% CI 1.68.4.92), ocrelizumab and rituximab associations with outcomes were ICU admission (aOR 3.23, 95% CI 1.17.8.91), and ventila- seen only among those ≤70 years of age. By MS phenotype, tion (aOR 5.52, 95% CI 1.71.17.84). Ocrelizumab showed ocrelizumab and rituximab consistently showed that asso- similar trends for hospitalization (aOR 1.86, 95% CI ciations with outcomes were roughly 2 to 3 times stronger in 1.13.3.07) but did not reach significance for ICU admission magnitude among those with the RRMS phenotype. By and was not associated with ventilation. Neither rituximab EDSS score, ocrelizumab and rituximab associations with (aOR 1.34, 95% CI 0.27.6.56) nor ocrelizumab (aOR 0.48, hospitalization were roughly comparable, but associations 95% CI 0.11.2.07) was associated with increased risk of with ICU admission and ventilation were 2 to 3 times death. Results were similar among confirmed-only COVID- stronger in magnitude among those of EDSS score ≤6. 19 (data not shown). These results indicate that the observed associations were a Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1877 Figure 4 Risk of Artificial Ventilation by DMT Among Suspected/Confirmed (Solid Line) and Confirmed-Only (Dashed Line) Cases Disease-modifying therapies (DMTs) were compared to dimethyl fumarate adjusted for age, sex, multiple sclerosis phenotype, and Expanded Disability Status Scale score. Other DMTs also includes siponimod. Null set denotes analyses that could not be undertaken due to no events occurring in the exposed group. aOR = adjusted odds ratio; CI = confidence interval. *p < 0.05. function of the DMT, not the underlying risk profile (data phenotype, or higher disability. We also found that older age, not shown). progressive MS phenotype, and higher disability were over- represented among patients with MS with more severe COVID-19. In subanalyses in which data were available, DMT associations were robust to further adjustment for comorbid Discussion conditions, BMI, and smoking status. Results were generally In the largest sample of people with MS with suspected and consistent on adjustment for multiple comparisons. confirmed COVID-19 to date, we demonstrated that the anti- CD20 DMTs rituximab and ocrelizumab were associated with An increasing array of highly effective DMTs with heteroge- more severe COVID-19. Compared to dimethyl fumarate, neous modes of effect have become available for people with MS pooled other DMTs, and natalizumab, anti-CD20 DMTs were but can increase infection risk.1,11 This has raised concern during associated with higher risks of hospitalization and ICU ad- the COVID-19 pandemic, and national studies have investigated mission, while only rituximab was associated with greater risk of risk factors for severe COVID-19 disease in people with MS. In requiring artificial ventilation. Comparison to natalizumab is the French Covisep study,3 of 347 patients with MS (42.1% particularly important, showing that anti-CD20 associations do confirmed COVID-19), older age, progressive MS phenotype, not likely reflect ascertainment bias. Regardless of comparator, higher disability, and comorbid conditions were associated rituximab consistently showed stronger associations with out- with COVID-19 severity. The authors’ pooled DMTs with comes than ocrelizumab for hospitalization and ICU admission moderate/high risk of systemic infection (fingolimod/ and was solely associated with requiring artificial ventilation. It ocrelizumab/rituximab/cladribine/alemtuzumab) were associ- is important to note that this was done as a post hoc explora- ated with 4.2 times higher COVID-19 severity score than DMTs tion, not an a priori plan, given the specific hypotheses of the with no systemic infection risk (interferon beta/glatiramer ace- study. At the same time, untreated patients had a significantly tate). This amalgamation of DMTs is a limitation because, while greater risk of death compared to persons treated with dimethyl comparable in terms of their infection risk, these DMTs have fumarate and pooled non–anti-CD20 DMTs, although this markedly different modes of action, especially in relation to the attenuated markedly on adjustment. Moreover, DMT associ- immunologic response to SARS-CoV-2.10 More recently, the ations were not merely driven by older age, progressive MS Italian MuSC-19 national registry study of 593 suspected and e1878 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N Table 3 Characteristics of COVID-19 Severity Outcomes, Anti-CD20 DMTs vs All Other DMTs, and Anti-CD20 DMTs vs Natalizumab, Suspected Plus Confirmed COVID-19 Hospitalization ICU admission a n (%) OR (95% CI) aOR (95% CI) n (%) OR (95% CI) aOR (95% CI)a Pooled other DMT 144/1,082 (13.3) 1.00 (Ref) 1.00 (Ref) 28/1,082 (2.6) 1.00 (Ref) 1.00 (Ref) Ocrelizumab 116/463 (25.1) 2.15 (1.62, 2.87)b,c 1.75 (1.29, 2.38)b,c 37/463 (8.0) 3.08 (1.83, 5.17)b,c 2.55 (1.49, 4.36)b,c Rituximab 70/251 (27.9) 3.16 (2.17, 4.61)b,c 2.76 (1.87, 4.07)b,c 24/251 (9.6) 4.69 (2.49, 8.83)b,c 4.32 (2.27, 8.23)b,c No DMT 97/261 (37.2) 3.49 (2.54, 4.80)b,c 2.05 (1.43, 2.94)b,c 17/261 (6.5) 2.69 (1.43, 5.04)b,c 1.52 (0.77, 3.02) Natalizumab 24/209 (11.5) 1.00 (Ref) 1.00 (Ref) 6/209 (2.9) 1.00 (Ref) 1.00 (Ref) Ocrelizumab 116/463 (25.1) 2.64 (1.64, 4.25)b,c 1.86 (1.13, 3.07)b,c 37/463 (8.0) 2.89 (1.18, 7.05)b,c 2.13 (0.85, 5.35) b,c b,c b,c Rituximab 70/251 (27.9) 3.16 (1.89, 5.26) 2.88 (1.68, 4.92) 24/251 (9.6) 4.17 (1.55, 11.18) 3.23 (1.17, 8.91)b,c Ventilation Death a n (%) OR (95% CI) aOR (95% CI) n (%) OR (95% CI) aOR (95% CI)a Pooled other DMT 24/1,082 (2.2) 1.00 (Ref) 1.00 (Ref) 17/1,082 (1.6) 1.00 (Ref) 1.00 (Ref) Ocrelizumab 18/463 (3.9) 1.99 (1.04, 3.80)b,c 1.60 (0.82, 3.14) 12/463 (2.6) 1.38 (0.65, 2.96) 0.76 (0.34, 1.69) Rituximab 24/251 (9.6) 6.95 (3.54, 13.64)b,c 6.15 (3.09, 12.27)b,c 7/251 (2.8) 2.94 (1.12, 7.72)c 1.90 (0.73, 4.93) b,c c b,c No DMT 18/261 (6.9) 3.42 (1.78, 6.57) 2.07 (1.01, 4.22) 27/261 (10.3) 6.47 (3.40, 12.33) 2.53 (1.24, 5.15)b,c Natalizumab 4/209 (1.9) 1.00 (Ref) 1.00 (Ref) 3/209 (1.4) 1.00 (Ref) 1.00 (Ref) Ocrelizumab 18/463 (3.9) 2.11 (0.69, 6.39) 1.34 (0.42, 4.24) 12/463 (2.6) 1.64 (0.45, 5.96) 0.53 (0.13, 2.24) b,c b,c Rituximab 24/251 (9.6) 7.52 (2.37, 23.81) 5.52 (1.71, 17.84) 7/251 (2.8) 2.96 (0.66, 13.29) 1.70 (0.38, 7.62) Abbreviations: aOR = adjusted OR; CI = confidence interval; COVID-19 = coronavirus disease 2019; DMT = disease-modifying therapy; EDSS = Expanded Disability Status Scale; ICU = intensive care unit; OR = odds ratio; Ref = referent. Analysis by multilevel mixed-effects logistic regression estimating OR (95% CI). Other DMT was queried as on another drug not listed, as well as aggregated with Siponimod, due to infrequency of persons on this DMT. a Multivariable models adjusted for age, sex, multiple sclerosis phenotype, and EDSS score. b Significant after family-wise Holm step-down multiple-comparisons adjustment. c Statistical significance (p < 0.05). 191 confirmed COVID-19 cases found that anti-CD20 DMT natalizumab. Dimethyl fumarate was identified as a suitable use was associated with 2.6 times greater risk of severe COVID- comparator, being common in the sample and also used for 19 compared to dimethyl fumarate use, adjusted for region, age, the MuSC-19 study.5 The untreated were not regarded as an sex, MS phenotype, and recent methylprednisolone use.5 Finally, appropriate primary comparator because these patients dif- the COViMS study found that, compared to untreated patients, fered markedly from the rest of the cohort in age, disability, rituximab- and ocrelizumab-treated patients had 4.6 and 1.6 and MS phenotype. This latter comparator is of particular times greater odds of hospitalization, although no significant importance because it assesses ascertainment bias as a po- associations were seen for other outcomes.2 This disparity with tential explanation for the associations seen for anti-CD20 our and previous results may reflect their use of the untreated as DMTs. Both the anti-CD20 DMTs and natalizumab require the comparator, as well as differences in the multivariable model patients to come in for DMT infusion at 6- and 3-month covariates. It should be acknowledged that the COViMS sample intervals, respectively. Thus, if the anti-CD20 DMT associa- comprised a meaningful proportion of our total sample (n = tions with COVID-19 severity were merely a function of more 1,161 of 2,340, 49.6%), and while the total cohort analyzed and regular hospital attendance and thus potential for COVID-19 analysis methods used here differ, results for this cohort have symptoms to be identified and treatment initiated, then there been described previously. The remainder of the cohort de- should be no difference from natalizumab. In fact, we found scribed here has not been previously described. that the associations of ocrelizumab and rituximab with hos- pitalization and ICU admission and rituximab with need for Indeed, the issue of comparator is a point of particular at- artificial ventilation persist. tention in our study; we compared individual DMTs to di- methyl fumarate, and then, because they showed significant Untreated patients showed consistent positive trends toward associations here, the anti-CD20 DMTs were compared associations with hospitalization, ICU admission, and re- against all other DMTs, and finally these were compared to quiring ventilation, albeit attenuating on adjustment for age, Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1879 sex, MS phenotype, and disability. This is in keeping with prior Thus, we could not assess other clinical features, particularly results. Louapre et al.3 found higher frequencies of severe prior MS clinical course and DMT use, paraclinical in- COVID-19 among the untreated vs treated (46.0% vs 15.5%), formation such as radiologic burden of MS, or the nuanced although this difference did not persist on adjustment. Sormani details of COVID-19 onset and evolution. Another limitation et al.5 compared untreated to dimethyl fumarate-treated pa- of our data is that they likely comprise greater proportions of tients, finding that they were 2.83 times more likely to have severe cases requiring medical attention. One particular ele- severe COVID-19, although this disappeared on adjustment ment lacking in our data is treatment duration or duration (aOR 1.04). The lack of independence of the untreated asso- since treatment; both may have