Available via license: CC BY-ND 4.0
Content may be subject to copyright.
1
Comparative effectiveness of natalizumab and fingolimod and injectable
therapies in patients with pediatric multiple sclerosis: A registry-based
retrospective cohort study
Tim Spelman,1 Gabrielle Simoneau,2 Robert Hyde,3 Robert Kuhelj,3 Raed Alroughani,4
Serkan Ozakbas,5 Rana Karabudak,6 Bassem Yamout,7 Samia J. Khoury,7 Murat Terzi,8
Cavit Boz,9 Dana Horakova,10 Eva Kubala Havrdova,10 Bianca Weinstock-Guttman,11
Francesco Patti,12 Ayse Altintas,13 Saloua Mrabet,14 Jihad Inshasi,15 Helmut
Butzkueven,16 on behalf of the MSBase Investigators
1Burnet Institute, Melbourne, Victoria, Australia; 2Biogen, Toronto, Canada; 3Biogen,
Baar, Switzerland; 4 Division of Neurology, Department of Medicine, Amiri Hospital,
Sharq, Kuwait; 5Dokuz Eylul University, Konak/Izmir, Turkey; 6Hacettepe University,
Ankara, Turkey; 7 Nehme and Therese Tohme Multiple Sclerosis Center, American
University of Beirut Medical Center, Beirut, Lebanon; 819 Mayis University, Samsun,
Turkey; 9KTU Medical Faculty Farabi Hospital, Trabzon, Turkey; 10Department of
Neurology and Center of Clinical Neuroscience, First Faculty of Medicine, Charles
University in Prague and General University Hospital, Prague, Czech Republic,
11Department of Neurology, Buffalo General Medical Center;12Department of Medical
and Surgical Sciences and Advanced Technologies, GF Ingrassia, Catania, Italy;13
Department of Neurology, School of Medicine and Koc University Research Center for
Translational Medicine (KUTTAM), Koc University, Istanbul, Turkey; 14Department of
Neurology, Razi University Hospital, Tunis, Tunisia; 15 Rashid Hospital, Dubai, United
Arab Emirates; 16 Department of Neuroscience, Central Clinical School, Monash
University, Melbourne, Victoria, Australia, and Department of Neurology, Box Hill
Hospital, Monash University, Box Hill, Victoria, Australia
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.
2
ABSTRACT
Background and Objectives: Patients with pediatric-onset multiple sclerosis (POMS)
typically experience higher levels of inflammation with more frequent relapses and
reach irreversible disability at a younger age than adult-onset patients. There have been
few randomized placebo-controlled clinical trials of multiple sclerosis (MS) disease-
modifying therapies (DMTs) in patients with POMS, and most available data are based
on observational studies of off-label use of DMTs approved for adults. We assessed the
effectiveness of natalizumab compared with fingolimod using injectable platform
therapies as a reference in pediatric patients in the global MSBase registry.
Methods: This retrospective study included patients with POMS who initiated treatment
with an injectable DMT, natalizumab, or fingolimod between January 1, 2006, and May
3, 2021 (N=1218). The primary outcome was the time to first relapse from index therapy
initiation. Secondary study outcomes included annualized relapse rate; proportions of
relapse-free patients at 1, 2, and 5 years post baseline; time to treatment
discontinuation; and times to 24-week confirmed disability worsening and confirmed
disability improvement.
Results: Patients treated with fingolimod had a significantly lower risk of relapse than
patients treated with injectable DMT (hazard ratio [HR], 0.49; 95% confidence interval
[CI], 0.29–0.83; P=0.008). After adjustment for prior DMT experience in the unmatched
sample, patients treated with natalizumab had a significantly lower risk of relapse than
patients treated either with injectable DMT (HR, 0.15; 95% CI, 0.07–0.31; P<0.001) or
fingolimod (HR, 0.37; 95% CI, 0.14–1.00; P=0.049). The adjusted secondary study
outcomes were generally consistent with the primary outcome or with previous
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
3
observations. The findings in the inverse probability treatment weighting–adjusted
patient populations were confirmed in multiple sensitivity analyses.
Discussion: Our results suggest that natalizumab and fingolimod have broadly
equivalent therapeutic efficacies in patients with POMS, consistent with previous studies
of natalizumab and fingolimod in adult-onset patients and POMS. However, analyses of
relapse outcomes suggest natalizumab is superior to fingolimod in the control of
relapses in this population with high rates of new inflammatory activity.
