Lehnert et al. Systematic Reviews (2022) 11:134 https://doi.org/10.1186/s13643-022-01997-2 PROTOCOL Open Access Immunotherapy for people with clinically isolated syndrome or relapsing‑remitting multiple sclerosis: treatment response by demographic, clinical, and biomarker subgroups (PROMISE)—a systematic review protocol Thomas Lehnert1, Christian Röver2 , Sascha Köpke3 , Jordi Rio4 , Declan Chard5,6 , Andrea V. Fittipaldo7 , Tim Friede2 , Christoph Heesen1 and Anne C. Rahn1,8*    Abstract Background: Multiple sclerosis (MS) is an inflammatory and degenerative disease of the central nervous system with an increasing worldwide prevalence. Since 1993, more than 15 disease-modifying immunotherapies (DMTs) have been licenced and have shown moderate efficacy in clinical trials. Based on the heterogeneity of the disease and the partial effectiveness of therapies, a personalised medicine approach would be valuable taking individual prognosis and suitability of a chosen therapy into account to gain the best possible treatment effect. The primary objective of this review is to assess the differential treatment effects of all approved DMTs in subgroups of adults with clinically isolated syndrome or relapsing forms of MS. We will analyse possible treatment effect modifiers (TEM) defined by baseline demographic characteristics (gender, age), and diagnostic (i.e. MRI measures) and clinical (i.e. relapses, disability level) measures of MS disease activity. Methods: We will include all published and accessible unpublished primary and secondary analyses of randomised controlled trials (RCTs) with a follow-up of at least 12 months investigating the efficacy of at least one approved DMT, with placebo or other approved DMTs as control intervention(s) in subgroups of trial participants. As the primary outcome, we will address disability as defined by the Expanded Disability Status Scale or multiple sclerosis functional composite scores followed by relapse frequency, quality of life measures, and side effects. MRI data will be analysed as secondary outcomes. MEDLINE, EMBASE, CINAHL, LILACS, CENTRAL and major trial registers will be searched for suitable studies. Titles and abstracts and full texts will be screened by two persons independently using Covidence. The risk of bias will be ana- lysed based on the Cochrane “Risk of Bias 2” tool, and the certainty of evidence will be assessed using GRADE. *Correspondence: anne.rahn@uksh.de 8 Institute for Social Medicine and Epidemiology, Nursing Research Unit, University of Lübeck, Lübeck, Germany Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Lehnert et al. Systematic Reviews (2022) 11:134 Page 2 of 13 Treatment effects will be reported as rate ratio or odds ratio. Primary analyses will follow the intention-to-treat princi- ple. Meta-analyses will be carried out using random-effects models. Discussion: Given that individual patient data from clinical studies are often not available, the review will allow to analyse the evidence on TEM in MS immunotherapy and thus support clinical decision making in individual cases. Systematic review registration: PROSPERO CRD42​02127​9665. Keywords: Multiple sclerosis, Systematic review, Meta-analysis, Immunotherapy, Treatment response, Subgroups Background [23–25]. However, their long-term benefit is still uncer- Multiple sclerosis (MS) is a chronic disease of the cen- tain [26–28]. Evidence from RCTs and observational tral nervous system, characterised by the destruction of studies suggests that starting a DMT soon after diagnosis myelin and axons, associated with focal inflammation potentially leads to better (long-term) outcomes in per- and diffuse neurodegeneration in the brain as well as spi- sons with RRMS, such as a reduced risk of relapses and nal cord grey and white matter [1]. Disease onset is typi- transition to SPMS, when compared to later initiation cally between the ages of 20–40 [2]. With a global median [29–32]. This favourable effect seems to be more pro- prevalence of 33 per 100,000 people and approximately nounced for DMTs with higher efficacy as alemtuzumab, 2.5 million people affected worldwide, MS is the most fingolimod, ocrelizumab, ozanimod, and natalizumab common neuroimmunological disorder and most fre- [33, 34]. DMTs with higher efficacy also exhibit higher quent cause of non-traumatic disability in young adults toxicity [24], which in rare cases may lead to severe and [3, 4]. While the pathogenesis of MS is not completely potentially life-threatening adverse effects, e.g. progres- understood, current evidence suggests that a multifacto- sive multifocal leukoencephalopathy (PML) [35, 36]. rial interplay of genetic, environmental and lifestyle risk The side-effect profile of a DMT affects the adherence factors results in a dysregulated immune response [5, 6]. to the DMT [37]. While treatment adherence is essential The prognosis and course of the disease are highly vari- to the effectiveness of DMTs [38], non-adherence was able and currently unpredictable at the individual patient reported in up to 70% of people using DMTs [39, 40]. level, particularly at disease onset [7, 8]. MS negatively Consequently, a substantial group of DMT users may not impacts many health-related outcomes, such as activi- experience the full benefits, thus having an increased risk ties of daily living [9], quality of life [10, 11] and mortality of new or enlarging brain lesions, relapses and worsening [12], and places a large financial burden on patients and disability [41, 42]. their families, healthcare systems and societies at large To improve care including immunotherapy manage- [13, 14]. ment, the concept of personalised (precision) medicine In most people, MS is relapsing at the beginning is highly relevant in MS. However, the implementation (relapsing-remitting MS (RRMS)), with exacerbations of such an approach in MS is currently limited by knowl- followed by partial or full remission of neurological edge gaps. The pathogenesis of MS is not fully under- symptoms and phases of relative stability [2, 15]. As evi- stood. Robust and validated biomarkers and validated dence has shown that neurodegeneration is present from treatment algorithms are lacking [43]. Hence, the cur- the earliest stages of MS [16], these clinical subtypes rent approach to personalised immunotherapy in MS should be considered as different points on a continuum, depends on an evidence-based estimate of the prognosis, with focal white matter inflammation being dominant a subsequent choice of therapy based on a shared deci- during the relapsing-remitting phase and neurodegenera- sion-making (SDM) approach and the evaluation of early tion during the progressive phase [2, 17, 18]. Thus, neu- immunotherapy responses [44]. rodegeneration is considered the main driver of disability Against the background of the heterogeneous clinical progression [19, 20]. presentation, progressive nature of MS and DMTs that Over the last three decades, more than 15 immuno- are only partially effective in controlling neurodegen- therapies (disease-modifying therapies (DMTs)) with dif- eration [19], it is crucial to optimise the use of available ferent mechanisms of action and profiles of benefits and DMTs during the early inflammatory phase of the disease risks have been approved for the treatment of relapsing to prevent the accumulation of irreversible neurological MS [21, 22]. Primarily targeting the focal inflammatory damage and concomitant accrual of disability [45]. The processes of the disease, these treatments reduce brain optimal treatment window may close early in the dis- white matter lesion formation, as seen using Magnetic ease course, i.e. before the walking distance is impaired resonance imaging (MRI), reduce relapse rates and possi- [34, 46]. Therefore, the reliable identification of patients bly slow disability accrual and overall disease progression with a poor response is crucial [25, 47]. One method to Lehnert et al. Systematic Reviews (2022) 11:134 Page 3 of 13 identify poor responders lies in the assessment of treat- progressive MS [59]. Further, ageing of the immune sys- ment effect modifiers (TEMs) in RCTs [47, 48], which tem leads to a decline in the regulation of innate and as are baseline variables associated with larger or smaller well as adaptive immune responses [58, 61], possibly also treatment effects [47, 49]. TEMs are defined in relation altering responses to DMTs. A recent meta-analysis of to a specific treatment and against a control group and 38 clinical trials that assessed the efficacy of DMTs on are typically assessed via subgroup analysis in RCTs [49]. disability progression found age to be an essential modi- The clinical rationale for the assessment of TEM is to fier of DMT efficacy [58]. The authors concluded that ascertain whether the effect of