bearing on the degree of B-cell ciations here and previously likely reflects the untreated com- depletion and thus on COVID-19 severity. This information prising, to variable degrees, people with a more benign MS was included in the core questionnaire, but the level of miss- course or other reasons not to use DMTs, so adjustment for ingness was too high to be a component of analyses. These MS phenotype and disability largely captures differences in data would have been useful in better describing and explaining COVID-19 severity. That said, it is important to acknowledge the differences in the associations between rituximab and that the heterogeneity of this patient population may extend ocrelizumab, but unfortunately, this information was not beyond differences in clinical phenotype and likely includes a available here. In addition, data on steroid use and other DMTs range of patient and region-level idiosyncrasies for which we that might have affected clinical progression or resulted in less were unable to account. severe COVID-19 were not available. The directions of effect in the associations of rituximab and Heterogeneity in the definitions of exposure and outcomes and ocrelizumab were consistent for the hospitalization and ICU in patient inclusion among the data sources is a known problem admission outcomes, although those for rituximab were in combining multiple data sources. Related to this are the dif- stronger and only rituximab showed an association with re- ferences in protocols for hospital and ICU admission and ini- quiring artificial ventilation. It is possible that differences in tiation of artificial ventilation between hospitals, as well as biology, due to differences in provenance or affinity for the differences in the availability/use of DMTs between countries.15 CD20 protein12,13 or differences in mechanisms of cytotox- To ensure that our results were not being driven by single icity,14 may underlie some of this difference. More likely, these influential data sources, we undertook all analyses using differences represent unmeasured confounding because the random-effects logistic regression, as well as serial-exclusion dataset, while large, was limited in the number of character- sensitivity analyses. These analyses showed that, while there was istics assessed, so potentially relevant factors like socioeco- some variation in the magnitudes and significance of associa- nomic status and access to care or factors affecting respiratory tions, trends tracked as seen for the whole cohort, indicating that health could not be assessed. In addition, parameters such as the results were not driven by a specific data source. time on treatment or serum immunoglobulin load, which would have been valuable to explore for explanations for Another issue lies in the anonymous nature of the data entry, differences in associations between these 2 DMTs, were not such that patients may be entered more than once in different available. That said, there was general internal consistency, data sources. We are unable to account for whether partici- and our results are broadly in line with those seen in other pants already participated or had their data entered in another studies,3,5 providing external consistency. This preliminary study because there is no identifying information to assess this observation is worth exploring in laboratory studies. or any query of prior participation in the survey. That anti-CD20 DMTs were not associated with death con- Another issue is the nature of the data aggregation, with some flicts with the results seen for the other outcomes, as well as data sources providing individual patient-level data but others with the MuSC-19 study, which found a positive trend be- only tabulations of discrete categorical terms. Thus, we were tween anti-CD20 DMTs and death. The issue may lie in obliged to use 3-level categories of age, 2-level EDSS scores, ascertainment bias, with fewer of the older patients included and 2-level BMIs rather than more exact values of each or in our sample: we had only 9.1% of confirmed COVID-19 2-level MS phenotype (RRMS/progressive) rather than in- cases >60 years of age vs 17.7% in the MuSC-19 cohort.5 It is dividual MS phenotype (RRMS/secondary progressive MS/ also possible that the infrequency of deaths in our cohort (n = primary progressive MS). That said, these levels are generally 73, 3.2%) may have limited our statistical power to assess this aligned with the levels of each associated with increased relationship. That said, our number of deaths was actually COVID-19 severity. greater than in previous studies (n = 12–54), and we did demonstrate similar deleterious associations of older age and This data aggregation process and resultant limitation of the higher disability with death seen previously.2,3,5 The potential number of dimensions of model covariates thus required a impacts of these DMTs on death caused by COVID-19 somewhat simplified definition of underlying clinical severity should be further explored. as used in the stratified analyses by age, MS phenotype, and disability. We are also limited in not having additional clinical In contrast to prior clinic-based studies, our cohort focused