This comparative effectiveness research study investigates the effectiveness of disease-modifying treatments in patients with primary progressive multiple sclerosis with or without active disease.
Key Points
Question
What is the effectiveness of disease-modifying treatments (DMTs) in patients with primary progressive multiple sclerosis (PPMS) with or without active disease?
Findings
In this comparative effectiveness research study of 409 matched patients with PPMS, DMTs were associated with reduced risk of becoming wheelchair dependent in patients with persistent inflammatory activity.
Meaning
Inflammatory activity may be a modifiable component of long-term disability outcomes in patients with PPMS.
Abstract
Importance
Except for ocrelizumab, treatment options in primary progressive multiple sclerosis (PPMS) are lacking.
Objective
To investigate the effectiveness of DMTs on the risk of becoming wheelchair dependent in a real-world population of patients with PPMS.
Design, Setting, and Participants
This was a multicenter, observational, retrospective, comparative effectiveness research study. Data were extracted on November 28, 2018, from the Italian multiple sclerosis register and analyzed from June to December 2021. Mean study follow-up was 11 years. Included in the study cohort were patients with a diagnosis of PPMS and at least 3 years of Expanded Disability Status Scale (EDSS) evaluations and 3 years of follow-up.
Main Outcomes and Measures
The risk of reaching an EDSS score of 7.0 was assessed through multivariable Cox regression models.
Exposures
Patients who received DMT before the outcome were considered treated. DMT was assessed as a time-dependent variable and by class of DMT (moderately and highly effective).
Results
From a total of 3298 patients with PPMS, 2633 were excluded because they did not meet the entry criteria for the phase 3, multicenter, randomized, parallel-group, double-blind, placebo-controlled study to evaluate the efficacy and safety of ocrelizumab in adults with PPMS (ORATORIO) trial. Among the remaining 665 patients (mean [SD] age, 43.0 [10.7] years; 366 female patients [55.0%]), 409 were further selected for propensity score matching (288 treated and 121 untreated patients). In the matched cohort, during the study follow-up, 37% of patients (152 of 409) reached an EDSS score of 7.0 after a mean (SD) follow-up of 10.6 (5.6) years. A higher EDSS score at baseline (adjusted hazard ratio [aHR], 1.32; 95% CI, 1.13-1.55; P < .001), superimposed relapses (aHR, 2.37; 95% CI, 1.24-4.54; P = .009), and DMT exposure (aHR, 1.75; 95% CI, 1.04-2.94; P = .03) were associated with a higher risk of an EDSS score of 7.0, whereas the interaction term between DMT and superimposed relapses was associated with a reduced risk of EDSS score of 7.0 (aHR, 0.33; 95% CI, 0.16-0.71; P = .004). Similar findings were obtained when treatment according to DMT class was considered and when DMT was included as a time-dependent covariate. These results were confirmed in the subgroup of patients with available magnetic resonance imaging data.
Conclusions and Relevance
Results of this comparative effectiveness research study suggest that inflammation also occurs in patients with PPMS, may contribute to long-term disability, and may be associated with a reduced risk of becoming wheelchair dependent by current licensed DMTs.
Introduction
Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system clinically characterized by 2 core phenomena: relapses and progression.1 Primary progressive MS (PPMS) accounts for approximately 10% of cases and is characterized by a clinically detectable disability progression from onset lasting at least 1 year.2 Superimposed relapses and/or radiologic activity may occur during the disease course and identify patients with active PPMS.1 Although several disease-modifying treatments (DMTs) are available for the treatment of relapsing-remitting MS,3,4 in the PPMS phenotype, with the exception of ocrelizumab,5 previous randomized clinical trials (RCTs) of current licensed DMTs were found to be ineffective.6
One possible explanation is that currently approved DMTs mostly target the inflammatory component of the disease, which is less prominent in this phenotype. For instance, in a preplanned subgroup analysis of the study to evaluate the safety and efficacy of rituximab in adults with PPMS (OLYMPUS) trial,7 treatment with rituximab slowed the disability progression in PPMS participants younger than 51 years and those with gadolinium-enhancing lesions at baseline.8 However, relapse activity in the progressive phenotype of MS still has a controversial role. Although some studies reported no influence on disability progression,9,10,11 in others, relapses accelerated time to severe disability.12 In a recent study on secondary progressive MS on the MSBase data set,13 DMT was associated with improvements in disability outcomes among patients with relapses during the secondary progressive phase of the disease. Likewise, a study on PPMS by the same group14 showed that relapses were associated with a lower risk of confirmed disability progression, probably owing to the association of superimposed relapses with DMT use.14 Collectively, these findings suggest that inflammatory disease activity remains a modifiable component in progressive MS.
With this background, by using the data collected within the Italian MS Register, we evaluated the potential effectiveness of DMTs in patients with PPMS with or without inflammatory activity. We used a reliable, robust, and clinically meaningful outcome, the risk of becoming wheelchair bound (Expanded Disability Status Scale [EDSS] score ≥ 7.0). We hypothesized that, among patients with PPMS, those who had clinically active disease in the year before study entry and/or experienced superimposed relapses during the follow-up would show an association between DMT use and a reduced risk of long-term disability accrual.
Methods
Standard Protocol Approvals and Patient Consent
The Italian iMedWeb network was approved by the Policlinico of Bari Ethics Committee and by the local ethics committees in all participating centers.15 Written informed consent was obtained from all enrolled patients in accordance with the Declaration of Helsinki. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
Population and Study Design
We conducted a multicenter, observational, retrospective comparative effectiveness research study on prospectively acquired data, with the aim to explore the potential effectiveness of DMTs on the risk of becoming wheelchair dependent in a cohort of patients with PPMS. Anonymized clinical records were extracted on November 28, 2018. We included patients with a clinical diagnosis of PPMS according to 2010 revised McDonald criteria,16 at least 3 EDSS evaluations, at least 3 years of follow-up, the available minimum data set (including sex, age at onset, MS course, first clinical presentation, date of symptom onset, diagnosis and first recorded visit, DMT type, and date of start), and the main ORATORIO17 inclusion criteria (a baseline functional systems pyramidal score ≥2, age at onset >18 years, and EDSS score ≥5.5 with disease duration ≤15 years or 3.0≤ EDSS score ≤5.0 with disease duration ≤10 years).5 Patient race and ethnicity data were not gathered in the minimum data set. We excluded patients who were already receiving DMT at the time of the first recorded visit, those who had already reached an EDSS score of 7.0 at study entry, and those participating in RCTs or who received hematopoietic stem cell transplant (eFigure in Supplement 1). Furthermore, to adhere to the ORATORIO exclusion criteria regarding patients with a progressive-relapsing course, we also performed a separate analysis where we further excluded those patients who had clinical relapses in the year before study entry.