Classification of Evidence: This study provides Class III evidence that natalizumab
may provide better disease control than fingolimod in patients with POMS.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
4
INTRODUCTION
Multiple sclerosis (MS) is a chronic inflammatory demyelinating and neurodegenerative
disease of the central nervous system, with an estimated worldwide prevalence in 2020
of more than 2.8 million cases.1 Although symptoms of MS usually first appear in adults
between 20 and 50 years of age,2 approximately 3%–5% of MS cases are of pediatric
onset, with first symptoms in childhood or, more commonly, in adolescence.3-5
Patients with pediatric-onset MS (POMS) typically experience higher levels of
inflammation with more frequent relapses than patients with adult-onset MS (AOMS).6-8
Although pediatric patients take longer than adults to reach irreversible disability, this
point still occurs at a younger age.9 A study of 394 patients with POMS found that
patients exhibited a median time of 28.9 years to reach an Expanded Disability Status
Scale (EDSS) score of 6, when patients were a median age of 42.2 years old,
approximately 10 years younger than the average age to EDSS 6 for AOMS.9
Additionally, patients with POMS generally have poorer cognitive performance and long-
term socioeconomic outcomes than do patients diagnosed with AOMS or healthy
controls, probably because brain inflammation can interfere with ongoing cerebral
maturation processes in adolescence.10,11 Early intervention with an appropriate
efficacious disease-modifying therapy (DMT) is therefore essential for reducing the risk
of persistent long-term disability in these patients.12
There have been few randomized placebo-controlled clinical trials of MS DMTs in
pediatric patients with POMS, and therefore many of the available data are based on
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
5
observational studies of off-label use of DMTs approved for adults.13 Patients with
pediatric- and adult-onset MS share similar genetic and environmental risk factors,
suggesting similar pathophysiology, and therefore pediatric patients typically show
similar responses to DMTs.14,15 The most widely used first-line therapies in POMS
patients have been injectable DMTs, including interferon-β and glatiramer acetate.16
However, these first-line agents may be poorly tolerated or fail to provide adequate
control of disease activity,17 prompting the need to escalate therapy to more efficacious
DMTs.18
Fingolimod was approved in the United States and European Union for pediatric
patients with MS in 2018, on the basis of the randomized clinical trial of fingolimod
versus interferon beta-1a in pediatric patients with MS.19-21 Observational studies of
pediatric patients with MS have reported effectiveness of other DMTs as well, including
dimethyl fumarate and natalizumab, though the latter has predominantly been studied
as a second-line therapy in patients with breakthrough disease.13,22-30
Comparative effectiveness data in POMS patients are also limited. In the absence of
comparative data from head-to-head randomized clinical trials, observational studies
can provide useful information for treatment decision-making. We therefore assessed
the effectiveness of natalizumab compared with fingolimod and with reference to the
injectable platform therapies (subcutaneous [SC] interferon beta-1b, intramuscular [IM]
or SC interferon beta-1a, IM or SC peginterferon beta-1a, or SC glatiramer acetate) in
patients with POMS in the global MSBase registry.31
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
6
METHODS
Data
The MSBase registry consists of anonymized patient-level data from contributing
member sites.32 Institutional review board and ethics committee approvals are required
for initiation of each site according to applicable local laws and regulations. Written
informed consent is obtained for each patient prior to their inclusion in the database in
accordance with the Declaration of Helsinki. Data were collected from 2006, when
natalizumab first became available, through 2021. A sensitivity analysis was performed
using data as of 2010, when fingolimod became available for use.
Study design and sample
This was a retrospective cohort study based on MSBase registry data from 100 centers
and 32 countries. Patients diagnosed with POMS and who initiated treatment with an
injectable DMT, natalizumab, or fingolimod between January 1, 2006, and May 3, 2021,
were included so as to have a contemporary sample since the availability of the second-
generation of MS treatments.
Inclusion and exclusion criteria
Eligible patients had a diagnosis of pediatric-onset relapsing MS (RMS) and initiated
index therapy treatment prior to the age of 18. Patients had to be naïve to prior therapy
or to have switched from an injectable DMT. Patients who initiated treatment with an
injectable DMT and who switched to either natalizumab or fingolimod during the follow-
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
7
up period were assigned to the natalizumab or fingolimod cohort, respectively.
Exclusion criteria included prior treatment with natalizumab, fingolimod, cladribine,
rituximab, ocrelizumab, alemtuzumab, dimethyl fumarate, or teriflunomide, or prior
treatment with a recognized immunosuppressive agent, such as azathioprine,
methotrexate, cyclophosphamide, or mitoxantrone.
Patient cohorts
Patients were assigned to 1 of 3 cohorts based on their index therapy: injectable DMT,
natalizumab, or fingolimod. Patients contributed to a cohort from baseline (therapy
initiation) until the patient discontinued the therapy or until the end of the follow-up.
Outcomes and Assessments
Primary outcome
The primary outcome was the time to first relapse from index therapy initiation.
Secondary outcomes
Secondary study outcomes included annualized relapse rate (ARR); proportions of
patients who were relapse-free at 1, 2, and 5 years post-baseline; time to treatment
discontinuation; and times to 24-week confirmed disability worsening (CDW) and 24-
week confirmed disability improvement (CDI) in EDSS scores. ARR was calculated
during the entire duration of follow-up. Relapses were defined as new or recurrent
neurologic symptoms occurring >30 days following the onset of a previous relapse.
Symptoms occurring ≤30 days following a previous relapse were counted only as 1
relapse, and the onset date used in the analysis was the onset date of the first relapse.
Time to treatment discontinuation was calculated during the entire duration of follow-up
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
8
time. Time to 24-week CDW or 24-week CDI was calculated from baseline until disease
progression or improvement, respectively. CDW was defined as a confirmed increase of
≥0.5 point in EDSS score for patients with a baseline EDSS score >5.5; ≥1.0 point for
those with a baseline EDSS score between 1.0 and 5.5, inclusive; and ≥1.5 points for
those with a baseline EDSS score of 0. For the confirmation of disability worsening,
EDSS scores recorded within 30 days after the onset of a relapse were excluded. Initial
and confirmatory disability progression had to be assessed on consecutive visits. CDI
was defined as a confirmed decrease in EDSS score of ≥1 point for patients with a
baseline EDSS score ≥2.0. CDI was not calculated for patients with a baseline EDSS
score <2.0. For the confirmation of disability improvement, EDSS scores recorded within
30 days after the onset of a relapse were excluded. Initial and confirmatory disability
improvement had to be assessed on consecutive visits.
Statistical analyses
Analysis populations
All patients fulfilling all the inclusion criteria and not meeting any of the exclusion criteria
were included in the analysis population. A multinomial logistic regression model was
used to calculate propensity scores and inverse probability of treatment weighting (PS-
IPTW) was used to balance groups by baseline patient characteristics (age, sex,
disease duration, baseline EDSS, country, prior DMT, relapse count in the past year
and past 2 years, magnetic resonance imaging [MRI] lesion count, and presence of
gadolinium-enhanced [Gd+] lesions). Index calendar year was not included as a
covariate to avoid positivity violation issues during modeling. This enabled a broader
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
9
analysis that included patients enrolled prior to the availability of some treatments in
their country.
Outcome assessments
For the primary and secondary outcome assessments, average treatment effect
weights—which may be interpreted as targeting a combined population of patients
treated with natalizumab, fingolimod, or injectable platform therapies—for each of the 3
treatment cohorts were estimated. Outcomes were assessed separately for pairwise
treatment comparison. For the primary outcome, the adjusted cumulative probability of
relapse post-baseline was estimated using a weighted Cox proportional hazard model
controlling for treatment arm, with adjustment for previous DMT use (naïve vs
experienced), count of prebaseline DMTs, and index year. The secondary ARR
endpoint was estimated using a weighted negative binomial regression model
controlling for treatment arm and with an offset for log-transformed follow-up time.