on a detail such as relapse activity or MRI parameters, which would predefined limited set of demographic and clinical characteristics.6 have allowed greater depth of assessment of underlying e1880 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N disease risk phenotype. That said, older age, progressive MS Study Funding phenotype, and higher disability have been consistently The authors disclosed receipt of the following financial support demonstrated as risk factors for COVID-19 severity across for the research, authorship, and/or publication of this article. our results and previous studies,2,3,5 so if the DMT associa- The operational costs linked to this study are funded by tions were merely a function of underlying clinical suscepti- the Multiple Sclerosis International Federation (MSIF) and the bility, it should be expected that analyses stratified by these Multiple Sclerosis Data Alliance (MSDA), acting under the covariates would show some signal. That they did not is in- umbrella of the European Charcot Foundation. The MSDA dicative of a true association. receives income from a range of corporate sponsors, recently including Biogen, Bristol-Myers Squibb (formerly Celgene), The information collected through the different sources does Canopy Growth Corp, Genzyme, Icometrix, Merck, Mylan, not provide us with detailed information about validation of Novartis, QMENTA, Quanterix, and Roche. MSIF receives the COVID-19 diagnosis. Whether patients remained in the income from a range of corporate sponsors, recently including suspected group because of discrepancy between clinical and Biogen, Bristol-Myers Squibb (formerly Celgene), Genzyme, laboratory assessment or other reasons is unknown. Because Med-Day, Merck, Mylan, Novartis, and Roche. This work was of these diagnostic uncertainties, we opted to perform 2 supported by the Flemish government under the Onderzoek- analyses, 1 among group with suspected and confirmed sprogramma Artificiële Intelligentie Vlaanderen programme COVID-19 and 1 in confirmed COVID-19 only. and the Research Foundation Fladers (FWO) for ELIXIR Belgium–Flanders (FWO) for ELIXIR Belgium. The central Another potential issue is the representativeness of the in- platform was provided by QMENTA, and the computational cluded patients from each of the noncentral data sources. resources used in this work were provided by Amazon. The While there were no inclusion or exclusion criteria beyond the statistical analysis was carried out at CORe, The University of patients needing to have MS and be at least 18 years of age, the Melbourne, with support from the National Health and clinicians entering data for their patients would likely bias to Medical Research Council (NHMRC; 1129189 and 1140766). enter data on patients who had suspected or confirmed COVID-19; this was reflected in the low proportion of the Disclosure sample without suspected/confirmed COVID-19 (data not T. Kalincik has served on scientific advisory boards for Roche, shown). However, such bias is typical of all the clinic-based Sanofi-Genzyme, Novartis, Merck, and Biogen; served on a studies of this sort. Our cohort characteristics are typical of steering committee for Brain Atrophy Initiative by Sanofi- people with MS, including female, RRMS phenotype, and low Genzyme; received conference travel support and/or speaker disability preponderance, so the extent to which this gener- honoraria from WebMD Global, Novartis, Biogen, Sanofi- alizability may affect the analyses here is likely minimal. Genzyme, Teva, BioCSL, and Merck; and received research support from Biogen. N. Rijke and C. Walton have no personal The fact that inverse trends were seen for smoking status and pecuniary interests to disclose other than being employees of COVID-19 is puzzling and not in line with expectations of a MSIF, which receives income from a range of corporate deleterious association with a respiratory condition. The fre- sponsors, recently including Biogen, BristolMyersSquibb quency of current smokers in the cohort is low compared to other (formerly Celgene), Genzyme, Med-Day, Merck, Mylan, cohorts,16-18 which may account for the absence of associations. Novartis, and Roche. J. Hillert has received honoraria for In addition, smoking status was not available from all data sour- serving on advisory boards for Biogen, Celgene, Sanofi- ces, so a material proportion (20.7%) of the sample had smoking Genzyme, Merck KGaA, Novartis, and Sandoz and speaker fees status missing. The Covisep study3 also found an inverse trend from Biogen, Novartis, Merck KGaA, Teva, and Sanofi- with COVID-19 severity, and the COViMS study2 found no Genzyme; he has served as principal investigator for projects or associations. The explanation for this absence of a deleterious received unrestricted research support from Biogen, Celgene, impact of smoking on COVID-19 severity may bear exploration. Merck KGaA, Novartis, Roche, and Sanofi-Genzyme; and his MS research was funded by the Swedish Research Council and In the largest population yet studied, we have shown that the Swedish Brain foundation. G. Edan has received patients with MS treated with the anti-CD20 DMTs rituximab consulting/speaking fees and research