The Italian MS Register protocol required biannual updates of the minimum data set, but patients with less frequent visits were not excluded from the analyses. To mitigate for immortal time bias, the study baseline for treated patients was the starting date of the first DMT, and for untreated patients, the date of the first EDSS score evaluation. MS duration at baseline was calculated from the first reported symptom. The follow-up time was defined as the time between baseline and the last available EDSS entry.
A relapse was defined as the occurrence of new symptoms or exacerbation of existing symptoms persisting for at least 24 hours, in the absence of concurrent illness or fever, and occurring at least 30 days after a previous relapse.18 The main analysis was also repeated in the subgroup of patients with a brain magnetic resonance imaging (MRI) assessment less than 1 year before the baseline and during the follow-up time. Radiologic activity was defined as the presence of new or enlarging T2 lesions and/or the presence of gadolinium-enhancing lesions. The study outcome was the risk of becoming wheelchair dependent, corresponding to an irreversible EDSS score of 7.0, and was reached if all the subsequent EDSS scores were greater or equal to 7.0.
Treatment Definitions
We considered patients to be treated when they started the DMT before reaching an EDSS score of 7.0. We categorized DMTs according to a largely accepted definition,3 in moderately effective treatments (interferon beta-1a/1b, glatiramer acetate, dimethyl fumarate, teriflunomide, methotrexate, and azathioprine) and highly effective treatments (ocrelizumab, rituximab, natalizumab, alemtuzumab, cyclophosphamide, mitoxantrone, cladribine, and fingolimod). As for the class of DMT, we divided patients into 4 categories: (1) patients who received only moderately effective treatments; (2) patients who switched, during the follow-up, to highly effective treatments; (3) patients who received, as first DMT, a highly effective treatment; and (4) patients who were never treated.
Statistical Analysis
Baseline and follow-up characteristics were expressed as mean and SD, median and IQR, or frequency and percentage for continuous and categorical covariates. Categorical variables were compared by using χ2 test, whereas continuous variables were compared by using Mann-Whitney U and t test, where appropriate.
To control for treatment indication bias, patients were matched on their propensity for receiving vs not receiving the treatment. Patients were matched on a 1:1 ratio, by nearest neighbor with replacement, within a caliper of 0.01 SD. Variables used to define treatment allocation were sex, age at onset, age at baseline, first EDSS score, visit density, and the presence of prebaseline relapses (yes vs no). The quality of the propensity score–matched cohorts was assessed with the standardized mean difference and the variance ratios for continuous variables. A standardized mean difference greater than 0.1 was considered a sign of imbalance.19
The risk of an EDSS score of 7.0 was assessed in the propensity score–matched cohorts, with a per-protocol analysis through multivariable Cox regression models adjusted for sex, age at the study baseline, disease duration, visit density (number of EDSS evaluations per years of follow-up), first EDSS score evaluation, baseline relapses (≤1 year), superimposed relapses, and the interaction term between DMT and superimposed relapses. DMT was assessed as both a dichotomous variable (yes/no) and as DMT class as explained previously. The origin of the models was the date of first DMT for treated patients and the date of first EDSS evaluation for untreated patients. Furthermore, DMT was introduced as a time-dependent covariate, with a value of zero until drug introduction and changes to 1 thereafter. In this last model, the baseline was the first EDSS visit date for all participants. In the subgroup of patients with an available brain MRI, we also evaluated the presence of radiologic activity. In this subgroup of patients, superimposed activity was defined as the presence of new relapses and/or MRI activity during the follow-up. Finally, as sensitivity analysis, we excluded patients registered before the year 2000 and reran the propensity score matching and survival analyses.
Patients who did not reach a sustained EDSS score of 7.0 were censored at the last EDSS available. Proportional hazard assumption was checked by graphical inspection of Schoenfeld residuals. Matching and statistical analyses were conducted using SPSS software for Windows, version 25.0 (IBM Corp) and R, version 4.1.2 (R Foundation). Statistical significance was set at a 2-sided P value < .05. Data were analyzed from June to December 2021.
Results
Baseline and Follow-up Characteristics
From a total of 3298 patients with PPMS, 2633 were excluded because they did not meet the entry criteria for the ORATORIO trial. We retained 665 patients with PPMS (mean [SD] age, 43.0 [10.7] years; 366 female patients [55.0%]; 299 male patients [45%]) of whom 452 (67%) were treated with DMT (eFigure and eTable 1 in Supplement 1).
At baseline among the unmatched patient group, compared with untreated patients, treated patients were younger (mean [SD] age, 46.5 [10.4] years vs 50.8 [11.2] years), had a shorter disease duration (mean [SD] duration, 4.8 [3.1] years vs 5.0 [3.6] years), and had, on average, a lower first EDSS score (median [IQR], 4.0 [3.0-5.0] vs 5.0 [4.0-6.0]). There were no differences in sex nor in symptoms at the onset of the disease (Table 1).