Proportions of relapse-free patients were evaluated using a weighted logistic regression
model controlling for treatment arm. Secondary time-to-event analyses were performed
using a weighted Cox proportional hazard model controlling for treatment arm. Errors
(95% confidence intervals [CIs]) for all primary and secondary outcome analyses were
calculated using robust sandwich estimation.
Sensitivity analyses
Five sensitivity analyses were conducted to assess the validity of assumptions and
robustness of the results (Supplementary Methods). For the propensity score–matching
analysis, patients were matched 1:1 with replacement using propensity scores instead
of IPTW. A weight-trimming analysis was performed to assess the impact of removing
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
10
outlier propensity score weights by removing the first quartile of scores from the
natalizumab-, fingolimod-, and injectable DMT–treated groups. For the alternative
weighting analyses, the outcome analyses were repeated using average treatment
effect among the treated (ATT) weighting to produce different target populations for
each treatment group or the combined (overlap) patient population. A cohort entry date
analysis was performed by restricting data to patients who were treated after January 1,
2010, when fingolimod was available for use. For the MRI lesion analysis, data were
restricted to patients with a baseline MRI and included number of MRI lesions and
presence of Gd+ lesions as covariates in the propensity score model.
A sensitivity analysis was also conducted to assess the effect of index calendar year on
outcomes.
Descriptive statistics
Continuous variables were assessed using mean, standard deviation, median, 25th and
75th percentiles, or minimum and maximum, as appropriate. Categorical variables were
summarized as frequencies and percentage. Descriptive statistics were tabulated for all
baseline characteristics by cohort. The number of visits with a EDSS measurement
after baseline were summarized separately by cohort as both a categorical and
continuous measure. The inter-visit time was calculated as the time in months between
consecutive visits with a EDSS measurement. The reasons for treatment
discontinuation was reported in a descriptive manner. No adjustments for multiplicity
were conducted, because the number of tests performed were determined to be small
enough as to not appreciably affect the findings.
Missing data
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
11
In general, missing values were not imputed. Patients with missing data in any variable
required for a given analysis were not included in that analysis. Tests were performed to
determine whether these exclusions caused a bias. Partial dates were imputed for dates
of relapse events, EDSS measurements, and therapy initiation or discontinuation.
Unknown days and months were imputed as the 1st and January, respectively.
Data Availability
The clinical data for this study were obtained under a license agreement with MSBase
(http://www.msbase.org). However, no patient-level data were disclosed as part of the
study. Therefore, all data relevant to the study are presented in this manuscript and the
Supplementary Materials.
RESULTS
Patients
As of May 3, 2021, there were 76,152 patients included in the MSBase registry, of
whom 5410 were diagnosed with POMS. Of these, 1218 had RMS; initiated treatment
with an injectable DMT, fingolimod, or natalizumab on or after January 1, 2006; and had
a baseline EDSS assessment within 6 months of the index date (Table 1). Although
baseline age, sex, and EDSS score were generally similar between groups, prior to
IPTW adjustment, 50% of covariates displayed standardized differences >0.20 (Table
1). Following propensity score–based IPTW, mean standardized differences between
treatment groups were <0.20 for 18 of 24 covariates (75%).
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
12
Patients who were prescribed injectable DMTs as the index therapy had shorter MS
disease duration and fewer relapses in the past year than did those who received
natalizumab or fingolimod.
Outcomes
Time to first relapse
Kaplan-Meier estimated proportions of relapse-free patients at years 1, 2, and 5 were
greatest among those treated with natalizumab and smallest among patients treated
with an injectable DMT (Figure 1). At year 1, the proportion of relapse-free patients
(95% CI) was highest for natalizumab (94.8% [86.6–98.0]), followed by fingolimod
(88.2% [76.9–94.2]) and injectable DMT (73.3% [69.8–76.6]). At year 2, the proportion
of relapse-free patients (95% CI) was highest for natalizumab (93.4% [84.8–97.1]),
followed by fingolimod (82.9% [70.4–90.4]) and injectable DMT (59.8% [55.6–63.7]). At
year 5, the proportion of relapse-free patients (95% CI) was highest for natalizumab
(90.0% [80.1–95.1]), followed by fingolimod (71.9% [55.5–83.1]) and injectable DMT
(35.8% [30.6–40.9]).
Patients treated with natalizumab demonstrated significantly lower risk of relapse than
did patients treated with an injectable DMT (adjusted HR, 0.15; 95% CI, 0.07–0.31;
P<0.001) and a lower risk of relapse than patients treated with fingolimod (adjusted HR,
0.37; 95% CI, 0.14–1.00; P=0.049) (Table 2). Patients treated with fingolimod also had
a significantly lower risk of relapse than those treated with injectable DMTs (adjusted
HR, 0.49; 95% CI, 0.29–0.83; P=0.008).
Secondary outcomes
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
13
The ARRs (95% CI) in patients treated with natalizumab (0.08 [0.05–0.11]) and in
patients treated with fingolimod (0.12 [0.08–0.16]) were both significantly lower than the
ARR in patients treated with an injectable DMT (0.35 [0.33–0.38]; P<0.0001 for both
comparisons; Table 3). ARR was not significantly different for natalizumab-treated
versus fingolimod-treated patients (RR [95% CI]: natalizumab vs fingolimod, 0.68 [0.41–
1.12] P=0.07).
The time to index treatment discontinuation was significantly longer in patients treated
with natalizumab or fingolimod versus injectable DMT (Figure 2; Table 4), as
demonstrated by a lower risk of discontinuing index treatment during follow-up (adjusted
HR [95% CI]: natalizumab vs injectable DMT, 0.24 [0.16–0.36]; fingolimod vs injectable
DMT, 0.24 [0.15–0.39]; P<0.001 for both comparisons). However, the time to
discontinuation of index therapy was similar for natalizumab- and fingolimod-treated
patients (adjusted HR, 1.00; 95%CI, 0.56–1.78; P=0.998).