support from Bayer, and ocrelizumab are at higher risk of more severe COVID-19 Novartis, Teva, Sanofi Genzyme, Merck Serono, Biogen Idec, compared to those treated with dimethyl fumarate, pooled and Roche. T. Spelman has served on scientific advisory boards other DMTs, and natalizumab. This risk is additional to the for Biogen. R. McBurney and H. Schmidt work for the Accel- risk associated with demographic and clinical characteristics, erated Cure Project for MS (ACP), which has received grants, with older age, progressive MS phenotype, and higher dis- collaboration funding, payments for use of assets, or in-kind ability all showing deleterious relationships with COVID-19 contributions from the following companies: EMD Serono, severity. These results agree with smaller cohort studies and Sanofi/Genzyme, Biogen, Genentech, AbbVie, Octave, Gly- suggest that the risk vs benefit of continued or new exposure coMinds, Pfizer, MedDay, AstraZeneca, Teva, Mallinckrodt, to CD20-depleting treatment strategies compared to other MSDx, Regeneron Genetics Center, BC Platforms, and Cel- DMTs needs to be considered in the context of the ongoing gene. ACP has also received funding from the Patient-Centered COVID-19 pandemic. Outcomes Research Institute and the National MS Society. A. Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1881 Glaser has received research support from Novartis. R. ECTRIMS; speaking honoraria from Sanofi and Merck; and is McBurney has received consulting payments from EMD a member of the International Women in Multiple Sclerosis Serono, which have been donated to ACP. A. Bergmann has network executive committee. G. Comi has received consulting received consulting fees from and is an advisory board/ and speaking fees from Novartis, Teva Pharmaceutical In- speaker/other activities for NeuroTransData, and has worked dustries Ltd, Teva Italia Srl, Sanofi Genzyme, Genzyme Corp, on project management/clinical studies for and received travel Genzyme Europe, Merck KGgA, Merck Serono SpA, Celgene expenses from Novartis and Servier. A. Stahmann has no per- Group, Biogen Idec, Biogen Italia Srl, F. Hoffman-La Roche, sonal pecuniary interests to disclose other than being the lead Roche SpA, Almirall SpA, Forward Pharma, Medday, and of the German MS-Registry, which receives (project) funding Excemed. L.M. Peeters has no personal pecuniary interests to from a range of public and corporate sponsors, recently in- disclose other than being the chair of the MSDA, which re- cluding the German Innovation Fund, the German MS Trust, ceives income from a range of corporate sponsors, recently Biogen, German MS Society, Celgene (BMS), Merck, Novar- including Biogen, BristolMyersSquibb (formerly Celgene), tis, Roche, and Sanofi. R. Middleton has received no personal Canopy Growth Corp, Genzyme, Icometrix, Merck, Mylan, funding from any sources but works for the UK MS Register, Novartis, QMENTA, Quanterix, Roche. S. Simpson-Yap, E. De which is funded by the UK MS Society and has received Brouwer, Y. Moreau, L. Geys, T. Parciak, C. Gautrais, N. funding for specific projects from Novartis, Sanofi-Genzyme, Lazovski, A. Pirmani, A. Ardeshirdavanai , L.E. Forsberg, S. and Merck KGaA. R.J. Fox has received personal consulting Braune, A. Salter, S. Ozakbas, I. van der Mei, N. Nag, R. Ivanov, fees from Actelion, Biogen, Celgene, EMD Serono, Genentech, A. Estavo Dias, D. Guimarães Brum, M.F. Mendes, J. Bauer, Immunic, Novartis, Sanofi, Teva, and TG Therapeutics; has A.S. Chertcoff, and A. Fidao report no disclosures relevant to served on advisory committees for Actelion, Biogen, Immunic, the manuscript. Go to Neurology.org/N for full disclosures. and Novartis; and has received clinical trial contract and re- search grant funding from Biogen and Novartis. A. van der Walt Publication History has received honoraria and unrestricted research funding from This manuscript was prepublished in MedRXiv: doi.org/10. Novartis, Biogen, Roche, Merck, and Sanofi. H. Butzkueven’s 1101/2021.02.08.21251316. Received by Neurology institution receives compensation for Advisory Board, Steering February 12, 2021. Accepted in final form August 27, 2021. Committee, and educational activities from Biogen, Roche, Merck, and Novartis. His institution receives research support from Roche, Novartis, Biogen, NHMRC, MRFF Australia, MS Research Australia, and the Trish MS Foundation. He receives Appendix Authors personal compensation from Oxford HPF for serving on the Name Location Contribution steering group of MS Brain Health. R. Alroughani received honoraria as a speaker for and for serving on scientific advisory Steve Simpson- CORe, Department of Drafting/revision of the Yap, PhD, MPH Medicine, The University manuscript for content, boards from Bayer, Biogen, GSK, Merck, Novartis, Roche, and of Melbourne, Australia; including medical writing Sanofi-Genzyme. J.I. Rojas has received honoraria from Neuroepidemiology Unit, for content; major role in Melbourne School of the acquisition of data; Novartis as a scientific advisor and has received travel grants Population & Global study concept or design; and attended courses and conferences on behalf of Merck- Health, The University of analysis