Progressive MS Treatment: 5 Critical Approaches Compared
Two BTK inhibitors read out in opposite directions, anti-CD20 reached a more disabled population, and simvastatin closed a door. What that leaves for the clinic.
One-Minute Takeaway
- Progressive MS treatment now has phase 3 randomised evidence across five mechanistic classes; three have a positive trial on confirmed disability progression, and every effect is modest.
- Tolebrutinib reduced 6-month confirmed disability progression in non-relapsing SPMS (HR 0.69, 95% CI 0.55–0.88; n=1,131) and is now EU-approved. The FDA declined it over liver injury.
- Ocrelizumab’s benefit extended to older and more disabled primary progressive patients (HR 0.70, 95% CI 0.57–0.86; n=1,013), widening who might reasonably be offered it.
- Siponimod remains the only agent with a positive phase 3 across secondary progressive MS as a whole (HR 0.79, 95% CI 0.65–0.95; n=1,651).
- High-dose simvastatin showed no effect (adjusted HR 1.13, 95% CI 0.91–1.39; n=964), ending the largest neuroprotection trial run in SPMS.
Why Progressive MS Treatment Matters
More than twenty disease-modifying therapies are licensed for multiple sclerosis, and almost all target relapse biology. None reaches the disability that accrues quietly between relapses, or after they stop.
That accrual is what progressive MS treatment has to address. The pathology sits behind a largely intact blood–brain barrier: iron-rimmed chronic active lesions, activated microglia, and meningeal ectopic lymphoid tissue in roughly 40% of secondary progressive cases.
Intravenous antibodies deplete circulating B cells but do not readily cross into that compartment. Brain-penetrant small molecules became the field’s principal bet.
Until recently the map was small: ocrelizumab for primary progressive disease, siponimod for active SPMS, nothing for non-relapsing SPMS.
A Practical Taxonomy of Progressive MS Treatment
Five mechanistic classes now carry randomised data in a progressive population. The table separates what has been tested from what has been shown.
| Study | Class | Design | n | Population | Comparison | Key result on primary endpoint | Tier |
|---|---|---|---|---|---|---|---|
| HERCULES Fox 2025 | BTK inhibitor | Phase 3 RCT, 2:1, event-driven | 1,131 | Non-relapsing SPMS, EDSS 3.0–6.5 | Tolebrutinib 60 mg/day vs placebo | 6-month CDP 22.6% vs 30.7%; HR 0.69 (95% CI 0.55–0.88); P=0.003 | V |
| PERSEUS Reich 2026 | BTK inhibitor | Phase 3 RCT, 2:1 | 767 | PPMS, EDSS 2.0–6.5 | Tolebrutinib vs placebo | 6-month composite CDP HR 1.01 (95% CI 0.81–1.26); P=0.94 — not met | S |
| FENtrepid Bar-Or 2026 | BTK inhibitor | Phase 3 RCT, 1:1, ≥120 weeks | 985 | PPMS, age 18–65, EDSS 3.0–6.5 | Oral fenebrutinib vs IV ocrelizumab | 12-week composite CDP HR 0.88 (95% CI 0.75–1.03); non-inferior, not superior | S |
| ORATORIO-HAND Giovannoni 2026 | Anti-CD20 | Phase 3b RCT, 1:1, 144 weeks | 1,013 | PPMS, age 18–65, EDSS 3.0–8.0 | Ocrelizumab 600 mg IV q6 months vs placebo | 12-week composite CDP 32.7% vs 40.4%; HR 0.70 (95% CI 0.57–0.86); P=0.0007 | V |
| ORATORIO Montalban 2017 | Anti-CD20 | Phase 3 RCT, 2:1, ≥120 weeks | 732 | PPMS, age ≤55, EDSS ≤6.5 | Ocrelizumab 600 mg IV q24 weeks vs placebo | 12-week CDP 32.9% vs 39.3%; HR 0.76 (95% CI 0.59–0.98); P=0.03 | V |
| OLYMPUS Hawker 2009 | Anti-CD20 | Phase 2/3 RCT, 2:1, 96 weeks | 439 | PPMS | Rituximab 1,000 mg q24 weeks vs placebo | 12-week CDP 30.2% vs 38.5%; P=0.14 — not met | V |
| EXPAND Kappos 2018 | S1P modulator | Phase 3 RCT, 2:1, event-driven | 1,651 | SPMS; 64% relapse-free ≥2 years | Siponimod 2 mg/day vs placebo | 3-month CDP 26% vs 32%; HR 0.79 (95% CI 0.65–0.95); P=0.013 | V |
| MS-STAT2 Chataway 2025 | Neuroprotection | Phase 3 RCT, 1:1 | 964 | Non-active SPMS, EDSS 4.0–6.5 | Simvastatin 80 mg/day vs placebo | 6-month CDP 40% vs 36%; adjusted HR 1.13 (95% CI 0.91–1.39); P=0.26 — no effect | V |
| SPRINT-MS Fox 2018 | Neuroprotection | Phase 2 RCT, 96 weeks | 255 | PPMS or SPMS, progressing | Ibudilast ≤100 mg/day vs placebo | Brain parenchymal fraction difference 0.0009/year (95% CI 0.00004–0.0017); P=0.04 | V |
| Boffa 2023 | Immune reconstitution | Propensity-matched cohort (Class III) | 2,054 (79 aHSCT / 1,975 DMT) | Active SPMS | aHSCT vs other anti-inflammatory DMTs | Time to 6-month CDP longer: HR 0.50 (95% CI 0.31–0.81); P=0.005 | V |
Tier key. V verified against the primary publication with a resolving DOI. S presented at ACTRIMS Forum 2026 with sponsor confirmation, not yet peer reviewed, no DOI.