The time to 24-week CDW was not significantly different in any treatment group;
however, natalizumab and fingolimod demonstrated a nominally reduced risk of 24-
week CDW in comparison with injectable DMTs (Figure 3; Table 4). Adjusted HRs (95%
CI) were 2.17 (0.81–5.85; P=0.124) for natalizumab versus injectable DMTs and 2.27
(0.73–7.06; P=0.156) for fingolimod versus injectable DMTs.
The cumulative proportion of patients demonstrating 24-week CDI was significantly
greater in patients treated with either natalizumab or fingolimod than in patients treated
with injectable DMTs (Figure 4; Table 4). Adjusted HRs (95% CI) for time to 24-week
CDI were 2.24 (1.33–3.76; p=0.002) for natalizumab versus injectable DMTs and 2.98
(1.70–5.23; p<0.001) for fingolimod versus injectable DMTs.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
14
Reasons for index treatment discontinuation
Treatment discontinuations occurred among 709, 76, and 72 patients treated with
injectable DMTs, fingolimod, and natalizumab, respectively. Within each treatment
group, the most common reported reason for treatment discontinuation was lack of
improvement (fingolimod, 29.0%; injectable DMT, 22.6%; natalizumab, 22.2%). The
reason for discontinuation was not reported for approximately a third of patients
(injectable DMT, 33.9%; natalizumab, 33.3%; fingolimod, 27.6%).
Description of monitoring times
Postbaseline visits with an EDSS measurement (mean [standard deviation [SD]) were
more common among patients treated with natalizumab (5.9 [7.6]) or fingolimod (4.3
[5.8]) than with an injectable DMT (2.0 [3.9]). However, the intervisit time (mean months
between visits with an EDSS evaluation [SD]) was similar between treatment groups
(natalizumab, 5.7 [4.9]; fingolimod, 5.0 [3.9]; injectable DMT, 5.3 [4.9]), indicating that
visits were conducted at a consistent rate regardless of treatment in patients for whom
EDSS was evaluated.
Sensitivity Analyses
Analyses of 1:1 propensity score–matched patients (n=54 patients in each treatment
group) produced similar results to the main IPTW analyses (Supplemental Tables 1–3).
The time to first relapse was nominally reduced in matched patients treated with
natalizumab or fingolimod in comparison with injectable DMT, as evidenced by a lower
risk of relapse during follow-up (HR [95% CI] natalizumab vs injectable DMTs: 0.44
[0.19–1.02], P=0.055; fingolimod vs injectable DMTs: 0.48 [0.21–1.11], P=0.085). ARR
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
15
(95% CI) was significantly reduced in propensity score–matched patients treated with
natalizumab (0.07 [0.04–0.11]) or fingolimod (0.08 [0.05–0.14]) in comparison with
injectable DMT (0.42 [0.27–0.62]) (P<0.001 for both comparisons). The time to index
treatment discontinuation was significantly longer in matched patients treated with
natalizumab or fingolimod than with injectable DMT, as demonstrated by a reduced risk
of relapse during follow-up (HR [95% CI]: natalizumab vs injectable DMT, 0.18 [0.10–
0.35]; fingolimod vs injectable DMT, 0.28 [0.15–0.52]; P<0.001 for both comparisons).
As in the primary analysis, time to 24-week CDW in EDSS was not significantly different
in any treatment group. Time to 24-week CDI was also similar to that in the primary
analysis, with a significantly increased risk of improvement observed with fingolimod
versus injectable DMT (HR, 8.22; 95% CI, 1.07–62.92; P=0.043) and nominally
increased risk of improvement in patients treated with natalizumab versus injectable
DMT (HR, 5.63; 95% CI, 0.72–43.92; P=0.099). In pairwise propensity score–matched
analyses, no 2 treatments demonstrated significant differences in 24-week CDI.
The results from other sensitivity analyses were also consistent with the primary
analysis. A trimmed analysis—which excluded 2 of 111 (1.8%) natalizumab-treated
patients, 6 of 104 (5.8%) fingolimod-treated patients, and 297 of 1003 (29.6%)
injectable DMT–treated propensity score outlier patients—also generated results
consistent with the main analyses (Supplemental Tables 4–6). Results of the ATT-
weighted analyses were consistent with the primary analysis regardless of whether
patients were weighted to be similar in baseline characteristics to those treated with
natalizumab, fingolimod, or injectable DMTs (Supplemental Table 7). Results from the
2010 data cutoff sensitivity analysis were consistent with the 2006 data cutoff, despite
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
16
the reduction in number of patients treated with natalizumab (n=91) or injectable DMTs
(n=724) in this time period (Supplemental Tables 8–10). Finally, results of a sensitivity
analysis in 83 of 111 (74.8%) natalizumab-treated patients, 62 of 104 (59.6%)
fingolimod-treated patients, and 536 of 1003 (53.4%) injectable DMT–treated patients
with baseline MRI data were also consistent with the main primary and secondary
analyses (Supplemental Tables 11–13).
DISCUSSION
Pediatric-onset MS is a rare disease, and patients under the age of 18 are excluded
from most randomized MS trials. Due to the relative rarity of the disease and frequent
off-label use, randomized trials require many centers and a long recruitment period, and
therefore are very expensive and difficult to conduct and complete.33 In the absence of
randomized clinical trial data, real-world evidence is increasingly used to investigate
important clinical questions including MS disease prognosis, predictors of treatment
response and long-term outcomes, therapeutic effectiveness, and comparative
effectiveness and safety of different DMTs. 34
We conducted a 3-way IPTW analysis of real-world clinical data from the MSBase
registry to compare the effectiveness of natalizumab, fingolimod and injectable
therapies (interferon-beta, glatiramer acetate) in patients with POMS. For the primary
study endpoint, patients with POMS who initiated treatment with natalizumab, or who
switched to natalizumab from an injectable therapy, showed a greater probability of
remaining relapse-free than did those who initiated treatment with fingolimod or who
switched to fingolimod from an injectable therapy. Patients treated with either
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
17
natalizumab or fingolimod had a significantly greater probability of remaining relapse-
free than those treated with injectable therapies.