a DMT differs in groups higher efficacy treatments exert their benefit over lower of patients with different baseline characteristics [50]. efficacy treatments only during the early stages of MS, Thus, TEM can help to identify groups of patients who and beyond the age of 53 years, the model suggests that have lower than average benefits from a DMT (poor for the average person with MS, DMTs have no added responders) or, at worst, no benefit at all (non-respond- value. Similarly, a meta-analysis of six RCTs by Signori ers), but also patients who benefit above-average from a and colleagues [51] found that treatment effects on the treatment. annualised relapse rate (ARR) were significantly higher To our knowledge, there is no systematic review con- in participants who were <40 years of age compared to sidering all presently available DMT options for all peo- those ≥40 years. ple with clinically isolated syndrome (CIS) or RRMS [51–54]. Disability level We will consider all DMTs currently approved by the Disability, as measured by the Expanded Disability Sta- European Medicines Agency (EMA) or the U.S. Food tus Scale (EDSS), correlates with disease duration, albeit and Drugs (FDA) for use in people with CIS or RRMS with substantial individual variation in longer-term as alemtuzumab, azathioprine, beta interferons (inter- clinical outcomes. Beyond an EDSS value of 4.0, MS is feron beta-1a, interferon beta-1b), cladribine, dimethyl, increasingly likely to be running a progressive rather fingolimod, glatiramer acetate, mitoxantrone, natali- than relapsing-remitting course [46]. It is well-recognised zumab, ofatumumab, ozanimod, ocrelizumab, ponesi- that many MS treatments have no significant effects on mod, siponimod and teriflunomide. A key goal of current disease progression (at least as assessed using EDSS) research is precision medicine, which is tailoring treat- without relapses [62]. Since it may take several years for ment to individuals rather than the blanket treatment of a progressive MS course to develop and manifest itself, groups. For people with MS, we now have a large range EDSS may be a more timely, if imperfect, proxy for this. of treatments; some with similar degrees of efficacy when However, to date, this has not thoroughly been looked assessed at a group level. At an individual level, treatment at. The multiple sclerosis functional composite (MSFC) effects are not homogeneous across trial participants as represents an integrated disability measure possibly more suggested by findings from subgroup analyses of individ- sensitive to change than the EDSS [63, 64]. While the ual RCTs [55, 56] but vary according to demographic and mobility-dependent EDSS shows floor effects at higher disease-related characteristics [51, 53, 57, 58]. Evidence disability levels [65], the MSFC and especially hand func- indicates that higher treatment benefit is associated tion [66] and information processing measures may show with baseline characteristics of earlier (i.e. younger age, more room for improvement. On the other hand, sub- lower levels of disability) and more inflammatory disease stantially impaired information processing may indicate activity as shown on MRI [51, 57, 58]. In the following, irreversible neurodegeneration [67] and thereby lowering we outline major possible TEMs, i.e. age, disability level, chances for a treatment benefit. Therefore, the impact of sex, relapse rate, MRI activity and previous DMT use (all these measures on treatment responses is meaningful. assessed at baseline). Sex Age MS is increasingly prevalent among women, but clini- In MS, the risk of a progressive disease course increases cally men are more likely to develop a progressive course with age [59]. Neuropathological evidence indicates that [15] and do so faster (both progression to SPMS from earlier inflammation, while still present, is more com- disease onset and age at SPMS diagnosis) [68]. Sex dif- partmentalised and diffuse later compared to early dis- ferences in immunological, neurodegenerative, and ease stages [60]. In addition, neurodegeneration is more neuroprotective mechanisms, which are increasingly apparent and may increasingly become the dominat- found in healthy individuals may also be seen in MS [69, ing factor determining clinical outcomes. These pro- 70]. While sex hormones and their dynamic interplay cesses most likely evolve rather than simply progress seem to be major drivers, other x-chromosomal factors on a continuum from relapsing-remitting to secondary might be also relevant. Women seem to have stronger Lehnert et al. Systematic Reviews (2022) 11:134 Page 4 of 13 inflammatory responses, but also a greater capacity for Previous DMT use repair in the CNS [71]. For these reasons, it seems likely Breakthrough disease on a low to moderate immunother- that responses to DMTs differ between females and apy may indicate another level of activity of the immune males. However, so far, no clear patterns for the effects of processes. In clinical practice, DMTs are often applied as sex on immunotherapy efficacy have been reported [57]. escalation therapies in case of previous treatment fail- ure. Very few studies have focussed on non-responders, Relapse rate though, such as CARE-MS2 investigating alemtuzumab Relapse rates are typically the main factor determining [83]. The pressing clinical question, if pretreatment or eligibility for treatments in clinical practice and entry previous treatment failure reduces the chances of a treat- to clinical trials and are the primary outcome of most ment benefit, remains unanswered. While this has not RRMS treatment studies. One to two relapses in the pre- been thoroughly addressed, registry data indicate that the ceding years are an inclusion criterion in almost every probability of converting to SPMS is significantly lower RRMS treatment trial, as there needs to be sufficient for patients escalated from glatiramer acetate or inter- disease activity in the comparator arms for a treatment feron beta to fingolimod, alemtuzumab or natalizumab effect to be demonstrable. Relapse activity correlates with within 5 years of disease onset compared with matched MRI inflammatory disease activity and relapse-associated patients who were escalated later [29]. Therefore, the disability accrual. The presence of superimposed relapses sequence of treatments might also be predictive for treat- and/or MRI inflammatory activity also appears to predict ment response, both in terms of relapses and disability treatment response in a subset of people with progres- progression. sive MS [72], albeit when using treatments that typically In summary, the identification of baseline factors have also been shown to be effective in RRMS, rather consistently associated with larger treatment effects than agents that may target non-inflammatory processes. (response) can help to optimise MS immunotreatment Overall, relapse rates have only a minor impact on disa- by channelling people with MS towards the most effec- bility evolution [73]. While in the earliest treatment trials tive DMT, thus not exposing them to ineffective and in MS, participants had substantially higher relapse rates potentially harmful treatments. This may in turn reduce prior to study inclusion [74] than in more recent studies non-adherence and premature treatment termination. [75], absolute relapse rate reduction and percentages of Identifying and quantifying the effect baseline TEMs relapse-free patients do not differ too much (15 versus should result in a more nuanced understanding of the 11 out of 100). Relapse rate before treatment initiation effect of a DMT and will help to set the overall effects of might have only a minor influence on subsequent relapse DMTs into context [84, 85]. Identification of factors pre- activity under treatment but may represent a TEM for dicting higher or lower than average response to specific the progression of disability. DMTs at treatment initiation can inform clinical treat- ment guidelines and ultimately contribute towards per- sonalised treatment in people with MS [86]. MRI activity MRI is a highly established diagnostic tool in MS [76]. Methods However, its prognostic value is a matter of ongoing The objective of this review is to assess differential treat- debate. A high number of T2 lesions at presentation as ment effects (response) of approved immunotherapies in well as an early increase of T2 lesion number has been subgroups of adults with CIS or RRMS, defined by base- regarded as an indicator of a greater risk of disability line age, sex, treatment history, relapse rate, disability progression [77]. Recent data underline the importance level, and MRI activity. of lesion location, i.e. early infratentorial and spinal cord lesions as possibly more predictive for later progression Criteria for