Each row carries its own trial’s primary endpoint; a 12-week composite is not a 6-month EDSS threshold. Treat the two BTK rows in PPMS as provisional. CDP = confirmed disability progression.
Brain-Penetrant BTK Inhibition
BTK sits downstream of the B-cell receptor and of Fc and Toll-like receptors on myeloid cells, including microglia. Inhibiting it modulates rather than depletes, and the molecules reach the CNS. On paper, the closest match yet to progressive MS biology.
What HERCULES showed
HERCULES randomised 1,131 adults with non-relapsing SPMS to tolebrutinib or placebo. Entry required progression in the previous year and no relapse for two — the cleanest non-relapsing cohort assembled so far.
Confirmed disability improvement roughly doubled, but as a nominal secondary endpoint outside the confirmatory hierarchy, so it reads as supportive rather than shown.
Clinical Pearl
A post-hoc analysis of HERCULES and the two GEMINI trials, presented at ACTRIMS 2025, found benefit tracked baseline paramagnetic rim lesion burden. The reduction in 6-month disability worsening reached 54% in participants with four or more rim lesions, and essentially vanished in those with none.
No confidence intervals were published in the conference material retrieved here. The finding is exploratory, but it is the most concrete biomarker candidate in progressive MS treatment.
Where the class split
PERSEUS tested the same drug in 767 adults with PPMS and found nothing. The hazard ratio sat on the null (HR 1.01, 95% CI 0.81–1.26), with favourable MRI signals that did not translate.
FENtrepid compared oral fenebrutinib against IV ocrelizumab in 985 PPMS patients and met non-inferiority (HR 0.88, 95% CI 0.75–1.03). The interval crosses one: equivalence, not superiority.
The strongest component signal was hand function, where fenebrutinib beat ocrelizumab on the nine-hole peg test (HR 0.74, 95% CI 0.56–0.98).
The safety problem
Hepatotoxicity now constrains the class. Across the placebo-controlled trials in a 2026 systematic review of six BTK inhibitor RCTs, serious hepatotoxicity affected 5.7% of treated participants versus 2.6% on placebo (RR 2.19, 95% CI 1.13–4.25).
The FDA’s redacted complete response letter identified six severe drug-induced liver injury cases among roughly 2,700 exposed patients, and judged the fatal-DILI risk among the highest in the class. One participant underwent liver transplantation and died. The agency also faulted efficacy: historical MRI had not been collected, so active SPMS could not be separated retrospectively.
Fenebrutinib’s liver signal looks milder: transaminase elevations in 13.3% versus 2.9% on ocrelizumab, all resolving, with no Hy’s law cases. Mortality is the open question. FENtrepid recorded 1.4% versus 0.2% fatal cases, all assessed as unrelated; the two relapsing trials recorded eight versus one, from varied causes.
The regulatory divergence
Tolebrutinib was approved in the European Union in June 2026, as Cenrifki, for secondary progressive MS without relapses in the preceding two years — the first EU licence for disability progression in that population. The UAE approved it in 2025.
The same dossier produced a complete response letter in the United States, where tolebrutinib remains unapproved. Two regulators read one trial and reached opposite conclusions about progressive MS treatment.
