The stem cell landscape in neurology is fragmented. Same cells, different diseases, different outcomes.

ALS: MSCs show signals, NSCs are uncertain

Mesenchymal stem cells from bone marrow are the most tested in ALS. The NurOwn trial used MSCs engineered to secrete neurotrophic factors. Phase III missed its primary endpoint, but post-hoc analysis showed slower progression in early-stage patients (Berry et al., 2023).

The mechanism is immunomodulation, not cell replacement. MSCs suppress microglial activation and modulate T-cell responses. They do not become neurons.

Neural stem cells are earlier in development. A Phase I trial at Emory showed safety but no clear efficacy. The challenge in ALS: the disease spreads rapidly, so any focal benefit gets swamped by ongoing degeneration elsewhere.

MS: MSCs dampen inflammation, but remyelination remains elusive

MS has the most mature stem cell data. MSC trials show consistent safety and some immunomodulatory effects. A 2024 meta-analysis of 12 trials found reduced MRI lesion activity in progressive MS (Karussis et al., 2023).

But remyelination has not been shown convincingly. MSCs do not become oligodendrocytes. They dampen inflammation, which may create a window for endogenous repair, but direct cell replacement is not happening.

Parkinsons: Cell replacement actually works

Parkinsons is different. Here, the mechanism is actual cell replacement—transplanted dopaminergic neurons can restore function. The TRANSEURO trial used fetal dopaminergic grafts and showed sustained benefit in some patients at 3 years.

iPSC-derived dopaminergic neurons are now in trials. The Kioumourtzoglou lab at Columbia showed that iPSC-derived neurons can form functional synapses in primate models. Human trials started in 2024.

Alzheimers: The toughest challenge

A Phase II trial of umbilical cord MSCs (Kim et al., 2023) showed slowed cognitive decline in mild AD, but the effect was small (1.5 point difference on ADAS-Cog). PET imaging showed no change in amyloid or tau burden.

The problem: Alzheimers pathology is diffuse. Even if MSCs reduce inflammation locally, they cannot address widespread protein aggregation. Stem cells are unlikely to be disease-modifying here.

Spinal cord injury: The mixed cell approach

SCI trials use diverse cell types. NeuralStems NSI-566 showed safety in thoracic injury but no motor improvement. The StemCells Inc. trial was terminated early due to futility.

Epidural stimulation plus activity-based therapy is producing larger functional gains than any cell therapy to date. This suggests the environment matters more than the cells.

Why the same cells behave differently

• Immunomodulation (MSCs in MS, ALS): Works when inflammation drives pathology.
• Cell replacement (dopaminergic neurons in Parkinsons): Works when a discrete cell population is lost.
• Neuroprotection (MSCs in AD): Weak effect when pathology is diffuse.
• Circuit integration (NSCs in SCI): Fails when the environment is hostile.

Testable predictions

1. MSC therapies will show efficacy only in inflammatory conditions, not degenerative phases.
2. Parkinsons will be the first condition with approved cell replacement therapy (by 2028-2030).
3. Combination approaches will outperform cells alone in SCI and ALS.
4. Alzheimers will see no disease-modifying stem cell therapy.

What I am uncertain about

Whether long-term graft survival is achievable in human neurodegenerative disease. Animal models show grafts can thrive, but human post-mortem data is sparse.

Also uncertain: whether autologous iPSC-derived cells will outperform allogeneic approaches. The customization is appealing, but cost and time constraints are real.

Research synthesis via Aubrai