Effectiveness of Siponimod with Up to 2 Years of Treatment in a Real-World Setting in People with SPMS: Analysis from the Novartis Global Managed Access Program

Introduction

Multiple sclerosis (MS) presents with a relapsing-remitting course in 80% to 90% of cases; however, within 30 years of disease onset, 93% of these patients will have progressed to secondary progressive MS (SPMS).^1,2 SPMS is characterized by gradual worsening of disability with or without relapse occurrence.³ Many of the disease-modifying treatments that are efficacious for relapsing MS have proved ineffective in SPMS; thus, therapeutic options for SPMS are limited.³

The Phase 3 EXPAND trial demonstrated the efficacy of siponimod in slowing disability progression in people living with SPMS (plwSPMS) compared with placebo.⁴ The EXPAND study recruited a broad SPMS population, ie, patients with both active (relapses in the 2 years prior to study initiation or with magnetic resonance imaging [MRI] disease activity) and non-active SPMS;⁴ however, most regions (including the United States [US] and European Union [EU]) approved siponimod for the treatment of active SPMS only. Other treatment options that are approved in the EU and US for patients with active or highly active relapsing MS including relapsing SPMS include ofatumumab, cladribine, and ocrelizumab. Mitoxantrone is approved for SPMS without regard to relapse activity.^3,5 However, it is associated with risks of congestive heart failure and secondary acute myeloid leukemia.

The inclusion criteria for clinical trials are usually narrow with generally a homogeneous patient population. Real-world studies provide evidence on the effectiveness and safety of a medicine under routine clinical conditions and in a broader patient population in which the frequency of comorbidities and concomitant medication is highly variable. Previous small observational and real-world studies confirmed the effectiveness of siponimod in stabilizing disease progression in plwSPMS.^6,7 However, more studies are warranted to confirm the effectiveness trends in the highly variable plwSPMS cohort receiving siponimod in routine clinical practice. Designed to provide access to treatment, data from compassionate use/managed access programs (MAPs) can also provide further insights into the effectiveness of siponimod in a population potentially different from that seen in clinical trials. Thus, MAP data may aid clinical decision-making in a broader patient population.

Novartis implemented the Global Siponimod MAP to facilitate patient access to siponimod for “compassionate use” when needed, requested, and allowed. Siponimod was made available via MAP in accordance with EMA guidance, local laws, and regulations in countries where marketing authorization is pending, and satisfactory alternative therapies are absent. MAPs are not clinical studies and limited observational data about the patient and the disease reported by the physicians are collected. Similarly, limited prospective data on the demographic and clinical characteristics of patients who are prescribed siponimod in the routine clinical setting under this program are collected.

Here, we describe the available demographic characteristics of patients who received siponimod as part of the global Novartis siponimod MAP cohort (BAF2001M cohort) and analyze the clinical outcomes data. Data from this cohort provided further insights into the effectiveness of siponimod in routine clinical practice.

Methods

Manage Access Program Design and Patient Population

Novartis started this MAP in February 2019 in countries where it was permitted and siponimod was not yet available for use in patients with MS. The global siponimod MAP cohort data (demographic and clinical characteristics at baseline and post-baseline data) were collected from physicians as the sole source of data using the internal Novartis database Grants, External Trials, and Managed Access System (GEMS). Post-baseline assessments were mapped to hypothetical 3-months (± 45 days) or 6-months (± 90 days) visit schedule as the re-supply visits did not follow a predefined schedule. Up until January 2021, to have access to siponimod via the MAP, plwSPMS must have been adults with a diagnosis of SPMS (active and non-active) and an Expanded Disability Status Scale (EDSS) score <7. From January 2021, access to the MAP required a diagnosis of SPMS with active disease (ie, relapses or on MRI in the prior 24 months) and an EDSS score of <7 according to the approved EU/US label. Inclusion of patients in the MAP and treatment selection were based on physician judgment. Patients’ data were collected between February 2019 and April 2022. Analysis of MAP data set was performed on anonymized aggregated data. Thus, it is exempted from approval from ethical committee/institutional review board. Details of relevant legislation are mentioned in Supplementary Information.

Data Collection

At baseline, MS disease status (active/non-active/unknown), presence of relapses in the last 24 months (yes/no), availability of an MRI scan in the last 24 months (yes/no) and if yes, presence of new, enlarged or active lesions, assessment of cognitive status (yes/no), and EDSS score, were collected. Post-baseline data including duration of siponimod treatment under the compassionate use program, reasons for closure of a request (reasons for a patients to exit the program), EDSS score, number of relapses, availability of MRI scan (if yes presence of new, enlarged, or active lesions), and cognitive status (worsened/stable/improved) were collected.