Anti-CD20 B-Cell Depletion
The only class with a licensed progressive indication, and the only one with replication.
OLYMPUS randomised 439 PPMS patients to rituximab and missed its primary endpoint. Prespecified subgroups suggested benefit in younger participants and those with gadolinium-enhancing lesions — an observation that shaped the ocrelizumab programme.
ORATORIO then enrolled 732 PPMS patients aged 55 or under, EDSS up to 6.5. Ocrelizumab reduced 12-week confirmed disability progression (HR 0.76, 95% CI 0.59–0.98) and slowed brain volume loss — an effect both real and small.
Why ORATORIO-HAND changes eligibility
ORATORIO-HAND is the most consequential progressive MS treatment readout of the past year, and not because of the hazard ratio. It enrolled 1,013 PPMS patients up to age 65 and EDSS 8.0 — the patients ORATORIO excluded.
Ocrelizumab reduced the 12-week composite over 144 weeks, and both components moved: EDSS worsening in 23.0% versus 30.8%, peg test progression in 16.7% versus 24.9%.
The consequence is practical: the common argument for withholding anti-CD20 therapy from an older, more disabled PPMS patient now has randomised evidence against it.
S1P Receptor Modulation
EXPAND remains the only trial to have shown superiority over placebo on disability across SPMS as a whole. Siponimod reduced 3-month confirmed disability progression in 1,651 randomised participants (HR 0.79, 95% CI 0.65–0.95).
The population was typical of established secondary progressive disease: nearly two thirds had not relapsed in two years, over half needed walking assistance.
The interpretive catch
Siponimod also reduced relapse risk substantially in EXPAND, and produced no significant effect on the timed 25-foot walk. A drug that suppresses relapses where a third of the cohort was still relapsing will reduce relapse-associated worsening. Disentangling that from progression biology has never been settled from the trial data alone.
The licensed indication in most jurisdictions reflects the ambiguity: active secondary progressive disease.
For a patient with genuinely non-relapsing, MRI-quiet SPMS, EXPAND is weaker support than its headline hazard ratio suggests. That is the gap HERCULES was built to fill.
Neuroprotection and Repurposed Agents
This class produced the field’s most instructive failure in progressive MS treatment. Two agents slowed brain atrophy in phase 2; neither slowed disability when tested properly.
The phase 2b MS-STAT trial randomised 140 SPMS patients and cut annualised whole-brain atrophy by 43% (adjusted difference −0.254% per year, 95% CI −0.422 to −0.087), launching a decade of expectation.
MS-STAT2 then randomised 964 patients with non-active SPMS, accumulating 365 progression events. Simvastatin produced no effect on 6-month confirmed disability progression, with the point estimate on the wrong side of unity.
SPRINT-MS randomised 255 patients to ibudilast and slowed brain parenchymal fraction loss (difference 0.0009 per year, 95% CI 0.00004–0.0017), with more gastrointestinal effects, headache and depression. A post-hoc analysis put the atrophy effect in the PPMS subgroup.
Brain atrophy is a reliable marker of accumulated injury and an unreliable surrogate for whether a drug changes a trajectory. Any progressive MS treatment programme reading a phase 2 atrophy signal as a green light is repeating an experiment already run twice.
Immune Reconstitution and Cell-Based Therapy
Autologous haematopoietic stem cell transplantation is the only progressive MS treatment that has produced sustained disability improvement. It is also the only one with no randomised trial confined to progressive disease.
The best comparative data come from an Italian registry: 79 aHSCT-treated patients with active SPMS, propensity-matched against 1,975 started on other anti-inflammatory DMTs.
Transplantation was associated with a longer time to confirmed disability progression, and with an EDSS trajectory that was flat rather than rising. This is a matched cohort, not a trial: the design supports association, and the authors classified it Class III.
A 2024 systematic review put a ceiling on enthusiasm: across 15 studies and 665 progressive MS patients, 5-year progression-free survival ranged from 0% to 78%.
The consistent signal is that response tracks inflammatory activity. Patients transplanted early, with demonstrable activity, do well; those transplanted late as a last resort largely do not, and carry the mortality risk regardless.
Comparative Analysis
Three observations survive when the five classes sit side by side.
First, the effect sizes converge. Every agent that worked landed in a narrow band of hazard ratios from 0.69 to 0.79 — a relative reduction of roughly a fifth to a third, six to eight percentage points in absolute terms.
