EUnetHTA Joint Action 3 WP4...V0.3 11/02/2020 Input from medical editor and manufacturer(s) has been...

Dec2015 ©EUnetHTA, 2015. Reproduction is authorised provided EUnetHTA is explicitly acknowledged 1

EUnetHTA Joint Action 3 WP4

Version 2.0, 03/03/2020

Relative effectiveness assessment of pharmaceutical technologies

SIPONIMOD FOR THE TREATMENT OF ADULT PATIENTS WITH

SECONDARY PROGRESSIVE MULTIPLE SCLEROSIS (SPMS) WITH ACTIVE

DISEASE EVIDENCED BY RELAPSES OR IMAGING FEATURES OF

INFLAMMATORY ACTIVITY.

Project ID: PTJA08

PTJA08 - Siponimod for SPMS with active disease

March 2020 EUnetHTA Joint Action 3 WP4 2

DOCUMENT HISTORY AND CONTRIBUTORS

Version Date Description

V0.1 06/01/2020 First draft

V0.2 27/01/2020 Input from dedicated reviewers has been processed

V0.3 11/02/2020 Input from medical editor and manufacturer(s) has been processed

V1.0 13/02/2020 Final assessment report

V2.0 03/03/2020 The following corrections were made:

- EXPAND subgroup numbers as reported in table 4.4 (p.45), p.74, table A14 (p.109)

- Half-life time of siponimod versus fingolimod is shorter (instead of longer) (p.10 and p.24)

In addition, the authors elaborated the MAIC assessment (p.19 and p.75).

Disclaimer

This Joint Assessment is part of the project / joint action ‘724130 / EUnetHTA JA3’ which has received funding from the European Union’s Health Programme (2014-2020). The content of this Project Plan rep-resents a consolidated view based on the consensus within the Authoring Team; it cannot be considered to reflect the views of the European Network for Health Technology Assessment (EUnetHTA), EUnetHTA’s participating institutions, the European Commission and/or the Consumers, Health, Agriculture and Food Executive Agency or any other body of the European Union. The European Commission and the Agency do not accept any responsibility for use that may be made of the information it contains.

Assessment team

Author(s) Autoridade Nacional do Medicamento e Produtos de Saúde, I.P. (INFARMED, I.P.), Portugal

Co-Author(s) National Centre for Pharmacoeconomics (NCPE), Ireland

Dedicated Reviewer(s)

Italian Medicines Agency (AIFA), Italy

Servicio de Evaluación del Servicio Canario de la Salud (SESCS) and Fundación Canaria Instituto de Investigación Sanitaria de Canarias (FIISC), Spain

Zorginstituut Nederland (ZIN), Netherlands

Regione Emilia-Romagna (RER), Italy



Further contributors

External experts

PD Dr. med. Clemens Warnke Answer specific question during the scoping phase

Manufacturer(s) [v0.2]

Novartis Preparation of the submission dossier

Factual accuracy check

Medical editor [v0.2]

Nextgenediting Medical editing of the 2nd draft assessment

Patient(s) / patient organisation(s) / citizens

MS Society of Slovakia

Russian MS Society

Smaragd Sclerosis Multiplexes Betegek Egyesülete (Hungary)

Združenie Sklerosis multiplex Nádej (Slovakia).

Provided input in response to the Open Call for Patient Input published on 28 March, 2019

Project Management

Zorginstituut Nederland (ZIN), Netherlands

Coordination between involved parties throughout the assessment

Conflict of interest

All authors and dedicated reviewers involved in the production of this assessment have declared that they have no conflicts of interest in relation to the technology assessed according to the EUnetHTA declaration of interest and confidentiality undertaking form.

How to cite this assessment

Please cite this assessment as follows:

EUnetHTA PTJA08. Authoring Team. Relative effectiveness assessment of pharmaceutical technologies. Siponimod for the treatment of adult patients with secondary progressive multiple sclerosis (SPMS) with active disease evidenced by relapses or imaging features of inflammatory activity. Joint Assessment. Die-men (The Netherlands): EUnetHTA; 2020. [date of citation]. 169 pages. Report No.: PTJA08. Available from: https //www.eunethta.eu



TABLE OF CONTENTS

DOCUMENT HISTORY AND CONTRIBUTORS .................................................................................... 2

TABLE OF CONTENTS......................................................................................................................... 4

LIST OF TABLES AND FIGURES ......................................................................................................... 5

LIST OF ABBREVIATIONS ................................................................................................................... 8

EXECUTIVE SUMMARY OF THE ASSESSMENT OF SIPONIMOD .................................................... 10 INTRODUCTION ................................................................................................................................ 10 OBJECTIVE AND SCOPE .................................................................................................................... 10 METHODS ....................................................................................................................................... 12 RESULTS ........................................................................................................................................ 13 DISCUSSION .................................................................................................................................... 19 CONCLUSION .................................................................................................................................. 20

1 BACKGROUND............................................................................................................................. 21 1.1 OVERVIEW OF THE DISEASE OR HEALTH CONDITION ..................................................................... 21 1.2 CURRENT CLINICAL PRACTICE ................................................................................................... 22 1.3 FEATURES OF THE INTERVENTION .............................................................................................. 24

2 OBJECTIVE AND SCOPE ............................................................................................................. 31

3 METHODS .................................................................................................................................... 33 3.1 INFORMATION RETRIEVAL .......................................................................................................... 33 3.2 DATA EXTRACTION ................................................................................................................... 35 3.3 RISK OF BIAS ASSESSMENT ....................................................................................................... 35 3.4 ANALYSES OF INCLUDED STUDIES .............................................................................................. 35

4 RESULTS ..................................................................................................................................... 39 4.1 INFORMATION RETRIEVAL .......................................................................................................... 39 4.2 STUDIES INCLUDED IN THE ASSESSMENT..................................................................................... 41 4.3 EXCLUDED STUDIES ................................................................................................................. 42 4.4 CHARACTERISTICS OF INCLUDED STUDIES .................................................................................. 43 4.5 OUTCOMES INCLUDED .............................................................................................................. 49 4.6 RISK OF BIAS/CERTAINTY OF THE EVIDENCE ................................................................................ 50 4.7 EXTERNAL VALIDITY ................................................................................................................. 62 4.8 RESULTS ON CLINICAL EFFECTIVENESS AND SAFETY .................................................................... 63

5 DISCUSSION ................................................................................................................................ 73

6 CONCLUSIONS ............................................................................................................................ 76

7 REFERENCES .............................................................................................................................. 77

APPENDIX 1: SLR search strategies ................................................................................................... 81

APPENDIX 2: Guidelines for diagnosis and management .................................................................. 106

APPENDIX 3: Evidence gaps ............................................................................................................ 107

APPENDIX 4: Studies included in the Ancillary analyses of the global SPMS population .................... 108

APPENDIX 5: EXPAND (Study A2304).............................................................................................. 114

APPENDIX 6: Ancillary analysis ........................................................................................................ 136



LIST OF TABLES AND FIGURES

Tables

Table 0.1. Scope of the assessment ..................................................................................................... 11 Table 0.2. Summary of relative outcomes in the relapsing SPMS population using Bucher ITCs ........... 16 Table 0.3. Adverse events in the global SPMS population ..................................................................... 18 Table 1.1. Features of the intervention and its comparators .................................................................. 25 Table 1.2. Administration and dosing of the technology and its comparators ......................................... 29 Table 2.1. Scope of the assessment ..................................................................................................... 32 Table 3.1. PICO for the systematic literature review .............................................................................. 34 Table 4.1. List of studies in the global SPMS population ....................................................................... 41 Table 4.2. Reported relapsing subgroups in SPMS ............................................................................... 42 Table 4.3. Studies that had no information on patients with relapsing disease ....................................... 43 Table 4.4. Design of the studies in the relapsing SPMS population ....................................................... 45 Table 4.5. Inclusion/exclusion criteria of studies in the relapsing SPMS population ............................... 46 Table 4.6. Outcome definitions of studies in the relapsing SPMS population ......................................... 47 Table 4.7. Patient baseline characteristics in the active population for studies included in the analyses

of the relapsing SPMS population .................................................................................................... 48 Table 4.8. Patient baseline characteristics in the global SPMS population for studies included in the

analyses of the relapsing SPMS population* .................................................................................... 49 Table 4.9. Risk of bias of included studies ............................................................................................ 51 Table 4.10. Evidence profile [siponimod vs interferon-β-1a 22 µg once a week, SC (Nordic SPMS

Study)] ............................................................................................................................................ 54 Table 4.11. Evidence profile [siponimod vs interferon-β-1b 250 µg every other day, SC (European

Study)] ............................................................................................................................................ 57 Table 4.12. Evidence profile [siponimod vs interferon-β-1a 22 µg or 44 µg three times weekly SC

(SPECTRIMS Study)] ...................................................................................................................... 61 Table 4.13. Summary of findings of siponimod for the treatment of adult patients with SPMS with

relapsing disease ............................................................................................................................ 65 Table 4.14. Overview of adverse events in the global SPMS population ................................................ 67 Table 4.15. Summary of ancillary analysis: indirect comparisons between siponimod and

comparators of interest in the global (active and non-active) SPMS population ................................ 70 Table A1. Original SLR search strategy for MEDLINE Daily, MEDINE Epub Ahead of Print and

Embase (via the Embase.com platform) .......................................................................................... 81 Table A2. Original SLR search strategy for MEDLINE® InProcess (via the PubMed.com platform) ....... 84 Table A3. Update SLR search terms for MEDLINE Daily, MEDLINE In-Process and Epub Ahead of

Print (searched via Ovid SP) ............................................................................................................ 86 Table A4. Update SLR search terms for Embase (searched via Ovid SP) ............................................. 89 Table A5. Original and Update SLR search strategy for CENTRAL and CDSR (via the Wiley Online

Platform) ......................................................................................................................................... 92 Table A6. Original SLR search strategy for DARE and HTAD (via the CRD platform) ............................ 94 Table A7. Congress searches for the clinical SLR update ..................................................................... 95 Table A8. Search terms used for other grey literature searches in the clinical SLR update .................... 95 Table A9. Publications included in the original SLR ............................................................................... 96 Table A10. SLR Update – Publications included in the SLR update .................................................... 104 Table A11. Rationale for exclusion of trials used for indirect or treatment comparisons from the SLR

results ........................................................................................................................................... 105 Table A12. Overview of guidelines used for this assessment .............................................................. 106 Table A13. Recommendations for research ........................................................................................ 107 Table A14. Study design of studies in the global SPMS population ..................................................... 109 Table A15. Inclusion/exclusion criteria of studies in the global SPMS population ................................. 110 Table A16. Outcome definitions of studies in the global SPMS population ........................................... 111 Table A17. Patient baseline characteristics of studies in the global SPMS population ......................... 112 Table A18. Placebo arm treatment effects of studies in the global SPMS population ........................... 113 Table A19. Endpoints and definitions .................................................................................................. 115 Table A20. Analysis sets .................................................................................................................... 119 Table A21. Patient disposition on double-blind study drug ................................................................... 120 Table A22. Analysis sets .................................................................................................................... 121 Table A23. Patient baseline characteristics ......................................................................................... 121 Table A24. Multiple sclerosis disease history ...................................................................................... 122



Table A25. Multiple sclerosis baseline characteristics ......................................................................... 123 Table A26. Prior medications .............................................................................................................. 124 Table A27. Most frequently reported adverse events (≥3%) ................................................................ 133 Table A28. Most frequently reported treatment emergent SAEs .......................................................... 134 Table A29. Most frequently reported TEAEs causing permanent study drug discontinuation ............... 135 Table A30. Matching inclusion/exclusion criteria – Nordic SPMS Study vs EXPAND ........................... 139 Table A31. Adjustment factors for CDP – Nordic SPMS Study ............................................................ 140 Table A32. Results of population matching and adjustment for CDP ................................................... 140 Table A33. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1a

22 µg once a week) ....................................................................................................................... 142 Table A34. Matching inclusion/exclusion criteria – SPECTRIMS vs EXPAND ..................................... 143 Table A35. Adjustment factors for CDP – SPECTRIMS ...................................................................... 144 Table A36. Adjustment factors for ARR – SPECTRIMS ...................................................................... 144 Table A37. Results of population matching and adjustment for CDP-3 ................................................ 145 Table A38. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1a

22 µg three times weekly) .............................................................................................................. 146 Table A39. Matching inclusion/exclusion criteria (SPECTRIMS vs. EXPAND) ..................................... 147 Table A40. Adjustment factors for CDP – SPECTRIMS ...................................................................... 148 Table A41. Adjustment factors for ARR – SPECTRIMS ...................................................................... 148 Table A42. Results of population matching and adjustment for CDP ................................................... 149 Table A43. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1a

44 µg three times weekly) .............................................................................................................. 150 Table A44. Matching inclusion/exclusion criteria – North American Study vs EXPAND ........................ 153 Table A45. Matching inclusion/exclusion criteria – European Study vs EXPAND ................................. 154 Table A46. Adjustment factors for CDP-6 – North American Study ...................................................... 155 Table A47. Adjustment factors for CDP-3 – European Study ............................................................... 155 Table A48. Results of population matching and adjustment for CDP-6– North American Study ........... 156 Table A49. Results of population matching and adjustment for CDP-3 – European Study ................... 156 Table A50. Results of population matching for ARR – North American Study ...................................... 157 Table A51. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1b

250 µg every other day) ................................................................................................................. 159 Table A52. Matching inclusion/exclusion criteria – ASCEND vs EXPAND ........................................... 160 Table A53. Adjustment factors for CDP – ASCEND ............................................................................ 161 Table A54. Adjustment factors for ARR – ASCEND ............................................................................ 161 Table A55. Results of population matching and adjustment for CDP ................................................... 162 Table A56. Summary of treatment effects on outcomes by ITC method (siponimod vs natalizumab

300 mg q4w) ................................................................................................................................. 164 Table A57. Pairwise comparison of inclusion/exclusion criteria – EXPAND vs IMPACT ....................... 165 Table A58. Adjustment factors for CDP – IMPACT .............................................................................. 166 Table A59. Adjustment Factors for ARR – IMPACT ............................................................................ 166 Table A60. Results of population matching and adjustment for CDP ................................................... 167 Table A61. Summary of treatment effects on outcomes by ITC method (siponimod vs interferon β-

1a 60 µg once a week) .................................................................................................................. 169



Figures

Figure 4.1. Original systematic literature review (flow of studies) ........................................................... 39 Figure 4.2. PRISMA flow diagram ......................................................................................................... 40 Figure A1. Study design of EXPAND study ......................................................................................... 114 Figure A2. Patients free of 3-month CDP ............................................................................................ 125 Figure A3. Time to 3-month CDP by subgroup (FAS) .......................................................................... 127 Figure A4. Time to 6-month CDP by subgroup (FAS) .......................................................................... 128 Figure A5. Patients free of 3-month confirmed worsening of at least 20% from baseline in T25FW ...... 129 Figure A6. Time to 3-month confirmed worsening of at least 20% from baseline in T25FW by

subgroup (FAS) ............................................................................................................................. 129 Figure A7. Change from baseline in T2 lesion volume by subgroup (FAS) .......................................... 130 Figure A8. Proportion of relapse-free patients over time ...................................................................... 131 Figure A9. Network diagram for time to confirmed disability progression at 6 Months .......................... 136 Figure A10. Network diagram for time to confirmed disability progression at 3 Months ........................ 137 Figure A11. Network diagram for annualised relapse rate ................................................................... 137 Figure A12. MAIC scenario analysis results for time to CDP-6 (siponimod vs IFNβ-1a 22 µg once a

week) ............................................................................................................................................ 141 Figure A13. MAIC results for ARR (siponimod vs IFNβ-1a 22 µg once a week) ................................... 141 Figure A14. MAIC scenario analysis results for time to CDP-3 – siponimod vs IFNβ-1a 22 µg three

times weekly .................................................................................................................................. 145 Figure A15. MAIC Results for ARR – Siponimod vs IFNβ-1a 22 µg three times weekly ....................... 146 Figure A16. MAIC scenario analysis results for Time to CDP-3 – siponimod vs IFN beta-1a 44 µg

three times weekly ......................................................................................................................... 149 Figure A17. MAIC results for ARR – siponimod vs IFNβ-1a 22 µg three times weekly ......................... 150 Figure A18. MAIC scenario analysis results for time to CDP-6 – siponimod vs IFN-1b 250 µg every

other day ....................................................................................................................................... 157 Figure A19. MAIC scenario analysis results for time to CDP-3 – Siponimod vs IFN-1b 250 µg every

other day ....................................................................................................................................... 158 Figure A20. MAIC results for ARR – siponimod vs IFNβ-1b 250 µg every other day (North American

Study) ........................................................................................................................................... 158 Figure A21. MAIC scenario analysis results for proportion of patients with CDP-6 (96w) – siponimod

vs natalizumab .............................................................................................................................. 163 Figure A22. MAIC scenario analysis results for ARR – siponimod vs natalizumab ............................... 163 Figure A23. MAIC scenario analysis results for time to CDP-3 – siponimod vs interferon β-1a 60 µg

once a week .................................................................................................................................. 168 Figure A24. MAIC scenario analysis results for ARR – siponimod vs interferon β-1a 60 µg once a

week ............................................................................................................................................. 168



LIST OF ABBREVIATIONS

AE Adverse event

ARR Annualised relapse rate

ATC Anatomical Therapeutic Chemical [classification system]

AUC Area under the curve

BAF312 Siponimod

BVMT-R Brief Visuospatial Memory Test Revised

CDP Confirmed Disability Progression

CDP-3 Confirmed Disability Progression at 3 months

CDP-6 Confirmed Disability Progression at 6 months

CHMP Committee for Medicinal Products for Human Use

CI Confidence interval

CNS Central nervous system

CRD Centre for Reviews and Dissemination

CSR Clinical study report

CYP2C9 Cytochrome P450 2C9

DMT Disease-modifying therapy

DOICU Declaration of interest and confidentiality undertaking

DR Dedicated reviewers

EAN European Academy of Neurology

ECTRIMS European Committee for Treatment and Research in Multiple Sclerosis

EDSS Expanded Disability Status Scale

EMA European Medicines Agency

EPAR European Public Assessment Report

EQ-5D 5-Dimension European Quality of Life Questionnaire

EUnetHTA European Network for Health Technology Assessment

F2F Face to face meeting

FSS Functional Systems Scores

Gd+ Gadolinium enhancing

GRADE Grading of Recommendations, Assessment, Development and Evaluation

HR Hazard ratio

HRQoL Health-related quality of life

HTAi Health Technology Assessment international

ICD International Classification of Diseases

IDSS Integrated Disability Status Score

IFN-β Interferon beta

IPD Individual patient data

IRT Interactive response technology

ITC Indirect treatment comparisons

ITT Intention-to-treat

IV Intravenous therapy

MAH Market Authorization Holder

MD Mean difference

MAIC Matched adjusted indirect comparisons

MeSH Medical Subject Headings

MRI Magnetic resonance imaging

MS Multiple sclerosis

MSFC Multiple Sclerosis Functional Composite

MSIS-29 Multiple Sclerosis Impact Scale

MSWS-12 12-Item Multiple Sclerosis Walking Scale

n/a Not applicable

NEDA No evidence of disease activity

Neff Effective sample size

NICE National Institute for Health and Care Excellence

NMA Network meta-analysis

OIS Optimal information size

OR Odds ratio

PASAT Progression of Paced Auditory Serial Addition Test

PICO Patient, intervention, comparator and outcome

MAH Marketing Authorisation Holder

PMS Progressive multiple sclerosis

PPMS Primary progressive multiple sclerosis

PRISMA Preferred reporting items for systematic reviews and meta-analyses



REA Relative effectiveness assessment

RCT Randomised controlled trial

RFSS Regional Functional System Score

RMS Relapsing multiple sclerosis

RR Relative risk

RRMS Relapsing-remitting multiple sclerosis

S1P Sphingosine-1-phosphate

SDMT Symbol Digit Modalities Test

SAE Serious adverse event

SC Subcutaneous

SD Standard deviation

SLR Systematic literature review

SMD Standardised mean difference

SmPC Summary of product characteristics

SOP Standard operating procedure

SPMS Secondary progressive multiple sclerosis

STC Simulated treatment comparisons

TEAEs Treatment emergent adverse events

T25FW Timed 25-foot Walk test

WP4 Work package 4



EXECUTIVE SUMMARY OF THE ASSESSMENT OF SIPONIMOD

Introduction

Secondary progressive multiple sclerosis

Multiple sclerosis (MS) is a chronic immune-mediated disorder of the central nervous system (CNS) char-acterised by inflammation, demyelination, and degenerative changes including neuroaxonal loss and pro-gressive atrophy [1]. The exact aetiology of MS is unknown, although a number of genetic and environmen-tal risk factors associated with developing MS have been identified [1]. Secondary progressive MS (SPMS) occurs after an initial relapsing disease course, and a large proportion of patients with relapsing-remitting multiple sclerosis (RRMS) will eventually go on to develop SPMS. The gradual worsening of SPMS can occur with or without additional inflammatory events. In most clinical contexts, SPMS is diagnosed retro-spectively from a history of gradual worsening after initial relapses, with or without acute exacerbations during progression. The transition from RRMS to SPMS usually occurs within 10 to 20 years of disease onset [2]. The SPMS disease course can be further categorised as either having active inflammation (so-called “active” SPMS) or not having active inflammation (so-called “non-active” SPMS) based upon the presence or absence of clinical relapses (assessed at least annually) and/or magnetic resonance imaging (MRI) signs of activity (contrast-enhancing lesions or new and unequivocally enlarging T2 lesions) [3]. Nu-merous disease-modifying therapies (DMTs) are indicated for the treatment of RRMS (glatiramer acetate, interferon-beta (β) preparations, cladribine, dimethyl fumarate, fingolimod, teriflunomide, alemtuzumab, mi-toxantrone, natalizumab, ocrelizumab), and some of these DMTs are licensed for “relapsing MS”, which includes “active” SPMS. Interferon-β-1b is the only DMT specifically licensed for the treatment of SPMS with active disease as evidenced by relapses. According to clinical practice surveys, DMTs are often used in patients with SPMS with evidence of disease activity, but there is no consensus on the optimal treatment choice. Overall, there are very few effective treatments available for SPMS.

Compound under assessment

Siponimod (Mayzent®) (Table 1.1 and Table 1.2) is indicated for the treatment of adult patients with SPMS with active disease evidenced by relapses or imaging features of inflammatory activity. Siponimod is a sphingosine-1-phosphate (S1P) receptor modulator that selectively binds two out of the five G-protein-cou-pled receptors (GPCRs) for S1P, namely S1P1 and S1P5. By acting as a functional antagonist on S1P1 receptors on lymphocytes, siponimod prevents egress from lymph nodes, reducing the recirculation of T cells into the CNS and thereby limiting central inflammation. Siponimod is similar in activity to fingolimod, another S1P receptor modulator approved for RRMS [4]. However, in contrast fingolimod, which at phar-macological doses is additionally selective for the S1P3 and S1P4 receptors, has a longer half-life and is administered as a prodrug. The clinical development programme for siponimod comprised two controlled studies including a single one phase III, placebo-controlled study in patients with SPMS. The efficacy of siponimod compared with placebo was demonstrated in a subgroup of patients with active disease but not in those without relapses and focal MRI activity. Siponimod is formulated as film-coated tablets, with treat-ment starting with a dose of 0.25 mg once daily on days one and two titrated up to a maintenance dose of 2 mg once daily starting on day six. Dose adjustments are required in certain patients with specific CYP2C9 metaboliser status. Details can be found in the summary of product characteristics (SmPC) [5].

Objective and scope

The aim of this EUnetHTA Joint Relative Effectiveness Assessment was to compare the clinical effective-ness and safety of siponimod in the target patient populations with relevant comparators [6]. The target patient populations and relevant comparators (based on the requirements of EUnetHTA Partners) are de-fined in the project scope in Table 0.1.



Table 0.1. Scope of the assessment

Description Assessment scope

PICO

Population Adult patients with SPMS classified as active as evidenced by relapses and/or MRI signs of activity (contrast-enhancing lesions or new and unequivocally enlarging T2 lesions)

Intervention Siponimod in combination with best supportive care*

Comparison

- Interferon-β-1a or -β-1b plus best supportive care* - Mitoxantrone plus best supportive care* - Ocrelizumab plus best supportive care* - Natalizumab plus best supportive care* - Fingolimod plus best supportive care* - Cladribine plus best supportive care* - Rituximab plus best supportive care*

Outcomes Clinical effectiveness Rate b Relative

importance

Confirmed disability progression at six monthsa 9 critical

Other measures of disability progressiona (e.g., confirmed disability progression (CDP) at three months, time to CDP, Timed 25-Foot Walk Test (T25FW), Nine-Hole Peg Test, Multiple Sclerosis Walking Scale (MSWS-12), rate of patients that become confined to wheelchair use, progression of Paced Auditory Serial Addition Test (PASAT), Symbol Digit Modalities Test (SDMT), or Brief Visuospatial Memory Test Revised (BVMT-R)

8 critical

Symptomsa (e.g., fatigue and cognitive, bowel, and bladder dysfunction)

7 important

Clinical relapse (e.g., annualised relapse rate, proportion of relapse-free patients)

7 important

Mortality 9 critical

Health-related quality of life (HRQoL) (e.g., European Quality of Life-5 Dimensions (EQ-5D), Multiple Sclerosis Impact Scale (MSIS-29)

9 critical

MRI-measured inflammatory disease activity and burden (T1 gadolinium-enhancing lesions, new or enlarging T2 lesions, brain volume)

5 important

No evidence of disease activity (NEDA; absence of progression, relapses, and MRI activity)

5 important

Safety

Adverse eventsa 6 important

Serious adverse events 8 critical

Adverse events leading to treatment discontinuation 7 critical

Treatment-related mortality 9 critical

Study design Randomised controlled trials * Best supportive care (BSC) is defined as symptomatic management through targeted physical therapy and symptomatic pharma-

cological interventions such as fampridine for gait difficulties and baclofen or tizanidine for spasticity. Disease-modifying therapeu-tics (DMTs) are excluded from BSC. a Outcomes that are related to issues particularly emphasised by patient organisations. b According to the GRADE methodology, authors of a drug assessment must, as a first step in a drug assessment process, make a preliminary classification of the importance of the selected outcomes. The outcomes were rated on a 1-9 scale, in which critical out-comes were scored 7-9, important outcomes were scored 4-6, and non-important outcomes were scored 1-3. Each author (from the

Authoring Team) scored each outcome separately, and the final score was the arithmetic mean (the average of the set of numerical values as calculated by adding them together and dividing by the number of terms in the set) of the scores assigned by three au-thors. Authors also took patients’ perspectives expressed during the scoping phase into account.



A broad systematic literature review (SLR) was conducted in order to retrieve all relevant studies that could inform this evaluation. The final assessment deviates from what was planned in the scoping document as follows:

Comparisons were not performed between siponimod and mitoxantrone. Owing to the risk of adverse events, including cardiotoxicity and risk of cancer, the Marketing Authorisation Holder (MAH) consid-ered that siponimod would not be an alternative to mitoxantrone in clinical practice and, therefore, mitoxantrone was excluded from the list of comparators. Mitoxantrone is indeed only indicated for the treatment of patients with MS where no alternative therapeutic options exist according to the outcome of a procedure under article 30 of Directive 2001/83/EC: referral of Novantrone and associated names [7]. Taking this into consideration, the Authoring Team decided to accept the exclusion of mitoxan-trone considering its extremely limited use in clinical practice in Europe.

Methods

The assessment is based on the data and analyses included in the submission dossier prepared by the MAH. During the assessment, the completeness of the data and analyses in the submission dossier was verified. Furthermore, the methods for data analysis and synthesis applied by the MAH were checked against the requirements for the submission dossier and applicable EUnetHTA guidelines and assessed with regard to scientific validity.

Information used for the assessment of clinical effectiveness and safety were extracted from the submission dossier, the clinical study report (CSR) of the EXPAND study, the technical report for the SLR, the technical reports for the indirect treatment comparisons (ITCs) and simulated treatment comparisons (STCs), the feasibility assessment, the report for the matching adjusted indirect comparisons (MAICs) including the analysis code provided by the MAH, and the European Public Assessment Report (EPAR).

Literature search and study selection

An SLR was initially undertaken by the MAH to identify all relevant randomised controlled trials (RCTs) evaluating the efficacy and safety of DMTs for the treatment of SPMS. The search period was from inception until 17 October 2018, with a supplemental search conducted covering the period up to 21 March 2019. The evidence base with regard to siponimod was reviewed by the Authoring Team. Search strategies and the results of information retrieval strategies were checked for appropriateness by the Authoring Team. Since no major flaws were identified in the MAH search strategy, no supplemental searches were per-formed.

Risk of bias/certainty of evidence

The quality rating tool developed by the Cochrane Collaboration (version 5.0.1; 2008 edition reprinted in August 2012) [8] was used to assess the quality of the evidence in RCTs. The risk of bias for six different domains was judged by two assessors working independently, who then reached a consensus.

The quality rating tool developed by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) was used to classify the certainty of the evidence by outcome for each comparison.

Data analysis

The efficacy and safety of siponimod was evaluated in the phase 3, randomised, double-blind, placebo-controlled EXPAND trial [9]. The comparator in this trial (placebo) was not included in the PICO of this assessment, and the trial was conducted in a broader population than the population of interest in this assessment. Post hoc subgroup analysis of EXPAND formed the basis of the evidence for siponimod in “active” SPMS. None of the comparator studies identified by the MAH’s SLR specified an “active” SPMS subgroup defined by the presence of clinical relapses assessed at least annually and/or MRI activity, the population of interest in this assessment. In the absence of direct comparisons between siponimod and relevant comparators in the population of interest, ITCs were required to assess the relative effectiveness of siponimod and comparators. The feasibility of performing ITCs to obtain relative effects of siponimod against the comparators through the placebo common comparator was assessed based on the number of available studies, their design, and the patient characteristics. Subgroup-specific Bucher ITCs [10] were conducted using “relapsing SPMS” subgroups derived from individual patient data (IPD) in EXPAND to

https://www.eunethta.eu/methodology-guidelines



match the subgroup and outcome definitions of comparator trials. In support of this approach, the MAH submitted an analysis which found that the “active” and “relapsing” subgroups were similar with regards to the outcomes for time to three-month and six-month confirmed disability progression (CDP) compared to placebo. While similar analyses were not available for comparator studies, the Authoring team considered this to be a reasonable approach. However, notwithstanding the reasonableness of the approach, the au-thors consider that this rationale may not hold for every comparison given the variable definition of “relaps-ing” among comparator trials and that different definitions hamper the interpretation and external validity of the results, specifically with regard to the populations to which they apply.

Network meta-analysis (NMA) can be used to make comparisons between multiple treatments [11]. NMA is particularly useful when treatments have been compared directly in RCTs but there is also indirect evi-dence through a common comparator that can be used to strengthen the estimated relative effects. How-ever, as siponimod could only be connected to the comparators via the placebo arm and no additional evidence was available, NMA results would be identical to those from the Bucher ITCs. Therefore, an NMA was not deemed necessary, and results of the ITC were deemed sufficient.

To adjust for any potential imbalance in patient characteristics across studies, the feasibility of population adjustment methods such as MAIC [12] or STC [11] was assessed. However, there was insufficient infor-mation on the baseline characteristics of the relapsing population subgroups in comparator trials to allow these analyses to be carried out, as they require these characteristics for adjustment. Therefore, no popu-lation adjustment methods could be undertaken in the “relapsing” SPMS subgroup.

The MAH undertook ancillary analyses in the global (“active” and “non-active”) SPMS population: compar-isons were performed between siponimod and comparators using MAICs [12], Bucher ITCs [10], NMA [11], and STCs [11].

Patient involvement

EUnetHTA conducted an open call for patient input. Questions were based on the Health Technology As-sessment international (HTAi) questionnaire template and elicited patients’ and caregivers’ views on living with the disease, important outcomes to be considered in this assessment, and expectations about the drug under assessment. Four patient organisations completed the survey, namely the MS Society of Slovakia, Russian MS Society, Smaragd Sclerosis Multiplexes Betegek Egyesülete (Hungary), and Združenie Skle-rosis multiplex Nádej (Slovakia).

In addition, EUnetHTA contacted the patient organisations who responded to the open call to ask for input on the ranking of outcomes. None of the patient organisations responded to this additional request for information.

The information gathered from the open call was used to inform the scope of this assessment, in particular the outcomes to be considered. In the PICO (Table 0.1), the outcomes related to issues particularly em-phasised by patient organisations are marked with a superscript “a”. The frequency and route of admin-istration were also mentioned as important points to take into account.

Comments were received regarding how SPMS affects carers/unpaid caregivers. Even though this assess-ment will focus on patient-relevant outcomes, the costs (humans and financial) of carers/unpaid caregivers may be captured at the national level in an economic assessment with a societal perspective.

Results

Information retrieval and studies included in the assessment

Twenty-three RCTs were identified in the SLR and were considered by the MAH for inclusion. No trials pertained specifically to the population of interest, “active” SPMS evidenced by relapses or imaging features of inflammatory activity. Four randomised, double-blind, placebo-controlled trials were conducted in pa-tients with SPMS (both active and non-active) and reported results for a relapsing subgroup for three DMTs compared to placebo: siponimod (EXPAND), interferon-β-1a (Nordic SPMS Study [13], SPECTRIMS [14,15]), and interferon-β-1b (European Study [16,17]) (Table 4.2). Information on baseline patient charac-teristics for the “active” SPMS population subgroup was not available for the comparator studies (Table 4.7), so we instead present patient characteristics for the global SPMS population (Table 4.8). Excluded



studies included patients with clinically isolated syndrome, RRMS, and RMS. Three additional studies (AS-CEND [18], North American Study [19], and IMPACT [20]; Table 4.3) included the SPMS population but had no information on patients with active disease so were included in ancillary analyses only (Section 4.8).

No relevant data could be retrieved to perform several of the planned comparisons. Therefore:

Comparisons were not possible between siponimod and ocrelizumab, natalizumab, fingolimod, cladribine, and rituximab in the population of interest;

The relative effectiveness of siponimod in terms of MS symptoms, mortality, HRQoL, MRI inflamma-tory disease activity, no evidence of disease activity, adverse events, serious adverse events, treat-ment discontinuation due to adverse events, and treatment-related mortality was not evaluated in the population of interest.

Risk of bias/certainty of the evidence

The quality of the evidence was rated by the Authoring Team as very low for all comparisons. The risk of bias due to unblinding of participants and/or outcome assessors was unclear for all studies. In the case of EXPAND, this was due to deviation from blinding procedures for Expanded Disability Status Scale (EDSS) raters, which resulted in temporary access to potentially unblinding information. While sensitivity analyses, excluding patients affected by potential unblinding, were reassuring, the potential for bias cannot be com-pletely excluded. In the case of the comparator studies, interferon side effects are well recognised, raising concerns about the possibility of unblinding of participants. Furthermore, it was not clear whether blinding of investigators was adequate. In the EXPAND trial, only 737/1105 (66.7%) patients in the siponimod group and 322/546 (59.0%) patients in the placebo group completed study participation on the double-blind study drug. Although the reasons for discontinuation seemed to be balanced between treatment arms in the EX-PAND study, the high rate of premature discontinuation, which was greater than the proportion of patients with CDP-6 events, was considered a matter of concern and a potential risk of bias. Both the Nordic SPMS Study and the European Study were terminated early following the release of negative results from the SPECTRIMS study in the case of the Nordic SPMS Study and following an interim analysis showing benefit in the case of the European Study. In both cases, this early termination was considered a potential risk of bias. The quality of the evidence in every study was also downgraded by the Authoring Team due to im-precision arising from the use of post hoc subgroup analyses, for which the studies were not powered. As none of the studies reported the baseline characteristics of the relapsing subgroups, it was not possible to assess the comparability of the relapsing populations between studies. The “similarity assumption”, on which the validity of the indirect comparisons rests, was therefore not possible to verify.

In the relapsing SPMS population, it was only possible to compare:

Siponimod vs interferon-β-1a 44 µg three times weekly, subcutaneous (SC) (SPECTRIMS), for the outcome time to CDP-3;

Siponimod vs interferon-β-1a 22 µg three times weekly or 44 µg three times weekly, SC (SPEC-TRIMS), for the outcome clinical relapses;

Siponimod vs interferon-β-1a 22 µg once a week, SC (Nordic SPMS Study) for the outcome time to CDP-6;

Siponimod vs interferon-β-1b 250 µg every other day, SC (European Study), for the outcome propor-tion of patients with CDP-3 at 33 months.

Due to the absence of outcome data in the population of interest in comparator studies, it was not possible to compare the treatment effects of siponimod vs comparators for any of the following outcomes: MS symp-toms, mortality, HRQoL, MRI inflammatory disease activity, no evidence of disease activity, adverse events, serious adverse events, adverse events leading to treatment discontinuation, or treatment-related mortality.



Clinical effectiveness results

The indirect comparisons using Bucher ITCs (Table 0.2) suggested that for siponimod:

In comparison with interferon-β-1a 22 µg once a week, there was no statistically significant difference in the risk of CDP-6 events (hazard ratio 0.70; 95%CI 0.43 to 1.15);

In comparison with interferon-β-1b 250 µg every other day, there was no statistically significant dif-ference in the proportion of patients with CDP-3 events (hazard ratio vs relative risk 0.81; 95%CI 0.57 to 1.15);

In comparison with interferon-β-1a 44 µg three times weekly, there was no statistically significant difference in the risk of CDP-3 events (hazard ratio 0.88; 95%CI 0.55 to 1.42);

In comparison with interferon-β-1a 22 µg three times weekly, there was no statistically significant difference in the risk of ARR (hazard ratio 1.10; 95%CI 0.65 to 1.87);

In comparison with interferon-β-1a 44 µg three times weekly, there was no statistically significant difference in the risk of annualised relapse rate (ARR; hazard ratio 0.94; 95%CI 0.59 to 1.49).



Table 0.2. Summary of relative outcomes in the relapsing SPMS population using Bucher ITCs

Comparator in-tervention

Regimen Study ID(s) Notable assumptions

Subgroup Bucher ITC results: Siponimod vs comparator (95% CI) Certainty of the

evidencea

Type Value No.

studies No. participants

Time to CDP-6 in subgroup: patients with relapses in the 4 years before study

Rebif® (SC IFN-β-1a)

22 µg once a week

Nordic SPMS Study

- HR 0.70

(0.43 - 1.15) 2

EXPAND – 1651 Nordic SPMS – 371 Total: 2022

Very low

Proportion with CDP-3 (33 months) in subgroup: patients with relapses in the 2 years before study

Betaferon® (SC IFN-β-1b)

250 µg every other day

European Study Imputation of censored data for EXPAND: last observa-tion carried forward*

OR 0.88

(0.53 - 1.47) 2

EXPAND – 1651 European – 718 Total: 2369

Very low

Betaferon® (SC IFN-β-1b)


European Study

Compared subgroup RR of European Study with sub-group time-to-event HR of EXPAND**

RR com-pared to

HR**

0.81 (0.57 - 1.15)

2

EXPAND – 1651 European – 718 Total: 2369

Very low



44 µg three times weekly‡

SPECTRIMS - HR 0.88

(0.55 - 1.42) 2

EXPAND – 1651 SPECTRIMS – 618 Total: 2269

Very low

ARR in subgroup: patients with relapses in the 2 years before study†


22 µg three times weekly

SPECTRIMS SPECTRIMS: assumed the rate ratio p-value=0.001 in order to calculate the 95% CI†

Rate ratio 1.10

(0.65 - 1.87) 2


Very low


44 µg three times weekly

SPECTRIMS Rate ratio 0.94

(0.59 - 1.49) 2


Very low

Source: adapted from the technical reports provided by the MAH. p-values not reported by the MAH. * Last observation carried forward: censored patients were assumed to have the same status at the cut-off time-point as they did on the day they were censored. Note that complete case could not be derived from the EXPAND IPD due to the volume of censoring (in the complete case, censored patients are removed from the data).

** Must be interpreted with caution. † Analysis of the ARR outcome for the relapsing subgroup in SPECTRIMS required making an assumption about the p-value (reported as "<0.001 for both doses") in order to calculate the 95% CI, which was not reported. The p-value was assumed to equal to 0.001 for each dose. ‡ Time to CDP-3 subgroup results not reported for the 22 µg dose in SPECTRIMS. a According to the GRADE methodology Abbreviations: ARR=annualised relapse rate; CI=confidence interval; HR=hazard ratio; IFN=interferon; OR=odds ratio; RR=relative risk ; SC IFNβ=subcutaneous interferon beta.



Diagnosing the transition from active SPMS to non-active SPMS can be challenging. Therefore, although the global SPMS population was not the population of interest in this assessment, deciding whether or not to discontinue siponimod in patients transitioning from active to non-active SPMS will be difficult for clini-cians in real-world practice. Thus, we present the results of studies in the global SPMS population to inform decisions on this combined population. These results are summarised in APPENDIX 6: Ancillary analysis.

Safety results

Comparison of siponimod vs comparators of interest in the active SPMS population

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments for adverse events, serious adverse events, adverse events leading to treatment discontinuation, or treatment-related mortality in the active SPMS population. There-fore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on these outcomes in the population of interest.

Comparison of siponimod vs comparators of interest in the global SPMS population

Due to the absence of data in comparator studies, for the safety outcomes in the population of interest a naïve indirect comparison was performed in the global SPMS population. Though its relevance is limited, a table was added with the available information on safety included in EXPAND and comparator studies (Table 0.3).

Comparison of siponimod vs placebo in the global SPMS population (EXPAND study)

As there were no relevant data on comparative safety between siponimod and comparators, the toxicity of siponimod can only be assessed using data from the EXPAND study (siponimod vs placebo; information available in APPENDIX 5: EXPAND (Study A2304)). Adverse events were more frequent in the siponimod group than in the placebo group (88.7% vs 81.5%). Serious adverse events were also reported more frequently in the siponimod group than in the placebo group (17.9% vs 15.2%). More patients discontinued the study drug permanently due to adverse events in the siponimod group (7.6% vs 5.1%). Adverse events of special interest that were reported more frequently in the siponimod group were herpes zoster reactiva-tions (2.2% vs 0.7%), lymphopenia (1.6% vs 0%), macular oedema (1.7% vs 0.2%), and increased liver transaminases (1.4% vs 0.6%). Globally, the safety information raised no specific safety concerns.

Other safety issues

Siponimod is mainly metabolised via CYP2C9. In a phase I study of siponimod, subjects with CYP2C9*2*3 or CYP2C9*3*3 genotypes had a two- or four-fold increase, respectively, in area under the curve (AUC) and T½ compared with patients with the CYP2C9*1*1 (extensive metaboliser) genotype. Likewise, the pharmacokinetic parameters for the metabolites M3 and M5 were delayed or lower in subjects with the CYP2C9*3*3 polymorphism compared with those with the CYP2C9*1*1 genotype. In phase I studies, pa-tients with the CYP2C9*3*3 genotype had an ~3 bpm lower heart rate on each day of up-titration compared to those with the CYP2C9*1*1 genotype. Also, patients with CYP2C9*2*3 and CYP2C9*3*3 genotypes had a lower percentage of lymphocytes than extensive metabolisers (CYP2C9*1*1). The risk of macular oedema was also higher in poor metabolisers. Based on the expected risk of high chronic exposure, those patients with the CYP2C9*3*3 (poor metaboliser) polymorphism were not included in the EXPAND study. Thus, siponimod should not be used in patients with a CYP2C9*3*3 polymorphism. In patients with a CYP2C9*2*3 or *1*3 genotype, the recommended maintenance dose is 1 mg once daily. Therefore, the use of siponimod requires the determination of CYP2C9 genotypes. This genotyping assay is not currently used in clinical practice and may not be available in some countries. The availability and cost of this gen-otyping assay will need to be evaluated locally.



Table 0.3. Adverse events in the global SPMS population

EXPAND SPECTRIMS European Study North American Study

ASCEND IMPACT Nordic SPMS Study

Safety outcome

Siponimod (n = 1099)

n (%)

Placebo (n = 546)

n (%)

IFN-β-1a 22 µg

(n = 209) n (%)

IFN-β-1a 44 µg

(n = 204) n (%)

Placebo (n = 205)

n (%)

IFN-β-1b (250 µg)

(n = 360) n (%)

Placebo (n = 358)

n (%)

IFN-β-1b 250 µg/m2

(n = 317) n (%)

Placebo (n = 308)

n (%)

Natalizumab (n = 439)

n (%)

Placebo (n = 440)

n (%)

IFN-β-1a 60 µg

(n = 217) n (%)

Placebo (n = 218)

n (%)

IFN -β-1a 22 µg

(once weekly)

(n = 186)

n (%)

Placebo (n = 178)

n (%)

Total number of adverse events

975 (88.7) 445 (81.5) NR NR NR NR NR NR

NR 401 (91.3) 410 (93.2)

215 (99.1)

215 (98.6)

NR NR

Total number of serious adverse events

197 (17.9) 83 (15.2) NR NR NR NR NR NR

NR 90 (20.5) 100 (22.3)

NR NR 79

(42.5)

72

(40.4)

Total number

of deaths

4 (0.4) 4 (0.7) 1 (0.5) 2 (1.0) 2 (1.0) 3 (0.8) 1 (0.3) 4 (1.3)

1 (0.3) 2 (0.5) 0 (0) 2 (0.9) 0 (0) 2 (1.1) 2 (1.1)

Total number of adverse events leading

to temporary or permanent treatment

withdrawal

160 (14.5) 44 (8.1) NR NR NR 5 (1.4) 4 (1.1) 30 (9.5)

12 (3.9) NR NR NR NR NR NR

Total number of withdrawals from the study

because of adverse events

84 (7.6) 28 (5.1) NR NR NR NR NR NR

NR 21 (4.8) 21 (4.8) 13 (6.0) 8 (3.6) 16 (8.6) 6 (3.4)

Source: Adapted from MAH submission dossier and Nordic SPMS Study [13]. Abbreviation: IFN β=interferon beta; NR=not reported



Discussion

The aim of this assessment was to assess the relative effectiveness of siponimod in comparison with the appropriate comparators in adult SPMS patients with active disease evidenced by relapses or imaging features of inflammatory activity. Significant challenges arose from the post hoc subgroup analysis of the EXPAND trial, from which the evidence for efficacy of siponimod was drawn, the lack of a head-to-head comparison with any comparator of interest, and the general lack of evidence for the comparator treat-ments in both the population and outcomes of interest. The information and analyses included in the MAH’s submission deviated significantly from the project scope, including only a small proportion of the compar-ators and outcomes of interest. The PICO defined during the scoping phase included the following treat-ments as comparators: interferon-β-1a or -β-1b, mitoxantrone, ocrelizumab, natalizumab, fingolimod, cladribine, and rituximab, all in combination with best supportive care. Relative effectiveness assessments were only conducted for siponimod and interferon-β-1a or -β-1b. The MAH’s submission did not address the majority of the outcomes included in the project scope, including many outcomes rated by partners as critical or important. The Authoring Team recognises that these omissions from the MAH’s submission arose from the absence of evidence for the comparator treatments in active or global SPMS and for many of the outcomes of interest.

In the absence of an active comparator in the EXPAND trial of siponimod, indirect comparisons were required to assess the relative effectiveness of siponimod and comparators. The quality of the evidence informing the ITCs was very low due to concerns regarding the use of post hoc subgroup analyses for which the studies were not powered, blinding, and incomplete data. None of the studies reported the baseline characteristics of the relapsing subgroups, so it was not possible to assess the comparability of the relapsing populations between studies. The validity of ITC results is dependent on whether the studies meet the assumption of consistency/similarity (meaning comparability, or assumption that the true treat-ment effect of each of the treatments in the contrast is comparable across the trials that contribute to the comparison). Since there may have been differences in the populations included in each trial, it is unclear whether the treatment effect was consistent between the studies compared. This limits the interpretation of the findings.

An “active” SPMS subgroup, as defined in the EXPAND study, was not reported for comparator studies. Instead, “relapsing” subgroups were defined variably in each study, requiring comparison-specific relaps-ing SPMS subgroups to be derived from the EXPAND IPD to match the subgroup and outcome definitions of the specific comparator study. The MAH submitted an analysis that showed that the “active” and “re-lapsing” subgroups of EXPAND were similar with regards to the outcomes of time to three-month and six-month CDP compared to placebo. While similar analyses were not available for comparator studies, the Authoring Team considered this to be a reasonable approach. However, notwithstanding the reasonable-ness of the approach, different definitions hampered the interpretation and external validity of the results, specifically with regard to the populations to which they apply.

The outcomes included in the relative effectiveness assessments in the MAH’s submission were limited to disability progression and relapse rate. At the request of the European Medicines Agency (EMA), the MAH performed a set of post hoc subgroup analyses to try to assess whether the benefit of siponimod on disease progression is independent of the presence of relapses. Based on these post hoc subgroup anal-yses, the Committee for Medicinal Products for Human Use (CHMP) considered that it remains very chal-lenging to disentangle the effect of siponimod on relapses and on general disease progression. The ITCs showed that, in patients with relapsing SPMS, there were no statistically significant differences between siponimod and interferon-β-1a 22 µg once a week, interferon-β-1b 250 µg every other day, and interferon-β-1a 44 µg three times weekly in relation to disability progression. Also, there were no statistically signifi-cant differences between siponimod and interferon-β-1a 22 or 44 µg three times weekly in relation to relapse rate. Furthermore, given the serious limitations associated with the available data, the individual trials, and the indirect comparison, these results should be interpreted with caution.

Ancillary analyses were conducted by the MAH using MAICs in the global (active and non-active) SPMS population. The Authoring team considers that MAIC is an inappropriate method for indirect treatment comparison in the setting of the current assessment. This is due to the inherent limitations of statistical techniques like MAIC [69], which has not been shown to produce less biased estimates than would be available through standard indirect comparisons, in the target population. Additionally, the alleged ad-vantage of matching and adjustment by the MAIC is not clear. Bucher ITCs were not deemed appropriate by the MAH due to significant differences between studies in inclusion/exclusion criteria and in the baseline characteristics of the populations included in those studies. However, at the Authoring Team’s request,



additional comparisons were performed using STCs, Bucher ITCs, and NMA. For each pairwise indirect comparison, different methodologies generally produced consistent results. Overall, in the global SPMS population, siponimod in comparison with natalizumab, interferon-β-1a 22 and 44 µg three times weekly, interferon β-1b 250 µg every other day, or interferon β-1a 60 µg once a week did not show statistically significant differences with respect to disease progression or clinical relapses. These results are consistent with those obtained in the relapsing SPMS population. Siponimod in comparison with interferon-β-1a 22 µg once a week showed a statistically significant difference favouring siponimod. As described for the primary analysis in this assessment, results of the ancillary analyses are highly uncertain and should be interpreted with caution.

Conclusion

For disability progression in patients with relapsing SPMS, indirect comparisons (using Bucher’s method [10]) of treatment effects from subgroup analyses showed no statistically significant differences between siponimod and interferon-β-1a 22 µg once a week (hazard ratio 0.70; 95%CI 0.43 to 1.15), interferon-β-1b 250 µg every other day (odds ratio 0.88; 95%CI 0.53 to 1.47), and interferon-β-1a 44 µg three times weekly (hazard ratio 0.88; 95%CI 0.55 to 1.42).