considering studies for this review [7, 78]. As all DMTs impact on lesion evolution, a treat- Types of studies ment effect on MRI-visible lesion activity is (at a group We will include all published and unpublished pri- level) a good predictor of clinical treatment efficacy [8]. mary and secondary analyses of RCTs including ran- Slowed lesion accumulation moreover correlates with domised crossover trials with a follow-up of at least 12 disability progression in the short term [79], albeit this months, in which the efficacy of at least one approved association is not without criticism [80]. Although the DMT, with placebo or other approved DMTs as control long-term predictive value of MRI measures is not clear, intervention(s), was investigated in subgroups of trial some studies have hypothesised that the accumulation participants. Subgroup characteristics need to be meas- of new active lesions in the first stages of therapy may ured at baseline, i.e. prior to randomisation before partic- impact long-term disability [81, 82]. ipants were exposed to the study drug. We will consider Lehnert et al. Systematic Reviews (2022) 11:134 Page 5 of 13 subgroups defined by age, disability level, sex, relapse published literature [94–96]. This list was adapted and history, MRI activity and treatment history, which are reduced in two rounds of structured group discussions described and operationalised under Data Synthesis with two patient representatives from the German Mul- below. We will include studies regardless of publication tiple Sclerosis Society [97]. In a final step, the representa- status and language. tives were asked to classify outcomes as critical outcomes Types of participants and important outcomes. We will include participants aged ≥18 years with CIS or RRMS, according to established diagnostic criteria, as applied in the original studies, and treated with FDA or Primary outcomes EMA approved DMTs. CIS is defined as a first monopha- Critical outcomes sic clinical episode typical of CNS demyelination sugges- tive of MS, while para-clinical findings do not justify a diagnosis of MS [87, 88]. Definition of RRMS has to be • Disability worsening: number or proportion of par- based on a diagnosis according to established diagnostic ticipants experiencing, or free from disability-wors- criteria, i.e. Schumacher [89], Poser [90], original 2001 ening at 12, 24, 36 or 48 months after randomisa- McDonald [91] or modified McDonald [92, 93] criteria. tion, yearly until the end of follow-up (if longer than We will exclude studies on people with progressive forms 48 months) or at the end of follow-up. Disability of MS [87]. In the case of trials quoting results for mixed worsening must have been assessed using either the populations of participants with relapsing as well as pro- Expanded Disability Status Scale (EDSS) [98] or the gressive forms of MS, they will be included if more than multiple sclerosis functional composite (MSFC) [63]. 75% of participants in the study had relapsing MS. The EDSS is a method used to quantify disability in MS based on a neurological examination [98]. The Types of interventions MSFC assesses disability with three different tests: We will include studies that evaluated one or more of the walking ability (25 foot walk test), hand functioning following DMT regimens as monotherapy: (9 hole peg test) and information processing (sym- bol digit modalities test) [63]. The MSFC can pro- Alemtuzumab (Lemtrada®) duce scores for each of the three individual measures Azathioprine (e.g. Imurek®) • as well as a composite score, which is a continuous Cladribine (Mavenclad®) • variable that can be used as any numerical variable in Dimethyl fumarate (Tecfidera®) • analyses. MSFC disability worsening is defined as an Fingolimod (Gilenya®) • increase in scores in at least one MSFC component Glatiramer acetate (Copaxone®, Clift®) • by 20% from baseline, sustained for at least 3 months Interferon beta-1a (Rebif®, Avonex®), pegylated • [99]. interferon beta-1a (Plegridy®) • • Relapses: number of relapses, ARR or proportion Interferon beta-1b (Betaferon®, Betaseron®, Exta- of participants with new relapses at 12, 24, 36 or 48 via®) • months after randomisation, yearly until the end of Mitoxantrone (Novantron®, Ralenova®) follow-up (if longer than 48 months) or at the end of Natalizumab (Tysabri®) • follow-up. While a relapse is typically defined as one Ofatumumab (Kesimpta®) • or more newly developing or reoccurring symptoms Ozanimod (Zeposia®) • of neurological dysfunction persisting for more than Ocrelizumab (Ocrevus®) • 24 h, preceded by a period of stability for at least 1 Ponesimode (Ponvory®) • month [91], other relapse definitions will be accepted Siponimod (Mayzent®) • as well. Teriflunomide (Aubagio®) • • • Quality of life (QoL): questionnaires assessing generic or MS-specific QoL (e.g. Leeds Multiple Comparators will be placebo or other approved DMTs. Sclerosis Quality of Life (LMSQoL), Functional Assessment of Multiple Sclerosis (FAMS), Hamburg Types of outcome measures Quality of Life Questionnaire in Multiple Sclerosis The selection and classification of outcomes were con- (HAQUAMS), Multiple Sclerosis International Qual- ducted in a structured process among the review authors ity of Life (MUSIQoL) questionnaire, Multiple Scle- and two patient representatives. Initially, a list of treat- rosis Quality of Life-54 (MSQoL-54), Health Status ment outcome domains used in studies on people with Questionnaire 36 (SF-36) and EuroQol-5 Dimension MS was compiled based on expert knowledge and (EQ-5D). Lehnert et al. Systematic Reviews (2022) 11:134 Page 6 of 13 • Adverse events (AE): total number of serious adverse • Latin American and Caribbean Health Science Infor- events or number of participants with at least one mation Database (LILACS) (Bireme) (1982 to search serious adverse event at the end of follow-up. date) • Clini​calTr​ials.​gov (http://​clini​caltr​ials.​gov) • World Health Organization (WHO) International Secondary outcomes Clinical Trials Registry Platform (https://​trial​search.​ who.​int/) Important outcomes The search keywords are listed in Additional file 1. Searching other resources • MRI activity: The number or proportion of par- In addition to the electronic searches, we will screen the ticipants with new/enlarging T2 lesions or new T1 reference lists of published reviews and of included full- contrast-enhancing lesions at 12, 24, 36 or 48 months text articles for additional references to identify relevant after randomisation, yearly until the end of follow-up trials. (if longer than 48 months) or at the end of follow-up. • No Evidence of Disease Activity (NEDA): number or Data collection and analysis proportion of participants achieving NEDA-3 at 12, 24, 36 or 48 months after randomisation, yearly until Selection of studies the end of follow-up (if longer than 48 months) or Data management and extraction will be conducted fol- at the end of follow-up months. NEDA-3 is defined lowing the methods described in the Cochrane Hand- as an absence of MRI activity (i.e. new or enlarging book for Systematic Reviews of Interventions [101]. To lesions), relapses and disability worsening (as meas- select search titles and abstracts considered relevant for ured by EDSS) [100]. the review, two review authors will independently assess • Activities of daily living (ADLs): questionnaires the results of the various searches using Covidence. Disa- assessing basic or instrumental ADLs or both (e.g. greements will be resolved by discussion with a third Functional Independence Measure (FIM), Katz ADL team member. Should more information be required to Index, Barthel Index). determine eligibility, we will contact study authors for • Fatigue: questionnaires assessing fatigue (e.g. Fatigue additional information and check related publications. Severity Scale (FFS), Modified Fatigue Impact Scale Multiple reports of the same study will be collated. All (MFIS), Multidimensional Fatigue Index (MFI), and potentially relevant reports will be retrieved in full-text visual analogue scale (VAS) for fatigue). and investigated further for inclusion independently by • Conversion to clinically definite MS: time to conver- two authors. We will resolve disagreements through dis- sion to clinically definite MS. cussion and/or consensus with another author. To assure • Conversion to SPMS: time to conversion to SPMS a reproducible process, we will screen full texts following [87]. the standardised inclusion and exclusion criteria. Rea- sons for study exclusion will be recorded and reported in detail in the “Characteristics of excluded studies” table Search methods for identification of studies and the process of study selection will be documented in We will not apply any language restrictions to the a flow chart according to PRISMA guidelines [102]. searches. Data extraction and management Electronic searches Data extraction from included studies will be performed We will search the following sources for eligible reports by two authors independently using a pre-defined data in any language: extraction form within an Excel spreadsheet. In case of disagreement not resolved by discussion, we will consult • Cochrane Central Register of Controlled Trials another review author for clarification; if not possible, (CENTRAL 2020) in the Cochrane Library the record will be assigned “awaiting assessment” and • MEDLINE (PubMed) (1946 to search date) study authors will be contacted for clarification. The data • Embase (Embase.