The adjusted secondary study outcomes were generally consistent with the primary
outcome or with previous observations. Proportions of patients remaining on index
therapy were significantly greater for patients treated with either natalizumab or
fingolimod than with an injectable DMT. The time to index therapy discontinuation was
essentially the same for natalizumab and fingolimod, and for all 3 groups the most
common reason for discontinuation was a lack of improvement.35
Time to 24-week CDW was not significantly different among the 3 treatment groups.
The lack of a differential treatment effect for natalizumab is consistent with previous
studies of natalizumab in adult patients with relapsing-recurring MS in MSBase, and is
in part due to the relative rarity of CDP events in treated real-world observational
studies .36
Time to 24-week CDI was significantly lower in MSBase patients with pediatric MS
treated with natalizumab or with fingolimod than in patients treated with an injectable
DMT, consistent with previous clinical and real-world observations of patients with
AOMS.37,38 These observations are also consistent with evidence that natalizumab and
fingolimod are effective anti-inflammatory agents, as assessed by reductions in relapse
rates, in patients with POMS.13
The findings in the IPTW-adjusted patient populations were confirmed in multiple
sensitivity analyses. Importantly, the analyses of propensity score–matched patients
were generally consistent with the main analyses, as were the analyses of the trimmed
IPTW-adjusted treatment groups.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
18
This was a real-world, retrospective cohort study of patients in the MSBase registry and,
as such is subject to the limitations typical of real-world analyses generally and of
registry-based studies specifically.34 Aside from the nonrandomized design inherent to
these studies, some residual indication bias can remain, especially in relation to the
results of MRI and their association with treatment choice. Head-to-head randomized
clinical trials are generally considered to provide the highest-quality evidence for
comparative studies of therapeutic agents. For rarer diseases such as POMS,
randomized clinical trial data can be limited or unavailable. Real-world studies offer a
powerful alternative to randomized clinical trials for these patient populations, and large
multicenter networks and patient registries such as MSBase can provide access to
more patients than could be enrolled in a randomized clinical trial. Therefore real-world
studies can answer practical questions related to treatment and outcomes for rarer
patient populations, and real-world data are being increasingly accepted and used by
regulatory agencies.39,40
When comparing patient cohorts from real-world data sources, patient groups must be
balanced to address potential differences in baseline demographic or disease
characteristics. Standard propensity score matching is not straightforward with 3
comparator groups, because the target population after matching is not easily defined.
Additionally, the algorithms used for matching generally need a larger sample size per
treatment arm than was available to us. When weighting with IPTW, variations in
baseline characteristics that predict an individual treatment are weighted so as to
calculate propensity scores independent of treatment assignment.41 Thus, each
treatment group may be weighted to mirror baseline characteristics of the overall treated
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
19
population. Furthermore, IPTW enables sensitivity analyses to compare weighting to
baseline characteristics of the full analysis sample and weighting to baseline
characteristic patterns present in each individual treatment group.
Even though the 3 treatment groups appeared to be well balanced following IPTW,
there is potential for residual bias due to unmeasured covariates. However, the
consistent results obtained with the sensitivity analyses argue that such bias, if present,
cannot account for the observed results. The similar results of the main and sensitivity
analyses are evidence of high internal validity of this study. It is not unreasonable to
assume that the pediatric patients in this study who were treated with natalizumab or
fingolimod have more severe disease than those treated with injectable DMTs and
might therefore be expected to have worse outcomes than those in the injectable group.
This was not observed, however, suggesting that indication bias is significantly
mitigated.
Overall, the results of this retrospective registry study are consistent with previous
studies of natalizumab 13,22-30 and fingolimod in patients with POMS, and therefore
entirely expected.21,42-46 The results are also generally consistent with comparative
studies of natalizumab and fingolimod in adult patients with MS, in concordance with the
general agreement that pediatric- and adult-onset MS have similar underlying
pathophysiology14 and that outcomes for patients <18 years are not fundamentally
different than those for patients >18 years, although data supporting this point are
limited.
Our results suggest that natalizumab and fingolimod have broadly equivalent
therapeutic effectiveness in this population of MSBase patients with POMS. However,
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
20
the consistent results of the analyses of relapse outcomes in these 2 treatment groups
suggests natalizumab could be superior to fingolimod on relapse outcomes. This is
supported by relapse outcomes from comparative studies of natalizumab and fingolimod
in patients with pediatric- or adult-onset MS. 38,47-49 It is highly likely that lower relapse
rates would also be associated with lower brain and spinal cord lesion accumulation,
with potential for differential long-term benefits. In particular, cognitive and productivity
outcomes, and risk of secondary progressive MS are potentially reduced with maximal
control of the early inflammatory phase of RRMS. MSBase is planning to enhance its
data collection for cognition and productivity outcomes in the future. Larger cohorts with
longer follow-up will be required to ultimately assess differential effects on protection
from secondary progressive MS, as has been demonstrated for early use of high-
efficacy therapies in AOMS.
Our findings also demonstrate the usefulness of large MS registries and networks in
general, and of the MSBase registry in particular, for comparative effectiveness studies
of MS DMTs in rare patient populations that are difficult to study in a randomized
setting. These results may also be helpful to healthcare providers and their patients in
optimizing relapse control and potentially reducing the risk of persistent long-term
disability in POMS.
Acknowledgments
Luke Ward, PhD, of Ashfield MedComms, an Ashfield Health Company (Middletown, CT, USA),
assisted with drafting the manuscript, and Celia Nelson of Ashfield MedComms edited and
styled the manuscript per journal requirements. A list of contributing members of the
MSBase Study Group is given in supplementary table 14.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
21
Funding
This work was supported by Biogen who provided funding for these analyses, which were
conducted by MSBase. Biogen also funded medical writing support in the development of this
manuscript. Biogen reviewed and provided feedback on the manuscript to the authors. The
authors had full editorial control and provided final approval of all content.