For relapse rate in patients with relapsing SPMS, an indirect comparison (using Bucher’s method [10]) of treatment effects from subgroup analyses showed no statistically significant differences between siponi-mod and interferon-β-1a 22 µg three times weekly (relative risk 1.10; 95%CI 0.65 to 1.87) or 44 µg three times weekly (relative risk 0.94; 95%CI 0.59 to 1.49).

The quality of evidence underlying these comparisons is very low, indicating very little confidence in the effect estimate. Due to the serious limitations of the evidence, these results should be interpreted with caution.

The use of siponimod requires the determination of CYP2C9 genotypes. This genotype test is not used in clinical practice and may not be available in some countries. The availability and cost of this genotype test will have to be evaluated locally.

There was no evidence, either direct or indirect, to evaluate the compared efficacy and safety of siponimod with most of the available comparators in use in current clinical practice in Europe for patients with active SPMS, namely ocrelizumab, natalizumab, fingolimod, cladribine and rituximab.

Taking into consideration the lack of comparative clinical evidence available at the time of the marketing authorisation and presented by the MAH in the submission dossier, the Authoring team suggest that further studies are warranted. Recommendations for further research may be found in Table A13 of APPENDIX 3: Evidence gaps.



1 BACKGROUND

1.1 Overview of the disease or health condition

Multiple sclerosis (MS) is a chronic immune-mediated disorder of the central nervous system (CNS) characterised by inflammation, demyelination, and degenerative changes including neuroaxonal loss and progressive atrophy [1]. The exact aetiology of MS is unknown, although a number of genetic and environmental risk factors have been identified [1].

Among studies that provided standardised estimates, MS prevalence in Europe varies from 170.5 per 105 of the population in the Swedish county of Värmland [21], 154.5 per 105 in Denmark [22], and 163 per 105 in Seinäjoki, a district of Finland [23] to 70.6 per 105 inhabitants in Las Palmas in the Canary Islands, Spain [24]. With respect to incidence, estimates vary from 7.6 per 105 of the popu-lation in Oppland County, Norway, between 1994 and 1998 [25] and 11.6 per 105 in Seinäjoki, Fin-land, between 1979 and 1993 to 4.1 per 105 between 1998 and 2002 in Las Palmas [24]. These results seem to suggest a north to south gradient in MS incidence and prevalence in the northern hemisphere [26].

In 1996, the US National Multiple Sclerosis Society Advisory Committee on Clinical Trials in Multiple Sclerosis defined four clinical subtypes of MS: relapsing-remitting (RRMS), secondary progressive (SPMS), primary progressive (PPMS), and progressive relapsing (PRMS) [27]. The RRMS pheno-type is the most common, as approximately 85% of those with MS initially experience relapses and remissions of neurological symptoms, with relapses often associated with new areas of CNS inflam-mation. The remaining patients diagnosed with MS, approximately 15%, gradually accrue disability from disease onset independently of relapses over time (PPMS). PRMS was defined as a progres-sive disease from onset with clear acute relapses, with or without full recovery, and periods between relapses characterised by continuing progression. Half of individuals with RRMS may evolve into SPMS [27], in whom gradual worsening can occur with or without additional inflammatory events. In most clinical contexts, SPMS is diagnosed retrospectively from a history of gradual worsening after an initially relapsing disease course, with or without acute exacerbations during the progressive course. To date, no clear clinical, imaging, immunological, or pathological criteria have been defined to determine the transition point from RRMS to SPMS; the transition is usually gradual [3]. The tran-sition from RRMS to SPMS usually occurs within 10 to 20 years following disease onset [2].

Disability progression can manifest in many different ways, and progression occurs over several years. Clinically, syndromes of neurological progression can include one or more symptoms or signs such as lower limb spasticity, ataxic paraparesis, sphincteric dysfunction, and cerebellar signs. The Expanded Disability Status Scale (EDSS) is the metric most commonly used to define disability pro-gression over time in clinical trials. The EDSS is an ordinal scale ranging from 0 to 10 in 0.5 point increments that assigns a severity score to the patient's clinical status [28]. The scores from grades 0–4.5 refer to people with MS who are fully ambulatory and are determined using Functional Systems Scores (FSS) that evaluate disability in the following eight neurological systems: pyramidal, cerebel-lar, brainstem, sensory, bladder, bowel, vision, and cerebral. EDSS steps 5.0 to 9.5 are defined by impaired walking. The Multiple Sclerosis Functional Composite (MSFC) is another quantitative meas-ure of MS-related disability that includes arm, leg, visual, and cognitive components [29].

In 2013, an international expert panel (International MS Phenotype Group) proposed a revision of the earlier classification criteria to further characterise the clinical course of progressive MS (PMS) [3]. These changes included categorisation of disease course in PMS as either having active inflam-mation (so-called “active” PMS) or not having active inflammation (so-called “non-active” PMS) based upon the presence or absence of clinical relapses, assessed at least annually, and/or mag-netic resonance imaging (MRI) signs of activity (contrast-enhancing lesions or new and unequivo-cally enlarging T2 lesions) [3]. The expert panel also recommended classifying PMS based on the presence or absence of gradual clinical disease progression as measured by clinical evaluation, assessed at least annually. Thus, disease "progression" is a measure of disability, independently quantified from relapses, and characteristic of PPMS and SPMS. The phenotype of progressive dis-ease (PPMS and SPMS) can be characterised as one of the following:



Active and with progression;

Active but without progression;

Not active but with progression;

Not active and without progression (stable disease).

For example, according to this classification, a patient with SPMS who has gradually worsened and has gadolinium (Gd)-enhancing lesions on MRI is classified as active and progressing SPMS.

More recently, the term relapsing MS (RMS) has also been used to describe both RRMS and SPMS patients with superimposed relapses [30]. Patients with RMS, whether or not they suffer from neuro-logic progression in the absence of relapses, have a common, inflammatory pathophysiology and, therefore, constitute a common target for treatment.

To evaluate the efficacy of a product designed to prevent disability progression in SPMS, it is rec-ommended to target only SPMS patients without a recent relapse, with no MRI activity suggestive of active inflammation, and with evidence of recent progression independent of relapses [30].

Interferon-beta-1b (interferon-β-1b) was the first disease-modifying therapy (DMT) for MS to receive regulatory approval and remains the only DMT specifically licensed for the treatment of SPMS with active disease evidenced by relapses. Since the introduction of interferon-β-1b there have been sig-nificant therapeutic advances, and several DMTs have been approved for the treatment of MS. Most approved therapies target various immune cells that contribute to the inflammatory cascade identified in MS. However, overall, MS treatments seem to be more efficacious in the earlier stages of RRMS and decreases over time. The following therapies are indicated for the treatment of RRMS: interferon-β preparations, glatiramer acetate, dimethyl fumarate, fingolimod, teriflunomide, alemtuzumab, na-talizumab. In RMS: interferon-β-1b, cladribine, mitoxantrone, and ocrelizumab are indicated. Accord-ing to surveys of clinical practice, these drugs are often also used in patients with SPMS with evi-dence of disease activity, but there is no agreement on the optimal treatment choice [31,32]. These surveys also acknowledge the difficulties in detecting when patients transition from RRMS to SPMS and in incorporating the recent revisions of the classification of disease course (into active and non-active forms of progressive MS) into clinical practice.

Only one therapy (ocrelizumab) has recently been approved to reduce the risk of progression in PPMS [33].

1.2 Current clinical practice

In the European trial of Betaferon® in SPMS [16], interferon-β-1b (Table 1.1 and Table 1.2) was associated with an increased time to onset of sustained progression of disability as measured by the EDSS. This treatment effect, independent of baseline EDSS score and previous relapses, led to the approval of interferon-β-1b for patients with SPMS in Europe and Canada [34]. However, similar benefits were not found in the other phase III trials: the North American trial of interferon-β-1b (Be-taseron®), the Secondary Progressive Efficacy Clinical Trial of Recombinant Interferon-β-1a in Mul-tiple Sclerosis (SPECTRIMS) trial of interferon-β-1a (Rebif®) [15] (Table 1.1 and Table 1.2), and the International Multiple Sclerosis Secondary Progressive Avonex Controlled Trial (IMPACT) of inter-feron-β-1a (Avonex®) [20] all showed no benefit. Some investigators have suggested that the differ-ent treatment effects measured by the EDSS in these clinical trials reflect fundamental differences in the cohorts of included patients. Patients in the European trial were, on average, younger and less likely to be free of exacerbations for two years before study enrolment than those in the other trials [34]. Consistent with this interpretation, the National Institute for Health and Care Excellence (NICE) in the UK recommended withdrawal of approval for interferon-β-1b for patients with SPMS without superimposed relapses. According to the European Committee for Treatment and Research in Mul-tiple Sclerosis (ECTRIMS)/European Academy of Neurology (EAN) guidelines on the pharmacolog-ical treatment of MS, treatment with interferon-β-1a (subcutaneously), interferon-β-1b, ocrelizumab, cladribine, or mitoxantrone should only be considered in patients with active SPMS taking into ac-count, in discussion with the patient, the uncertain efficacy and the safety and tolerability profiles of these drugs [35] (APPENDIX 2: Guidelines for diagnosis and management, Table A12).

Mitoxantrone (Table 1.1 and Table 1.2) is an anthracycline analogue that is used as a chemothera-peutic agent in some cancer patients. Small randomised controlled trials (RCTs) found that mitoxan-trone is effective in patients with worsening RRMS or SPMS, particularly in patients with superim-posed relapses. However, the risks of cardiotoxicity and potential for leukaemia development with



mitoxantrone limit its utility. The largest trial of mitoxantrone in MS was a single multicentre, double-blind trial of 194 patients with worsening RRMS or SPMS (where worsening was defined as 1.0 or more EDSS points during the 18 months before enrolment) [36]. Patients were randomly assigned to treatment with intravenous (IV) mitoxantrone or placebo every three months for two years. Treat-ment with mitoxantrone was associated with significant clinical benefits compared to placebo in mul-tivariate analysis, reducing progression of disability and clinical exacerbations. According to the EC-TRIMS/EAN guidelines [35], treatment with mitoxantrone should be considered in patients with active SPMS taking into account, in discussion with the patient, the efficacy and specifically the safety and tolerability profile of this agent (Table 1.1 and Table 1.2).

Ocrelizumab (Table 1.1 and Table 1.2) is a recombinant human anti-CD20 monoclonal antibody that binds to a different but overlapping CD20 epitope from that of rituximab. Evidence from large ran-domised trials has shown that, in patients with relapsing MS with at least two documented clinical relapses within the two years prior to screening or one clinical relapse in the year prior to screening, ocrelizumab was more effective than interferon-β-1a in reducing relapses and, in the pooled analysis, disability progression. This trial did not specifically include patients with SPMS [37]. Ocrelizumab was also the first drug approved in the EU for the treatment of adult patients with early PPMS in terms of disease duration and level of disability with imaging features characteristic of inflammatory activity [33] (Table 1.1 and Table 1.2).

Natalizumab (Table 1.1 and Table 1.2) is a recombinant monoclonal antibody directed against the integrin alpha-4 subunit, thereby blocking integrin association with vascular receptors and limiting leukocyte adhesion and transmigration. Natalizumab is highly effective for the treatment of RRMS. However, in a phase III placebo-controlled trial in patients with SPMS for at least two years and disability progression unrelated to relapses in the previous year (ASCEND), natalizumab did not sig-nificantly reduce disability progression as assessed by the primary multicomponent endpoint and secondary endpoints [18]. However, progression of the upper limb component of the primary disabil-ity endpoint was reduced [18].

Cladribine (Table 1.1 and Table 1.2) is a nucleoside analogue of deoxyadenosine that selectively depletes dividing and non-dividing T and B cells implicated in the pathogenesis of MS. Cladribine is approved for patients with RMS. Controlled trials of cladribine have not shown consistent benefits, specifically for patients with the progressive forms of MS, although some data suggest benefit in patients with SPMS. Cladribine was associated with a reduction in the imaging signs of activity in two trials of patients with progressive forms of MS [38,39]. No significant treatment effects were found, however, for cladribine in terms of disability progression.

Fingolimod (Table 1.1 and Table 1.2) is a sphingosine analogue that modulates the S1P receptor and thereby alters lymphocyte migration, resulting in sequestration of lymphocytes in lymph nodes. There is evidence from several RCTs that fingolimod effectively reduces the relapse rate in patients with RRMS. In a trial of patients with PPMS (INFORMS), fingolimod compared with placebo failed to slow disease progression [27,40]. No studies have assessed the efficacy of fingolimod in SPMS.

Although treatment of patients with PPMS with rituximab (Table 1.1 and Table 1.2) was not associ-ated with delayed time to confirmed disease progression in the OLYMPUS trial [41], the available information on treatment practices for MS in Europe suggests that rituximab is occasionally used to treat SPMS [31].

The study of dimethyl fumarate in SPMS (INSPIRE) was terminated early by the sponsor for com-mercial reasons. Efficacy, patient-reported outcomes, and pharmacodynamic data were not ana-lysed (in ClinicalTrials.gov, accessed 29th March 2019). Teriflunomide and alemtuzumab have not been studied in SPMS. A Cochrane review concluded that glatiramer acetate is ineffective in pro-gressive MS patients [42]. The available information on treatment practices for MS in Europe sug-gests that dimethyl fumarate, teriflunomide, and glatiramer acetate are seldom used to treat SPMS [31]. Due to its safety profile, alemtuzumab is usually reserved for patients with highly active RRMS who have had an inadequate response to one or more first-line DMTs. This restricted use is recom-mended by the European Medicines Agency (EMA) [43].

Older immunomodulators including steroids [44], azathioprine [45], methotrexate [46], and cyclo-phosphamide [47] have also been studied in patients with MS, but the results have not shown con-vincing benefits with respect to progression.

Finally, in a trial, a synthetic peptide analogue of myelin basic protein (MBP8298) was found to be ineffective in SPMS [48].

https://www.uptodate.com/contents/cladribine-drug-information?search=progressive+multiple&topicRef=1686&source=see_link

http://clinicaltrials.gov/



According to a recent survey of diagnostic and treatment practices for MS in Europe, most respond-ents agreed that RRMS patients converting to a clinically apparent progressive course with evidence of MRI activity should have changes made to their ongoing injectable DMTs. However, there was no agreement on what the new treatment should be. Second-line treatments (fingolimod and natali-zumab) received the highest proportion of responses (~30% for each). However, as this survey was published in 2018, the reported patterns might have since changed, particularly after more recent label approvals for MS (e.g., ocrelizumab and cladribine). Furthermore, as a general rule, DMT treat-ment was only initiated in patients with SPMS who experienced Gd-positive lesions or exacerbations [31]. Additionally, there was general agreement that DMT treatment should only be initiated in pa-tients with SPMS who experience contrast-enhancing lesions or exacerbations [31].

1.3 Features of the intervention

Siponimod (Table 1.1 and Table 1.2) is a sphingosine-1-phosphate (S1P) receptor modulator that binds selectively to two out of five G-protein-coupled receptors (GPCRs) for S1P, namely S1P1 and S1P5. By acting as a functional antagonist on S1P1 receptors on lymphocytes, siponimod prevents egress from lymph nodes, reducing the recirculation of T cells into the CNS and thereby limiting central inflammation. It is similar in activity to fingolimod, another S1P receptor modulator approved for RRMS [4]. In contrast fingolimod, which at pharmacological doses is additionally selective for the S1P3 and S1P4 receptors, has a longer half-life and is administered as a prodrug.



Table 1.1. Features of the intervention and its comparators

Non-proprie-tary name

siponimod interferon-β-1a interferon-β-1b

mitoxantrone ocrelizumab natalizumab fingolimod cladribine rituximab

Proprietary name

Mayzent Avonex/Rebif/Plegridy Betaferon/Ex-tavia

Nationally au-thorised prod-ucts

Ocrevus Tysabri Gilenya Mavenclad MabThera

Registered EMA indica-tion

Mayzent is indicated for the

treatment of adult patients

with SPMS with active dis-

ease evidenced by relapses

or imaging features of inflam-

matory activity.

Avonex is indicated for the

treatment of:

- Patients diagnosed with re-

lapsing MS. In clinical trials,

this was characterised by two

or more acute exacerbations

(relapses) in the previous

three years without evidence

of continuous progression be-

tween relapses; Avonex

slows the progression of disa-

bility and decreases the fre-

quency of relapses.

- Patients with a single demy-

elinating event with an active

inflammatory process, if it is

severe enough to warrant

treatment with intravenous

corticosteroids, if alternative

diagnoses have been ex-

cluded, and if they are deter-

mined to be at high risk of de-

veloping clinically definite MS.

Avonex should be discontin-

ued in patients who develop

progressive MS.

Rebif is indicated for the

treatment of:

- Patients with a single demy-

elinating event with an active

inflammatory process, if alter-

native diagnoses have been

excluded, and if they are de-

termined to be at high risk of

developing clinically definite

MS.

- RMS. In clinical trials, this

was characterised by two or

Betaferon/Ex-

tavia are indi-

cated for the

treatment of:

- Patients with

a single demy-

elinating event

with an active

inflammatory

process, if it is

severe enough

to warrant

treatment with

intravenous

corticosteroids,

if alternative di-

agnoses have

been excluded,

and if they are

determined to

be at high risk

of developing

clinically defi-

nite MS

- Patients with

RRMS and two

or more re-

lapses within

the last two

years,

- Patients with

SPMS with ac-

tive disease,

evidenced by

relapses.

Mitoxantrone

is indicated for

the treatment

of patients

with highly ac-

tive RMS as-

sociated with

rapidly evolv-

ing disability

where no al-

ternative ther-

apeutic op-

tions exist.

Ocrevus is in-

dicated for the

treatment of

adult patients

with relapsing

forms of MS

with active

disease de-

fined by clini-

cal or imaging

features.

Ocrevus is in-

dicated for the

treatment of

adult patients

with early

PPMS in

terms of dis-

ease duration

and level of

disability and

with imaging

features char-

acteristic of in-

flammatory

activity.

Tysabri is indicated*

as a single DMT in

adults with highly ac-

tive RRMS for the fol-

lowing patient

groups:

- Patients with highly

active disease de-

spite a full and ade-

quate course of treat-

ment with at least

one DMT;

or

- Patients with rapidly

evolving severe

RRMS defined by

two or more disabling

relapses in one year,

and with one or more

Gd-enhancing le-

sions on brain MRI or

a significant increase

in T2 lesion load as

compared to a previ-

ous recent MRI.

Gilenya is indicated

as a single DMT in

highly active RRMS

for the following

groups of adult pa-

tients and paediatric

patients aged 10

years and older:

- Patients with highly

active disease de-

spite a full and ade-

quate course of

treatment with at

least one DMT.

or

- Patients with rap-

idly evolving severe

RRMS defined by

two or more disa-

bling relapses in one

year, and with one

or more Gd-enhanc-

ing lesion(s) on

brain MRI or a sig-

nificant increase in

T2 lesion load as

compared to a previ-

ous recent MRI.

Mavenclad is

indicated for

the treatment

of adult pa-

tients with

highly active

RRMS as de-

fined by clinical

or imaging fea-

tures.

Multiple sclerosis (off-label).

SmPC indications: MabThera is indicated

in adults for the following indications:

- Non-Hodgkin’s lymphoma (NHL): previ-

ously untreated patients with stage III-IV

follicular lymphoma in combination with

chemotherapy. MabThera maintenance

therapy is indicated for the treatment of

follicular lymphoma patients responding to

induction therapy. MabThera monother-

apy is indicated for the treatment of pa-

tients with stage III-IV follicular lymphoma

who are chemoresistant or are in their

second or subsequent relapse after

chemotherapy. MabThera is indicated for

the treatment of patients with CD20-posi-

tive diffuse large B cell NHL in combina-

tion with CHOP (cyclophosphamide, dox-

orubicin, vincristine, prednisolone) chem-

otherapy.

- Chronic lymphocytic leukaemia (CLL):

MabThera in combination with chemother-

apy is indicated for the treatment of pa-

tients with previously untreated and re-

lapsed/refractory CLL. Only limited data

are available on efficacy and safety for

patients previously treated with monoclo-

nal antibodies including MabThera or pa-

tients refractory to previous MabThera

plus chemotherapy.

- Rheumatoid arthritis: MabThera in com-

bination with methotrexate is indicated for

the treatment of adult patients with severe

active rheumatoid arthritis who have had

an inadequate response or intolerance to

other disease-modifying anti-rheumatic

drugs (DMARDs) including one or more

tumour necrosis factor (TNF) inhibitor

therapies. MabThera has been shown to






Proprietary name




more acute exacerbations in

the previous two years.

Efficacy has not been demon-

strated in patients with SPMS

without ongoing relapse activ-

ity.

Plegridy is indicated in adult

patients for the treatment of

RRMS.

reduce the rate of progression of joint

damage as measured by X-ray and to im-

prove physical function when given in

combination with methotrexate.

- Granulomatosis with polyangiitis and mi-

croscopic polyangiitis: MabThera, in com-

bination with glucocorticoids, is indicated

for the treatment of adult patients with se-

vere, active granulomatosis with polyan-

giitis (Wegener’s; GPA) and microscopic

polyangiitis (MPA).

- Pemphigus vulgaris: MabThera is indi-

cated for the treatment of patients with

moderate to severe pemphigus vulgaris

(PV).

Marketing

authorisa-

tion holder

Novartis Europharm Limited Biogen Netherlands

B.V./Merck Europe B.V.

Bayer Pharma

A.G./Novartis

Europharm

Limited

Nationally

authorised

products

Roche Regis-

tration GmbH

Biogen Netherlands

B.V.

Novartis Europharm

Limited

Merck Europe

B.V.

Roche Registration GmbH

Contraindi-

cations - Hypersensitivity to the active

substance or to peanut, soya, or

any of its excipients;

- Immunodeficiency syndrome;

- History of progressive multifo-

cal leukoencephalopathy or

cryptococcal meningitis;

- Active malignancies;

- Severe liver impairment (Child-

Pugh class C);

- Patients who in the previous

six months had a myocardial in-

farction (MI), unstable angina

pectoris, stroke/transient is-

chaemic attack (TIA), decom-

pensated heart failure (requiring

inpatient treatment), or New

York Heart Association (NYHA)

class III/IV heart failure;

- Initiation of treatment in

pregnancy;

− Patients with a history of

hypersensitivity to natural or

recombinant interferon-β,

peginterferon, human albu-

min, or to any of its excipi-

ents;

− Patients with current severe

depression and/or suicidal

ideation;

− Patients with decompen-

sated liver disease.

- Initiation of

treatment in

pregnancy;

− Patients with

a history of hy-

persensitivity to

natural or re-

combinant in-

terferon-β, hu-

man albumin,

or to any of its

excipients;

− Patients with

current severe

depression

and/or suicidal

ideation;

− Patients with

decompen-

sated liver dis-

ease.

- Hypersensitiv-

ity to the active

substance or to

any of its excipi-

ents listed in-

cluding sulphites

that may be pro-

duced during the

manufacturing of

mitoxantrone;

- Mitoxantrone is

contraindicated

in women who

are breast-feed-

ing;

- Mitoxantrone must not be used in treat-

ment of MS in pregnant women.

- Hypersensitiv-

ity to the active

substance or to

any of its excipi-

ents;

- Current active

infection;

- Patients in a

severely immun-

ocompromised

state;

- Known ac-

tive malignan-

cies.

- Hypersensitivity to

natalizumab or to any

of its excipients;

- Progressive multifo-

cal leukoencephalo-

pathy (PML);

- Patients with in-

creased risk for op-

portunistic infections

including immuno-

compromised pa-

tients (including

those currently re-

ceiving immunosup-

pressive therapies or

those immunocom-

promised by prior

therapies);

- Combination with

other DMTs;

- Immunodeficiency

syndrome.

- Patients with in-

creased risk of op-

portunistic infections

including immuno-

compromised pa-

tients (including

those currently re-

ceiving immunosup-

pressive therapies

or those immuno-

compromised by

prior therapies);

- Severe active in-

fections, active

chronic infections

(hepatitis, tuberculo-

sis);

- Active malignan-

cies;

- Hypersensi-

tivity to the ac-

tive substance

or to any of its

excipients;

- Infection with

human immu-

nodeficiency vi-

rus (HIV);

- Active chronic

infection (tu-

berculosis or

hepatitis);

- Initiation of

cladribine treat-

ment in immun-

ocompromised

patients, in-

cluding pa-

tients currently

Contraindications for use in NHL and

CLL: hypersensitivity to the active sub-

stance or to murine proteins or to any of

its excipients. Active, severe infections.

Patients in a severely immunocompro-

mised state.

Contraindications for use in rheumatoid

arthritis: granulomatosis with polyangiitis,

microscopic polyangiitis, and pemphigus

vulgaris. Hypersensitivity to the active

substance or to murine proteins or to any

of the other excipients listed. Active, se-

vere infections. Patients in a severely im-

munocompromised state. Severe heart

failure (NYHA Class IV) or severe, uncon-

trolled cardiac disease.






Proprietary name




- Patients with a history of sec-

ond-degree Mobitz type II atrio-

ventricular (AV) block, third-de-

gree AV block, sino-atrial heart

block, or sick sinus syndrome if

they do not wear a pacemaker;

- Patients homozygous for the

CYP2C9*3 (CYP2C9*3*3) gen-

otype (poor metaboliser);

- During pregnancy and in

women of childbearing poten-

tial not using effective contra-

ception.

- Known active malig-

nancies, except for

patients with cutane-

ous basal cell carci-

noma.

- Severe liver impair-

ment (Child-Pugh

class C);

- Patients who in the

previous six months

had MI, unstable an-

gina pectoris,

stroke/TIA, decom-

pensated heart fail-

ure (requiring inpa-

tient treatment), or

NYHA class III/IV

heart failure;

- Patients with se-

vere cardiac arrhyth-

mias requiring anti-

arrhythmic treatment

with class Ia or class

III anti-arrhythmic

medicinal products;

- Patients with sec-

ond-degree Mobitz

type II AV block,

third-degree AV

block, or sick sinus

syndrome if they do

not wear a pace-

maker;

- Patients with a

baseline QTc inter-

val ≥ 500 msec;

- During pregnancy

and in women of

childbearing poten-

tial not using effec-

tive contraception;

- Hypersensitivity to

the active substance

or to any of its excip-

ients.

receiving im-

munosuppres-

sive or myelo-

suppressive

therapy;

- Active malig-

nancy;

- Moderate or

severe renal

impairment

(creatinine

clearance <60

mL/min);

- Pregnancy

and breast-

feeding.






Proprietary name




Pharmaceu-

tical formu-

lation(s)

Tablets Powder and solvent for solu-

tion for injection; solution for

injection in pre-filled pen; so-

lution for injection in pre-filled

syringe; solution for injection

in cartridge

Powder and

solvent for so-

lution for injec-

tion

Concentrate for solution for infusion; solu-tion for injec-tion

Concentrate

for solution for

infusion

Concentrate for solu-

tion for infusion

Hard capsules Tablets Concentrate for solution for infusion; solu-

tion for injection

ATC code L04AA42 siponimod L03AB07 interferon-β-1a/ L03AB13 peginterferon-β-1a

L03AB08 inter-

feron beta-1b

L01DB07 mi-

toxantrone

L04AA36 oc-

relizumab

L04AA23 natali-

zumab

L04AA27 fingolimod L04AA40

cladribine

L01XC02 rituximab

In vitro diag-

nostics re-

quired

Yes: Patients must be gen-

otyped for CYP2C9 to deter-

mine their CYP2C9 metabo-

liser status.

Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable Not applicable

Monitoring

required Yes No No No Yes Yes Yes No No

Sources: MAH submission dossier of Mayzent, SmPC for Mayzent [5], SmPC for Avonex [49], Rebif [50], Plegridy [51], Betaferon [52], Extavia [53], Mitoxantrone [54], Ocrevus [55], Tysabri [56], Gilenya [57], Mavenclad [58] and Mabthera [59]. * Based on the referral on alemtuzumab (article 20 procedure) recently concluded, the EMA has recommended restriction of the use of alemtuzumab due to reports of rare but serious side-effects including deaths. Taking

into consideration the final recommendation, this medicine should now only be used to treat RRMS if the disease is highly active despite treatment with at least one DMT or if the disease is worsening rapidly. Alemtuzumab must also no longer be used in patients with certain heart, circulation, or bleeding disorders or in patients who have autoimmune disorders other than MS. The medicine should only be given in a hospital with ready access to intensive care facilities and specialists who can manage serious adverse reactions [43].

Abbreviations: ATC= Anatomical Therapeutic Chemical; AV=atrioventricular; CLL=chronic lymphocytic leukaemia; DMARD=disease-modifying anti-rheumatic drug; DMT=disease-modifying therapeutic; GPA=granulo-matosis with polyangiitis; HIV=human immunodeficiency virus; MI=myocardial infarction; MPA=microscopic polyangiitis; MRI=magnetic resonance imaging; MS=multiple sclerosis; NHL=non-Hodgkin’s lymphoma; NYHA=New York Heart Association; PV=pemphigus vulgaris; RMS=relapsing multiple sclerosis; PPMS=primary progressive multiple sclerosis; SmPC=summary of product characteristics; SPMS=secondary progressive

multiple sclerosis; TIA=transient ischaemic attack.



Table 1.2. Administration and dosing of the technology and its comparators

Non-proprietary

name

siponimod interferon-β-1a interferon-β-1b mitoxantrone ocrelizumab natalizumab fingolimod cladribine rituximab

Method of ad-

ministration

Oral use Intramuscular injection Subcutaneous

injection

Intravenous infusion Intravenous infu-

sion

Intravenous infu-

sion

Oral use Oral use intravenous infu-

sion

Doses 0.25 mg and 2 mg 6 M.U.I./0.5 ml

6 M.U.I./3 ml 12 M.U.I./0.5 ml 12 M.U.I./ml 24 M.U.I./ml 2.4 M.U.I./0.2 ml 63 microgram/0.5 ml 94 microgram/0.5 ml

125 microgram/0.5 ml

250 μg/ml 2 mg/ml 300 mg of ocreli-

zumab in 10 mL (30 mg/ml). After dilution: 1.2 mg/ml

20 mg/ml. After

dilution: 2.6 mg/ml

0.25 mg and

0.5 mg

10 mg Concentrate for so-

lution for infusion: 100 mg/10 ml and 500 mg/50 ml

Solution for injec-tion: 1400 mg/11.7 ml and 1600 mg/13.4 ml

Dosing fre-

quency

Once daily.

Dose titration

starts with 0.25

mg once daily on

days 1 and 2, fol-

lowed by once

daily doses of 0.5

mg (two tablets of

0.25 mg) on day

3, 0.75 mg (three

tablets of 0.25

mg) on day 4, and

1.25 mg (five tab-

lets of 0.25 mg)

on day 5, to reach

the maintenance

dose of siponi-

mod starting on

day 6.

The maintenance

dose of siponi-

mod is 2 mg once

daily, although a

1 mg mainte-

nance dose is

recommended for

certain CYP gen-

otypes

(CYP2C9*2*3 or

*1*3).

Avonex 30 μg:

30 micrograms (1 ml solution) ad-

ministered by intramuscular (IM) in-

jection once a week.

Titration: Initiating therapy in ¼

dose increments per week reach-

ing the full dose (30 μg/week) by

the fourth week or ½ dose of

Avonex once a week before in-

creasing to the full dose.

Rebif:

44 μg given three times per week

by subcutaneous injection. A lower

dose of 22 μg, also given three

times per week by subcutaneous

injection, is recommended for pa-

tients who cannot tolerate the

higher dose in view of the treating

specialist.

When first starting treatment with

Rebif, the dose should be gradually

escalated.

Plegridy:

The recommended dose of Ple-

gridy is 125 μg injected subcutane-

ously every 2 weeks (14 days). It is

generally recommended that pa-

tients start treatment with 63 μg at

dose 1 (on day 0), increasing to 94

μg at dose 2 (on day 14), reaching

the full dose of 125 μg by dose 3

250 μg (8.0 mil-

lion IU), con-

tained in 1 ml of

the reconstituted

solution, to be

injected subcuta-

neously every

other day.

Titration: started

at 62.5 μg (0.25

ml) subcutane-

ously every

other day, and

increased slowly

to a dose of 250

μg (1.0 ml) every

other day.

The recommended

dosage of mitoxan-

trone is usually 12

mg/m2 body surface

area given as a short

(approximately 5 to

15 minutes) intrave-

nous infusion that

may be repeated

every 1-3 months.

The maximum life-

time cumulative dose

should not exceed 72

mg/m2.

If mitoxantrone is ad-

ministered repeat-

edly, dosing adjust-

ments should be

guided by extent and

duration of bone mar-

row suppression.

Initial dose: the

initial 600 mg

dose is adminis-

tered as two sep-

arate intravenous

infusions; first as

a 300 mg infu-

sion, followed two

weeks later by a

second 300 mg

infusion

Subsequent

doses: subse-

quent doses of

Ocrevus thereaf-

ter are adminis-

tered as a single

600 mg intrave-

nous infusion

every 6 months.

Intravenous infu-

sion once every

4 weeks

0.5 mg cap-

sule taken

orally once

daily

In paediatric

patients (10

years of age

and above),

the recom-

mended dose

is dependent

on body

weight: Pae-

diatric pa-

tients with

body weight

≤40 kg: one

0.25 mg cap-

sule taken

orally once

daily. Paedi-

atric patients

with body

weight >40

kg: one 0.5

mg capsule

taken orally

once daily.

The recommended cumulative

dose of Mavenclad is 3.5 mg/kg

body weight over two years, ad-

ministered as one treatment

course of 1.75 mg/kg per year.

Each treatment course consists

of 2 treatment weeks, one at the

beginning of the first month and

one at the beginning of the sec-

ond month of the respective

treatment year. Each treatment

week consists of 4 or 5 days on

which a patient receives 10 mg

or 20 mg (one or two tablets) as

a single daily dose, depending

on body weight. Following com-

pletion of the two treatment

courses, no further cladribine

treatment is required in years 3

and 4.

Multiple sclerosis

(off label): 1000

mg intravenous in-

fusions of rituxi-

mab every 24

weeks, on days 1

and 15 [41, 60].

SmPC indications:

Concentrate for so-

lution for infusion;

- multiple. Please

see SmPC.

Solution for injec-

tion:

- Subcutaneous in-

jection at a fixed

dose of 1600 mg

irrespective of the

patient’s body sur-

face area.



Non-proprietary

name

siponimod interferon-β-1a interferon-β-1b mitoxantrone ocrelizumab natalizumab fingolimod cladribine rituximab

(on day 28) and continuing with the

full dose (125 μg) every 2 weeks

(14 days) thereafter

Standard length

of a course of

treatment

It is anticipated

that siponimod is

taken continu-

ously.

Not known. Not known. The maximum life-



mg/m2.

Not known. 2 years. The ef-

ficacy of re-ad-

ministration has

not been estab-

lished.

Not known. Re-initiation of therapy after year

four has not been studied.

Not known.

Standard inter-

val between

courses of treat-

ments

One day. One week. Every other day. Every 1-3 months. Every 6 months. Every 4 weeks. Once daily. Two treatment weeks at the be-

ginning of the first and second

month of the respective treat-

ment year. Following completion

of the two treatment courses, no

further cladribine treatment is re-

quired in years 3 and 4.

Every 24 weeks.

Standard num-

ber of repeat

courses of treat-

ments

It is anticipated

that siponimod is

taken continu-

ously.

Not applicable/not known. Not applica-

ble/not known.

The maximum life-



mg/m2.

Not applicable/not

known.

Not applica-

ble/not known.

Not applica-

ble/not

known.

Not applicable/not known. Not applicable/not

known.

Dose adjust-

ments

In patients with a

CYP2C9*2*3 or

*1*3 genotype,

the recommended

maintenance

dose is 1 mg

taken once daily.

Yes. Please see SmPC. No/not known. Yes. Please see

SmPC.

Yes. please see

SmPC.

Yes. please see

SmPC.

No/not

known.

No/not known. No dose reduc-

tions of MabThera

are recommended

for the SmPC indi-

cations.

Sources: MAH submission dossier of Mayzent, SmPC for Mayzent [5], SmPC for Avonex [49], Rebif [50], Plegridy [51], Betaferon [52], Extavia [53], Mitoxantrone [54], Ocrevus [55], Tysabri [56], Gilenya [57], Mavenclad [58] and Mabthera [59].

Abbreviation: SmPC=summary of product characteristics.



2 OBJECTIVE AND SCOPE

The aim of this EUnetHTA Joint Relative Effectiveness Assessment is to compare the clinical effective-ness and safety of siponimod in the target patient population with relevant comparators [61, 62,63,64]. The target patient population and relevant comparators (based on the requirements of EUnetHTA Part-ners) are defined in the project scope in Table 2.1.

One agency requested RRMS as an additional population in the PICO due to the national reimburse-ment system. The authors acknowledge that different systems may have valid requirements for the PICO; however, in this case, RRMS was considered out of the scope of both the EMA and the EU-netHTA submission and was therefore not included.

The assessment was based on a dossier submitted by the Marketing Authorisation Holder (MAH) No-vartis.

The scope of the assessment deviates from the scope (Table 2.1) described in the project plan as follows:

Comparisons were not performed between siponimod and mitoxantrone. Owing to the risk of adverse events, including cardiotoxicity and risk of cancer, the MAH considered that siponimod would not be an alternative to mitoxantrone in clinical practice and, therefore, mitoxantrone was excluded from the list of comparators. Mitoxantrone is indeed only indicated for the treatment of patients with MS where no alternative therapeutic options exist according to the outcome of a procedure under article 30 of Directive 2001/83/EC: referral of Novantrone and associated names [7]. Taking this into consideration, the Authoring Team decided to accept the exclusion of mito-xantrone considering the extremely limited use in clinical practice in Europe.

Although the global SPMS population is not the population of interest in this assessment, the decision of whether or not to discontinue siponimod in patients transitioning from active to non-active SPMS will be difficult for clinicians in real-world practice. Thus, results of ancillary analyses in the global SPMS were also summarised (APPENDIX 6: Ancillary analysis).



Table 2.1. Scope of the assessment

Description Assessment scope

PICO

Population Adult patients with SPMS classified as active as evidenced by relapses and/or MRI signs of activity (contrast-enhancing lesions or new and unequivocally enlarging T2 lesions)

Intervention Siponimod in combination with best supportive care*

Comparison

- Interferon-β-1a or -β-1b plus best supportive care* - Mitoxantrone plus best supportive care* - Ocrelizumab plus best supportive care* - Natalizumab plus best supportive care* - Fingolimod plus best supportive care* - Cladribine plus best supportive care* - Rituximab plus best supportive care*

Outcomes Clinical effectiveness Rate b Relative

importance

Confirmed disability progression at six monthsa 9 critical

Other measures of disability progressiona (e.g., confirmed disability progression (CDP) at three months, time to CDP, Timed 25-Foot Walk Test (T25FW), Nine-Hole Peg Test, Multiple Sclerosis Walking Scale (MSWS-12), rate of patients that become confined to wheelchair use, progression of Paced Auditory Serial Addition Test (PASAT), Symbol Digit Modalities Test (SDMT), or Brief Visuospatial Memory Test Revised (BVMT-R)

8 critical

Symptomsa (e.g., fatigue and cognitive, bowel, and bladder dysfunction)

7 important

Clinical relapse (e.g., annualised relapse rate, proportion of relapse-free patients)

7 important

Mortality 9 critical

Health-related quality of life (HRQoL) (e.g., European Quality of Life-5 Dimensions (EQ-5D), Multiple Sclerosis Impact Scale (MSIS-29)

9 critical

MRI-measured inflammatory disease activity and burden (T1 gadolinium-enhancing lesions, new or enlarging T2 lesions, brain volume)

5 important

No evidence of disease activity (NEDA; absence of progression, relapses, and MRI activity)

5 important

Safety

Adverse eventsa 6 important

Serious adverse events 8 critical

Adverse events leading to treatment discontinuation 7 critical

Treatment-related mortality 9 critical

Study design Randomised controlled trials * Best supportive care (BSC) is defined as symptomatic management through targeted physical therapy and symptomatic phar-macological interventions such as fampridine for gait difficulties and baclofen or tizanidine for spasticity. Disease-modifying

therapeutics (DMTs) are excluded from BSC. a Outcomes that are related to issues particularly emphasised by patient organisations. b According to the GRADE methodology, authors of a drug assessment must, as a first step in a drug assessment process,

make a preliminary classification of the importance of the selected outcomes. The outcomes were rated on a 1-9 scale, in which critical outcomes were scored 7-9, important outcomes were scored 4-6, and non-important outcomes were scored 1-3. Each author (from the Authoring Team) scored each outcome separately, and the final score was the arithmetic mean (the av-

erage of the set of numerical values as calculated by adding them together and dividing by the number of terms in the set) of the scores assigned by three authors. Authors also took patients’ perspectives expressed during the scoping phase into ac-count.



3 METHODS

This assessment was based on the data and analyses included in the submission dossier prepared by the MAH. During the assessment, the completeness of data and analyses in the submission dossier was verified. Furthermore, the methods for data analysis and synthesis applied by the MAH were checked against the requirements of the submission dossier and applicable EUnetHTA Guidelines and assessed with regard to scientific validity.

Information used for the assessment of clinical effectiveness and safety were extracted from the sub-mission dossier, the clinical study report (CSR) of the EXPAND study, the technical report for the sys-tematic literature review (SLR), the technical reports for the indirect treatment comparisons (ITCs) and simulated treatment comparisons (STCs), the feasibility assessment and report for the matching ad-justed indirect comparisons (MAICs) including the analysis codes provided by the MAH and the Euro-pean Public Assessment Report (EPAR).

3.1 Information retrieval

During the assessment, the evidence base with regard to the drug under assessment provided by the MAH was reviewed by the Authoring Team [65]. Search strategies were checked for appropriateness, and the results of information retrieval included in the MAH’s submission dossier were checked for completeness as well as the rationale indicated for excluding studies following full text review.

An SLR for the treatment of SPMS was initially undertaken by the MAH (Table A1, Table A2, Table A5 and Table A6). The aim of the SLR was to identify all relevant RCTs evaluating the clinical effectiveness, safety, and tolerability of pharmacological treatments for patients with SPMS and to conduct a feasibility analysis to determine the suitability of each trial identified in the SLR for inclusion in a meta-analysis. The technical report of this SLR was provided by the MAH, which followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [66]. According to the technical report of the original systematic review, each citation was screened by two independent reviewers, and any discrepancies between reviewers were reconciled by a third independent reviewer. The eligibility criteria were applied to the full-text citations. Likewise, each full text article was screened by two inde-pendent reviewers, and any discrepancies between reviewers reconciled by a third independent re-viewer.

The MAH review used the following databases: MEDLINE, MEDLINE In-Process, the Cochrane Library, Centre for Reviews and Dissemination (CRD; University of York), and Embase. The search period was from inception until 17 October 2018. In addition, abstracts from various conference proceedings that took place over the past three years (2016-2018) were manually searched.

An updated search was conducted by the MAH on 21 March 2019 (Table A3, Table A4 and Table A5). The following databases were searched by the MAH: MEDLINE, including MEDLINE In-Process, MED-LINE Daily, Epub Ahead of Print, and Versions (via Ovid SP) from 1946 to 21 March 2019, Embase (via Ovid SP) from 1974 to 21 March 2019, CDSR up to Issue 3 of 12 March 2019 (via the Wiley Online platform), and CENTRAL up to issue 3 of 12 March 2019 (via the Wiley Online platform). An update of the conference abstract books was also conducted in March 2019 (Table A7). The complete search strategies are included in APPENDIX 1: SLR search strategies. For the purpose of the SLR, compara-tors were not restricted to those within the scope of this submission but were kept deliberately broad. The MAH also undertook a grey literature search of trial registries included in ClinicalTrials.gov on 13 August 2018, with no date restrictions (Table A8). In addition, the studies included in a Cochrane Re-view assessing treatment with interferon-β-1a or -β-1b in patients with SPMS were cross-referenced and were evaluated for inclusion in the current assessment. The PICO criteria used by the MAH for the SLR are shown in Table 3.1.

https://www.eunethta.eu/methodology-guidelines

http://clinicaltrials.gov/



Table 3.1. PICO for the systematic literature review

Category Inclusion criteria Exclusion criteria

Population

Adults (≥18 years) diagnosed with SPMS

Children (<18 years)

Relevant outcomes were not presented separately for adults with SPMS*

*Studies reporting results for mixed MS population were included if >80% of the population met the inclusion criteria for the disease

Interven-tions

Siponimod

Fingolimod

Interferon-β

Ocrelizumab

MIS416

Glatiramer acetate

Natalizumab

Masitinib

Peginterferon-β

Stem cell transplantation

Alemtuzumab

Dimethyl fumarate

Imilecleucel T

Idebenone

Simvastatin

Mitoxantrone

Teriflunomide

Ibudilast

Opicinumab

Fluoxetine

Rituximab

Cladribine

Biotin

Riluzole

Amiloride

Studies not investigating at least one of the relevant interventions

Compara-tors

Any intervention listed above

Placebo

Best supportive care

Dose-ranging studies were included; extension studies were only included if they contained a placebo or comparator arm

Any other comparator

Outcomes Any efficacy, HRQoL, or safety outcomes including:

Disability (e.g., EDSS, CDP at three months, confirmed disa-bility progression at six months);

Timed 25-foot walk test, 9-hole peg test, MSFC;

Relapses (e.g., annualised relapse rate, time to first relapse, proportion of patients relapse-free);

MRI parameters (e.g., number of new/enlarging T2 lesions, number of T1 Gd+ lesions, T1 lesion volume, T2 lesion vol-ume, brain volume);

Cognition (e.g., SDMT, PASAT, BVMTR);

HRQoL (e.g., MSIS-29, MSWS-12, EQ-5D).

Safety and tolerability (e.g., adverse events, serious adverse events, specific adverse events, treatment discontinuation)

Studies not reporting any eligible outcomes

Studies reporting eligible outcomes in a mixed popu-lation, without separately reporting data for the pop-ulation of interest (unless >80% of the study popula-tion are adults with SPMS)

Study

design

RCTs

Non-randomised interventional studies

- Prospective observational studies

- Retrospective observational studies

- Cross-sectional studies

- Case-control studies

Any other study design

SLR/NMAs were included at the abstract stage but subsequently excluded at the full text stage and their bibliographies hand searched for additional articles of relevance to this review

Source: MAH submission dossier. Abbreviations: BVMTR=Brief Visuospatial Memory Test-Revised; CDP=confirmed disability progression; EDSS=Expanded Disability Status Scale; EQ-5D=EuroQol five dimension scale; Gd=gadolinium; HRQoL=health-related quality of life; MRI=mag-

netic resonance imaging; MSFC=Multiple Sclerosis Functional Composite; MSIS-29=Multiple Sclerosis Impact Scale-29; MSWS-12=Multiple Sclerosis Walking Scale-12; PASAT=Paced Auditory Serial Addition Test; RCT=randomised controlled trial; SDMT=Symbol Digit Modalities Test; SPMS=secondary progressive multiple sclerosis.



3.2 Data extraction

Information used for the assessment of clinical effectiveness and safety was extracted from the sub-mission dossier, the CSR of the EXPAND study, the technical report for the first SLR for the treatment of SPMS, the ITC technical report, and other original documentation provided in the submission dossier, as well as the EPAR. The technical report of the systematic review update was not provided. The tech-nical report of ITCs included a systematic review performed in 2017. As this review was not mentioned in the submission dossier and was out-dated, it was not further considered in this analysis.

3.3 Risk of bias assessment

The quality rating tool developed by the Cochrane Collaboration (version 5.0.1; 2008 edition reprinted in August 2012 [8]) was used to assess the risk of bias in randomised trials. Risk of bias at study level was assessed for six different domains:

Method used to generate the sequence of randomisation (random sequence generation);

Method used to mask the sequence of allocation to treatment (allocation concealment);

Measures used to ensure the ‘blindness’ of the study with respect to treatment assignment (blind-ing of participants, medical personnel, and outcome assessors);

Completeness of the data for each outcome considered (incomplete outcome data);

Selective description of the results (selective outcome reporting);

Other sources of bias (e.g., bias due to the early interruption of the study because of the benefits without an appropriate stopping rule, use of a non-validated measurement instrument, incorrect statistical analysis).

For each domain, two independent assessors (from the author and co-author team) judged the risk of bias as ‘low risk’, ‘high risk’, or ‘unclear’ on the basis of the information retrieved from the full-text pub-lications, the protocols, and the submission dossier. A summary of the risk of bias assessment is de-picted in Table 4.9. A detailed description of the risk of bias assessment can be found in Section 4.6.

3.4 Analyses of included studies

The information in the submission dossier on the study design, study methods, populations, endpoints, and study results was evaluated. The results of this evaluation are presented in Section 4 and were used to identify relevant analyses and were considered in the conclusions (Section 7) of the assessment report.

Meta-analysis

Multiple comparators were of interest and no pair of interventions were compared in more than one study, so standard meta-analyses were not conducted. ITCs were conducted and are described below. A network meta-analysis (NMA) [11] was not deemed necessary, as detailed below.

Indirect comparisons

In the absence of a direct comparator in the EXPAND trial of siponimod, indirect comparisons were required to assess the relative effectiveness of siponimod and comparators.

The Bucher method for ITC can be used to compare outcomes between two treatments, which may not have been directly compared in a clinical trial, but which are compared to a common comparator [10]. This method of indirect comparison contrasts with naïve or unadjusted indirect comparisons as it allows randomisation between originally assigned patient groups to be preserved. The methodology assumes that the trials are sufficiently similar with regard to study design, population, outcome measures and the distribution of treatment effect-modifiers. In the case of siponimod, relative treatment effects versus comparators of interest may be derived from trials which compare siponimod versus placebo, and trials which compare comparators of interest versus placebo, for example.



However, none of the comparator studies identified by the MAH’s SLR reported results on an “active” SPMS subgroup (the population of interest in this assessment). Only results on “relapsing SPMS” sub-groups were reported. Bucher ITCs were performed using the relapsing subgroups of comparator stud-ies with corresponding subgroups in the EXPAND study by using the EXPAND IPD to match the “re-lapsing SPMS” and outcome definitions stated in the comparator trials.

To check if comparisons between relapsing SPMS sub-populations of EXPAND and comparator trials were an appropriate indicator of relative effectiveness in the active SPMS population, the MAH carried out a post hoc subgroup analysis of EXPAND to compare the active and relapsing subgroups with regards to the outcomes for time to three-month and six-month CDP compared to placebo. As the results were similar, the relapsing subgroup was used as a proxy for the active subgroup.

The feasibility of performing ITCs to obtain relative effects of siponimod against the comparators through the placebo common comparator was assessed by evaluating the number of available studies and their design and patient characteristics. Where possible, ITCs were carried out by applying the formulae described by Bucher et al. [10]. Where more than one study made the same comparison to placebo, a fixed effect meta-analysis was carried out to pool the studies, and its results were incorpo-rated in the ITC.