​com) (1974 to search date) extraction sheet will be designed to extract study-level • Cumulative Index to Nursing and Allied Health Lit- and subgroup-level data, i.e. for the subgroups which this erature (CINAHL) (EBSCO host) (1981 to search review aims to summarise: age, sex, treatment history, date) relapse rate, disability level and MRI activity (see “Data synthesis” section). The data extraction sheet will be Lehnert et al. Systematic Reviews (2022) 11:134 Page 7 of 13 piloted and assessed for usability within the review team. the intended interventions, (3) bias due to missing out- Study names and authors will not be masked. The follow- come data, (4) bias in the measurement of the outcome ing information will be extracted from individual studies: and (5) bias in the selection of the reported result. Within each domain, a series of “signalling questions”, having the • Study details: name/acronym, registration number(s), response options: “yes”, “probably yes”, “probably no”, “no” date, country, reports, and authors and “no information”, will be used to elicit information • Eligibility criteria (inclusion/exclusion) about trial features relevant to the risk of bias of a par- • Study design and clinical setting (e.g. mono or multi- ticular outcome. Based on the responses to the signalling centric) questions, risk of bias judgement (i.e. “low risk of bias”, • Number and characteristics of participants in the “some concerns” or “high risk of bias”) will be made for overall trial and each treatment arm (e.g. age, gender, each of the five domains. Finally, using the domain-level years since diagnosis, MS type, level of disability) and judgements, we will reach one of the following three number of participants in each treatment arm of the overall risk of bias judgements for a specific outcome of subgroups to be summarised in this review a trial: • Intervention details: description of the intervention(s), dosage, regimen, frequency, duration 1. “Low risk of bias”, if the trial was judged as low risk of • Subgroup details (for all subgroups reported in a bias in all five domains considered for the outcome. study): rationale for inclusion, definition, operation- 2. “Some concerns”, if the trial was judged to raise some alisation (e.g. categories, thresholds), statistical analy- concerns in at least one domain for the outcome, but ses not at high risk of bias for any domain. • Length and description of participant follow-up 3. “High risk of bias”, if the trial was judged to be at high • Data analysis: statistical models used to conduct sub- risk of bias in at least one domain for the outcome, or group analysis, adjustment: co-variables (potentially) if it was judged to raise some concerns for multiple considered in the derivation of subgroup estimates domains. • Outcomes: outcomes measured and outcomes reported in the main analyses and subgroup analyses Measures of treatment effect • Declaration of interest and funding source Treatment effects on rates (e.g. relapses, AE) will be quantified using rate ratios (RRs) when event rates are Extracted data will be compared and checked for accu- reported or computable, or using odds ratios (OR) when racy; disagreements will be resolved through discussion only proportions of affected patients are reported. Treat- and consensus. ment effects on disability worsening will be quantified in terms of OR, based on the proportions of patients Assessment of risk of bias in included studies experiencing a sustained or confirmed disability worsen- To assess the risk of bias of included studies, we will fol- ing. Treatment effects on binary endpoints (e.g. disabil- low the guidance in the Cochrane Handbook for System- ity worsening, NEDA) will be quantified in terms of OR, atic Reviews of Interventions [101]. Two review authors based on the proportions of affected patients. Treatment will independently assess the included studies consider- effects expressed in terms of rating scales (e.g. quality of ing also the main publications of the respective phase two life, fatigue) will be analysed based on standardised mean and three studies using the Cochrane “Risk of Bias 2” tool differences (SMD). Treatment effects expressed in terms (RoB2) for randomised trials [103], which will be imple- of continuous endpoints (e.g. time to conversion) will be mented using the Rob2 Excel tool (https://​sites.​google.​ analysed based on mean differences. We will focus on dif- com/​ s ite/​ r isko​ f bias​ tool/​ w elco​ m e/​ rob-2-​ 0 -​ tool/ ​ c urre​ ferences in treatment effects between subgroups (tech- nt-​versi​on-​of-​rob-2). Assessments will be conducted for nically an interaction effect of treatment and subgroup). each outcome included in the “Summary of findings” Interpretation of SMD values will be based on the cat- table. We will resolve disagreements in the risk of bias egories suggested by Cohen 1988 [104] (i.e., 0.2=”small”, assessments by discussion and/or through consensus 0.5=”medium”, 0.8=”large”). Estimates of overall mean with a third author; if not possible, study authors will be effects will be quoted along with 95% confidence inter- contacted for clarification. The review is concerned with vals (CI). the effect of assignment to the interventions at baseline and will therefore follow the intention-to-treat (ITT) Unit of analysis issues principle. In the case of crossover trials, we will use only data from RoB2 includes five domains: (1) bias arising from the the first (pre-crossover) period. We will treat studies randomisation process, (2) bias due to deviations from with more than two arms that would contribute to a joint Lehnert et al. Systematic Reviews (2022) 11:134 Page 8 of 13 analysis by adjusting standard errors to mimic a splitting Subsequently, we will address the main objectives. In of the placebo group into independent groups. When tri- particular, we will pool differences in effect estimates als include several arms using the same DMT but with between the subgroup categories (e.g. effect differences different modes of administration or different doses, we between the categories “male” and “female” in the sub- will consider only the arm with the approved dose or group “sex”). In a second step, we will look into individ- mode of administration. ual DMTs, at first with placebo as control and next with active controls (i.e. other approved DMTs), whereas pla- Dealing with missing data cebo and active controls will be included in separate anal- In case results are reported both in terms of ITT and yses. In case subgroups have more than two categories, per-protocol analysis sets, the ITT results will be pre- we will dichotomise them. If not possible, we will con- ferred. Care will be taken to track the sample size cor- duct more than one analysis, based on reasonable sum- responding to quoted estimates to be able to correctly maries and integrations of reported subgroup categories. quantify uncertainties. In case of missing or inadequate If subgroup category thresholds used in the identified or inconsistent trial-level data, we will contact the princi- studies are not alike, but similar enough to be reason- pal investigator and primary authors to obtain additional ably combined, we will summarise respective studies (e.g. information. combine age subgroup categories of <38 (vs. >38) years and <40 (vs. >40) years into a younger (vs. older) group). If studies used different thresholds and thus cannot be Assessment of heterogeneity summarised as described below, we will summarise stud- Between-trial heterogeneity will be accounted for by ies based on the most frequently used thresholds for a using random-effects models, where the heterogeneity particular subgroup. parameter (the standard deviation tau) is estimated using We will summarise data for the following subgroups: the Paule-Mandel method [105]. We will judge heteroge- neity (tau) on its absolute scale. In case there is evidence • Age: Baseline age will be dichotomised into a younger for the presence of “fairly extreme” heterogeneity [106], (<40 years) and older group (>40 years). considering the heterogeneity estimate and its 95% CI, • Sex: Sex will be dichotomised into female and male we will refrain from a quantitative synthesis and report a groups. qualitative summary only. • Treatment history: We will dichotomise treatment history into a treatment-naive group and a group in which participants had been treated with DMT(s) Assessment of reporting biases before entering the study (regardless of timeframe We will investigate reporting bias visually using funnel and type of DMT). plots if we identify at least ten clinically and methodo- • Relapse rate: We will dichotomise participants based logically homogeneous studies that would contribute to on the number of relapses per year before randomi- a meta-analysis for the critical outcomes (concerning sation into a group with <1 relapse and a group with the “main analyses” described under the Data Synthesis >2 or more relapses. In case the number of relapses section). If these visual assessments suggest asymmetry, in the previous 2 years are reported, these will be we will use formal tests as suggested by Sterne and col- annualised by assuming a uniform distribution (e.g. leagues [103]. 