Abstract
Background and purpose
Therapeutic management of relapsing–remitting multiple sclerosis (RRMS) has evolved towards early treatment. The objective was to assess the impact of early treatment initiation on disability progression amongst RRMS first‐line‐treated patients.
Methods
This study included all incident RRMS cases starting interferon or glatiramer acetate for the first time from 1 January 1996 to 31 December 2012 (N = 5279) from 10 MS expert Observatoire Français de la Sclérose en Plaques centres. The delay from treatment start to attaining an irreversible Expanded Disability Status Scale (EDSS) score of 3.0 was compared between the early group (N = 1882; treated within 12 months following MS clinical onset) and the later group using propensity score weighted Kaplan–Meier methods, overall and stratified by age.
Results
Overall, the restricted mean time before reaching EDSS 3.0 from treatment start was 11 years and 2 months for patients treated within the year following MS clinical onset and 10 years and 7 months for patients treated later. Thus, early treated patients gained 7 months (95% confidence interval [CI] 4–11 months) in the time to reach EDSS 3.0 compared to patients treated later (treatment start delayed by 28 months). The difference in restricted mean time was respectively 6 months (95% CI 1–10 months) and 14 months (95% CI 4–24 months) in the ≤40 years age group and in the >40 years age group, in favour of the early group.
Conclusions
Early treatment initiation resulted in a significant reduction of disability progression amongst patients with RRMS, and also amongst older patients.
Keywords: beta‐interferon, disability progression, early treatment, glatiramer acetate, multiple sclerosis, observational studies, propensity score
Using the French MS Registry (Observatoire Français de la Sclérose En Plaques, OFSEP), the real‐life effectiveness of early first‐line treatment initiation was assessed amongst 5279 patients included over the period 1996–2012. Early treated patients, that is, patients in the first year following MS clinical onset, were compared to later treated patients, and the two groups were made comparable using propensity scores. Our work confirmed the overall benefit of early treatment on disability progression but also demonstrated such benefit in the population of patients developing MS after 40 years of age.
INTRODUCTION
In the last 20 years, treatment options for relapsing–remitting multiple sclerosis (RRMS) have improved due to the increased numbers of approved disease‐modifying therapies (DMTs). These drugs are administered in many ways including injections, pills or in‐hospital infusions and the treatment choices are driven by information concerning the risk–benefit balance and the expanding volume of experience from their use by neurologists in real‐life settings [1]. Choosing a DMT is complex, taking into account its efficacy and the severity of the proven and/or potential risks. Some available DMTs, such as first‐line injectable therapies, have milder efficacies and better safety profiles, whereas other therapies such as natalizumab, fingolimod, mitoxantrone or alemtuzumab are more effective but are associated with a higher risk profile which may exhibit rare but severe adverse events such as progressive multifocal leukoencephalopathy, infection or leukaemia. Choosing a first‐line DMT and its initiation timing is a complex process undertaken according to patient characteristics, particularly age, gender and comorbidity. Additionally, disease severity, drug safety profiles and the accessibility of relevant drugs are taken into account [2].
There is accumulating evidence that suggests that action should be initiated as soon as the diagnosis of MS is made [3]. Natural history and immunopathological studies have shown that what happens early during the inflammation phase of the disease influences the long‐term outcome, particularly regarding the development of irreversible neurological disability [3, 4].
Numerous clinical trials (including their extensions) have pointed out that early treatment is beneficial in increasing the time of conversion from a first attack to a clinically definite case of MS compared with delayed treatment [5, 6, 7, 8, 9, 10]. Moreover, early therapeutic intervention is also strongly recommended by the European Committee of Treatment and Research in Multiple Sclerosis and the European Academy of Neurology [2].
Regarding the benefits of early intervention on disability progression, available findings show that the initiation of early treatment seems to delay disability progression [11, 12, 13, 14, 15, 16, 17]. However, results from real‐life studies remain poorly documented [18, 19, 20]. In that context, the objective of the present study was to assess the impact of early treatment initiation on disability progression, both overall and stratified by age, amongst RRMS patients treated initially with first‐line DMTs.
METHODS
Data collection
Data were obtained routinely from the Observatoire Français de la Sclérose en Plaques (OFSEP), a French network of MS expert centres, using the European Database for Multiple Sclerosis (EDMUS) software to collect prospective standardized data on MS patients [21, 22]. Data from the 10 oldest MS expert centres in France (data collection started before 2001; Lyon, Nancy, Bordeaux, Rennes, Toulouse, Nice, Clermont‐Ferrand, Dijon, Besançon and Nîmes) were extracted on 15 December 2018 and produced an initial cohort of 34,582 patients.
Study population
All incident RRMS cases starting an approved first‐line injectable treatment (interferon or glatiramer acetate) for the first time from 1 January 1996 to 31 December 2012 were included in our study. A minimum of three visits and a follow‐up of 5 years from MS clinical onset were also required for inclusion. Patients starting natalizumab, fingolimod and mitoxantrone as a first treatment were excluded. The study period started in 1996 because it corresponded to the approval of the first DMT (beta‐interferon 1b) in France. The period ended in 2012 to get sufficient follow‐up after treatment initiation to observe any disability progression. During this period, only beta‐interferon and glatiramer acetate were available as first‐line DMTs. All patients who had a sustained 6‐month confirmed Expanded Disability Status Scale (EDSS) score of 3.0 or more before starting treatment were also excluded.