Second, the dividing line is inflammation, not drug class. Every positive result sits where inflammatory activity was demonstrable; every failure where it was not.
- OLYMPUS missed overall and separated in the gadolinium-enhancing subgroup.
- EXPAND worked where a third were still relapsing, and did not move the timed walk.
- HERCULES’s post-hoc benefit concentrated in rim-lesion participants, and vanished without them.
- Simvastatin, tested deliberately in non-active SPMS, produced nothing.
Third, the phenotype label still carries information. Tolebrutinib is the cleanest evidence: HR 0.69 in non-relapsing SPMS, 1.01 in PPMS, from one drug at one dose in one programme.
One structural caveat governs this. FENtrepid is the single head-to-head trial here; every other comparison is placebo-anchored across trials run years apart. Cross-trial hazard ratios are not a ranking.
Quality and Consistency of the Evidence
Certainty differs sharply by claim. The labels below refer to propositions, not drugs.
- High — anti-CD20 slows disability in PPMS. Two independent phase 3 trials, consistent direction, over 1,700 participants. Strong evidence, modest effect.
- High — simvastatin does not work in non-active SPMS. An adequately powered phase 3, 365 events, point estimate above unity. Strong evidence of no benefit.
- Moderate — BTK inhibition slows disability in non-relapsing SPMS. Evidence suggests benefit: one positive phase 3, no replication, a regulatory split.
- Moderate — siponimod slows disability in SPMS. Evidence suggests benefit, from one phase 3 whose relapse effect confounds attribution.
- Moderate — BTK inhibition does not help PPMS. Evidence suggests no benefit: one null trial, one non-inferiority result, both unpublished.
- Low — ibudilast slows progressive MS. Preliminary evidence suggests a surrogate-endpoint effect only, no phase 3 follow-on.
- Low — aHSCT slows disability in active SPMS. Preliminary evidence suggests benefit, from Class III matched-cohort data with wide variability.
- Insufficient — remyelination changes progressive disability. No trial has tested a remyelinating agent against a disability endpoint in progressive disease.
Two structural weaknesses cut across the table. Placebo-controlled trials here span nearly two decades of shifting diagnostic criteria. The 2026 BTK meta-analysis is dominated by relapsing-MS trials, limiting what it says about progressive disease.
What the Evidence Does Not Show
Named gaps
No head-to-head in non-relapsing SPMS. Tolebrutinib has never been compared with ocrelizumab or siponimod. FENtrepid, the only comparator trial, ran in PPMS.
No validated response biomarker. The paramagnetic rim lesion analysis is post-hoc, presented at a conference, reported without confidence intervals, and never prospectively tested.
No randomised aHSCT trial in progressive MS. That evidence base is entirely observational, and the 0% to 78% spread in progression-free survival reflects selection rather than treatment.
Remyelination has no progressive-disease evidence. CCMR-Two, presented at ECTRIMS 2025, shortened visual evoked potential latency with metformin and clemastine — over six months, in relapsing-remitting patients, changing neither vision nor disability.
The hepatotoxicity question is open at class level. Whether fatal liver injury is specific to tolebrutinib’s covalent chemistry or shared across the class is unresolved, and fenebrutinib’s mortality imbalance keeps it so.
What comes next is already enrolled. FREVIVA is testing frexalimab, an anti-CD40L antibody that modulates without depleting lymphocytes, against placebo in a planned 858 participants with non-relapsing SPMS.
Practical Implications
What the evidence supports, by phenotype:
- Non-relapsing SPMS. Tolebrutinib is licensed in the EU, the first agent with randomised support here. Access is jurisdictional: US patients have no approved option.
- PPMS. Ocrelizumab remains the only approved therapy. ORATORIO-HAND supports offering it up to EDSS 8.0 and age 65.
- Active SPMS. Siponimod has randomised support. aHSCT at an experienced centre is defensible in younger patients with demonstrable activity, framed as Class III evidence carrying roughly 2% treatment-related mortality.
- Simvastatin. The question is answered. Statins in MS belong to vascular indications.
Document inflammatory activity while you can. The FDA’s central efficacy criticism of the tolebrutinib dossier was that historical MRI had not been collected. The same gap limits any individual patient’s case for treatment.
Liver monitoring is not a formality. Alanine aminotransferase elevations with tolebrutinib clustered within 90 days of starting, and severe injury has been fatal. A BTK inhibitor prescription carries a monitoring schedule to be adhered to rather than intended.