REFERENCES
1. Multiple Sclerosis International Federation. Atlas of MS 3rd edition. Part 2: Clinical
management of multiple sclerosis around the world. 2021. https://www.msif.org/wp-
content/uploads/2021/05/Atlas-3rd-Edition-clinical-management-report-EN-5-5-21.pdf.
2. McGinley MP, Goldschmidt CH, Rae-Grant AD. Diagnosis and treatment of multiple
sclerosis: a review. JAMA. 2021;325(8):765-779.
3. Boiko A, Vorobeychik G, Paty D, Devonshire V, Sadovnick D. Early onset multiple
sclerosis: a longitudinal study. Neurology. 2002;59(7):1006-1010.
4. Chitnis T, Glanz B, Jaffin S, Healy B. Demographics of pediatric-onset multiple sclerosis
in an MS center population from the Northeastern United States. Mult Scler.
2009;15(5):627-631.
5. Harding KE, Liang K, Cossburn MD, et al. Long-term outcome of paediatric-onset
multiple sclerosis: a population-based study. J Neurol Neurosurg Psychiatry.
2013;84(2):141-147.
6. Gorman MP, Healy BC, Polgar-Turcsanyi M, Chitnis T. Increased relapse rate in
pediatric-onset compared with adult-onset multiple sclerosis. Arch Neurol.
2009;66(1):54-59.
7. Alroughani R, Boyko A. Pediatric multiple sclerosis: a review. BMC Neurol.
2018;18(1):27.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
22
8. Benson LA, Healy BC, Gorman MP, et al. Elevated relapse rates in pediatric compared
to adult MS persist for at least 6 years. Mult Scler Relat Disord. 2014;3(2):186-193.
9. Renoux C, Vukusic S, Mikaeloff Y, et al. Natural history of multiple sclerosis with
childhood onset. N Engl J Med. 2007;356(25):2603-2613.
10. Wallach AI, Waltz M, Casper TC, et al. Cognitive processing speed in pediatric-onset
multiple sclerosis: Baseline characteristics of impairment and prediction of decline. Mult
Scler. 2019;26(14):1938-1947.
11. McKay KA, Friberg E, Razaz N, Alexanderson K, Hillert J. Long-term Socioeconomic
Outcomes Associated With Pediatric-Onset Multiple Sclerosis. JAMA Neurol.
2021;78(4):478-482.
12. Baroncini D, Simone M, Iaffaldano P, et al. Risk of Persistent Disability in Patients With
Pediatric-Onset Multiple Sclerosis. JAMA Neurol. 2021;78(6):726-735.
13. Jakimovski D, Awan S, Eckert SP, Farooq O, Weinstock-Guttman B. Multiple Sclerosis
in Children: Differential Diagnosis, Prognosis, and Disease-Modifying Treatment. CNS
Drugs. 2022;36(1):45-59.
14. Bar-Or A, Hintzen RQ, Dale RC, Rostasy K, Bruck W, Chitnis T.
Immunopathophysiology of pediatric CNS inflammatory demyelinating diseases.
Neurology. 2016;87(9 Suppl 2):S12-19.
15. Waubant E, Ponsonby AL, Pugliatti M, Hanwell H, Mowry EM, Hintzen RQ.
Environmental and genetic factors in pediatric inflammatory demyelinating diseases.
Neurology. 2016;87(9 Suppl 2):S20-27.
16. Margoni M, Rinaldi F, Perini P, Gallo P. Therapy of Pediatric-Onset Multiple Sclerosis:
State of the Art, Challenges, and Opportunities. Front Neurol. 2021;12:676095.
17. Baroncini D, Zaffaroni M, Moiola L, et al. Long-term follow-up of pediatric MS patients
starting treatment with injectable first-line agents: A multicentre, Italian, retrospective,
observational study. Mult Scler. 2019;25(3):399-407.
18. Krysko KM, Graves J, Rensel M, et al. Use of newer disease-modifying therapies in
pediatric multiple sclerosis in the US. Neurology. 2018;91(19):e1778-e1787.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
23
19. Gilenya Prescribing Information. Gilenya® (fingolimod) [prescribing information]
In. East Hanover, NJ: Novartis Pharmaceuticals Corp; 2019.
20. Gilenya Summary of Product Characteristics. Gilenya® (fingolimod) [summary of product
characteristics]. Nuremberg, Germany: Novartis Pharma GmbH; 2018. In:2018.
21. Chitnis T, Arnold DL, Banwell B, et al. Trial of Fingolimod versus Interferon Beta-1a in
Pediatric Multiple Sclerosis. N Engl J Med. 2018;379(11):1017-1027.
22. Pohl D, Rostasy K, Gartner J, Hanefeld F. Treatment of early onset multiple sclerosis
with subcutaneous interferon beta-1a. Neurology. 2005;64(5):888-890.
23. Banwell B, Reder AT, Krupp L, et al. Safety and tolerability of interferon beta-1b in
pediatric multiple sclerosis. Neurology. 2006;66(4):472-476.
24. Ghezzi A, Amato MP, Capobianco M, et al. Treatment of early-onset multiple sclerosis
with intramuscular interferon beta-1a: long-term results. Neurol Sci. 2007;28(3):127-132.
25. Kornek B, Aboul-Enein F, Rostasy K, et al. Natalizumab Therapy for Highly Active
Pediatric Multiple Sclerosis. JAMA Neurol. 2013;70(4):469-475.
26. Tenembaum SN, Banwell B, Pohl D, et al. Subcutaneous interferon Beta-1a in pediatric
multiple sclerosis: a retrospective study. J Child Neurol. 2013;28(7):849-856.
27. Ghezzi A, Moiola L, Pozzilli C, et al. Natalizumab in the pediatric MS population: results
of the Italian registry. BMC Neurol. 2015;15:174.
28. Krysko KM, Graves JS, Rensel M, et al. Real-World Effectiveness of Initial Disease-
Modifying Therapies in Pediatric Multiple Sclerosis. Ann Neurol. 2020;88(1):42-55.
29. Palavra F, Figueiroa S, Correia AS, et al. TyPed study: Natalizumab for the treatment of
pediatric-onset multiple sclerosis in Portugal. Mult Scler Relat Disord. 2021;51:102865.