Table 1. Baseline and Follow-up Characteristics of the Unmatched and Matched Cohort.
| Characteristics | Total (N = 665 | Unmatched | Matched | ||
|---|---|---|---|---|---|
| Treated (n = 452) | Untreated (n = 213) | Treated (n = 288) | Untreated (n = 121) | ||
| Baseline characteristics | |||||
| Female patients, No. (%) | 366 (55) | 249 (55) | 117 (55) | 150 (52) | 64 (53) |
| Male patients, No. (%) | 299 (45) | 203 (45) | 96 (45) | 138 (48) | 57 (47) |
| Age at onset, mean (SD), y | 43.0 (10.7) | 41.2 (10.2) | 45.8 (11.2) | 43.0 (10.3) | 44.2 (11.2) |
| Age at baseline, mean (SD), y | 47.8 (10.8) | 46.5 (10.4) | 50.8 (11.2) | 47.7 (10.3) | 49.5 (11.3) |
| Disease duration, mean (SD), y | 4.8 (3.4) | 4.8 (3.1) | 5.0 (3.6) | 4.8 (3.0) | 5.2 (3.9) |
| Symptom at onset, No. (%) | |||||
| Optic neuritis | 41 (6) | 29 (6) | 12 (6) | 20 (7) | 10 (8) |
| Brainstem | 137 (21) | 92 (20) | 45 (21) | 58 (20) | 28 (23) |
| Spinal cord | 299 (45) | 212 (47) | 87 (41) | 139 (48) | 48 (40) |
| Supratentorial | 343 (52) | 225 (50) | 118 (55) | 141 (49) | 63 (31) |
| First EDSS score, median (IQR) | 4.5 (3.5-5.5) | 4.0 (3.0-5.0) | 5.0 (4.0-6.0) | 4.0 (3.0-5.0) | 4.5 (3.5-5.5) |
| Follow-up characteristics | |||||
| Follow-up time, mean (SD), y | 11.0 (5.8) | 11.2 (5.4) | 10.4 (6.6) | 11.1 (5.5) | 9.5 (5.7) |
| Visit density per year, mean (SD) | 0.8 (0.6) | 1.0 (0.8) | 0.7 (0.5) | 0.7 (0.5) | 0.7 (0.4) |
| Last EDSS score, median (IQR) | 6.5 (6.0-7.5) | 6.5 (6.0-7.5) | 6.5 (6.0-7.0) | 6.5 (6.0-7.5) | 6.5 (6.0-7.5) |
| Relapses | |||||
| Relapsing 12 mo before baseline | 123 (18) | 99 (22) | 24 (11) | 52 (18) | 19 (16) |
| Superimposed relapse | 229 (34) | 167 (37) | 62 (27) | 103 (36) | 44 (36) |
| Treatments at baseline, No. (%) | |||||
| Interferon beta | NA | 156 (23) | NA | 86 (21) | NA |
| Azathioprine | 116 (17) | 78 (19) | |||
| Methotrexate | 51 (8) | 36 (9) | |||
| Mitoxantrone | 50 (8) | 34 (8) | |||
| Glatiramer acetate | 40 (6) | 28 (7) | |||
| Cyclophosphamide | 22 (3) | 16 (4) | |||
| Fingolimod | 11 (2) | 7 (2) | |||
| Cyclosporin | 1 (0.5) | 1 (0.3) | |||
| Cladribine | 1 (0.5) | NA | |||
| Dimethyl fumarate | 1 (0.5) | 1 (0.3) | |||
| Natalizumab | 3 (1.5) | 1 (0.3) | |||
| Treatments during follow-up, No. (%) | |||||
| Moderately effective treatmenta | NA | 242 (53) | NA | 157 (54) | NA |
| Switched to highly effective treatmentb | 126 (28) | 73 (25) | |||
| Starting with highly effective treatment | 89 (19) | 59 (21) | |||
| Outcome | |||||
| EDSS score of 7.0, No. (%) | 272 (41) | 190 (42) | 82 (39) | 112 (39) | 40 (33) |
| Age at EDSS score of 7.0, mean (SD), y | 55.8 (10.8) | 54.5 (10.5) | 58.7 (11.1) | 55.8 (10.2) | 57.0 (11.7) |
Abbreviations: DMT, disease-modifying treatment; EDSS, Expanded Disability Status Scale; NA, not applicable.
Moderately effective treatments included interferon beta-1a/1b, glatiramer acetate, dimethyl fumarate, teriflunomide, methotrexate, and azathioprine.
Highly effective treatments included ocrelizumab, rituximab, natalizumab, alemtuzumab, mitoxantrone, cyclophosphamide, cladribine, and fingolimod.
The mean (SD) study follow-up was 11.0 (5.8) years and was similar in treated (11.2 [5.4] years) and untreated (10.4 [6.6] years) patients. Treated patients were evaluated more often than untreated patients (mean [SD] visit density per year, 1.0 [0.8] vs 0.7 [0.5]) whereas the last EDSS score did not differ between groups.
Regarding clinical inflammatory activity, a higher percentage of treated patients had baseline relapse activity (99 of 452 [22%] vs 24 of 213 [11%]) and superimposed relapse activity during follow-up occurred in 167 of 452 treated patients (37%) and 62 of 213 untreated patients (27%) (Table 1). Regarding DMT class, approximately one-half of treated patients (242 of 452 [53%]) received only moderately effective DMT, 126 (28%) switched during follow-up to a highly effective DMT, whereas 89 (19%) received as first treatment a highly effective DMT (Table 1).
Based on their probability of receiving vs not receiving a DMT, we matched on a 1:1 ratio, according to their sex, age at onset, age at baseline, prebaseline relapses, first EDSS score, and visit density, 409 patients (288 [70%] treated and 121 [30%] untreated) (Table 1). Variables significantly associated with DMT prescription were age at onset (odds ratio [OR], 0.98; 95% CI, 0.95-0.98; P = .07), a lower EDSS score (OR, 0.77; 95% CI, 0.65-0.91; P = .003), a higher visit density (OR, 1.31; 95% CI, 1.22-1.44; P = .001), and the presence of baseline relapses (OR, 2.2; 95%CI, 1.28-3.70; P = .003). After propensity score matching, there were no statistically significant differences of baseline characteristics between the treated and untreated cohort. Table 1 summarizes the baseline and follow-up characteristics of the unmatched and matched cohort.
A brain MRI was available at baseline and during follow-up in 277 patients. Treated patients had an active MRI at baseline more frequently than untreated patients (47 of 196 [24%] vs 8 of 81 [10%]). There were no significant differences in the proportion of active MRI between treated and untreated patients over the follow-up period. Table 2 summarizes the baseline and follow-up characteristics of the cohort with available MRI data.