Clinical Definitions

Active MS disease was defined at baseline as the presence of a relapse in the last 24 months and/or presence of (a) new or active lesion(s) on the MRI scan. Disability progression was defined as an EDSS increase of ≥1.0 points if the baseline EDSS was ≤5.0 or an increase of ≥0.5 points if the baseline EDSS was >5.0. Disability improvement was defined as an EDSS decrease of ≥1.0 points if the baseline EDSS was ≤5.0 or a decrease of ≥0.5 points if the baseline EDSS was >5.0. Six-month confirmed disability progression (6m-CDP) was defined as disability progression confirmed at 2 assessments at least 6-months apart. Annualized relapse rate (ARR) was defined as the number of relapses per year. A resupply request was defined as the request for a second or subsequent prescription for siponimod for patients who had filled at least 1 prescription for siponimod on entering the program. The treated cohort included patients with evidence of at least 1 prescription filled for siponimod on entering the program, for whom baseline characteristics were collected and the patients who have shown benefit to treatment with siponimod.

Statistical Analysis

Baseline demographic characteristics and clinical parameters were summarized descriptively by the numbers and percentages for categorical variables, and by means with standard deviations (SD) for continuous variables. Change from baseline in EDSS score and 95% confidence intervals (CI) were estimated at 6-monthly time points using a mixed model for repeated measures adjusted for baseline EDSS and with time as categorical factor. The cumulative percentage of patients with disability progression confirmed at 6 months was summarized over time using Kaplan-Meier curves.

Results

Patients and Treatment Duration

Data from 632 patients were collected. Documented evidence of the start of treatment with siponimod in the MAP was available for 516 of 632 patients (81.6%) and this cohort is referred to as the “treated cohort.” Of the 516 patients in the treated cohort, 423 (82%) had at least 1 resupply request post-baseline and had data availability post-baseline and were referred as the “effectiveness cohort.” In the “effectiveness cohort”, the primary reason for discontinuation was transition to commercial supply (n=166; 39.2%) and discontinuation of the treatment by the patients (n=51, 12.1%) (Figure 1).

Figure 1 Participant disposition. Notes:: aTreated cohort: Participants with evidence of at least 1 prescription filled for siponimod on entering the program and baseline characteristics collected. bEffectiveness cohort: Participants who subsequently had at least one resupply and contributed to effectiveness analysis, as they had one baseline and at least one post-baseline assessment.

Baseline Demographics

The mean (±SD) age of the patients in the treated cohort (N=516) was 52.5 (±8.7) years and 311 patients (60.4%) were female (Table 1). Approximately, half of the patients (239 [46.3%]) did not have a disease classification, whereas 198 patients (38.4%) were classified as having active disease and 79 patients (15.3%) were classified as non-active. The majority of patients with active SPMS were from Italy (156 of 198, 78.8%), and the country with the highest number of patients with non-active SPMS was Greece (46 of 79, 58.2%). The median EDSS score at baseline was 5.5 (interquartile range [IQR]: 4.5–6). For 298 of the patients, information regarding relapse in last 24 months was available and half of those patients (140 of 298 [46.9%]) had a relapse in the last 2 years. For 298 (58%) patients, information regarding MRI at the baseline was available and MRI was performed for 254 (85.2%) patients. Of whom, 112 patients (44.1%) showed new, enlarged, or active lesions on MRI.

Table 1 Baseline and Demographic Characteristics

Outcomes

EDSS Score

No difference in EDSS score from baseline was observed at any of the assessments time points, including month 6 (mean change from baseline: −0.04 [95% CI: −0.10, 0.01) and month 24 (mean change from baseline: 0.03 [95% CI: −0.03, 0.10]). At the end of 24 months, 86.6% (n=129 of the 149) of participants were stable and 7.4% (n=11 of the 149) of the participants showed and EDSS improvement (Figure 2). Patients with active disease had more variability in their EDSS score, whereas those with non-active disease had a relatively more stable EDSS. (Supplementary Figure 1) However, this assessment may be confounded with differential schedule of assessment in these two subgroups. In addition, the number of patients for this observation was small.

Figure 2 Change in EDSS score from baseline. Abbreviation: EDSS, Expanded Disability Status Scale. Notes: Disability progression: EDSS increase by ≥1.0 points if baseline EDSS≤5.0 or increase by ≥0.5 points. If baseline EDSS>5.0. Disability improvement: EDSS decrease by ≥1.0 points if baseline EDSS≤5.0 or decrease by ≥0.5 points if baseline EDSS>5.0.

Time to 6m-CDP in the Effectiveness Cohort

The rate of progression, as measured by time to 6m-CDP, was very low regardless of disease status at baseline (Figure 3) and regardless of prior relapse status (Figure 4).