or interpretation Melbourne, Australia; of data Serono Argentina and Novartis Argentina. G. Sciascia do Olival Menzies Institute for has received honoraria for lecturing and support for congress Medical Research, University of Tasmania, participation from Biogen, Merck, Novartis, Sanofi/Genzyme, Australia EMS, and Roche. M. Magyari has served on scientific advisory Edward De ESAT-STADIUS, KU Drafting/revision of the board for Biogen, Sanofi, Roche, Novartis, Merck, and Abbvie; Brouwer, MSc Leuven, Belgium manuscript for content, has received honoraria for lecturing from Biogen, Merck, including medical writing Novartis, Sanofi, and Genzyme; and has received research for content; major role in the acquisition of support and support for congress participation from Biogen, data; study concept Genzyme, Roche, Merck, and Novartis. R. Alonso has received or design; analysis or interpretation honoraria from Novartis as a scientific advisor and received of data travel grants and attended courses and conferences on behalf of Tomas Kalincik, CORe, Department of Drafting/revision Merck-Serono Argentina, Biogen Argentina, Genzyme Argen- MD, PhD, Medicine, The University of the manuscript for tina, Roche Argentina, and Novartis Argentina. R. Nicholas has PGCertBiostat of Melbourne, Australia; content, including Melbourne MS Centre, medical writing for received honoraria from Novartis, Roche, and Biogen for ad- Department of Neurology, content; study concept visory boards. A. Zabalza has received travel expenses for sci- Royal Melbourne or design; analysis Hospital, Australia or interpretation entific meetings from Biogen, Novartis, and Genzyme; of data speaking honoraria from Eisai; and a study grant from Novartis. Nick Rijke, MA MS International Drafting/revision of the G. Arrambide has received compensation for consulting ser- Federation, UK manuscript for content, vices or participation in advisory boards from Sanofi, Merck, including medical writing for content; study concept and Roche; research support from Novartis; travel expenses for or design scientific meetings from Novartis, Roche, Stendhal, and e1882 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N Appendix (continued) Appendix (continued) Name Location Contribution Name Location Contribution Jan A Hillert, MD, Department of Clinical Drafting/revision of the Lars Forsberg, Department of Clinical Drafting/revision of the PhD Neuroscience, Swedish manuscript for content, MSc Neuroscience, Swedish manuscript for content, MS Registry, Sweden including medical writing MS Registry, Sweden including medical writing for content; major role in for content; study concept the acquisition of data; or design study concept or design; analysis or interpretation Anna Glaser, PhD Department of Clinical Drafting/revision of the of data Neuroscience, Swedish manuscript for content, MS Registry, Sweden including medical writing Clare Walton, MS International Drafting/revision of the for content; major role in PhD Federation, UK manuscript for content, the acquisition of data; including medical writing analysis or interpretation for content; study concept of data or design Robert iConquerMS People- Drafting/revision of the Gilles Edan, MD Department of Neurology, Drafting/revision of the McBurney, PhD Powered Research manuscript for content, CHU Pontchaillou, France manuscript for content, Network, Accelerated including medical writing including medical writing Cure Project for MS, for content; major role in for content; study concept Waltham, MA the acquisition of data or design; analysis or interpretation of data Hollie Schmidt, iConquerMS People- Drafting/revision of the MS Powered Research manuscript for content, Yves Moreau, ESAT-STADIUS, KU Drafting/revision of the Network, Accelerated including medical writing PhD Leuven, Belgium manuscript for content, Cure Project for MS, for content; major role in including medical writing Waltham, MA the acquisition of data for content; major role in the acquisition of data; Arnfin B. NeuroTransData Study Drafting/revision of the study concept or design; Bergmann, MD, Group, NeuroTransData, manuscript for content, analysis or interpretation PhD Germany including medical writing of data for content; major role in the acquisition of data Tim Spelman, Department of Clinical Drafting/revision of the PhD Neuroscience, Swedish manuscript for content, Stefan Braune, NeuroTransData Study Drafting/revision of the MS Registry, Sweden; including medical PhD Group, NeuroTransData, manuscript for content, Karolinska Institutet, writing for content; Germany including medical writing Sweden study concept or design; for content; major role in analysis or interpretation the acquisition of data of data Alexander German MS-Register by Drafting/revision of the Lotte Geys, PhD Biomedical Research Drafting/revision of the Stahmann, MSc the National MS Society, manuscript for content, Institute–Data Science manuscript for content, MS Forschungs- und including medical writing Institute, Hasselt including medical writing Projektentwicklungs- for content; major role in University, Belgium for content; study concept gGmbH, Germany the acquisition of data or design Rodden UK MS Register, Swansea