Table 2. Baseline and Follow-up Characteristics of the Cohort With Available MRI Data at Baseline (<1 Year) and During Follow-up.
| Characteristics | Total (N = 277) | Treated (n = 196) | Untreated (n = 81) |
|---|---|---|---|
| Baseline characteristics | |||
| Female patients, No. (%) | 150 (54) | 106 (54) | 44 (54) |
| Male patients, No. (%) | 127 (46) | 90 (46) | 37 (46) |
| Age at onset, mean (SD), y | 43.2 (10.2) | 43.0 (10.1) | 43.7 (10.7) |
| Age at baseline, mean (SD), y | 48.0 (10.3) | 47.7 (10.1) | 48.8 (10.9) |
| Disease duration, mean (SD) | 4.8 (3.2) | 4.7 (2.9) | 5.1 (3.9) |
| Symptom at onset, No. (%) | |||
| Optic neuritis | 20 (7) | 12 (6) | 8 (10) |
| Brainstem | 59 (21) | 38 (19) | 21 (26) |
| Spinal cord | 123 (44) | 94 (48) | 29 (36) |
| Supratentorial | 144 (52) | 100 (51) | 44 (54) |
| First EDSS score, median (IQR) | 4.0 (3.0-5.0) | 4.0 (3.0-5.0) | 4.5 (3.0-5.5) |
| Follow-up characteristics | |||
| Follow-up time, mean (SD), y | 10.9 (5.8) | 11.3 (5.6) | 10.0 (5.9) |
| Visit density per year, mean (SD) | 0.7 (0.4) | 0.7 (0.5) | 0.7 (0.3) |
| Last EDSS score, median (IQR) | 6.5 (6.0-7.0) | 6.5 (6.0-7.0) | 6.5 (6.0-7.0) |
| Inflammatory activitya | |||
| Relapsing 12 mo before baseline | 47 (17) | 32 (16) | 15 (18) |
| MRI activity in the year before baseline | 55 (20) | 47 (24) | 8 (10) |
| MRI and/or clinical activity in the year before baseline | 91 (33) | 71 (36) | 20 (25) |
| Superimposed | |||
| Relapse | 105 (38) | 73 (37) | 32 (39) |
| MRI activity | 96 (35) | 72 (37) | 24 (30) |
| MRI and/or clinical activity | 122 (44) | 88 (45) | 34 (41) |
| Treatments at baseline, No. (%) | |||
| Interferon beta | NA | 63 (32) | NA |
| Azathioprine | 52 (26) | ||
| Methotrexate | 26 (13) | ||
| Glatiramer acetate | 21 (11) | ||
| Mitoxantrone | 20 (10) | ||
| Cyclophosphamide | 10 (5) | ||
| Fingolimod | 3 (2) | ||
| Natalizumab | 1 (1) | ||
| Time to first treatment, mean (SD), y | NA | 4.7 (2.9) | NA |
| Treatments during follow-up, No. (%) | |||
| Moderately effective treatmentb | NA | 107 (54) | NA |
| Switched to highly effective treatmentc | 55 (28) | ||
| Starting with highly effective treatment | 34 (17) | ||
| Outcome | |||
| EDSS score of 7.0, No. (%) | 104 (37) | 75 (38) | 29 (36) |
| Age at EDSS score of 7.0 (y), mean (SD) | 56.2 (10.5) | 56.0 (10.2) | 56.6 (11.3) |
Abbreviations: DMT, disease-modifying treatment; EDSS, Expanded Disability Status Scale; MRI, magnetic resonance imaging; NA, not applicable.
MRI activity was defined as either the presence of new or enlarging T2 lesions and/or the presence of new T1 gadolinium-enhancing lesions.
Moderately effective treatments included interferon beta-1a/1b, glatiramer acetate, dimethyl fumarate, teriflunomide, methotrexate, and azathioprine.
Highly effective treatments included ocrelizumab, rituximab, natalizumab, alemtuzumab, mitoxantrone, cyclophosphamide, cladribine, and fingolimod.
Survival Analyses
In the matched cohort, during the study follow-up, 37% of patients (152 of 409) reached the EDSS score of 7.0 after a mean (SD) follow-up time of 10.6 (5.6) years, without differences between groups. Treated patients were wheelchair bound at a mean (SD) age of 55.8 (10.2) years whereas untreated patients were wheelchair bound at a mean (SD) age of 57.0 (11.7) years (Table 1).
Table 3 shows the results of the multivariable Cox regression models for the outcome EDSS score of 7.0 in the matched sample (n = 409). A higher EDSS score at baseline (adjusted hazard ratio [aHR], 1.32; 95% CI, 1.13-1.55; P < .001), the presence of superimposed relapses (aHR, 2.37; 95% CI, 1.24-4.54; P = .009), and DMT exposure (aHR, 1.75; 95% CI, 1.04-2.94; P = .03) were associated with a higher risk of an EDSS score 7.0. The only variable associated with a reduced risk of becoming wheelchair bound was the interaction term between DMT exposure and superimposed relapses (aHR, 0.33; 95% CI, 0.16-0.71; P = .004). Similar findings were obtained when we considered treatment exposure according to DMT class (moderately effective DMT: aHR, 0.41; 95% CI, 0.18-0.95; P = .03; switch to highly effective DMT: aHR, 0.32; 95% CI, 0.12-0.87; P = .03; first treatment with highly effective DMT: aHR, 0.31; 95% CI, 0.10-0.95; P = .04) (Table 3).
Table 3. Multivariable Cox Regression Model for the Risk of Reaching Expanded Disability Status Scale (EDSS) Score of 7.0, Matched Cohort (n = 409)a.
| Variable | Reference class | aHR (95% CI) | P value |
|---|---|---|---|
| Male sex | Female sex | 1.17 (0.84-1.62) | .35 |
| Age at baseline | NA | 0.99 (0.98-1.01) | .43 |
| Disease duration at baseline | NA | 1.03 (0.98-1.08) | .22 |
| Visit density | NA | 0.82 (0.57-1.20) | .31 |
| First EDSS score | NA | 1.32 (1.13-1.55) | <.001 |
| Relapsing 12 mo before baseline | No relapse | 1.01 (0.81-1.25) | .95 |
| Superimposed relapses | No relapse | 2.37 (1.24-4.54) | .009 |
| DMT (yes/no) | No therapy | 1.75 (1.04-2.94) | .03 |
| DMT × superimposed relapse (IT) | No therapy | 0.33 (0.16-0.71) | .004 |
| DMT class × superimposed relapse | |||
| Moderately effective treatmentb (IT) | No therapy | 0.41 (0.18-0.95) | .04 |
| Switched to highly effective (IT)c | No therapy | 0.32 (0.12-0.87) | .03 |
| Starting with highly effective (IT) | No therapy | 0.31 (0.10-0.95) | .04 |
| DMT as time-dependent variable × superimposed relapse (IT)d,e | No therapy | 0.41 (0.21-0.81) | .01 |
Abbreviations: aHR, adjusted hazard ratio; DTM, disease-modifying therapy; IT, interaction term; NA, not applicable.