30. Simpson A, Mowry EM, Newsome SD. Early Aggressive Treatment Approaches for
Multiple Sclerosis. Curr Treat Options Neurol. 2021;23(7):19.
31. Kalincik T, Butzkueven H. The MSBase registry: Informing clinical practice. Mult Scler.
2019;25(14):1828-1834.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
24
32. Butzkueven H, Chapman J, Cristiano E, et al. MSBase: an international, online registry
and platform for collaborative outcomes research in multiple sclerosis. Mult Scler.
2006;12(6):769-774.
33. Waubant E, Banwell B, Wassmer E, et al. Clinical trials of disease-modifying agents in
pediatric MS: Opportunities, challenges, and recommendations from the IPMSSG.
Neurology. 2019;92(22):e2538-e2549.
34. Cohen JA, Trojano M, Mowry EM, et al. Leveraging real-world data to investigate
multiple sclerosis disease behavior, prognosis, and treatment. Mult Scler. 2020;26(1):23-
37.
35. Schwartz CE, Grover SA, Powell VE, et al. Risk factors for non-adherence to disease-
modifying therapy in pediatric multiple sclerosis. Mult Scler. 2018;24(2):175-185.
36. Butzkueven H, Kappos L, Wiendl H, et al. Long-term safety and effectiveness of
natalizumab treatment in clinical practice: 10 years of real-world data from the Tysabri
Observational Program (TOP). J Neurol Neurosurg Psychiatry. 2020;91:660-668.
37. Kalincik T, Horakova D, Spelman T, et al. Switch to natalizumab versus fingolimod in
active relapsing-remitting multiple sclerosis. Ann Neurol. 2015;77(3):425-435.
38. He A, Spelman T, Jokubaitis V, et al. Comparison of switch to fingolimod or interferon
beta/glatiramer acetate in active multiple sclerosis. JAMA Neurol. 2015;72(4):405-413.
39. US Food and Drug Administration Center for Drug Evaluation and Research.
Considerations for the Use of Real-World Data and Real-World Evidence to Support
Regulatory Decision-Making for Drug and Biological Products. Guidance for Industry. In.
Silver Spring, MD2021.
40. European Medicines Agency. A vision for use of real-world evidence in EU medicines
regulation. 2021; https://www.ema.europa.eu/en/news/vision-use-real-world-evidence-
eu-medicines-regulation. Accessed March 6, 2022.
41. Allan V, Ramagopalan SV, Mardekian J, et al. Propensity score matching and inverse
probability of treatment weighting to address confounding by indication in comparative
effectiveness research of oral anticoagulants. J Comp Eff Res. 2020;9(9):603-614.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
25
42. Arnold DL, Banwell B, Bar-Or A, et al. Effect of fingolimod on MRI outcomes in patients
with paediatric-onset multiple sclerosis: results from the phase 3 PARADIGMS study. J
Neurol Neurosurg Psychiatry. 2020;91(5):483-492.
43. Deiva K, Huppke P, Banwell B, et al. Consistent control of disease activity with
fingolimod versus IFN β-1a in paediatric-onset multiple sclerosis: further insights from
PARADIGMS. J Neurol Neurosurg Psychiatry. 2020;91(1):58-66.
44. Feng J, Rensel M. Review of the Safety, Efficacy and Tolerability of Fingolimod in the
Treatment of Pediatric Patients With Relapsing-Remitting Forms of Multiple Sclerosis
(RRMS). Pediatric Health Med Ther. 2019;10:141-146.
45. Zaffaroni M. Fingolimod in pediatric-onset multiple sclerosis. Neurol Sci. 2021;42(Suppl
1):1-4.
46. Ziemssen T, Albrecht H, Haas J, et al. Descriptive Analysis of Real-World Data on
Fingolimod Long-Term Treatment of Young Adult RRMS Patients. Front Neurol.
2021;12:637107.
47. Cohen M, Mondot L, Bucciarelli F, et al. BEST-MS: A prospective head-to-head
comparative study of natalizumab and fingolimod in active relapsing MS. Mult Scler.
2021;27(10):1556-1563.
48. Butzkueven H, Licata S, Jeffery D, et al. Natalizumab versus fingolimod for patients with
active relapsing-remitting multiple sclerosis: results from REVEAL, a prospective,
randomised head-to-head study. BMJ Open. 2020;10(10):e038861.
49. Guerra T, Caputo F, Orlando B, Paolicelli D, Trojano M, Iaffaldano P. Long-term
comparative analysis of no evidence of disease activity (NEDA-3) status between
multiple sclerosis patients treated with natalizumab and fingolimod for up to 4 years.