Figure 3 Kaplan-Meier analysis of time to 6m-CDP* by MS disease status at baseline. Abbreviations: 6m-CDP, 6-month confirmed disability progression; EDSS, Expanded Disability Status Scale; MS, multiple sclerosis. Note: *6m-CDP as assessed in participants with at least 2 post-baseline EDSS assessments separated by >66 days.

Figure 4 6m-CDP in participants with relapses prior to baseline. Abbreviation: 6m-CDP, 6-month confirmed disability progression.

Baseline Characteristics of Patients with/Without 6m-CDP

6m-CDP occurred in 17 of the 423 patients in the “effectiveness cohort”. Baseline demographic and disease characteristics of patients with and without 6m-CDP are presented in Table 2. However, as the number of patients with 6m-CDP was small, we cannot draw any conclusion from the differences in baseline characteristics observed between these groups with and without 6m-CDP.

Table 2 Baseline Characteristics of Patients with 6m-CDP During the Study

Annualized Relapse Rate (ARR)

The ARR (number of relapses/total number of follow-up years) was 0.023 (15 relapses in 423 patients) for the overall cohort, 0.031 for the patients with active disease (5 relapses in 141 patients), and 0.009 for those with non-active disease (1 relapse in 67 patients).

Cognition Status

Cognition was stable in almost all patients for whom data were available (month 6, n=190; month 12, n=209; month 18, n=190; and month 24, n=141), with a small number (n=10) of patients showing improvement at any timepoints (Figure 5).

Figure 5 Cognition status at post-baseline time points.

MRI

The mean duration of follow-up between MRI assessments per patient was 165.8 days (±81.07). Of the 165 patients with ≥1 post-baseline MRI assessments, 154 (93.3%) had no findings of disease activity, 10 patients (6.1%) had new, enlarged or active MRI lesions on 1 MRI scan, and 1 patient (0.6%) had findings of new, enlarged or active MRI lesions on 2 MRI scans. The pattern of MRI assessments by country and by disease status is shown in Supplementary Figure 2. Since this was an analysis of the data set built in the context of the MAP, all the assessments, including MRI, were conducted according to local routine clinical practice, not at any specified time point. A great degree of diversity was observed for the time of MRI assessment among different countries and among different patients.

Discussion

This data set analysis evaluated the demographic and clinical characteristics, along with the effectiveness of siponimod, in a cohort of plwSPMS who received siponimod under a compassionate use/MAP for up to 2 years. The cohort comprised of 516 plwSPMS receiving siponimod under MAP program criteria. The majority of patients for whom the data were collected (or available) were patients with active SPMS. At month 24 of siponimod treatment, 87% of the patients had a stable EDSS score, 98.6% had a stable cognitive status, 93% of patients had no new, enlarged or active MRI lesion, and the mean ARR was very low (0.023). No major difference was noted between patients with active and non-active disease.

With the treatment options for SPMS being limited, siponimod represents a confirmed effective treatment already available for the treatment of adult plwSPMS.⁸ The EXPAND study demonstrated significant clinical benefits of siponimod versus placebo with a 21% of reduction in the risk of 3-month CDP (3m-CDP; hazard ratio [HR]: 0.79; 95% CI: 0.65–0.95; P = 0.013) and a 26% of reduced risk of 6m-CDP (HR: 0.74; 95% CI: 0.60–0.92; P = 0.0058).^3,4,9,10 These efficacy outcomes were sustained in the extension part of the study up to >5 years.¹⁰ A post hoc subgroup analysis of the EXPAND study population found that the benefits observed for the overall EXPAND population were more obvious in those with active SPMS.⁹ Interim findings from the EXCHANGE study, a prospective, interventional phase 3b study, which included 163 people living with advanced RMS, showed that treatment with siponimod for 6 months was associated with an overall stabilizing effect on clinical and cognitive outcome measures.⁷

Similar to other clinical trials, patients in the EXPAND and EXCHANGE trials were recruited according to strict inclusion/exclusion criteria. Real-world data can provide evidence of effectiveness across a broader range of patients.¹¹ Although, the indication for siponimod in some regions is restricted to patients with active SPMS, some real-world data indicate that we may also expect disease stability for up to 2 years in patients with non-active SPMS treated with siponimod.⁸ The AMASIA study is the first real-world study to provide evidence on the long-term effectiveness and safety of siponimod as well as the impact of siponimod treatment on quality of life and socioeconomic conditions.⁶ AMASIA is following 1200 people living with active SPMS who have been treated with siponimod for up to 3 years. In an interim analysis, EDSS remained stable during the first 12 months of siponimod treatment, regardless of age or time since diagnosis at study start and regardless of the last moderately effective therapy taken by the patients.⁶ Another real-world study including 227 plwSPMS treated with siponimod showed that 64.8% of patients experienced sustained clinical disease stability at month 12, and 21.4% of these patients showed evidence of improvement.¹²

The findings of the present analysis are in line with these other real-word studies and show most patients (86.6%) maintained a stable EDSS at the end of 24 months, regardless of active/non-active disease status. Pre-clinical studies suggest a mechanistic rationale for the disease stabilization associated with siponimod.^13–15 Siponimod is known to penetrate the central nervous system (CNS) and has the potential to exert direct beneficial effects on compartmentalized CNS inflammation, neurodegeneration, and remyelination.^16–20

Relapses in our data set analysis were reported in 15 of 423 patients (3.5%) during the follow-up period similar to the real-world study of Regner-Nelke et al in which 8/108 (74%) patients experienced relapses.¹² In the EXPAND study, which included people living with active and non-active SPMS, siponimod was associated with a 55% reduction in the mean ARR. The mean ARR in the MAP cohort was similar to the ARR in the EXPAND study cohort (0.023 vs 0.07, respectively) but the relatively low patient numbers and difference in this dataset and design preclude any meaningful comparisons.⁴

It is important to note that in the present analysis, the data on cognition status are reported by the physician and there is no information on the tools used by physicians to assess the patient’s cognition status. Considering limitations, this MAP cohort analysis could suggest that siponimod may be associated with a stabilization of cognition status in plwSPMS. Although no other real-world studies have assessed the effect of siponimod on cognition, secondary analyses from the EXPAND study population have shown that siponimod-treated patients are at a significantly lower risk of having a ≥4-point decrease in the Symbol Digit Modalities Test (SDMT) and have a significantly higher chance of having a ≥4-point sustained increase in the SDMT score versus placebo.²¹ The beneficial effect of siponimod on cognition was also observed in post hoc analyses patients with active disease in the EXPAND study.⁹ Siponimod increased the likelihood of 6-month confirmed SDMT improvement by 62% (P = 0.007) and reduced the risk of 6-month confirmed SDMT worsening by 27% (P = 0.060).⁹

In the present analyses, a high proportion of patients (93.3%) had no findings of disease activity on MRI, despite the majority of patients in the MAP being classified as having active disease at baseline. Benefits on MRI endpoints associated with siponimod were demonstrated in the EXPAND study, including lower T2 lesion volumes, fewer new or enlarging T2 lesions, and a higher proportion of patients free from gadolinium-enhancing lesions.⁴ In a further real-world retrospective study (N=227) of a population with a high prevalence of previous use of high-efficacy anti-inflammatory agents, and therefore representing a difficult-to-treat population, siponimod was found to suppress inflammatory MRI disease activity in all but 24.1% patients after 6 months of treatment and in all but 29.6% patients at 12 months follow-up.¹² Other radiological outcomes including volume of white (P = 0.35) and gray matter (P = 0.47) remained stable throughout the first 12 months of treatment. Lesion count (P = 0.74) also did not change significantly within the first 12 months of treatment.¹² It is important to note that safety data were not collected in this MAP, as safety information on these programs is collected via the routine safety surveillance systems.

There are a number of limitations when analyzing data from a compassionate use setting. There is considerable heterogeneity among countries in terms of clinical practice (eg, the frequency of resupply request and EDSS assessments) and in the profile of patients who entered the program (patients with non-active SPMS are over-represented in some countries and vice versa). Patient selection and the treatment were based on physician judgment, and no predefined schedule of monitoring visits was established by the physician and no case report forms were used to record the data. Physicians have requested resupply of siponimod only for patients who were considered to be deriving benefits from the treatment. Another limitation of the compassionate setting is that limited information is collected on the baseline disease characteristics including the previous use of DMTs, making it difficult to put results into perspective with results obtained in other real-word or randomized studies.

In terms of cognition, the program did not capture information regarding the methods/tools used to assess cognition status. Physicians were only asked about the cognition status at baseline and then to report any changes during the follow-up based on 3 categories, ie, stable, improved or worsened. The improvement/worsening status of the cognition is as reported by the physicians and is based on their judgement. Further, these analyses are limited due to the observational nature of the program and the lack of data on all patients.

Conclusion

In summary, MAP cohort analysis presented here provide data on the demographics and clinical characteristics of plwSPMS taking siponimod in routine clinical practice and according to MAP criteria and of the effectiveness that can be attained. The findings support sustained effectiveness of siponimod in a cohort of plwSPMS closer to routine clinical practice than a study population of a randomized clinical trial. This data can help physicians make informed choices in the face of the limited treatment options available for plwSPMS.