2 relapses in 24 months will be considered 1 relapse per year). Relapses will be defined as described Data synthesis under Types of outcome measures. If studies are sufficiently similar concerning participants, • Disability level: Based on baseline EDSS values, we eligibility criteria and outcomes, we will pool effect esti- will dichotomise participants into a group with low mates from all DMTs for the overall trial population with disability (EDSS <3.5) and a group with higher levels placebo as a control and conduct meta-analyses for each of disability (EDSS >3.5). Based on baseline MSFC critical and important outcome. Following the proce- scores, we will dichotomise participants into a group dure for the overall trial population, we will pool effect with below median scores and a group with equal or estimates from all DMTs, but this time for each subgroup above median scores. with placebo as a control and conduct meta-analyses. We • MRI activity: We will dichotomise participants based will then look into individual DMTs, at first with placebo on the number of baseline T2 hypointense lesions as control and next with active controls. This analysis will into a group with <9 lesions and a group with >9 precede the main analyses and illustrate the (combined) T2 lesions and based on baseline T2 lesion volume interventions’ overall effects. into a group with below median (or mean) lesion Lehnert et al. Systematic Reviews (2022) 11:134 Page 9 of 13 volume and a group with equal or above median (or Summary of findings and assessment of the certainty mean) lesion volume. We will dichotomise partici- of the evidence pants based on the number of contrast-enhancing T1 We will assess the certainty of the evidence for the criti- hyperintense lesions into a group without lesions and cal outcomes (concerning the "main analyses" described a group with >1 lesions, and based on baseline con- under Data synthesis) using the GRADE approach, which trast-enhancing T1 hyperintense lesion volume into a considers issues related to both internal validity (risk of group with below median lesion volume and a group bias, inconsistency, imprecision, publication bias) and with equal or above median lesion volume. We will external validity (directness of the results). Two review dichotomise participants based on baseline lesion authors will independently rate the quality of the evi- location into a group with infratentorial lesions and a dence for each outcome as “high”, “moderate”, “low” or group with supratentorial lesions. “very low”; discrepancies (see above). The main results of this review will be presented in one or more “Summary Analyses will be carried out using random-effects of findings” table(s) in a transparent tabular format. For models based on a normal-normal hierarchical model, the application of the GRADE system and preparation of combining effect (or interaction) estimates using generic the “Summary of findings” table(s), we will follow the rec- inverse-variance methods and Paule-Mandel estimates ommendations provided in Chapter 14 in the Cochrane for the heterogeneity variance. We will use “R” and the Handbook for Systematic Reviews of Interventions [110]. “metafor” packages [107–109]. In case a quantitative meta-analysis is not possible, we will Discussion attempt to summarise effect estimates and present a struc- A wealth of clinical trials has been conducted over the tured tabulation of the results, with outcomes ordered by past decades in relapsing MS [111, 112]. For many RCTs e.g. risk of bias, according to the guidelines given in Chap- investigating the efficacy of DMTs, sub-group analyses ter 12 of the Cochrane Handbook for Systematic Reviews of were conducted to identify groups of patients with larger Interventions [107]. The main analyses of this review will treatment efficacy. While most individual trials found include all eligible studies and will be concerned with out- differences in the treatment effects for the examined comes reported 2 years post-randomisation. subgroups, effects were mostly relatively small and did not challenge the overall trial’s findings (i.e. DMTs were Subgroup analysis and investigation of heterogeneity significantly more efficacious than placebo in almost all If we identify a sufficient number of studies, we will con- subgroups). While subgroup analyses of RCTs thus were, duct the following subgroup analyses: for most baseline factors considered, unable to identify TEM, this may be because clinical trials are rarely suf- • Different subtypes of MS, i.e. CIS or RRMS; ficiently powered to detect subgroup-specific effects • Different diagnostic criteria, e.g. Poser, original [113]. Combining subgroup data from different studies McDonald or modified McDonald criteria in a meta-analysis could therefore identify patient groups • DMTs grouped by level of efficacy, e.g. low efficacy consistently showing smaller/larger effects, due to the DMTs (i.e. interferon-beta, glatiramer acetate, teri- higher power of the meta-analytic approach. Despite the flunomide, dimethyl fumarate), moderate efficacy large number of RCTs conducted, the respective empiri- DMTs (i.e. cladribine, fingolimod, ozanimod) or high cal evidence from subgroup analyses of clinical trials efficacy DMTs (i.e. ocrelizumab, mitoxantrone, alem- investigating DMT efficacy in people with RRMS has tuzumab, natalizumab) not been comprehensively and systematically assessed • Different definitions of disability worsening, e.g. and summarised to date [51]. Evidence-based informa- increase in the EDSS of at least 1 point from baseline tion on factors predicting significant differential treat- sustained for at least 3 or 6 months over a relapse- ment effects in people with RRMS are lacking for most free period for baseline scores 5.5 or less, or of at approved DMTs and for people with highly active MS. least 0.5 points for baseline scores of more than 5.5 A tailored treatment approach that accounts for indi- • Different follow-up periods, e.g. < 2 years or ≥ 2 viduals’ characteristics and preferences, environmental years. aspects, disease features and biomolecular traits, will be particularly beneficial to improve and maximise out- comes in persons with MS [72, 114]. Solid knowledge Sensitivity analysis about effect modification is therefore an essential ele- We will re-run the analyses (as described in the Data Syn- ment in personalised treatment algorithms [115]. thesis section) excluding studies judged to be at high risk There are some limitations to our planned review. of bias in RoB2. We do not provide an extensive search to identify grey Lehnert et al. Systematic Reviews (2022) 11:134 Page 10 of 13 literature and might potentially miss some relevant stud- DC: is a consultant for Biogen and Hoffmann-La Roche. In the last 3 years, he has received research funding from Hoffmann-La Roche, the International ies. However, we will perform forward and backward ref- Progressive MS Alliance, the MS Society, and the National Institute for Health erence screening and perform an extensive search in all Research (NIHR) University College London Hospitals (UCLH) Biomedical relevant databases. Also, as we have no individual patient Research Centre, and speaker’s honorarium from Novartis. He co-supervises a clinical fellowship at the National Hospital for Neurology and Neurosurgery, data access, this review will depend on the quality and London, which is supported by Merck. comparability of the reported subgroup analyses. MS: has received Grants (Roche, Novartis, Biogen), fees for educational events We plan to publish the review in a peer-reviewed jour- and lectures (Roche, Novartis, Biogen, Merck, Sanofi, Celgene), support for