Treatment exposure
All patients who were exposed for at least 1 day to a first‐line DMT (beta‐interferon and glatiramer acetate) during the study period were considered as treated. The delay from MS clinical onset to first treatment initiation was used to define groups. Early treatment was defined as an initiation within 12 months from MS clinical onset, whereas later treatment was defined as an initiation after 12 months.
Outcome
Outcome was the time from the start of treatment (baseline) to the attainment of a 6‐month confirmed EDSS score of 3.0, sustained until the final visit. EDSS scores were collected outside of relapses.
Variables
The following variables were considered in the analysis: sex, age at MS clinical onset and at treatment start, year of MS clinical onset, number of relapses in the 12 months preceding treatment initiation, EDSS scores (outside of relapses) within 12 months before treatment start and centre. In the case of multiple EDSS entries, the nearest to treatment start was selected. The variable related to EDSS score before treatment start included a ‘missing’ category.
Statistical analysis
Characteristics were described using the usual indicators. The proportion of patients considered as treated early was calculated for each MS clinical onset year over the study period. The median follow‐up durations were computed using the reverse Kaplan–Meier method [23]. Patients treated early were compared to patients treated later. To control indication bias, a propensity score (PS) was used and obtained from a logistic regression including variables associated with the outcome (level of significance at 20%) [24]. All variables were considered eligible and the variables finally included in each PS model are summarized in Table S1. Two PS models were performed, that is, one which included age at MS clinical onset and another that included age at treatment start. The outcome was then studied using Kaplan–Meier methods weighted by a stabilized inverse probability of treatment weights. The weights were stabilized by multiplication by the marginal probability of receiving the treatment [25]. According to both PS models, the restricted mean survival time (RMST) [26], that is, the restricted mean time before reaching the outcome, between the two groups was reported and compared. A difference in the RMST showed the gain or loss of time before reaching an outcome in one group compared to the other. The outcome was also studied using Cox proportional hazard models weighted by the stabilized inverse probability of treatment weights with robust estimation of variance [27]. Confidence intervals were obtained by bootstrap (k = 1000) [28]. Analyses were conducted based on an intention‐to‐treat framework meaning that the follow‐up was not censored in the case of treatment cessation. Lastly, an analysis stratified on age was performed at MS clinical onset with a cut‐off set at 40 years. This analysis followed the same statistical plan. As a sensitivity analysis, the analysis was replicated after excluding patients with missing EDSS score within 12 months before treatment start. Analyses were performed using R software (R 4.0.3), and the package RISCA [29] was used to obtain the weighted Kaplan–Meier estimators.
RESULTS
Characteristics of the study population
Overall, 5279 patients fulfilled the inclusion criteria and were included in the study (Figure 1). The first treatment was beta‐interferon for 82% of the patients and glatiramer acetate for the remaining 18%. Table 1 provides patient characteristics at treatment start. The median age at MS clinical onset was 31 years (interquartile range [IQR] 25–38). One‐third of patients had at least two relapses in the 12 months before treatment start. The baseline EDSS score was 2.0 or more for 42% of patients.
FIGURE 1.
Study flowchart to assess the disability progression in multiple sclerosis (MS) patients using early first‐line treatment
TABLE 1.
Multiple sclerosis (MS) patient characteristics at treatment initiation: overall and according to age at MS clinical onset
| All | Age at MS clinical onset | ||
|---|---|---|---|
| ≤40 years old group | >40 years old group | ||
| N = 5279 | N = 4197 | N = 1082 | |
| Sex (female) a | 4010 (76%) | 3159 (75.3%) | 8.51 (78.7%) |
| Age at MS clinical onset b | 30.9 (25.2–38.4) | 28.4 (24.1–33.5) | 44.5 (41.9–48.2) |
| Age at treatment start b | 33.9 (27.7–41.2) | 31.3 (26.5–36.6) | 47.2 (44–51.4) |
| MS duration at treatment start b | 1.6 (0.7–3.7) | 1.7 (0.7–3.9) | 1.5 (0.7–3.3) |
| Number of relapses within 12 months before treatment start a | |||
| 0 | 755 (14.3%) | 583 (13.9%) | 172 (15.9%) |
| 1 | 2716 (51.4%) | 2135 (50.9%) | 581 (53.7%) |
| 2 | 1410 (26.7%) | 1135 (27%) | 275 (25.4%) |
| ≥3 | 398 (7.5%) | 344 (8.2%) | 54 (5%) |
| EDSS within 12 months before treatment start a , c | |||
| [0.0; 1.5] | 1740 (33%) | 1438 (34.3%) | 302 (27.9%) |
| [2.0; 3.5] | 1161 (22%) | 859 (20.5%) | 302 (27.9%) |
| ≥4 | 111 (2.1%) | 88 (2.1%) | 23 (2.1%) |
| Missing | 2267 (42.9%) | 1812 (43.2%) | 455 (42.1%) |
| Year of MS onset a | |||
| 1996–2000 | 1620 (30.7%) | 1620 (30.7%) | 280 (25.9%) |
| 2001–2005 | 1780 (33.7%) | 1780 (33.7%) | 359 (33.2%) |
| 2006–2012 | 1879 (35.6%) | 1879 (35.6%) | 443 (40.9%) |
Abbreviation: EDSS, Expanded Disability Status Scale.
N (%).
Median (quartiles).
One‐time and unconfirmed measurement.
