Protecting nerves and repairing myelin

Two different aims

Protecting nerves means keeping nerve fibres and nerve cells alive when they are under stress, whatever the cause. Repairing myelin means helping the brain rebuild the sheath that has been lost. They overlap, because a fibre without myelin is more vulnerable and new myelin helps it survive, but they are tested in different ways.

Repair is the more hopeful aim, because the brain already does it. Oligodendrocyte precursor cells are present throughout the adult brain, and many early lesions are partly remyelinated (lesson 3.1). The problem is that repair becomes slower and less complete with age and with each new attack. Most repair medicines try to wake these precursor cells up or help them finish the job.

How repair is measured

Disability changes too slowly to show repair in a trial of a few months, so repair trials use measures of myelin itself:

• The speed of the signal from the eye to the back of the brain, measured as the latency of visual evoked potentials (lesson 7.3). New myelin should make the signal faster again.
• MRI methods that are sensitive to myelin, such as magnetisation transfer ratio.
• The thickness of the retina, measured by optical coherence tomography, as a sign of how many fibres survive.

Optic neuritis is often chosen as the setting, because the damage is in one known place, starts on a known day and can be measured precisely.

Clemastine: an old antihistamine, and a warning

In 2014 researchers grew precursor cells around tiny pillars that imitate nerve fibres and tested about 1,000 approved medicines to see which ones helped the cells wrap them in myelin. Several antihistamines and related medicines did, and clemastine, an antihistamine sold for allergies since the 1960s, was among the strongest.^[1]

In ReBUILD, 50 people with MS and long-standing damage to the optic nerve took clemastine and placebo in turn. The signal from the eye became 1.7 milliseconds faster on clemastine. It was small, but it was the first evidence in people that a medicine could repair myelin.^[2]

The warning came in 2024. In TRAP-MS, a trial in progressive MS at the US National Institutes of Health, three of nine people taking clemastine worsened about five times faster than before, and that arm of the trial was stopped. Studies of their spinal fluid suggested that clemastine, in the conditions of a damaged brain, can push cells into an inflammatory form of cell death called pyroptosis.^[3] A medicine that helps repair in one setting may harm in another, and repair medicines need the same care over safety as any other.

Opicinumab and bexarotene

Opicinumab is an antibody against LINGO-1, a protein that holds back precursor cells. It worked well in animals, but in people with relapsing MS it did not improve the combined measure of disability in the SYNERGY trial, and a further trial, AFFINITY, was stopped in 2020 for lack of effect.^[4]

Bexarotene, a cancer medicine that acts on a receptor that drives myelin formation in animals, was tested in CCMR One in Cambridge. It did not meet its main MRI goal, although the signal from the eye became slightly faster, and its side effects, including an underactive thyroid and raised blood fats, rule it out for long-term use.^[5]

Metformin, and making old cells young

Precursor cells lose their ability to respond as the brain ages. In 2019 researchers showed in old rats that fasting, or the diabetes medicine metformin, restored the young cells' ability to make myelin.^[6] In CCMR Two, 70 people with relapsing MS took metformin together with clemastine, or placebo, for six months alongside their usual treatment. The signal from the eye slowed on placebo and held steady on the combination. The result, presented in 2025, is the first of its kind for a combination, and longer trials are needed to show whether it helps people's function.^[7]

Protecting nerve fibres

Several medicines approved for other diseases have been tested for protection:

• Phenytoin, an epilepsy medicine that blocks sodium channels, was given to 86 people within two weeks of optic neuritis. Six months later they had lost about 30 per cent less of the nerve fibre layer of the retina than those on placebo.^[8] Blocking sodium channels may reduce the energy crisis of damaged fibres (lesson 1.2).
• Simvastatin, a cholesterol medicine, slowed brain shrinkage by 43 per cent in the phase 2 MS-STAT trial of 140 people with secondary progressive MS.^[9] The phase 3 trial, MS-STAT2, with almost 1,000 people, found no effect on disability worsening.^[10]
• High-dose biotin, a vitamin thought to help nerve cells make energy and myelin, seemed to help in a first trial, but the larger MS-SPI2 trial of 642 people found no benefit.^[11]
• Amiloride, fluoxetine and riluzole, tested together against placebo in the MS-SMART trial of 445 people with secondary progressive MS, did not slow brain shrinkage.^[12]
• Ibudilast, an anti-inflammatory medicine used in Japan for asthma, slowed brain shrinkage by about half compared with placebo in the SPRINT-MS trial of 255 people with progressive MS. Its effect on disability has not yet been tested in a larger trial.^[13]
• Lipoic acid, an antioxidant sold as a supplement, slowed brain shrinkage in a small trial in 2017. A larger trial of 115 people with progressive MS, published in 2025, found that it did not slow the decline in walking speed, although brain volume held steadier than on placebo.^[14]

What these trials have taught

The pattern repeats:

• A medicine helps in the laboratory or slows brain shrinkage in phase 2.
• It then fails to change disability in phase 3.

Brain shrinkage measured over one or two years has turned out to be a weak guide to disability measured later. This has pushed researchers towards better early measures, such as neurofilament light in the blood, slowly expanding lesions and the visual pathway, and towards platform trials that can test several medicines at once and drop those that do not work.

The experience with clemastine adds a second lesson: repair is not automatically safe. The brain's environment in progressive MS, with its smouldering inflammation, may turn a helpful signal into a harmful one. This is one reason many researchers think that repair will work best together with a medicine that controls inflammation, and in people whose lesions are still young.