travelling and participation to meetings (Roche, Novartis, Biogen, Merck, nal with an open access option. Sanofi, Celgene) and payment for participation on a Data Safety Monitoring Board, Advisory Board, or Guideline Panel (Roche, Novartis, Biogen, Merck, Sanofi, Celgene, GeNeuro, Medday) and that they have been involved in con- Abbreviations ducting a study funded by Merck that is eligible for inclusion in the review. ADLs: Activities of daily living; AE: Adverse events; ARR​: Annualised relapse TF: has received within the past 3 years personal fees for participation in the rate; CI: Confidence interval; CIS: Clinically isolated syndrome; CNS: Central Scientific Board and lectures of the JUMPstart programme from Fresenius nervous system; DMT: Disease-modifying therapy; EDSS: Expanded Disability Kabi, for participation in data monitoring committees from Novartis, Bayer, Status Scale; EMA: European Medicines Agency; EQ-5D: EuroQol-5 Dimen- Janssen, Roche, Daiichi-Sankyo, Boehringer Ingelheim, Coherex Medical and sion; FAMS: Functional Assessment of Multiple Sclerosis; FDA: U.S. Food and BiosenseWebster and for statistical consultancies from Novartis, Bayer, CSL Drugs; FFS: Fatigue Severity Scale; FIM: Functional Independence Measure; Behring, Galapagos, Vifor, Mediconomics, and Penumbra, but all outside the HAQUAMS: Hamburg Quality of Life Questionnaire in Multiple Sclerosis; ITT: submitted work. Intention to treat; JC: John Cunningham virus; LMSQoL: Leeds Multiple Sclero- CH: has received research grants from Roche, Merck, Bristol Myers Squibbs, sis Quality of Life; MFI: Multidimensional Fatigue Index; MFIS: Modified Fatigue Sanofi, honoraries for talks from Merck and a travel grant from Celgene. TL, CR, Impact Scale; MRI: Magnetic resonance imaging; MS: Multiple sclerosis; MSFC: SK, AF and AR declare that they have no competing interests. Multiple sclerosis functional composite; MSQoL-54: Multiple Sclerosis Quality of Life-54; MUSIQoL: Multiple Sclerosis International Quality of Life question- Author details naire; NEDA: No Evidence of Disease Activity; OR: Odds ratio; PML: Progres- 1 Institute of Neuroimmunology and Multiple Sclerosis, University Medical sive multifocal leukoencephalopathy; QoL: Quality of life; RCT​: Randomised Center Hamburg-Eppendorf (UKE), Hamburg, Germany. 2Department of Medical controlled trial; RRMS: Relapsing-remitting MS; RR: Rate ratio; SDM: Shared Statistics, University Medical Center Göttingen, Göttingen, Germany. 3Institute decision-making; SF-36: Health Status Questionnaire 36; SMD: Standardised of Nursing Science, Faculty of Medicine and University Hospital Cologne, Uni- mean difference; SPMS: Secondary progressive multiple sclerosis; TEM: Treat- versity of Cologne, Cologne, Germany. 4Neurology/Neuroimmunology, Centre ment effect modifier; VAS: Visual analogue scale for fatigue. d’Esclerosi Multiple de Catalunya (Cemcat), Hospital Universitari Vall d’Hebron, Universitat Autonoma de Barcelona, Barcelona, Spain. 5Department of Neuroin- flammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, Supplementary Information University College London, London, UK. 6National Institute for Health Research The online version contains supplementary material available at https://​doi.​ (NIHR), University College London Hospitals (UCLH) Biomedical Research Centre, org/​10.​1186/​s13643-​022-​01997-2. London, UK. 7Department of Oncology, Istituto Ricerche Farmacologiche “Mario Negri” IRCCS, Milano, Italy. 8Institute for Social Medicine and Epidemiology, Nurs- Additional file 1. Search strings. ing Research Unit, University of Lübeck, Lübeck, Germany. Received: 7 December 2021 Accepted: 28 May 2022 Acknowledgements We would like to acknowledge the contributions of Edeltraud Faßhauer, Clau- dia Schilewski and Markus van der Loo. Their work as consumer representa- tives helped to define outcomes and to progress the protocol. References Authors’ contributions 1. Dendrou CA, Fugger L, Friese MA. Immunopathology of multiple sclero- AF, AR, CR, CH, TF, SK and TL drafted and edited the protocol. All authors sis. Nat Rev Immunol. 2015;15(9):545–58. reviewed and approved the protocol before publication. 2. Dobson R, Giovannoni G. Multiple sclerosis - a review. Eur J Neurol. 2019;26(1):27–40. Funding 3. GBD. 2016 Neurology Collaborators. Global, regional, and national bur- Open Access funding enabled and organized by Projekt DEAL. Federal Ministry den of multiple sclerosis 1990-2016: a systematic analysis for the Global of Education and Research (BMBF), Germany (grant 01KG1804). The funding Burden of Disease Study 2016. Lancet Neurol. 2019;18(3):269–85. body had no influence on the design of the protocol. 4. Wallin MT. Global, regional, and national burden of multiple sclerosis 1990-2016: a systematic analysis for the Global Burden of Disease Study Availability of data and materials 2016. Lancet Neurol. 2019;(3):269–85. Not applicable 5. Belbasis L, Bellou V, Evangelou E, Tzoulaki I. Environmental factors and risk of multiple sclerosis: findings from meta-analyses and Mendelian randomization studies. Mult Scler. 2020;26(4):397–404. Declarations 6. Olsson T, Barcellos LF, Alfredsson L. Interactions between genetic, lifestyle and environmental risk factors for multiple sclerosis. Nat Rev Ethics approval and consent to participate Neurol. 2017;13(1):25–36. Not applicable. 7. Chung KK, Altmann D, Barkhof F, Miszkiel K, Brex PA, O’Riordan J, et al. A 30-year clinical and magnetic resonance imaging observational study Consent for publication of multiple sclerosis and clinically isolated syndromes. Ann Neurol. Not applicable. 2020;87(1):63–74. 8. Sormani MP, De Stefano N. Defining and scoring response to IFN-β in Competing interests multiple sclerosis. Nat Rev Neurol. 2013;9(9):504–12. JR: has received honoraria for consultancy from Mylan, Novartis and Sanofi- 9. Einarsson U, Gottberg K, Fredrikson S, von Koch L, Holmqvist LW. Activi- Aventis and compensation for lectures and educational presentations from, ties of daily living and social activities in people with multiple sclerosis Merck-Serono, Novartis, Teva and Sanofi-Aventis in Stockholm County. Clin Rehabil. 2006;20(6):543–51. Lehnert et al. Systematic Reviews (2022) 11:134 Page 11 of 13 10. Berrigan LI, Fisk JD, Patten SB, Tremlett H, Wolfson C, Warren S, et al. Pooled Analysis from the FREEDOMS/FREEDOMS II Trials. Neurol Ther. Health-related quality of life in multiple sclerosis: direct and indirect 2019;8(2):461-475. https://​doi.​org/​10.​1007/​s40120-​019-​0146-z. Erratum effects of comorbidity. Neurology. 2016;86(15):1417–24. in: Neurol Ther. 2020;9(1):193-195. 11. Hawton A, Green C. Health utilities for multiple sclerosis. Value Health 32. Giovannoni G, Butzkueven H, Dhib-Jalbut S, Hobart J, Kobelt G, Pepper J Int Soc Pharmacoeconomics Outcomes Res. 2016;19(4):460–8. G, et al. Brain health: time matters in multiple sclerosis. Mult Scler Relat 12. Manouchehrinia A, Tanasescu R, Tench CR, Constantinescu CS. Mor- Disord. 2016;9(Suppl 1):S5–48. tality in multiple sclerosis: meta-analysis of standardised mortality 33. Harding K, Williams O, Willis M, Hrastelj J, Rimmer A, Joseph F, et al. Clini- ratios. J Neurol Neurosurg Psychiatry. 2016;87(3):324–31. cal outcomes of escalation vs early intensive disease-modifying therapy 13. García-Domínguez JM, Maurino J, Martínez-Ginés ML, Carmona O, in patients with multiple sclerosis. JAMA Neurol. 2019;76(5):536–41. Caminero AB, Medrano N, et al. Economic burden of multiple scle- 34. He A, Merkel B, Brown JWL, Zhovits Ryerson L, Kister I, Malpas CB, et al. rosis in a population with low physical disability. BMC Public Health. Timing of high-efficacy therapy for multiple sclerosis: a retrospective 2019;19(1):609. observational cohort study. Lancet Neurol. 2020;19(4):307–16. 14. Kobelt G, Thompson A, Berg J, Gannedahl M, Eriksson J. New insights 35. Ayrignac X, Bilodeau P-A, Prat A, Girard M, Labauge P, Le Lorier J, et al. into the burden and costs of multiple sclerosis in Europe. Mult Scler. Assessing the risk of multiple sclerosis disease-modifying therapies. 2017;23(8):1123–36. Expert Rev Neurother. 2019;19(7):695–706. 15. Degenhardt A, Ramagopalan SV, Scalfari A, Ebers GC. Clinical prog- 36. Berger JR. Classifying PML risk with disease modifying therapies. Mult nostic factors in multiple sclerosis: a natural history review. Nat Rev Scler Relat Disord. 2017;12:59–63. Neurol. 2009;5(12):672–82. 37. Higuera L, Carlin CS, Anderson S. Adherence to disease-modify- 16. Azevedo CJ, Overton E, Khadka S, Buckley J, Liu S, Sampat M, et al. ing therapies for multiple sclerosis. J Manag Care Spec Pharm. Early CNS neurodegeneration in radiologically isolated syndrome. 2016;22(12):1394–401. Neurol Neuroimmunol Neuroinflammation. 2015;2(3):e102. 