Overall, the median delay from MS clinical onset to treatment start was 1.6 years (IQR = 0.7–3.7) and the median follow‐up duration from treatment start was 9.5 years (95% confidence interval [CI] 9.3–9.7). The definition of early led us to consider 1882 patients (36%) as treated early. As shown in Figure 2, the percentage of patients treated early increased over time. After treatment initiation, 923 of the early treated patients (49%) and 1256 (36%) of the later treated patients escalated to higher‐efficacy agents (natalizumab, fingolimod, alemtuzumab or mitoxantrone). Over 15 years, the restricted mean times from start of the initial treatment before escalation were respectively 9.6 years (95% CI 9.3–9.9) and 10.6 years (95% CI 10.4–10.8) in the two groups.
FIGURE 2.
Percentage of multiple sclerosis (MS) patients treated early (defined as an initiation within the year following MS clinical onset) according to year of MS clinical onset
Comparison between early and later treated patients on disability progression
The PS distributions showed a noticeable overlap between the two groups, as shown in Figure S1. After weighting, the two groups were well balanced with standardized mean differences or Mahalanobis distances of around 10% for all patients' characteristics (Table S2).
Overall, the restricted mean survival time before reaching an EDSS score of 3.0 was 10 years and 10 months (10 years 7 months to 10 years 11 months) over the 15‐year study period. Figure 3 shows the Kaplan–Meier estimates for each PS definition, Table 2 summarizes the RMSTs and Figure 5 presents the hazard ratio. After PS weighting, when considering comparable age at MS clinical onset, the RMST before reaching EDSS 3.0 from treatment start was 11 years and 2 months for patients treated within the year following MS clinical onset and 10 years and 7 months for patients treated later. In the early group, the median delay before treatment start from MS clinical onset was 6 months, whereas it was 35 months for patients treated later. Therefore, compared with patients with a median delayed treatment of 28 months, early treated patients gained 7 months (95% CI 4–11 months) in the time to reach an EDSS score of 3.0. Early treated patients showed a 16% decreased hazard of reaching EDSS 3.0 compared with later treated patients (hazard ratio 0.84; 95% CI 0.76–0.92). When patients had a comparable age at treatment start, early treated patients gained 4 months (95% CI 0.5–8 months) in comparison to patients treated later.
FIGURE 3.
Kaplan–Meier estimates of the time before reaching an EDSS score of 3.0 from treatment start for multiple sclerosis (MS) patients treated by a first‐line treatment as an initial treatment within the year following MS clinical onset (early) or later (late), according to two propensity score models
TABLE 2.
Restricted mean time (RMST) before reaching EDSS 3.0 in early and late treated groups of multiple sclerosis (MS) patients: overall, according to age at MS clinical onset, and in the sensitivity analysis from two propensity score (PS) models
| Group of patients | Early | Late | Median delay in treatment start (months) | Difference in RMST (months) | ||
|---|---|---|---|---|---|---|
| Delay before treatment start | RMST | Delay before treatment start | RMST | |||
| PS model adjusted for age at MS clinical onset | ||||||
| All | 6 m (4–9 m) | 11 y and 2 m | 35 m (20–61 m) | 10 y and 7 m | 28 | 7 [4; 11] |
| ≤40 years old group | 6 m (4–9 m) | 11 y and 6 m | 36 m (21–62 m) | 11 y | 30 | 6 [1; 10] |
| >40 years old group | 6 m (4–9 m) | 9 y and 11 m | 32 m (19–58 m) | 8 y and 8 m | 26 | 14 [4; 24] |
| Sensitivity analysis a | 7 m (5–9 m) | 11 y and 7 m | 35 m (20–62 m) | 11 y and 1 m | 28 | 6 [0.4; 12] |
| PS model adjusted for age at treatment start | ||||||
| All | 6 m (4–9 m) | 11 y | 35 m (20–61 m) | 10 y and 8 m | 28 | 4 [0.5; 8] |
| ≤40 years old group | 6 m (4–9 m) | 11 y and 5 m | 36 m (21–62 m) | 11 y and 1 m | 30 | 4 [−0,3; 9] |
| >40 years old group | 6 m (4–9 m) | 9 y and 8 m | 32 m (19–58 m) | 8 y and 9 m | 26 | 11 [2; 21] |
| Sensitivity analysis a | 7 m (5–9 m) | 11 y and 6 m | 11 y and 2 m | 28 | 3 [−2; 9] | |
Notes: Delay before treatment start: median (quartiles). RMST, restricted mean survival time, corresponds to the restricted mean time before reaching the outcome. For instance, over 15 years restricted, the mean time for patients treated within the year following MS onset to reach an EDSS score of 3 confirmed at 6 months and sustained was 11 years and 2 months, and it was 10 years and 7 months for patients treated later; thus patients treated early reached the outcome around 7 months later than patients treated after 1 year following MS onset. Early treated patients gained 7 months on the time before moderate disability in comparison to patients treated later.
Abbreviations: m, month; MS, multiple sclerosis; PS, propensity score; y; year.
Sensitivity analysis, excluding patients without EDSS 12 months before treatment start.
FIGURE 5.
Estimated hazard ratio (HR) of early treatment initiation effect on attainment of EDSS 3.0 in multiple sclerosis (MS) patients: overall, according to age at MS clinical onset, and in the sensitivity analysis from two propensity score (PS) models
Stratification by age
Multiple sclerosis started in patients before the age of 40 amongst 4197 (79%) patients (≤40 years old group) and after 40 years old for 1082 patients (21%) (>40 years old group). The median delays from MS clinical onset to treatment start were 1.7 years (IQR = 0.7–3.9) and 1.5 years respectively (0.7–3.3) in the two subgroups, whose characteristics are shown in Table 1.