Neurol Sci. 2021;42(11):4647-4655.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
26
Table 1. Baseline characteristics for patients with pediatric-onset RMS
Characteristic
Natalizumaba
(n=111)
Fingolimodb
(n=104)
Injectable DMTc
(n=1003)
Standardized difference, unweighted Standardized difference, weighted
Natalizumab
vs
fingolimod
Natalizumab
vs
injectables
Fingolimod
vs
injectables
Natalizumab
vs
fingolimod
Natalizumab
vs
injectables
Fingolimod
vs
injectables
Age, mean
(SD), years 15.80 (2.28) 16.00 (2.89) 16.06 (1.98) 0.000 0.000 −0.027 0.000 0.008 −0.008
Female, n (%) 83 (74.8) 75 (72.1) 721 (71.9) 0.060 0.065 0.005 0.047 0.063 0.049
Country, n (%)
Australia 14 (12.6) 21 (20.2) 63 (6.3)
0.214 −0.166 −0.279 0.194 −0.071 −0.250
Italy 10 (9.0) 2 (1.9) 68 (6.8)
Kuwait 32 (28.8) 11 (10.6) 74 (7.4)
Spain 9 (8.1) 14 (13.5) 51 (5.1)
Turkey 5 (4.5) 25 (24.0) 282 (28.1)
Czech
Republic 5 (4.5) 2 (1.9) 39 (3.9)
Iran 0 (0.0) 0 (0.0) 46 (4.6)
Belgium 3 (2.7) 3 (2.9) 29 (2.9)
Canada 5 (4.5) 1 (1.0) 24 (2.4)
Other 28 (25.2) 25 (24.0) 327 (32.6)
MS duration,
mean (SD),
years
1.74 (1.71) 1.91 (1.75) 1.32 (1.65) −0.096 0.248 0.343 −0.016 0.031 0.148
BL EDSS
score,d median
(IQR)
1.5 (1, 2.5) 1 (0, 2) 1.5 (1, 2.5) 0.292 0.130 −0.161 0.155 0.063 −0.137
Prior DMT
use, n (%)
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
27
Naïve 64 (57.7) 48 (46.2) 865 (86.2)
0.231 −0.669 −0.932 0.135 −0.265 −0.420
Experienced 47 (42.3) 56 (53.9) 138 (13.8)
Number of
prior DMTs,
mean (SD)
0.60 (0.83) 0.69 (0.80) 0.15 (0.39) −0.108 0.699 0.864 −0.009 0.216 0.341
Relapses in
year prior to
BL, mean (SD)
1.53 (1.23) 1.12 (1.17) 1.07 (0.95) 0.346 0.420 0.044 0.083 0.037 −0.146
Index year, n
(%)
2006–2010 22 (19.8) 1 (1.0) 359 (35.8)
−0.397 0.421 0.881 0.076 0.514 0.519 2011–2015 49 (44.1) 59 (56.7) 433 (43.2)
2016+ 40 (36.0) 44 (42.3) 211 (21.0)
Follow-up
time, yearse
Mean (SD) 3.91 (2.94) 3.03 (2.79) 2.19 (2.45)
Median
(IQR) 3.21 (1.59, 5.55) 2.38 (0.59, 4.73) 1.37 (0.22, 3.41)
aPatients did not have prior natalizumab treatment. bPatients did not have prior fingolimod treatment. cIncludes IM or SC interferon
beta-1a, SC interferon beta-1b, SC glatiramer acetate, and IM or SC peginterferon beta-1a. dNearest EDSS score within 6 months of
BL. eFollow-up time while on index DMT.
BL, baseline; DMT, disease-modifying therapy; EDSS, Expanded Disability Status Scale; IM, intramuscular; IQR, interquartile range;
MS, multiple sclerosis; RMS, relapsing multiple sclerosis; SC, subcutaneous; SD, standard deviation.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
28
Table 2. PS-IPTW adjusted Kaplan-Meier estimates of relative risk of relapse
Comparison PS-IPTW adjusted HR (95% CI)
a
P value
Natalizumab vs fingolimod 0.37 (0.14–1.00) 0.049
Natalizumab vs injectable DMT 0.15 (0.07–0.31) <0.001
Fingolimod vs injectable DMT 0.49 (0.29–0.83) 0.008
aAdjusted for prior DMT (naïve vs experienced), count of prebaseline DMTs, and index year.
CI, confidence interval; DMT, disease-modifying therapy; HR, hazard ratio; PS-IPTW, propensity score inverse probability of
treatment weighting.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
29
Table 3. PS-IPTW adjusted annualized relapse rates
Index therapy N Total relapses Follow-up years
Adjusted ARR
(95% CI)a
P values
Natalizumab
vs fingolimod
Natalizumab vs
injectable DMT
Fingolimod vs
injectable DMT
Injectable DMT 1003 771 2199.82 0.351 (0.326–0.376)
0.067 <0.001 <0.001 Natalizumab 111 33 434.27 0.076 (0.052–0.107)
Fingolimod 104 37 314.91 0.118 (0.083–0.162)
aAdjusted for prior DMT (naïve vs experienced), count of prebaseline DMTs, and index year.
ARR, annualized relapse rate; CI, confidence interval; DMT, disease-modifying therapy; PS-IPTW, propensity score inverse
probability of treatment weighting.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
30
Table 4. PS-IPTW adjusted hazard ratio estimates of secondary outcomes
Outcome PS-IPTW adjusted HR (95% CI)a P value
Risk of persisting on index DMT
Natalizumab vs fingolimod 1.00 (0.56–1.78) 0.998
Natalizumab vs injectable DMT 0.24 (0.16–0.36) <0.001
Fingolimod vs injectable DMT 0.24 (0.15–0.39) <0.001
Risk of not demonstrating 24-week CDW
Natalizumab vs fingolimod 0.84 (0.26–2.75) 0.782
Natalizumab vs injectable DMT 2.17 (0.81–5.85) 0.124
Fingolimod vs injectable DMT 2.27 (0.73–7.06) 0.156
Risk of reaching 24-week CDI
Natalizumab vs fingolimod 0.97 (0.52, 1.83) 0.936
Natalizumab vs injectable DMT 2.24 (1.33, 3.76) 0.002
Fingolimod vs injectable DMT 2.98 (1.70, 5.23) <0.001
aAdjusted for prior DMT (naïve vs experienced), count of prebaseline DMT, and index year.
CDI, confirmed disability improvement; CI, confidence interval; CDW, confirmed disability worsening; DMT, disease-modifying
therapy; HR, hazard ratio; PS-IPTW, propensity score inverse probability of treatment weighting.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
31
Figure 1. Cumulative probabilities of remaining relapse-free for patients treated with natalizumab, fingolimod, or injectable therapies.
DMT, disease-modifying therapy.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
32
Figure 2. Cumulative probability of remaining on index DMT. DMT, disease-modifying therapy.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
33
Figure 3. Cumulative probability of remaining free of 24-week CDW for patients treated with natalizumab, fingolimod, or injectable
therapies. CDP, confirmed disability progression; CDW, confirmed disability worsening; DMT, disease-modifying therapy.
Proportion patients CDP free
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
34
Figure 4. Cumulative probability of reaching 24-week CDI for patients treated with natalizumab, fingolimod, or injectable therapies.
CDI, confirmed disability improvement; DMT, disease-modifying therapy.
. CC-BY-ND 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted October 13, 2022. ; https://doi.org/10.1101/2022.10.12.22280969doi: medRxiv preprint