Drafting/revision of the Tina Parciak, MSc Department of Medical Drafting/revision of the Middleton, MBA University, UK manuscript for content, Informatics, University manuscript for content, including medical writing Medical Center Göttingen, including medical writing for content; major role in Germany for content; study concept the acquisition of data or design Amber Salter, COViMS, US; Division of Drafting/revision of the Clement Department of Computer Drafting/revision of the PhD Biostatistics, Washington manuscript for content, Gautrais, PhD Science and AI, KU Leuven, manuscript for content, University in St. Louis, MO including medical writing Belgium including medical writing for content; major role in for content; study concept the acquisition of data or design Robert J. Fox, MD COViMS, US; Mellen Drafting/revision of the Nikola Lazovski, QMENTA, Spain Drafting/revision of the Center for Multiple manuscript for content, PDEng manuscript for content, Sclerosis, Cleveland Clinic, including medical writing including medical writing OH for content; major role in for content; study concept the acquisition of data or design Anneke van der Department of Drafting/revision of the Ashkan Pirmani, Medpace Reference Drafting/revision of the Walt, PhD Neuroscience, Central manuscript for content, MSc Laboratories, Molecular manuscript for content, Clinical School, Monash including medical writing Unit, Leuven, Belgium including medical writing University, Australia for content; major role in for content; study concept the acquisition of data or design Helmut Department of Drafting/revision of Amin Medpace Reference Drafting/revision Butzkueven, PhD Neuroscience, Central the manuscript for Ardeshirdavanai, Laboratories, Molecular of the manuscript for Clinical School, Monash content, including PhD Unit, Leuven, Belgium content, including University, Australia medical writing for medical writing for content; major role content; study in the acquisition concept or design of data Continued Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1883 Appendix (continued) Appendix (continued) Name Location Contribution Name Location Contribution Raed Alroughani, Al-Amiri Hospital, Kuwait Drafting/revision of the Ricardo N. RELACOEM, Argentina; Drafting/revision of the PhD manuscript for content, Alonso, MD, MSc Multiple Sclerosis manuscript for content, including medical writing University Center, Ramos including medical writing for content; major role in Mejia Hospital–EMA, for content; major role in the acquisition of data Argentina the acquisition of data Serkan Ozakbas, Dokuz Eylul University, Drafting/revision of the Richard S. UK MS Register, Swansea Drafting/revision of the PhD Turkey manuscript for content, Nicholas, PhD University, UK; Imperial manuscript for content, including medical writing College London, UK; including medical writing for content; major role in Swansea University, UK for content; major role in the acquisition of data the acquisition of data Juan I. Rojas, MSc Neurology Department, Drafting/revision of the Johana Bauer, Mental Health Area, EMA, Drafting/revision of the Hospital Universitario de manuscript for content, PhD Argentina manuscript for content, CEMIC, Argentina; including medical writing including medical writing RELACOEM, Argentina for content; major role in for content; major role in the acquisition of data the acquisition of data Ingrid van der Australian MS Drafting/revision of the Anı́bal Sebastián MS and Demyelinating Drafting/revision of the Mei, PhD Longitudinal Study, manuscript for content, Chertcoff, MD Diseases. Hospital manuscript for content, Menzies Institute for including medical writing Británico de Buenos Aires, including medical writing Medical Research, for content; major role in EMA, Argentina for content; major role in University of Tasmania, the acquisition of data; the acquisition of data Australia analysis or interpretation of data Anna Zabalza, MD Servei de Neurologia- Drafting/revision of the Neuroimmunologia, manuscript for content, Nupur Nag, PhD Neuroepidemiology Unit, Drafting/revision of the Centre d’Esclerosi including medical writing Melbourne School of manuscript for content, Múltiple de Catalunya, for content; major role in Population & Global including medical writing (Cemcat); Vall d’Hebron the acquisition of data Health, The University of for content; major role in Institut de Recerca, Vall Melbourne, Australia the acquisition of data d’Hebron Hospital Universitari; Universitat Rumen Ivanov, Bulgarian SmartMS Drafting/revision of the Autònoma de Barcelona, MSc COVID-19 Dataset, manuscript for content, Spain Bulgaria including medical writing for content; major role in Georgina Servei de Neurologia- Drafting/revision of the the acquisition of data Arrambide, MD, Neuroimmunologia, manuscript for content, PhD Centre d’Esclerosi including medical writing Guilherme ABEM–Brazilian MS Drafting/revision of the Múltiple de Catalunya, for content; major role in Sciascia do Olival, Patients Association, manuscript for content, (Cemcat); Vall d’Hebron the acquisition of data PhD Brazil including medical writing Institut de Recerca, Vall for content; major role in d’Hebron Hospital the acquisition of data Universitari; Universitat Autònoma de Barcelona, Alice Estavo Dias, ABEM–Brazilian MS Drafting/revision of the Spain PhD Patients Association, manuscript for content, Brazil including medical writing Alexander Fidao, Neuroepidemiology Unit, Drafting/revision of the for content; major role in MPH Melbourne School of manuscript for content, the acquisition of data Population & Global including medical writing Health, The University of for content; analysis or Melinda Magyari, Danish Multiple Sclerosis Drafting/revision of the Melbourne, Australia interpretation of data PhD Registry, Department of manuscript for content, Neurology, University including medical writing Giancarlo Comi, Institute of Experimental Drafting/revision of the Hospital Rigshospitalet, for content; major role in MD Neurology, Ospedale San manuscript for content, Denmark the acquisition of data Raffaele, Italy including medical writing for content; major role in Doralina Brum, Universidade Estadual Drafting/revision of the the acquisition of data; PhD Paulista, Unesp, manuscript for content, study concept or design Faculdade de Medicina, including medical writing Botucatu, Brazil; for content; major role in Liesbet Peeters, Biomedical Research Drafting/revision of the REDONE.br–Brazilian the acquisition of data PhD Institute–Data Science manuscript for content, Registry of Multiple Institute, Hasselt including medical writing Sclerosis and University, Belgium for content; study concept Neuromyelitis Optica or design Spectrum Disorders, Brazil Maria Fernanda REDONE.br–Brazilian Drafting/revision of the References Mendes, PhD Registry of Multiple manuscript for content, 1. Luna G, Alping P, Burman J, et al. Infection risks among patients with multiple Sclerosis and including medical writing sclerosis treated with fingolimod, natalizumab, rituximab, and injectable therapies. Neuromyelitis Optica for content; major role in JAMA Neurol. 2019;77(2):184-191. Spectrum Disorders, the acquisition of data 2. Salter A, Fox RJ, Newsome SD, et al. Outcomes and risk factors associated with SARS- Brazil; Irmandade da CoV-2 infection in a North American registry of patients with multiple sclerosis. Santa Casa de JAMA Neurol. 2021;78(6):699-708. Misericórdia de São Paulo, 3. Louapre C, Collongues N, Stankoff B, et al. Clinical characteristics and outcomes in Brazil patients with coronavirus disease 2019 and multiple sclerosis. JAMA Neurol. 2020; 77(9):1079-1088. e1884 Neurology | Volume 97, Number 19 | November 9, 2021 Neurology.org/N 4. Evangelou N, Garjani A, dasNair R, et al. Self-diagnosed COVID-19 in people with 11. Wijnands JMA, Zhu F, Kingwell E, et al. Disease-modifying drugs for multiple sclerosis multiple sclerosis: a community-based cohort of the UK MS Register. J Neurol and infection risk: a cohort study. J Neurol Neurosurg Psychiatry. 2018;89(10):1050-1056. Neurosurg Psychiatry. 2020;92(1):107-109. 12. Myhr KM, Torkildsen Ø, Lossius A, Bø L, Holmøy T. B cell depletion in the treatment 5. Sormani MP, De Rossi N, Schiavetti I, et al. Disease modifying therapies and covid-19 of multiple sclerosis. Expert Opin Biol Ther. 2019;19(3):261-271. severity in multiple sclerosis. Ann Neurol. 2021;89(4):780-789. 13. Klein C, Lammens A, Schäfer W, et al. Epitope interactions of monoclonal antibodies 6. Peeters LM, Parciak T, Walton C, et al. COVID-19 in people with multiple sclerosis: a targeting CD20 and their relationship to functional properties. mAbs. 2013;5(1):22-33. global data sharing initiative. Mult Scler. 2020;26(10):1157-1162. 14. Gelfand JM, Cree BAC, Hauser SL. Ocrelizumab and other CD20(+) B-cell-depleting 7. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an Expanded Dis- therapies in multiple sclerosis. Neurotherapeutics. 2017;14(4):835-841. ability Status Scale (EDSS). Neurology. 1983;33(11):1444-1452. 15. Atlas of MS, 3rd ed. Multiple Sclerosis International Federation;2020. 8. D’Souza M, Yaldizli Ö, John R, et al. Neurostatus e-scoring improves consistency of 16. Marrie RA, Cutter G, Tyry T, Campagnolo D, Vollmer T. Smoking status over two Expanded Disability Status Scale assessments: a proof of concept study. Mult Scler. years in patients with multiple sclerosis. Neuroepidemiology. 2009;32(1):72-79. 2017;23(4):597-603. 17. Ponsonby AL, Lucas RM, Dear K, et al. The physical anthropometry, lifestyle habits 9. Mehta D, Miller C, Arnold DL, et al. Effect of dimethyl fumarate on lymphocytes in and blood pressure of people presenting with a first clinical demyelinating event RRMS: implications for clinical practice. Neurology. 2019;92(15):e1724–e1738. compared to controls: the Ausimmune study. Mult Scler. 2013;19(13):1717-1725. 10. Baker D, Amor S, Kang AS, Schmierer K, Giovannoni G. The underpinning biology 18. Pittas F, Ponsonby AL, van der Mei IA, et al. Smoking is associated with progressive relating to multiple sclerosis disease modifying treatments during the COVID-19 disease course and increased progression in clinical disability in a prospective cohort pandemic. Mult Scler Relat Disord. 2020;43:102174. of people with multiple sclerosis. J Neurol. 2009;256(4):577-585. Neurology.org/N Neurology | Volume 97, Number 19 | November 9, 2021 e1885