An NMA can be used to make comparisons between multiple treatments and is particularly advanta-geous when treatments have been compared directly in RCTs but there is also indirect evidence through a common comparator that can be used to strengthen the estimated relative effects. However, as siponimod could only be connected to the comparators via the placebo arm and no additional evi-dence was available, results of an NMA would be identical to those from the Bucher ITCs. Therefore, an NMA was not deemed necessary and ITC results were deemed sufficient.

To adjust for any potential imbalance in patient characteristics across studies, the feasibility of popula-tion adjustment methods was assessed. However, there was insufficient information on the baseline characteristics of the relapsing population subgroups in comparator trials to allow these analyses to be carried out, as they require availability of these characteristics for adjustment. Therefore, no population adjustment methods such as MAIC [12] or STC [67] could be undertaken in the “relapsing” SPMS sub-group.

Ancillary analyses

The MAH undertook ancillary analyses in the global (active and non-active) SPMS population. Com-parisons were performed between siponimod and comparators using MAICs, Bucher ITCs, NMA, and STCs.

The MAH considered that Bucher ITCs were not appropriate due to differences between studies in inclusion/exclusion criteria and proposed MAIC as an alternative analytical approach. However, insuf-ficient justification was provided for the use of MAIC. Therefore, the authors requested that comparisons between siponimod and comparators be made using Bucher ITCs and NMA, with MAICs and STCs provided as supporting analyses where considered feasible and valid. Comparisons were therefore performed between siponimod and comparators using MAICs, Bucher ITCs, NMA, and STCs.

Anchored MAICs were carried out to make the patients in EXPAND match the characteristics of the patients in each of the comparator studies by first excluding patients in EXPAND that would not be eligible for the comparator study (the matching step). Then, the remaining patients were re-weighted in order to match the reported covariate characteristics of the patients in each of the comparator studies (the adjustment step). Although not fully described in the MAH submission, the codes provided for the MAIC analyses suggests that matching and adjustment were performed on both the mean and variance of the covariates adjusted for, which is the recommended approach [68] [69]. The resulting relative effects of siponimod are only applicable to the comparator study’s population.

Bucher ITCs were also carried out as described above. Fixed-effect NMAs were also carried out. How-ever, given that the only loops in the networks were formed by a single multi-arm study (see APPENDIX 6: Ancillary analysis), this approach is exactly equivalent to Bucher ITC. Therefore, the resulting relative effects of siponimod compared to the other interventions and their validity were identical to those ob-tained through Bucher ITCs and will not be commented on further.



STCs are an alternative population adjustment method to MAIC, which aim to adjust for differences in population characteristics [67]. Anchored STCs were conducted as sensitivity analyses for the MAICs using the methods outlined in the NICE DSU TSD18 [69]. STCs are designed to reduce cross-trial differences in baseline patient characteristics and reduce sensitivity to effect modifiers. Regression models were applied to individual patient data from the EXPAND study, and the fitted model was used to simulate (i.e., predict) the effect of siponimod in the populations represented in the comparator stud-ies. The regression models were specified to incorporate prognostic effects and treatment effect modi-fiers, and the resulting relative effects apply to the comparator study populations.

Subgroup analysis and other effect modifiers

The MAH performed a systematic review searching for studies including siponimod and the compara-tors of interest in adult patients with SPMS. None of the comparator studies retrieved by the SLR iden-tified an “active” SPMS subgroup defined by the presence of clinical relapses assessed at least annually and/or MRI activity. Three comparator studies reported a relapsing subgroup and included the out-comes of interest. Therefore, the MAH extracted data from the relevant subgroups of the comparator studies and performed an analysis of SPMS patients with active disease, defined as the presence of clinical relapses assessed at least annually.

The MAH reported results of the post hoc subgroup analysis of EXPAND, which found that the active subgroup and relapsing subgroup were similar with regards to the outcomes for time to three-month and six-month CDP compared to placebo (three-month CDP: hazard ratio (HR) 0.68; 95%CI 0.52 to 0.89; p=0.0049) in the active SPMS subgroup, HR 0.67; 95%CI 0.50 to 0.91; p=0.0108 in the relapsing SPMS subgroup). Therefore, the MAH considered that comparisons between the relapsing SPMS sub-populations of EXPAND and comparator trials are an appropriate indicator of the relative effectiveness in the active SPMS population

Certainty of the evidence

The quality rating tool developed by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) [70-83] was used to assess the quality of the evidence in RCTs. GRADE judg-ments refer not only to individual studies but to a body of evidence, and the certainty of the evidence as used in GRADE means more than risk of bias. According to GRADE, the certainty of the evidence reflects our confidence that the effect estimates are true. The GRADE approach results in an assess-ment of the quality of a body of evidence as high, moderate, low, or very low. Randomised trials begin as high-quality evidence and then five reasons can be used to possibly down-rate the quality of evi-dence: risk of bias, imprecision, inconsistency, indirectness of study results, and publication bias. GRADE recommends moving from six risk of bias criteria for each individual study to a judgment about rating down the quality of evidence across a group of studies addressing a particular outcome. For indirect comparisons NMA, GRADE recommends assessing each comparison separately using the GRADE’s approach for assessing the certainty of evidence from conventional (direct) meta-analysis. The main analysis included only pairwise comparisons using Bucher ITC. As Bucher ITC is generally considered to be a simple form of NMA, the GRADE’s guideline for network meta-analysis was followed. This assessment is made in Section 4.6.

Risk of bias at outcome level was assessed for five different domains:

Method used to mask the sequence of allocation to treatment (allocation concealment);

Measures used to ensure the ‘blindness’ of the study with respect to treatment assignment (blind-ing of participants, medical personnel, and outcome assessors);

Completeness of the data for each outcome considered (incomplete outcome data);

Selective description of the results (selective outcome reporting);

Other sources of bias (e.g., bias due to the early interruption of the study because of the benefits without an appropriate stopping rule, use of a non-validated measurement instrument, incorrect statistical analysis).

The assessment of the quality of the evidence also included imprecision, inconsistency, and indirect-ness.



For each outcome, two independent assessors (from the author and co-author team) judged the quality of evidence (certainty of results) as “very low”, ‘low’, ‘moderate’, or ‘high’ on the basis of the information retrieved from the full-text publications, the protocols, and the submission dossier.

Patient involvement

At the start of this Joint Assessment, an open call for patient input was published on the EUnetHTA website. This open call specifically asked patient organisations to answer the questions posed, as they are in a position to collect and present patients’ and caregivers’ views and experiences by engaging with a wide range of patients and their carers.

The open call used by EUnetHTA asked general questions to elicit patients’ views on living with the disease, important outcomes to be considered in this assessment, and expectations about the drug under assessment. The questions were based on the Health Technology Assessment international (HTAi) questionnaire template. For more information on the development of the HTAi questionnaire template, please see the HTAi website.

European and national patient organisations were asked to provide an organisational perspective on the questions in English. For all parts of the open call, the term “patient” referred to anyone living with, or who has lived with, the condition for which the new medicine is indicated.

Four patient organisations completed the survey, namely the MS Society of Slovakia, the Russian MS Society, Smaragd Sclerosis Multiplexes Betegek Egyesülete (Hungary), and Združenie Sklerosis mul-tiplex Nádej (Slovakia).

In addition, EUnetHTA contacted the patient organisations who responded to the open call to ask for input on the ranking of outcomes.

https://htai.org/interest-groups/pcig/resources/for-patients-and-patient-groups/



4 RESULTS

4.1 Information retrieval

The MAH search retrieved 2655 citations from MEDLINE, 71 citations from MEDLINE In-Process, 657 citations from the Cochrane Library, and 95 citations from the CRD, totalling 3478 citations. After re-moving duplicates (n=266), the abstracts from 3212 citations were screened, and 2871 records were excluded for several reasons. Thus, 341 full-text articles were assessed for eligibility, with 248 articles being excluded. The most frequent reason for exclusion was “population not relevant” and “no subgroup SPMS”. Therefore, 23 studies from 97 publications (Table A9) were available for assessment.

The flow of studies through the systematic review process performed by the MAH is depicted in Figure 4.1. There are minor inconsistencies between the flowchart provided in the technical report of the review and the flowchart provided in the final submission dossier of the MAH, particularly the total number of records identified through searching the CRD (96 vs 95), number of duplicates (265 vs 266), conference searches (4 vs 3), and final number of publications (97 vs 98). However, the final number of studies is consistent (n=23), and therefore this was not considered a major flaw.

Figure 4.1. Original systematic literature review (flow of studies)

Source: MAH submission dossier. * One blank record was obtained and removed. Abbreviations: CRD=Centre for reviews and dissemination; CSR=Cochrane Database of Systematic Reviews; Embase=Ex-

cerpta Medica Database; MEDLINE=Medical Literature Analysis and Retrieval System Online; SLR= systemic literature review; SPMS=secondary progressive multiple sclerosis.



The systematic review update performed by the MAH retrieved 4178 citations. Records included in the previous review, as well as duplicates, were then removed, leaving 407 abstracts for screening. Cita-tions identified through supplementary searches (conference proceedings, reference lists, and website searches) were also screened again. Finally, the systematic review yielded five additional publications on two previously identified studies (Table A10). The flow of studies through this systematic review process undertaken by the MAH is presented in Figure 4.2.

Figure 4.2. PRISMA flow diagram Source: MAH submission dossier. Abbreviations: CDSR=Cochrane Database of Systematic Reviews; CENTRAL=Cochrane Controlled Register of Trials; Em-base=Excerpta Medica Database; MEDLINE=Medical Literature Analysis and Retrieval System Online; SPMS=secondary pro-

gressive multiple sclerosis.

Overall, from the original and updated systematic review, the MAH considered 23 studies for inclusion, which were then further refined to those studying a relevant comparator and reporting relevant out-comes. The reasons for exclusion were provided by the MAH (Table A11), particularly the lack of the intervention of interest in nine studies and the lack of relevant outcomes in seven studies. During the assessment, the evidence base with regard to the drug under assessment provided by the MAH was reviewed. Since no major flaws were identified in the MAH search strategy, no supplemental searches were performed. The MAH proposed the exclusion of one additional study based on the unlicensed regimen of interferon-β-1a in the Nordic SPMS Study, which was not accepted by the Authoring Team. Therefore, seven studies (Table 4.1) were included in the subsequent analysis.



4.2 Studies included in the assessment

Of the 23 trials identified in the SLR, no trials pertained specifically to the population of interest, SPMS with active disease evidenced by relapses and/or imaging features of inflammatory activity. Seven trials pertained to the global SPMS population (active with relapses and/or evidence of new MRI activity and not active), and four of them reported results for a relapsing subgroup. Table 4.1 presents the list of the seven trials retrieved.

Table 4.1. List of studies in the global SPMS population

Study ID Author (Year) NCT Numberb Citation Reference

EXPAND Kappos (2018) NCT01665144 Kappos L, Bar-Or A, Cree BAC, Fox RJ, Giovannoni G et al. (2018) Siponimod vs. placebo in secondary

progressive multiple sclerosis (EXPAND): a double-blind, randomised, phase III study. Lancet 391

(10127): 1263-1273.

[9]

ASCEND Kapoor (2018) NCT01416181 Kapoor R, Ho PR, Campbell N, Chang I, Deykin A et al. (2018) Effect of natalizumab on disease progres-sion in secondary progressive multiple sclerosis (AS-

CEND): a Phase III, randomised, double-blind, pla-cebo-controlled trial with an open-label extension. Lancet Neurol 17 (5): 405-415.

[18]

SPECTRIMS SPECTRIMS Study Group (2001) a

- SPECTRIMS Study Group (2001) Randomized con-trolled trial of interferon-beta-1a in secondary pro-gressive MS: Clinical results. Neurology 56 (11):

1496-1504.

[15]

Li (2001) Li DK, Zhao GJ, Paty DW (2001) Randomized con-trolled trial of interferon-beta-1a in secondary pro-gressive MS: MRI results. Neurology 56 (11): 1505-

1513.

[14]

North Ameri-can Study

Panitch (2004) - Panitch H, Miller A, Paty D, Weinshenker B (2004) In-terferon beta-1b in secondary progressive MS: results from a 3-year controlled study. Neurology 63 (10):

1788-1795.

[19]

European

Study

European Study

Group (1998)

- European Study Group (1998) Placebo-controlled

multicentre randomised trial of interferon beta-1b in treatment of secondary progressive multiple sclerosis. European Study Group on interferon beta-1b in sec-ondary progressive MS. Lancet 352 (9139): 1491-

1497.

[16]

Kappos (2001) a Kappos L, Polman C, Pozzilli C, Thompson A, Beck-

mann K et al. (2001) Final analysis of the European multicenter trial on IFNbeta-1b in secondary-progres-sive MS. Neurology 57 (11): 1969-1975.

[17]

Nordic SPMS

Study

Andersen

(2004)

- Andersen O, Elovaara I, Farkkila M, Hansen HJ,

Mellgren SI et al. (2004) Multicentre, randomised, double blind, placebo-controlled, phase III study of weekly, low dose, subcutaneous interferon beta-1a in secondary progressive multiple sclerosis. J Neurol Neurosurg Psychiatry 75 (5): 706-710.

[13]

IMPACT Cohen (2002) - Cohen JA, Cutter GR, Fischer JS, Goodman AD, Hei-

denreich FR et al. (2002) Benefit of interferon beta-1a on MSFC progression in secondary progressive MS. Neurology 59 (5): 679-687.

[20]

Source: Adapted from the technical reports provided by the MAH. a Indicates the pivotal publication for RCTs for which there are also supporting publications. b Note that ClinicalTrials.gov became available in 2008, and so trials published before this date will not have NCT numbers

available.

A comparison between siponimod and the comparators of interest in the active SPMS population was only possible for the comparator studies reporting the results of an “active” or relapsing sub-group. The studies identified by the SLR included patients with SPMS (active and non-active) including different proportions of patients with “active disease”. None of the comparator studies included only an “active” SPMS population. Moreover, none of the comparator studies identified by the SLR specified an “active” SPMS subgroup, the population of interest in this assessment, defined by the presence of clinical re-lapses assessed at least annually and/or MRI activity. Three studies reported outcomes in a relapsing SPMS subgroup, defined variably between studies: the Nordic SPMS Study (interferon-β-1a 22 µg once a week), SPECTRIMS (interferon-β-1a 22 µg three times weekly), and the European Study (interferon-



β-1b 250 µg every other day). The MAH decided to exclude the Nordic SPMS Study from the main analysis due to the use of an unlicensed regimen of interferon-β-1a. However, as the interferon-β-1a dose/regimen used in the Nordic SPMS Study may be used in clinical practice in some countries, it was decided to include the Nordic SPMS Study in the analyses. The information on the presence/absence of a relapsing subgroup in the comparator studies is depicted in Table 4.2.

Table 4.2. Reported relapsing subgroups in SPMS

Study ID Intervention Contains a rel-evant “active” or “relapsing” subgroup?

Reported sub-groups

Reported subgroup-specific outcomes

Requirement to conduct ITC

Nordic SPMS Study

IFN-β-1a (Rebif®)

Yes Patients with and without relapses in the 4 years prior to baseline

Time to CDP-6 Derive sub-group from EXPAND IPD

SPECTRIMS IFN-β-1a (Rebif®)

Yes Patients with and without relapses in the 2 years before study

ARR

Time to CDP-3

Calculate out-comes from EXPAND IPD

European Study

IFN-β-1b (Betaferon®)

Yes Patients with and without relapses in the 2 years before study

Proportion of patients with CDP-3 at 33 months

Calculate out-come from EXPAND IPD; requires im-putation of censored pa-tients

EXPAND Siponimod (Mayzent)

Yes Patients with and without relapses in the 2 year be-fore study; Num-ber of Gd-en-hancing T1 le-sions at base-line; others.

Time to CDP-3, time to CDP-6, time to 3-month confirmed worsening of at least 20% from baseline in T25FW, change from baseline in T2 lesion volume

n/a

Source: Adapted from the technical reports provided by the MAH. Abbreviations: ARR=annualised relapse rate; CDP-3=confirmed disability progression at 3 months; CDP-6=confirmed disabil-ity progression at 6 months; IFN-β=interferon beta; ID=identification; IPD=individual patient data; ITC=indirect treatment com-

parison; n/a=not applicable; SPMS=secondary progressive multiple sclerosis.

Thus, four studies (including EXPAND) were available for assessment of the population of interest.

4.3 Excluded studies

Sixteen studies did not include the population of interest and instead included patients with clinically isolated syndrome, RRMS, and RMS, and were therefore excluded from the assessment. Of note, trials focusing on relapsing MS did not report information on specific outcomes for patients with SPMS. Three additional studies (ASCEND, North American, IMPACT) (Table 4.3) that included the SPMS population but had no outcome information on patients with active disease were only included in sensitivity anal-yses (APPENDIX 4: Studies included in the Ancillary analyses of the global SPMS population).



Table 4.3. Studies that had no information on patients with relapsing disease

Study ID Intervention Contains a rele-vant “active” or “relapsing” sub-group?

Reported sub-groups

Reported sub-group-specific outcomes

Requirement to conduct ITC

North

American

Study

IFN-β-1b

(Betaferon®)

No None n/a n/a

ASCEND Natalizumab

(Tysabri®)

Yes Patients with and

without relapses

in the prior 12 or

24 months

None n/a

IMPACT IFN-β-1a

(Avonex®)

Yes Patients with and

without relapses

in the 1 year be-

fore study

None n/a

Source: Adapted from the technical reports provided by the MAH.

Abbreviation: IFN=interferon.

4.4 Characteristics of included studies

EXPAND was a phase 3, randomised, double-blind, placebo-controlled trial conducted at 292 centres in 31 countries for up to three years or until the occurrence of at least 374 three-month CDP events. The study included 1,651 patients aged between 18 and 60 years old, with SPMS, with an EDSS be-tween 3.0 and 6.5, who were randomly assigned in a 2:1 ratio to oral siponimod (n=1,105) 2 mg once daily or placebo (n=546) for up to three years or until the occurrence of at least 374 three-month CDP events. The primary endpoint was three-month CDP. The key secondary endpoints were time to three-month confirmed worsening of at least 20% from baseline in the timed 25-foot walk test (T25FW) and change in T2 lesion volume from baseline.

SPECTRIMS was a randomised, double-blind, placebo-controlled trial conducted at 22 centres in Eu-rope, Canada, and Australia, which tested two doses of interferon-β-1a (IFN-β-1a 22 or 44 µg SC three times weekly). The study included 618 patients aged between 18 and 55 years old with SPMS, with an EDSS at study entry between 3.0 and 6.5, who were randomly assigned in a 1:1:1 ratio to receive 22 µg (n=209) or 44 µg (n=204) interferon-β-1a three times weekly (n=1,105) or placebo (n=205) for up to three years. The primary endpoint was three-month CDP. Secondary endpoints were proportion of pa-tients progressing, exacerbation count, time to first exacerbation, time between first and second exac-erbations, number of moderate and severe exacerbations, number of steroid courses for MS, number of hospitalisations for MS, and Integrated Disability Status Score (IDSS; defined by area under an EDSS time-curve adjusted for baseline).

Nordic SPMS study was a randomised, double-blind, placebo-controlled trial conducted at 32 centres in Denmark, Finland, Norway, and Sweden. The study included 371 patients aged between 18 and 65 years old, with SPMS, with an EDSS ≤6.5 at baseline, who were randomly assigned in a 1:1 ratio to interferon-β-1a 22 µg once weekly (n=186) or placebo (n=178) for up to three years. The primary end-point was six-month CDP. The secondary endpoint was time to progression on the Regional Functional System Score (RFSS), defined as an increase of 2% or more from baseline on the RFSS. The study was terminated early during year three following the release of results from SPECTRIMS using inter-feron-β-1a 22 or 44 µg three times weekly, which showed no statistically significant difference between groups with respect to three-month CDP.

The European Study was a randomised, double-blind, placebo-controlled trial conducted at 32 Euro-pean centres. The study included 718 patients aged between 18 and 55 years old with SPMS, with an EDSS at study entry between 3.0 and 6.5, who were randomly assigned in a 1:1 ratio to interferon-β-1b 250 µg every other day (n=360) or placebo (n=358) for up to three years. The primary endpoint was three-month CDP. Secondary endpoints were time to becoming wheelchair-bound, proportion of pa-tients becoming wheelchair-bound, and EDSS at the primary endpoint. The study was terminated early for benefit.



In the absence of a direct comparator in the EXPAND trial of siponimod, indirect comparisons were required to assess the relative effectiveness of siponimod and comparators. As no studies were identi-fied for comparators in the active SPMS population, studies reporting treatment effects in “relapsing SPMS” populations (variably defined) were selected by the MAH and are presented in Table 4.4. Table 4.4 also presents the design of the studies included in the assessment for the relapsing SPMS popula-tion. For assessing comparability, “study design” included: study phase, study design, sample size, population, study duration, comparator treatment, and intervention. Table 4.5 depicts the inclusion/ex-clusion criteria of the four studies included in the assessment. Table 4.6 shows the outcome definitions used by the different studies. Table 4.7 shows the patient baseline characteristics in the active SPMS population for the studies identified by the SLR. Data was only available for EXPAND. Table 4.8 pre-sents patient baseline characteristics in the global SPMS population for studies included in the analyses of the relapsing SPMS population.



Table 4.4. Design of the studies in the relapsing SPMS population

Study ID Author (year) Phase Study design Sample size relapsing/ total (%)

MS popu-lation

Study du-ration

Compara-tor

Treatment

EXPAND Kappos (2018) 3 Randomised, double-blind,

parallel group 590/1651

(35.7) SPMS 3 years

Placebo PO each

day

Siponimod 2 mg PO each day

Nordic SPMS study Andersen (2004) 3 Randomised, double-blind,

parallel group 217/371 (58.5) SPMS 3 years

Placebo SC once a

week

IFN-β-1a 22 µg SC once a week

SPECTRIMS

SPECTRIMS Study Group

(2001) Li (2001)

NA Randomised, double-blind,

parallel group 293/618 (47.4)

SPMS 3 years

Placebo SC three

times weekly

IFN-β-1a 22 µg SC three times weekly IFN-β-1a 44 µg SC three times weekly

European study European Study

Group (1998) Kappos (2001)

3 Randomised, double-blind,

parallel group 502/718 (69.9) SPMS 3 years*

Placebo SC every other day

IFN-β-1b 250 µg SC every other day

Source: Adapted from the technical reports provided by the MAH. * Early termination at month 33. Mean duration of follow-up was 1054 and 1068 days in the placebo and IFN-β-1b group, respectively. Abbreviations: IFN=interferon; PO=per os (oral); SC=subcutaneous; SPMS=secondary progressive multiple sclerosis.



Table 4.5. Inclusion/exclusion criteria of studies in the relapsing SPMS population

Study ID EXPAND Nordic SPMS Study SPECTRIMS European Study

MS Population SPMS SPMS SPMS SPMS

Baseline EDSS range 3.0-6.5 ≤6.5 3.0-6.5 3.0-6.5

Age range 18-60 18-65 18-55 18-55

Prior IFN therapy Allowed No prior IFN use No prior IFN use No prior IFN use

No relapses in X months prior 3 months 2 months 2 months 1 month

Recently documented progression For ≥6 months in past 24

months For ≥6 months in past 4

years For ≥6 months in past 2

years Progression in past 2 years or ≥2

relapses in past 2 years

History of RRMS Required Required Required Required

Duration of MS No restriction ≥1 year NR ≥2 years

Duration of SPMS No restriction NR NR NR

MS severity score No restriction NR NR NR

T25FW test No restriction NR NR NR

Source: Adapted from the technical reports provided by the MAH.

Abbreviations: IFN=interferon; MS=multiple sclerosis; RRMS=relapsing remitting multiple sclerosis; SPMS=secondary progressive multiple sclerosis; T25FW=Timed 25-foot Walk test.



Table 4.6. Outcome definitions of studies in the relapsing SPMS population

ARR Time to CDP-3 Time to CDP-6 Discontinuation

EXPAND Total number of relapses

per patient-years 1.0-point inc. in EDSS score: 3.0-5.0 0.5-point inc. in EDSS score: 5.5-6.5

1.0-point inc. in EDSS score: 3.0-5.0 0.5-point inc. in EDSS score: 5.5-6.5

The proportion of randomised patients who discontinued treatment for any reason

Nordic SPMS study

Total number of relapses per patient-years

NR 1.0-point inc. in EDSS score: 3.0-5.0 0.5-point inc. in EDSS score: 5.5-6.5


SPECTRIMS Total number of relapses

per patient-years 1.0-point inc. in EDSS score: 3.0-5.0 0.5-point inc. in EDSS score: 5.5-6.5

NR The proportion of randomised patients who

discontinued treatment for any reason

European study


1.0-point inc. in EDSS score: 3.0-5.5 0.5-point inc. in EDSS score: 6.0-6.5

NR The proportion of randomised patients who

discontinued treatment for any reason

Source: Adapted from the technical reports provided by the MAH. Abbreviations: CDP=confirmed disability progression; EDSS=Expanded Disability Status Scale; NR=not recorded.



Table 4.7. Patient baseline characteristics in the active population for studies included in the analyses of the relapsing SPMS population

Study ID EXPAND (active SPMS

group)

Nordic SPMS Study

SPECTRIMS European Study

Age (mean years) 46.2 NR NR NR

Proportion female (%) 64 NR NR NR

Mean EDSS score 5.4 NR NR NR

Proportion of patients with EDSS score ≥6.0 (%) 56 NR NR NR

Time since onset of MS symptoms (mean years) 15.6 NR NR NR

Duration of MS (mean years) 11.7 NR NR NR

Duration of SPMS (mean years) 3.2 NR NR NR

Normalised brain volume (mean cm3) 1421 NR NR NR

Proportion of patients with Gd+ lesions of T1-weighted images (%)

46.7 NR NR NR

Total volume of T2 lesions on T2-weighted images (mean mm3)

17,864 NR NR NR

Proportion of patients without previous use of a DMT (%)

23,7 NR NR NR

Mean Timed 25-Foot Walk Test (seconds) NR NR NR NR

Time since most recent relapse (months) 32 NR NR NR

Proportion of patients relapse-free in prior year (%) 43.5 NR NR NR

Proportion of patients relapse-free in prior 2 years (%)

75.4 NR NR NR

Proportion of patients with relapses in prior 2 years (%)

24.6 NR NR NR

Number of relapses per patient in the prior year (mean)

NR NR NR NR

Number of relapses per patient in the previous 2 years (mean)

NR NR NR NR

Source: Adapted from the technical reports provided by the MAH and submission dossier. * There is a lack of information on most baseline characteristics for the subgroup of interest (relapsing SPMS). Abbreviations: DMT=disease-modifying therapy; EDSS=Expanded Disability Status Scale; Gd=gadolinium; MS=multiple scle-

rosis; NR=not recorded; SPMS=secondary progressive multiple sclerosis.



Table 4.8. Patient baseline characteristics in the global SPMS population for studies included in the analyses of the relapsing SPMS population*

Study ID EXPAND (active SPMS

group)

Nordic SPMS Study

SPECTRIMS European Study

Age (mean years) 48.0 45.7 42.8 41.0

Proportion female (%) 60 60 63 61

Mean EDSS score 5.4 4.8 5.4 5.1

Proportion of patients with EDSS score ≥6.0 (%) 56 NR NR 45

Time since onset of MS symptoms (mean years) 16.8 NR NR NR

Duration of MS (mean years) 12.6 14.3 13.3 13.1

Duration of SPMS (mean years) 3.8 5.4 4.0 2.2

Normalised brain volume (mean cm3) 1423 NR NR NR

Proportion of patients with Gd+ lesions of T1-weighted images (%)

21 NR NR NR

Total volume of T2 lesions on T2-weighted images (mean mm3)

15,322 NR NR NR

Proportion of patients without previous use of a DMT (%)

22 NR NR NR

Mean Timed 25-Foot Walk Test (seconds) 16.7 NR NR NR

Time since most recent relapse (months) 59 NR NR NR

Proportion of patients relapse-free in prior year (%) 78 NR NR NR

Proportion of patients relapse-free in prior 2 years (%)

64 NR 53 30

Proportion of patients with relapses in prior 2 years (%)

36 NR 47 70

Number of relapses per patient in the prior year (mean)

0.2 NR NR NR

Number of relapses per patient in the previous 2 years (mean)

0.7 NR 0.9 NR

Source: Adapted from the technical reports provided by the MAH and submission dossier.

* There is a lack of information on most baseline characteristics for the subgroup of interest (relapsing SPMS). Abbreviations: DMT=disease-modifying therapy; EDSS=Expanded Disability Status Scale; Gd=gadolinium; MS=multiple scle-rosis; NR=not recorded; SPMS=secondary progressive multiple sclerosis.

4.5 Outcomes included

In the relapsing SPMS population, it was only possible to compare:

Siponimod vs interferon-β-1a 44 µg three times weekly, SC (SPECTRIMS) for the outcome time to CDP-3;

Siponimod vs interferon-β-1a 22 µg once a week, SC (Nordic SPMS Study) for the outcome time to CDP-6;

Siponimod vs interferon-β-1b 250 µg every other day, SC (European Study) for the outcome pro-portion of patients with CDP-3 at 33 months;

Siponimod vs interferon-β-1a 22 µg three times weekly or 44 µg three times weekly, SC (SPEC-TRIMS) for the outcome clinical relapses.



Due to the absence of outcome data in comparator studies, it was not possible to compare the treatment effects of siponimod vs comparators for any of the following outcomes: MS symptoms, mortality, HRQoL, MRI inflammatory disease activity, no evidence of disease activity, adverse events, serious adverse events, adverse events leading to treatment discontinuation, or treatment-related mortality. In the ancillary analyses (APPENDIX 6: Ancillary analysis) in the global SPMS population, it was only possible to compare:

Siponimod vs interferon-β-1a 44 µg three times weekly, SC (SPECTRIMS) for the outcome time to CDP-3;

Siponimod vs interferon-β-1a 22 µg three times weekly or 44 µg three times weekly, SC (SPEC-TRIMS) for the outcome clinical relapses;

Siponimod vs interferon-β-1a 22 µg once a week, SC (Nordic SPMS Study) for the outcomes time to CDP-6 and ARR;

Siponimod vs interferon-β-1b 250 µg every other day, SC for the outcomes CDP-3 (European Study), CDP-6 (North American Study), and ARR;

Siponimod vs natalizumab 300 mg q4w (ASCEND) for the outcomes proportion of patients with CDP-6 and ARR;

Siponimod vs interferon-β-1a 60 µg once a week, SC (IMPACT) for the outcomes proportion of patients with CDP-3 and ARR.

4.6 Risk of bias/certainty of the evidence

The Cochrane Collaboration recommends the use of six domains for assessing risk of bias in each study included in the assessment [84]. This includes a judgement for each entry in a risk of bias table, which involves answering a question with ‘Yes‘ indicating low risk of bias, ‘No‘ indicating high risk of bias, or ‘Unclear‘ indicating either uncertainty over the potential bias or insufficient information. The risk of bias in each of the included studies is shown in the Table 4.9.

The sections below detail the classification of the certainty of the evidence by outcome for each com-parison using the GRADE methodology but, with the exception of the domain “sequence generation”, includes the explanatory information for the classification of the risk of bias in each individual study.

The explanatory information for the classification of the risk of bias in each study for the domain “se-quence generation” is as follows:

EXPAND study: A randomisation list was produced by the Interactive Response Technology (IRT) provider using a validated system that automated the random assignment of patient numbers to ran-domisation numbers. Low risk of bias.

Nordic SPMS Study: Patients were randomised, no further details. Unclear risk of bias.

European Study: A central randomisation schedule assigned placebo or interferon-β-1b to blocks of six patients in a 1/1 ratio. Unclear risk of bias.

SPECTRIMS: Treatment was assigned using a computer-generated randomisation list provided by Ser-ono, stratified by centre; treatments were equally allocated with a block size of six. Low risk of bias.



Table 4.9. Risk of bias of included studies

Studies

Seq

uen

ce g

en

era

tion

Allo

catio

n c

on

ceal-

men

t

Blin

din

g o

f partic

i-

pan

ts (p

erfo

rman

ce

bia

s)

Blin

din

g o

f ou

tco

me

assessm

en

t

(dete

ctio

n b

ias)*

Blin

din

g o

f ou

tco

me

assessm

en

t (dete

c-

tion

bia

s)**

Inco

mp

lete

ou

tco

me

data

Sele

ctiv

e o

utc

om

e

rep

ortin

g

Oth

er p

ote

ntia

l

so

urc

es o

f bia

s

EXPAND Low Low Low Unclear Unclear High Low Low

Nordic SPMS Study Unclear Unclear Unclear Unclear NA Low Low High

European Study Unclear Low Unclear Unclear Unclear Low Low High

SPECTRIMS Study Low Unclear Unclear Unclear Unclear Low Low Low

* For investigator-evaluated outcomes. ** For EDSS-rater evaluated outcomes. Abbreviations: NA=not applicable.

Comparison of siponimod vs interferon-β-1a 22 µg once a week

The comparison of siponimod vs interferon-β-1a 22 µg once a week included two studies (EXPAND and the Nordic SPMS Study).

Outcome: Confirmed disability progression at six months (time to CDP-6)

The quality of evidence for the outcome CDP-6 was rated as very low. A summary of the assessment of the certainty of the evidence is depicted in Table 4.10.

Risk of bias

Allocation concealment

EXPAND: Patients were randomly assigned in a 2:1 ratio via IRT to receive siponimod or matching placebo. Thus, allocation was adequately concealed. Risk was considered to be low.

Nordic SPMS study: The published report of the study provides no information on the method of patient allocation to treatment arms. Therefore, whether allocation was properly concealed in the Nordic SPMS Study is unclear.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with allocation concealment was not serious, and it was therefore decided not to rate down the evidence quality for this reason.

Blinding

EXPAND: This was a double-blind study. Study drug and placebo were identical in appearance, taste, and odour. The risk of bias for participants (blinding of participants) was therefore considered to be low. However, blinding procedures were potentially not followed for 11.3% of patients in the siponimod group and 11.85 in the placebo group. The potential unblinding resulted from temporary incorrect access rights for site staff (EDSS raters and investigators) to the different databases, leading to potential unblinding of some cases. While sensitivity analyses, excluding patients affected by potential unblinding, were reassuring, the potential for bias cannot be completely excluded.

Nordic SPMS Study: This was a double-blind study. However, interferon side-effects are well recog-nised, raising concerns about the possibility of unblinding of participants. Therefore, the blinding of participants was considered unclear. Regarding the blinding of investigators, the only information given



in the published report was that patients were instructed to cover injection sites and to discuss only neurological matters during neurological evaluations. Thus, whether blinding of investigators was ade-quate in the Nordic SPMS Study is unclear. Of note, the investigators, instead of independent EDSS raters, performed EDSS evaluations. No blinding questionnaire was conducted at study termination. Therefore, the risk of bias was considered unclear for the outcome assessors (investigators).

Comparison: Given the pooled information on both studies in the comparison, it was considered that there was a serious risk of bias associated with the domain ‘blinding’. Therefore, it was decided to rate down the evidence quality by one level due to lack of blinding.

Incomplete outcome data

EXPAND: In this study, efficacy analyses were conducted in the Full Analysis Set, which was defined as all randomised patients who took at least one dose of study medication. Patients were analysed according to treatment assignment (i.e., irrespective of the treatment received).

No data were available on the number of patients in the relapsing population who discontinued prema-turely from the study. In the global SPMS population, in the siponimod group, six patients were excluded from the efficacy analyses for the following reasons: five patients never took the study medication, and one patient did not provide signed informed consent. Only 737/1105 patients (66.7%) in the siponimod group and 322/546 patients (59.0%) in the placebo group completed study participation on the double-blind study drug. The main reasons for premature discontinuation from the double-blind study drug were: subject decision (10.3% vs 13.0%), disease progression (9.1% vs 14.8%), and adverse events (8.5% vs 5.1%) for the siponimod and placebo groups, respectively. Discontinuation due to lack of efficacy was more common in placebo patients (4.9%) than siponimod patients (3.3%). In the EXPAND study, in the global SPMS population, 228/1096 six-month CDP events (19.9%) were observed in the siponimod group, and 139/545 six-month CDP events (25.5%) were observed in placebo group. Alt-hough the reasons for discontinuation seem balanced between treatment arms in EXPAND, the high rate of premature discontinuation, which was greater than the proportion of patients with CDP-6 events, was a matter of concern. In the relapsing population, CDP-6 events were observed in 74/388 patients (19.1%) in the siponimod group and in 58/202 patients (28.7%) in the placebo group. The high rate of premature discontinuation observed in the relapsing population was therefore also a matter of concern. Therefore, the risk of bias due to incomplete outcome data was considered to be high.

Nordic SPMS Study: In this study, 153/178 patients (86.0%) in the placebo arm and 148/186 patients (79.6%) in the IFN arm completed the double-blind phase of the study. Reasons for discontinuation of the study drug included patient decision (placebo: 13 patients; IFN: 17 patients), adverse events (pla-cebo: 6 patients; IFN: 16 patients), disease progression (placebo: 2 patients; IFN: 3 patients), death (placebo: 2 patients; IFN: 2 patients), protocol violations (placebo: 1 patient; IFN: 0 patients), and preg-nancy (placebo: 1 patient; IFN: 0 patients). In the global SPMS population, confirmed disability progres-sion at six months was observed in 77/186 patients (41%) in the IFN group and 68/178 patients (38%) in the placebo group. In the Nordic SPMS Study, in the relapsing population, CDP at six months was observed in 47/111 patients (42%) in the IFN group and 42/106 patients (40%) in the placebo group. The reasons for discontinuation seem well balanced between treatment arms, and the rate of premature discontinuation was lower than the proportion of patients with CDP-6 events. Therefore, the risk of bias due to incomplete outcome data was considered to be low.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with the domain ‘incomplete outcome data’ was not serious. Thus, it was decided not to down-rate the evidence quality for this reason.

Selective outcome reporting

EXPAND: The study protocol was available, and all of the study’s pre-specified outcomes were re-ported. Selective outcome reporting was not an issue and therefore the risk of bias due to selective outcome reporting was considered to be low.

Nordic SPMS study: The study protocol was not available but the published report included all expected outcomes and therefore the risk of bias due to selective outcome reporting was considered to be low.



Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with selective outcome reporting was not serious, and therefore it was de-cided not to rate down the evidence quality for this reason.

Other limitations

EXPAND: This study did not have other limitations that might increase the risk of bias. The risk of bias due to other limitations was considered to be low.

Nordic SPMS study: This study was terminated early following the release of the SPECTRIMS study results, which showed no benefit of interferon-β-1a 22 µg or 44 µg three times weekly on disability. This fact may have overestimated the reported treatment effect. The risk of bias due to other limitations was considered to be high.

Comparison: None of the studies reported the baseline characteristics of the relapsing subgroups. Thus, it was not possible to assess the comparability of the relapsing populations between studies. Due to the uncertainty around this issue, we down-rated the quality of the evidence by one level.

Imprecision

GRADE methodology recommends the combined use of 95% confidence intervals (CI) of treatment effect estimates and calculation of “optimal information size” (OIS). The studies (EXPAND and Nordic SPMS Study) were not powered to detect a difference in CDP between treatment groups in the relaps-ing population, and therefore these results are imprecise. Taking into account the width of the 95% CI and the IOS criteria, these results are imprecise. Thus, it was decided to down-rate the evidence quality for imprecision.

Inconsistency

This comparison included only two studies and, therefore, this criterion was not applicable.

Indirectness

The estimate of treatment effect on CDP-6 was the result of an anchored indirect comparison. There is no information on the baseline patient characteristics of the relapsing population. Although it was not possible to assess the comparability of patient characteristics in the relapsing populations between groups, it was decided not to down-rate the evidence quality for indirectness, as it was already rated down for this in ‘other limitations’.

Other outcomes

It was not possible to compare treatment effect of siponimod vs interferon-β-1a 22 µg once a week for any other outcome.



Table 4.10. Evidence profile [siponimod vs interferon-β-1a 22 µg once a week, SC (Nordic SPMS Study)]

Risk of bias

Outcomes Allocation

concealment

Blinding Incomplete

outcome

data

Selective

outcome

reporting

Others Indirect-

ness

Impreci-

sion

Incon-

sistency

Quality of

evidencea

No studies

Confirmed disability pro-

gression (time to CDP-6)

No Yes No No Yes No Yes NA Very low 2

Other measures of disabil-

ity Progression

NR NR NR NR NR NR NR NR NA 0

Clinical relapse NR NR NR NR NR NR NR NR NA 0

MS symptoms NR NR NR NR NR NR NR NR NA 0

Mortality NR NR NR NR NR NR NR NR NA 0

Health-related quality of life NR NR NR NR NR NR NR NR NA 0

MRI inflammatory disease

activity


No evidence of disease ac-

tivity


Adverse events NR NR NR NR NR NR NR NR NA 0

Serious adverse events NR NR NR NR NR NR NR NR NA 0

Discontinuation due to ad-

verse events


Treatment-related mortality NR NR NR NR NR NR NR NR NA 0 a According to the GRADE methodology Abbreviations: CDP=confirmed disability progression; MRI-magnetic resonance imaging; MS=multiple sclerosis; NR=not reported; NA=not applicable.



Comparison of siponimod vs interferon-β-1b 250 µg every other day

Outcome: Confirmed disability progression at 3 months (proportion of patients with CDP-3)

The quality of evidence for the outcome CDP-3 was rated as very low. A summary of the assessment of the certainty of the evidence is presented in Table 4.11.

Risk of bias


EXPAND: As detailed above, the risk of bias due to allocation concealment was considered to be low.

European Study: The published report states that a central randomisation schedule assigned placebo or interferon-β-1b to blocks of six patients in a 1:1 ratio. Therefore, allocation seems adequately con-cealed. Accordingly, there were no concerns over the lack of allocation concealment.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with allocation concealment was not serious, and therefore it was decided not to down-rate the evidence quality for this reason.

Blinding

EXPAND: As detailed above, the risk of bias due to blinding was considered to be unclear.

European Study: This was a double-blind study. However, interferon side-effects are well recognised. Therefore, the risk of bias associated with potential unblinding of participants was considered unclear. Regarding investigator blinding, the only information given in the published report was that patients were instructed to cover injection sites. Additionally, to avoid unmasking as a result of the interferon-β-1b side effects, EDSS raters were not supposed to receive any potentially unmasking information from treating physicians. At the end of the study, a questionnaire was administered to patients, treating phy-sicians, and EDSS raters. This questionnaire was received from 84% to 86% of patients and study personnel. Treating physicians often guessed correctly whether patients were on placebo (48.4%) or interferon-β-1b (56.2%), and did not know or guessed incorrectly in only 36.3% of patients in the placebo group and 11.3% in the interferon-β-1b group. Likewise, patients often guessed correctly whether they were on placebo (54.3%) or interferon-β-1b (65.2%) and guessed incorrectly in only 23.4% of placebo cases and 11.7% of interferon-β-1b cases. Of note, EDSS raters guessed 18.6% of patients correctly in the placebo group and 20.8% of the patients in the interferon-β-1b group. They stated ‘do not know’ for 66.5% of patients. Thus, whether blinding of outcome assessors (investigators and EDSS raters) was adequate was unclear.

Comparison: Given the pooled information on both studies in the comparison, it was considered that there was a serious risk of bias due to lack of adequate blinding. Therefore, it was decided to down-rate the evidence quality by one level due to the likely lack of blinding.


EXPAND: As stated above, the risk of bias was considered to be high.

European Study: In this study, 270/360 patients (75.0%) in the IFN arm and 261/358 patients (72.9%) in the placebo arm completed the double-blind phase of the study. Reasons for discontinuation of study drug included adverse events (placebo: 4 patients; IFN: 5 patients), disease progression (placebo: 10 patients; IFN: 5 patients), death (placebo: 1 patient; IFN: 3 patients), lost to follow-up (placebo: 4 pa-tients; IFN: 8 patients), and others (placebo: 12 patients; IFN: 5 patients). In the global SPMS popula-tion, confirmed disability progression at three months was observed in 140/360 patients (38.9%) in the IFN group and 178/358 patients (49.8%) in the placebo group. In the European Study, in the relapsing population, confirmed disability progression at three months was observed in 94/245 patients (38.4%) in the IFN group and 128/257 patients (49.8%) in the placebo group. The reasons for discontinuation seem balanced between treatment arms, and the rate of premature discontinuation was lower than the



proportion of patients with CDP-3 events. Therefore, the risk of bias due to incomplete outcome data was considered to be low.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with incomplete outcome data was not serious. Therefore, it was decided not to rate down the evidence quality for this reason.


EXPAND: As stated above, the risk of bias was considered to be low.

European Study: The study protocol was not available, but the published report included all expected outcomes. Therefore, the risk of bias due to selective outcome reporting was considered to be low.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with selective outcome reporting was not serious. Therefore, it was decided not to down-rate the evidence quality for this reason.

Other limitations

EXPAND: As stated above, this study did not have other limitations that might increase the risk of bias. The risk of bias due to other limitations was considered to be low.

European Study: This study was stopped early for benefit at 33 months, increasing the risk of overesti-mating the treatment effect. The risk of bias due to other limitations was considered to be high.

Comparison: As none of the studies reported the baseline characteristics of the relapsing subgroups, it was not possible to assess the comparability of the relapsing populations between studies. Thus, be-cause of uncertainty around this issue, we down-rated the quality of the evidence by one level.

Imprecision

GRADE methodology recommends the combined use of 95% CI of treatment effect estimates and cal-culation of OIS. Siponimod in comparison with interferon-β-1b 250 µg every other day SC was associ-ated with a risk reduction of 19% for CDP-3 (hazard ratio vs relative risk 0.81; 95%CI 0.57 to 1.15). Of note, in the relapsing population, two different effect sizes were compared between EXPAND (hazard ratio) and European Study (relative risk). The studies (EXPAND and European Study) were not pow-ered to detect a difference in CDP between treatment groups in the relapsing population. Taking into account the width of the 95% CI and the IOS criteria, these results were imprecise. Thus, it was decided to down-rate the evidence quality for imprecision.

Inconsistency

This comparison includes only two studies and, therefore, this criterion was not applicable.

Indirectness

The estimate of treatment effect on CDP-3 was the result of an anchored indirect comparison. There is no information on the baseline patient characteristics of the relapsing population. Although it was not possible to assess the comparability of patient characteristics in the relapsing populations between groups, it was decided not to down-rate the evidence quality for indirectness, as it was already rated down for this in “other limitations”.

Other outcomes

It was not possible to compare treatment effect of siponimod vs interferon-β-1b 250 µg every other day SC for any other outcomes.



Table 4.11. Evidence profile [siponimod vs interferon-β-1b 250 µg every other day, SC (European Study)]

Risk of bias

Outcomes Allocation

concealment

Blinding Incomplete

outcome

data

Selective

outcome

reporting

Others Indirect-

ness

Impreci-

sion

Incon-

sistency

Quality of

evidencea

No studies


gression at 6 months



gression (proportion of pa-

tients with CDP-3)


Clinical relapse NR NR NR NR NR NR NR NR NA 0





activity



tivity





verse events


Treatment-related mortality NR NR NR NR NR NR NR NR NA 0 a According to the GRADE methodology

Abbreviations: CDP=confirmed disability progression; MRI-magnetic resonance imaging; MS=multiple sclerosis; NR=not reported; NA=not applicable.



Comparison of siponimod vs interferon-β-1a 22 or 44 µg three times weekly

The comparison of siponimod vs interferon-β-1a 22 or 44 µg three times weekly SC included two studies (EXPAND and SPECTRIMS). The overall quality of evidence was rated as very low. A summary of the assessment of the certainty of the evidence is presented in Table 4.12.

Outcome: Confirmed disability progression at three months (time to CDP-3)

The quality of evidence for the outcome CDP-3 was rated as very low.

Risk of bias


EXPAND: As detailed above, the risk of bias due to allocation concealment was considered to be low.

SPECTRIMS: The published report of the study stated that treatment was assigned using a computer-generated list provided by a third party, patients were allocated with a block size of 6, and that the block size was not revealed to the investigators. This information raised concerns about allocation conceal-ment, because if investigators had access to the randomisation list (and it is not clear that they did not), they would have known the allocation of the next randomised patient. Therefore, it was unclear whether allocation was adequately concealed.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with allocation concealment was not serious, and therefore it was decided not to down-rate the evidence quality for this reason.

Blinding

EXPAND: As stated above, the risk of bias was considered unclear for the outcome assessors (inves-tigators and EDSS raters).

SPECTRIMS: This was a double-blind study. However, interferon side-effects are well recognised. Re-garding blinding of investigators, the information given by the published report is that during EDSS assessments, all potential injection sites were covered. Additionally, to avoid unmasking as a result of the interferon-β-1a side effects, disability assessment was performed by EDSS raters and clinical and neurological data were recorded in separate binders. Solutions of IFN and placebo were physically indistinguishable, and the containers of the study medication were labelled with patients’ identification numbers assigned at the time of randomisation. At the end of the study, a questionnaire was adminis-tered to patients and evaluating physicians to determine the success of blinding. Fifty-nine percent of patients on active treatment and 52% on placebo correctly guessed their treatment. In the high-dose group, 64% of patients guessed correctly, whereas in the low-dose group, 54% of patients guessed correctly. Patients were incorrect in 21% of treatment cases and 27% of placebo cases. Evaluating physicians (there were no separate data for investigators and outcome assessors) guessed treatment assignments correctly for 29% of patients in the interferon group and 26% of patients in the placebo group. Treating physicians were incorrect for 16% of patients in the interferon group and 18% of patients in the placebo group, and gave no opinion for the remainder. As it is unlikely that so many investigators had no opinion (could not guess) on which arm patients were allocated, the risk of bias associated with potential unblinding of participants was considered unclear. Also, whether outcome assessment (inves-tigators and EDSS raters) blinding was adequately maintained or whether it might have been broken due to IFN-associated adverse events is unclear, but it is likely that it was broken.

Comparison: Given the pooled information on both studies in the comparison, it was considered that there was a serious risk of bias due to lack of adequate blinding. Therefore, it was decided to down-rate the evidence quality by one level due to the likely lack of blinding.




EXPAND: As detailed above, the risk of bias was considered to be high.

SPECTRIMS: In this study, 173/205 patients (84.4%) in the placebo arm, 172/209 patients (82.3%) in the IFN 22 µg arm, and 161/204 patients (78.9%) in the IFN 44 µg arm completed the double-blind phase of the study. Reasons for discontinuation of the study drug included adverse events (placebo: 4 patients; IFN: 5 patients), disease progression (placebo: 9 patients; IFN low-dose: 7 patients; IFN high-dose: 4 patients), death (placebo: 1 patients; IFN low-dose: 1 patient; IFN high-dose: 2 patients), lost to follow-up (placebo: 19 patients; IFN low-dose: 14 patients; IFN high-dose: 14 patients), protocol de-viation (placebo: 0 patients; IFN low-dose: 1 patient; IFN high-dose: 3 patients), and patient decision (placebo: 17 patients; IFN low-dose: 12 patients; IFN high-dose: 15 patients). In the SPECTRIMS study (interferon-β-1a 44 µg three times weekly SC), either in the global or in the relapsing population, the number of CDP-3 events by treatment group was not reported. However, taking into account the pro-portion of patients with CDP-3 events observed in other IFN studies, it is plausible that the rate of premature discontinuation was small and therefore that incomplete outcome data was not an issue. Therefore, the risk of bias due to incomplete outcome data was considered to be low.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with incomplete outcome data was not serious. Therefore, it was decided not to down-rate the evidence quality for this reason.