After PS weighting, when patients from the ≤40 years old group had comparable age at MS clinical onset, this study showed that, compared with patients with a median delayed treatment of 30 months, early treated patients gained 6 months (95% CI 1–10 months) before reaching an EDSS of 3.0 (Table 2, Figure 4). Early treated patients showed a 13% decreased hazard of reaching EDSS 3.0 compared with later treated patients (Figure 5). No difference was shown for patients who had a comparable age at treatment start.
FIGURE 4.
Kaplan–Meier estimates of the time before reaching an EDSS score of 3.0 from treatment start for multiple sclerosis (MS) patients treated by a first‐line treatment as initial treatment within the year following MS clinical onset (early) or later (late), according to propensity score models and stratified by age at MS clinical onset
In the >40 years old group and using the PS including age at MS clinical onset, this study showed that, compared with patients with a median delayed treatment of 26 months, early treated patients gained 14 months (95% CI 4–24 months) before reaching EDSS 3.0. Early treated patients showed 24% decreased hazard of reaching EDSS 3.0 compared with later treated patients (Figure 5). Including age at treatment start in the PS, early treated patients gained approximately 11 months (95% CI 2–21 months) before reaching an EDSS score of 3.0 compared to patients treated later.
Sensitivity analysis
Results from sensitivity analysis, that is, excluding patients without an EDSS score before treatment start, were close to the main results but somewhat deflated. They also failed to achieve statistical significance, probably due to a lack of power related to the smaller population size (Table 2 and Figure 5).
DISCUSSION
This study aimed to assess the impact of early treatment initiation on disability progression, and a beneficial effect was found. Indeed, by accelerating the median initiation of treatment by 2 years and 4 months, the time to reach 6‐month confirmed and sustained EDSS 3.0 was delayed by about 7 months. This significant impact was observed overall and after stratification by age at MS clinical onset (≤40 years old, 6 months for 2 and half years of treatment delayed; >40 years old, 14 months for 2 years of treatment delayed).
The risk of faster disability progression with increased age at MS clinical onset has been shown in large cohort studies [4] which reinforces the need for early control of disease activity overall as well as with older RRMS patients.
In addition, it should be noted that early active MS patients who are more at risk of sequelae after relapse are expected to be seen rapidly by a neurologist and then treated early. Contrarily, MS patients with mild symptoms at first relapse might not be seen by a neurologist and therefore have decreased chances of being treated early. In such patients, treatment initiation may occur several years later. Such differences in MS management may minimize the magnitude of early intervention. Beyond clinical parameters (such as symptom severity and magnetic resonance imaging [MRI] data), an initial consultation with a neurologist is the starting point for treatment initiation. Unfortunately, it was not possible to study the delay between the first consultation with a neurologist and treatment start since the date of a first consultation was not systematically recorded in the OFSEP database.
This study included patients with at least 5 years of follow‐up in an MS expert centre, either exclusively or in collaboration with a private neurologist either from disease onset or later in the course of the disease. Although follow‐up can be supervised by both MS experts or private neurologists, it is thought that our study probably focused on patients with particular characteristics that required clinical follow‐up in MS expert centres. A previous study in Lorraine, France, which compared the natural history of an MS expert centre population with patients followed outside an MS centre, showed that patients in MS expert centres were younger and had earlier disability accumulation [30]. However, the comparison of the overall OFSEP cohort to the MS Lorraine population‐based registry (ReLSEP) did not highlight any differences. It cannot be excluded that a potential recruitment bias may have minimized the benefit of early intervention.
In our study, the focus was on only one definition of early treatment initiation. Two other definitions have been considered (within 6 months and within 24 months). The definition studied in this work was chosen as a compromise between concepts and practical considerations. Our study aimed to measure the effect of such therapeutic decision. Treatment changes over follow‐up were not considered to avoid adjustment on post‐baseline variables.
When comparing this study with the literature, our findings seem less favourable but our result should be interpreted with regard to the delay of treatment. Indeed, in an Italian study focusing on early beta‐interferon initiation, a 40% reduction in the risk of reaching EDSS 4.0 was shown for a median delayed treatment of around 5 years [16]. Another recent study performed in Sweden showed that patients who started treatment 3 years after MS onset reached EDSS 4 sooner (hazard ratio 2.64, 95% CI 1.71–4.08) than patients who started treatment within a year from MS onset [14]. Given that these authors confirmed late older age at MS clinical onset as a bad prognostic factor, our group performed a stratified analysis by age at MS clinical onset. A Danish study showed a 42% higher hazard ratio for reaching an EDSS score of 6 compared with early treatment (median treatment delay of 4 years) but did not find consistent results according to age. Indeed, the benefit was significant amongst patients treated before 40 years of age whereas the result was not significant after 40 years old [17]. Several differences could explain this discrepancy between studies such as data sources (MS expert centres vs. a population‐based registry) and outcome (EDSS score of 3 vs. EDSS score of 6). Similar to the study from Chalmer et al. [17], our study included patients over a long time frame, from 1996 to 2013, during which treatment practices have evolved. To limit a potential issue, the year of MS onset was considered in our analyses and was included in the PS models.
One of our expected findings was the evidence of an earlier initiation of DMTs over time which is probably related to several factors or changes related to MS management. First, changes in MS diagnostic criteria enabled earlier MS diagnosis defined as the first clinical demyelinating event [31]. Secondly, treatment indications changed over time allowing treatment from the first relapse instead of waiting for the occurrence of two or three clinical events [5, 6, 7, 8, 9, 10]. Finally, the increasing knowledge regarding MS prognosis [1, 3, 4] combined with a broader experience using DMTs [2] led neurologists to initiate treatment early amongst patients with MS.
