Further controlled and randomized prospective studies in Multiple sclerosis, analyzing the potential impact of treatment discontinuation on disability progression, focal disease activity and quality of life are needed. The optimum patient age and duration of inactive SPMS before treatment withdrawal and the monitoring procedures also need to be specified, the ultimate goal being to provide evidence-based recommendations for clinical practice. Following the previous retrospective experience, we decided to drive a multicenter prospective study in France based on the hypothesis that stopping disease modifying therapy will not induce an increased risk of disability progression and relapse in selected SPMS patients (older patients without lesion activity) but will improve the quality of life and may reduce treatment-related costs.

Multiple sclerosis (MS) usually evolves over decades and can present several phenotypes. Approximately 85% of newly diagnosed Multiple Sclerosis (MS) patients present the Relapsing-Remitting MS (RRMS) phenotype. After a mean time of approximatively 20 years, a large majority of these patients evolve to the so-called "Secondary Progressive MS" (SPMS) phase. SPMS is characterized by an irreversible disability progression not related to relapses, although relapses could be superimposed. Nevertheless, the shift in-between RRMS and SPMS is not clear. Different subtypes of SPMS have been recently defined by F Lublin et al. This classification takes into account persistent focal inflammatory activity (active vs inactive SPMS) along with disease progression (progressing vs non-progressing SPMS). In clinical routine, it is important to identify these stages of MS as they differently respond to the disease modifying therapies (DMTs).

Introducing DMTs during the RRMS phase had consistently demonstrated a significant impact on the annual relapse rate (ARR) and on the short-term disability progression. Conversely, during the SPMS phase, the impact of DMTs remained uncertain on disability progression, especially in older patients, with "inactive" disease. As a matter of fact, the DMTs are considered to be anti-inflammatory by nature, but the focal inflammation reduces with age and disease duration.

In addition, the DMTs have side effects and cost approximately 10,000 euros per year and per patient. In this context, the usefulness of continuing DMTs in "inactive" SPMS patients older than 50 years is questionable.

In a preliminary retrospective study conducted at our Institute which enrolled 100 SPMS patients, the ARR remained stable 3 years after treatment withdrawal (0.07, 95% CI [0.05, 0.11]), relative to the 3 years prior to treatment withdrawal (0.12, [0.09, 0.16]). EDSS scores were available for 94 patients The percentage of patients experiencing a significant increase of their EDSS score during the 3 years after treatment withdrawal also remained stable compared to the 3 years prior treatment withdrawal. These preliminary data support the safety of DMTs withdrawal in selected SPMS patients. However, further prospective studies are needed to provide evidence-based guidelines for daily practice.

This randomized controlled clinical trial thus aims to compare SPMS patients older than 50 years without evidence of focal inflammatory activity for 3 years, stopping DMTs versus patients with the same criteria still receiving treatment. We hypothesize that stopping DMTs will not induce an increased risk of disability progression or relapse in SPMS patients but will improve their quality of life and have an impact on treatment-related costs.

So far, the impact of DMTs withdrawal in a selected SPMS population has not been explored. Having evidence-based recommendations on the treatment management of these patients is essential, considering the consequences in terms of disability, relapses, side effects, quality of life and costs. DMTs in MS are now available since 20 years, with an increasing number of approved molecules. As a matter of fact, this question concerns a large number of patients: a retrospective analysis of patients included in the Rennes EDMUS database allowed to identify 71 SPMS patients older than 50 years and without evidence of focal inflammatory activity for 3 years actually undergoing a DMT.

For evident conflict of interests, the pharmaceutical firms will not promote or fund clinical trials on treatment withdrawal. A randomized controlled study initiated by academia and financed by public funding should be performed to explore these questions. We will evaluate the impact of these changes from the patient and the health system's points of view. The results of this clinical trial will lead to a concrete change in clinical practice.

Inclusion Criteria:

• Patients > 50 years old;
• Secondary progressive phenotype for at least 3 years; The secondary progressive phenotype will be defined as progressive deterioration of disability not due to relapse, with an increase of at least 1 EDSS point since the beginning of the progressive phase (or 0.5 EDSS point if EDSS score ≥ 5.5).
• Disease modifying therapy of MS for at least 3 years (interferon, glatiramer acetate, teriflunomide, dimethyl fumarate, cyclophosphamide, azathioprine, methotrexate, mycophenolate mofetil, rituximab, ocrelizumab); Both patients with the same DMT or with successive DMTs during 3 years can be included. It is important to note that patients could have been treated with fingolimod or natalizumab 2 or 3 years before inclusion, but not during the year before inclusion ;
• No evidence of focal inflammatory activity for at least 3 years (no clinical relapse and no gadolinium enhancement on an MRI scan);
• EDSS≥3.

Concomitant medications with Fampridine are allowed throughout the study, provided they have been introduced at least 1 months before inclusion.

Natalizumab and fingolimod during the year before inclusion were excluded because of the risk of recurrence of inflammatory activity or even rebound of inflammatory activity after withdrawal.

Both patients with the same DMT or with successive DMTs during 3 years can be included, as for example, cyclophosphamide is used for 1 or 2 years, sometimes followed by mycophenolate mofetil.

For Rituximab and Ocrelizumab, inclusion in STOP-I-SEP will be at 6 months from the last infusion to take into account the mode of action of these treatments and their specific administration scheme.

Exclusion Criteria:

• Patients treated with mitoxantrone or alemtuzumab, during the previous 3 years before inclusion;
• Patients treated with natalizumab or fingolimod during the year before inclusion;
• Change of disease modifying therapy of MS for less than a year
• Other neurological or systemic disease ;
• Incapacity to understand or sign the consent form ;
• Contraindication to MRI ;
• Pregnancy or breast-feeding ;
• Patient in another clinical trial
• Persons referred to in Articles L. 1121-5 to L. 1121-8 and L. 1122-1-2 of the Public Health Code (eg minors, protected adults, …).