Therapeutic Potential of Stem Cells in Pediatric Neurology: Insights From Clinical Trials

Major Points and Findings:

In this review, two authors from Lebanon collected the published clinical trials of stem cell therapy across the whole of paediatric neurology and appraised them one by one. It is included on the SMA page because it shows how little clinical evidence exists for cell therapy in spinal muscular atrophy specifically. The review covers autism, cerebral palsy, epilepsy, brain and spinal cord injury, muscular dystrophy and the lysosomal storage diseases. Within it, the entire clinical evidence base for SMA is a single study of five infants.

By comparison, the trial tables for nusinersen, risdiplam and gene therapy run to thousands of patients. The review does not overstate the case for cell therapy.

Aim:

To summarise which stem cell therapies have been tested in children with neurological disorders, what happened, and how reliable those results are. The authors deliberately included studies of low methodological quality in order to show early feasibility signals, and they appraise the weaknesses of each one.

Methods:

A narrative review with structured appraisal tables. For each disorder the authors list every clinical study they identified, with sample size, age, cell type, cell origin (autologous or allogeneic), route, dose schedule and outcome, followed by a paragraph on the risk of bias. Cell types covered include mesenchymal stem cells from bone marrow, adipose tissue and umbilical cord, umbilical cord blood and cord blood mononuclear cells, haematopoietic stem cells, neural stem and progenitor cells, and bone marrow mononuclear cells. Because this is a narrative review, there is no formal search protocol, no pooled analysis and no quantitative risk-of-bias score.

Results:

Spinal muscular atrophy: One clinical study is listed, by Mohseni and colleagues (2022). Five infants with SMA type 1, aged 2 to 10 months, received three intrathecal doses of allogeneic side population adipose-derived mesenchymal stem cells, with five untreated infants for comparison. The reviewers summarise the findings as follows:

• There was a significant increase in the motor response amplitude of the tibial nerve on electromyography after the third injection, which the reviewers read as a possible neurotrophic effect.
• Only one of the five treated infants survived the study period. The reviewers note that untreated SMA type 1 has a median life expectancy under two years, so the deaths cannot be attributed to the treatment. They also state that the intervention can be described as feasible and safe but not as effective.
• Their appraisal lists an extremely small sample, no randomisation, no control group beyond observation, no power calculation, and survival outcomes confounded by disease severity and the timing of treatment. The findings “inform feasibility and safety rather than efficacy”.

On the laboratory side they note that embryonic stem cells can be turned into motor neurons with modest gains in mouse survival, and that genetically modified induced pluripotent stem cells extended lifespan in treated mice by up to 40%, but they add that these results “remain modest compared to the transformative effects of gene therapies”.

Other conditions: In cerebral palsy the evidence is much larger and includes several randomised controlled trials of umbilical cord blood and umbilical cord mesenchymal stem cells, with reported gains in gross motor function at 6 to 12 months. In autism the results are mixed. An open-label cord blood trial reported improvements, a placebo-controlled crossover trial found none, and a later trial found inconsistent results. In drug-resistant epilepsy there are two small uncontrolled studies. In severe paediatric traumatic brain injury, one randomised trial of autologous bone marrow mononuclear cells reported 50.9% less white matter volume loss at one year against controls.

Safety: Most adverse events across the field were mild. The authors flag ectopic tissue growth and tumour formation as concerns for some cell types, and graft-versus-host disease for allogeneic haematopoietic transplants. Allogeneic procedures require monitoring, with fever a recognised event. Cells from umbilical cord blood carry a lower risk of graft-versus-host disease because they are immunologically immature, which is why a perfect tissue match is not strictly required for cord sources. Higher mesenchymal cell doses are sometimes associated with better outcomes, but dose escalation can raise the risk of embolic events. Transplanted mesenchymal stem cells generally persist for only 2 to 3 months. The authors give this as the reason single doses may not give durable benefit and repeat dosing continues to be tried.

Conclusions:

Across paediatric neurology the results vary by disease, cell type, timing, dose and route, and a one-size-fits-all cell therapy is unlikely to work. Age and disease stage affect the outcome. Results were consistently better when treatment was given early, before irreversible loss of neurons. The authors’ recommendations include proper dose-response and biodistribution studies before larger clinical use, standardised manufacturing to Good Manufacturing Practice with validated potency assays, multicentre placebo-controlled trials with prespecified endpoints, central registries, and a minimum follow-up of five years to catch late adverse events and to see whether any benefit lasts.

For SMA their conclusion is a narrow one. Rigorously controlled future trials are needed, and the existing data address feasibility and safety in a handful of infants. They do not address efficacy.

Background Information:

Two points from the review are relevant to any discussion of cell therapy for SMA.

The first is the proposed mechanism. Mesenchymal stem cells are not expected to replace lost motor neurons. Their proposed value is immunomodulation and the secretion of neurotrophic factors, delivered through what the authors call the secretome. The same reasoning underlies the interest in cell-free approaches. The review points out that extracellular vesicles and the secretome may offer a safer route that keeps the paracrine benefit while reducing the risk of tumour formation.

The second is timing. The clearest pattern across the whole review is that early intervention works better, and late intervention often does not work at all. In Krabbe disease, cord blood transplant before symptoms changed the disease course, while the same transplant after symptoms began produced minimal neurological improvement. The same logic applies to SMA, where motor neuron loss is irreversible and starts early. It is also the main argument for newborn screening and for starting approved disease-modifying treatment as soon as the diagnosis is made.