With regard to the statistical analysis, two different PS models were chosen to consider two different kinds of information related to patient age. The model adjusted for age at MS clinical onset makes patients comparable in age at the first clinical symptoms which is an important prognostic factor of disability progression. Secondly, the model adjusted for age at treatment start allowed patients to be comparable in age at the time of the treatment decision and therefore combined both age at MS onset and disease duration before treatment start. The difference between the two models is moderate in the group treated early, which is not surprising as there was a maximum of 1 year between the two measurements. The two analyses are presented because they complement each other. Age at treatment start had to be considered since it reflected the age of the treatment decision. However, the analysis based on the age at MS onset reinforced the idea that age is a predominant prognostic factor. Indeed, the beneficial effect of early treatment may be stronger when patients were compared by age at MS clinical onset which was 7 months compared to 4 months gained. Regarding PS model specification, the choice of variables to be included is often discussed in the literature. A statistical criterion was chosen [24, 32, 33, 34]. Based on that, the number of relapses before treatment start was not included in the PS models. In our opinion, there is no doubt that this parameter is associated with treatment start, but it was not shown to be associated with disability progression in our dataset.
The main limitations of the present study were a lack of MRI data as well as missing data related to the EDSS score at treatment start. During the study period, access to MRI was limited, and this is the reason why relapse events and disability scores were most often used to make therapeutic decisions. Moreover, cerebral and spinal sequences have changed over time and this lack of standardization may limit the use of such data in observational studies.
The second limitation led us to create a ‘missing EDSS’ category to minimize a potential selection bias coming from EDSS availability. When comparing patients according to baseline EDSS availability, patients with missing baseline EDSS had fewer relapses in the previous year and a longer disease duration (Table S3). As a sensitivity analysis, when excluding patients with missing EDSS, results were close to the main results but not statistically significant.
Outcome was defined from the EDSS scores measured during visits to the MS expert centres. This study did not require any specific rhythm in follow‐up consultations besides the request of three visits as inclusion criteria. Visit frequency differed from one MS patient to another; therefore, the risk of inaccurate outcomes due to irregular visits cannot be excluded. In addition, as with all long‐term observational studies, it cannot be ruled out that time is running out to observe the EDSS 6.0 outcome, and so this disability hallmark was not considered in this observational study.
In conclusion, the findings of the present study indicate that early treatment initiation (within 12 months following MS clinical onset) becomes more frequent over time in France and results in a significant delayed disability progression amongst patients with relapsing MS. This benefit was found overall and occurs even amongst patients over the age of 40 years. It is considered that our study is in line with the concept of ‘time is brain’ in MS since such results encourage promoting an early diagnosis of MS and an early treatment. Further research should be considered to compare two strategies: the effect of an early escalation strategy (using both clinical and MRI criteria for DMT monitoring) versus systematic early ‘high efficacy’ treatment.
AUTHOR CONTRIBUTIONS
Mathilde LEFORT: Conceptualization (equal); data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); software (equal); validation (equal); writing – original draft (equal); writing – review and editing (equal). Sandra Vukusic: Methodology (equal); writing – review and editing (equal). Romain Casey: Data curation (equal); methodology (equal); writing – review and editing (equal). Gilles Edan: Conceptualization (equal); methodology (equal); project administration (equal); writing – review and editing (equal). Emmanuelle Leray: Conceptualization (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); supervision (equal); validation (equal); writing – original draft (equal); writing – review and editing (equal).
CONFLICT OF INTEREST
M. Lefort reports that she had travel grants from Roche SAS. S. Vukusic reports that she had consulting and lecturing fees, travel grants and unconditional research support from Biogen, Celgène, Geneuro, Genzyme, MedDay, Merck Serono, Novartis, Roche, Sanofi Aventis and Teva Pharma. R. Casey reports that he had no conflict of interest. G. Edan reports that he had personal honoraria for lectures or consulting from Bayer, Biogen, LFB, Merck, Novartis, Roche, Sanofi; research support from Bayer, Biogen, Genzyme, Merck, Novartis, Roche and Teva Pharma. E. Leray reports that she had consulting and lecture fees or travel grants from Biogen, Genzyme, MedDay Pharmaceuticals, Merck Serono, Novartis, Sanofi Aventis and Roche.
PATIENT CONSENT
All patients enrolled in the OFSEP database provided informed consent for participation and that their data be stored and used for research. Data confidentiality and safety were ensured according to the recommendations of the French organization Commission Nationale Informatique et Libertés (CNIL). OFSEP received approval for storing clinical, biological and imaging data for research purposes. The study did not require any additional procedures in accordance with French legislation. OFSEP was registered at clinicaltrials.gov under NCT02889965.
Supporting information
ACKNOWLEDGEMENTS
The authors would like to thank the Observatoire Français de la Sclérose en Plaques (OFSEP, http://www.ofsep.org/fr/), which is supported by a grant provided by the French government and administered by the Agence Nationale de la Recherche, within the framework of the ‘Investments for the Future’ programme (reference ANR‐10‐COHO‐002), the Eugène Devic EDMUS Foundation against Multiple Sclerosis and the ARSEP Foundation. The authors would also like to thank Sarina Yaghobian on behalf of AcaciaTools for medical editing services.
APPENDIX A.