EXPAND: As stated above, the risk of bias was considered to be low.

SPECTRIMS: The study protocol was not available, but the published report included all expected out-comes and therefore the risk of bias due to selective outcome reporting was considered to be low.

Comparison: Given the pooled information on both studies in the comparison, it was considered that the risk of bias associated with selective outcome reporting was not serious. Therefore, it was decided not to down-rate the evidence quality for this reason.

Other limitations

As none of the studies reported baseline patient characteristics of the relapsing subgroups, it was not possible to assess the comparability of the relapsing populations between studies.

Comparison: Because of the uncertainty around this issue, we down-rated the quality of the evidence by one level.

Imprecision

GRADE methodology recommends the combined use of 95% CI of treatment effect estimates and cal-culation of OIS for defining the presence of imprecision. Siponimod in comparison with interferon-β-1a 44 µg three times weekly reduced the risk of CDP-3 events by 12% (HR 0.88; 95%CI 0.55 to 1.42), but the difference did not reach statistical significance. The studies (EXPAND and SPECTRIMS) were not powered to detect a difference in CDP between treatment groups in the relapsing population. Taking into account the width of the 95% CI and the IOS criteria, these results were imprecise. Thus, it was decided to down-rate the evidence quality for imprecision.

Inconsistency

This comparison included only two studies and, therefore, this criterion was not applicable.

Indirectness

The estimate of treatment effect on CDP-3 was the result of an anchored indirect comparison. There is no information on the baseline patient characteristics of the relapsing population. Although it was not possible to assess the comparability of patient characteristics in the relapsing populations between



groups, it was decided not to rate down the evidence quality for indirectness as it was already rated down for this in “other limitations”.

Outcome: Annualised relapse rate (ARR)

The quality of evidence for the outcome ARR was rated as very low.

Risk of bias


Please see risk of bias for the outcome time to CDP-3.

Blinding

Please see blinding for the outcome time to CDP-3.


Please see incomplete outcome data for the outcome time to CDP-3.


For the outcome ARR, selective outcome reporting was not an issue, either in the EXPAND or SPEC-TRIMS studies.

Other limitations

Please see other limitations for the outcome time to CDP-3.

Imprecision

GRADE methodology recommends the combined use of 95% CI of treatment effect estimates and cal-culation of OIS for defining the presence of imprecision. The studies (EXPAND and SPECTRIMS) were not powered to detect a difference in ARR between treatment groups. Therefore, the criteria for OIS was not met in both trials. Furthermore, siponimod in comparison with interferon-β-1a 22 µg three times weekly increased the risk of ARR by 10% (HR 1.10; 95%CI 0.65 to 1.87), but the difference did not reach statistical significance. Additionally, siponimod, in comparison with interferon-β-1a 44 µg three times weekly, decreased the risk of ARR by 6% (HR 0.94; 95%CI 0.59 to 1.49), but the difference did not reach statistical significance. Taking into account the width of the 95% CI and the IOS criteria, these results were imprecise. Thus, it was decided to down-rate the evidence quality for imprecision.

Inconsistency

This comparison includes only two studies and, therefore, this criterion was not applicable.

Indirectness

The estimate of treatment effect on ARR was the result of an anchored indirect comparison. There is no information on the baseline patient characteristics of the relapsing populations. Although it was not possible to assess the comparability of patient characteristics in the relapsing populations between groups, it was decided not to down-rate the evidence quality for indirectness as it was already down-rated for this in “other limitations”.

Other outcomes

It was not possible to compare treatment effect of siponimod vs interferon-β-1a 22 µg three times weekly or interferon-β-1a 44 µg three times weekly for any other outcome.



Table 4.12. Evidence profile [siponimod vs interferon-β-1a 22 µg or 44 µg three times weekly SC (SPECTRIMS Study)]

Risk of bias

Outcomes Allocation

concealment

Blinding Incomplete

outcome

data

Selective

outcome

reporting

Others Indirect-

ness

Impreci-

sion

Incon-

sistency

Quality of

evidencea

No studies


gression at 6 months



gression (time to CDP-3)


Clinical relapse No Yes No No Yes No Yes NA Very low 2





activity



tivity





verse events


Treatment-related mortality NR NR NR NR NR NR NR NR NA 0 a According to the GRADE methodology Abbreviations: CDP=confirmed disability progression; MRI-magnetic resonance imaging; MS=multiple sclerosis; NR=not reported; NA=not applicable.



4.7 External validity

The external validity of the included trials was assessed using EUnetHTA guidelines [85] on applicability of evidence in the context of a relative effectiveness assessment of pharmaceuticals considering the following elements: population, intervention, comparator, outcomes, and setting (PICO(S)).

Population

The randomised population in the siponimod study (EXPAND) constituted a global SPMS population, i.e., active and inactive SPMS. The population of interest for this assessment was, however, ”active SPMS“, which aligned with the marketing authorisation for siponimod. Efficacy results from the com-parison of siponimod versus the comparators of interest were based on a post hoc subgroup analysis, for which the study was not powered. These problems with internal validity may have influenced the external validity of the ITCs. Therefore, the conclusions derived from the indirect comparisons may not apply to all patients with active SPMS.

The EXPAND study did not include patients older than 61 years. Overall, the patients participating in the trials included in this assessment were generally younger and more likely to have active disease than most patients with progressive MS seen in clinical practice. Additionally, an older patient with less active progressive disease may be less likely to respond to a DMT and be more prone to develop infections and other complications of immunosuppressive therapy. Therefore, the conclusions of this assessment may not be applicable to older patients with progressive MS seen in clinical practice.

The studies included in the analysis for this assessment included a “relapsing SPMS” population due to the absence of data on the population of interest (“active” SPMS) from comparator studies. The MAH reported results of a post hoc subgroup analysis of EXPAND, which showed that the active and relaps-ing subgroup were similar with regards to the outcomes time to three-month and six-month CDP com-pared to placebo (see Section 3.4). Although similar data were not available for the comparator studies, it seemed reasonable to consider that comparisons between the relapsing SPMS sub-populations of EXPAND and comparator trials may be a reasonable indicator of relative effectiveness in the active SPMS population. Of note, the definition of the relapsing population varied across studies: in the Nordic SPMS Study (interferon-β-1a 22 µg once a week), the relapsing subpopulation was defined as patients with relapses during the four years prior to the start of the study, while in the SPECTRIMS (interferon-β-1a 22 µg three times weekly or 44 µg three times weekly) and the European Study (interferon-β-1b 250 µg every other day), the relapsing population was defined as patients with relapses during the two years prior to study commencement. These different definitions hampered the interpretation of the re-sults and, therefore, to which populations they apply.

Interventions

The dose of interferon-β-1a used in the Nordic Study (22 µg once weekly) is not currently used in clinical practice in most countries. The MAH decided to exclude the Nordic SPMS Study from the main analysis due to the use of an unlicensed interferon-β-1a regimen. Although the study was conducted before 2004 and it is unlikely that such a low dose is still being used, considering that the low-dose interferon-β-1a regimen could be used in clinical practice in some countries, it was decided to include the Nordic SPMS Study in the analyses. The indirect comparisons technical report also included the Nordic SPMS study. However, we note that the treatment effect of interferon observed in these analyses applied to this specific dose of interferon-β-1a and may therefore not apply to many of the patients with relapsing SPMS receiving a higher dose.

Comparators

The comparator included in the siponimod study (EXPAND) (i.e., best supportive care) was not a rele-vant comparator for this assessment. As such, indirect comparisons were required to assess the relative effectiveness of siponimod versus comparators. However, due to an absence of data on comparators in the population of interest, most of the comparators identified in the PICO are not included in this assessment.



Outcomes

For comparison with placebo, disability endpoints (CDP) were reported in the comparator studies with confirmation at three months (SPECTRIMS and European Study) and at six months (Nordic Study). Thus, these disability outcomes were available for the indirect comparisons. Of note, six-month CDP is usually considered a more relevant measure to assess treatment effect and, therefore, the results of treatment effect on six-month CDP (Nordic Study) would have the highest external validity. However, the interpretation of the results of the Nordic Study were hampered by the low-dose regimen used in the study.

4.8 Results on clinical effectiveness and safety

Results presented in this section are based on the “relapsing SPMS” subgroups derived from the EX-PAND IPD to match the subgroup and outcome definitions stated in the comparator trials (Table 4.2).

Effectiveness of siponimod vs comparators of interest in patients with “active” SPMS

A summary of the indirect comparisons performed in the relapsing population is presented in Table 4.13.

Comparison of siponimod vs interferon-β-1a 22 µg once a week

According to the CSR of the EXPAND study, in the relapsing population, CDP-6 events were observed in 74/388 patients (19.1%) in the siponimod group and in 58/202 patients (28.7%) in the placebo group (HR 0.65; 95%CI 0.44 to 0.89). However, in the MAIC Report (Table D.2), this hazard ratio was reported as 0.71 (95%CI 0.54 to 0.93). In the Nordic SPMS Study relapsing population, CDP-6 events were observed in 47/111 patients (42%) in the IFN-β-1a 22 µg once a week SC group and in 42/106 patients (40%) in the placebo group (HR 1.01; 95%CI 0.68 to 1.56).

The indirect comparison using Bucher ITC including EXPAND and the Nordic SPMS Study showed no statistically significant difference between siponimod and interferon-β-1a 22 µg once a week SC (HR 0.70; 95%CI 0.43 to 1.15). Note that the hazard ratio estimate of 0.71 was used in the ITC, as shown in the MAIC report Table D.2.

Comparison of siponimod vs interferon-β-1b 250 µg every other day

In the EXPAND study relapsing population, CDP-3 events were observed in 98/388 patients (19.1%) in siponimod group and in 72/202 patients (35.6%) in the placebo group (HR 0.67; 95%CI 0.49 to 0.91). In the European Study relapsing population, CDP-3 events were observed in 109/245 patients (44.5%) in interferon-β-1b 250 µg every other day SC group and in 132/257 patients (53.7%) in the placebo group (RR 0.83; 95%CI 0.69 to 0.99).

The indirect comparison using Bucher ITC including EXPAND and the European Study showed no statistically significant difference between siponimod and interferon-β-1b 250 µg every other day (HR vs RR 0.81; 95%CI 0.57 to 1.15).

Comparison of siponimod vs interferon-β-1a 22 or 44 µg three times weekly

In the EXPAND study relapsing population, CDP-3 events were observed in 98/388 patients (19.1%) in siponimod group and in 72/202 patients (35.6%) in the placebo group (HR 0.67; 95%CI 0.49 to 0.91). In the SPECTRIMS study (interferon-β-1a 44 µg three times weekly SC) relapsing population, the num-ber of CDP-3 events by treatment group was not reported (HR 0.76; 95%CI 0.53 to 1.10).

The indirect comparison using Bucher ITC including EXPAND and SPECTRIMS showed no statistically significant difference between siponimod and interferon-β-1a 44 µg three times weekly (HR 0.88; 95%CI 0.55 to 1.42).

In the EXPAND study relapsing population, ARR events were observed in 85/388 patients (0.139; 95%CI 0.108 to 0.78) in the siponimod group and in 76/202 patients (0.240; CI not reported) in the



placebo group (rate reduction 42%; ARR ratio 0.579; 95%CI 0.399 to 0.839; nominal p=0.0039). In the EXPAND relapsing population, the relative risk for ARR was 0.58 (95%CI 0.40 to 0.84). In the SPEC-TRIMS relapsing population treated with interferon-β-1a 22 µg three times weekly SC, there was a 47% decrease in the risk of ARR (RR 0.53; 95%CI 0.36 to 0.77), and in the relapsing population treated with interferon-β-1a 44 µg three times weekly SC, there was a 38% decrease in the risk of ARR (RR 0.62; 95%CI 0.47 to 0.82).

The indirect comparison using Bucher ITC including EXPAND and SPECTRIMS showed no statistically significant difference between siponimod and interferon-β-1a 22 µg three times weekly (HR 1.10; 95%CI 0.65 to 1.87). Siponimod in comparison with interferon-β-1a 44 µg three times weekly also showed no statistically significant difference (HR 0.94; 95%CI 0.59 to 1.49).

A summary of the indirect comparisons performed in the relapsing population is presented in Table 4.13.



Table 4.13. Summary of findings of siponimod for the treatment of adult patients with SPMS with relapsing disease

Comparator intervention

Regimen Study ID(s) Notable assumptions

Siponimod vs placebo subgroup from EXPAND IPD

(95% CI)

Comparator vs placebo subgroup from publica-

tion (95% CI)

Subgroup Bucher ITC results Siponimod vs comparator

(95% CI)

Type Value Type Value Type Value


Rebif® 22 µg once a week

Nordic SPMS Study - HR 0.71

HR 1.01

HR 0.70

(SC IFN-β-1a) (0.54 - 0.93) (0.68 - 1.56) (0.43 - 1.15)

Proportion with CDP-3 (33 months) in subgroup: patients with relapses in the 2 years before study

Betaferon®

(SC IFN-β-1b)


European Study Imputation of censored

data for EXPAND: last ob-

servation carried forward*

OR 0.61

OR 0.69

OR 0.88

(0.42 - 0.88) (0.49 - 0.98) (0.53 - 1.47)

Betaferon®

(SC IFN-β-1b)


European Study

Compared subgroup RR of European Study with subgroup time-to-event

HR of EXPAND**

HR 0.67

RR 0.83 RR compared to

HR**

0.81

(0.50 - 0.91) (0.69 - 0.99) (0.57 - 1.15)


Rebif® 44 µg three times weekly‡

SPECTRIMS - HR 0.67

HR 0.76

HR 0.88

(SC IFN-β-1a) (0.49 - 0.91) (0.53 - 1.10) (0.55 - 1.42)

ARR in subgroup: patients with relapses in the 2 years before study

Rebif® 22 µg three times weekly

SPECTRIMS SPECTRIMS: assumed

the rate ratio p-

value=0.001 in order to

calculate the 95% CI†

Rate ratio 0.58

Rate ratio 0.53

Rate ratio 1.10

(SC IFN-β-1a) (0.40 - 0.84) (0.36 - 0.77)† (0.65 - 1.87)

Rebif® 44 µg three times weekly

SPECTRIMS Rate ratio 0.58

Rate ratio 0.62

Rate ratio 0.94

(SC IFN-β-1a) (0.40 - 0.84) (0.47 - 0.82)† (0.59 - 1.49)

Source: Adapted from the technical reports provided by the MAH. Statistically significant values are in bold. p-values not reported by the MAH. * Last observation carried forward: censored patients are assumed to have the same status at the cut-off time-point as they did on the day they were censored. Note that complete case could not be

derived from the EXPAND IPD due to the volume of censoring (in complete case, censored patients are removed from the data). ** Must be interpreted with caution. † Analysis of the ARR outcome for the relapsing subgroup in SPECTRIMS required making an assumption about the p-value (reported as <0.001 :for both doses") in order to calculate the 95% CI,

which was not reported. The p-value was assumed to equal 0.001 for each dose. ‡ Time to CDP-3 subgroup results not reported for the 22 µg dose in SPECTRIMS. Abbreviations: CI=confidence interval; HR=hazard ratio; IFN=interferon; IPD=individual patient data; NR=not reported; OR=odds ratio; RR=relative risk; SC=subcutaneously.



Safety of siponimod vs comparators of interest

Comparison of siponimod vs comparators of interest in the active SPMS population

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on adverse events, serious adverse events, adverse events leading to treatment discontinuation, or treatment-related mortality. Therefore, it was not possible to evaluate how siponimod compared with the selected comparators in relation to effect on these out-comes in the population of interest (“active” SPMS).

Comparison of siponimod vs comparators of interest the global SPMS population

Due to the absence of data in comparator studies, for the safety outcomes in the population of interest, a naïve indirect comparison was performed in the global SPMS population. Though its relevance is limited, a table with the available information on safety included in EXPAND and comparator studies has been included (Table 4.14)

Comparison of siponimod vs placebo in the global SPMS population (EXPAND study)

As there were no relevant data on comparative safety between siponimod and comparators, the toxicity of siponimod can only be assessed using data from the EXPAND study (siponimod vs placebo; infor-mation available in APPENDIX 5: EXPAND (Study A2304)). Adverse events were more frequent in the siponimod than the placebo group (88.7% vs 81.5%). Serious adverse events were also reported more frequently in the siponimod than the placebo group (17.9% vs 15.2%). More patients discontinued from the study drug permanently due to adverse events in the siponimod group (7.6% vs 5.1%). Adverse events of special interest that were reported more frequently in the siponimod group were herpes zoster reactivations (2.2% vs 0.7%), lymphopenia (1.6% vs 0%), macular oedema (1.7% vs 0.2%), and in-creased liver transaminases (1.4% vs 0.6%). Globally, the safety information raised no specific safety concerns.

Other safety issues

Siponimod is mainly metabolised via CYP2C9. In a phase I study of siponimod, subjects with the CYP2C9*2*3 or CYP2C9*3*3 genotype showed a two- or four-fold increase, respectively, in the area under the curve (AUC) and T½ compared with CYP2C9*1*1. Likewise, the pharmacokinetic parameters for the metabolites M3 and M5 have been shown to be delayed or lower in subjects with the CYP2C9*3*3 polymorphism compared with those with the CYP2C9*1*1 genotype. In phase I studies, patients with CYP2C9*3*3 had an ~3 bpm lower heart rate on each day of up-titration compared to those with the CYP2C9*1*1 genotype. Also, in patients with CYP2C9*2*3 and CYP2C9*3*3, the per-centage of lymphocytes was lower compared with extensive metabolisers (CYP2C9*1*1). The risk of macular oedema was also higher in poor metabolisers. Based on the expected risk of high chronic exposure, those patients with the CYP2C9*3*3 (poor metaboliser) polymorphism were not to be in-cluded in the EXPAND study. Thus, siponimod should not be used in patients with a CYP2C9*3*3 polymorphisms. In patients with a CYP2C9*2*3 or *1*3 genotype, the recommended maintenance dose is 1 mg taken once daily. Therefore, the use of siponimod requires the determination of CYP2C9 gen-otypes. This test is not used in clinical practice and may not be available in some countries. The avail-ability and cost of this genotyping assay will need to be evaluated locally.



Table 4.14. Overview of adverse events in the global SPMS population

EXPAND SPECTRIMS European Study North American Study

ASCEND IMPACT Nordic SPMS Study

Safety outcome Siponi-mod

(n = 1099)

n (%)

Placebo (n = 546)

n (%)

IFN-β-1a 22 µg

(n = 209)

n (%)

IFN-β-1a 44 µg

(n = 204)

n (%)

Placebo (n = 205

)

n (%)

IFN-β-1b (250 µg) (n = 360)

n (%)

Placebo (n = 358)

n (%)

IFN-β-1b 250

µg/m2

(n = 317) n (%)

Placebo (n = 308) n (%)

Natalizumab (n = 439) n (%)

Placebo (n = 440) n (%)

IFN-β-1a 60 µg

(n = 217)

n (%)

Placebo (n = 218) n (%)

IFN-β-1a 22 µg (once

weekly) (n = 186)

n (%)

Pla-cebo (n =

178) n (%)

Total number of

adverse events

975 (88.7) 445 (81.5) NR NR NR NR NR NR NR 401 (91.3) 410

(93.2)

215

(99.1)

215

(98.6)

NR NR

Total number of

serious adverse events

197 (17.9) 83 (15.2) NR NR NR NR NR NR NR 90 (20.5) 100

(22.3)

NR NR 79

(42.5)

72

(40.4)

Total number of deaths 4 (0.4) 4 (0.7) 1 (0.5) 2 (1.0) 2 (1.0) 3 (0.8) 1 (0.3) 4 (1.3) 1 (0.3) 2 (0.5) 0 (0) 2 (0.9) 0 (0) 2 (1.1) 2 (1.1)

Total number of adverse events leading to temporary or

permanent treatment withdrawal

160 (14.5) 44 (8.1) NR NR NR 5 (1.4) 4 (1.1) 30 (9.5) 12 (3.9) NR NR NR NR NR NR

Total number of

withdrawals from the study because of adverse events

84 (7.6) 28 (5.1) NR NR NR NR NR NR NR 21 (4.8) 21 (4.8) 13 (6.0) 8 (3.6) 16 (8.6) 6 (3.4)

Source: Adapted from MAH submission dossier and Nordic SPMS Study [13]. Abbreviations: IFB=interferon; NR=not reported



Summary of treatment effect on the selected outcomes in the relapsing population

In the relapsing population, the comparative effectiveness of siponimod was assessed for each out-come.

Confirmed disability progression

In patients with relapsing SPMS, the ITCs showed no statistically significant difference between siponi-mod and:

Interferon-β-1b 250 µg every other day (HR 0.81; 95%CI 0.57 to 1.15) for the outcome time to CDP-6 (EXPAND vs Nordic SPMS Study);

Interferon-β-1a 22 µg once a week (HR 0.70; 95%CI 0.43 to 1.15) for the outcome proportion of patients with CDP-3 at 33 months (EXPAND vs European Study);

Interferon-β-1a 44 µg three times weekly (HR 0.88; 95%CI 0.55 to 1.42) for the outcome time to CDP-3 (EXPAND vs SPECTRIMS).

However, the quality of the available evidence was considered to be too low to derive an added benefit or a lesser benefit in the absence of dramatic effects. Therefore, these results should be interpreted with caution.

MS symptoms

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on MS symptoms. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on MS symptoms.

Clinical relapse

In patients with relapsing SPMS, the ITCs showed no statistically significant differences between siponi-mod and interferon-β-1a 22 µg three times weekly (HR 1.10; 95%CI 0.65 to 1.87) or 44 µg three times weekly (0.94; 95%CI 0.59 to 1.49) in relation to relapse rate. The SLR carried out by the MAH did not retrieve any information on the treatment effect on relapse rate of siponimod versus interferon-β-1a 22 µg once a week or interferon-β-1b 250 µg every other day, and therefore it was not possible to evaluate how siponimod compares with these comparators in relation to relapse rate.

However, the quality of the available evidence was considered to be too low to derive an added benefit or a lesser benefit in the absence of dramatic effects. Therefore, these results should be interpreted with caution.

Mortality

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on mortality. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on mortality.

Health-related quality of life

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparators treatments on HRQoL. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on HRQoL.

MRI inflammatory disease activity

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on MRI inflammatory disease activity. Therefore, it was not



possible to evaluate how siponimod compares with the selected comparators in relation to effect on MRI inflammatory disease activity.

No evidence of disease activity

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on no evidence of disease activity. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on this outcome.

Adverse events

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on adverse events. Therefore, it was not possible to evalu-ate how siponimod compares with the selected comparators in relation to effect on adverse events.

Serious adverse events

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on serious adverse events. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on serious adverse events.

Adverse events leading to treatment discontinuation

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on adverse events leading to treatment discontinuation. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on adverse events leading to treatment discontinuation.

Treatment-related mortality

The SLR carried out by the MAH did not retrieve any information on the treatment effect of siponimod versus the selected comparator treatments on treatment-related mortality. Therefore, it was not possible to evaluate how siponimod compares with the selected comparators in relation to effect on treatment-related mortality.

Direct comparison versus placebo

EXPAND was the only study of siponimod in the SPMS population. Though the comparator in this study (placebo) was not included in the PICO of this assessment and the study was conducted in a broader population than the population of interest in this assessment, the authors have included a brief overview of the effects of siponimod vs placebo in the “active” and the “relapsing” populations, as shown in Sec-tion 3.4. A detailed description of the EXPAND study is presented in APPENDIX 5: EXPAND (Study A2304).

Ancillary analysis in the global SPMS population

As an ancillary analysis, the MAH performed analyses in the global (active and non-active) SPMS pop-ulation.

Although the global SPMS population is not the population of interest, the results of studies in this population may be informative to the reader. Therefore, the results in the global population are summa-rised in Table 4.15. An in-depth review of the ancillary analysis in the global SPMS population is pre-sented in APPENDIX 6: Ancillary analysis.



Table 4.15. Summary of ancillary analysis: indirect comparisons between siponimod and com-parators of interest in the global (active and non-active) SPMS population

Outcome ITC method

MAIC STC Bucher ITC NMA

Comparison of siponimod vs interferon-β-1a 22 µg once a week (Nordic Study)

CDP-3 ------ ------ ------ ------

CDP-6

HR (95% CI)

0.43

(0.20 to 0.93)

0.50

(0.29 to 0.87)

0.66

(0.44 to 0.97)

0.65

(0.44 to 0.97)

ARR

Rate Ratio: (95% CI)

0.59

(0.32 to 1.07)

0.64

CI: (0.36 to 1.16)

0.50

(0.32 to 0.78)

0.50

(0.32 to 0.78)

Comparison of siponimod vs interferon-β-1a 22 µg three times weekly (SPECTRIMS)

CDP-3

HR (95% CI)

0.80

(0.46 to 1.38)

0.72

(0.47 to 1.10)

0.90

(0.66 to 1.22)

p-value (0.4919)

0.90

(0.66 to 1.22)

CDP-6 ------ ------ ------ ------

ARR


0.73

(0.40 to 1.32)

0.83

(0.49 to 1.40)

0.65

(0.46 to 0.92)

p-value (0.0143)

0.65

(0.46 to 0.92)

Comparison of siponimod vs interferon-β-1a 44 µg three times weekly (SPECTRIMS)

CDP-3

HR (95% CI)

0.84

(0.49 to 1.47)

0.76

(0.50 to 1.17)

0.95

(0.70 to 1.30)

0.95

(0.70 to 1.30)

CDP-6

HR (95% CI)

------ ------ ------ ------

ARR


0.73

(0.40 to 1.32)

0.83

(0.49 to 1.40)

0.65

(0.46 to 0.92)

p-value (0.7573)

0.65

(0.46 to 0.92)

Comparison of siponimod vs interferon-β-1b 250 µg every other day (European Study

and North American Study)

CDP-3 (European Study)

HR (95% CI)

0.82

(0.42 to 1.63)

0.81

(0.53 to 1.23)

1.07

(0.81 to 1.41)

p-value (0.6460)

1.06

(0.81 to 1.41)

CDP-6 (North American

Study)

HR (95% CI)

0.55

(0.33 to 0.91)

0.63

(0.39 to 1.01)

0.80

(0.57 to 1.12)

p-value (0.1953)

0.80

(0.57 a 1.13)

ARR

Rate Ratio (95% CI)

0.90

(0.51 to 1.59)

0.89

(0.50 a 1.56)

0.69

(0.46 to 1.04)

p-value (0.0734)

0.69

(0.46 a 1.04)

Comparison of siponimod vs natalizumab 300 mg q4w (ASCEND)

CDP-3 ------ ------ ------ ------

Proportion of patients

with CDP-6

OR (95% CI)

0.76

(0.44 to 1.30)

0.69

(0.44 to 1.09)

0.74

(0.48 to 1.14)

------



Outcome ITC method

MAIC STC Bucher ITC NMA

ARR

Rate Ratio (95% CI)

1.43

(0.78 to 2.61)

1.11

(0.69 to 1.78)

0.99

(0.64 to 1.54)

p-value (0. 9761)

0.99

(0.64 to 1.53)

Comparison of siponimod vs interferon-β-1a 60 µg once a week (IMPACT)

CDP-3

HR (95% CI)

0.42

(0.20 to 0.88)

0.64

(0.38 to 1.08)

0.81

(0.53 to 1.23)

p-value (0.3107)

0.81

(0.54 to 1.22)

CDP-6 ------ ------ ------ ------

ARR

Rate Ratio (95% CI)

0.997

(0.46 to 2.18)

0.93

(0.52 to 1.66)

0.68

(0.45 to 1.01)

p-value (0.0586)

0.67

(0.45 to 1.01)

Source: Adapted from the technical reports provided by the MAH. p-values available for Bucher ITC are are reported in the table. p-values not available for other comparisons Abbreviations: CDP=confirmed disability progression; ARR=annualised relapse rate; ITC=indirect treatment comparisons;

MAIC=matching adjusted indirect comparisons; STC=simulated treatment comparisons; NMA=network meta-analysis; q4w=every four weeks; CI=confidence interval; HR=hazard ratio; OR=odds ratio.

Subgroup analyses

No additional subgroup analyses were carried out.



5 PATIENT INVOLVEMENT

At the start of this Joint Assessment, an open call for patient input was published on the EUnetHTA website. This open call specifically asked patient organisations to answer the questions, as they are best placed to collect and present patients’ and caregivers’ views and experiences by engaging with a wide range of patients and their carers.

The open call used by EUnetHTA asked general questions to elicit patients’ views on living with the disease, important outcomes to be considered in this assessment, and expectations about the drug under assessment. The questions were based on the HTAi questionnaire template. For more infor-mation on the development of the HTAi questionnaire template, please see their website.

European and national patient organisations had to provide an organisational perspective on the ques-tions in English. In all parts of the open call, the term ‘patient’ refers to anyone living with, or who has lived with, the condition for which the new medicine is indicated. Four patient organisations completed the survey, namely the MS Society of Slovakia, the Russian MS Society, Smaragd Sclerosis Multiplexes Betegek Egyesülete (Hungary), and Združenie Sklerosis multiplex Nádej (Slovakia).

In addition, EUnetHTA contacted the patient organisations who responded to the open call to ask for input into the ranking of outcomes. None of the patient organisations responded to this additional re-quest for information.

The information gathered from the open call was used to inform the scope of this assessment, in par-ticular the outcomes to be considered. In the PICO table (Table 0.1), the outcomes that are related to issues particularly emphasised by patient organisations are marked by a superscript “a”. The frequency and route of administration were also mentioned as important points to take into account.

Comments were received regarding how SPMS affects carers/unpaid caregivers. Even though this as-sessment focuses on patient-relevant outcomes, the costs (humans and financial) of carers/unpaid caregivers may be captured at the national level in an economic assessment with a societal perspective.

https://htai.org/interest-groups/pcig/resources/for-patients-and-patient-groups/



6 DISCUSSION

The present assessment aimed to to assess the relative effectiveness and safety of siponimod in com-parison with appropriate comparators in adult patients with SPMS with active disease evidenced by relapses or imaging features of inflammatory activity.

The PICO defined during the scoping phase selected the following treatments as comparators: inter-feron-β-1a or -β-1b, mitoxantrone, ocrelizumab, natalizumab, fingolimod, cladribine, and rituximab, all in combination with best supportive care. Owing to the risk of adverse events, including cardiotoxicity and risk of cancer, the MAH considered that siponimod would not be an alternative to mitoxantrone in clinical practice and, therefore, mitoxantrone was excluded from the list of comparators. Indeed, mito-xantrone is only indicated for the treatment of MS patients in whom no alternative therapeutic options exist, according to the outcome of a procedure under article 30 of Directive 2001/83/EC: referral of Novantrone and associated names [7]. Taking that into consideration, the Authoring Team decided to accept the exclusion of mitoxantrone considering the extremely limited use in clinical practice in Europe.

The MAH performed a SLR to search for studies including siponimod and the comparators of interest in adult patients. The search retrieved seven studies in patients with SPMS: one study of siponimod, two studies of interferon-β-1b, three studies of interferon-β-1a, and one study of natalizumab. These studies were: EXPAND (siponimod), ASCEND (natalizumab), SPECTRIMS (IFN-β-1a 22 and 44 µg three times weekly), North American Study (interferon-β-1b 250 µg every other day), European Study (interferon-β-1b 250 µg every other day), Nordic SPMS study (IFN-β-1a 22 µg once a week), and IM-PACT (interferon-β-1a 60 µg once a week). No studies of ocrelizumab, fingolimod, cladribine, or rituxi-mab in the SPMS population were identified. No study directly compared siponimod to the comparators of interest.

In the absence of an active comparator in the only trial of siponimod (EXPAND), indirect comparisons were required to assess the relative effectiveness of siponimod and comparators.

However, none of the comparator studies identified by the MAH’s SLR reported results on an “active” SPMS subgroup defined by the presence of clinical relapses assessed at least annually and/or MRI activity (the population of interest in this assessment); only results on “relapsing SPMS” subgroups were reported. Relapsing subgroups were defined variably in each study, and Bucher ITCs were performed using treatment effect on this subgroup from EXPAND and compar-ator trials by using the EXPAND IPD to match the “relapsing SPMS” subgroup and outcome definitions stated in the comparator trials. In this way, the relapsing subgroup was used as a proxy for the active subgroup due to the absence of comparator data in active SPMS. The MAH reported results of the post hoc subgroup analysis of EX-PAND, which found that the active and relapsing subgroups were similar with regards to the outcomes for time to three-month and six-month CDP compared to placebo (3-month CDP: HR 0.68; 95%CI 0.52 to 0.89; p=0.0049 in the active SPMS subgroup and HR 0.67; 95% CI 0.50 to 0.91; p=0.0108 in the relapsing SPMS subgroup). The MAH considered that comparisons between relapsing SPMS sub-pop-ulations of EXPAND and comparator trials may therefore be an appropriate indicator of the relative effectiveness in the active SPMS population. While similar analyses were not available for comparator studies, the Authoring Team considered this to be a reasonable approach. However, notwithstanding the reasonableness of the approach, the authors consider that this rationale may not hold for every comparison given the variable definition of “relapsing” among comparator trials and that different defi-nitions hamper the interpretation and external validity of the results, specifically with regard to the pop-ulations to which they apply.

Three comparator studies reported a relapsing subgroup and included the outcomes of interest. This was the case for the Nordic SPMS Study (interferon-β-1a 22 µg once a week), which reported time to CDP-6 in the relapsing subpopulation (defined as patients with relapses during the four years prior to study start); SPECTRIMS (interferon-β-1a 22 µg three times weekly or 44 µg three times weekly), which reported time to CDP-3 (only for 44 µg) and ARR in the relapsing population (defined as patients with relapses during the two years prior to study start); and the European Study (interferon-β-1b 250 µg every other day), which reported the treatment effect on the proportion of patients with CDP-3 at 33 months in the relapsing population (defined as patients with relapses during the two years prior to study start). The MAH extracted data from the relevant subgroups of the comparator studies and performed



an analysis of SPMS patients with active disease defined by the presence of clinical relapses assessed at least annually.

In the main report, the MAH decided to exclude the Nordic SPMS Study from the analysis due to the use of an unlicensed regimen of interferon-β-1a. However, as the interferon-β-1a dose and regimen used in the Nordic SPMS Study may still be used in clinical practice in some countries, it was decided to include the Nordic SPMS Study in the assessment.

The ITCs showed that in patients with relapsing SPMS, there were no statistically significant differences between siponimod and interferon-β-1a 22 µg once a week, interferon-β-1b 250 µg every other day, and interferon-β-1a 44 µg three times weekly in relation to disability progression. Also, there were no statistically significant differences between siponimod and interferon-β-1a 22 or 44 µg three times weekly in relation to relapse rate.

The validity of results generated by ITCs is dependent on whether studies meet the assumption of similarity (meaning comparability, or assumption that the true treatment effect of each of the treatments in the contrast is comparable across trials that contribute to the comparison). None of the comparator studies reported the baseline characteristics of the relapsing subgroups. Thus, it was not possible to assess the comparability of the relapsing populations between studies. Since there may have been differences in the MS populations included in each trial, it was unclear whether the treatment effect was consistent between the studies compared. This limits the interpretation of the findings.

For comparisons with the European Study, the outcome proportion of patients with CDP-3 at 33 months was estimated for the EXPAND study, but it was assumed that these patients had the same status at 33 months as they did on the day that they were censored (imputation method: last observation carried forward). This imputation method may overestimate the efficacy of siponimod and, thus, a more con-servative approach would be preferable (for example, assuming that censored patients had CDP-3 events).

The imputation method for missing values used in EXPAND was last observation carried forward. Since SPMS is a progressive disease, this assumption is likely to be flawed and favour the intervention the imputed data referred to.

The data from the population with active disease came from a post hoc subgroup analysis. There are several limitations to subgroup analyses. Using the EXPAND data, the MAH performed analyses to determine whether the effect of siponimod was independent of relapses. A pre-planned subgroup anal-ysis revealed that the hazard ratio for patients without relapses in the two years prior to study start was 0.87 (95%CI 0.68 to 1.11) for three-month CDP, while the hazard ratio was 0.67 (95%CI 0.49 to 0.91) for those with relapses in the two years prior to study start. Similar findings were reported for six-month CDP: hazard ratio 0.63 (95%CI 0.44 to 0.89) for those with relapses in the two years prior to study start and hazard ratio 0.82 (95%CI 0.62 to 1.08) for those without them. At the request of the EMA, the MAH performed post hoc subgroup analyses to assess whether the benefit of siponimod on disease progres-sion was independent of the presence of relapses. Based on these post hoc subgroup analyses, the CHMP considered that it remains very challenging to disentangle the effect of siponimod on relapses and on general disease progression, and concluded that the “CHMP is of the opinion that efficacy of siponimod has been demonstrated in patients with SPMS with active disease evidenced by relapses or imaging feature of inflammatory activity”.

However, when comparing siponimod with the comparators of interest in the relapsing SPMS popula-tion, the relapsing population represented a subgroup of the population in each of the included studies (EXPAND (35.7%), Nordic SPMS (58.5%), European (69.9%), and SPECTRIMS (47.4%)), and the studies were not powered to make the comparisons of interest. Indeed, in subgroup analyses, there was a reduction in sample size relative to the ITT population that may not have enough statistical power to detect meaningful differences. This issue is also inherent to MAICs, following the matching and ad-justment process. Therefore, these results should be interpreted with caution.

According to the EMA, siponimod is mainly metabolised via CYP2C9. In study A2128, subjects with CYP2C9*2*3 or CYP2C9*3*3 genotypes had a two- or four-fold increase, respectively, in AUC and T½ compared with those with a CYP2C9*1*1 genotype. In phase I studies, patients with CYP2C9*3*3 pol-ymorphisms had an ~3 bpm lower heart rate on each day of up-titration compared to those with



CYP2C9*1*1. Also, in patients with CYP2C9*2*3 and CYP2C9*3*3, the percentage of lymphocytes was lower compared with extensive metabolisers (CYP2C9*1*1). The risk of macular oedema was also higher in poor metabolisers. Based on the expected risk of high chronic exposure, those patients with CYP2C9*3*3 (poor metaboliser) polymorphisms were not to be included in the EXPAND study. Thus, siponimod should not be used in patients with a CYP2C9*3*3 genotype. In patients with a CYP2C9*2*3 or *1*3 genotype, the recommended maintenance dose is 1 mg taken once daily. Therefore, the use of siponimod requires the determination of CYP2C9 genotypes. This genotyping test is not currently used in clinical practice and may not be available in some countries. The availability and cost of this geno-typing test will have to be evaluated locally.

The MAH performed ancillary analyses in the global (active and non-active) SPMS population. The MAH stated that there were significant differences between studies in inclusion/exclusion criteria, as well as in the baseline characteristics of the populations included in those studies. The MAH considered, therefore, that Bucher ITCs were not appropriate, and proposed MAIC as an alternative analytical ap-proach. However, insufficient justification was provided for the use of MAIC. Therefore, the authors requested that comparisons be made between siponimod and comparators using Bucher ITCs and NMA, with MAICs and STCs provided as supporting analyses where considered feasible and valid. Of note, for each pairwise comparison, the treatment effect on CDP and ARR using different methodolo-gies (whether from Bucher ITC, NMA, STC, or MAIC analyses) generally gave consistent results. The Authoring team considers that MAIC is an inappropriate method for indirect treatment comparison in the setting of the current assessment. This is due to the inherent limitations of statistical techniques like MAIC [69], which has not been shown to produce less biased estimates than would be available through standard indirect comparisons, in the target population. Additionally, the alleged advantage of matching and adjustment by the MAIC is not clear. Overall, in the global SPMS population, siponimod in comparison with natalizumab, interferon-β-1a 22 µg and 44 µg three times weekly, INF-β-1b 250 µg every other day, or interferon-β-1a 60 µg once a week did not show statistically significant differences in relation to disease progression or clinical relapses. These results are consistent with those obtained in the relapsing SPMS population. Siponimod in comparison with interferon-β-1a 22 µg once a week showed a statistically significant difference favouring siponimod, in the global SPMS population.

As there were no data on comparative safety between siponimod and comparators, the toxicity of siponi-mod could only be assessed using data from the EXPAND study (siponimod vs placebo). Adverse events were more frequent in the siponimod than the placebo group (88.7% vs 81.5%). Serious adverse events were also reported more frequently in the siponimod than the placebo group (17.9% vs 15.2%). More patients discontinued the study drug permanently due to adverse events in the siponimod group (7.6% vs 5.1%). Adverse events of special interest that were reported more frequently in the siponimod group were herpes zoster reactivations (2.2% vs 0.7%), lymphopenia (1.6% vs 0%), macular oedema (1.7% vs 0.2%), and increased liver transaminases (1.4% vs 0.6%). Globally, the safety information raised no specific safety concerns.

The scope of the assessment deviated from the scope described in the project plan as follows:

Comparisons were not performed between siponimod and mitoxantrone. Owing to the risk of ad-verse events including cardiotoxicity and risk of cancer, the MAH considered that siponimod would not be an alternative to mitoxantrone in clinical practice and, therefore, mitoxantrone was excluded from the list of comparators. Mitoxantrone is indeed only indicated for the treatment of patients with MS where no alternative therapeutic options exist, according to the outcome of a procedure under article 30 of Directive 2001/83/EC: referral of Novantrone and associated names [7]. Taking this into consideration, the Authoring Team decided to accept the exclusion of mito-xantrone considering the extremely limited use in clinical practice in Europe.

No relevant data could be retrieved to perform several of the planned comparisons, therefore:

Comparisons were not possible between siponimod and ocrelizumab, natalizumab, fingolimod, cladribine, and rituximab in the population of interest;

The relative effectiveness of siponimod in terms of MS symptoms, mortality, HRQoL, MRI inflam-matory disease activity, no evidence of disease activity, adverse events, serious adverse events, treatment discontinuation due to adverse events, and treatment-related mortality were not evalu-ated in the population of interest.



7 CONCLUSIONS

In patients with relapsing SPMS, indirect comparisons (using Bucher ITCs) of treatment effects from subgroup analyses showed no statistically significant difference between siponimod and interferon-β-1a 22 µg once a week (HR 0.70; 95%CI 0.43 to 1.15), interferon-β-1b 250 µg every other day (OR 0.88; 95%CI 0.53 to 1.47), and interferon-β-1a 44 µg three times weekly (HR 0.88; 95%CI 0.55 to 1.42) in relation to disability progression.

In patients with relapsing SPMS, an indirect comparison (using Bucher ITCs) of treatment effects from subgroup analyses showed no statistically significant differences between siponimod and interferon-β-1a 22 µg three times weekly (RR 1.10; 95% CI 0.65 to 1.87) or 44 µg three times weekly (RR 0.94; 95%CI 0.59 to 1.49) in relation to relapse rate.

The quality of the available evidence was too low to derive an added benefit or a lesser benefit in the absence of dramatic effects. Therefore, these results should be interpreted with caution due to the serious limitations of the evidence.

The use of siponimod requires the determination of CYP2C9 genotypes. This genotype test is not used in clinical practice and may not be available in some countries. The availability and cost of this genotype test will have to be evaluated locally.

There was no evidence, either direct or indirect, to evaluate the compared efficacy and safety of siponi-mod with most of the available comparators in use in current clinical practice in Europe for patients with active SPMS, namely ocrelizumab, natalizumab, fingolimod, cladribine and rituximab.

Taking into consideration the lack of comparative clinical evidence available at the time of the marketing authorisation and presented by the MAH in the submission dossier, the Authoring team suggest that further studies are warranted. Recommendations for further research may be found in Table A13 of APPENDIX 3: Evidence gaps.



8 REFERENCES

1. Compston A, Coles A. Multiple sclerosis. Lancet 2008;372:1502–15717. 2. Eriksson M, Andersen O, Runmarker B. Long-term follow up of patients with clinically isolated

syndromes, relapsing-remitting and secondary progressive multiple sclerosis. Mult Scler 2003;9(3):260–74.

3. Lublin FD. New multiple sclerosis phenotypic classification. Eur Neurol 2014;72(suppl 1):1–5. 4. Dumitrescu L, Constantinescu CS, Tanasescu R. Siponimod for the treatment of secondary

progressive multiple sclerosis. Expert Opin Pharmacother 2019;20(2):143–50. 5. European Medicines Agency. Summary of Product Characteristics: Mayzent.

https://www.ema.europa.eu/en/documents/product-information/mayzent-epar-product-infor-mation_en.pdf (accessed 7 Feb 2020).

6. European Network for Health Technology Assessment (EUnetHTA). Guideline. Comparators & Comparisons: Direct and indirect comparisons. 2015;16(2):34. https://eunethta.eu/wp-content/uploads/2018/01/Comparators-Comparisons-Direct-and-indirect-comparisons_Amended-JA1-Guideline_Final-Nov-2015.pdf (accessed 1 Feb 2020).

7. European Medicines Agency. Questions and answers on Novantrone and associated names (mitoxantrone 2 mg/ml concentrate for solution for infusion). https://www.ema.europa.eu/en/documents/referral/questions-answers-novantrone-associated-names-mitoxantrone-2-mg/ml-concentrate-solution-infusion_en.pdf (accessed 1 Feb 2020).

8. Higgins JPT GS (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.1. The Cochrane Collaboration and John Wiley & Sons Ltd; 2012.

9. Kappos L, Bar-Or A, Cree BAC, et al. Siponimod versus placebo in secondary progressive multiple sclerosis (EXPAND): a double-blind, randomised, phase 3 study. Lancet 2018;391(10127):1263–73.

10. Bucher HC, Guyatt GH, Griffith LE, Walter SD. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J Clin Epidemiol 1997;50(6):683–91.

11. Dias S, Welton NJ, Sutton AJ AA. NICE DSU Technical Support Document 2: A generalised linear modelling framework for pair-wise and network meta-analysis of randomised controlled trials. http://nicedsu.org.uk/wp-content/uploads/2017/05/TSD2-General-meta-analysis-corrected-2Sep2016v2.pdf (accessed 1 Feb 2020).

12. Signorovitch JE, Sikirica V, Erder MH, et al. Matching-adjusted indirect comparisons: A new tool for timely comparative effectiveness research. Value Heal 2012;15(6):940–7.

13. Andersen O, Elovaara I, Färkkilä M, et al. Multicentre, randomised, double blind, placebo controlled, phase III study of weekly, low dose, subcutaneous interferon beta-1a in secondary progressive multiple sclerosis. J Neurol Neurosurg Psychiatry 2004;75(5):706–10.

14. Li DKB, Zhao GJ, Paty DW. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: MRU results. Neurology 2001;56(11):1505–513.

15. SPECTRIMS Study Group. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: Clinical results. Neurology 2001;56(11):1496–504.

16. Kappos L, Polman C, Pozzilli C, Thompson A, Dahlke F. Placebo-controlled multicentre randomised trial of interferon β-1b in treatment of secondary progressive multiple sclerosis. Lancet 1998;352(9139):1491–7.

17. Kappos L, Polman C, Pozzilli C, Thompson A, Beckmann K, Dahlke F. Final analysis of the European multicenter trial on IFNβ-1b in secondary-progressive MS. Neurology 2001;57(11):1969–75.

18. Kapoor R, Ho PR, Campbell N, et al. Effect of natalizumab on disease progression in secondary progressive multiple sclerosis (ASCEND): a phase 3, randomised, double-blind, placebo-controlled trial with an open-label extension. Lancet Neurol 2018;17(5):405–15.

19. Panitch H. Interferon beta-1b in secondary progressive MS: Results from a 3-year controlled study. Neurology 2004;63(10):1788–95.

20. Cohen JA, Cutter GR, Fischer JS, et al. Benefit of interferon β-1a on MSFC progression in secondary progressive MS. Neurology 2002;59(5):679–87.

21. Boström I, Callander M, Kurtzke JF, Landtblom AM. High prevalence of multiple sclerosis in the Swedish county of Värmland. Mult Scler 2009;15(11):1253–62.

22. Bentzen J, Meulengracht Flachs E, Stenager E, Brønnum-Hansen H, Koch-Henriksen N. Prevalence of multiple sclerosis in Denmark 1950-2005. Mult Scler 2010;16(5):520–5.

23. Sarasoja T, Wikström J, Paltamaa J, Hakama M, Sumelahti ML. Occurrence of multiple sclerosis in central Finland: A regional and temporal comparison during 30 years. Acta Neurol Scand



2004;110(5):331–6. 24. Aladro Y, Alemany MJ, Pérez-Vieitez MC, et al. Prevalence and incidence of multiple sclerosis

in Las Palmas, Canary Islands, Spain. Neuroepidemiol 2005;24(1–2):70–5. 25. Risberg G, Aarseth JH, Nyland H, Lauer K, Myhr KM, Midgard R. Prevalence and incidence of

multiple sclerosis in Oppland County - a cross-sectional population-based study in a landlocked county of Eastern Norway. Acta Neurol Scand. 2011;124(4):250–7.

26. Leray E, Moreau T, Fromont A, Edan G. Epidemiology of Multiple Sclerosis. Rev Neurol 2016;172(1):3–13.

27. Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis : Neurology 1996;46:907–11.

28. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983;33(11):1444–52.

29. Fischer JS, Rudick RA, Cutter GR, Reingold SC. The Multiple Sclerosis Functional Composite measure (MSFC): an integrated approach to MS clinical outcome assessment. Mult Scler 1999;5(4):244–50.

30. European Medicines Agency. Guideline on Clinical Investigation of MPs for the treatment of Multiple Sclerosis (EMA/CHMP/771815/2011, Rev. 2). Guideline. 2015;44(2):20. http://www.ema.europa.eu/ema/index.jsp?curl=pages/regulation/general/general_content_000425.jsp&mid=WC0b01ac0580034cf5 (accessed 1 Feb 2020).

31. Fernández O, Delvecchio M, Edan G, et al. Survey of diagnostic and treatment practices for multiple sclerosis (MS) in Europe. Part 2: Progressive MS, paediatric MS, pregnancy and general management. Eur J Neurol 2018;25(5):739–46.

32. Khan O, Miller AE, Tornatore C, Theodore Phillips J, Barnes CJ. Practice patterns of US neurologists in patients with SPMS and PPMS: A consensus study. Neurol Clin Pract 2012;2(1):58–66.

33. Montalban X, Hauser SL, Kappos L, et al. Ocrelizumab versus placebo in primary progressive multiple sclerosis. N Engl J Med 376(3):280–2.

34. Kappos L, Weinshenker B, Pozzilli C, et al. Interferon beta-1b in secondary progressive MS. Neurology 2004;63(10):1788 LP – 1795.

35. 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.

36. Hartung H, Gonsette R, König N, et al. Mitoxantrone in progressive multiple sclerosis : a placebo-controlled, double-blind, randomised, multicentre trial. Lancet 2018;360:2018–25.

37. Hauser SL, Bar-Or A, Comi G, et al. Ocrelizumab versus Interferon Beta-1a in Relapsing Multiple Sclerosis. N Engl J Med 2016;376(3):221–34.

38. Sipe JC, Romine JS, Koziol JA, et al. Cladribine in treatment of chronic progressive multiple sclerosis. Lancet 1994;344(8914):9–13.

39. Rice GP, Filippi M, Comi G. Cladribine and progressive MS: clinical and MRI outcomes of a multicenter controlled trial. Cladribine MRI Study Group. Neurology 2000;54(5):1145–55.

40. Lublin F, Miller DH, Freedman MS, et al. Oral fingolimod in primary progressive multiple sclerosis (INFORMS): A phase 3, randomised, double-blind, placebo-controlled trial. Lancet 2016;387(10023):1075–84.

41. Hawker K, O’Connor P, Freedman MS, et al. Rituximab in patients with primary progressive multiple sclerosis: Results of a randomized double-blind placebo-controlled multicenter trial. Ann Neurol 2009;66(4):460–71.

42. La Mantia L, Munari LM, Lovati R. Glatiramer acetate for multiple sclerosis. Cochrane Database Syst Rev 2010;(5).

43. European Medicines Agency. Lemtrada for multiple sclerosis: measures to minimise risk of serious side effects. 2019;31(November). www.ema.europa.eu/contact (accessed 1 Feb 2020).

44. Goodkin DE, Kinkel RP, Weinstock-Guttman B, et al. A phase II study of IV methylprednisolone in secondary-progressive multiple sclerosis. Neurology 1998;51(1):239–45.

45. Casetta I, Iuliano G, Filippini G. Azathioprine for multiple sclerosis (Review). Cochrane Database Syst Rev 2010;(4):1–75.

46. Gray OM, McDonnell G V., Forbes RB. A systematic review of oral methotrexate for multiple sclerosis. Mult Scler 2006;12(4):507–10.

47. La Mantia L, Milanese C, Mascoli N, D’Amico R, Weinstock-Guttman B. Cyclophosphamide for multiple sclerosis. Cochrane Database Syst Rev 2007;(1).

48. Freedman, Bar-Or A, Oger J, et al. A phase III study evaluating the efficacy and safety of MBP8298 in secondary progressive MS. Neurology 2011;77(16):1551–60.

49. European Medicines Agency. Summary of Product Characteristics: Avonex.



https://www.ema.europa.eu/en/documents/product-information/avonex-epar-product-information_en.pdf (accessed 1 Feb 2020).

50. European Medicines Agency. Summary of Product Characteristics: Rebif. https://www.ema.europa.eu/en/documents/product-information/rebif-epar-product-information_en.pdf (accessed 1 Feb 2020).

51. European Medicines Agency. Summary of Product Characteristics: Plegridy. https://www.ema.europa.eu/en/documents/product-information/plegridy-epar-product-information_en.pdf (accessed 1 Feb 2020).

52. European Medicines Agency. Summary of Product Characteristics: Betaferon. https://www.ema.europa.eu/en/documents/product-information/betaferon-epar-product-information_en.pdf (accessed 1 Feb 2020).

53. European Medicines Agency. i https://www.ema.europa.eu/en/documents/product-information/extavia-epar-product-information_en.pdf (accessed 1 Feb 2020).

54. Electronic Medicines Compendium. Mitoxantrone 2 mg / ml concentrate for solution for infusion. https://www.medicines.org.uk/emc/product/1694/smpc (accessed 1 Feb 2020).55. European Medicines Agency. Summary of Product Characteristics: Ocrevus. https://www.ema.europa.eu/en/documents/product-information/ocrevus-epar-product-information_en.pdf (accessed 1 Feb 2020).

56. European Medicines Agency. Summary of Product Characteristics: Tysabri. https://www.ema.europa.eu/en/documents/product-information/tysabri-epar-product-information_en.pdf (accessed 1 Feb 2020).

57. European Medicines Agency. Summary of Product Characteristics: Gilenya. https://www.ema.europa.eu/en/documents/product-information/gilenya-epar-product-information_en.pdf (accessed 1 Feb 2020).

58. European Medicines Agency. Summary of Product Characteristics: Mavenclad. https://www.ema.europa.eu/en/documents/product-information/mavenclad-epar-product-information_en.pdf (accessed 1 Feb 2020).

59. European Medicines Agency. Summary of Product Characteristics: Mabthera. https://www.ema.europa.eu/en/documents/product-information/mabthera-epar-product-information_en.pdf (accessed 1 Feb 2020).

60. Hauser SL, Waubant E, Arnold DL, et al. B-cell depletion with rituximab in relapsing-remitting multiple sclerosis. N Engl J Med 2008;358(7):676–88.

61. European Network for Health Technology Assessment (EUneHTA). Comparators & comparisons. Criteria for the choice of the most appropriate comparator(s). Summary of current policies and best practice recommendations. 2015:23. https://eunethta.eu/wp-content/uploads/2018/01/WP7-SG3-GL-choice_of_comparator_amend2015.pdf (accessed 1 Feb 2020).

62. European Network for Health Technology Assessment (EUnetHTA). Guideline. Endpoints used for Relative Effectiveness Assessment: Clinical Endpoints. 2015:20. https://eunethta.eu/wp-content/uploads/2018/01/WP7-SG3-GL-clin_endpoints_amend2015.pdf (accessed 1 Feb 2020).

63. European Network for Health Technology Assessment (EUnetHTA). Guideline. Endpoints used for REA: Health-related Quality of Life and Utility Measures. 2015:1–53. https://eunethta.eu/wp-content/uploads/2018/01/Endpoints-used-for-Relative-Effectiveness-Assessment-Health-related-quality-of-life-and-utility-measures_Amended-JA1-Guideline_Final-Nov-2015.pdf (accessed 1 Feb 2020).

64. European Network for Health Technology Assessment (EUnetHTA). Guideline. Endpoints used in relative effectiveness assessment of pharmaceuticals Surrogate Endpoints. 2013:1–23. https://eunethta.eu/wp-content/uploads/2018/01/Endpoints-used-in-Relative-Effectiveness-Assessment-Surrogate-Endpoints_Amended-JA1-Guideline_Final-Nov-2015.pdf (accessed 1 Feb 2020).

65. European Network for Health Technology Assessment (EUnetHTA). Guideline. Process of information retrieval for systematic reviews and health technology assessments on clinical effectiveness. 2017:75. https://eunethta.eu/wp-content/uploads/2018/01/Guideline_Information_Retrieval_V1-2_2017.pdf (accessed 1 Feb 2020).

66. Moher D, Liberati A, Tetzlaff J, Altman DG, Group TP. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 2009 Jul;6(7).

67. Caro JJ, Ishak KJ. No head-to-head trial? Simulate the missing arms. Pharmacoeconomics. 2010;28(10):957–67.



68. Phillippo DM, Ades AE, Dias S, Palmer S, Abrams KR, Welton NJ. Methods for Population-Adjusted Indirect Comparisons in Health Technology Appraisal. Med Decis Mak 2018;38(2):200–11.

69. Phillippo DM, Ades AE, Dias S, Palmer S, Abrams KR, Welton NJ. Methods for Population-Adjusted Indirect Comparisons in Submissions To NICE. NICE DSU Tech Support Doc 18. 2016:1–82. http://www.nicedsu.org.uk/Populationadjusted-ICs-TSD(3026862).htm (accessed 1 Feb 2020).

70. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. Introduction - GRADE evidence profiles and summary of findings tables. J Clin Epidemiol 2011;64(4):383–94.

71. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 2. Framing the question and deciding on important outcomes. J Clin Epidemiol. 2011;64(4):395–400.

72. Balshem H, Helfand M, Schünemann HJ, et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol 2011;64(4):401–6.

73. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the quality of evidence - Study limitations (risk of bias). J Clin Epidemiol 2011;64(4):407–15.

74. Guyatt GH, Oxman AD, Montori V, et al. GRADE guidelines: 5. Rating the quality of evidence - Publication bias. J Clin Epidemiol 2011;64(12):1277–82.

75. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines 6. Rating the quality of evidence - Imprecision. J Clin Epidemiol 2011;64(12):1283–93.

76. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7. Rating the quality of evidence - Inconsistency. J Clin Epidemiol 2011;64(12):1294–302.

77. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 8. Rating the quality of evidence - Indirectness. J Clin Epidemiol 2011;64(12):1303–10.

78. Guyatt GH, Oxman AD, Sultan S, et al. GRADE guidelines: 9. Rating up the quality of evidence. J Clin Epidemiol 2011;64(12):1311–6.

79. Guyatt G, Oxman AD, Sultan S, et al. GRADE guidelines: 11. Making an overall rating of confidence in effect estimates for a single outcome and for all outcomes. J Clin Epidemiol 2013;66(2):151–7.

80. Guyatt GH, Oxman AD, Santesso N, et al. GRADE guidelines: 12. Preparing Summary of Findings tables - Binary outcomes. J Clin Epidemiol 2013;66(2):158–72.

81. Guyatt GH, Thorlund K, Oxman AD, et al. GRADE guidelines: 13. Preparing Summary of Findings tables and evidence profiles - Continuous outcomes. J Clin Epidemiol 2013;66(2):173–83.

82. Guyatt GH, Ebrahim S, Alonso-Coello P, et al. GRADE guidelines 17: assessing the risk of bias associated with missing participant outcome data in a body of evidence. J Clin Epidemiol 2017;87(2017):14–22.

83. Zhang Y, Alonso-Coello P, Guyatt GH, et al. GRADE Guidelines: 19. Assessing the certainty of evidence in the importance of outcomes or values and preferences—Risk of bias and indirectness. J Clin Epidemiol 2019;111:94–104.

84. Higgins JPT and Green S E. Cochrane Handbook for systematic reviews of interventions. New Jersey: Wiley-Blackwell; 2012.

85. European Network for Health Technology Assessment (EUnetHTA). Guideline. Applicability of evidence in the context of a relative effectiveness assessment of pharmaceuticals. 2013:1–35. https://eunethta.eu/wp-content/uploads/2018/01/Applicability.pdf (accessed 1 Feb 2020).



APPENDIX 1: SLR search strategies

Table A1. Original SLR search strategy for MEDLINE Daily, MEDINE Epub Ahead of Print and Embase (via the Embase.com platform)

# Search terms

1 'clinical trial'/exp

2 'randomized controlled trial'/de

3 'controlled clinical trial'/de

4 'phase 3 clinical trial'/de OR 'phase 2 clinical trial'/de OR 'phase 4 clinical trial'/de

5 'randomization'/de

6 'controlled study'/de

7 'comparative study'/de

8 'single blind procedure'/de

9 'double blind procedure'/de

10 'crossover procedure'/de

11 'placebo'/de OR placebo*

12 'clinical trial' OR 'clinical trials'

13 'controlled clinical trial' OR 'controlled clinical trials'

14 'randomised controlled trial' OR 'randomized controlled trial' OR 'randomised controlled trials' OR 'randomized controlled trials'

15 'randomisation' OR 'randomization'

16 rct

17 'random allocation' OR 'randomly allocated' OR 'allocated randomly'

18 (allocated NEAR/2 random) OR (random* NEAR/1 assign*) OR random*

19 'prospective study'/exp

20 (single OR double OR triple OR treble) NEAR/1 (blind* OR mask*)

21 'cohort analysis'/exp OR cohort*:ab,ti

22 'longitudinal study'/exp

23 'multicenter study'/exp

24 'follow up'/exp

25 'major clinical study'/exp

26 'case control study'/exp OR ((case* NEXT/1 control*):ab,ti)

27 'clinical article'/exp

28 'survival'/exp

29 (('follow up' OR followup) NEXT/1 (study OR studies)):ab,ti

30 (clinical NEXT/1 trial*):ab,ti

31 'retrospective study'/exp

32 'case control study'/exp

33 ((observational OR cohort) NEXT/1 (study OR studies)):ab,ti

34 'intervention study'/exp

35 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34



# Search terms

36 'case study'/de

37 'case report'

38 'abstract report'/de

39 'letter'/de

40 #36 OR #37 OR #38 OR #39

41 #35 NOT #40

42 ((((disseminated OR insular OR multiple) NEAR/4 sclerosis):ab,ti) OR ms:ab,ti) AND ('secondary progressive':ab,ti OR progressive:ab,ti OR 'non-relapsing':ab,ti)

43 ('secondary progressive' OR 'secondary-progressive' OR progressive OR secondary OR deteriorat* OR 'non relapsing' OR 'non relapsing') NEAR/3 (ms OR 'multiple sclerosis' OR 'disseminated sclerosis' OR 'encephalomyelitis disseminata' OR 'chariot disease' OR 'insular sclerosis')

44 spms OR cpms

45 'progressive secondary multiple sclerosis' OR 'secondary progressive multiple sclerosis'

46 #42 OR #43 OR #44 OR #45

47 'disease modifying therapies':ab,ti OR dmt*:ab,ti OR 'disease modifying therapy':ab,ti OR 'disease modifying drugs':ab,ti OR 'disease modifying drug':ab,ti OR dmd*:ab,ti

48 'glatiramer'/syn OR glatiramer OR 'cop 1' OR copaxone OR 'copolymer 1' OR 'copolymer cop 1' OR 'copolymer i' OR (glatiramer NEAR/1 (acetate OR sodium)) OR glatopa OR 'tv 5010' OR tv5010

49 'alemtuzumab'/syn OR alemtuzumab OR 'campath 1' OR 'campath 1h' OR (cd52 NEAR/1 'monoclonal antibody') OR 'ldp 103' OR ldp103 OR lemtrada OR mabcampath

50 'mitoxantrone'/syn OR mitoxantrone OR 'cl 232, 315' OR 'cl 232315' OR 'cl232, 315' OR cl232315 OR dhad OR dhaq OR domitrone OR elsep OR formyxan OR genefadrone OR misostol OR mitoxanthrone OR mitoxantron OR mitoxantrona OR mitoxgen OR mitozantrone OR mitroxantrone OR mitroxone OR neotalem OR norexan OR novanthron OR novantron OR novantrone OR 'now 85 34' OR 'now 8534' OR now8534 OR 'nsc 279836' OR 'nsc 301739' OR 'nsc 301739d' OR nsc279836 OR nsc301739 OR nsc301739d OR oncotron OR onkotrone OR ralenova

51 'rituximab'/syn OR rituximab OR blitzima OR 'ct p10' OR 'ctp10' OR 'idec 102' OR 'idec c2b8' OR 'idec102' OR idecc2b8 OR mabthera OR 'r 105' OR r105 OR reditux OR 'rg 105' OR rg105 OR ritemvia OR rituxan OR rituxin OR rituzena OR rixathon OR riximyo OR 'ro 452294' OR ro452294 OR truxima OR tuxella

52 'fingolimod'/syn OR fingolimod OR 'fty 720' OR fty720 OR gilenia OR gilenya

53 'natalizumab'/syn OR natalizumab OR 'an 100226' OR an100226 OR antegren OR tysabri

54 'fumaric acid dimethyl ester'/syn OR 'bg 12' OR bg12 OR 'dimethyl fumarate' OR dimethylfumarate OR 'fag 201' OR fag201 OR panaclar OR psorinovo OR skilarence OR tecfidera OR dmf

55 'teriflunomide'/syn OR teriflunomide OR 'a 771726' OR 'a77 1726' OR 'a77-1726' OR a771726 OR aubagio OR 'hmr 1726' OR hmr1726 OR 'rs 61980' OR rs61980 OR 'su 0020' OR su0020

56 'cladribine'/syn OR cladribine OR intocel OR leustat OR leustatin OR leustatine OR litak OR litax OR mavenclad OR movectro OR mylinax OR 'rwj 26251' OR rwj26251020

57 'beta interferon'/syn OR 'recombinant beta interferon'/exp OR ((interferon OR ifn) NEAR/2 beta) OR belerofon OR ifn?beta OR (beta1* NEAR/2 interferon) OR 'interferon beta1' OR 'beta-1 interferon' OR 'beta 1 interferon'

58 'beta1a interferon'/syn OR 'beta interferon 1a'/exp OR avonex OR ((interferon OR ifn) NEAR/1 ('beta 1a' OR 'beta-1a' OR beta1a OR 'beta 1b' OR beta1b OR 'beta 1b')) OR rebif OR 'rifn beta'

59 'interferon beta serine'/syn OR 'beta interferon 1b'/exp OR beneseron OR betaferon OR betaseron OR extavia OR 'rifn beta-1b' OR 'sh 579' OR sh579 OR 'zk 157046' OR zk157046

60 'masitinib'/syn OR masitinib OR 'ab 1010' OR ab1010 OR kinaction OR masatinib OR masican OR masipro OR masivet OR masiviera

61 'siponimod'/syn OR siponimod OR 'baf 312' OR baf312 OR mayzent

62 mis416



# Search terms

63 'imilecleucel t'/syn OR 'imilecleucel t' OR tcelna OR tovaxin

64 'biotin'/exp OR biotin OR biotine OR md1003

65 'ibudilast'/syn OR ibudilast OR 'av 411' OR av411 OR 'kc 404' OR kc404 OR ketas OR 'mn-166'

66 'ocrelizumab'/syn OR ocrelizumab OR ocrevus OR 'pro 70769' OR pro70769 OR 'rhumab 2h7'

67 'peginterferon'/exp OR 'peginterferon beta1a'/exp OR peginterferon OR 'beta 1a peginterferon' OR 'beta1a peginterferon' OR 'biib 017' OR 'biib017' OR 'peginterferon beta 1a' OR 'peginterferon beta-1a' OR 'pegylated human interferon beta 1a' OR 'pegylated interferon beta 1a' OR 'pegylated interferon beta-1a' OR 'pegylated interferon beta1a' OR plegridy

68 'idebenone'/syn OR idebenone OR avan OR cerestabon OR 'cv 2619' OR cv2619 OR mnesis OR 'qsa 10' OR qsa10 OR raxone OR 'snt mc17' OR sovrima

69 'opicinumab'/syn OR opicinumab OR 'biib 033' OR biib033

70 'stem cell transplantation'/exp OR 'hematopoietic stem cell transplantation'/syn OR ('stem cell' NEAR/2 (therap* OR transplant*)) OR hsct

71 'simvastatin'/syn OR simvastatin OR avastinee OR belmalip OR cholestat OR clinfar OR colastatina OR colemin OR colestricon OR covastin OR denan OR epistatin OR esvat OR ethicol OR eucor OR ifistatin OR jabastatina OR kavelor OR klonastin OR kolestevan OR 'l 644128' OR l644128 OR lipcut OR lipecor OR lipex OR lipinorm OR liponorm OR lipovas OR lodales OR medipo OR mersivas OR 'mk 733' OR mk733 OR 'nor vastina' OR normofat OR orovas OR pantok OR rechol OR simbado OR simcard OR simchol OR simovil OR simtin OR simva OR simvacor OR simvahex OR simvalord OR simvastar OR simvastatina OR simvastatine OR simvata OR simvatin OR simvor OR simvotin OR sinvacor OR sinvastatin OR sinvinolin OR sivastin OR starzoco OR synvinolin OR torio OR valemia OR vasilip OR vasotenal OR vazim OR vidastat OR zimmex OR zocor OR zocord OR zovast

72 'riluzole'/syn OR riluzole OR 'pk 26124' OR pk26124 OR rilutek OR 'rp 54274' OR rp54274

73 'fluoxetine'/syn OR fluoxetine OR actan OR adofen OR 'alzac 20' OR andep OR ansilan OR 'atd 20' OR auroken OR auscap OR captaton OR 'compound 110140' OR daforin OR depren OR deprexin OR deprizac OR deproxin OR elizac OR floxet OR fluctin OR fluctine OR fludac OR flufran OR fluketin OR flunil OR flunirin OR fluohexal OR fluox OR 'fluox puren' OR fluoxac OR fluoxeren OR fluoxifar OR fluoxil OR fluronin OR flusac OR flutin OR flutine OR fluxen OR fluxet OR fluxetil OR fluxetin OR fontex OR foxetin OR foxtin OR fropine OR fuloren OR lanclic OR 'lilly 110140' OR lilly110140 OR lorien OR lovan OR 'ly 110140' OR ly110140 OR magrilan OR margrilan OR modipran OR nopres OR nuzak OR oxedep OR plinzene OR pragmaten OR prizma OR proctin OR prodep OR prosac OR prozac OR prozamin OR qualisac OR rapiflux OR rowexetina OR salipax OR sanzur OR sarafem OR selfemra OR sinzac OR zactin OR zepax

74 'amiloride'/syn OR amiloride OR amiclaran OR amikal OR 'amilo 5' OR amilorid OR amiloridehydrochlorhydrate OR amiloridine OR amipramidine OR amyloride OR arumil OR berkamil OR colectril OR guanamprazine OR kaluril OR medamor OR midamor OR 'mk 870' OR modamide OR nirulid OR pandiuren

75 #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 OR #57 OR #58 OR #59 OR #60 OR #61 OR #62 OR #63 OR #64 OR #65 OR #66 OR #67 OR #68 OR #69 OR #70 OR #71 OR #72 OR #73 OR #74

76 #41 AND #46 AND #75

77 #76 AND ([conference review]/lim OR [editorial]/lim OR [letter]/lim OR [review]/lim)

78 #76 AND [animals]/lim NOT ([animals]/lim AND [humans]/lim)

79 #77 OR #78

80 #76 NOT #79

81 #80 AND [english]/lim

Sources: Adapted from MAH submission dossier



Table A2. Original SLR search strategy for MEDLINE® InProcess (via the PubMed.com platform)

# Search terms

1 Search "Multiple Sclerosis, Chronic Progressive" OR "chronic progressive multiple sclerosis"

2 Search (((disseminated[Title/Abstract] or insular[Title/Abstract] or multiple[Title/Abstract])) AND (sclerosis[Title/Abstract])) AND ("secondary progressive" OR progressive OR secondary or non-relapsing)

3 Search ("secondary progressive" OR "secondary-progressive" OR progressive OR secondary OR deteriorat* OR "non relapsing" OR non-relapsing) AND (ms OR "multiple sclerosis" OR "disseminated sclerosis" OR "encephalomyelitis disseminate" OR "chariot disease" OR "insular sclerosis")

4 Search spms or cpms

5 Search "progressive secondary multiple sclerosis" or "secondary progressive multiple sclerosis"

6 Search #1 OR #2 OR #3 OR #4 OR #5

7 Search "clinical trial" OR "clinical trials"

8 Search "randomized controlled trial"

9 Search "Random Allocation"

10 Search "Double Blind"

11 Search "Single Blind"

12 Search "phase i" OR phasei OR "Phase 1" OR phase1

13 Search "phase ii" OR phaseii OR "Phase 2" OR phase2

14 Search "phase iii" OR phaseiii OR "Phase 3" OR phase3

15 Search "phase iv" OR phaseiv OR "Phase 4" OR phase4

16 Search "controlled clinical trial" OR "controlled clinical trials"

17 Search "multicenter study"

18 Search "Observational Study" OR "Comparative Study" OR "Cross-Over Studies" OR "Cross-Over Study" OR "Prospective Studies" OR "Prospective Study"

19 Search "Cohort Studies" OR "Cohort Study" OR "Longitudinal Studies" OR "Longitudinal Study" OR "Follow-Up Studies" Or "Follow-Up Study"

20 Search "Clinical Study" OR "Historically Controlled Study" OR "Retrospective Study" OR "Retrospective Studies"

21 Search Survival

22 Search placebo*

23 Search "clinical trial"[Title/Abstract]

24 Search ((singl*[Title/Abstract] or doubl*[Title/Abstract] or treb*[Title/Abstract] or tripl*[Title/Abstract])) AND (blind* or mask*[Title/Abstract])

25 Search "randomly allocated"[Title/Abstract]

26 Search (allocated AND random*)

27 Search "randomised controlled trial" OR "randomized controlled trial" OR "randomised controlled trials" OR "randomized controlled trials"

28 Search "randomisation" OR "randomization" OR random*

29 Search rct

30 Search ((case*[Title/Abstract])) AND (control*[Title/Abstract])

31 Search (("follow up"[Title/Abstract] or followup[Title/Abstract])) AND (study[Title/Abstract] or studies[Title/Abstract])

32 Search ((clinical[Title/Abstract])) AND (trial*[Title/Abstract])

33 Search ((observational[Title/Abstract] or cohort[Title/Abstract])) AND (study[Title/Abstract] or studies[Title/Abstract])



# Search terms

34 Search #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #33 OR #33

35 Search case report[Title/Abstract]

36 Search letter

37 Search #35 OR #36

38 Search #34 NOT #37

39 Search “disease modifying therapies” [Title/Abstract] OR dmt*[Title/Abstract] OR "disease modifying therapy"[Title/Abstract] OR "disease modifying drugs"[Title/Abstract] OR "disease modifying drug"[Title/Abstract] OR dmd*[Title/Abstract]

40 Search "Glatiramer Acetate" OR “cop 1” or copaxone or “copolymer 1” or “copolymer cop 1” or “copolymer i” or glatiramer or glatopa or “tv 5010” or tv5010

41 Search "Alemtuzumab" OR “campath 1” or “campath 1h” or (cd52 AND “monoclonal antibody”) or “ldp 103” or ldp103 or lemtrada or mabcampath

42 Search "Mitoxantrone" OR “cl 232, 315” or “cl 232315” or “cl232, 315” or cl232315 or dhad or dhaq or domitrone or elsep or formyxan or genefadrone or misostol or mitoxanthrone or mitoxantron or mitoxantrona or mitoxgen or mitozantrone or mitroxantrone or mitroxone or neotalem or norexan or novanthron or novantron or novantrone or “now 85 34” or “now 8534” or now8534 or “nsc 279836” or “nsc 301739” or “nsc 301739d” or nsc279836 or nsc301739 or nsc301739d or oncotron or onkotrone or ralenova

43 Search "Rituximab" OR blitzima or “ct p10” or “ctp10” or “idec 102” or “idec c2b8” or “idec102” or idecc2b8 or mabthera or “r 105” or r105 or reditux or “rg 105” or rg105 or ritemvia or rituxan or rituxin or rituzena or rixathon or riximyo or “ro 452294” or ro452294 or truxima or tuxella

44 Search Fingolimod OR “fty 720” or fty720 or gilenia or gilenya

45 Search Natalizumab OR “an 100226” or an100226 or antegren or tysabri

46 Search "Dimethyl Fumarate" OR dimethylfumarate or “fag 201” or fag201 or panaclar or psorinovo or skilarence or tecfidera or dmf

47 Search Cladribine OR intocel or leustat or leustatin or leustatine or litak or litax or mavenclad or movectro or mylinax or “rwj 26251” or rwj26251

48 Search teriflunomide or “a 771726” or “a77 1726” or a771726 or aubagio or “hmr 1726” or hmr1726 or “rs 61980” or rs61980 or “su 0020” or su0020

49 Search "Interferon-beta" OR “recombinant beta interferon” or ((interferon or ifn) AND beta) or belerofon or ifn?beta or (beta1* AND interferon) or “interferon beta1” or “beta-1 interferon” or “beta 1 interferon”

50 Search "Interferon beta-1a" OR “beta1a interferon” or “beta interferon 1a” or avonex or ((interferon or ifn) AND (“beta 1a” or “beta-1a” or beta1a or “beta 1b” or beta1b or “beta 1b”)) or rebif or “rifn beta”

51 Search "Interferon beta-1b" OR “interferon beta serine” or “beta interferon 1b” or beneseron or betaferon or betaseron or extavia or “rifn beta-1b” or “sh 579” or sh579 or “zk 157046” or zk157046

52 Search masitinib or “ab 1010” or ab1010 or kinaction or masatinib or masican or masipro or masivet or masiviera

53 Search siponimod or “baf 312” or baf312 or mayzent

54 Search mis416

55 Search “imilecleucel t” or tcelna or tovaxin

56 Search biotin OR biotine OR md1003

57 Search “ibudilast” or “av 411” or av411 or “kc 404” or kc404 or ketas OR mn-166

58 Search ocrelizumab or ocrevus or “pro 70769” or pro70769 or “rhumab 2h7”

59 Search "peginterferon beta-1a" OR peginterferon or “peginterferon beta1a” or “beta 1a peginterferon” or “beta1a peginterferon” or “biib 017” or “biib017” or “peginterferon beta 1a” or “peginterferon beta-1a” or “pegylated human interferon beta 1a” or “pegylated interferon beta 1a” or “pegylated interferon beta-1a” or “pegylated interferon beta1a” or plegridy



# Search terms

60 Search idebenone OR avan OR cerestabon OR "cv 2619" OR cv2619 OR mnesis OR "qsa 10" OR qsa10 OR raxone OR "snt mc17" OR sovrima

61 Search opicinumab OR "biib 033" OR biib033

62 Search “Stem Cell Transplantation” OR “Hematopoietic Stem Cell Transplantation” OR “Haematopoietic Stem Cell Transplantation”

63 Search ("stem cell" AND (therap* OR transplant*)) OR hsct

64 Search simvastatin OR avastinee OR belmalip OR cholestat OR clinfar OR colastatina OR colemin OR colestricon OR covastin OR denan OR epistatin OR esvat OR ethicol OR eucor OR ifistatin OR jabastatina OR kavelor OR klonastin OR kolestevan OR "l 644128" OR l644128 OR lipcut OR lipecor OR lipex OR lipinorm OR liponorm OR lipovas OR lodales OR medipo OR mersivas OR "mk 733" OR mk733 OR "nor vastina" OR normofat OR orovas OR pantok OR rechol OR simbado OR simcard OR simchol OR simovil OR simtin OR simva OR simvacor OR simvahex OR simvalord OR simvastar OR simvastatina OR simvastatine OR simvata OR simvatin OR simvor OR simvotin OR sinvacor OR sinvastatin OR sinvinolin OR sivastin OR starzoco OR synvinolin OR torio OR valemia OR vasilip OR vasotenal OR vazim OR vidastat OR zimmex OR zocor OR zocord OR zovast

65 Search riluzole OR "pk 26124" OR pk26124 OR rilutek OR "rp 54274" OR rp54274

66 Search fluoxetine OR actan OR adofen OR "alzac 20" OR andep OR ansilan OR "atd 20" OR auroken OR auscap OR captaton OR "compound 110140" OR daforin OR depren OR deprexin OR deprizac OR deproxin OR elizac OR floxet OR fluctin OR fluctine OR fludac OR flufran OR fluketin OR flunil OR flunirin OR fluohexal OR fluox OR 'fluox puren' OR fluoxac OR fluoxeren OR fluoxifar OR fluoxil OR fluronin OR flusac OR flutin OR flutine OR fluxen OR fluxet OR fluxetil OR fluxetin OR fontex OR foxetin OR foxtin OR fropine OR fuloren OR lanclic OR "lilly 110140" OR lilly110140 OR lorien OR lovan OR "ly 110140" OR ly110140 OR magrilan OR margrilan OR modipran OR nopres OR nuzak OR oxedep OR plinzene OR pragmaten OR prizma OR proctin OR prodep OR prosac OR prozac OR prozamin OR qualisac OR rapiflux OR rowexetina OR salipax OR sanzur OR sarafem OR selfemra OR sinzac OR zactin OR zepax

67 Search amiloride OR amiclaran OR amikal OR "amilo 5" OR amilorid OR amiloridehydrochlorhydrate OR amiloridine OR amipramidine OR amyloride OR arumil OR berkamil OR colectril OR guanamprazine OR kaluril OR medamor OR midamor OR "mk 870" OR modamide OR nirulid OR pandiuren

68 Search #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 OR #57 OR #58 OR #59 OR #60 OR #61 OR #62 OR #63 OR #64 OR #65 OR #66 OR #67

69 Search #6 AND #38 AND #68

70 Search #69 AND (inprocess[sb] OR pubstatusaheadofprint)


Table A3. Update SLR search terms for MEDLINE Daily, MEDLINE In-Process and Epub Ahead of Print (searched via Ovid SP)

# Search terms

1 exp clinical trial/

2 randomized controlled trial/

3 controlled clinical trial/

4 random allocation/

5 comparative study/

6 single-blind method/

7 double blind method/

8 crossover studies/

9 Placebo/ or placebo$.mp.

10 (clinical trial or clinical trials).mp.

11 (controlled clinical trial or controlled clinical trials).mp.



# Search terms

12 (randomised controlled trial or randomized controlled trial or randomised controlled trials or randomized controlled trials).mp.

13 (randomisation or randomization).mp.

14 rct.mp.

15 (random allocation or randomly allocated or allocated randomly).mp.

16 ((allocated adj2 random) or (random$ adj1 assign) or random$).mp.

17 exp prospective studies/

18 ((single or double or triple or treble) adj1 (blind$ or mask$)).mp.

19 exp cohort studies/ or cohort$.ti,ab.

20 exp longitudinal studies/

21 exp multicenter study/

22 exp follow up studies/

23 exp case control studies/ or (case$ adj control$).ti,ab.

24 exp survival/

25 ((follow up or followup) adj (study or studies)).ti,ab.

26 (clinical adj trial$).ti,ab.

27 exp retrospective studies/

28 exp case control studies/

29 ((observational or cohort) adj (study or studies)).ti,ab.

30 or/1-29

31 case reports/

32 case report.mp.

33 letter/

34 or/31-33

35 30 not 34

36 ((((disseminated or insular or multiple) adj4 sclerosis) or ms) and (secondary progressive or progressive or non-relapsing)).ti,ab.

37 (Secondary progressive or secondary-progressive or progressive or secondary or deteriorat$ or non relapsing or non-relapsing) adj3 (ms or multiple sclerosis or disseminated sclerosis or encephalomyelitis disseminata or chariot disease or insular sclerosis).ti,ab,kf.

38 (spms or cpms).mp.

39 (progressive secondary multiple sclerosis or secondary progressive multiple sclerosis).mp.

40 or/36-39

41 disease modifying therapies.ti,ab. or dmt$.ti,ab. or disease modifying therapy.ti,ab. or disease modifying drugs.ti,ab. or disease modifying drug.ti,ab. or dmd$.ti,ab.

42 glatiramer acetate/ or (cop 1 or copaxone or copolymer 1 or copolymer cop 1 or copolymer i OR (glatiramer adj1 (acetate or sodium)) or glatopa or 'tv 5010' or tv5010).ti,ab,kf.

43 alemtuzumab/ or (alemtuzumab or campath 1 or campath 1h or (cd52 adj1 monoclonal antibody) or ldp 103 or ldp103 or lemtrada or mabcampath).ti,ab,kf.

44 mitoxantrone/ or (mitoxantrone or 'cl 232, 315' or 'cl 232315' or 'cl232, 315' or cl232315 or dhad or dhaq or domitrone or elsep or formyxan or genefadrone or misostol or mitoxanthrone or mitoxantron or mitoxantrona or mitoxgen or mitozantrone or mitroxantrone or mitroxone or neotalem or norexan or novanthron or novantron or novantrone or 'now 85 34' or 'now 8534' or now8534 or 'nsc 279836' or 'nsc 301739' or 'nsc 301739d' or nsc279836 or nsc301739 or nsc301739d or oncotron or onkotrone or ralenova).ti,ab,kf.



# Search terms

45 rituximab/ or (rituximab or blitzima or ct p10 or ctp10 or 'idec 102' or idec c2b8 or idec102 or idecc2b8 or mabthera or r 105 or r105 or reditux or 'rg 105' or rg105 or ritemvia or rituxan or rituxin or rituzena or rixathon or riximyo or 'ro 452294' or ro452294 or truxima or tuxella).ti,ab,kf.

46 fingolimod/ or (fingolimod or 'fty 720' or fty720 or gilenia or gilenya).ti,ab,kf.

47 natalizumab/ or (natalizumab or 'an 100226' or an100226 or antegren or Tysabri).ti,ab,kf.

48 dimethyl fumarate/ or (bg 12 or bg12 or dimethyl fumarate or dimethylfumarate or fag 201 or fag201 or panaclar or psorinovo or skilarence or tecfidera or dmf).ti,ab,kf.

49 (teriflunomide or a 771726 or a77 1726 or a77-1726 or a771726 or aubagio or hmr 1726 or hmr1726 or rs 61980 or rs61980 or su 0020 or su0020).ti,ab,kf.

50 cladribine/ or (cladribine or intocel or leustat or leustatin or leustatine or litak or litax or mavenclad or movectro or mylinax or rwj 26251 or rwj26251020).ti,ab,kf.

51 interferon-beta/ or (recombinant beta interferon or ((interferon or ifn) adj2 beta) or belerofon or ifn?beta or (beta1$ adj2 interferon) or interferon beta1 or beta-1 interferon or 'beta 1 interferon).ti,ab,kf.

52 (beta1a interferon or beta interferon 1a or avonex or (((interferon or ifn) adj1 beta 1a or beta-1a or beta1a or beta 1b or beta1b or beta 1b)) or rebif or rifn beta).ti,ab,kf.

53 (interferon beta serine or beta interferon 1b or beneseron or betaferon or betaseron or extavia or rifn beta-1b or sh 579 or sh579 or zk 157046 or zk157046).ti,ab,kf.

54 (masitinib or ab 1010 or ab1010 or kinaction or masatinib or masican or masipro or masivet or masiviera).ti,ab,kf.

55 siponimod/ or (siponimod or 'baf 312' or baf312 or mayzent).ti,ab,kf.

56 mis416.ti,ab,kf.

57 (imilecleucel t or tcelna or tovaxin).ti,ab,kf.

58 exp biotin/ or (biotin or biotine or md1003).ti,ab,kf.

59 (ibudilast or av 411 or av411 or kc 404 or kc404 or ketas or mn-166).ti,ab,kf.

60 (ocrelizumab or ocrevus or pro 70769 or pro70769 or rhumab 2h7).ti,ab,kf.

61 (peginterferon or peginterferon beta1a or peginterferon or beta 1a peginterferon or beta1a peginterferon or biib 017 or biib017 or peginterferon beta 1a or peginterferon beta-1a or 'pegylated human interferon beta 1a or pegylated interferon beta 1a or pegylated interferon beta-1a or pegylated interferon beta1a or plegridy).ti,ab,kf.

62 (idebenone or avan or cerestabon or cv 2619 or cv2619 or mnesis or qsa 10 or qsa10 or raxone or snt mc17 or sovrima).ti,ab,kf.

63 (opicinumab or biib 033 or biib033).ti,ab,kf.

64 exp stem cell transplantation/ or hematopoietic stem cell transplantation/ or ((stem cell adj2 (therap$ or transplant$)) or hsct).ti,ab,kf.

65 simvastatin/ or (simvastatin or avastinee or belmalip or cholestat or clinfar or colastatina or colemin or colestricon or covastin or denan or epistatin or esvat or ethicol or eucor or ifistatin or jabastatina or kavelor or klonastin or kolestevan or 'l 644128' or l644128 or lipcut or lipecor or lipex or lipinorm or liponorm or lipovas or lodales or medipo or mersivas or 'mk 733' or mk733 or 'nor vastina' or normofat or orovas or pantok or rechol or simbado or simcard or simchol or simovil or simtin or simva or simvacor or simvahex or simvalord or simvastar or simvastatina or simvastatine or simvata or simvatin or simvor or simvotin or sinvacor or sinvastatin or sinvinolin or sivastin or starzoco or synvinolin or torio or valemia or vasilip or vasotenal or vazim or vidastat or zimmex or zocor or zocord or zovast).ti,ab,kf.

66 riluzole/ or (riluzole or pk 26124 or pk26124 or rilutek or rp 54274 or rp54274).ti,ab,kf.

67 fluoxetine/ or (fluoxetine or actan or adofen or 'alzac 20' or andep or ansilan or 'atd 20' or auroken or auscap or captaton or 'compound 110140' or daforin or depren or deprexin or deprizac or deproxin or elizac or floxet or fluctin or fluctine or fludac or flufran or fluketin or flunil or flunirin or fluohexal or fluox or 'fluox puren' or fluoxac or fluoxeren or fluoxifar or fluoxil or fluronin or flusac or flutin or flutine or fluxen or fluxet or fluxetil or fluxetin or fontex or foxetin or foxtin or fropine or fuloren or lanclic or 'lilly 110140' or lilly110140 or lorien or lovan or 'ly 110140' or ly110140 or magrilan or margrilan or modipran or nopres or nuzak or oxedep or plinzene or pragmaten or prizma or proctin or prodep or prosac or prozac or prozamin or qualisac or rapiflux or rowexetina or salipax or sanzur or sarafem or selfemra or sinzac or zactin or



# Search terms

zepax).ti,ab,kf.

68 amiloride/ or (amiloride or amiclaran or amikal or 'amilo 5' or amilorid or amiloridehydrochlorhydrate or amiloridine or amipramidine or amyloride or arumil or berkamil or colectril or guanamprazine or kaluril or medamor or midamor or 'mk 870' or modamide or nirulid or pandiuren).ti,ab,kf.

69 or/41-68

70 35 and 40 and 69

71 ("conference review" or "editorial" or "letter" or "review").pt.

72 exp animals/ not (exp animals/ and exp humans/)

73 71 or 72

74 70 not 73

75 Limit 74 to english


Table A4. Update SLR search terms for Embase (searched via Ovid SP)

# Search terms

1 exp clinical trial/

2 randomized controlled trial/

3 controlled clinical trial/

4 phase 3 clinical trial/ or phase 2 clinical trial/ or phase 4 clinical trial/

5 randomization/

6 controlled study/

7 comparative study/

8 single blind procedure/

9 double blind procedure/

10 crossover procedure/

11 Placebo/ or placebo$.mp.

12 (clinical trial or clinical trials).mp.

13 (controlled clinical trial or controlled clinical trials).mp.

14 (randomised controlled trial or randomized controlled trial or randomised controlled trials or randomized controlled trials).mp.

15 (randomisation or randomization).mp.

16 rct.mp.

17 (random allocation or randomly allocated or allocated randomly).mp.

18 ((allocated adj2 random) or (random$ adj1 assign) or random$).mp.

19 exp prospective study/

20 (single or double or triple or treble) adj1 (blind$ or mask$).mp.

21 exp cohort analysis/ or cohort$.ti,ab.

22 exp longitudinal study/

23 exp multicentre study/

24 exp follow up/

25 major clinical study/



# Search terms

26 exp case control study/ or (case$ adj control$).ti,ab.

27 exp clinical article/

28 exp survival/

29 ((follow up or followup) adj (study or studies)).ti,ab.

30 (clinical adj trial$).ti,ab.

31 exp retrospective study/

32 exp case control study/

33 ((observational or cohort) adj (study or studies)).ti,ab.

34 exp intervention study/

35 or/1-34

36 case study/

37 case report.mp.

38 abstract report/

39 letter/

40 or/36-39

41 35 not 40

42 ((((disseminated or insular or multiple) adj4 sclerosis).ti,ab.) or ms.ti,ab.) and (secondary progressive.ti,ab. or progressive.ti,ab. or non-relapsing.ti,ab.)

43 (Secondary progressive or secondary-progressive or progressive or secondary or deteriorat$ or non relapsing or non-relapsing) adj3 (ms or multiple sclerosis or disseminated sclerosis or encephalomyelitis disseminata or chariot disease or insular sclerosis).mp.

44 (spms or cpms).mp.

45 (progressive secondary multiple sclerosis or secondary progressive multiple sclerosis).mp.

46 or/42-45

47 disease modifying therapies.ti,ab. or dmt$.ti,ab. or disease modifying therapy.ti,ab. or disease modifying drugs.ti,ab. or disease modifying drug.ti,ab. or dmd$.ti,ab.

48 glatiramer acetate/ or (cop 1 or copaxone or copolymer 1 or copolymer cop 1 or copolymer i or (glatiramer adj1 (acetate or sodium)) or glatopa or 'tv 5010' or tv5010).ti,ab,kw.

49 alemtuzumab/ or (alemtuzumab or campath 1 or campath 1h or (cd52 adj1 monoclonal antibody) or ldp 103 or ldp103 or lemtrada or mabcampath).ti,ab,kw.

50 mitoxantrone/ or (mitoxantrone or 'cl 232, 315' or 'cl 232315' or 'cl232, 315' or cl232315 or dhad or dhaq or domitrone or elsep or formyxan or genefadrone or misostol or mitoxanthrone or mitoxantron or mitoxantrona or mitoxgen or mitozantrone or mitroxantrone or mitroxone or neotalem or norexan or novanthron or novantron or novantrone or 'now 85 34' or 'now 8534' or now8534 or 'nsc 279836' or 'nsc 301739' or 'nsc 301739d' or nsc279836 or nsc301739 or nsc301739d or oncotron or onkotrone or ralenova).ti,ab,kw.

51 rituximab/ or (rituximab or blitzima or ct p10 or ctp10 or 'idec 102' or idec c2b8 or idec102 or idecc2b8 or mabthera or r 105 or r105 or reditux or 'rg 105' or rg105 or ritemvia or rituxan or rituxin or rituzena or rixathon or riximyo or 'ro 452294' or ro452294 or truxima or tuxella).ti,ab,kw.

52 fingolimod/ or (fingolimod or 'fty 720' or fty720 or gilenia or gilenya).ti,ab,kw.

53 natalizumab/ or (natalizumab or 'an 100226' or an100226 or antegren or Tysabri).ti,ab,kw.

54 dimethyl fumarate/ or (bg 12 or bg12 or dimethyl fumarate or dimethylfumarate or fag 201 or fag201 or panaclar or psorinovo or skilarence or tecfidera or dmf).ti,ab,kw.

55 (teriflunomide or a 771726 or a77 1726 or a77-1726 or a771726 or aubagio or hmr 1726 or hmr1726 or rs 61980 or rs61980 or su 0020 or su0020).ti,ab,kw.

56 cladribine/ or (cladribine or intocel or leustat or leustatin or leustatine or litak or litax or mavenclad or



# Search terms

movectro or mylinax or rwj 26251 or rwj26251020).ti,ab,kw.

57 interferon-beta/ or (recombinant beta interferon or ((interferon or ifn) adj2 beta) or belerofon or ifn?beta or (beta1$ adj2 interferon) or interferon beta1 or beta-1 interferon or 'beta 1 interferon).ti,ab,kw.

58 beta1a interferon/ or (avonex or ((interferon or ifn) adj1 (beta 1a or beta-1a or beta1a or beta 1b or beta1b or beta 1b)) or rebif or rifn beta).ti,ab,kw.

59 interferon beta serine/ or (interferon beta serine or beneseron or betaferon or betaseron or extavia or rifn beta-1b or sh 579 or sh579 or zk 157046 or zk157046).ti,ab,kw.

60 (masitinib or ab 1010 or ab1010 or kinaction or masatinib or masican or masipro or masivet or masiviera).ti,ab,kw.

61 siponimod/ or (siponimod or 'baf 312' or baf312 or mayzent).ti,ab,kw.

62 mis416.ti,ab,kw.

63 (imilecleucel t or tcelna or tovaxin).ti,ab,kw.

64 exp biotin/ or (biotin or biotine or md1003).ti,ab,kw.

65 (ibudilast or av 411 or av411 or kc 404 or kc404 or ketas or mn-166).ti,ab,kw.

66 (ocrelizumab or ocrevus or pro 70769 or pro70769 or rhumab 2h7).ti,ab,kw.

67 peginterferon/ or peginterferon beta1a/ or (peginterferon or peginterferon beta1a or peginterferon or beta 1a peginterferon or beta1a peginterferon or biib 017 or biib017 or peginterferon beta 1a or peginterferon beta-1a or 'pegylated human interferon beta 1a or pegylated interferon beta 1a or pegylated interferon beta-1a or pegylated interferon beta1a or plegridy).ti,ab,kw.

68 idebenone/ or (idebenone or avan or cerestabon or cv 2619 or cv2619 or mnesis or qsa 10 or qsa10 or raxone or snt mc17 or sovrima).ti,ab,kw.

69 opicinumab/ or (opicinumab or biib 033 or biib033).ti,ab,kw.

70 exp stem cell transplantation/ or hematopoietic stem cell transplantation/ or ((stem cell adj2 (therap$ or transplant$)) or hsct).ti,ab,kw.

71 simvastatin/ or (simvastatin or avastinee or belmalip or cholestat or clinfar or colastatina or colemin or colestricon or covastin or denan or epistatin or esvat or ethicol or eucor or ifistatin or jabastatina or kavelor or klonastin or kolestevan or 'l 644128' or l644128 or lipcut or lipecor or lipex or lipinorm or liponorm or lipovas or lodales or medipo or mersivas or 'mk 733' or mk733 or 'nor vastina' or normofat or orovas or pantok or rechol or simbado or simcard or simchol or simovil or simtin or simva or simvacor or simvahex or simvalord or simvastar or simvastatina or simvastatine or simvata or simvatin or simvor or simvotin or sinvacor or sinvastatin or sinvinolin or sivastin or starzoco or synvinolin or torio or valemia or vasilip or vasotenal or vazim or vidastat or zimmex or zocor or zocord or zovast).ti,ab,kw.

72 riluzole/ or (riluzole or pk 26124 or pk26124 or rilutek or rp 54274 or rp54274).ti,ab,kw.

73 fluoxetine/ or (fluoxetine or actan or adofen or 'alzac 20' or andep or ansilan or 'atd 20' or auroken or auscap or captaton or 'compound 110140' or daforin or depren or deprexin or deprizac or deproxin or elizac or floxet or fluctin or fluctine or fludac or flufran or fluketin or flunil or flunirin or fluohexal or fluox or 'fluox puren' or fluoxac or fluoxeren or fluoxifar or fluoxil or fluronin or flusac or flutin or flutine or fluxen or fluxet or fluxetil or fluxetin or fontex or foxetin or foxtin or fropine or fuloren or lanclic or 'lilly 110140' or lilly110140 or lorien or lovan or 'ly 110140' or ly110140 or magrilan or margrilan or modipran or nopres or nuzak or oxedep or plinzene or pragmaten or prizma or proctin or prodep or prosac or prozac or prozamin or qualisac or rapiflux or rowexetina or salipax or sanzur or sarafem or selfemra or sinzac or zactin or zepax).ti,ab,kw.

74 amiloride/ or (amiloride or amiclaran or amikal or 'amilo 5' or amilorid or amiloridehydrochlorhydrate or amiloridine or amipramidine or amyloride or arumil or berkamil or colectril or guanamprazine or kaluril or medamor or midamor or 'mk 870' or modamide or nirulid or pandiuren).ti,ab,kw.

75 or/47-74

76 41 and 46 and 75

77 ("conference review" or "editorial" or "letter" or "review").pt.

78 exp animals/ not (exp animals/ and exp humans/)

79 77 or 78



# Search terms

80 76 not 79

81 Limit 80 to english


Table A5. Original and Update SLR search strategy for CENTRAL and CDSR (via the Wiley Online Platform)

# Search terms

1 MeSH descriptor: [Multiple Sclerosis, Chronic Progressive] explode all trees

2 ((((disseminated OR insular OR multiple) NEAR/4 sclerosis):ab,ti,kw) OR ms:ab,ti,kw) AND (“secondary progressive”:ab,ti,kw OR progressive:ab,ti,kw OR non-relapsing:ab,ti,kw)

3 ("secondary progressive" OR "secondary-progressive" OR progressive OR secondary OR deteriorat* OR "non relapsing" OR non-relapsing) NEAR/3 (ms OR "multiple sclerosis" OR "disseminated sclerosis" OR "encephalomyelitis disseminate" OR "chariot disease" OR "insular sclerosis")

4 spms OR cpms

5 progressive secondary multiple sclerosis OR "secondary progressive multiple sclerosis"

6 #1 OR #2 OR #3 OR #4 OR #5

7 disease modifying therapies:ab,ti,kw OR dmt*:ab,ti,kw OR "disease modifying therapy":ab,ti,kw OR "disease modifying drugs":ab,ti,kw OR "disease modifying drug":ab,ti,kw OR dmd*:ab,ti,kw

8 MeSH descriptor: [Glatiramer Acetate] explode all trees

9 glatiramer OR “cop 1” or copaxone or “copolymer 1” or “copolymer cop 1” or “copolymer i” or (glatiramer near/1 (acetate or sodium)) or glatopa or “tv 5010” or tv5010

10 MeSH descriptor: [Alemtuzumab] explode all trees

11 Alemtuzumab OR “campath 1” or “campath 1h” or (cd52 near/1 “monoclonal antibody”) or “ldp 103” or ldp103 or lemtrada or mabcampath

12 MeSH descriptor: [Mitoxantrone] explode all trees

13 Mitoxantrone or “cl 232, 315” or “cl 232315” or “cl232, 315” or cl232315 or dhad or dhaq or domitrone or elsep or formyxan or genefadrone or misostol or mitoxanthrone or mitoxantron or mitoxantrona or mitoxgen or mitozantrone or mitroxantrone or mitroxone or neotalem or norexan or novanthron or novantron or novantrone or “now 85 34” or “now 8534” or now8534 or “nsc 279836” or “nsc 301739” or “nsc 301739d” or nsc279836 or nsc301739 or nsc301739d or oncotron or onkotrone or ralenova

14 MeSH descriptor: [Rituximab] explode all trees

15 Rituximab or blitzima or “ct p10” or “ctp10” or “idec 102” or “idec c2b8” or “idec102” or idecc2b8 or mabthera or “r 105” or r105 or reditux or “rg 105” or rg105 or ritemvia or rituxan or rituxin or rituzena or rixathon or riximyo or “ro 452294” or ro452294 or truxima or tuxella

16 MeSH descriptor: [Fingolimod Hydrochloride] explode all trees

17 Fingolimod or “fty 720” or fty720 or gilenia or gilenya

18 MeSH descriptor: [Natalizumab] explode all trees

19 Natalizumab or “an 100226” or an100226 or antegren or tysabri

20 MeSH descriptor: [Dimethyl Fumarate] explode all trees

21 “Dimethyl Fumarate” or dimethylfumarate or “fag 201” or fag201 or panaclar or psorinovo or skilarence or tecfidera or dmf

22 teriflunomide or “a 771726” or “a77 1726” or a771726 or aubagio or “hmr 1726” or hmr1726 or “rs 61980” or rs61980 or “su 0020” or su0020

23 MeSH descriptor: [Cladribine] explode all trees

24 Cladribine or intocel or leustat or leustatin or leustatine or litak or litax or mavenclad or movectro or mylinax or “rwj 26251” or rwj26251



# Search terms

25 MeSH descriptor: [Interferon-beta] explode all trees

26 “recombinant beta interferon” or ((interferon or ifn) near/2 beta) or belerofon or ifn?beta or (beta1* near/2 interferon) or “interferon beta1” or “beta-1 interferon” or “beta 1 interferon”

27 MeSH descriptor: [Interferon beta-1a] explode all trees

28 “beta1a interferon” or “beta interferon 1a” or avonex or ((interferon or ifn) near/1 (“beta 1a” or “beta-1a” or beta1a or “beta 1b” or beta1b or “beta 1b”)) or rebif or “rifn beta”

29 MeSH descriptor: [Interferon beta-1b] explode all trees

30 “interferon beta serine” or “beta interferon 1b” or beneseron or betaferon or betaseron or extavia or “rifn beta-1b” or “sh 579” or sh579 or “zk 157046” or zk157046

31 masitinib or “ab 1010” or ab1010 or kinaction or masatinib or masican or masipro or masivet or masiviera

32 siponimod or “baf 312” or baf312 or mayzent

33 mis416

34 imilecleucel-t OR tcelna OR tovaxin

35 MeSH descriptor: [Biotin] explode all trees

36 biotin OR biotine OR md1003

37 ibudilast OR "av 411" OR av411 OR "kc 404" OR kc404 OR ketas OR mn-166

38 ocrelizumab OR ocrevus OR "pro 70769" OR pro70769 OR "rhumab 2h7"

39 peginterferon or “peginterferon beta1a” or “beta 1a peginterferon” or “beta1a peginterferon” or “biib 017” or “biib017” or “peginterferon beta 1a” or “peginterferon beta-1a” or “pegylated human interferon beta 1a” or “pegylated interferon beta 1a” or “pegylated interferon beta-1a” or “pegylated interferon beta1a” or plegridy

40 idebenone OR avan OR cerestabon OR "cv 2619" OR cv2619 OR mnesis OR "qsa 10" OR qsa10 OR raxone OR "snt mc17" OR sovrima

41 opicinumab OR "biib 033" OR biib033

42 MeSH descriptor: [Stem Cell Transplantation] explode all trees

43 MeSH descriptor: [Hematopoietic Stem Cell Transplantation] explode all trees

44 ("stem cell" NEAR/2 (therap* OR transplant*)) OR hsct

45 MeSH descriptor: [Simvastatin] explode all trees

46 simvastatin OR avastinee OR belmalip OR cholestat OR clinfar OR colastatina OR colemin OR colestricon OR covastin OR denan OR epistatin OR esvat OR ethicol OR eucor OR ifistatin OR jabastatina OR kavelor OR klonastin OR kolestevan OR "l 644128" OR l644128 OR lipcut OR lipecor OR lipex OR lipinorm OR liponorm OR lipovas OR lodales OR medipo OR mersivas OR "mk 733" OR mk733 OR "nor vastina" OR normofat OR orovas OR pantok OR rechol OR simbado OR simcard OR simchol OR simovil OR simtin OR simva OR simvacor OR simvahex OR simvalord OR simvastar OR simvastatina OR simvastatine OR simvata OR simvatin OR simvor OR simvotin OR sinvacor OR sinvastatin OR sinvinolin OR sivastin OR starzoco OR synvinolin OR torio OR valemia OR vasilip OR vasotenal OR vazim OR vidastat OR zimmex OR zocor OR zocord OR zovast

47 MeSH descriptor: [Riluzole] explode all trees

48 riluzole OR "pk 26124" OR pk26124 OR rilutek OR "rp 54274" OR rp54274

49 MeSH descriptor: [Fluoxetine] explode all trees

50 fluoxetine OR actan OR adofen OR "alzac 20" OR andep OR ansilan OR "atd 20" OR auroken OR auscap OR captaton OR "compound 110140" OR daforin OR depren OR deprexin OR deprizac OR deproxin OR elizac OR floxet OR fluctin OR fluctine OR fludac OR flufran OR fluketin OR flunil OR flunirin OR fluohexal OR fluox OR 'fluox puren' OR fluoxac OR fluoxeren OR fluoxifar OR fluoxil OR fluronin OR flusac OR flutin OR flutine OR fluxen OR fluxet OR fluxetil OR fluxetin OR fontex OR foxetin OR foxtin OR fropine OR fuloren OR lanclic OR "lilly 110140" OR lilly110140 OR lorien OR lovan OR "ly 110140" OR ly110140 OR magrilan OR margrilan OR modipran OR nopres OR nuzak OR oxedep OR plinzene OR pragmaten OR prizma OR proctin OR prodep OR prosac OR prozac OR prozamin OR qualisac OR rapiflux OR rowexetina OR salipax OR sanzur OR sarafem OR selfemra OR sinzac OR



# Search terms

zactin OR zepax

51 MeSH descriptor: [Amiloride] explode all trees

52 amiloride OR amiclaran OR amikal OR "amilo 5" OR amilorid OR amiloridehydrochlorhydrate OR amiloridine OR amipramidine OR amyloride OR arumil OR berkamil OR colectril OR guanamprazine OR kaluril OR medamor OR midamor OR "mk 870" OR modamide OR nirulid OR pandiuren

53 #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52

54 #6 and #53 (Word variations have been searched)

55 #54 in Trials (Word variations have been searched)

56 #54 in Cochrane Reviews, Cochrane Protocols (Word variations have been searched)


Table A6. Original SLR search strategy for DARE and HTAD (via the CRD platform)

# Search terms

1 MeSH DESCRIPTOR Multiple Sclerosis, Chronic Progressive EXPLODE ALL TREES

2 ((disseminated or insular or multiple) near4 sclerosis) AND (secondary progressive OR progressive OR secondary)

3 ((secondary progressive OR secondary-progressive OR progressive OR secondary OR deteriorat* OR non relapsing OR non-relapsing) NEAR3 (ms OR multiple sclerosis OR disseminated sclerosis OR encephalomyelitis disseminate OR chariot disease OR insular sclerosis))

4 spms OR cpms

5 progressive secondary multiple sclerosis OR secondary progressive

multiple sclerosis

6 #1 OR #2 OR #3 OR #4 OR #5 (NHS EED/DARE/HTA)

7 (#6) IN DARE

8 (#6) IN HTA

Sources: Adapted from MAH submission dossier *One empty record was retrieved so 62 hits



Table A7. Congress searches for the clinical SLR update

Congress Site Search Strategy

Americas Committee for Treatment and Re-search in Multiple Sclerosis (ACTRIMS)

https://actrims.confex.com/actrims/2019/meet-ingapp.cgi/Home/0

Using the search function: “SPMS or secondary or relapsing”


Table A8. Search terms used for other grey literature searches in the clinical SLR update

Source Site Search Strategy

All Wales Medicines Strategy Group (AWMSG)

http://www.awmsg.org/awmsgonline/app/browsebnf?execution=e13s1

Using the 'Browse by BNF category' function in combination with searching for "multiple sclerosis" using browser Find function; results published since the original search were reviewed

Canadian Agency for Drugs and Technologies in Health (CADTH)

https://www.cadth.ca/ Searches conducted for “secondary progressive multiple sclerosis” and terms for each intervention

Centre for Reviews and Dissemination (CRD) https://www.york.ac.uk/crd/) Searches conducted for “secondary progressive multiple sclerosis” and terms for each intervention

ClinicalTrials.gov https://www.clinicaltrials.gov/ Disease area: secondary progressive multiple sclerosis

Last updated from: 17.10.18

EU Clinical Trials Register (EU CTR) https://www.clinicaltrialsregister.eu/ctr-search/search

Search Terms:

Multiple Sclerosis

Date limit: Posted since 17.10.18

European Medicines Agency (EMA)

https://www.ema.europa.eu Searches conducted for medicines listed in the ‘Multiple Sclerosis’ section of the database and revisions since the 17th October 2019

Haute Autorite de Sante (HAS) https://www.has-sante.fr/ Searches conducted for “multiple sclerosis” and terms for each inter-vention

Institute for Quality and Efficiency in Health Care (IQWiG)

https://www.iqwig.de/ Searches conducted for “multiple sclerosis” and terms for each inter-vention

http://www.ema.europa.eu/ema/index.jsp?curl=pages%2Fmedicines%2Flanding%2Fsmop_search.jsp&mid=WC0b01ac058001d127&searchkwByEnter=false&alreadyLoaded=true&isNewQuery=true&status=Positive&status=Negative&applicationType=Initial+authorisation&applicationType=



Source Site Search Strategy

National Institute for Health and Care Excel-lence (NICE)

https://www.nice.org.uk/ Searches conducted for "secondary progressive multiple sclerosis", results published since the original search were reviewed

NIHR HTA Programme https://www.nihr.ac.uk/ Searches for “multiple sclerosis”, after September 2018

Scottish Medicines Consortium (SMC) https://www.scottishmedicines.org.uk/ Searches conducted for "secondary progressive multiple sclerosis" and each eligible intervention

Utility-weight collection collated by Tufts New England Medical Center's Catalogue of Prefer-ence Scores

http://healtheconomics.tuftsmedi-calcenter.org/cear4/aboutus/whatisthecearegis-try.aspx

Searches for each relevant intervention

WHO ICTRP http://apps.who.int/trialsearch/ Searches conducted for “multiple sclerosis” in title or condition, 17/10/18 to 02/04/19


Studies included following full-text review

Table A9. Publications included in the original SLR

# Author, Year Reference

EXPAND trial (NCT01665144)

1 Kappos 2018 Kappos L, Bar-Or A, Cree BAC, et al. Siponimod versus placebo in secondary progressive multiple sclerosis (EXPAND): a double-blind, randomised, phase 3 study. The Lancet 2018;391:1263-1273.

2 Cree 2018 Cree B, Fox R, Giovannoni G, et al. Siponimod affects disability progression in SPMS patients independent of relapse activity: Results from the phase III EXPAND study. European Journal of Neurology 2018;25:70.

3 Benedict 2018a Benedict RH, Cree B, Tomic D, et al. Siponimod improves cognitive processing speed in patients with SPMS: Results from Phase 3 EXPAND Study. European Journal of Neurology 2018;25:432.

4 Benedict 2018b Benedict RHB, Cree B, Tomic D, et al. Impact of siponimod on cognition in patients with secondary progressive multiple sclerosis: Results from phase 3 expand study. Neurology 2018;90.

5 Kuhle 2018 Kuhle J, Kropshofer H, Barro C, et al. Siponimod reduces neurofilament light chain blood levels in secondary progressive multiple sclerosis patients. Neurology 2018;90.



6 Cree 2018 Cree B, Fox R, Giovannoni G, et al. Uncoupling the impact on relapses and disability progression: Siponimod in relapsing and non-relapsing patients with secondary progressive multiple sclerosis in the phase III expand study. Neurology 2018;90.

7 Bar 2017 Bar-Or A, Derfuss T, Vermersch P, et al. Longitudinal changes in lymphocyte subsets of siponimodtreated patients with SPMS. Multiple Sclerosis Journal 2017;23:660.

8 Mao 2017 Mao-Draayer Y, Wu Q, Wang Q, et al. Basic immunological profile changes of secondary progressive multiple sclerosis patients treated with BAF312 (SIPONIMOD). Journal of the Neurological Sciences 2017;381:783.

9 Gold 2017 Gold R, Giovannoni G, Cree B, et al. Impact of primary endpoint definitions and patient baseline characteristics on study outcomes in progressive multiple sclerosis. Multiple Sclerosis Journal 2017;23:660-661.

10 Vermersch 2017 Vermersch P, Bar-Or A, Cree B, et al. The EXPAND study results: Efficacy of siponimod in secondary progressive multiple sclerosis. European Journal of Neurology 2017;24:44.

11 Giovannoni 2017 Giovannoni G, Baror A, Cree B, et al. The EXPAND study results: Safety and tolerability of siponimod in patients with secondary progressive multiple sclerosis. European Journal of Neurology 2017;24:495.

12 Fox 2017a Fox R, Kappos L, Bar-Or A, et al. Safety and tolerability of siponimod in patients with secondary progressive multiple sclerosis. Neurology 2017;88.

13 Kappos 2016a Kappos L, Bar-Or A, Cree B, et al. Baseline subgroup characteristics of expand: A phase 3 study of siponimod (BAF312) for the treatment of secondary progressive multiple sclerosis. Neurology 2016;86.

14 Kappos 2014 Kappos L, Bar-Or A, Cree B, et al. Siponimod (BAF312) for the treatment of secondary progressive multiple sclerosis: Design of the phase 3 expand trial. Multiple Sclerosis 2014;20:927-928.

15 Kappos 2013a Kappos L, Bar-Or A, Cree B, et al. Siponimod (BAF312) for the treatment of secondary progressive multiple sclerosis: Design of the phase 3 expand trial. Neurology 2013;80.

16 Kappos 2015 Kappos L, Bar-Or A, Cree B, et al. Siponimod (BAF312) for the treatment of secondary progressive multiple sclerosis (SPMS): baseline characteristics of the EXPAND study population. Multiple sclerosis. Volume 23, 2015:317‐318.

17 Kappos 2017a Kappos L, Bar-Or A, Cree B, et al. Efficacy of Siponimod in Secondary Progressive Multiple Sclerosis: Results of the Phase 3 Study. Neurology 2017;88.

18 Kappos 2013b Kappos L, Bar-Or A, Cree B, et al. Siponimod (BAF312) for the treatment of secondary progressive multiple sclerosis: design of the phase 3 expand trial. Volume 80, 2013.

19 Fox 2017b Fox R, Arnold D, Bar-Or A, et al. Effects of siponimod on MRI outcomes in patients with secondary progressive multiple sclerosis: results of the phase 3 EXPAND study. Multiple sclerosis journal. Conference: 7th joint ECTRIMS-ACTRIMS, MSPARIS2017. France. Volume 23, 2017:34‐35.



20 CSR CSR EXPAND

21 Kappos 2017b Kappos L, Vermersch P, Bar-Or A, et al. Efficacy of siponimod on disability progression in SPMS patients with and without on-study relapses. Multiple sclerosis journal. Conference: 7th joint ECTRIMS-ACTRIMS, MSPARIS2017. France. Volume 23, 2017:397‐398.

22 Kappos 2016b Kappos L, Bar-Or A, Cree B, et al. Efficacy and safety of siponimod in secondary progressive multiple sclerosis-Results of the placebo controlled, double-blind, Phase III EXPAND study. Multiple Sclerosis 2016;22:828-829.

ASCEND trial (NCT01416181)

23 Kapoor 2018a Kapoor R, Ho PR, Campbell N, et al. Effect of natalizumab on disease progression in secondary progressive multiple sclerosis (ASCEND): a phase 3, randomised, double-blind, placebo-controlled trial with an open-label extension. The Lancet Neurology 2018;17:405-415.

24 Giovannoni 2017a Giovannoni G, Freedman MS, Hartung HP, et al. Natalizumab improves walking and upper-limb disability compared with placebo in patients with secondary progressive multiple sclerosis: An integrated, post hoc area under the outcome-time curve analysis from the ASCEND trial. Multiple Sclerosis Journal 2017;23:36-37.

25 Giovannoni 2016 Giovannoni G, Steiner D, Sellebjerg F, et al. Sustained disability improvement in patients with secondary progressive multiple sclerosis (SPMS) assessed by a multicomponent endpoint: A post hoc analysis from the ASCEND study. Multiple Sclerosis 2016;22:671-672.

26 Kapoor 2016 Kapoor R, Steiner D, Miller A, et al. Subgroup analyses of natalizumab treatment response in ASCEND, a multicenter, double-blind, placebo-controlled, randomized phase 3 clinical trial in patients with secondary progressive multiple sclerosis (SPMS). Multiple Sclerosis 2016;22:874-875.

27 Steiner 2016 Steiner D, Arnold D, Freedman M, et al. Natalizumab versus placebo in patients with secondary progressive multiple sclerosis (SPMS): Results from ASCEND, a multicenter, double-blind, placebo-controlled, randomized phase 3 clinical trial. Neurology 2016;87:e22.

28 Cano 2015 Cano S, Cleanthous S, Marquis P, et al. Measuring the impact of secondary progressive multiple sclerosis (SPMS) in the ASCEND trial: Equating the MSIS-29, MSWS-12, ABILHAND-56 and SF-36. Value in Health 2015;18:A713.

29 Steiner 2015 Steiner D, Hartung HP, Kapoor R, et al. Increasing levels of disability on objective measures of ambulation and upper extremity function are associated with increasing levels of patient-reported impairment in secondary progressive multiple sclerosis patients: Baseline data from ASCEND. Multiple Sclerosis 2015;23:310-311.

30 Kapoor 2015 Kapoor R, Arnold D, Miller A, et al. Gray matter volume correlates with information processing as measured by the symbol digit modalities test (SDMT) but not with physical disability as measured by the expanded disability status scale (EDSS) in patients with secondary progressive multiple sclerosis (SPMS): Analysis of baseline correlations from the ascend natalizumab study. Neurology 2015;84.



31 Giovannoni 2017b Giovannoni G, Steiner D, Sellebjerg F, et al. Sustained disability improvement as assessed by a multicomponent endpoint in secondary progressive multiple sclerosis (SPMS) Patients: A post hoc analysis from ASCEND. Neurology 2017;88.

32 Cadavid 2013 Cadavid D, Brochet B, Mancardi G, et al. The MS-COG, a novel endpoint for measurement of cognitive function in multiple sclerosis clinical trials: baseline characteristics of the cognitive substudy of the ASCEND natalizumab secondary progressive multiple sclerosis study. Multiple sclerosis. Volume 19, 2013:508.

33 Mikol 2013 Mikol D, Freedman M, Goldman M, et al. ASCEND study of natalizumab efficacy on reducing disability in patients with secondary progressive multiple sclerosis: baseline demographics and disease characteristics. Multiple sclerosis. Volume 19, 2013:507‐508.

34 Kapoor 2018b Kapoor R, Sellebjerg F, Hartung HP, et al. Natalizumab Reduced Serum Levels of Neurofilament Light Chain in Secondary Progressive Multiple Sclerosis Patients From the Phase 3 ASCEND Study. Multiple sclerosis. Volume 24, 2018:988.

MS-SPI trial (NCT02220933)

35 Morteau 2018 Morteau O, Lasser R, Sedel F, et al. Annual relapse rates (ARR) in patients with spinal progressive multiple sclerosis treated with MD1003 (high-dose pharmaceutical biotin). Multiple Sclerosis Journal 2018;24:31.

36 Laplaud 2018 Laplaud DA, Gout O, Clavelou P, et al. Effect of MD1003 (High-Dose Pharmaceutical Biotin) in Spinal Progressive Multiple Sclerosis (MS-SPI): Subgroup Analyses. Multiple sclerosis. Volume 24, 2018:33.

37 Laplaud 2017 Laplaud DA, Gout O, Clavelou P, et al. Effect of MD1003 (High-Dose Biotin) in spinal progressive multiple sclerosis (MS-SPI): subgroup analyses. Multiple sclerosis. Volume 23, 2017:402-403.

38 Tourbah 2015 Tourbah A, Lebrun-Frenay C, Edan G, et al. MD1003 (high doses of biotin) in progressive multiple sclerosis: Subgroup analyses of the MS-SPI trial. Multiple Sclerosis 2015;23:785.

39 Tourbah 2016 Tourbah A, Lebrun-Frenay C, Edan G, et al. MD1003 (high-dose biotin) for the treatment of progressive multiple sclerosis: A randomised, double-blind, placebo-controlled study. Multiple Sclerosis 2016;22:1719-1731.

MS-STAT trial (NCT00647348)

40 MS-STAT trial Chataway J, Schuerer N, Alsanousi A, et al. Effect of high-dose simvastatin on brain atrophy and disability in secondary progressive multiple sclerosis (MS-STAT): a randomised, placebo-controlled, phase 2 trial. Lancet (London, England). Volume 383, 2014:2213‐2221.

41 Chan 2017 Chan D, Binks S, Nicholas JM, et al. Effect of high-dose simvastatin on cognitive, neuropsychiatric, and health-related quality-of-life measures in secondary progressive multiple sclerosis: secondary analyses from the MS-STAT randomised, placebo-controlled trial. The Lancet Neurology 2017;16:591-600.

42 Chan 2016 Chan D, Binks S, Nicholas J, et al. Effect of high-dose simvastatin on cognition in secondary progressive multiple sclerosis (MS-STAT cognitive): A randomised, placebo-controlled, Phase 2 trial. Multiple Sclerosis 2016;22:82-83.



43 Chataway 2015 Chataway J, Nicholas J, Wych J, et al. Smaller baseline brain volume and higher atrophy rate over two years is associated with poorer clinical outcomes: Post hoc analysis of the MS-STAT trial in secondary progressive MS. Neurology 2015;84.

44 Chataway 2012a Chataway JS, Alsanousi A, Chan D, et al. The ms-stat trial: A phase II trial of high dose simvastatin for secondary progressive multiple sclerosis (SPMS): Initial results. Annals of Neurology 2012;72:S111.

45 Chataway 2011 Chataway J, Awad M, Meadmore K, et al. Cognitive and neuropsychiatric status in a large cohort of patients with secondary progressive multiple sclerosis. Journal of Neurology, Neurosurgery and Psychiatry 2011;82:e4.

46 Chataway 2010 Chataway J, Anderson V, Chan D, et al. The MS-Stat Trial: A phase II trial of high-dose simvastatin for secondary progressive multiple sclerosis: Baseline trial profile. Journal of Neurology, Neurosurgery and Psychiatry 2010;81:e55.

47 Chataway 2013 Chataway J, Schuerer N, Alsanousi A, et al. The ms-stat trial: High dose simvastatin slows brain atrophy and delays disability in secondary progressive multiple sclerosis: A phase II placebo-controlled trial. Neurology 2013;80.

48 Chataway 2012b Chataway J, Alsanousi A, Chan D, et al. The MS-STAT trial: A randomised placebocontrolled phase II trial of high dose simvastatin in secondary progressive multiple sclerosis (SPMS). European Journal of Neurology 2012;19:87.

49 Chataway 2012 Chataway J, Alsanousi A, Chan D, et al. The MS-STAT trial: high dose simvastatin demonstrates neuroprotection without immune-modulation in secondary progressive multiple sclerosis (SPMS)-a phase II trial. Multiple sclerosis. Volume 18, 2012:509.

Morales 2017

50 Morales 2017 Morales IB, Eleftheriou E, Maranda L, et al. The safety and efficacy of rituximab use in secondary-progressive multiple sclerosis (SPMS) at UMMHC: Five years follow up data. Neurology 2017;88.

51 Morales 2016 Morales IB, Eleftheriou E, Maranda L, et al. The safety and efficacy of rituximab use in secondary progressive multiple sclerosis (SPMS) at UMMHC: five years follow up data. Five years follow up data. Multiple Sclerosis 2016;22:805.

EUSPMS

52 Kappos 1998 Kappos L, Polman C, Pozzilli C, et al. Placebo-controlled multicentre randomised trial of interferon β-1b in treatment of secondary progressive multiple sclerosis. Lancet 1998;352:1491-1497.

53 Kuhle 2016 Kuhle J, Hardmeier M, Disanto G, et al. A 10-year follow-up of the European multicenter trial of interferon β-1b in secondary-progressive multiple sclerosis. Multiple sclerosis (houndmills, basingstoke, england). Volume 22, 2016:533‐543.

54 Sormani 2005 Sormani MP, Bruzzi P, Beckmann K, et al. The distribution of magnetic resonance imaging response to interferonβ-1b in multiple sclerosis. Journal of Neurology 2005;252:1455-1458.

55 Polman 2005 Polman CH, Kappos L, Dahlke F, et al. Interferon beta-1b treatment does not induce autoantibodies. Neurology 2005;64:996-1000.



56 Polman 2003 Polman C, Kappos L, White R, et al. Neutralizing antibodies during treatment of secondary progressive MS with interferon β-1b. Neurology 2003;60:37-43.

57 Molyneux 2001 Molyneux PD, Barker GJ, Barkhof F, et al. Clinical-MRI correlations in a European trial of interferon beta-1b in secondary progressive MS. Neurology 2001;57:2191-2197.

58 Kappos 2001 Kappos L, Polman C, Pozzilli C, et al. Final analysis of the European multicenter trial on IFNβ-1b in secondary-progressive MS. Neurology 2001;57:1969-1975.

59 Barkhof 2001 Barkhof F, Van Waesberghe JHTM, Filippi M, et al. T1 hypointense lesions in secondary progressive multiple sclerosis: Effect of interferon beta-1b treatment. Brain 2001;124:1396-1402.

60 Molyneux 2000 Molyneux PD, Kappos L, Polman C, et al. The effect of interferon beta-1b treatment on MRI measures of cerebral atrophy in secondary progressive multiple sclerosis. Brain 2000;123:2256-2263.

61 Sormani 2003 Sormani M, Bruzzi P, Beckmann K, et al. MRI metrics as surrogate endpoints for EDSS progression in SPMS patients treated with IFN beta-1b. Neurology. Volume 60, 2003:1462‐1466.

62 Freeman 2001 Freeman J, Thompson A, Fitzpatrick R, et al. Interferon-beta1b in the treatment of secondary progressive MS: impact on quality of life. Neurology. Volume 57, 2001:1870‐1875.

63 Miller 1999 Miller D, Molyneux P, Barker G, et al. Effect of interferon-beta1b on magnetic resonance imaging outcomes in secondary progressive multiple sclerosis: results of a European multicenter, randomized, double-blind, placebo-controlled trial. European Study Group on Interferon-beta1b in secondary progressive multiple sclerosis. Annals of neurology. Volume 46, 1999:850‐859.

64 Polman 1995 Polman C, Dahlke F, Thompson A, et al. Interferon beta-1b in secondary progressive multiple sclerosis--outline of the clinical trial. Multiple sclerosis (houndmills, basingstoke, england). Volume 1 Suppl 1, 1995:S51‐4.

65 Miller 1997 Miller D, Polman C, Pozzilli C, et al. MRI protocol for the European trial of Beta interferon-1b in secondary progressive multiple sclerosis. Journal of the neurological sciences. Volume 150, 1997:S251.

66 Thompson 1998 Thompson A, Kappos L, Polman C, et al. Interferon beta-1b delays progression of disability in secondary progressive multiple sclerosis: final results of the european multicentre study. Multiple sclerosis (houndmills, basingstoke, england). Volume 4, 1998:392.

67 Brex 2001 Brex P, Molyneux P, Smiddy P, et al. The effect of IFNbeta-1b on the evolution of enhancing lesions in secondary progressive MS. Neurology. Volume 57, 2001:2185‐2190.

68 Kuhle 2004 Kuhle J, Hardmeier M, Rio J, et al. Long-term follow-up of the European Study of Interferon beta-1b (EUSPMS) in secondary progressive MS: predictors of treatment response. Multiple sclerosis (houndmills, basingstoke, england). Volume 10, 2004:S247.



NA SPMS

69 Panitch 2004 Panitch H. Interferon beta-1b in secondary progressive MS: Results from a 3-year controlled study. Neurology 2004;63:1788-1795.

70 Goodkin 2000 Goodkin D. Interferon Beta-1b in secondary progressive MS: clinical and MRI results of a 3-Year randomized controlled trial. Neurology. Volume 54, 2000:2352.

NORDIC SPMS Study

71 Andersen 2004 Andersen O, Elovaara I, Färkkilä M, et al. Multicentre, randomised, double blind, placebo controlled, phase III study of weekly, low dose, subcutaneous interferon beta-1a in secondary progressive multiple sclerosis. Journal of Neurology, Neurosurgery and Psychiatry 2004;75:706-710.

72 Beiske 2007 Beiske AG, Naess H, Aarseth JH, et al. Health-related quality of life in secondary progressive multiple sclerosis. Multiple Sclerosis 2007;13:386-392.

SPECTRIMS

73 Francis 2001 Francis G. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: Clinical results. Neurology 2001;56:1496-1504.

74 Li 2001 Li D, Zhao G, Paty D. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: MRI results. Neurology. Volume 56, 2001:1505‐1513.

75 Patten 2002 Patten S, Metz L. Interferon beta1a and depression in secondary progressive MS: data from the SPECTRIMS Trial.

Neurology. Volume 59, 2002:744‐746.

76 Sormani 2010 Sormani M, Stubinski B, Cornelisse P, et al. Magnetic resonance active lesions as individual-level surrogate for relapses in multiple sclerosis. Multiple sclerosis (houndmills, basingstoke, england), 2010:[Epub ahead of print].

Beutler 1996

77 Beutler 1996 Beutler E, Sipe JC, Romine JS, et al. The treatment of chronic progressive multiple sclerosis with cladribine. Proceedings of the National Academy of Sciences of the United States of America 1996;93:1716-1720.

78 Sipe 1994a Sipe JC, Romine JS, Koziol JA, et al. Cladribine in treatment of chronic progressive multiple sclerosis. Lancet 1994;344:9-13.

79 Sipe 1995 Sipe J, Romine J, Koziol J, et al. Cladribine treatment of chronic progressive (C/P) MS: a double-blind, crossover study with 2+ years' observation. Neurology. Volume 45 Suppl 4, 1995:A418.

80 Sipe 1994b Sipe J, Romine J, Zyroff J, et al. Cladribine favorably alters the clinical course of progressive multiple sclerosis (MS). Neurology. Volume 44 Suppl 2, 1994:A357.



81 Wajgt 1997 Wajgt A, Strzyzewska S, Ochudlo S. The treatment of chronic progressive multiple sclerosis with cladribine. Journal of the neurological sciences. Volume Suppl, 1997:S116.

82 Rice 2000 Rice GPA, Filippi M, Comi G. Cladribine and progressive MS: Clinical and MRI outcomes of a multicenter controlled trial. Neurology 2000;54:1145-1155.

83 Bornstein 1991 Bornstein M, Miller A, Slagle S, et al. A placebo-controlled, double-blind, randomized, two-center, pilot trial of Cop 1 in chronic progressive multiple sclerosis. Neurology. Volume 41, 1991:533‐539.

84 Fernandez 2018 Fernandez O, Izquierdo G, Fernandez V, et al. Adipose-derived mesenchymal stem cells (AdMSC) for the treatment of secondary-progressive multiple sclerosis: a triple blinded, placebo controlled, randomized phase I/II safety and feasibility study. Plos one. Volume 13, 2018.

IMPACT

85 Cohen 2002 Cohen J, Cutter G, Fischer J, et al. Benefit of interferon beta-1a on MSFC progression in secondary progressive MS. Neurology. Volume 59, 2002:679‐687.

86 Miller 2006 Miller D, Cohen J, Kooijmans M, et al. Change in clinician-assessed measures of multiple sclerosis and subject-reported quality of life: results from the IMPACT study. Multiple sclerosis (houndmills, basingstoke, england). Volume 12, 2006:180‐186.

87 Krishnan 2012 Krishnan A, Goodman A, Potts J, et al. Slower walking speed is associated with reduced health-related quality of life in patients with SPMS. Neurorehabilitation and neural repair. Volume 26, 2012:661.

88 Cadavid 2010 Cadavid D, Lee S, Lucas N, et al. Effect of natalizumab on ambulatory improvement in relapsing-remitting and secondary progressive multiple sclerosis. Multiple Sclerosis 2010;16:S142.

Individual studies

89 Patti 1999 Patti F, L'Episcopo MR, Cataldi ML, et al. Natural interferon-β treatment of relapsing-remitting and secondary- progressive multiple sclerosis patients. A two-year study. Acta Neurologica Scandinavica 1999;100:283-289.

90 Perrone 2014 Perrone C, Berriosmorales I, Beretich B, et al. Rituximab in the treatment of secondary-progressive multiple sclerosis. Multiple Sclerosis 2014;20:194.

91 Gunduz 2016 Gunduz T, Ozcan G, Çakar A, et al. Comparison of Mitoxantrone versus cyclophosphamide in patients with secondary progressive multiple sclerosis. European Journal of Neurology 2016;23:812.

92 Perini 2006 Perini P, Calabrese M, Tiberio M, et al. Mitoxantrone versus cyclophosphamide in secondary-progressive multiple sclerosis: A comparative study. Journal of Neurology 2006;253:1034-1040.

93 MS231 trial (International Natalizumab Multiple Sclerosis Trial)

Miller DH, Khan OA, Sheremata WA, et al. A controlled trial of natalizumab for relapsing multiple sclerosis. New England Journal of Medicine 2003;348:15-23.




Table A10. SLR Update – Publications included in the SLR update

# Author, Year Reference

EXPAND trial (NCT01665144)

1 Adlard, 2018 Adlard NE, Rendas-Baum R, Bjorner JB, et al. Responder Definition of the Multiple Sclerosis Impact Scale (Msis)-29 V2 among Patients with Secondary Progressive Multiple Sclerosis. Value in Health 2018;21 (Supplement 3):S390.

2 Department of Error, 2018 Erratum: Siponimod versus placebo in secondary progressive multiple sclerosis (EXPAND): a double-blind, randomised, phase 3 study (The Lancet (2018) 391(10127) (1263-1273), (S0140673618304756) (10.1016/S0140-6736(18)30475-6)). The Lancet 2018;392:2170.

3 Kappos, 2018 Kappos L, Vermersch P, Fox R, et al. Longer-term Safety with Siponimod Treatment in Multiple Sclerosis: Pooled Analysis of Data from the Bold and Expand Trials and their Extensions. Multiple Sclerosis and Related Disorders 2018;26:255-256.

SPECTRIMS

4 Freedman, 2016 Freedman MS, Hayward B, Warth JD, et al. Clinical and MRI efficacy of IFN beta-1a SC tiw in MS patients with more advanced disease (EDSS 4.0-6.0). Multiple Sclerosis 2016;1):18.

5 Traboulsee, 2017 Traboulsee A, Li D, Tam R, et al. Subcutaneous interferon beta-1a three times weekly and the natural evolution of gadolinium-enhancing lesions into chronic black holes in relapsing and progressive multiple sclerosis: Analysis of PRISMS and SPECTRIMS trials. Multiple Sclerosis Journal Experimental Translational & Clinical 2017;3:2055217317745340.


94 Mostert 2013 Mostert J, Heersema T, Mahajan M, et al. The effect of fluoxetine on progression in progressive multiple sclerosis: A double-blind, randomized, placebo-controlled trial. ISRN Neurology 2013;1.

95 Vermersch 2012 Vermersch P, Benrabah R, Schmidt N, et al. Masitinib treatment in patients with progressive multiple sclerosis: a randomized pilot study. BMC Neurology 2012;12.

96 Bosco 1997 Bosco A, Cazzato G, Monti F, et al. Double-blind, placebo-controlled, randomized study of idebenone in patients with chronic progressive MS. Multiple sclerosis (houndmills, basingstoke, england). Volume 3 Suppl, 1997:349.

97 Wang 2018 Wang L, Qi CH, Zhong R, et al. Efficacy of alemtuzumab and natalizumab in the treatment of different stages of multiple sclerosis patients. Medicine (United States) 2018;97:e9908.



Rationale for exclusion of trials used for indirect treatment comparisons from the SLR

results

Table A11. Rationale for exclusion of trials used for indirect or treatment comparisons from the SLR results

Trial Reference Inclusion status and reason

EXPAND trial Included – Siponimod

ASCEND trial Included – Natalizumab

MS-SPI trial Excluded – Biotin

Wang 2018 Excluded – Alemtuzumab vs. Natalizumab

MS-STAT trial Excluded – Simvastatin

Morales 2017 Excluded – Lack of comparable outcomes

Perrone 2014 Excluded – Lack of comparable outcomes

Gunduz 2016 Excluded – Mitoxantrone vs. cyclophosphamide

Perini 2006 Excluded – Mitoxantrone vs. cyclophosphamide

MS231 trial (International Natalizumab Multiple Sclerosis Trial)

Excluded – did not report ARR or CDP

Mostert 2013 Excluded – Fluoxetine

Vermersch 2012 Excluded – Masitinib

EUSPMS Included – Interferon β

NA SPMS Included – Interferon β

Nordic SPMS Study Excluded – Unlicensed regimen of Interferon β-1a

SPECTRIMS trial Included – Interferon β

Patti 1999 Excluded – did not report ARR or CDP

Beutler 1996 Excluded – Lack of comparable outcomes

Rice 2000 Excluded – Lack of comparable outcomes

Bornstein 1991 Excluded – Non-comparable outcome definitions

Fernandez 2018 Excluded – Stem cell therapy

IMPACT study Included – Interferon β

Bosco 1997 Excluded – Idebenone




APPENDIX 2: Guidelines for diagnosis and management

Table A12. Overview of guidelines used for this assessment

Name of society/organisation issuing guidance

Date of issue

Country/ies to which applicable

Summary of recommendation Level of evidence

European Committee of Treatment and Research in Multiple Sclerosis (ECTRIMS) / European Academy of Neurology (EAN)

2018 European countries

- Consider treatment with interferon beta-1a (subcutaneously) or interferon beta-1b in patients with active secondary progressive MS;

- Consider treatment with mitoxantrone in patients with active secondary progressive MS taking into account, in discussion with the patient, the efficacy and specifically the safety and tolerability profile of this agent;

- Consider treatment with ocrelizumab or cladribine for patients with active secondary-progressive MS.

Weak

European Committee of Treatment and Research in Multiple Sclerosis (ECTRIMS) / European Academy of Neurology (EAN)

2018 European countries

- Always consult the Summary of Product Characteristics (SPC) for dosage, special warnings and precautions for use, contraindications and monitoring of side effects and potential harms.

Consensus statement

Sources: Montalban, X., Gold, R., Thompson, A. J., Otero-Romero, S., Amato, M. P., Chandraratna, D et al. (2018). EC-TRIMS/EAN Guideline on the pharmacological treatment of people with multiple sclerosis. Multiple Sclerosis, 24(2), 96–120.

https://doi.org/10.1177/1352458517751049 [35] Abbreviations: MS- multiple sclerosis



APPENDIX 3: EVIDENCE GAPS

Table A13. Recommendations for research

Research question: What is the relative clinical effectiveness and safety of siponimod compared to other interventions, in adult patients with SPMS with active disease evidenced by relapses or imaging features of inflammatory activity?

Rationale: No direct comparison of siponimod to any active comparator for active SPMS is available.

Evidence Direct evidence is needed.

Population Patients with SPMS with active disease evidenced by relapses or imaging features of inflammatory activity.

Intervention Siponimod

Comparators Interferon-β-1a, Interferon-β-1b, mitoxantrone, ocrelizumab, natalizumab, fingolimod, cladribine, rituximab.

Outcome(s) Confirmed disability progression at 6 months, other measures of disability progression, symptoms, clinical relapse, health-related quality of life, MRI inflammatory measured disease activity and burden of disease, no evidence of disease activity, adverse events, serious adverse events, adverse events leading to treatment discontinuation and treatment-related mortality.

Time stamp Beyond 1 year.

Burden of disease High

Study design Pragmatic RCTs and prospective cohort studies.



APPENDIX 4: STUDIES INCLUDED IN THE ANCILLARY ANALYSES OF THE

GLOBAL SPMS POPULATION

Characteristics of included studies

Table A14 presents the study design of the studies included in the assessment for the global SPMS population. For assessing comparability, “study design” included: study phase, study design, sample size, population, study duration, comparator treatment, and intervention. Table A15 depicts the inclu-sion/exclusion criteria of the 7 studies included in the assessment. Table A16 shows the outcome defi-nitions used by the different studies. Table A17 presents the baseline patient characteristics of the studies included in the global SPMS population analyses. Table A18 presents the placebo arm treat-ment effects of studies in the global SPMS population.



Table A14. Study design of studies in the global SPMS population

Study ID Author / Year Phase Study design

Contains data from a relapsing

population

Sample size relapsing/ total (%)

MS popula-

tion

Study duration

Comparator Treatment

EXPAND Kappos 2018 3 Randomised, double-blind, parallel group

Yes 590/1651

(35.7) SPMS 3 years

Placebo PO each day

Siponimod 2 mg PO each day

Nordic SPMS study

Andersen 2004 3 Randomised, double-blind, parallel group

Yes 217/371 (58.5)

SPMS 3 years Placebo SC once a week

IFNβ-1a 22 µg SC once a week

SPECTRIMS

SPECTRIMS Study Group 2001

Li 2001

NA Randomised, double-blind, parallel group

Yes 293/618

(47.4) SPMS 3 years

Placebo SC three times weekly

IFNβ-1a 22 µg SC three times weekly

IFNβ-1a 44 µg SC three times weekly

North Ameri-can Study

Panitch 2004 NA Randomised, double-blind, parallel group

No unk/939 SPMS 3 years Placebo SC every other day

IFNβ-1b 160 µg/m2 SC every other day

IFNβ-1b 250 µg SC every other day

European study

European Study Group 1998

Kappos 2001

3 Randomised, double-blind, parallel group

Yes 502/718 (69.9)

SPMS 3 years* Placebo SC every other day

IFNβ-1b 250 µg SC every other day

ASCEND Kapoor 2018 3 Randomised, double-blind, parallel group

No unk/889 SPMS 2 years** Placebo IV q4w Natalizumab 300 mg IV q4w

IMPACT Cohen 2002 NA Randomised, double-blind, parallel group

No unk/436 SPMS 2 years Placebo IM once a week

IFNβ-1a 60 µg IM once a week

Source: adapted from the technical reports provided by the MAH

* Early termination at month 33. Mean duration of follow up was 1054 and 1068 days in the placebo and IFNβ-1b group, respectively ** Outcomes were assessed over a 96-week treatment period Highlighted studies with data on the outcomes of interest in the relapsing population



Table A15. Inclusion/exclusion criteria of studies in the global SPMS population

Study ID EXPAND Nordic SPMS study

SPECTRIMS North American Study

European study ASCEND IMPACT

MS Population SPMS SPMS SPMS SPMS SPMS SPMS SPMS

Baseline EDSS range 3.0-6.5 ≤6.5 3.0-6.5 3.0-6.5 3.0-6.5 3.0-6.5 3.5-6.5

Age range 18-60 18-65 18-55 18-65 18-55 18-58 18-60

Prior IFN therapy Allowed No prior IFN use No prior IFN use No prior IFN use No prior IFN use No prior IFN use 4 weeks prior to study

No prior IFNβ use

No relapses in X months prior

3 months 2 months 2 months 2 months 1 month 3 months NR

Recently documented progression

For ≥6 months in past 24 months

For ≥6 months in past 4 years



Progression in past 2 years or ≥2 relapses in past 2 years

In the past 1 year

In the past 1 year

History of RRMS Required Required Required Required Required NR NR

Duration of MS No restriction ≥ 1 year NR ≥ 2 years ≥ 2 years NR NR

Duration of SPMS No restriction NR NR NR NR ≥ 2 years NR

MS severity score No restriction NR NR NR NR Score of 4 or higher

NR

T25FW test No restriction NR NR NR NR <30 seconds NR

Source: adapted from the technical reports provided by the MAH Highlighted studies with data on the outcomes of interest in the relapsing population



Table A16. Outcome definitions of studies in the global SPMS population

ARR Time to CDP-3 Time to CDP-6 Discontinuation

EXPAND Total number of relapses per patient-years

1.0-point inc. in EDSS score: 3.0-5.0





Nordic SPMS study


NR 1.0-point inc. in EDSS score: 3.0-5.0



SPECTRIMS Total number of relapses per patient-years



NR The proportion of randomised patients who discontinued treatment for any reason

North American Study


NR 1.0-point inc. in EDSS score: 3.0-5.5



European study





ASCEND Total number of relapses per patient-years

NR 1.0-point inc. in EDSS

score: 3.0-5.5

0.5-point inc. in EDSS

score: 6.0-6.5

Inc. of ≥20% in T25FW

Inc. ≥20% in 9-HPT


IMPACT Total number of relapses per patient-years







Table A17. Patient baseline characteristics of studies in the global SPMS population

Study ID EXPAND Nordic SPMS study

SPECTRIMS North American Study

European study

ASCEND IMPACT

Age (mean years) 48.0 45.7 42.8 46.8 41.0 47.2 47.6

Proportion female (%) 60 60 63 63 61 62 64

Mean EDSS score 5.4 4.8 5.4 5.1 5.1 5.6 5.2

Proportion of patients with EDSS score ≥6.0 (%) 56 NR NR NR 45 63 48

Time since onset of MS symptoms (mean years) 16.8 NR NR NR NR 16.5 NR

Duration of MS (mean years) 12.6 14.3 13.3 14.7 13.1 12.1 16.5

Duration of SPMS (mean years) 3.8 5.4 4.0 4.0 2.2 4.8 NR

Normalised brain volume (mean cm3) 1423 NR NR NR NR 1423 NR

Proportion of patients with Gd+ lesions of T1-weighted images (%) 21 NR NR NR NR 24 36

Total volume of T2 lesions on T2-weighted images (mean mm3) 15,322 NR NR NR NR 16,793 NR

Proportion of patients without previous use of a DMT (%) 22 NR NR NR NR 23 a NR

Mean Timed 25-Foot Walk Test (seconds) 16.7 NR NR NR NR 11.2 b 14.5

Time since most recent relapse (months) 59 NR NR NR NR 57 44.4

Proportion of patients relapse-free in prior year (%) 78 NR NR NR NR 84 61

Proportion of patients relapse-free in prior 2 years (%) 64 NR 53 55 30 71 NR

Proportion of patients with relapses in prior 2 years (%) 36 NR 47 45 70 29 NR

Number of relapses per patient in the prior year (mean) 0.2 NR NR NR NR NR 0.6

Number of relapses per patient in the previous 2 years (mean) 0.7 NR 0.9 0.8 NR NR NR


Highlighted studies with data on the outcomes of interest in the relapsing population a Gold R et al. (2017) Impact of Primary Endpoint Definitions and Patient Baseline Characteristics on Study Outcomes in Progressive Multiple Sclerosis. Poster presented at the 7th Joint Congress of the European Committee for Treatment and Research in Multiple Sclerosis-Americas Committee for Treatment and Research in Multiple Sclerosis; 25-28 October 2017; Paris, France. P1239. Value

derived using DigitizeIt: Bormann I (Web Page) DigitizeIt V 2.3.3. Updated 2016. Available online at: http://www.digitizeit.de. Accessed: 2019 July 3. b Median value. Abbreviations: DMT=disease-modifying treatment; EDSS=Expanded Disability Status Scale; Gd+=gadolinium-enhancing; MS=multiple sclerosis; NR=not reported; n/a=not applicable; SPMS=sec-

ondary progressive multiple sclerosis.



Table A18. Placebo arm treatment effects of studies in the global SPMS population

Study Placebo administration ARR Discontinuation

EXPAND PO each day 0.16 0.084

Nordic SPMS study SC once a week 0.27 0.060

SPECTRIMS SC three times weekly 0.71 0.057

North American Study SC every other day 0.28 0.093

European study SC every other day 0.57 0.132

ASCEND IV q4w 0.17 0.186

IMPACT IM once a week 0.30 0.142




APPENDIX 5: EXPAND (STUDY A2304)

EXPAND (Study A2304) was the only study with siponimod in the SPMS population. Therefore, we performed a detailed description of this study.

Study design

EXPAND was a phase 3, randomised, double-blind, placebo controlled trial, conducted at 292 centres, in 31 countries. The study included 1,651 patients, aged between 18 and 60 years old, with SPMS, who were randomly assigned, in a 2:1 ratio, to oral siponimod (n= 1,105) 2 mg once daily or placebo (n= 546) for up to 3 years or until the occurrence of at least 374 3-month CDP events. The primary endpoint was 3-month CDP.

The study included a Core Part and an Extension Part. Patients who showed 6-month CDP during treatment had the option of starting treatment with open-label siponimod as rescue medication. Patients who prematurely discontinued double-blind treatment were asked to remain in the study and follow a simplified visit schedule. The study design for patients who participated in the Core and extended treat-ment part is depicted in Figure A1.

Figure A1. Study design of EXPAND study

Source: Clinical Study Report of EXPAND study

Inclusion and exclusion criteria (Table 4.5)

Key eligibility criteria included: age between 18 and 60 years; prior history of RRMS according to the 2010 Revised McDonald criteria; a diagnosis of SPMS, defined by a progressive increase in disability of at least 6 months duration, in the absence of relapses; patients had to have an Expanded Disability Status Scale (EDSS) score between 3.0 and 6.5, and documented progression in the two years prior to the study (defined as at least an 1-point increase for patients with EDSS<6.0, and as at least a 0.5-point increase for patients with EDSS≥6.0, assessed at screening); and no evidence of relapse within 3 months prior to randomization.

Exclusion criteria included: patients with an active chronic disease of the immune system (other than MS) or with a known immunodeficiency syndrome; history of malignancy, treated or untreated, within the prior 5 years; diabetes mellitus with renal insufficiency, retinal disease or neuropathy; diagnosis of macular edema at the screening visit; active systemic infection, including acquired immunodeficiency syndrome, positive human immunodeficiency virus antibody, and hepatitis A, B, C or E acute or chronic infection; patients negative for varicella-zoster virus IgG antibodies at screening;

Patients were also excluded if they received live or live-attenuated vaccines within 2 months prior to randomisation; received prior treatment with siponimod; received fingolimod for more than 6 months or



within 2 months prior to randomization; received dimethyl fumarate within 2 months prior to randomiza-tion; treatment with natalizumab within 6 months prior to randomization; received immunosuppres-sive/chemotherapeutic agents within 6 months prior to randomization; treatment with rituximab, ofatu-mumab, ocrelizumab or cladribine within 2 years prior to randomization; prior treatment with alemtuzumab; mitoxantrone within 2 years prior to randomization; teriflunomide within 2 years prior to randomization.

Additional exclusion criteria included conditions or treatments that may have affected cardiovascular function; and presence of inadequate liver, kidney, or blood function.

Randomisation and treatment assignment

Patients were randomly assigned, in a 2:1 ratio, via Interactive Response Technology (IRT) to receive siponimod or matching placebo by blocked randomization with a block of 6. The IRT assigned a ran-domisation number to the patient, which was used to link the patient to a treatment arm. The randomi-zation numbers were generated by the IRT provider, and were not communicated to the investigators. Randomisation was stratified by geographic region. Patients, investigators and assessors were masked to treatment assignment for the duration of the study.

Study drug and placebo were identical in appearance, taste and odour.

Procedures

Patients were randomised to receive oral siponimod 2 mg once daily (OD) or matching placebo. A dose reduction to 1 mg OD was required in patients with lymphocyte counts below 0.2x109 /L. Following a dose reduction, the lower dose was maintained until the end of the study regardless of any increase in lymphocyte counts.

The study included a Core Part and an Extension Part. Patients who showed 6-month CDP during treatment had the option of starting treatment with open-label siponimod as rescue medication. Rescue treatment began with titration to the 2-mg dose of siponimod, independently of the dose taken by the patient at the end of the Core Part (double-blind treatment Epoch). Dose reduction criteria used during the Core Part of the study applied to the Extension Part as well.

Patients maintaining plasma lymphocyte counts below 0.2x109 /L in spite of dose reduction, were re-quired to discontinue study drug until lymphocyte level reached 0.6x109/L.

Patients with 6-month CDP had the option of continuing the blind treatment, switching to open-label siponimod, or starting any other treatment for MS. For all 3 options, treatment allocation remained blinded until the end of the Core Part.

If MS relapses occur during study period, a standard course of intravenous corticosteroids for 3-5 days, at investigator discretion, was allowed.

Patients receiving beta-blockers, could only initiate siponimod if heart rate was over 50 bpm.

Outcomes

A detailed description of the EXPAND endpoints is presented in Table A19.

Table A19. Endpoints and definitions

Primary Endpoint

Time to 3-month CDP Time to 3-month Confirmed Disability Progression (3mCDP) defined as an increase from baseline of EDSS score of:

- 1 point in patients with a Baseline EDSS score of 3.0 to 5.0, or

- 0.5 point in patients with a Baseline EDSS score of 5.5 to 6.5.

Criteria must be met at visits at least 3 months after onset and at any interim



assessment.

Confirmation cannot be during a relapse

Key Secondary Endpoint

Time to 3-month confirmed worsening of T25W

Time to 3-month confirmed worsening of Timed 25 Walking test by 20% compared to baseline.

Criteria must be met at visits 3 months after onset and at any interim assessment. Confirmation cannot be during a relapse.

Change from baseline in T2 lesion volume

Secondary Endpoint

Time to 6-month CDP Time to 6-month Confirmed Disability Progression (6mCDP). Progression defined as for 3 month CDP. Criteria must be met at visits 6 months after onset and at any interim assessment. Confirmation cannot be during a relapse.

Annualised relapse rate (ARR)

Defined as the average number of confirmed relapses per year. Confirmed relapse: associated with an increase of at least 0.5 points on the EDSS score, or an increase of 1 point in two Functional System scores or 2 points in one Functional System score

Change in Multiple Sclerosis Walking Scale (MSWS-12)

Patient Reported Outcome

Change in other measures by conventional MRI

Number of Gd enhancing T1 lesions

Number of new or enlarging T2 lesions

T1 hypointense lesions

Percentage of Brain volume change

Exploratory objectives

Evaluate the Health-related quality of life (QoL) on the following patient reported outcomes: MSIS-29 and EQ-5D

Cognitive tests: Paced Auditory Serial Addition Test (PASAT), Symbol Digit Modalities Test (SDMT) and Brief Visuospatial Memory Test Revised (BVMT-R)

To evaluate the evolution of acute lesions into chronic black holes (MRI)

To evaluate the efficacy of siponimod on the Multiple Sclerosis Functional Composite (MSFC) z-score

To evaluate the efficacy of siponimod in delaying the time to: 3m-CDP of at least 20% from baseline in the T25W or 3m-CDP as measured by EDSS score or 3-month confirmed worsening of at least 20% from baseline in the 9-Hole Peg Test (9-HPT) in either one of the hands (dominant or non-dominant)

To explore baseline characteristics which are associated with a positive treatment response to define clinically relevant responder subgroups

To explore the relationship between disability progression endpoints and drug concentration/lymphocyte count

To explore the relationship between selected safety parameters and drug concentration/lymphocyte count

To evaluate the pharmacokinetics (PK) of siponimod

To evaluate the effects of siponimod on 3m-CDP as measured by EDSS in the following subgroups: patients previously treated, or not, with interferon beta-1b and treatment-naïve and patients with prior treatment with disease-modifying drugs.

Source: Adapted from EPAR



The primary efficacy variable was the time to 3-month confirmed disability progression (3-month CDP), defined as the time from baseline to the onset of disability progression (confirmed after at least 3 months). Confirmed disability progression (CDP) was defined as a 1-point increase in EDSS when baseline score was between 3.0 and 5.0, or a 0.5 increase in EDSS when baseline score was between 5.5 and 6.5. The Expanded Disability Status Scale (EDSS) was assessed every 3 months and following a relapse, based on neurological examination, with scores in each of 7 functional systems (visual, brain stem, pyramidal, cerebellar, sensory, bowel and bladder, and cerebral) being combined to determine EDSS, with total scores ranging from 0 to 10.

Progression was confirmed if, following the onset of disability progression, progression was confirmed at least 3 months later, and every EDSS score obtained between the onset and confirmation also met the progression criterion. Confirmation of progression was obtained from EDSS assessments at sched-uled visits, in the absence of relapses, whether confirmed or unconfirmed.

According to the study protocol, a relapse was an event that could not last longer than 90 days and, therefore, if the end date of a relapse was missing or showed a duration longer than 90 days, a 90 days relapse duration was assumed. Progression occurring out of this time period was assumed to take place in the absence of relapse.

The confirmation of EDSS score could not take place during a relapse. There was no imputation for missing values, when patients discontinued without having a CDP (confirmed progression).

For Kaplan-Meier estimates, the percentage of patients free of 3-month CDP at 12, 24, and 36 months was used, including the FAS population.

Key secondary efficacy variables included time to 3-month confirmed worsening of at least 20% from baseline in T25FW test and change from baseline in T2 lesion volume.

Each assessment included two T25FW tests, and the T25FW score was calculated as the mean of both tests. Patients with physical limitation who were unable to complete the tests were considered to have worsened at that visit. In cases where only one test was available, this test was used for the determi-nation of disability progression, even when the other test was not available due to physical limitation. A 3-month confirmed worsening of at least 20% was defined as a decrease from baseline that was con-firmed at least 3 months later.

Secondary efficacy variables were the following: Time to 6-month CDP, Annualised relapse rate (ARR) and Change in Multiple Sclerosis Walking Scale (MSWS-12).

In time to CDP-6, progression was confirmed if, following the onset of disability progression, progression was confirmed at least 6 months later.

Walking ability (ambulation) was self-assessed by patients using the MSWS-12. The 12-item Multiple Sclerosis Walking Scale (MSWS-12) is a patient-reported outcome measure assessing the extent to which MS impacts an individual’s walking ability.

The MSWS-12 comprises 12 questions concerning different aspects of walking function and quality that are rated on a scale ranging between 1 (not at all) and 5 (extremely). Individual item scores are summed to achieve a total score that ranges from 12 to 60. This total score is then transformed to achieve a range of 0–100, with higher scores reflecting a greater impact of MS on walking.

Exploratory outcomes, assessing cognitive function were the following: the Paced Auditory Serial Addition Test (PASAT), the Symbol Digit Modalities Test (SDMT), and the Brief Visuospacial Memory Test-Revised (BVMT-R).

The PASAT is a test for assessing cognitive function, measuring auditory information processing speed as well as calculation ability. The PASAT is presented using audiocassette tape or compact disk to ensure standardization in the rate of stimulus presentation. Single digits are presented every 3 seconds and the patient must add each new digit to the one immediately prior to it. The score is calculated based on the number of correct answers provided during 3 minutes (possible range 0-60). The PASAT is one of the components of the Multiple Sclerosis Functional Composite (MSFC).



SDMT assessed information processing speed.The SDMT involves a simple substitution task. Using a reference key, the patient has 90 seconds to pair specific numbers with given geometric figures. Ad-ministration time is 5 minutes.

The BVMT-R is a test that measures the visuospacial memory. In three learning trials, patients viewed a page with six geometric figures printed in a 2 X 3 array for 10 seconds and are asked to draw as many figures as possible in their correct location on a distinct page. A delayed recall trial is administered 25 minutes later.

The Symbol Digit Modalities Test (SDMT) and the Brief Visuospacial Memory Test-Revised (BVMT-R) were performed by the independent EDSS rater or another qualified individual every 6 months.

The Health-related quality of life (QoL) was assessed in patient reported outcomes using the following questionnaires: MSIS-29 and EQ-5D. The MSIS-29 is a 29-item, self-administered questionnaire that includes 2 domains (physical and psychological). Responses are scored on a 4-point ordinal scale ranging from 1 (not at all) to 4 (extremely), with higher scores reflecting greater impact on daily living activities. The questions assess the impact of MS on their day-to-day life during the prior 2 weeks.

The EQ-5D-3L comprises five dimensions (mobility, self-care, usual activities, pain/discomfort and anx-iety/depression) and each dimension has three levels (no problems, some problems, extreme prob-lems).

Multiple Sclerosis Functional Composite (MSFC) is a composite measure that assesses ambulation (timed 25 foot walk, T25FW), upper extremity function (9-hole peg test, 9HPT), and cognitive function (Paced Auditory Serial Addition Test, PASAT). T25FT and 9-HPT were assessed every 3 months, and PASAT was assessed every 6 months. The T25FW measures the time, in seconds, to walk 25 feet (7,62 meters). The 9-HPT measures the time, in seconds, required to insert and remove 9 pegs, with the right and left arm. Time is then converted into a score.

Protocol amendments

Following the original protocol (12 July 2012), 4 amendments were performed at a later time. Amend-ment 3 (23 March 2016) included an Extension Part in the study design, following which were added the designations of Core Part and Extension Part. Then, the protocol was amended. The original plan of stopping the Core Part when 374 3-month CDP events were observed, which was expected to occur at 2 years following randomization, was changed to approximately 3 years following the randomization of the first patient, and this was expected to compensate for the power lost due to alpha adjustment for the additional interim analysis. No additional sample size estimation was performed.

Statistical analysis

Continuous data are presented as means, standard deviations, medians, 25th and 75th percentiles, and range. Categorical data are presented as frequencies and percentages.

It was estimated that observation of at least 374 3-month CDP events was needed to detect a 30% reduction in the risk of 3-month CDP, with a 90% study power, using a log-rank test with a 5% signifi-cance level (two-sided). Assuming a proportion of 0.30 of patients with disability progression in the placebo group during the 2-year period, and a drop-out rate of 20%, an estimated sample size of 1,530 patients was required to observe at least 374 disability progression events.

A hierarchical testing was used to analyse primary and secondary endpoints, which were tested in the following order: Time to 3-month CDP, time to 3-month confirmed worsening of at least 20% from base-line in T25FW, change from baseline in T2 lesion volume.

A futility interim analysis was to be performed when 50% of the 3-month CDP events estimated as necessary had already occurred. In order to achieve a cumulative type I error below 5% (for a two-sided test), analysis was adjusted for multiplicity. For the primary endpoint, analysis was performed at a two-sided significance level adjusted according to the O’Brien-Fleming alpha correction level which was



estimated to be 0.0434. The second and third endpoints were tested sequentially at a two-sided signif-icance level of 0.05. All other secondary endpoints were tested at a nominal significance level of 0.05 without correction for multiplicity.

Efficacy analysis were conducted in the Full Analysis Set defined as all randomised patients who took at least one dose of study medication. Patients were analysed according to treatment assignment (i.e. irrespective of the treatment received).

The primary hypothesis (that there was a difference between groups in the time to 3-month CDP) was tested using a Cox proportional hazards model with treatment, country, and baseline EDSS and SPMS group as covariates. Kaplan-Meier estimates were used as supportive analysis.

In some cases, EDSS assessment was performed by the principal investigator, and not by the inde-pendent EDSS rater, or by the independent EDSS rater having access to the cardiac monitoring data-base. This protocol deviation and the associated risk of bias are discussed further in section 4.6. A sensitivity analysis was conducted excluding these patients.

An exploratory analysis was performed, evaluating the effect of treatment on 3-month CDP not influ-enced by a relapse. This analysis included patients in whom 3-month CDP did not occur during a re-lapse. For purposes of progression definition in patients with a relapse, if the EDSS score did not return to baseline value after the relapse, the EDSS score after the relapse was considered to be the new baseline EDSS score.

The first secondary hypothesis (that there was a difference between groups in time to 3-month con-firmed worsening of at least 20% from baseline in T25FW test) was tested using a Cox proportional hazards model with treatment, country, and baseline EDSS and SPMS group (with or without relapses) as covariates. Kaplan-Meier estimates were used as supportive analysis. Kaplan-Meier and log-rank test analyses did not include patients with missing baseline data.

The second secondary hypothesis (that there was a difference between groups in change from baseline in T2 lesion volume) was analysed using a mixed model for repeated measures, with visit as the cate-gorical factor, adjusted for treatment, country, age, baseline volume of T2 lesion, number of gadolinium-enhanced lesions, and SPMS group (with or without relapses).

The number of patients included in the distinct populations is depicted in Table A20.

Table A20. Analysis sets

Source: Clinical Study Report of Expand Study and EPAR BAF312 (siponimod); Full Analysis Set (FAS): comprised all randomised patients with assigned treatments who took at least

one dose of study medication. All available efficacy assessments were used, irrespective of the study treatment received. The FAS was used for all efficacy analyses; Modified Full Analysis Set (MFAS): comprised all randomised patients with assigned treatments who took at least one dose of study medication. If a patient prematurely discontinued study treatment and started a

new MS-DMT or open-label siponimod, efficacy assessments were only used up to the start of the new MS-DMT or open-label siponimod; Per-protocol Set (PPS): consisted of all patients in the FAS who did not have any major protocol deviations that could have confounded the interpretation of analyses conducted on the FAS; Safety Set (SAF): comprised all patients who

received at least one dose of study medication. Patients were analysed according to the actual treatment received, using all available data up to and including 30 days after last dose of study drug or the day before start of open-label siponimod, what-ever came first. The SAF was used for all safety analyses; Open-label Set (OLS): included all patients who received at least

one dose of open-label siponimod in the Core Part; Follow-up Set (FUS): consisted of all patients who received at least one dose of study medication and had follow-up assessments (i.e., had off-drug evaluations/assessments after end of study drug). However, for patients who received at least one dose of open-label siponimod, off-drug evaluations/assessments were only

included in this analysis if performed prior to the first dose of open-label siponimod; likewise, off-drug assessments after expo-sure to open-label siponimod were not included.



Protocol deviations

Blinding procedures were not followed in 11.3% of the patients in siponimod group, and 11.5% in pla-cebo group. The potential unblinding resulted from temporary incorrect access rights for site staff to the different databases.

Patient disposition

The inclusion period took place between 5 February 2015 and 2 June 2015. During this period, 1,651 patients were randomly assigned, in a 2:1 ratio, to siponimod (n= 1,105) or to placebo (n= 546). In the siponimod group, 6 patients were excluded from efficacy analysis for the following reasons: 5 patients never took the study medication, and 1 patient did not provide signed informed consent.

Of note, 903/1105 patients (81.7%) in siponimod group and 424/546 patients (77.7%) in placebo group completed study participation (double-blind, open-label treatment, and abbreviated visit schedule). However, only 737/1105 patients (66.7%) in siponimod group and 322/546 patients (59.0%) completed study participation on double-blind study drug. The main reasons for premature discontinuation from double blind study drug were subject decision (10.3% vs 13.0%), disease progression (9.1% vs 14.8%), and adverse events (8.5% vs 5.1%), for the siponimod and placebo groups, respectively. Table A21 summarizes patient disposition.

Table A21. Patient disposition on double-blind study drug


Following premature discontinuation of double-blind treatment, some patients continued in the so-called Treatment Epoch which included open-label siponimod (10.5% of patients in siponimod group, and 17.2% of patients in placebo group) and abbreviated visit schedule (12.2% siponimod; 10.4% placebo).



Datasets analysed

The seven analysis sets used in this study are shown in Table A22. The Full Analysis Set was used for efficacy analysis. In the FAS population, patients were analysed according to the groups to which they were assigned, independently of the study treatment they received. Therefore, the efficacy data of pa-tients receiving open-label siponimod or other DMTs (abbreviated visit schedule) were analysed using the treatment groups as randomised.

Table A22. Analysis sets


Baseline characteristics

Patients had a median age of 49 years, and most of the patients (94.7%) were white. Baseline demo-graphic characteristics were generally balanced between groups summarizes demographic character-istics by treatment group and are shown in Table A23.

Table A23. Patient baseline characteristics




Patients had multiple sclerosis for a median time of 16 years following the first symptom, and for a median of 11.7 years since diagnosis. Patients had converted to SPMS 2.6 years (median time) prior to study entrance. The mean number of relapses over the last two years was 0.7±1.19 (median value 0). Most patients did not have relapses either in the two years prior to study start (63.9%), or within the year prior to study start (78.4%). Table A24 summarizes multiple sclerosis disease history.

Table A24. Multiple sclerosis disease history




Approximately 50% of the patients had either at least one relapse in the previous 2 years of inclusion or one or more Gd-enhancing lesion at baseline

Mean EDSS score was 5.4±1.1. Median EDSS score was 6.0. Fifty six percent of the patients had a mean baseline EDSS score of at least 6.0.

Mean time to perform 25-foot walk test was 16.7±21.3 seconds. The baseline median time to perform 25-foot walk test was 10.1 seconds.

Table A25 summarizes multiple sclerosis baseline characteristics.

Table A25. Multiple sclerosis baseline characteristics




Prior medications

Table A26 summarizes the prior medications. The most frequent prior treatments included interferon beta-1A (41.7%), interferon beta-1B (27.4%), and glatiramer acetate (26.8%). In more than half of the patients, the reason for discontinuation was inadequate response.

Twenty two percent of the patients had not been on immunomodulatory treatments prior to study start.

Table A26. Prior medications




Analysis of efficacy

Time to 3-month CDP

At cut-off date for the interim CSR report (29-Apr-2016), 3-month CDP events were observed in 288/1096 (26.3%) of patients in the siponimod group, and 173/545 CDP events (31.7%) were observed in placebo group (hazard ratio 0,79; 95%CI 0.65 to 0.95; p=0.013).

Figure A2 depicts Kaplan-Meier curves for the proportion of patients free of 3-month CDP.

Figure A2. Patients free of 3-month CDP


In the full analysis set (comprised all randomised patients with assigned treatments who took at least one dose of study medication. All available efficacy assessments were used, irrespective of the study treatment received) time to 3-month CDP showed a risk reduction for siponimod in comparison to pla-cebo of 21.2% (p=0.0134).

A sensitivity analysis was conducted to assess the potential impact of potential unblinding in the primary outcome - time to 3-month CDP. This complete analysis can be consulted in the EPAR. Excluding patients at sites where the EDSS rater had temporary access to potentially unblinding information (213 patients potentially unblinded) the HR was 0.85 95% (0.69-1.05). Excluding data only from patients for whom EDSS blinded assessment could have been compromised (65 patients), the HR was 0.80 95%CI (0.66-0.97).

Treatment effect on 3-month CDP was evaluated in predefined subgroups.

In patients without superimposed relapses in the 2 years prior to study start, 3-month CDP events were observed in 190/708 patients (26.8%) in the siponimod group, and in 101/343 patients (29.5%) in the placebo group (hazard ratio 0.87; 95%CI 0.68 to 1.11). In patients with superimposed relapses in the 2 years prior to study start, 3-month CDP events were observed in 98/388 patients (26.8%) in the siponi-mod group, and in 72/202 patients (35.6%) in the placebo group (hazard ratio 0.67; 95%CI 0.49 to 0.91).



In patients without gadolinium-enhancing T1 lesion at baseline, 3-month CDP events were observed in 231/857 patients (26.9%) in the siponimod group, and in 137/432 patients (31.7%) in the placebo group (hazard ratio 0.82; 95%CI 0.66 to 1.01). In patients with at least one gadolinium-enhancing T1 lesion at baseline, 3-month CDP events were observed in 61/236 patients (25.8%) in the siponimod group, and in 40/114 patients (35.1%) in the placebo group (hazard ratio 0.64; 95%CI 0.42 to 0.95).

The MAH defined rapidly evolving patients as subjects with 1.5 or greater EDSS change in the 2 years prior to or at study start, and in whom disability progression in the 2 years prior to study start was not adjudicated. In rapidly evolving patients, 3-month CDP events were observed in 82/264 patients (31.1%) in the siponimod group, and in 60/145 patients (43.4%) in the placebo group (hazard ratio 0.65; 95%CI 0.46 to 0.91). In not rapidly evolving patients at baseline, 3-month CDP events were observed in 206/835 patients (24.7%) in the siponimod group, and in 113/401 patients (28.2%) in the placebo group (hazard ratio 0.86; 95%CI 0.69 to 1.09).

Results from the patients with active disease, defined as presence of relapses in the prior 2 years to screening and/or presence of T1 Gad lesion at baseline can be found in the EPAR. The hazard ratio for siponimod compared to placebo for 3m-CDP (primary endpoint) was 0.693 (0.525; 0.914) p=0.0094.

Time to 6-month CDP

At cut-off date (29-Apr-2016), 6-month CDP events were observed in 228/1096 (19.9%) patients in the siponimod group, and 139/545 CDP events (25.5%) were observed in placebo group (risk reduction 25.9%; hazard ratio 0.74; 95%CI 0.60 to 0.92; nominal p=0.0058).

Figure A3 and Figure A4 depicts a forest plot of time to 3-month CDP by subgroup (FAS) and time to 6-month CDP by subgroup (FAS), respectively.

In patients without superimposed relapses in the 2 years prior to study start, 6-month CDP events were observed in 144/708 patients (20.3%) in the siponimod group, and in 81/343 patients (23.6%) in the placebo group (hazard ratio 0.82; 95%CI 0.62 to 1.08). In patients with superimposed relapses in the 2 years prior to study start, 6-month CDP events were observed in 74/388 patients (19.1%) in the siponi-mod group, and in 58/202 patients (28.7%) in the placebo group (hazard ratio 0.63; 95%CI 0.44 to 0.89).

Overall, 10.7% of the patients in siponimod group, and 18.9% in placebo group had confirmed relapses during the study. In patients without superimposed relapses during the study, 6-month CDP events were observed in 178/986 patients (18.1%) in the siponimod group, and in 102/444 patients (23.0%) in the placebo group (hazard ratio 0.76; 95%CI 0.60 to 0.97).

In patients with superimposed relapses during the study, 6-month CDP events were observed in 40/113 patients (35.4%) in the siponimod group, and in 37/102 patients (36.3%) in the placebo group (hazard ratio 0.86; 95%CI 0.54 to 1.36).

In patients without gadolinium-enhancing T1 lesion at baseline, 6-month CDP events were observed in 165/828 patients (19.9%) in the siponimod group, and in 102/415 patients (24.6%) in the placebo group (hazard ratio 0.78; 95%CI 0.61 to 1.00). In patients with at least one gadolinium-enhancing T1 lesion at baseline, 6-month CDP events were observed in 47/236 patients (19.9%) in the siponimod group, and in 33/114 patients (28.9%) in the placebo group (hazard ratio 0.59; 95%CI 0.38 to 0.93).

In rapidly evolving patients, 6-month CDP events were observed in 65/264 patients (24.6%) in the siponimod group, and in 51/145 patients (35.1%) in the placebo group (hazard ratio 0.61; 95%CI 0.42 to 0.89). In not rapidly evolving patients at baseline, 6-month CDP events were observed in 153/835 patients (18.3%) in the siponimod group, and in 88/401 patients (21.9%) in the placebo group (hazard ratio 0.82; 95%CI 0.63 to 1.07).

Among patients with new or enlarging T2 lesions and relapses, 30% of placebo and 25% of siponimod patients experienced 6m-CDP based on EDSS after 3 years in the trial.



Results from the patients with active disease, defined as presence of relapses in the prior 2 years to screening and/or presence of T1 Gad lesion at baseline can be found in the EPAR. The hazard ratio for siponimod compared to placebo for 6m-CDP was 0.653 (0.466; 0.865) p=0.0040.

Figure A3. Time to 3-month CDP by subgroup (FAS)




Figure A4. Time to 6-month CDP by subgroup (FAS)


Time to 3-month confirmed worsening of at least 20% from baseline in T25FW

At cut-off date (29-Apr-2016), 3-month confirmed worsening of at least 20% from baseline in T25FW events were observed in 432/1087 (39.7%) patients in the siponimod group, and 225/543 3-month con-firmed worsening of at least 20% from baseline in T25FW (41.4%) were observed in placebo group (risk reduction 6.2%; hazard ratio 0.94; 95%CI 0.80 to 1.10; p=0.4398).

Figure A5 depicts Kaplan-Meier curves for the proportion of patients free of 3-month confirmed wors-ening of at least 20% from baseline in T25FW.



Figure A5. Patients free of 3-month confirmed worsening of at least 20% from baseline in T25FW


Figure A6 depicts a forest plot of 3-month confirmed worsening of at least 20% from baseline in T25FW in the various subgroups.

Figure A6. Time to 3-month confirmed worsening of at least 20% from baseline in T25FW by subgroup (FAS)




Change from baseline in T2 lesion volume

From baseline to month 12, the average change in T2 lesion volume was 204.9 mm3 in siponimod group, and 818.0 mm3 in placebo group (difference between groups: -613.1±95.39 [SE]; 95%CI -800.2 to -426.0; p<0.0001).

From baseline to month 24, the average change in T2 lesion volume was 162.9 mm3 in siponimod group, and 940.4 mm3 in placebo group (difference between groups: -777.5±108.62 [SE]; 95%CI -990.6 to -564.4; p<0.0001).

Figure A7 depicts a forest plot of change from baseline in T2 lesion volume in the various subgroups. In patients without superimposed relapses in the 2 years prior to study start, the change from baseline in T2 lesion volume was 168.5 in the siponimod group, and 641.7 in the placebo group (difference between groups -473.2; 95%CI -703.9 to -242.5). In patients with superimposed relapses in the 2 years prior to study start, the change from baseline in T2 lesion volume was 143.1 in the siponimod group, and 1209.0 in the placebo group (difference between groups -1065.8; 95%CI -1364.6 to -767.1).

Figure A7. Change from baseline in T2 lesion volume by subgroup (FAS)


Annualised relapse rate (ARR)

The adjusted ARR (negative binomial regression) was 0.071 (95%CI 0.055 to 0.092) for the siponimod group, and 0.160 (95%CI 0.123 to 0,207) for the placebo group (rate reduction 55.5%; ARR ratio 0.445; 95%CI 0.337 to 0.587; nominal p<0.0001).

Time to first confirmed relapse

At cut-off date (29-Apr-2016), 113/1067 relapses (10.7%) were observed in the siponimod group, and 100/528 relapses (18.9%) were observed in the placebo group (risk reduction 46.4%; hazard ratio 0.54; 95%CI 0.41 to 0.70; nominal p<0.0001).

Figure A8 depicts Kaplan-Meier curves for the proportion of relapse-free patients over time.



Figure A8. Proportion of relapse-free patients over time


Proportion of patients with relapse

Confirmed relapses were observed in 113/1099 patients (10.3%) in the siponimod group, and 102/546 patients (18.7%) in the placebo group.

Multiple Sclerosis Walking Scale (MSWS-12)

Analysis of change from baseline in MSWS-12 converted scores, showed that, at month 12, the mean change for the siponimod group was 1.53±0.678 (SE), and for the placebo group was 3.36±0.908 (dif-ference between groups: -1.83±1.030; 95%CI -3.85 to 0.19; nominal p= 0.0764). At month 24, the mean change for the siponimod group was 4.16±0.848 (SE), and for the placebo group was 5.38±1.167 (dif-ference between groups: -1.23±1.359; 95%CI -3.89 to 1.44; nominal p= 0.3671).

Other MRI-related variables

At month 12, the proportion of patients free of T1 gadolinium-enhancing lesions was 93.6% (954/1019) for the siponimod group, and 77.1% (391/507) for placebo group. At month 24, the proportion of patients free of T1 gadolinium-enhancing lesions was 95.3% (593/622) for the siponimod group, and 82.2% (250/304) for the placebo group.

At month 12, the number of T1 gadolinium-enhancing lesions per patient per scan, was 0.08 for the siponimod group, and 0.64 for the placebo group (rate reduction 84.4%; rate ratio 0.126; 95%CI 0.083 to 0.191; nominal p<0.0001). At month 24, the number of T1 gadolinium-enhancing lesions per patient per scan, was 0.0074 for the siponimod group, and 0.418 for the placebo group (rate reduction 82.2%; rate ratio 0.178; 95%CI 0.087 to 0.362; nominal p<0.0001).

At month 12, the number of new or enlarging T2 lesions, was 1.003 for the siponimod group, and 3.776 for the placebo group (rate reduction 73.3%; rate ratio 0.266; 95%CI 0.215 to 0.328; nominal p<0.0001). At month 24, the number of new or enlarging T2 lesions, was 0.489 for the siponimod group, and 3.437 for the placebo group (rate reduction 85.8%; rate ratio 0.142; 95%CI 0.103 to 0.196; nominal p<0.0001).

Percent brain volume change

At month 12, the percent brain volume change relative to baseline was -0.283 for the siponimod group, and -0.458 for the placebo group (difference between groups: 0.175; 95%CI 0.103 to 0.247; nominal p<0.0001).



At month 24, the percent brain volume change relative to baseline was -0.711 for the siponimod group, and -0.839 for the placebo group (difference between groups: 0.128; 95%CI 0.021 to 0.236; nominal p= 00196).

Symbol Digit Modalities Test (SDMT) oral score

At month 12, the mean adjusted score change from baseline was 0.148±0.288 for the siponimod group, and -0.945±0.384 for the placebo group (difference between groups 1.092 (95%CI 0.234 to 1.950; p= 0.0126).

At month 24, the mean adjusted score change from baseline was 0.142±0.418 for the siponimod group, and -1.059±0.593 for the placebo group (difference between groups 2.478; 95%CI 1.110 to 3.847; p= 0.0004).

Paced Auditory Addition Test (PASAT)

At month 12, the mean adjusted score change from baseline was 0.00±0.014 for the siponimod group, and -0.02±0.020 for the placebo group (difference between groups 0.02 (95%CI -0.03 to 0.07; nominal p= 0.374).

At month 24, the mean adjusted score change from baseline was 0.07±0.021 for the siponimod group, and -0.07±0.03 for the placebo group (difference between groups -0.01; 95%CI -0.08 to 0.07; nominal p= 0.856).

Brief Visuospatial Memory Test Revised (BVMT-R)

At month 12, the mean adjusted score change from baseline was 0.21±0.223 for the siponimod group, and 0.56±0.298 for the placebo group (difference between groups -0.35 (95%CI -1.01 to 0.32; nominal p= 0.307).

At month 24, the mean adjusted score change from baseline was 1.78±0.319 for the siponimod group, and 1.33±0.452 for the placebo group (difference between groups -0.45; 95%CI -0.59 to 1.49; nominal p= 0.394).

Multiple Sclerosis Impact Scale (MSIS-29)

At month 12, the mean adjusted score change from baseline was 1.33±0.649 for the siponimod group, and 4.22±0.869 for the placebo group (difference between groups -2.89; 95%CI -4.83 to -0.96; nominal p= 0.0034).

At month 24, the mean adjusted score change from baseline was 3.31±0.816 for the siponimod group, and 4.67±1.123 for the placebo group (difference between groups -1.36; 95%CI -3.93 to 1.21; nominal p= 0.300).

Health-Related Quality of Life (EQ-5D-3L)

At month 12, the mean adjusted utility index score change from baseline was -0.023±0.007 for the siponimod group, and -0.049±0.010 for the placebo group (difference between groups 0.026 (95%CI 0.002 to 0.050; nominal p= 0.0368).

At month 24, the mean adjusted score change from baseline was -0.040±0.011 for the siponimod group, and -0.072±0.0158 for the placebo group (difference between groups 0.033 (95%CI -0.005 to 0.070; nominal p= 0.088).

Safety

The median duration of exposure to double-blind study drug was 18.5±8.4 months in siponimod group, and 17.7 months in placebo group.

Adverse events were observed in 975/1099 patients (88.7%) in the siponimod group, and in 445/546 patients (81.5%) in placebo group. Serious adverse events were observed in 197/1099 patients (17.9%) in the siponimod group, and in 83/546 patients (15.2%) in placebo group. In the safety set population,



84/1099 patients (7.6%) in siponimod group, and 28/546 patients (5.1%), discontinued from study drug permanently due to adverse events. Overall, 4/1099 patients (0.4%) in siponimod group, and 4/546 patients (0.7%) in placebo group died during double-blind study.

Adverse events most frequently reported (≥3%) are depicted in Table A27.

Table A27. Most frequently reported adverse events (≥3%)


Most frequently reported treatment emergent serious adverse events on double-blind study drug are depicted in Table A28.



Table A28. Most frequently reported treatment emergent SAEs




Adverse events by preferred term, that were more frequently reported in the siponimod group were headache (14.5% vs 13.0%), hypertension (10.5% vs 7.5%), dizziness (6.8% vs 4.8%), nausea (6.7% vs 3.5%), increased ALT (5.9% vs 1.5%), bradycardia (4.5% vs 2.6%), peripheral oedema (4.5% vs 2.4%), and increased gamma-glutamyltransferase (3.9% vs 1.1%).

The most frequently reported treatment emergent adverse events (TEAEs) causing permanent study drug discontinuation were, in the siponimod and placebo groups, respectively: macular oedema (1.0% vs 0.2%), increased alanine aminotransferase (0.5% vs 0%), bradycardia (0.4% vs 0%), increased as-partate aminotransferase (0.3% vs 0%), depression (0.3% vs 0.2%), dizziness (0.3% vs 0%), fatigue (0.3% vs 0.7%), increased gamma-glutamyltransferase (0,3% vs 0%), and decreased pulmonary func-tion test (0.3% vs 0%).

The most frequently reported TEAEs causing permanent study drug discontinuation are depicted in Table A29.

Table A29. Most frequently reported TEAEs causing permanent study drug discontinuation


Adverse events of special interest that were reported more frequently in the siponimod group were herpes zoster reactivations (2.2% vs 0.7%), lymphopenia (1.6% vs 0%), macular oedema (1.7% vs 0.2%), and increased liver transaminases (1.4% vs 0.6%).

Genetic polymorphism of CYP2C9 has a great influence on the overall metabolism, therefore knowledge of metaboliser status (extensive, intermediate or poor) is required before initiation of treat-ment. The influence of the genotypes *1/*3 and *2/*3 are reflected in the SmPC with mandated reduction in dose to 1 mg EACH DAY maintenance treatment. The combination of siponimod with moderate or strong CYP3A4 inhibitors or moderate CYP2C9/CYP3A4 dual inhibitor should be avoided irrespective of the CYP2C9 genotype.



APPENDIX 6: ANCILLARY ANALYSIS

Indirect comparisons between siponimod and comparators of interest in the global

SPMS population

A summary of the results of the ancillary analysis on the global population is available in Section 4.8 and depicted in Table 4.15.

The SLR search retrieved seven studies in patients with SPMS (with and without relapses): one study of siponimod, two studies of interferon beta-1b, three studies of interferon beta-1a, and one study of natalizumab. These studies were: EXPAND (siponimod), ASCEND (natalizumab), SPECTRIMS (IFNβ-1a 22 and 44 µg three times weekly), North American Study (interferon β-1b 250 µg every other day), European Study (interferon β-1b 250 µg every other day), Nordic SPMS study (IFN β-1a 22 µg once a week), and IMPACT study (interferon β-1a 60 µg once a week) (Table A18). No studies of mitoxantrone, ocrelizumab, fingolimod, cladribine, or rituximab, in the SPMS population were identified.

Although the global SPMS population is not the population of interest, the results of studies in this population may contribute to a better understanding of treatment effects in the relapsing SPMS popu-lation. Therefore, the results in the global population are summarised below.

The evidence networks for the outcomes of CDP-6, CDP-3 and ARR are presented in Figure A9, Figure A10 and Figure A11, respectively.

Figure A9. Network diagram for time to confirmed disability progression at 6 Months

Source: adapted from the technical reports provided by the MAH Abbreviations: IFNβ-1b250=interferon-beta-1b 250 micrograms; IV=intravenous; PBO=placebo; PO=oral; once a week=once

weekly; REB22=Rebif® (IFNβ-1a) 22 micrograms; SC=subcutaneous; SIP2=siponimod 2 milligrams.



Figure A10. Network diagram for time to confirmed disability progression at 3 Months


Abbreviations: AVO60=Avonex® (IFNβ-1a) 60 micrograms; IFNβ-1b250=interferon-beta-1b 250 micrograms; IM=intramuscular; PBO=placebo; PO=oral; REB22=Rebif® (IFNβ-1a) 22 micrograms; REB44=Rebif® (IFNβ-1a) 44 micrograms; SC=subcutaneous; SIP2=siponimod 2 milligrams; three times weekly=three times weekly.

Figure A11. Network diagram for annualised relapse rate


Abbreviations: AVO60=Avonex® (IFNβ-1a) 60 micrograms; IFNβ-1b250=interferon-beta-1b 250 micrograms; IM=intramuscular; IV=intravenous; NAT300=natalizumab 300 milligrams; PBO=placebo; PO=oral; once a week=once weekly; REB22=Rebif® (IFNβ-1a) 22 micrograms; REB44=Rebif® (IFNβ-1a) 44 micrograms; SC=subcutaneous; SIP2=siponimod 2 milligrams; three times

weekly=three times weekly.



Comparison of siponimod vs interferon beta-1a 22 µg once a week

Comments on ITC feasibility assessment

As a general comment we would like to emphasize that although the MAH assessed plausibility of effect modifiers (main dossier Tables 35 and 36), in practice the variables assessed were not always rated important in their own analysis. It is therefore unclear whether adjustment was needed in some cases, and whether it improved the comparability of the study results.

The comparison of siponimod vs interferon beta-1a 22 µg once a week included two studies (EXPAND and Nordic SPMS study).

Study design: there were no significant differences between the two studies. The MAH underlines that the Nordic SPMS study terminated early following the release of the non-significant results from the SPECTRIMS Study Group (2001) using IFNβ-1a with higher and more frequent dosing. The difference in study duration between studies could be adjusted for, by the MAIC. However, the mean time on treatment for IFNβ-1a and siponimod was 35.0 months and 21 months, respectively. The clinical rele-vance of the difference between groups is uncertain.

Inclusion/exclusion criteria: Based on the pre-defined threshold of 10%, the MAH states that there was a difference in the baseline EDSS range between studies, as the EXPAND study included patients with an EDSS range between 3.0 and 6.5, and the Nordic SPMS study included patients with an EDSS score of 6.5 or lower (the published report reports a threshold of 7.0). The mean EDSS score at baseline was, for the siponimod and IFNβ-1a groups, 5.4 and 4.8, respectively, a score difference of 0.6. Addi-tionally, the MAH underlines that the age range of patients included in siponimod and IFNβ-1a studies, which are 18-60 years and 18-65 years, respectively, is significantly different. However, this difference is unlikely to be clinically relevant. Furthermore, patients eligible for the EXPAND study could not have any relapse in the 3 months prior to study start, while patients in the Nordic SPMS study could not have any relapse in the 2 months prior to study start. The MAH states that this is a relevant difference, and indicates the presence of heterogeneity. However, for these three characteristics, using a MAIC had no advantage over a conventional NMA, as the population included in EXPAND study could not be matched to the population included in Nordic SPMS. According to the MAH, in the EXPAND study patients were allowed to receive prior IFN therapy while in the Nordic SPMS study there was no prior IFN use. However, this information is not included in the published report of the Nordic SPMS study. In fact, prior to study start, patients included in the EXPAND study received several DMTs therapies (in-cluding 40.5% interferon beta-1a and 28.5% interferon beta-1b), and it is unlikely that patients included in the Nordic SPMS study had not received, prior to randomisation, interferon, which was the only treat-ment available at the time to treat MS. The MAH also underlines the difference between groups in the duration of MS, as there was no restriction in EXPAND, while in Nordic SPMS Study patients had to have MS with at least 1-year duration. At baseline, the mean duration of MS was, for the siponimod and interferon beta-1a groups, 12.6 and 14.3 years. The difference in MS duration between studies could be overcome by matching the population included in EXPAND study to that included in Nordic SPMS, but the clinical relevance of this difference is uncertain.

Outcome definitions: as per scope, the outcome measures to be considered in this assessment in-cluded disability, symptoms, clinical relapse, mortality, health-related quality of life, MRI inflammatory measured disease activity, adverse events, serious adverse events, and adverse events leading to treatment discontinuation. According to the MAH, “The outcomes of interest were determined based on key primary and secondary outcomes evaluated in the EXPAND trial. The primary outcome of the EX-PAND trial was the time to CDP-3. Key secondary outcomes in the EXPAND trial included time to CDP-6, and annualised relapse rate (ARR). Measures of CDP and ARR are common outcomes amongst trials in SPMS. Discontinuation (all-cause) was explored in the feasibility assessment for potential Bucher ITCs only; discontinuation was not a target outcome for any potential matching and/or adjusted comparison methodologies because treatment effect modifiers related to adverse events and discon-tinuation are not well-reported, thereby precluding a valid matched and/or adjusted comparison”. The MAH considered the outcomes of interest to be CDP-3, CDP-6, and ARR. There were no major differ-ences between studies in outcome definitions.



Baseline patient characteristics: for baseline patient characteristics which were available in both studies, mean duration of MS (14.3 vs 12.6 years), and mean duration of SPMS (5.4 vs 3.8) were longer in IFNβ-1a patients than in siponimod patients, and the difference could be adjusted for, by the MAIC. Furthermore, the mean EDSS score at baseline was, for the siponimod and IFNβ-1a groups, 5.4 and 4.8, respectively, a score difference of 0.6. This characteristic could also be adjusted for, by the MAIC.

Treatment effect on the placebo arm: treatment effect on the outcomes ARR and annualised rate of treatment discontinuation were dissimilar (>10% different) in the two studies. ARR was 0.16 and 0.27, in siponimod and IFNβ-1a groups, respectively. Discontinuation was 0.084 and 0.060, in siponimod and IFNβ-1a groups, respectively. However, when performing ITCs, treatment effect was not adjusted for differences in treatment effect on placebo arms.

Indirect treatment comparison using MAIC in the global SPMS population

Siponimod was compared with IFNβ-1a 22 µg once a week by MAIC. The matching process for inclu-sion/exclusion criteria differences is summarised in Table A30.

Table A30. Matching inclusion/exclusion criteria – Nordic SPMS Study vs EXPAND

Inclusion / Exclusion Criteria Category

Nordic SPMS Study

EXPAND

(after matching)

Comment

Adults with SPMS Yes Yes Identical criteria

Baseline EDSS Range ≤6.5 3.0-6.5 Removed patients with baseline EDSS score >6.5

Nordic Study did not specify a lower boundary for EDSS

Age Range 18-65 18-60 Cannot match since Nordic study includes broader age range than EXPAND

Prior IFN Therapy Excluded Excluded Removed patients with prior IFNβ therapy*

No Recent Relapse in Specified Time-Frame

<2 months <3 months Cannot match since Nordic study’s criteria on relapses within a time-frame is broader than EXPAND.

Documented Progression in Specified Time-Frame

Progression for 6 months

in past 4 years

Progression in prior 24 months;

progression for at least 6 months

Nordic Study was broader than EXPAND, preventing matching the EXPAND IPD to the Nordic criterion

Duration of MS ≥1 year ≥1 year Removed patients with MS duration <1 year

Duration of SPMS - - -

MS Severity Score - - -

T25FW Test Score - - -

Source: adapted from the technical reports provided by the MAH *Authors state inclusion criteria are “similar” to SPECTRIMS. Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; SPMS=secondary progressive multiple sclerosis.



The adjustment process for CDP is presented in Table A31. No adjustment was made for differences in baseline patient characteristics for the outcome ARR. The MAH did not show that these patient char-acteristics were effect modifiers.

Table A31. Adjustment factors for CDP – Nordic SPMS Study

CDP

Rank Adjustment Factor (Treatment Effect Modifier)

Nordic SPMS Study

1 Age ✓

2 EDSS score at screening ✓

3 Duration of MS since diagnosis ✓

4 Treatment experience (IFN or DMT history) n/a (matched)

5 Normalised brain volume X

6 Gadolinium-enhancing lesions on T1-weighted images X

7 Duration of SPMS ✓

8 Total volume of T2 lesions on T2-weighted images X

9 Number of relapses in prior 2 years (or any other relapse variable) X

10 Sex ✓


✓=reported (i.e., could be adjusted); X=not reported (i.e., could not be adjusted).Abbreviations: CDP=confirmed disability pro-

gression; DMT=disease-modifying therapy; EDSS=Expanded Disability Status Scale; IFN=interferon; SPMS=secondary pro-gressive multiple sclerosis.

Results

Sample size of the EXPAND population decreased from 1645 to 157 patients as a result of matching and adjustment. This makes any results unreliable. The results of population matching and adjustment for CDP is summarised in Table A32.

Table A32. Results of population matching and adjustment for CDP

Variables Nordic Study a EXPAND (unmatched)

EXPAND (matched and unadjusted)

EXPAND (matched and

adjusted)

N (Neff) 364 1642 578 157

Age (mean years [SD]) 45.7 (7.1) 48.02 (7.86) 49.29 (7.75) 45.7 (7.11)

EDSS score at screening (mean [SD])

4.8 (1.1) 5.42 (1.06) 5.26 (1.11) 4.8 (1.1)

MS duration since diagnosis (mean years [SD])

14.3 (7.1) 12.61 (7.76) 12.23 (8.44) 14.3 (7.11)

Duration of SPMS (mean years [SD])

5.4 (3) 3.76 (3.51) 3.69 (3.48) 5.4 (3.01)

Sex (proportion female) 60.0% 59.99% 60.38% 60.0%

Source: adapted from the technical reports provided by the MAH Matched sample excludes patients with EDSS scores >0.5, MS duration <1 year, and those with prior IFNβ therapy. The Nordic

Study did not report standard deviations for adjusting factors. Results use values from SPECTRIMS. a Nordic SPMS Study did not report standard deviation for baseline patient characteristics, to match and adjust the EXPAND patient population to the Nordic population, standard deviation was imputed from SPECTRIMS.

Abbreviations: CDP=confirmed disability progression; EDSS=Expanded Disability Status Scale; Neff=effective sample size SD=standard deviation; SPMS=secondary progressive multiple sclerosis.



The MAIC, suggested that siponimod, in comparison with IFNβ-1a 22 µg once a week, reduced CDP-6 by 57% (hazard ratio 0.43; 95%CI 0.20 to 0.93). Similar results were obtained using Bucher ITC (hazard ratio 0.66; 95%CI 0.44 to 0.97). As expected, the 95%CI was narrower with Bucher ITC due to the larger sample size. Results are depicted in Figure A12, but note that matched and adjusted results apply only to the population in the Nordic study.

Figure A12. MAIC scenario analysis results for time to CDP-6 (siponimod vs IFNβ-1a 22 µg once a week)

Source: adapted from the technical reports provided by the MAH The matched sample excludes patients with EDSS scores >6.5, MS duration < 1 year, and those prior interferon therapy. Sce-

nario A adjusts for all available ranked characteristics, subsequent scenarios drop the least important characteristics from ad-justment. Abbreviations: Neff=effective sample size.

The MAIC, including EXPAND matched but unadjusted, suggested that siponimod, in comparison with IFNβ-1a 22 µg once a week, reduced ARR by 41% (hazard ratio 0.59; 95%CI 0.32 to 1.07), but the difference did not reach statistical meaning. Bucher ITC results were similar, although more precise (hazard ratio 0.50; 95%CI 0.32 to 0.78). This data is depicted in Figure A13.

Figure A13. MAIC results for ARR (siponimod vs IFNβ-1a 22 µg once a week)

Source: adapted from the technical reports provided by the MAH The matched sample excludes patients with EDSS scores >6.5, MS duration < 1 year, and those prior interferon therapy.

Abbreviations: Neff=effective sample size; µg=microgram.



Indirect treatment comparison using network meta-analysis

As expected results from the NMA were identical to the Bucher ITC (see Section 3.4).

Indirect treatment comparison using simulated treatment comparison

The simulated treatment comparison (STC), indicated that siponimod, in comparison with IFNβ-1a 22 µg once a week, reduced CDP-6 by 50% (rate ratio 0.50; 95%CI 0.29 to 0.87). Siponimod, in compari-son with IFNβ-1a 22 µg once a week, reduced ARR by 36% (rate ratio 0.64; 95%CI 0.36 to 1.16), but the difference did not reach statistical significance. These results are comparable to the MAIC and Bucher ITC results, although more uncertain.

Table A33. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1a 22 µg once a week)

ITC method

Outcome MAIC STC Bucher ITC NMA

CDP-3 ------ ------ ------ ------

CDP-6 0.43

(0.20 to 0.93)

0.50

(0.29 to 0.87)

0.66

(0.44 to 0.97)

0.65

(0.44 to 0.97)

ARR 0.59

(0.32 to 1.07)

0.64

(0.36 to 1.16)

0.50

(0.32 to 0.78)

0.50

(0.32 to 0.78)

Source: adapted from the technical reports provided by the MAH p-values not available

Abbreviations: MAIC=matched adjusted indirect comparison; STC=simulated treatment comparison; ITC=indirect treatment comparison; NMA=network meta-analysis; results are hazard ratio (95% confidence interval) for MAIC and Bucher ITC, rate ratio (95% confidence interval) for STC, and rate ratio (95% credible interval) for NMA

Comparison of siponimod vs interferon beta-1a 22 µg three times weekly


The comparison of siponimod vs interferon beta-1a 22 µg three times weekly included two studies (EX-PAND and SPECTRIMS study).

Study design: there were no significant differences between the two studies.

Inclusion/exclusion criteria: patients eligible for the EXPAND study could not have any relapse in the 3 months prior to study start, while patients in the SPECTRIMS study could not have any relapse in the 2 months prior to study start. The MAH states that this is a relevant difference, and indicates the pres-ence of heterogeneity. However, for this characteristic, a using a MAIC had no advantage over a con-ventional NMA, as the difference in this characteristic between studies could not be adjusted for. Fur-thermore, the MAH states that the age range of patients included in EXPAND and SPECTRIMS, which are 18-60 years and 18-55 years, respectively, is significantly different, and matching was possible (MAIC). However, this difference is unlikely to be clinically relevant, so it is doubtful whether this should have been adjusted for. According to the MAH, in the EXPAND study patients were allowed to receive prior IFN therapy while in the SPECTRIMS study there was no prior IFN use. In fact, prior to study start, patients included in EXPAND received several DMTs therapies (including 40.5% interferon beta-1a and 28.5% interferon beta-1b). For this characteristic matching was possible as the population in EXPAND was broader. There was no information on the duration of MS, and duration of SPMS for the SPEC-TRIMS study and, therefore, these characteristics could not be taken into account.

Outcome definitions: There were no major differences between studies in the outcomes definitions.



Baseline patient characteristics: patient characteristics which were available in both studies, mean age (48.0 vs 42.8 years), and proportion of patients relapse-free in prior 2 years (64% vs 53% ) were higher in siponimod patients than in IFNβ-1a patients, and the difference could be adjusted for by the MAIC. There was no difference in the proportion of patients relapse-free in prior year.

Treatment effect on the placebo arm: treatment effect on the outcomes ARR and annualised rate of treatment discontinuation were dissimilar (>10% different) in the two studies. ARR was 0.16 and 0.71, in siponimod and IFNβ-1a groups, respectively. Discontinuation was 0.084 and 0.057, in siponimod and IFNβ-1a groups, respectively. However, when performing ITCs, treatment effect was not adjusted for differences in treatment effect on placebo arms.

Indirect treatment comparison using Matching Adjusted Indirect Comparison

Siponimod was compared with IFNβ-1a 22 µg three times weekly by MAIC. The matching process for inclusion/exclusion criteria differences is summarised in Table A34.

Table A34. Matching inclusion/exclusion criteria – SPECTRIMS vs EXPAND


SPECTRIMS EXPAND

(after matching)

Comment


Baseline EDSS Range 3.0-6.5* 3.0-6.5 Removed patients with baseline EDSS score <3.0 and >6.5*

Age Range 18-55 18-55 Removed patients >55 years old

Prior IFN Therapy Excluded Excluded Removed patients with prior IFNβ therapy


<2 months <3 months Cannot match since SPECTRIMS’s criteria on relapses within a time-frame is broader than EXPAND.


Progression for 6 months

in past 2 years

Progression in prior 24 months; progression for at least 6 months

Very similar criteria, did not require matching

Duration of MS - - -




Source: adapted from the technical reports provided by the MAH Although the EXPAND inclusion criteria specified an EDSS of 3.0-6.5, eleven patients participated in the trial despite a baseline EDSS outside this range and were included in the published analyses.

Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; MS=multiple sclerosis; SPMS=secondary progres-sive multiple sclerosis; T25FW=timed 25-foot walk test.

The adjustment process for CDP is presented in Table A35, and for ARR is depicted in Table A36. For ARR, most baseline patient characteristics which were reported in EXPAND study were not reported in SPECTRIMS and, therefore, no comparison was possible.



Table A35. Adjustment factors for CDP – SPECTRIMS

CDP


SPECTRIMS

1 Age ✓








9 Number of relapses in prior 2 years (or any other relapse variable)

✓

10 Sex ✓


✓=reported (i.e., could be adjusted); X=not reported (i.e., could not be adjusted).

Abbreviations: CDP=confirmed disability progression; DMT=disease-modifying therapy; EDSS=Expanded Disability Status Scale; IFN=interferon; SPMS=secondary progressive multiple sclerosis.

Table A36. Adjustment factors for ARR – SPECTRIMS

ARR

Rank Adjustment Factor

(Treatment Effect Modifier)

SPECTRIMS

1 Time since onset of most recent relapse X

2 Number of relapses per patient in one year prior to study X

3 Number of relapses per patient in two years prior to study ✓


5 Total volume of lesions on T2-weighted images X



Abbreviations: ARR=annualised relapse rate.

Results

Sample size of the EXPAND population decreased from 1,645 to 237 patients as a result of matching and adjustment. This makes any results unreliable. The results of population matching and adjustment for CDP are summarised in Table A37, but note that matched and adjusted results apply only to the population in the SPECTRIMS study.



Table A37. Results of population matching and adjustment for CDP-3

Variables SPECTRIMS EXPAND (unmatched)

EXPAND (matched and

unadjusted)

EXPAND (matched and

adjusted)

N (Neff) 618 1638 455 237

Age (mean years [SD]) 42.8 (7.1) 48.03 (7.84) 46.43 (6.81) 42.8 (7.11)


5.4 (1.1) 5.42 (1.06) 5.19 (1.11) 5.4 (1.1)


13.3 (7.1) 12.62 (7.77) 11.06 (7.91) 13.3 (7.11)


4 (3) 3.77 (3.51) 3.42 (3.19) 4 (3)

Number of relapses in prior 2 years (mean [SD])

0.9 (1.3) 0.67 (1.19) 0.71 (1.08) 0.9 (1.3)


Source: adapted from the technical reports provided by the MAH Matched sample excludes patients >55 years old, EDSS <3 or >6.5, and those with prior IFNβ therapy.

Abbreviations: CDP=confirmed disability progression; EDSS=Expanded Disability Status Scale; Neff=effective sample size

SD=standard deviation; SPMS=secondary progressive multiple sclerosis.

The MAIC, suggested that siponimod, in comparison with IFNβ-1a 22 µg three times weekly, reduced CDP-3 by 20% (hazard ratio 0.80; 95%CI 0.46 to 1.38), but the difference did not reach statistical sig-nificance. Similar results were obtained using Bucher ITC (hazard ratio 0.90; 95%CI 0.66 to 1.22). As expected, 95%CI was narrower with Bucher ITC. This data is depicted in Figure A14.

Figure A14. MAIC scenario analysis results for time to CDP-3 – siponimod vs IFNβ-1a 22 µg three times weekly

Source: adapted from the technical reports provided by the MAH A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients >55 years old, baseline EDSS <3 or >6.5, and those with prior interferon therapy. Scenario A adjusts for all available

ranked characteristics, subsequent scenarios drop the lowest-ranked characteristic from adjustment. Abbreviations: CDP-3=confirmed disability progression at three months; CI=confidence interval; MAIC=matching-adjusted

indirect comparison; Neff=effective sample size; µg=microgram.



The MAIC, including EXPAND matched and adjusted, suggested that siponimod, in comparison with IFNβ-1a 22 µg three times weekly, reduced ARR by 27% (hazard ratio 0.73; 95%CI 0.40 to 1.32), but the difference did not reach statistical meaning. Using Bucher ITC results were more precise but in broad agreement (hazard ratio 0.65; 95%CI 0.46 to 0.92). As expected, the 95%CI was narrower with Bucher ITC. These results are depicted in Figure A15, but note that matched and adjusted results apply only to the population in the SPECTRIMS study.

Figure A15. MAIC Results for ARR – Siponimod vs IFNβ-1a 22 µg three times weekly


A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients >55 years old, baseline EDSS <3 or >6.5, and those with prior interferon therapy. Abbreviations: ARR=annualised relapse rate; CI=confidence interval; MAIC=matching-adjusted indirect comparison; Neff=ef-

fective sample size; µg=microgram.


As expected results from the NMA were identical to the Bucher ITC (see Section3.4).


The simulated treatment comparison (STC), indicated that siponimod, in comparison with IFNβ-1a 22 µg three times weekly, reduced CDP-3 by 28% (rate 0.72; 95% CI 0.47 to 1.10), but the difference did not reach statistical significance. Siponimod, in comparison with IFNβ-1a 22 µg three times weekly, reduced ARR by 17% (rate ratio 0.83; 95%CI 0.49 to 1.40), but the difference did not reach statistical significance.

Table A38. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1a 22 µg three times weekly)

ITC method


CDP-3 0.80

(0.46 to 1.38)

0.72

(0.47 to 1.10)

0.90

(0.66 to 1.22)

p-value (0.4919)

0.90

(0.66 to 1.22)

CDP-6 ------ ------ ------ ------

ARR 0.73

(0.40 to 1.32)

0.83

(0.49 to 1.40)

0.65

(0.46 to 0.92)

p-value (0.0143)

0.65

(0.46 to 0.92)


p-values available for Bucher ITC are are reported in the table. p-values not available for other comparisons Abbreviations: MAIC=matched adjusted indirect comparison; STC=simulated treatment comparison; ITC=indirect treatment comparison; NMA=network meta-analysis; results are hazard ratio (95% confidence interval) for MAIC and Bucher ITC, rate

ratio (95% confidence interval) for STC, and rate ratio (95% credible interval) for NMA



Comparison of siponimod vs interferon beta-1a 44 µg three times weekly


The comparison of siponimod vs interferon beta-1a 44 µg three times weekly included two studies (EX-PAND and SPECTRIMS).

Study design: please see the comparison of siponimod vs interferon beta-1a 22 µg three times weekly.

Inclusion/exclusion criteria: please see the comparison of siponimod vs interferon beta-1a 22 µg three times weekly.

Outcome definitions: please see the comparison of siponimod vs interferon beta-1a 22 µg three times weekly.

Baseline patient characteristics: please see the comparison of siponimod vs interferon beta-1a 22 µg three times weekly.


Siponimod was compared with IFNβ-1a 44 µg three times weekly by MAIC. The matching process for inclusion/exclusion criteria differences is summarised in Table A39.

Table A39. Matching inclusion/exclusion criteria (SPECTRIMS vs. EXPAND)


SPECTRIMS EXPAND

(after matching)

Comment


Baseline EDSS Range 3.0-6.5* 3.0-6.5 Removed patients with baseline EDSS score <3.0 and >6.5*




<2 months <3 months Cannot match since SPECTRIMS’s criteria on relapses within a time-frame is broader than EXPAND.


Progression for 6 months in

past 2 years

Progression in prior 24 months; progression for at least 6 months

Very similar criteria, did not require matching





Source: adapted from the technical reports provided by the MAH Although the EXPAND inclusion criteria specified an EDSS of 3.0-6.5, eleven patients participated in the trial despite a baseline EDSS outside this range and were included in the published analyses.

Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; MS=multiple sclerosis; SPMS=secondary progres-sive multiple sclerosis; T25FW=timed 25-foot walk test.



The adjustment process for CDP is presented in Table A40, and for ARR is depicted in Table A41. For ARR, most baseline patient characteristics which were reported in EXPAND study were not reported in SPECTRIMS and, therefore, no comparison was possible.

Table A40. Adjustment factors for CDP – SPECTRIMS

CDP


SPECTRIMS

1 Age ✓









✓

10 Sex ✓




Table A41. Adjustment factors for ARR – SPECTRIMS

ARR


SPECTRIMS

1 Time since onset of most recent relapse X


3 Number of relapses per patient in two years prior to study ✓






Results

Sample size of the EXPAND population decreased from 1645 to 237 patients as a result of matching and adjustment. This makes the results unreliable. The results of population matching and adjustment for CDP are summarised in Table A42.




Variables SPECTRIMS EXPAND (unmatched)


EXPAND (matched and

adjusted)

N (Neff) 618 1638 455 237

Age (mean years [SD]) 42.8 (7.1) 48.03 (7.84) 46.43 (6.81) 42.8 (7.11)

EDSS score at screening (mean [SD]) 5.4 (1.1) 5.42 (1.06) 5.19 (1.11) 5.4 (1.1)


13.3 (7.1) 12.62 (7.77) 11.06 (7.91) 13.3 (7.11)

Duration of SPMS (mean years [SD]) 4 (3) 3.77 (3.51) 3.42 (3.19) 4 (3)


0.9 (1.3) 0.67 (1.19) 0.71 (1.08) 0.9 (1.3)


Source: adapted from the technical reports provided by the MAH Matched sample excludes patients >55 years old, EDSS <3 or >6.5, and those with prior IFNβ therapy.



The MAIC, suggested that siponimod, in comparison with IFNβ-1a 44 µg three times weekly, reduced CDP-3 by 16% (hazard ratio 0.84; 95%CI 0.49 to 1.47), but the difference did not reach statistical sig-nificance. Similar results were obtained using Bucher ITC (hazard ratio 0.95; 95%CI 0.70 to 1.30). As expected, 95%CI was narrower with Bucher ITC. This data is depicted in Figure A16, but note that matched and adjusted results apply only to the population in the SPECTRIMS study.

Figure A16. MAIC scenario analysis results for Time to CDP-3 – siponimod vs IFN beta-1a 44 µg three times weekly


A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients >55 years old, EDSS <3 or >6.5, and those with prior interferon therapy. Scenario A adjusts for all available ranked characteristics, subsequent scenarios drop the lowest ranked characteristics from adjustment.

Abbreviations: CDP-3=confirmed disability progression at three months; CI=confidence interval; ITC=indirect treatment com-parison; MAIC=matching-adjusted indirect comparison; Neff=effective sample size; µg=microgram.



The MAIC, including EXPAND matched and adjusted, suggested that siponimod, in comparison with IFNβ-1a 44 µg three times weekly, reduced ARR by 27% (hazard ratio 0.73; 95%CI 0.40 to 1.32), but the difference did not reach statistical significance. Using Bucher ITC results were more precise but in broad agreement (hazard ratio 0.65; 95%CI 0.46 to 0.92). As expected, 95%CI was narrower with Bucher ITC. These results are depicted in Figure A17, but note that matched and adjusted results apply only to the population in the SPECTRIMS study.

Figure A17. MAIC results for ARR – siponimod vs IFNβ-1a 22 µg three times weekly


A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients >55 years old, baseline EDSS <3 or >6.5, and those with prior interferon therapy. Abbreviations: ARR=annualised relapse rate; CI=confidence interval; MAIC=matching-adjusted indirect comparison; Neff=ef-

fective sample size; µg=microgram.




The simulated treatment comparison (STC), indicated that siponimod, in comparison with IFNβ-1a 22 µg three times weekly, reduced CDP-3 by 24% (rate 0.76; 95%CI 0.50 to 1.17), but the difference did not reach statistical significance. Siponimod, in comparison with IFNβ-1a 44 µg three times weekly, reduced ARR by 17% (rate ratio 0.83; 95%CI 0.49 to 1.40), but the difference did not reach statistical significance.

Table A43. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1a 44 µg three times weekly)

ITC method


CDP-3 0.84

(0.49 to 1.47)

0.76

(0.50 to 1.17)

0.95

(0.70 to 1.30)

p-value (0.7573)

0.95

(0.70 to 1.30)

CDP-6 ------ ------ ------ ------

ARR 0.73

(0.40 to 1.32)

0.83

(0.49 to 1.40)

0.65

(0.46 to 0.92)

p-value (0.0154)

0.65

(0.46 to 0.92)

Source: adapted from the technical reports provided by the MAH p-values available for Bucher ITC are are reported in the table. p-values not available for other comparisons Abbreviations: MAIC=matched adjusted indirect comparison; STC=simulated treatment comparison; ITC=indirect treatment

comparison; NMA=network meta-analysis; results are hazard ratio (95% confidence interval) for MAIC and Bucher ITC, rate ratio (95% confidence interval) for STC, and rate ratio (95% credible interval) for NMA



Comparison of siponimod vs interferon beta-1b 250 µg every other day


The comparison of siponimod vs interferon beta-1b 250 µg every other day included three studies (EX-PAND, the North American Study and the European Study). The North American Study only reported time to CDP-6 and the European Study only reported time to CDP-3.

EXPAND vs North American Study

Study design: there were no significant differences between the two studies.

Inclusion/exclusion criteria: Based on the threshold of 10%, the MAH states that the age range of patients included in EXPAND and North American studies, which are 18-60 years and 18-65 years, respectively, is significantly different. However, this difference is unlikely to be clinically relevant. Fur-thermore, patients eligible for the EXPAND study should not have any relapse in the 3 months prior to study start, while patients in the North American study should not have any relapse in the 2 months prior to study start. The MAH states that this is a relevant difference, which indicates the presence of heterogeneity. However, for these two characteristics, using a MAIC had no advantage over a conven-tional NMA, as the population included in EXPAND study could not be matched to the population in-cluded in the North American Study for the above variables.

Additionally, in the EXPAND study, patients were allowed to receive prior IFN therapy while in the North American Study there was no prior IFN use. The MAH also underlines the difference between studies in the duration of MS, as there was no restriction in EXPAND, while in North American Study patients had to have MS with at least 2-year duration. At baseline, the mean duration of MS was, for the siponi-mod and interferon beta-1b groups, 12.6 and 14.7 years. Matching was possible for prior IFN use and duration of MS, but the clinical relevance of those differences is uncertain.

Outcome definitions: There were no major differences between studies in the outcomes definitions.

Baseline patient characteristics: as far as patient characteristics which were available in both studies, duration of MS (12.6 vs 14.7 years) was higher in the North American Study, the proportion of patients relapse-free in the previous 2 years (64% vs 55% ) was higher in siponimod patients, and the difference between studies could be adjusted for, by the MAIC.

Treatment effect on the placebo arm: treatment effect on the outcomes ARR and annualised rate of treatment discontinuation were dissimilar (>10% different) in the two studies. ARR was 0.16 and 0.28, in siponimod and IFNβ-1b groups, respectively. Discontinuation was 0.084 and 0.093, in siponimod and IFNβ-1b groups, respectively. However, when performing ITCs, treatment effect was not adjusted for differences in treatment effect on placebo arms.



EXPAND vs European Study

Study design: although the anticipated study duration was 3 years, the European Study had an early termination at month 33. The reported duration of follow up was unsuitable for comparison, due to different reported measures: the mean duration of follow up for the IFNβ-1b group, was 35.6 months, and the median time on study was 21 months for the siponimod group. Therefore, the clinical relevance of the difference between groups is uncertain. The population included in EXPAND could be matched to the European Study.

Inclusion/exclusion criteria: Based on the 10% threshold, the MAH states that the age range of pa-tients included in the EXPAND and European studies, which are 18-60 years and 18-55 years, respec-tively, is significantly different. However, this difference is unlikely to be clinically relevant. Patients eli-gible for the EXPAND study were allowed to have received prior IFN therapy, while for patients included in the European study this was an exclusion criterion. For these characteristics, the population included in EXPAND could be matched to the European Study. Furthermore, patients eligible for the EXPAND study should not have any relapse in the 3 months prior to study start, while patients in the European study should not have any relapse within the month prior to study start. The MAH states that this is a relevant difference, which indicates the presence of heterogeneity. However, for the number of relapses prior to study start, using a MAIC had no advantage over a conventional NMA, as the population in-cluded in EXPAND study could not be matched to the population included in the European Study.

Of note, the MAH uses inconsistent criteria for the clinical relevance of differences in the duration of MS: while the MAH underlines the difference between studies in the duration of MS between the EX-PAND and North American studies, as there was no restriction in EXPAND, while in the North American Study patients had to have MS with at least 2-year duration, in the comparison between EXPAND and the European Study, the MAH considers as “non-applicable” the same difference between studies.

Outcome definitions: The EXPAND study defined progression of disease as a 1.0-point increase in EDSS if the baseline score was 3.0-5.0 and a 0.5-point increase in EDSS if the baseline score was 5.5-6.5. The European Study defined progression of disease as a 1.0-point increase in EDSS if the baseline score was 3.0-5.5 and a 0.5-point increase in EDSS if the baseline score was 6.0-6.5.

Baseline patient characteristics: as far as patient characteristics which were available in both studies, age (48.0 vs 41.0 years), proportion of patients with an EDSS score ≥6.0 (56% vs 45%), duration of SPMS (3.8 vs 2.2 years), and proportion of patients relapse-free in the prior 2 years (64% vs 30%), were higher in the EXPAND study than in the European Study. The differences between studies could be adjusted for, by the MAIC.

Treatment effect on the placebo arm: treatment effect on the outcomes ARR and annualised rate of treatment discontinuation were dissimilar (>10% different) in the two studies. ARR was 0.16 and 0.57, in siponimod and IFNβ-1b groups, respectively. Discontinuation was 0.084 and 0.132, in siponimod and IFNβ-1b groups, respectively. However, when performing ITCs, treatment effect was not adjusted for differences in treatment effect on placebo arms.




Siponimod was compared with interferon beta-1b 250 µg every other day by MAIC. The matching pro-cess for inclusion/exclusion criteria differences for the North American Study and European Study are summarised in Table A44 and Table A45, respectively.

Table A44. Matching inclusion/exclusion criteria – North American Study vs EXPAND


North American Study EXPAND

(after matching) Comment


Baseline EDSS Range 3.0-6.5 3.0-6.5 Removed patients with baseline EDSS score <3.0 and >6.5

Age Range 18-65 18-60 North American Study includes ages 18-65 which is broader than EXPAND (18-60)



- - -


Progression in the 2 years prior to screening; progression for at least 6

months

Progression in prior 24 months; progression for at

least 6 months

Nearly identical criteria, did not require matching

Duration of MS ≥2 years ≥2 years Removed patients with MS duration <2 years




Source: adapted from the technical reports provided by the MAH Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; MS=multiple sclerosis; SPMS=secondary progres-sive multiple sclerosis; T25FW=timed 25-foot walk test.



Table A45. Matching inclusion/exclusion criteria – European Study vs EXPAND


European Study EXPAND (after matching)

Comment


Baseline EDSS Range

3.0-6.5 3.0-6.5 Removed patients with baseline EDSS score <3.0 and >6.5




<1 month <3 months Cannot match because European study’s criteria on relapses within a time-frame is broader than EXPAND

European Study required no relapse-related neurological deterioration within one month prior to study (not reported in EXPAND IPD)


≥2 relapses OR

disability increase in prior 2 years

Progression in prior 24 months; progression for at

least 6 months

Criteria were considered similar, although European Study was slightly broader than EXPAND


Duration of SPMS ≥6 months ≥6 months a Identical criteria



Source: adapted from the technical reports provided by the MAH ”a” as per definition of SPMS Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; MS=multiple sclerosis; SPMS=secondary progres-

sive multiple sclerosis; T25FW=timed 25-foot walk test.

The adjustment process for CDP-6 and CDP-3 is presented in Table A46TableA6_17 and Table A47, respectively. Some baseline patient characteristics which were reported in EXPAND study were not reported in the North American Study and European Study and, therefore, no adjustment was possible for these variables.



Table A46. Adjustment factors for CDP-6 – North American Study

CDP-6


North American Study

1 Age ✓









✓

10 Sex ✓




Table A47. Adjustment factors for CDP-3 – European Study

CDP-3



European Study

1 Age ✓









✓

10 Sex ✓






Results

For CDP-6 (North American Study), sample size of the EXPAND population decreased from 1645 to 410 patients as a result of matching and adjustment. The results of population matching and adjustment for CDP-6 are summarised in Table A48.

For CDP-3 (European Study), sample size of the EXPAND population decreased from 1645 to 140 patients as a result of matching and adjustment. The results of population matching and adjustment for CDP-3 are summarised in Table A49 and Table A50.

For ARR, sample size of the EXPAND population decreased from 1645 to 606 patients as a result of matching. No adjustment was possible.

Table A48. Results of population matching and adjustment for CDP-6– North American Study

Variables North American

Study

EXPAND

(unmatched)


EXPAND (matched and

adjusted)

N (Neff) 939 1638 543 410

Age (mean years [SD]) 46.83 (8.14) 48.03 (7.84) 49.4 (7.74) 46.83 (8.15)

EDSS score at screening (mean [SD]) 5.13 (1.18) 5.42 (1.06) 5.27 (1.11) 5.13 (1.18)


14.66 (8.32) 12.62 (7.77) 12.92 (8.24) 14.66 (8.33)

Duration of SPMS (mean years [SD]) 4.03 (3.48) 3.77 (3.51) 3.84 (3.53) 4.03 (3.48)


0.83 (1.32) 0.67 (1.19) 0.65 (1.1) 0.83 (1.32)



Matched sample excludes patients >55 years old, EDSS <3 or >6.5, and those with prior IFNβ therapy



Table A49. Results of population matching and adjustment for CDP-3 – European Study

Variables European Study

EXPAND (unmatched)

EXPAND (matched and

unadjusted)

EXPAND (matched and

adjusted)

N (Neff) 718 1638 455 140

Age (mean years [SD]) 41 (7.2) 48.03 (7.84) 46.43 (6.81) 41 (7.22)


5.15 (1.1) 5.42 (1.06) 5.19 (1.11) 5.15 (1.1)


13.1 (7.06) 12.62 (7.77) 11.06 (7.91) 13.1 (7.08)


2.15 (2.3) 3.77 (3.51) 3.42 (3.19) 2.15 (2.31)


30.4% 64.04% 59.78% 30.4%



Matched sample excludes patients >55 years old, EDSS <3 or >6.5, and those with prior IFNβ therapy.





Table A50. Results of population matching for ARR – North American Study

Variables North American Study and European Study

EXPAND (unmatched) EXPAND

(matched and unadjusted)

N (Neff) 1343 1645 606

Source: adapted from the technical reports provided by the MAH The matched sample excludes patients with EDSS scores <3 or > 6.5, and those with prior IFNβ therapy.

Abbreviations: Neff=effective sample size.

The MAIC, suggested that siponimod, in comparison with interferon beta-1b 250 µg every other day, reduced the risk for CDP-6 (North American Study) by 45% (hazard ratio 0.55; 95%CI 0.33 to 0.91). A smaller effect was obtained using Bucher ITC and the difference did not reach statistical significance (hazard ratio 0.80; 95%CI 0.57 to 1.12). This data is depicted in Figure A18, but note that matched and adjusted results apply only to the population in the North American study.

Figure A18. MAIC scenario analysis results for time to CDP-6 – siponimod vs IFN-1b 250 µg every other day

Source: adapted from the technical reports provided by the MAH A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients with duration of MS <2 years, baseline EDSS score <3.0 and >6.5, and patients with prior interferon therapy. Scenario

A adjusts for all available ranked characteristics, subsequent scenarios drop the least important characteristics from adjust-ment. Abbreviations: CDP-6=confirmed disability progression at six months; CI=confidence interval; EDSS=Expanded Disability Sta-

tus Score; ITC=indirect treatment comparison; MAIC=matching-adjusted indirect comparison; Neff=effective sample size; µg=microgram.

The MAIC, suggested that siponimod, in comparison with interferon beta-1b 250 µg every other day, reduced the risk for CDP-3 (European Study) by 18% but the difference did not reach statistical signifi-cance (hazard ratio 0.82; 95%CI 0.42 to 1.63). Using Bucher ITC, siponimod, in comparison with inter-feron beta-1b 250 µg every other day, increased the risk for CDP-3 by 7% but the difference did not reach statistical meaning (hazard ratio 1.07; 95%CI 0.81 to 1.41). This data is depicted in Figure A19, but note that matched and adjusted results apply only to the population in the European study.



Figure A19. MAIC scenario analysis results for time to CDP-3 – Siponimod vs IFN-1b 250 µg every other day

Source: adapted from the technical reports provided by the MAH A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes

patients >55 years old, baseline EDSS score <3.0 and >6.5, and patients with prior interferon therapy. Scenario A adjusts for all available ranked characteristics, subsequent scenarios drop the least important characteristics from adjustment. Abbreviations: CDP-3=confirmed disability progression at three months; CI=confidence interval; EDSS=Expanded Disability

Status Score; ITC=indirect treatment comparison; MAIC=matching-adjusted indirect comparison; Neff=effective sample size.

The MAIC, including a matched EXPAND population, suggested that siponimod, in comparison with IFNβ-1b 250 µg every other day, reduced ARR by 10% (hazard ratio 0.90; 95%CI 0.51 to 1.59), but the difference did not reach statistical significance. Using Bucher ITC, gave similar results (hazard ratio 0.69; 95%CI 0.46 to 1.04). These results are depicted in Figure A20.

Figure A20. MAIC results for ARR – siponimod vs IFNβ-1b 250 µg every other day (North Amer-ican Study)

Source: adapted from the technical reports provided by the MAH A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients with EDSS scores <3 or > 6.5, and those with prior IFNβ therapy.

Abbreviations: ARR=annualised relapse rate; CI=confidence interval; IFNβ=interferon-beta; ITC=indirect treatment compari-son; MAIC=matching-adjusted indirect comparison; Neff=effective sample size.






The simulated treatment comparison (STC), indicated that siponimod, in comparison with IFN-1b 250 µg every other day, reduced CDP-3 by 37% (rate 0.63; 95%CI 0.39 to 1.01), but the difference did not reach statistical significance. Siponimod, in comparison with IFNβ-1b 250 µg every other day, reduced ARR by 11% (rate ratio 0.89; 95%CI 0.50 to 1.56), but the difference did not reach statistical signifi-cance. This results are presented in Table A51.

Table A51. Summary of treatment effects on outcomes by ITC method (siponimod vs IFN beta-1b 250 µg every other day)

ITC method


CDP-3 (European Study) 0.82

(0.42 to 1.63)

0.81

(0.53 to 1.23)

1.07

(0.81 to 1.41)

p-value (0.6460)

1.06

(0.81 to 1.41)

CDP-6 (North American Study) 0.55

(0.33 to 0.91)

0.63

(0.39 to 1.01)

0.80

(0.57 to 1.12)

p-value (0.1953)

0.80

(0.57 a 1.13)

ARR 0.90

(0.51 to 1.59)

0.89

(0.50 a 1.56)

0.69

(0.46 to 1.04)

p-value (0.0734)

0.69

(0.46 a 1.04)

Source: adapted from the technical reports provided by the MAH p-values available for Bucher ITC are are reported in the table. p-values not available for other comparisons

Abbreviations: MAIC=matched adjusted indirect comparison; STC=simulated treatment comparison; ITC=indirect treatment comparison; NMA=network meta-analysis; results are hazard ratio (95% confidence interval) for MAIC and Bucher ITC, rate ratio (95% confidence interval) for STC, and rate ratio (95% credible interval) for NMA

Comparison of siponimod vs natalizumab 300 mg q4w


The comparison of siponimod vs natalizumab q4w included two studies (EXPAND and ASCEND).

Study design: the study duration of EXPAND was 3 years, while the study duration of ASCEND was 2 years. For this characteristic matching was possible as the population in EXPAND was broader.

Inclusion/exclusion criteria: The MAH states that the age range of patients included in EXPAND and ASCEND, which are 18-60 years and 18-58 years, respectively, is significantly different, and matching was possible (MAIC). However, this difference is unlikely to be clinically relevant. In the EXPAND study, patients were allowed to receive prior IFN therapy, while in the ASCEND study patients were not al-lowed to receive IFN within 4 weeks prior to study start. The MAH considered this characteristic to be similar between the two studies. Indeed, the difference in IFN use prior to study start, between EXPAND and ASCEND, does not seem to be clinically relevant. There was no information on the duration of MS for the ASCEND study and, therefore, patients included in EXPAND study could not be matched for this characteristic. Thus, in this case, a MAIC had no advantage over a conventional NMA, as the popula-tions could not be matched for differences that might exist between studies in the duration of MS.

Outcome definitions: The EXPAND study defined disability progression as a 1.0-point increase in EDSS if the baseline score was 3.0-5.0 and a 0.5-point increase in EDSS if the baseline score was 5.5-6.5. The ASCEND study defined disability progression as a 1.0-point increase in EDSS if the baseline score was 3.0-5.5 and a 0.5-point increase in EDSS if the baseline score was 6.0-6.5. However, for the definition of disability progression, using a MAIC had no advantage over a conventional NMA, as the results could not be adjusted for the difference between studies in the definition of disability progression.

Baseline patient characteristics: patient characteristics which were available in both studies, propor-tion of patients with EDSS≥6 (56% vs 63%), duration of SPMS (3.8 vs 4.8 years), were higher in AS-CEND patients than in EXPAND patients, and the difference could be adjusted for, by the MAIC.



Treatment effect on the placebo arm: treatment effect on the outcome annualised rate of treatment discontinuation were dissimilar (>10% different) in the two studies. Discontinuation was 0.084 and 0.186, in the EXPAND and ASCEND studies, respectively. However, when performing ITCs, treatment effect was not adjusted for differences in treatment effect on placebo arms.


Siponimod was compared with natalizumab by MAIC. The matching process for inclusion/exclusion criteria differences is summarised in Table A52.

Table A52. Matching inclusion/exclusion criteria – ASCEND vs EXPAND


ASCEND EXPAND (after

matching)

Comment


Baseline EDSS Range

3.0-6.5 3.0-6.5 Removed patients with baseline EDSS score <3.0 and >6.5


Prior IFN Therapy

Excluded if within 4

weeks prior to randomization

Allowed

ASCEND excluded patients who received IFNβ therapy within 4 weeks prior to randomization, cannot match since time of IFNβ therapy is not recorded in EXPAND IPD

Note that ASCEND also excluded natalizumab-experienced patients; this data was not available in the EXPAND IPD


≥3 months ≥3 months

Removed patients with most recent relapse within 3 months

ASCEND required no relapse-related neurological deterioration within one month prior to study; not captured in EXPAND IPD. However, this was still considered similar.


Prior year Prior 2 years

ASCEND required disability progression not related to relapses in prior year; although EXPAND was ‘broader’ (2 years), time since disability progression was not captured in EXPAND IPD


Duration of SPMS ≥2 years ≥2 years Removed patients with SPMS onset <2 years

MS Severity Score

≥4 ≥4 Removed patients with MS severity score of <4

T25FW Test Score

≤30s ≤30s Remove patients with timed 25-foot walk >30 seconds


Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; IPD=individual patient data; MS=multiple sclerosis; SPMS=secondary progressive multiple sclerosis; T25FW=timed 25-foot walk test.



The adjustment process for CDP is presented in the Table A53, and for ARR is depicted in Table A54.

Table A53. Adjustment factors for CDP – ASCEND

CDP


ASCEND

1 Age ✓



4 Treatment experience (IFN or DMT history) ✓

5 Normalised brain volume ✓

6 Gadolinium-enhancing lesions on T1-weighted images ✓


8 Total volume of T2 lesions on T2-weighted images ✓


✓

10 Sex ✓




Table A54. Adjustment factors for ARR – ASCEND

ARR


ASCEND

1 Time since onset of most recent relapse ✓


3 Number of relapses per patient in two years prior to study X


5 Total volume of lesions on T2-weighted images ✓






Results

Sample size of the EXPAND population decreased from 1,645 to 516 patients as a result of matching and adjustment. The results of population matching and adjustment for CDP are summarised in Table A55. The MAH did not submit any evidence demonstrating that patient characteristics with differences between treatment groups were effect modifiers.


Variables ASCEND EXPAND (unmatched)

EXPAND (matched and

unadjusted)

EXPAND (matched

and adjusted)

N (Neff) 887 1584 608 516

Age (mean years [SD]) 47.25 (7.61) 48.07 (7.84) 47.77 (6.82) 47.25 (7.61)


5.6 (0.9) 5.41 (1.07) 5.75 (0.83) 5.6 (0.9)


12.14 (6.88) 12.69 (7.8) 13.28 (6.93) 12.14 (6.89)

Prior DMT (proportion) 77.0% 78.41% 83.55% 77.0%

Normalised brain volume (mean cm3

[SD]) 1423.37 (82.95) 1422.95 (86.76) 1429.17 (83.49) 1423.37 (83.02)

No Gd-enhancing T1 lesions (proportion)

76.2% 78.09% 78.12% 76.2%

Duration of SPMS (mean years [SD]) 4.8 (3.37) 3.77 (3.51) 5.2 (3.32) 4.8 (3.38)

Total volume of T2 lesions (mean mm3 [SD])

16793.21 (17003.8)

15231.14 (15942.01) 14961.27 (16181.56)

16793.2 (17018.97)

Relapse-free in prior 2 years (mean

[SD])

70.7% 63.83% 68.91% 70.7%


Source: adapted from the technical reports provided by the MAH Matched sample excludes patients >58 years old, SPMS onset within previous 2 years of enrolment, baseline EDSS <3 or

>6.5, MS severity score of <4, most recent relapses within 3 months, and patients with T25FW test of >30s during screening period. Abbreviations: ARR=annualised relapse rate; CDP=confirmed disability progression; CI=confidence interval; EDSS=Ex-panded Disability Status Scale; DMT=disease-modifying therapy; Gd=gadolinium; IFN=interferon; ITC=indirect treatment com-

parison; MS=multiple sclerosis; Neff=effective sample size; SD=standard deviation; SPMS=secondary progressive multiple sclerosis; T25FW=timed 25-foot walk test.

MAIC suggested that siponimod, in comparison with natalizumab, reduced the proportion of patients with CDP-6 by 24% (hazard ratio 0.76; 95%CI 0.44 to 1.30), but the difference did not reach statistical significance. Similar results were obtained using Bucher ITC (hazard ratio 0.74; 95%CI 0.48 to 1.14). This data is depicted in Figure A21, but note that matched and adjusted results apply only to the popu-lation in the ASCEND study.



Figure A21. MAIC scenario analysis results for proportion of patients with CDP-6 (96w) – siponi-mod vs natalizumab

Source: adapted from the technical reports provided by the MAH Odds ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes pa-tients >58 years old, baseline EDSS score <3.0 and >6.5, SPMS onset within 2 years of enrolment, MS severity score of <4,

patients with recent relapses within 3 months, and patients with T25FW test of >30 seconds during screening period. Scenario A adjusts for all available ranked characteristics, subsequent scenarios drop the least important characteristics from adjust-ment.

Abbreviations: CDP-6=confirmed disability progression at six months; CI=confidence interval; EDSS=Expanded Disability Sta-tus Score; ITC=indirect treatment comparison; MAIC=matching-adjusted indirect comparison; MS=multiple sclerosis; Neff=ef-fective sample size; SPMS=secondary progressive multiple sclerosis; T25FW=Timed 25-Foot Walk; w=week(s).

MAIC including EXPAND matched and adjusted, suggested that siponimod, in comparison with natali-zumab, increased the risk of ARR by 43% (hazard ratio 1.43; 95%CI 0.78 to 2.61), but the difference did not reach statistical significance. Using Bucher ITC suggested that siponimod, in comparison with natalizumab, decreased the risk of ARR by 1% (hazard ratio 0.99; 95%CI 0.64 to 1.54) but the difference did not reach statistical significance. These results are depicted in Figure A22, but note that matched and adjusted results apply only to the population in the ASCEND study.

Figure A22. MAIC scenario analysis results for ARR – siponimod vs natalizumab

Source: adapted from the technical reports provided by the MAH RR <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients >58 years old, baseline EDSS score <3.0 and >6.5, SPMS onset within 2 years of enrolment, MS severity score of <4, patients with

recent relapses within 3 months, and patients with T25FW test of >30 seconds during screening period. Scenario A adjusts for all available ranked characteristics, subsequent scenarios drop the least important characteristics from adjustment. Abbreviations: ARR=annualised relapse rate; CI=confidence interval; EDSS=Expanded Disability Status Score; ITC=indirect

treatment comparison; MAIC=matching-adjusted indirect comparison; MS=multiple sclerosis; Neff=effective sample size; SPMS=secondary progressive multiple sclerosis; T25FW=Timed 25-Foot Walk.






The simulated treatment comparison (STC), indicated that siponimod, in comparison with natalizumab, reduced the proportion of patients with CDP-6 by 31% (rate ratio 0.69; 95%CI 0.44 to 1.09), but the difference did not reach statistical significance. Siponimod, in comparison with natalizumab, increased the risk of ARR by 11% (rate ratio 1.11; 95%CI 0.69 to 1.78), but the difference did not reach statistical significance. This results are presented in Table A56.

Table A56. Summary of treatment effects on outcomes by ITC method (siponimod vs natali-zumab 300 mg q4w)

ITC method


CDP-3 ------ ------ ------ ------

Proportion of patients with CDP-6

0.76

(0.44 to 1.30)

0.69

(0.44 to 1.09)

0.74

(0.48 to 1.14)

p-value (0.9761)

------

ARR 1.43

(0.78 to 2.61)

1.11

(0.69 to 1.78)

0.99

(0.64 to 1.54)

p-value (0.9761)

0.99

(0.64 to 1.53)


p-values available for Bucher ITC are are reported in the table. p-values not available for other comparisons Abbreviations: MAIC=matched adjusted indirect comparison; STC=simulated treatment comparison; ITC=indirect treatment comparison; NMA=network meta-analysis; results are hazard ratio (95% confidence interval) for MAIC and Bucher ITC, rate

ratio (95% confidence interval) for STC, and rate ratio (95% credible interval) for NMA

Comparison of siponimod vs interferon β-1a 60 µg once a week


The comparison of siponimod vs interferon β-1a 60 µg once a week included two studies (EXPAND and IMPACT).

Study design: the study duration of EXPAND study was 3 years, while the study duration of IMPACT was 2 years. For this characteristic matching was possible as the population in EXPAND was broader.

Inclusion/exclusion criteria: In the EXPAND study, patients were allowed to receive prior IFN therapy, while in the ASCEND study patients were not allowed to receive IFN prior to study start. Furthermore, the EXPAND study included patients with an EDSS range between 3.0 and 6.5, while the IMPACT study included patients with an EDSS range between 3.5 and 6.5. For these characteristics matching was possible as the population in EXPAND was broader. There was no information on the duration of MS, or duration of SPMS for the IMPACT study and, therefore, patients included in EXPAND study could not be matched for these characteristics. Thus, for these characteristics, using a MAIC had no advantage over a conventional NMA, as the populations could not be matched for possible differences.

Outcome definitions: The EXPAND study defined disability progression as a 1.0-point increase in EDSS if the baseline score was 3.0-5.0 and a 0.5-point increase in EDSS if the baseline score was 5.5-6.5. The IMPACT study defined disability progression as a 1.0-point increase in EDSS if the baseline score was 3.0-5.5 and a 0.5-point increase in EDSS if the baseline score was 6.0-6.5. However, for the definition of disability progression, using a MAIC had no advantage over a conventional NMA, as the results could not be adjusted for the difference between studies in the definition of disability progression.



Baseline patient characteristics: proportion of patients with EDSS≥6 (56% vs 48%), mean timed 25-foot walk test (16.7 vs 14.5 seconds), time since most recent relapse (59 vs 44 months), proportion of patients relapse-free in prior year (78% vs 61%), duration of MS (12.6 vs 16.5 months), and number of relapses per patient in the prior year (0.2 vs 0.6), were higher in the IMPACT study than in EXPAND. These differences between studies could be adjusted for by the MAIC. However, duration of SPMS could not be adjusted for as details are not reported in IMPACT.

Treatment effect on the placebo arm: treatment effect on the outcomes ARR and annualised rate of treatment discontinuation were dissimilar (>10% different) in the two studies. ARR was 0.16 and 0.30, in EXPAND and IMPACT studies, respectively. Discontinuation was 0.084 and 0.142, in the EXPAND and IMPACT studies, respectively. However, when performing ITCs, treatment effect was not adjusted for differences in treatment effect on placebo arms.


Siponimod was compared with interferon β-1a 60 µg once a week by MAIC. The matching process for inclusion/exclusion criteria differences is summarised in Table A57.

Table A57. Pairwise comparison of inclusion/exclusion criteria – EXPAND vs IMPACT

Criteria EXPAND IMPACT Comparability

MS Population SPMS SPMS ✓

Baseline EDSS range 3.0-6.5 3.5-6.5 !

Age range 18-60 18-60 ✓

Prior IFN therapy Allowed No prior IFNβ use !

No relapses in X months prior 3 months NR n/a

Recently documented progression In the past 24 months In the past 1 year ! a

History of RRMS Required NR n/a

Duration of MS No restriction NR n/a

Duration of SPMS No restriction NR n/a

MS severity score No restriction NR n/a

T25FW test No restriction NR n/a


✓=Criterion is identical; !=Differences exist between trials and the EXPAND patient population is broader (i.e. matching may be

possible); X=Differences exist between the trials and the EXPAND patient population is not broader (i.e. matching is not possi-ble). a The included population of EXPAND is broader, however this criterion cannot be matched because the data is not available in

the IPD. Abbreviations: EDSS=Expanded Disability Status Scale; IFN=interferon; IFNβ=interferon beta; IPD=individual patient data; MS=multiple sclerosis; n/a=not applicable; NR=not reported; RRMS=relapsing-remitting multiple sclerosis; SPMS=secondary

progressive multiple sclerosis; T25FW=Timed 25-foot Walk Test.

The adjustment process for CDP is presented in Table A58, and for ARR is depicted in Table A59.



Table A58. Adjustment factors for CDP – IMPACT

CDP


IMPACT

1 Age ✓






7 Duration of SPMS X



✓

10 Sex ✓




Table A59. Adjustment Factors for ARR – IMPACT

ARR



IMPACT

1 Time since onset of most recent relapse ✓

2 Number of relapses per patient in one year prior to study ✓

3 Number of relapses per patient in two years prior to study X








Results

Sample size of the EXPAND population decreased from 1,645 to 113 patients as a result of matching and adjustment. This makes results unreliable. The results of population matching and adjustment for CDP are summarised in Table A60. The MAH did not submit any evidence demonstrating that patient characteristics with differences between treatment groups were effect modifiers.


Variables IMPACT EXPAND (unmatched)

EXPAND (matched and

unadjusted)

EXPAND (matched and

adjusted)

N (Neff) 436 1590 563 113

Age (mean years [SD]) 47.55 (7.95) 48.05 (7.87) 49.31 (7.81) 47.55 (7.97)


5.2 (1.1) 5.41 (1.07) 5.33 (1.03) 5.2 (1.1)


16.45 (9) 12.68 (7.79) 11.76 (8.57) 16.45 (9.02)

1 Gd-enhancing T1 lesion (proportion)

16.5% 10.88% 11.37% 16.5%

2 Gd-enhancing T1 lesions (proportion)

5.8% 3.4% 2.84% 5.8%

3 Gd-enhancing T1 lesions (proportion)

3.6% 2.2% 1.78% 3.6%

≥4 Gd-enhancing T1 lesions (proportion)

10.3% 5.47% 5.68% 10.3%

Number of relapses in prior 1 year (mean [SD])

0.55 (1) 0.26 (0.55) 0.26 (0.51) 0.55 (1.01)

Sex (proportion female) 64% 60.25% 61.81% 64%

Source: adapted from the technical reports provided by the MAH Matched sample excludes patients EDSS <3.5 or >6.5, and those with prior IFNβ therapy.

Abbreviations: EDSS=Expanded Disability Status Scale; IFNβ=interferon-beta; Gd=gadolinium; MS=multiple sclerosis; Neff=effective sample size; SD=standard deviation; SPMS=secondary progressive multiple sclerosis.

The MAIC, suggested that siponimod, in comparison with interferon β-1a 60 µg once a week, reduced the proportion of patients with CDP-3 by 58% (hazard ratio 0.42; 95%CI 0.20 to 0.88). A smaller effect was obtained using Bucher ITC (hazard ratio 0.81; 95%CI 0.53 to 1.23), and the difference did not reach statistical significance This data is depicted in Table A44, but note that matched and adjusted results apply only to the population in the IMPACT study.



Figure A23. MAIC scenario analysis results for time to CDP-3 – siponimod vs interferon β-1a 60 µg once a week

Source: adapted from the technical reports provided by the MAH A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients with a baseline EDSS score <3.5 and >6.5 and those with prior IFNβ therapy. Scenario A adjusts for all available ranked characteristics, subsequent scenarios drop the least important characteristics from adjustment.

Abbreviations: CDP-3=confirmed disability progression at three months; CI=confidence interval; EDSS=Expanded Disability Status Score; ITC=indirect treatment comparison; MAIC=matching-adjusted indirect comparison; Neff=effective sample size.

MAIC, including EXPAND matched and adjusted, suggested that siponimod, in comparison with inter-feron β-1a 60 µg once a week, decreased ARR by 0.3% (hazard ratio 0.997; 95%CI 0.46 to 2.18), but the difference did not reach statistical significance. Using Bucher ITC suggested that siponimod, in comparison with interferon β-1a 60 µg once a week, decreased ARR by 32% (hazard ratio 0.68; 95%CI 0.45 to 1.01). These results are depicted in Table A45, but note that matched and adjusted results apply only to the population in the IMPACT study.

Figure A24. MAIC scenario analysis results for ARR – siponimod vs interferon β-1a 60 µg once a week

Source: adapted from the technical reports provided by the MAH A hazard ratio <1 indicates siponimod has a favourable outcome relative to the comparator. The matched sample excludes patients with a baseline EDSS score <3.5 and >6.5 and those with prior IFNβ therapy. Scenario A adjusts for all available

ranked characteristics, subsequent scenarios drop the least important characteristics from adjustment. Abbreviations: ARR=annualised relapse rate; CI=confidence interval; EDSS=Expanded Disability Status Score; ITC=indirect treatment comparison; MAIC=matching-adjusted indirect comparison; Neff=effective sample size; µg=microgram.






The simulated treatment comparison (STC), suggested that siponimod, in comparison with interferon β-1a 60 µg once a week, reduced CDP-3 by 36% (rate ratio 0.64; 95%CrI 0.38 to 1.08) but this result was not statistically significant.

The simulated treatment comparison (STC), suggested that siponimod, in comparison with interferon β-1a 60 µg once a week, reduced ARR by 7% (rate ratio 0.93; 95%CrI 0.52 to 1.66) but the difference did not reach statistical meaning. This results are presented in Table A61.

Table A61. Summary of treatment effects on outcomes by ITC method (siponimod vs interferon β-1a 60 µg once a week)

ITC method


CDP-3 0.42

(0.20 to 0.88)

0.64

(0.38 to 1.08)

0.81

(0.53 to 1.23)

p-value (0.3107)

0.81

(0.54 to 1.22)

CDP-6 ------ ------ ------ ------

ARR 0.997

(0.46 to 2.18)

0.93

(0.52 to 1.66)

0.68

(0.45 to 1.01)

p-value (0.0586)

0.67

(0.45 to 1.01)

Source: adapted from the technical reports provided by the MAH p-values available for Bucher ITC are are reported in the table. p-values not available for other comparisons Abbreviations: MAIC=matched adjusted indirect comparison; STC=simulated treatment comparison; ITC=indirect treatment

comparison; NMA=network meta-analysis; results are hazard ratio (95% confidence interval) for MAIC and Bucher ITC, rate ratio (95% confidence interval) for STC, and rate ratio (95% credible interval) for NMA

EUnetHTA Joint Action 3 WP4...V0.3 11/02/2020 Input from medical editor and manufacturer(s) has been...

Documents

Transcript of EUnetHTA Joint Action 3 WP4...V0.3 11/02/2020 Input from medical editor and manufacturer(s) has been...