38. Lizán L, Comellas M, Paz S, Poveda JL, Meletiche DM, Polanco C. Treat- 17. Confavreux C, Vukusic S. The clinical course of multiple sclerosis. ment adherence and other patient-reported outcomes as cost determi- Handb Clin Neurol. 2014;122:343–69. nants in multiple sclerosis: a review of the literature. Patient Prefer 18. Leray E, Yaouanq J, Le Page E, Coustans M, Laplaud D, Oger J, et al. Adherence. 2014;8:1653–64. Evidence for a two-stage disability progression in multiple sclerosis. 39. Bruce JM, Lynch SG. Multiple sclerosis: MS treatment adherence--how Brain J Neurol. 2010;133(Pt 7):1900–13. to keep patients on medication? Nat Rev Neurol. 2011;7(8):421–2. 19. Bross M, Hackett M, Bernitsas E. Approved and emerging disease 40. Hansen K, Schüssel K, Kieble M, Werning J, Schulz M, Friis R, et al. modifying therapies on neurodegeneration in multiple sclerosis. Int J Adherence to disease modifying drugs among patients with mul- Mol Sci. 2020;21(12):4312. tiple sclerosis in Germany: a retrospective cohort study. PloS One. 20. Cree BAC, Hollenbach JA, Bove R, Kirkish G, Sacco S, Caverzasi E, et al. 2015;10(7):e0133279. Silent progression in disease activity-free relapsing multiple sclerosis. 41. Burks J, Marshall TS, Ye X. Adherence to disease-modifying therapies Ann Neurol. 2019;85(5):653–66. and its impact on relapse, health resource utilization, and costs among 21. Li H, Hu F, Zhang Y, Li K. Comparative efficacy and acceptability of patients with multiple sclerosis. Clin Outcomes Res CEOR. 2017;9:251–60. disease-modifying therapies in patients with relapsing-remitting 42. Menzin J, Caon C, Nichols C, White LA, Friedman M, Pill MW. Narrative multiple sclerosis: a systematic review and network meta-analysis. J review of the literature on adherence to disease-modifying therapies Neurol. 2020;267(12):3489–98. among patients with multiple sclerosis. J Manag Care Pharm JMCP. 22. Tintore M, Vidal-Jordana A, Sastre-Garriga J. Treatment of mul- 2013;19(1 Suppl A):S24–40. tiple sclerosis - success from bench to bedside. Nat Rev Neurol. 43. Giovannoni G. Personalized medicine in multiple sclerosis. Neurode- 2019;15(1):53–8. gener Dis Manag. 2017;7(6s):13–7. 23. Rae-Grant A, Day GS, Marrie RA, Rabinstein A, Cree BAC, Gronseth 44. Rotstein D, Montalban X. Reaching an evidence-based prognosis GS, et al. Comprehensive systematic review summary: disease- for personalized treatment of multiple sclerosis. Nat Rev Neurol. modifying therapies for adults with multiple sclerosis: Report of 2019;15(5):287–300. the Guideline Development, Dissemination, and Implementation 45. Smith AL, Cohen JA, Hua LH. Therapeutic targets for multiple sclerosis: Subcommittee of the American Academy of Neurology. Neurology. current treatment goals and future directions. Neurother J Am Soc Exp 2018;90(17):789–800. Neurother. 2017;14(4):952–60. 24. Tramacere IBM Capobussi M, Castellini G, Citterio A, Del Giovane 46. Confavreux C, Vukusic S, Adeleine P. Early clinical predictors and C, et al. Adverse effects of immunotherapies for multiple sclerosis: progression of irreversible disability in multiple sclerosis: an amnesic a network meta-analysis [Protocol]. Cochrane Database Syst Rev. process. Brain J Neurol. 2003;126(Pt 4):770–82. 2016;5:CD012186. https://​doi.​org/​10.​1002/​14651​858.​CD012​186. 47. Gasperini C, Prosperini L, Tintoré M, Sormani MP, Filippi M, Rio J, et al. 25. Ziemssen T, De Stefano N, Sormani MP, Van Wijmeersch B, Wiendl H, Unraveling treatment response in multiple sclerosis: a clinical and MRI Kieseier BC. Optimizing therapy early in multiple sclerosis: an evidence- challenge. Neurology. 2019;92(4):180–92. based view. Mult Scler Relat Disord. 2015;4(5):460–9. 48. Schmidt AF, Klungel OH, Nielen M, de Boer A, Groenwold RHH, Hoes 26. Coret F, Pérez-Miralles FC, Gascón F, Alcalá C, Navarré A, Bernad A, et al. AW. Tailoring treatments using treatment effect modification. Pharma- Onset of secondary progressive multiple sclerosis is not influenced by coepidemiol Drug Saf. 2016;25(4):355–62. current relapsing multiple sclerosis therapies. Mult Scler J - Exp Transl 49. Sormani MP. Subgroup analysis in MS trials. Mult Scler. 2017;23(1):34–5. Clin. 2018;4(2):2055217318783347. 50. Corraini P, Olsen M, Pedersen L, Dekkers OM, Vandenbroucke JP. Effect 27. Sormani MP, Bruzzi P. Can we measure long-term treatment effects in modification, interaction and mediation: an overview of theoretical multiple sclerosis? Nat Rev Neurol. 2015;11(3):176–82. insights for clinical investigators. Clin Epidemiol. 2017;9:331–8. 28. Wingerchuk DM, Weinshenker BG. Disease modifying therapies for 51. Signori A, Schiavetti I, Gallo F, Sormani MP. Subgroups of multiple relapsing multiple sclerosis. BMJ. 2016;354:i3518. sclerosis patients with larger treatment benefits: a meta-analysis of 29. Brown JWL, Coles A, Horakova D, Havrdova E, Izquierdo G, Prat A, et al. randomized trials. Eur J Neurol. 2015;22(6):960–6. Association of initial disease-modifying therapy with later conversion to 52. Río J, Ruiz-Peña J. Short-term suboptimal response criteria for predict- secondary progressive multiple sclerosis. JAMA. 2019;321(2):175–87. ing long-term non-response to first-line disease modifying therapies in 30. Filippini G, Del Giovane C, Clerico M, Beiki O, Mattoscio M, Piazza F, et al. multiple sclerosis: a systematic review and meta-analysis. J Neurol Sci. Treatment with disease-modifying drugs for people with a first clinical 2016;361:158–67. attack suggestive of multiple sclerosis. Cochrane Database Syst Rev. 53. Lu Y, Zhao J, Zhan Q. Effect of interferon-β1α therapy on multi- 2017;4(4):CD012200. https://​doi.​org/​10.​1002/​14651​858.​CD012​200.​ ple sclerosis based on gadolinium-enhancing or active T2 mag- pub2. netic resonance imaging outcomes: a meta-analysis. Neurol Res. 31. Ghezzi A, Chitnis T, K-Laflamme A, Meinert R, Häring DA, Pohl D. 2016;38(10):909–15. Long-Term Effect of Immediate Versus Delayed Fingolimod Treatment 54. Dobson R, Rudick R, Turner B. Assessing treatment response to in Young Adult Patients with Relapsing-Remitting Multiple Sclerosis: interferon-β: is there a role for MRI? Neurology. 2014;82(3):248–54. Lehnert et al. Systematic Reviews (2022) 11:134 Page 12 of 13 55. Bar-Or A, Gold R, Kappos L, Arnold DL, Giovannoni G, Selmaj K, et al. clinically isolated syndrome: a systematic review. Acta Neurol Scand. Clinical efficacy of BG-12 (dimethyl fumarate) in patients with relapsing- 2019;139(1):18–32. remitting multiple sclerosis: subgroup analyses of the DEFINE study. J 78. Galassi S, Prosperini L, Logoteta A, Hirsch MN, Fanelli F, De Giglio L, Neurol. 2013;260(9):2297–305. et al. A lesion topography-based approach to predict the outcomes of 56. Giovannoni G, Cook S, Rammohan K, Rieckmann P, Sørensen PS, patients with multiple sclerosis treated with Interferon Beta. Mult Scler Vermersch P, et al. Sustained disease-activity-free status in patients with Relat Disord. 2016;8:99–106. relapsing-remitting multiple sclerosis treated with cladribine tablets in 79. Wattjes MP, Rovira À, Miller D, Yousry TA, Sormani MP, de Stefano MP, the CLARITY study: a post-hoc and subgroup analysis. Lancet Neurol. et al. Evidence-based guidelines: MAGNIMS consensus guidelines on 2011;10(4):329–37. the use of MRI in multiple sclerosis--establishing disease prognosis and 57. Li R, Sun X, Shu Y, Mao Z, Xiao L, Qiu W, et al. Sex differences in monitoring patients. Nat Rev Neurol. 2015;11(10):597–606. outcomes of disease-modifying treatments for multiple sclerosis: a 80. Daumer M, Neuhaus A, Morrissey S, Hintzen R, Ebers GC. MRI systematic review. Mult Scler Relat Disord. 2017;12:23–8. as an outcome in multiple sclerosis clinical trials. Neurology. 58. Weideman AM, Tapia-Maltos MA, Johnson K, Greenwood M, Bielekova 2009;72(8):705–11. B. Meta-analysis of the age-dependent efficacy of multiple sclerosis 81. Bermel RA, You X, Foulds P, Hyde R, Simon JH, Fisher E, et al. Predic- treatments. Front Neurol. 2017;8:577. tors of long-term outcome in multiple sclerosis patients treated with 59. Confavreux C, Vukusic S. Age at disability milestones in multiple sclero- interferon β. Ann Neurol. 2013;73(1):95–103. sis. Brain J Neurol. 2006;129(Pt 3):595–605. 82. Río J, Rovira À, Tintoré M, Otero-Romero S, Comabella M, Vidal-Jordana 60. Lassmann H, van Horssen J, Mahad D. Progressive multiple sclerosis: Á, et al. Disability progression markers over 6-12 years in interferon-β- pathology and pathogenesis. Nat Rev Neurol. 2012;8(11):647–56. treated multiple sclerosis patients. Mult Scler. 2018;24(3):322–30. 61. Denkinger MDM, Leins H, Schirmbeck R. HSC aging and senescent 83. Coles AJ, Twyman CL, Arnold DL, Cohen JA, Confavreux C, Fox EJ, immune remodeling. Trends Immunol. 2015;36(12):815–24. et al. Alemtuzumab for patients with relapsing multiple sclerosis after 62. Díaz C, Zarco LA, Rivera DM. Highly active multiple sclerosis: an update. disease-modifying therapy: a randomised controlled phase 3 trial. Mult Scler Relat Disord. 