List of OFSEP investigators
Steering Committee
Romain Casey, PhD, Observatoire français de la sclérose en plaques (OFSEP), Centre de coordination national, Lyon/Bron, France;
François Cotton, MD, Hospices civils de Lyon, Hôpital Lyon sud, Service d'imagerie médicale et interventionnelle, Lyon/Pierre‐Bénite, France;
Jérôme De Sèze, MD, Hôpitaux universitaire de Strasbourg, Hôpital de Hautepierre, Service des maladies inflammatoires du système nerveux—neurologie, Strasbourg, France;
Pascal Douek, MD, Union pour la lutte contre la sclérose en plaques (UNISEP), Ivry‐sur‐Seine, France;
Francis Guillemin, MD, CIC 1433 Epidémiologie Clinique, Centre hospitalier régional universitaire de Nancy, Inserm et Université de Lorraine, Nancy, France;
David Laplaud, MD, Centre hospitalier universitaire de Nantes, Hôpital nord Laennec, Service de neurologie, Nantes/Saint‐Herblain, France;
Christine Lebrun‐Frenay, MD, Centre hospitalier universitaire de Nice, Université Nice Côte d'Azur, Hôpital Pasteur, Service de neurologie, Nice, France;
Alexandre Pachot, PhD, Hospices civils de Lyon, Direction de la Recherche en Santé, Lyon, France;
Thibault Moreau, MD, Centre hospitalier universitaire Dijon Bourgogne, Hôpital François Mitterrand, Service de neurologie, maladies inflammatoires du système nerveux et neurologie générale, Dijon, France;
Javier Olaiz, PhD, Université Claude Bernard Lyon 1, Lyon ingéniérie projets, Lyon, France;
Jean Pelletier, MD, Assistance publique des hôpitaux de Marseille, Centre hospitalier de la Timone, Service de neurologie et unité neuro‐vasculaire, Marseille, France;
Claire Rigaud‐Bully, Fondation Eugène Devic EDMUS contre la sclérose en plaques, Lyon, France;
Bruno Stankoff, MD, Assistance publique des hôpitaux de Paris, Hôpital Saint‐Antoine, Service de neurologie, Paris, France;
Sandra Vukusic, MD, Hospices civils de Lyon, Hôpital Pierre Wertheimer, Service de neurologie A, Lyon/Bron, France;
Hélène Zephir, MD, Centre hospitalier universitaire de Lille, Hôpital Salengro, Service de neurologie, Lille, France.
Investigators
Romain Marignier, MD, Hospices civils de Lyon, Hôpital Pierre Wertheimer, Service de neurologie A, Lyon/Bron, France;
Marc Debouverie, MD, Centre hospitalier régional universitaire de Nancy, Hôpital central, Service de neurologie, Nancy, France;
Gilles Edan, MD, Centre hospitalier universitaire de Rennes, Hôpital Pontchaillou, Service de neurologie, Rennes, France;
Jonathan Ciron, MD, Centre hospitalier universitaire de Toulouse, Hôpital Purpan, Service de neurologie inflammatoire et neuro‐oncologie, Toulouse, France;
Aurélie Ruet, MD, Centre hospitalier universitaire de Bordeaux, Hôpital Pellegrin, Service de neurologie, Bordeaux, France;
Nicolas Collongues, MD, Hôpitaux universitaire de Strasbourg, Hôpital de Hautepierre, Service des maladies inflammatoires du système nerveux—neurologie, Strasbourg, France;
Catherine Lubetzki, MD, Assistance publique des hôpitaux de Paris, Hôpital de la Pitié‐Salpêtrière, Service de neurologie, Paris, France;
Hélène Zephir, MD, Centre hospitalier universitaire de Lille, Hôpital Salengro, Service de neurologie, Lille, France;
Pierre Labauge, MD, Centre hospitalier universitaire de Montpellier, Hôpital Gui de Chauliac, Service de neurologie, Montpellier, France;
Gilles Defer, MD, Centre hospitalier universitaire de Caen Normandie, Service de neurologie, Hôpital Côte de Nacre, Caen, France;
Mikaël Cohen, MD, Centre hospitalier universitaire de Nice, Université Nice Côte d'Azur, Hôpital Pasteur, Service de neurologie, Nice, France;
Agnès Fromont, MD, Centre hospitalier universitaire Dijon Bourgogne, Hôpital François Mitterrand, Service de neurologie, maladies inflammatoires du système nerveux et neurologie générale, Dijon, France;
Sandrine Wiertlewsky, MD, Centre hospitalier universitaire de Nantes, Hôpital nord Laennec, Service de neurologie, Nantes/Saint‐Herblain, France;
Eric Berger, MD, Centre hospitalier régional universitaire de Besançon, Hôpital Jean Minjoz, Service de neurologie, Besançon, France;
Pierre Clavelou, MD, Centre hospitalier universitaire de Clermont‐Ferrand, Hôpital Gabriel‐Montpied, Service de neurologie, Clermont‐Ferrand, France;
Bertrand Audoin, MD, Assistance publique des hôpitaux de Marseille, Centre hospitalier de la Timone, Service de neurologie et unité neuro‐vasculaire, Marseille, France;
Claire Giannesini, MD, Assistance publique des hôpitaux de Paris, Hôpital Saint‐Antoine, Service de neurologie, Paris, France;
Olivier Gout, MD, Fondation Adolphe de Rothschild de l'œil et du cerveau, Service de neurologie, Paris, France;
Eric Thouvenot, MD, Centre hospitalier universitaire de Nîmes, Hôpital Carémeau, Service de neurologie, Nîmes, France;
Olivier Heinzlef, MD, Centre hospitalier intercommunal de Poissy Saint‐Germain‐en‐Laye, Service de neurologie, Poissy, France;
Abdullatif Al‐Khedr, MD, Centre hospitalier universitaire d'Amiens Picardie, Site sud, Service de neurologie, Amiens, France;
Bertrand Bourre, MD, Centre hospitalier universitaire Rouen Normandie, Hôpital Charles‐Nicolle, Service de neurologie, Rouen, France;
Olivier Casez, MD, Centre hospitalier universitaire Grenoble‐Alpes, Site nord, Service de neurologie, Grenoble/La Tronche, France;
Philippe Cabre, MD, Centre hospitalier universitaire de Martinique, Hôpital Pierre Zobda‐Quitman, Service de Neurologie, Fort‐de‐France, France;
Alexis Montcuquet, MD, Centre hospitalier universitaire Limoges, Hôpital Dupuytren, Service de neurologie, Limoges, France;
Abir Wahab, MD, Assistance publique des hôpitaux de Paris, Hôpital Henri Mondor, Service de neurologie, Créteil, France;
Jean‐Philippe Camdessanché, MD, Centre hospitalier universitaire de Saint‐Étienne, Hôpital Nord, Service de neurologie, Saint‐Étienne, France;
Serge Bakchine, MD, Centre hospitalier universitaire de Reims, Hôpital Maison‐Blanche, Service de neurologie, Reims, France;
Aude Maurousset, MD, Centre hospitalier régional universitaire de Tours, Hôpital Bretonneau, Service de neurologie, Tours, France;
Haifa Ben Nasr, MD, Centre hospitalier sud francilien, Service de neurologie, Corbeil‐Essonnes, France;
Karolina Hankiewicz, MD, Centre hospitalier de Saint‐Denis, Hôpital Casanova, Service de neurologie, Saint‐Denis, France;
Corinne Pottier, MD, Centre hospitalier de Pontoise, Service de neurologie, Pontoise, France;
Nicolas Maubeuge, MD, Centre hospitalier universitaire de Poitiers, Site de la Milétrie, Service de neurologie, Poitiers, France;
Céline Labeyrie, MD, Assistance publique des hôpitaux de Paris, Hôpital Bicêtre, Service de neurologie, Le Kremlin‐Bicêtre, France;
Chantal Nifle, MD, Centre hospitalier de Versailles, Hôpital André‐Mignot, Service de neurologie, Le Chesnay, France.
Lefort M, Vukusic S, Casey R, Edan G, Leray E, OFSEP investigators . Disability progression in multiple sclerosis patients using early first‐line treatments. Eur J Neurol. 2022;29:2761‐2771. doi: 10.1111/ene.15422
The OFSEP investigators are listed in the Appendix A.
Funding information
This work was funded by the Eugène Devic EDMUS Foundation against Multiple Sclerosis and by the ARSEP Foundation (Fondation pour l'Aide à la Recherche sur la Sclérose en Plaques)
DATA AVAILABILITY STATEMENT
The individual data from this study can be obtained upon request and after validation from the OFSEP scientific committee (see: http://www.ofsep.org/fr/)
REFERENCES
- 1. Thompson AJ, Baranzini SE, Geurts J, Hemmer B, Ciccarelli O. Multiple sclerosis. Lancet. 2018;391(10130):1622‐1636. doi: 10.1016/S0140-6736(18)30481-1 [DOI] [PubMed] [Google Scholar]
- 2. Montalban X, Gold R, Thompson AJ, et al. ECTRIMS/EAN guideline on the pharmacological treatment of people with multiple sclerosis. Mult Scler. 2018;24(2):96‐120. [DOI] [PubMed] [Google Scholar]
- 3. Tintoré M. New options for early treatment of multiple sclerosis. J Neurol Sci. 2009;277(SUPPL. 1):9‐11. [DOI] [PubMed] [Google Scholar]
- 4. Leray E, Yaouanq J, Le Page E, et al. Evidence for a two‐stage disability progression in multiple sclerosis. Brain. 2010;133(7):1900‐1913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kappos L, Freedman MS, Polman CH, et al. Long‐term effect of early treatment with interferon beta‐1b after a first clinical event suggestive of multiple sclerosis: 5‐year active treatment extension of the phase 3 BENEFIT trial. Lancet Neurol. 2009;8(11):987‐997. doi: 10.1016/S1474-4422(09)70237-6 [DOI] [PubMed] [Google Scholar]
- 6. Comi G, Martinelli V, Rodegher M, et al. Effects of early treatment with glatiramer acetate in patients with clinically isolated syndrome. Mult Scler. 2013;19:1074‐1083. [DOI] [PubMed] [Google Scholar]
- 7. Edan G, Kappos L, Montalbán X, et al. Long‐term impact of interferon beta‐1b in patients with CIS: 8‐year follow‐up of BENEFIT. J Neurol Neurosurg Psychiatry. 2014;85(11):1183‐1189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Fujihara K. BENEFIT 8‐year results provide further support for the long‐term value of early treatment of multiple sclerosis. J Neurol Neurosurg Psychiatry. 2014;85(11):1179. [DOI] [PubMed] [Google Scholar]
- 9. Kappos L, Edan G, Freedman M, et al. The 11‐year long‐term follow‐up study from the randomized BENEFIT CIS trial. Neurology. 2016;87(10):978‐987. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Hartung HP, Graf J, Kremer D. Long‐term follow‐up of multiple sclerosis studies and outcomes from early treatment of clinically isolated syndrome in the BENEFIT 11 study. J Neurol. 2019;267:308‐316. [DOI] [PubMed] [Google Scholar]
- 11. Reiffel JA. Propensity score matching: the ‘devil is in the details’ where more may be hidden than you know. Am J Med. 2020;133(2):178‐181. doi: 10.1016/j.amjmed.2019.08.055 [DOI] [PubMed] [Google Scholar]
- 12. Rollot F, Casey R, Leray E, et al. Cumulative effects of therapies on disability in relapsing multiple sclerosis. Mult Scler J. 2021;27:1‐11. [DOI] [PubMed] [Google Scholar]
