How the data is built
Where the data comes from, how it is checked and what it does not yet cover.
MS research has identified effective ways to reduce relapsing disease activity, a detailed map of genetic susceptibility, and strong evidence implicating EBV in disease onset. Preventing continuing disability and restoring lost function remain major goals.[7][6][5][8]
Two central questions are why MS starts and whether it can be prevented, and why disability can continue to accumulate despite control of relapses. The research roadmap addresses stopping disease, restoring function and prevention together.[4][8]
A proposed mechanism-driven framework describes inflammatory and neurodegenerative processes as relevant early in MS, rather than only in late disease. Disability worsening outside defined relapse windows is studied as progression independent of relapse activity (PIRA), which contributes substantially to disability accumulation even in relapsing MS. Definitions vary between studies, and PIRA is not a direct measurement of a specific biological mechanism.[2][1][4]
References
- [1]Müller et al. · Harmonizing definitions for progression independent of relapse activity (2023)
- [2]Kappos et al. · Contribution of relapse-independent progression vs relapse-associated worsening to overall confirmed disability accumulation in typical relapsing multiple sclerosis. JAMA Neurology (2020)
- [3]Vollmer et al. · Multiple sclerosis phenotypes as a continuum: the role of neurologic reserve. Neurology: Clinical Practice (2021)
- [4]Kuhlmann et al. · Multiple sclerosis progression: time for a new mechanism-driven framework (2023)
- [5]Bjørnevik et al. · Longitudinal analysis of Epstein-Barr virus and MS (2022)
- [6]International MS Genetics Consortium · Multiple sclerosis genomic map implicates peripheral immune cells and microglia (2019)
- [7]Hauser et al. B-cell depletion with rituximab in relapsing–remitting multiple sclerosis (HERMES). NEJM, 2008
- [8]Bebo et al. Pathways to cures for multiple sclerosis: a research roadmap (2022)
How hypotheses are judged
A hypothesis is Confirmed when two or more controlled phase 3 trials met their main goal, or when one did and a medicine of the same approach is approved in the EU or the US for that use. With one positive phase 3 trial and no approval it is Supported by one trial. Analyses that only compare an active comparator with placebo, such as interferon beta against placebo inside a laquinimod trial, do not count as the tested medicine’s result.
- Confirmed
- Two or more controlled phase 3 trials met their main goal, or one did and a medicine of this approach is approved for this use.
- Supported by one trial
- One controlled phase 3 trial met its main goal. It needs a second trial or regulatory approval to count as confirmed.
Lab evidence
Lab studies in animals or cells. Results in animals often do not hold in people, and these studies never change the status of an idea.
Lab studies are found in Europe PMC every week: original research since 2015 that uses an animal model of MS, such as experimental autoimmune encephalomyelitis (EAE), cuprizone, lysolecithin or Theiler's virus, or human cells and organoids. Reviews are excluded. Software reads each title and abstract and records what was tested, in which model, when treatment started, and whether it helped on the main outcome. Names of treatments and genes are kept only if they appear in the abstract, and the quoted conclusion only if it appears word for word. Whether the abstract mentions randomisation or blinding is recorded by a fixed search, because studies that report them tend to find smaller effects.[1] Treatments are matched to the Trial Bank through ChEMBL, and genes to human genetic evidence through Open Targets.
Lab evidence is shown apart from trials and never changes the status of an idea. Animal models reproduce only parts of MS. EAE, the most used, is driven by an immune attack that researchers start on purpose, and treatment is often given before the disease appears, which is not possible in people. A systematic review of 1,152 animal studies of MS treatments found that few reported measures against bias, and that those that did found smaller effects.[1]
- Vesterinen HM, et al. Improving the translational hit of experimental treatments in multiple sclerosis. Mult Scler 2010. doi:10.1177/1352458510379612
Explore the evidence
Hypotheses
Every hypothesis tested in MS, and how it has fared.
Live count unavailable
By type of MS
What researchers are working on for relapsing, progressive and early MS, and for NMOSD and MOGAD.
Trial Bank
Every registered MS trial, with its status and results.
Live count unavailable
Medicines
Approved medicines and those in development, with their trials.
Live count unavailable
Missing results
Finished MS trials whose results we could not find, a year or more after they ended.
Live count unavailable
How the layers fit
Frontiers
Eight research questions, with the trials and hypotheses behind each one.
8 frontiers
Hypotheses
Each hypothesis tested in MS, judged by fixed rules from its controlled trials.
Live count unavailable · Live count unavailable
Results
Results posted in registries, and results read by software from published papers, each with the sentence it rests on.
Live count unavailable
Trial Bank
Every MS trial in one place, with duplicates merged, phase 2 and phase 3 trials of the same medicine linked, and each trial tagged by type of MS, approach and phase.
Live count unavailable
Registries
ClinicalTrials.gov, the EU's two registers and ISRCTN. Public, but scattered, duplicated and written for regulators.
Live count unavailable
Everyone can read the registries. MSFrontier adds the layers above them, built from the same records by the same rules every week, so a correction at any level carries through to the rest.
Changes to the rules
- 1 November 2026: Supported by one trial added between Confirmed and Promising. Confirmed requires two or more controlled phase 3 trials that met their main goal, or one plus an approved medicine for the same use.
- 2 November 2026: two or more phase 3 trials that met their goal, with none against, now give a verdict even when an extension study at the same phase has no readable result.
What we do not cover yet
- Trial registries outside the US and EU (for example Australia and New Zealand, China, India and Japan) are not yet included, except for trials also registered in ClinicalTrials.gov, the EU registers or ISRCTN.
- Results that appear only at conferences, in full-text papers without an abstract, or in languages other than English may be missed.
- Trials that ended before registries existed (before about 2005) appear only if they were registered later.
- Statuses, tags and summaries are assigned by software and are not reviewed by people.
Assembled by software from public sources and not checked by people. Each record links to its source. Report an error