Number of events = 152; DMT, n = 288; control, n = 121.
Moderately effective treatments included interferon beta-1a, interferon beta-1a/1b, glatiramer acetate, dimethyl fumarate, teriflunomide, methotrexate, and azathioprine,
Highly effective treatments included ocrelizumab, rituximab, natalizumab, alemtuzumab, mitoxantrone, cyclophosphamide, cladribine, and fingolimod.
Models were adjusted for sex, age at baseline, disease duration at baseline, visit density, first EDSS score, and active primary progressive multiple sclerosis phenotype.
The value of the time-dependent covariate is zero until drug introduction and changes to 1 thereafter. For patients who never received a DMT during the follow-up time, the value remains zero. In this model, the baseline was the first EDSS score for both cohorts.
In the subgroup with available MRI data (n = 277) in Table 2, we also considered the presence of MRI activity at baseline and during follow-up. In this model, the risk of an EDSS score of 7.0 was associated with a higher EDSS score at baseline (aHR, 1.35; 95% CI, 1.12-1.64; P = .002), DMT exposure (aHR, 3.61; 95% CI, 1.52-8.56; P = .004) and, marginally, with superimposed clinical and/or MRI activity (aHR, 2.14; 95% CI, 0.94-4.88; P = .07). On the other hand, the interaction term between DMT and superimposed clinical and/or MRI activity during follow-up was associated with a reduced risk of EDSS score of 7.0 (aHR, 0.30; 95% CI, 0.11-0.82; P = .02) (Table 4). Similar results were obtained when using the other treatment definitions (Table 4).
Table 4. Multivariable Cox Regression Model for the Risk of Reaching Expanded Disability Status Scale (EDSS) Score of 7.0, Magnetic Resonance Imaging (MRI) Cohort (n = 277)a.
| Variable | Reference class | aHR (95% CI) | P value |
|---|---|---|---|
| Male sex | Female sex | 1.13 (0.75-1.70) | .57 |
| Age at baseline | NA | 0.99 (0.97-1.02) | .57 |
| Disease duration at baseline | NA | 1.03 (0.96-1.10) | .39 |
| Visit density | NA | 0.58 (0.35-0.99) | .04 |
| First EDSS score | NA | 1.35 (1.12-1.64) | .002 |
| Relapsing 12 mo before baseline | No relapse | 1.10 (0.85-1.42) | .49 |
| GdE lesion 1 y before baseline | No GdE lesion | 0.93 (0.55-1.57) | .78 |
| MRI and/or clinical superimposed activity | No MRI and/or clinical activity | 2.14 (0.94-4.88) | .07 |
| DMT (yes/no) | No therapy | 3.61 (1.52-8.56) | .004 |
| DMT × MRI and/or clinical superimposed activity (IT) | No therapy | 0.30 (0.11-0.82) | .02 |
| DMT class × MRI and/or clinical superimposed activity | |||
| Moderately effective treatment (IT)b | No therapy | 0.49 (0.17-1.46) | .20 |
| Switched to highly effective (IT)c | No therapy | 0.12 (0.03-0.43) | .001 |
| Starting with highly effective (IT) | No therapy | 0.20 (0.04-0.96) | .04 |
| DMT as time-dependent variable × MRI and/or clinical superimposed activity (IT)d,e | No therapy | 0.38 (0.14-1.00) | .05 |
Abbreviations: aHR, adjusted hazard ratio; DTM, disease-modifying therapy; GdE, gadolinium enhancing; IT, interaction term; NA, not applicable.
Number of events = 104; DMT, n = 196; control, n = 81.
Moderately effective treatments included interferon beta-1a, interferon beta-1a/1b, glatiramer acetate, dimethyl fumarate, teriflunomide, methotrexate, and azathioprine.
Highly effective treatments included ocrelizumab, rituximab, natalizumab, alemtuzumab, mitoxantrone, cyclophosphamide, cladribine, and fingolimod.
Models were adjusted for sex, age at baseline, disease duration at baseline, visit density, first EDSS score, and active primary progressive multiple sclerosis phenotype.
The value of the time-dependent covariate is zero until drug introduction and changes to 1 thereafter. For patients who never received a DMT during the follow-up time, the value remains zero. In this model the baseline was the first EDSS for both cohorts.
Focusing the analysis on those patients who did not exhibit any relapse activity in the year before baseline (n = 338; 54% female patients; mean [SD] EDDS score, 4.5 [1.0]; mean [SD] disease duration, 14.6 [6.6] years), the association between superimposed relapses over the follow-up period and DMT effectiveness remained, although it did not reach statistical significance (aHR, 0.44; 95% CI, 0.19-1.05; P = .06) (Table 5). By including patients registered after the year 2000, we matched a cohort of 316 patients with PPMS (228 treated and 88 untreated). Sensitivity analyses confirmed the main results and are available as supplementary material (eTable 2 in Supplement 1).
Table 5. Multivariable Cox Regression Model for the Risk of Reaching Expanded Disability Status Scale (EDSS) Score of 7.0 in Patients Without Relapse Activity in the Year Before Baseline (n = 338)a.
| Variable | Reference class | aHR (95% CI) | P value |
|---|---|---|---|
| Male sex | Female sex | 1.10 (0.76-1.58) | .62 |
| Age at baseline | NA | 0.99 (0.97-1.01) | .32 |
| Disease duration at baseline | NA | 1.04 (0.98-1.10) | .21 |
| Visit density | NA | 0.76 (0.50-1.17) | .21 |
| First EDSS | NA | 1.28 (1.07-1.52) | .006 |
| Superimposed relapses | No relapse | 1.95 (0.93-4.08) | .08 |
| DMT (yes/no) | No therapy | 1.62 (0.94-2.79) | .08 |
| DMT × superimposed relapse (IT) | No therapy | 0.44 (0.19-1.05) | .06 |
| DMT class × superimposed relapse | |||
| Moderately effective treatment (IT)b | No therapy | 0.52 (0.19-1.38) | .18 |
| Switched to highly effective (IT)c | No therapy | 0.59 (0.19-1.78) | .34 |
| Starting with highly effective (IT) | No therapy | 0.34 (0.09-1.23) | .01 |
| DMT as time-dependent variable × superimposed relapse (IT)d,e | No therapy | 0.47 (0.20-1.11) | .09 |
Abbreviations: aHR, adjusted hazard ratio; DTM, disease-modifying Therapy; IT, interaction term; NA, not applicable.
Number of events = 123; DMT, n = 236; control, n = 102.
Moderately effective treatments included interferon beta-1a, interferon beta-1a/1b, glatiramer acetate, dimethyl fumarate, teriflunomide, methotrexate, and azathioprine.
Highly effective treatments included ocrelizumab, rituximab, natalizumab, alemtuzumab, mitoxantrone, cyclophosphamide, cladribine, and fingolimod.
Models were adjusted for sex, age at baseline, disease duration at baseline, visit density, first EDSS score, and active primary progressive multiple sclerosis phenotype.
The value of the time-dependent covariate is zero until drug introduction and changes to 1 thereafter. For patients who never received a DMT during the follow-up time, the value remains zero. In this model the baseline was the first EDSS for both cohorts.
Discussion
In this observational, multicenter, comparative effectiveness research study in Italy, we found that DMT was associated with a delayed time to wheelchair use in patients with PPMS fulfilling the ORATORIO trial inclusion criteria and with superimposed inflammatory activity during the follow-up period.
Despite the progressive onset, approximately 18% of the PPMS cohort experienced relapses in the year before baseline, and 34% had superimposed relapse activity over the follow-up period. The percentage of active patients increased to 33% in the year before baseline and to 44% during the follow-up period in the subgroup of patients with MRI data available. These proportions are in line with the findings of a recent large registry study,14 previous population-based studies,9,11 and with the baseline characteristics of the ORATORIO trial, where the percentage of patients with PPMS with gadolinium-enhancing lesions at baseline was 27%.5
Although some studies showed scarce influence on disability accrual,12,20 others found relapses as an independent determinant of disability accumulation.9,21 Our study results suggested that superimposed relapses were a risk factor for becoming wheelchair dependent; this finding remained consistent after propensity score matching. Moreover, we found that DMT was associated with a reduction in the risk of long-term disability accrual in patients with superimposed relapses during the follow-up period. In line with our findings, previous registry-based studies found that longer treatment persistency in patients with superimposed relapses was associated with a lower risk of confirmed disability progression in PPMS14 and of reaching an EDSS score of 7.0 in secondary progressive MS.13
Taken as a whole, the results of our and previous studies13,14 suggest that the occurrence of inflammatory activity is relatively frequent and can represent a modifiable determinant of long-term disability accrual even in progressive MS. Although the association between relapse occurrence and better outcomes could appear counterintuitive, it should be noted that relapse activity while receiving DMT cannot be compared with that occurring without treatment. For instance, it is possible that the number of relapses without treatment would have been higher than that observed during DMT. Moreover, treatments can mitigate the consequences of relapses. For example, in the Natalizumab Safety and Efficacy in Relapsing-Remitting Multiple Sclerosis (AFFIRM) study, natalizumab treatment decreased the clinical severity of relapses and improved recovery from disability induced by relapses.22 Likewise, fingolimod, in the placebo-controlled trial of Efficacy and Safety of Fingolimod in Patients With Relapsing-Remitting Multiple Sclerosis (FREEDOMS) study, decreased the likelihood that baseline gadolinium-enhancing lesions convert to persistent black holes.23
Beside disease activity, another factor possibly associated with response to DMTs is a younger patient age.4,8 In our cohort, age at baseline was included in all the multivariable models and was not significantly associated with treatment response. We do not have a clear explanation for this finding. We can hypothesize that the long-term follow-up and the potential for a therapeutic-lag phenomenon in treatment effectiveness24,25,26,27 may have mitigated age-related differences in treatment response.
Our findings suggest that in patients with PPMS who fulfill the main inclusion criteria for the ORATORIO trial and have superimposed activity over the follow-up period, many DMTs (both moderately and highly effective treatments) can be therapeutically effective in preventing progression to loss of ambulatory function. To date, the only approved drug in PPMS is ocrelizumab, which, despite a good safety profile, might not be available for all patients owing to safety concerns or cost-effectiveness considerations.28 In some patients, there may be a role for any DMT rather than just the single-approved therapy for this type of MS at this time. This would be important for patients with PPMS who live in places where ocrelizumab is still unaffordable. Moreover, the COVID-19 infection has spread worldwide, and some studies have suggested that severe COVID-19 outcomes were higher among patients treated with anti-CD20.29,30 In the future, COVID-19 infection may reshape the benefit-risk balance of some drugs. Our study results suggest that the off-label use of other DMTs in patients with PPMS with persistent inflammatory activity may be taken into consideration if ocrelizumab is contraindicated.
On the other hand, in patients with nonactive PPMS, DMT exposure was associated with higher risk of EDSS score of 7.0. This may reflect the choice of neurologists to treat patients with more aggressive disease (for instance, rapidly progressing MS), despite the absence of clear inflammatory activity, and may reflect an indication bias and the lack of efficacy of DMTs in inactive PPMS. Therefore, it would be crucial to identify, at the time of treatment introduction, patients who may experience persistent inflammatory activity in order to maximize the benefit-risk balance. It has to be noted that, in our cohort, approximately 80% of patients with baseline MRI and/or clinical activity also had superimposed inflammatory events over time (data not shown). Therefore, the presence of baseline disease activity remains, to date, the main factor strongly supporting the introduction of DMT.
Strengths and Limitations
Our study had several strengths. One point of strength was the choice of a long-term and robust outcome (reaching an EDSS score of 7.0), which could explain the consistency of our findings with those of Lizak et al.13 Interestingly, the efficacy of ocrelizumab was recently confirmed in a real-world sample of PPMS, having time to wheelchair as a main study outcome.31
Another strength of the study was the long-term follow-up. Indeed, RCTs may be too short to accurately assess the effect of DMT. DMT effects, should they exist in progressive MS, may occur several years after the patient has been exposed to the therapy owing to a therapeutic lag phenomenon.24,25,26,27 Therefore, progressive MS trials with the EDSS score as a primary outcome may need to be larger and longer to overcome these limitations and point out the importance of evaluating extension phases of RCTs and observational real-world cohort studies.24
The observational nature of this study and the limited availability of MRI data represent the principal limitations. Regarding the class of DMT (moderately or highly effective), we did not find consistent and significant differences, although we were not able to perform powerful head-to-head comparisons between different DMTs. In patients with relapsing-remitting MS, strong evidence suggests that the early (within 2 years) initiation of highly effective DMTs, especially in patients with highly active MS, is superior to the escalation approach.32 It has to be noted, however, that in our cohort patients received the first DMT, on average, after 7 years from disease onset, a time lag after which the advantages of intensive immunosuppression could be lost.
Moreover, our cohort may be not fully representative of the PPMS population as 18% of patients with PPMS experienced relapse activity in the year before baseline. However, this proportion is line with the 27% of patients with PPMS with baseline active MRI in the ORATORIO trial5 and with data from registries14 and population-based studies.9,11 Furthermore, repeating the analysis in the subgroup of patients without relapses in the year before baseline, the association between superimposed relapses over the follow-up period and DMT effectiveness remained, although it was marginally significant (eTable 2 in Supplement 1), possibly owing to the limited statistical power. It also should be noted that more than one-half of treated patients received a moderately effective DMT, which could be less effective in patients with PPMS without clinical disease activity.
Although RCTs that evaluate the long-term effect of DMTs are impractical, future MRI studies can improve the accuracy in determining the role of DMTs in patients with PPMS with persistent inflammatory activity. Another potential study weakness was that the presumed effectiveness of DMT may have been attributable to a milder disease course rather than to the beneficial effect of treatment. However, we confirmed our results after having matched patients according to their treatment status. Further, the availability of just 1 outcome measure, the preplanned evaluation of upper limbs, cognitive, and walking function, may improve the sensitivity in detecting treatment effect. Finally, we did not assess the rate of adverse events, which could represent one of the reasons for early discontinuation of treatment in some patients.
Conclusions
Together with previous RCTs,5 observational studies,13,33 and subgroup analyses,8 results of this comparative effectiveness research study suggest that inflammatory activity is a modifiable risk factor of long-term disability in PPMS. In these patients, DMT exposure may be associated with a decreased risk of becoming wheelchair bound. Future RCTs should be performed to confirm our study findings that suggest the treatment benefit of both moderate- and high-efficacy DMTs for patients with PPMS. Further, future studies should focus on the identification of early and reliable markers of inflammatory activity in the progressive phases of the disease to inform treatment decision-making.
eFigure. Patient Disposition
eTable 1. Baseline Characteristics of the Excluded and Included Patients
eTable 2. Multivariable Cox Regression Model for the Risk of Reaching EDSS 7.0 in Patients Included After the Year 2000 and Propensity Score Matched (n = 316)
Nonauthor Collaborators
References
- 1.Lublin FD, Reingold SC, Cohen JA, et al. Defining the clinical course of multiple sclerosis: the 2013 revisions. Neurology. 2014;83(3):278-286. doi: 10.1212/WNL.0000000000000560 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Thompson AJ, Banwell BL, Barkhof F, et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol. 2018;17(2):162-173. doi: 10.1016/S1474-4422(17)30470-2 [DOI] [PubMed] [Google Scholar]
- 3.Rotstein D, Montalban X. Reaching an evidence-based prognosis for personalized treatment of multiple sclerosis. Nat Rev Neurol. 2019;15(5):287-300. doi: 10.1038/s41582-019-0170-8 [DOI] [PubMed] [Google Scholar]
- 4.Amato MP, Fonderico M, Portaccio E, et al. Disease-modifying drugs can reduce disability progression in relapsing multiple sclerosis. Brain. 2020;143(10):3013-3024. doi: 10.1093/brain/awaa251 [DOI] [PubMed] [Google Scholar]
- 5.Montalban X, Hauser SL, Kappos L, et al. ; ORATORIO Clinical Investigators . Ocrelizumab vs placebo in primary progressive multiple sclerosis. N Engl J Med. 2017;376(3):209-220. doi: 10.1056/NEJMoa1606468 [DOI] [PubMed] [Google Scholar]
- 6.Tur C, Montalban X. Progressive MS trials: lessons learned. Mult Scler. 2017;23(12):1583-1592. doi: 10.1177/1352458517729460 [DOI] [PubMed] [Google Scholar]
- 7.A study to evaluate the safety and efficacy of rituximab in adults with primary progressive multiple sclerosis (OLYMPUS). ClinicalTrials.gov identifier: NCT00087529. Updated September 9, 2015. Accessed December 15, 2021. https://clinicaltrials.gov/ct2/show/NCT00087529
- 8.Hawker K, O’Connor P, Freedman MS, et al. ; OLYMPUS Trial Group . Rituximab in patients with primary progressive multiple sclerosis: results of a randomized double-blind placebo-controlled multicenter trial. Ann Neurol. 2009;66(4):460-471. doi: 10.1002/ana.21867 [DOI] [PubMed] [Google Scholar]
- 9.Confavreux C, Vukusic S, Moreau T, Adeleine P. Relapses and progression of disability in multiple sclerosis. N Engl J Med. 2000;343(20):1430-1438. doi: 10.1056/NEJM200011163432001 [DOI] [PubMed] [Google Scholar]
- 10.Tremlett H, Yousefi M, Devonshire V, Rieckmann P, Zhao Y; UBC Neurologists . Impact of multiple sclerosis relapses on progression diminishes with time. Neurology. 2009;73(20):1616-1623. doi: 10.1212/WNL.0b013e3181c1e44f [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kremenchutzky M, Cottrell D, Rice G, et al. The natural history of multiple sclerosis: a geographically based study. 7. progressive-relapsing and relapsing-progressive multiple sclerosis: a re-evaluation. Brain. 1999;122(pt 10):1941-1950. doi: 10.1093/brain/122.10.1941 [DOI] [PubMed] [Google Scholar]
- 12.Paz Soldán MM, Novotna M, Abou Zeid N, et al. Relapses and disability accumulation in progressive multiple sclerosis. Neurology. 2015;84(1):81-88. doi: 10.1212/WNL.0000000000001094 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lizak N, Malpas CB, Sharmin S, et al. ; MSBase Study Group . Association of sustained immunotherapy with disability outcomes in patients with active secondary progressive multiple sclerosis. JAMA Neurol. 2020;77(11):1398-1407. doi: 10.1001/jamaneurol.2020.2453 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hughes J, Jokubaitis V, Lugaresi A, et al. ; MSBase Study Group . Association of inflammation and disability accrual in patients with progressive-onset multiple sclerosis. JAMA Neurol. 2018;75(11):1407-1415. doi: 10.1001/jamaneurol.2018.2109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Trojano M, Bergamaschi R, Amato MP, et al. ; Italian Multiple Sclerosis Register Centers Group . The Italian multiple sclerosis register. Neurol Sci. 2019;40(1):155-165. doi: 10.1007/s10072-018-3610-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol. 2011;69(2):292-302. doi: 10.1002/ana.22366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.A study of ocrelizumab in participants with primary progressive multiple sclerosis. ClinicalTrials.gov identifier: NCT01194570. Updated May 9, 2022. Accessed December 15, 2021. https://clinicaltrials.gov/ct2/show/NCT01194570
- 18.McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol. 2001;50(1):121-127. doi: 10.1002/ana.1032 [DOI] [PubMed] [Google Scholar]
- 19.Zhang Z, Kim HJ, Lonjon G, Zhu Y; written on behalf of AME Big-Data Clinical Trial Collaborative Group . Balance diagnostics after propensity score matching. Ann Transl Med. 2019;7(1):16. doi: 10.21037/atm.2018.12.10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tremlett H, Paty D, Devonshire V. The natural history of primary progressive MS in British Columbia, Canada. Neurology. 2005;65(12):1919-1923. doi: 10.1212/01.wnl.0000188880.17038.1d [DOI] [PubMed] [Google Scholar]
- 21.Ebers GC, Koopman WJ, Hader W, et al. The natural history of multiple sclerosis: a geographically based study: 8: familial multiple sclerosis. Brain. 2000;123(Pt 3):641-649. doi: 10.1093/brain/123.3.641 [DOI] [PubMed] [Google Scholar]
- 22.Lublin FD, Cutter G, Giovannoni G, Pace A, Campbell NR, Belachew S. Natalizumab reduces relapse clinical severity and improves relapse recovery in MS. Mult Scler Relat Disord. 2014;3(6):705-711. doi: 10.1016/j.msard.2014.08.005 [DOI] [PubMed] [Google Scholar]
- 23.Radue EW, Sprenger T, de Vera A, et al. Effect of fingolimod on evolution of baseline enhancing MRI lesions into persistent T1 hypointense lesions: post hoc analysis of the FREEDOMS study. Mult Scler. 2014;20:112. doi: 10.1177/1352458513488235 [DOI] [PubMed] [Google Scholar]
- 24.Giovannoni G, Cutter G, Sormani MP, et al. Is multiple sclerosis a length-dependent central axonopathy? the case for therapeutic lag and the asynchronous progressive MS hypotheses. Mult Scler Relat Disord. 2017;12:70-78. doi: 10.1016/j.msard.2017.01.007 [DOI] [PubMed] [Google Scholar]
- 25.Montalban X, Sastre-Garriga J, Tintoré M, et al. A single-center, randomized, double-blind, placebo-controlled study of interferon beta-1b on primary progressive and transitional multiple sclerosis. Mult Scler. 2009;15(10):1195-1205. doi: 10.1177/1352458509106937 [DOI] [PubMed] [Google Scholar]
- 26.Tur C, Montalban X, Tintoré M, et al. Interferon beta-1b for the treatment of primary progressive multiple sclerosis: 5-year clinical trial follow-up. Arch Neurol. 2011;68(11):1421-1427. doi: 10.1001/archneurol.2011.241 [DOI] [PubMed] [Google Scholar]
- 27.Roos I, Leray E, Frascoli F, et al. ; MSBase; OFSEP investigators . Delay from treatment start to full effect of immunotherapies for multiple sclerosis. Brain. 2020;143(9):2742-2756. doi: 10.1093/brain/awaa231 [DOI] [PubMed] [Google Scholar]
- 28.Zimmermann M, Brouwer E, Tice JA, et al. Disease-modifying therapies for relapsing-remitting and primary progressive multiple sclerosis: a cost-utility analysis. CNS Drugs. 2018;32(12):1145-1157. doi: 10.1007/s40263-018-0566-9 [DOI] [PubMed] [Google Scholar]
- 29.Sormani MP, De Rossi N, Schiavetti I, et al. ; Musc-19 Study Group . Disease-modifying therapies and Coronavirus Disease 2019 severity in multiple sclerosis. Ann Neurol. 2021;89(4):780-789.doi: 10.1002/ana.26028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sahraian MA, Azimi A, Navardi S, Ala S, Naser Moghadasi A. Evaluation of the rate of COVID-19 infection, hospitalization, and death among Iranian patients with multiple sclerosis. Mult Scler Relat Disord. 2020;46:102472. doi: 10.1016/j.msard.2020.102472 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Butzkueven H, Spelman T, Horakova D, et al. Risk of requiring a wheelchair in primary progressive multiple sclerosis: data from the ORATORIO trial and the MSBase registry. Eur J Neurol. 2022;29(4):1082-1090. doi: 10.1111/ene.14824 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.He A, Merkel B, Brown JWL, et al. ; MSBase study group . Timing of high-efficacy therapy for multiple sclerosis: a retrospective observational cohort study. Lancet Neurol. 2020;19(4):307-316. doi: 10.1016/S1474-4422(20)30067-3 [DOI] [PubMed] [Google Scholar]
- 33.Naegelin Y, Naegelin P, von Felten S, et al. Association of rituximab treatment with disability progression among patients with secondary progressive multiple sclerosis. JAMA Neurol. 2019;76(3):274-281. doi: 10.1001/jamaneurol.2018.4239 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure. Patient Disposition
eTable 1. Baseline Characteristics of the Excluded and Included Patients
eTable 2. Multivariable Cox Regression Model for the Risk of Reaching EDSS 7.0 in Patients Included After the Year 2000 and Propensity Score Matched (n = 316)
Nonauthor Collaborators