- 13. Landfeldt E, Castelo‐Branco A, Svedbom A, Löfroth E, Kavaliunas A, Hillert J. The long‐term impact of early treatment of multiple sclerosis on the risk of disability pension. J Neurol. 2018;265:701‐707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kavaliunas A, Manouchehrinia A, Stawiarz L, et al. Importance of early treatment initiation in the clinical course of multiple sclerosis. Mult Scler J. 2016;23:1233‐1240. [DOI] [PubMed] [Google Scholar]
- 15. Cocco E, Sardu C, Spinicci G, et al. Influence of treatments in multiple sclerosis disability: a cohort study. Mult Scler J. 2015;21(4):433‐441. doi: 10.1177/1352458514546788 [DOI] [PubMed] [Google Scholar]
- 16. Trojano M, Pellegrini F, Paolicelli D, et al. Real‐life impact of early interferon beta therapy in relapsing multiple sclerosis. Ann Neurol. 2009;66(4):513‐520. [DOI] [PubMed] [Google Scholar]
- 17. Chalmer TA, Baggesen LM, Nørgaard M, Koch‐Henriksen N, Magyari M, Sorensen PS. Early versus later treatment start in multiple sclerosis: a register‐based cohort study. Eur J Neurol. 2018;25(10):1262‐e110. [DOI] [PubMed] [Google Scholar]
- 18. Freedman MS, Comi G, De Stefano N, et al. Moving toward earlier treatment of multiple sclerosis: findings from a decade of clinical trials and implications for clinical practice. Mult Scler Relat Disord. 2014;3(2):147‐155. doi: 10.1016/j.msard.2013.07.001 [DOI] [PubMed] [Google Scholar]
- 19. Merkel B, Butzkueven H, Traboulsee AL, Havrdova E, Kalincik T. Timing of high‐efficacy therapy in relapsing–remitting multiple sclerosis: a systematic review. Autoimmun Rev. 2017;16(6):658‐665. doi: 10.1016/j.autrev.2017.04.010 [DOI] [PubMed] [Google Scholar]
- 20. Fernández Ó. Is there a change of paradigm towards more effective treatment early in the course of apparent high‐risk MS? Mult Scler Relat Disord. 2017;17:75‐83. doi: 10.1016/j.msard.2017.07.003 [DOI] [PubMed] [Google Scholar]
- 21. Vukusic S, Casey R, Rollot F, et al. Observatoire français de la sclérose en plaques (OFSEP): a unique multimodal nationwide MS registry in France. Mult Scler J. 2018;26:118‐122. [DOI] [PubMed] [Google Scholar]
- 22. Confavreux C, Compston DAS, Hommes R, McDonald WI, Thompson AJ. EDMUS, a European database for multiple sclerosis. J Neurol Neurosurg Psychiatry. 1992;55:671‐676. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Shuster JJ. Median follow‐up in clinical trials. J Clin Oncol. 1991;9(1):191‐192. [DOI] [PubMed] [Google Scholar]
- 24. Brookhart MA, Schneeweiss S, Rothman KJ, Glynn RJ, Avorn J, Stürmer T. Variable selection for propensity score models. Am J Epidemiol. 2006;163(12):1149‐1156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Vaughan AS, Kelley CF, Luisi N, Del Rio C, Sullivan PS, Rosenberg ES. An application of propensity score weighting to quantify the causal effect of rectal sexually transmitted infections on incident HIV among men who have sex with men data analysis, statistics and modelling. BMC Med Res Methodol. 2015;15(1):1‐9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Zhao L, Claggett B, Tian L, et al. On the restricted mean survival time curve in survival analysis. Biometrics. 2016;72(1):215‐221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Cole SR, Hernán MA. Adjusted survival curves with inverse probability weights. Comput Methods Programs Biomed. 2004;75:45‐49. [DOI] [PubMed] [Google Scholar]
- 28. Austin PC. Variance estimation when using inverse probability of treatment weighting (IPTW) with survival analysis. Stat Med. 2016;35(30):5642‐5655. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Foucher Y, Le Borgne F, Dantan E, Gillaizeau F, Chatton A, Combescure C. RISCA: causal inference and prediction in cohort‐based analyses. 2020.
- 30. Debouverie M, Pittion‐Vouyovitch S, Louis S, Guillemin F. Natural history of multiple sclerosis in a population‐based cohort. Eur J Neurol. 2008;15(9):916‐921. [DOI] [PubMed] [Google Scholar]
- 31. Sorensen PS, Sellebjerg F, Hartung HP, Montalban X, Comi G, Tintoré M. The apparently milder course of multiple sclerosis: changes in the diagnostic criteria, therapy and natural history. Brain. 2020;143(9):2637‐2652. [DOI] [PubMed] [Google Scholar]
- 32. Witte J, Didelez V. Covariate selection strategies for causal inference: classification and comparison. Biom J. 2018;61:1‐20. [DOI] [PubMed] [Google Scholar]
- 33. Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399‐424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Austin PC, Grootendorst P, Anderson GM. A comparison of the ability of different propensity score models to balance measured variables between treated and untreated subjects: a Monte Carlo study. Stat Med. 2007;26(4):734‐753. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The individual data from this study can be obtained upon request and after validation from the OFSEP scientific committee (see: http://www.ofsep.org/fr/)