Calibrate to absolute numbers. The honest figure is roughly an eight percentage-point reduction in the chance of a confirmed progression event over two to three years, with no data beyond that horizon.
Evidence Grade and Bottom Line
Moderate for the proposition that progressive MS is modifiable at the margin, resting on replicated phase 3 data in PPMS and one positive phase 3 in non-relapsing SPMS. High for the proposition that current effects are small, and that agents targeting neurodegeneration without an inflammatory target have not worked.
- Progressive MS treatment is no longer an empty category, but every effect that held up is modest, and concentrated where inflammation is demonstrable.
- For non-relapsing SPMS, tolebrutinib is the first agent with a positive phase 3, and whether a patient can have it depends on jurisdiction rather than data.
- For PPMS, ocrelizumab remains the only approved therapy, and the eligibility ceiling has moved: older, more disabled patients now have randomised evidence behind them.
- BTK inhibition has not worked in PPMS: one trial null, the other only matching an existing drug, both unpublished.
- Neuroprotection routed through atrophy surrogates has not delivered. The next credible signal is frexalimab in non-relapsing SPMS.
Article Information and References
For Educational Purposes Only. This is an original evidence synthesis informed by the studies listed below. It does not replace clinical judgement. Drug dosages should be verified against current prescribing information.
- Raj A, Patel V, Dang M, et al. Emerging role of BTK inhibitors in multiple sclerosis: from immunobiology to clinical translation. Brain Sci. 2026;16(6):634. DOI: https://doi.org/10.3390/brainsci16060634
- Braun B, Fischbach F, Pfeffer LK, et al. Exploring the therapeutic potential of autologous hematopoietic stem cell transplantation in progressive multiple sclerosis—a systematic review. Eur J Neurol. 2024;31(12):e16427. DOI: https://doi.org/10.1111/ene.16427
- Fox RJ, Bar-Or A, Traboulsee A, et al. Tolebrutinib in nonrelapsing secondary progressive multiple sclerosis. N Engl J Med. 2025;392(19):1883–1892. DOI: https://doi.org/10.1056/NEJMoa2415988
- Giovannoni G, Airas L, Bove R, et al. Efficacy and safety of ocrelizumab in primary progressive multiple sclerosis, including older patients and those with more advanced disease (ORATORIO-HAND): a multicentre, double-blind, randomised, placebo-controlled, phase 3b study. Lancet. 2026;407(10544):2195–2207. DOI: https://doi.org/10.1016/S0140-6736(26)00617-3
- Montalban X, Hauser SL, Kappos L, et al. Ocrelizumab versus placebo in primary progressive multiple sclerosis. N Engl J Med. 2017;376(3):209–220. DOI: https://doi.org/10.1056/NEJMoa1606468
- 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):1263–1273. DOI: https://doi.org/10.1016/S0140-6736(18)30475-6
- Chataway J, Williams T, Blackstone J, et al. Effect of repurposed simvastatin on disability progression in secondary progressive multiple sclerosis (MS-STAT2): a phase 3, randomised, double-blind, placebo-controlled trial. Lancet. 2025;406(10512):1611–1624. DOI: https://doi.org/10.1016/S0140-6736(25)01039-6
- Oh J, Arnold DL, Cree BAC, et al. Tolebrutinib versus teriflunomide in relapsing multiple sclerosis. N Engl J Med. 2025;392(19):1893–1904. DOI: https://doi.org/10.1056/NEJMoa2415985
- Fox RJ, Coffey CS, Conwit R, et al. Phase 2 trial of ibudilast in progressive multiple sclerosis. N Engl J Med. 2018;379(9):846–855. DOI: https://doi.org/10.1056/NEJMoa1803583
- Boffa G, Signori A, Massacesi L, et al. Hematopoietic stem cell transplantation in people with active secondary progressive multiple sclerosis. Neurology. 2023;100(11):e1109–e1122. DOI: https://doi.org/10.1212/WNL.0000000000206750
- 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 (OLYMPUS). Ann Neurol. 2009;66(4):460–471. DOI: https://doi.org/10.1002/ana.21867
- 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. 2014;383(9936):2213–2221. DOI: https://doi.org/10.1016/S0140-6736(13)62242-4
Conference sources without a DOI. Three findings come from meeting presentations with sponsor confirmation, not yet peer reviewed, and flagged as provisional in the text. They are PERSEUS and FENtrepid (ACTRIMS Forum 2026), and the paramagnetic rim lesion analysis (ACTRIMS Forum 2025).