2019;30:215–24. Lancet Lond Engl. 2012;380(9856):1829–39. 63. Cutter GR, Baier ML, Rudick RA, Cookfair DL, Fischer JS, Petkau J, et al. 84. He D, Zhang C, Zhao X, Zhang Y, Dai Q, Li Y, et al. Teriflunomide for Development of a multiple sclerosis functional composite as a clinical multiple sclerosis. Cochrane Database Syst Rev. 2016;22(3):CD009882. trial outcome measure. Brain J Neurol. 1999;122(Pt 5):871–82. 85. La Mantia L, Di Pietrantonj C, Rovaris M, Rigon G, Frau S, Berardo 64. Kappos L, Wolinsky JS, Giovannoni G, Arnold DL, Wang Q, Bernasconi C, F, et al. Interferons-beta versus glatiramer acetate for relapsing- et al. Contribution of relapse-independent progression vs relapse-asso- remitting multiple sclerosis. Cochrane Database Syst Rev. ciated worsening to overall confirmed disability accumulation in typical 2016;11(11):CD009333. relapsing multiple sclerosis in a pooled analysis of 2 randomized clinical 86. Gafson A, Craner MJ, Matthews PM. Personalised medicine for multiple trials. JAMA Neurol. 2020;77(9):1132–40. sclerosis care. Mult Scler. 2017;23(3):362–9. 65. Sharrack B, Hughes RA, Soudain S, Dunn G. The psychometric proper- 87. Lublin FD, Reingold SC, Cohen JA, Cutter GR, Sørensen PS, Thompson ties of clinical rating scales used in multiple sclerosis. Brain J Neurol. AJ, et al. Defining the clinical course of multiple sclerosis: the 2013 revi- 1999;122(Pt 1):141–59. sions. Neurology. 2014;83(3):278–86. 66. Fox EJ, Markowitz C, Applebee A, Montalban X, Wolinsky JS, Belachew 88. Miller DH, Chard DT, Ciccarelli O. Clinically isolated syndromes. Lancet S, et al. Ocrelizumab reduces progression of upper extremity Neurol. 2012;11(2):157-69. impairment in patients with primary progressive multiple sclerosis: 89. Schumacher GA, Beebe G, RF K, LT K, JF K, Mcdowell F, et al. Problems of findings from the phase III randomized ORATORIO trial. Mult Scler. experimental trials of therapy in multiple sclerosis: report by the panel 2018;24(14):1862–70. on the evaluation of experimental trials of therapy in multiple sclerosis. 67. DeLuca J, Chiaravalloti ND, Sandroff BM. Treatment and management Ann N Y Acad Sci. 1965;122:552–68. of cognitive dysfunction in patients with multiple sclerosis. Nat Rev 90. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, Neurol. 2020;16(6):319–32. et al. New diagnostic criteria for multiple sclerosis: guidelines for 68. Koch M, Kingwell E, Rieckmann P, Tremlett H. The natural history of research protocols. Ann Neurol. 1983;13(3):227–31. secondary progressive multiple sclerosis. J Neurol Neurosurg Psychiatry. 91. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, 2010;81(9):1039–43. et al. Recommended diagnostic criteria for multiple sclerosis: guidelines 69. Ramien C, Taenzer A, Lupu A, Heckmann N, Engler JB, Patas K, et al. Sex from the International Panel on the diagnosis of multiple sclerosis. Ann effects on inflammatory and neurodegenerative processes in multiple Neurol. 2001;50(1):121–7. sclerosis. Neurosci Biobehav Rev. 2016;67:137–46. 92. Polman CH, Reingold SC, Edan G, Filippi M, Hartung H-P, Kappos L, et al. 70. Ysrraelit MC, Correale J. Impact of sex hormones on immune function Diagnostic criteria for multiple sclerosis: 2005 revisions to the "McDon- and multiple sclerosis development. Immunology. 2019;156(1):9–22. ald Criteria". Ann Neurol. 2005;58(6):840–6. 71. Du S, Itoh N, Askarinam S, Hill H, Arnold AP, Voskuhl RR. XY sex chromo- 93. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, et al. some complement, compared with XX, in the CNS confers greater Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDon- neurodegeneration during experimental autoimmune encephalomy- ald criteria. Ann Neurol. 2011;69(2):292–302. elitis. Proc Natl Acad Sci U S A. 2014;111(7):2806–11. 94. Karabudak R, Dahdaleh M, Aljumah M, Alroughani R, Alsharoqi IA, AlTa- 72. Montalban X, Gold R, Thompson AJ, Otero-Romero S, Amato MP, Chan- han AM, et al. Functional clinical outcomes in multiple sclerosis: current draratna D, et al. ECTRIMS/EAN guideline on the pharmacological treat- status and future prospects. Mult Scler Relat Disord. 2015;4(3):192–201. ment of people with multiple sclerosis. Eur J Neurol. 2018;25(2):215–37. 95. Khurana V, Sharma H, Afroz N, Callan A, Medin J. Patient-reported 73. Scalfari A, Neuhaus A, Degenhardt A, Rice GP, Muraro PA, Daumer M, outcomes in multiple sclerosis: a systematic comparison of available et al. The natural history of multiple sclerosis: a geographically based measures. Eur J Neurol. 2017;24(9):1099–107. study 10: relapses and long-term disability. Brain J Neurol. 2010;133(Pt 96. Tur C, Moccia M, Barkhof F, Chataway J, Sastre-Garriga J, Thompson AJ, 7):1914–29. et al. Assessing treatment outcomes in multiple sclerosis trials and in 74. IFNB Multiple Sclerosis Study Group. Interferon beta-1b is effective in the clinical setting. Nat Rev Neurol. 2018;14(2):75–93. relapsing-remitting multiple sclerosis. Neurology. 1993;43(4):655. 97. German Multiple Sclerosis Society [Deutsche Multiple Sklerose Gesells- 75. O’Connor P, Wolinsky JS, Confavreux C, Comi G, Kappos L, Olsson TP, chaft (DMSG)]. Available from: https://​www.​dmsg.​de/. Accessed 12 et al. Randomized trial of oral teriflunomide for relapsing multiple June 2022. sclerosis. N Engl J Med. 2011;365(14):1293–303. 98. Kurtzke JF. Rating neurologic impairment in multiple scle- 76. Thompson AJ, Banwell BL, Barkhof F, Carroll WM, Coetzee T, Comi G, rosis: an expanded disability status scale (EDSS). Neurology. et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald 1983;33(11):1444–52. criteria. Lancet Neurol. 2018;17(2):162–73. 99. Rudick RA, Polman CH, Cohen JA, Walton MK, Miller AE, Confavreux C, 77. Rahn AC, Köpke S, Stellmann J-P, Schiffmann I, Lukas C, Chard D, et al. et al. Assessing disability progression with the Multiple Sclerosis Func- Magnetic resonance imaging as a prognostic disability marker in tional Composite. Mult Scler. 2009;15(8):984–97. Lehnert et al. Systematic Reviews (2022) 11:134 Page 13 of 13 100. Rotstein DL, Healy BC, Malik MT, Chitnis T, Weiner HL. Evaluation of No Evidence of Disease Activity in a 7-year longitudinal multiple sclerosis cohort. JAMA Neurol. 2015;72(2):152–8. 101. Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.3 (updated February 2022). Cochrane. 2022. Available from: www.​train​ing.​cochr​ane.​org/​handb​ook. Accessed 12 June 2022. 102. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1–1. 103. Sterne JAC, Sutton AJ, Ioannidis JPA, Terrin N, Jones DR, Lau J, et al. Rec- ommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002. 104. Cohen 1988 J. Statistical power analysis for the behavioural sciences. 2nd ed. New York: Routledge; 1988. 105. Veroniki AA, Jackson D, Viechtbauer W, Bender R, Bowden J, Knapp G, et al. Methods to estimate the between-study variance and its uncer- tainty in meta-analysis. Res Synth Methods. 2016;7(1):55–79. 106. Spiegelhalter DJ, Abrams KR, Myles JP. Bayesian approaches to clinical trials and health-care evaluation. Chichester: Wiley; 2004. 107. McKenzie JE, Brennan SE. Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editors. Cochrane Handbook for Systematic Reviews of Interventions version 6.3 (updated February 2022). Cochrane. 2022. Available from: www.​train​ing.​cochr​ane.​org/​ handb​ook. Accessed 12 June 2022. 108. R Core Team. R. A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2020. 109. Viechtbauer W. Conducting meta-analyses in R with the metafor pack- age. J Stat Softw. 2010;36(3):1–48. 110. Schünemann HJ, Higgins JP, Vist GE, Glasziou P, Akl EA, Skoetz N, et al. Completing ‘Summary of findings’ tables and grading the certainty of the evidence. Cochrane Handbook Syst Rev Intervent. 2019:375–402. 111. Röver C, Nicholas R, Straube S, Friede T. Changing EDSS progression in placebo cohorts in relapsing MS: a systematic review and meta-regres- sion. PloS One. 2015;10(9):e0137052. 112. Steinvorth SM, Röver C, Schneider S, Nicholas R, Straube S, Friede T. Explaining temporal trends in annualised relapse rates in placebo groups of randomised controlled trials in relapsing multiple sclerosis: systematic review and meta-regression. Mult Scler. 2013;19(12):1580–6. 113. Tanniou J, van der Tweel I, Teerenstra S, Roes KCB. Subgroup analyses in confirmatory clinical trials: time to be specific about their purposes. BMC Med Res Methodol. 2016;16:20. 114. Kalincik T, Manouchehrinia A, Sobisek L, Jokubaitis V, Spelman T, Horakova D, et al. Towards personalized therapy for multiple sclerosis: prediction of individual treatment response. Brain J Neurol. 2017;140(9):2426–43. 115. Chitnis T, Prat A. A roadmap to precision medicine for multiple sclerosis. Mult Scler. 2020;26(5):522-32. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions