By condition

Conditions

The honest answer changes completely depending on what your child has. Below, each condition is set out with the context that decides whether any of this is plausible, then the therapies being tried for it.

Jump to: Cerebral palsy · Autism · Newborn brain injury · Genetic epilepsies and encephalopathies

Cerebral palsy

The condition with the most paediatric trial data in this field — and the one where the gap between trial results and clinic promises is widest.

Cerebral palsy follows an injury to the developing brain that has already happened. No therapy here reverses that injury; the realistic target is to help the surviving network work better.

That is also why the evidence that matters most is still intensive, goal-directed motor therapy. Every trial on this page delivers its regenerative or stimulation arm on top of that, never instead of it.

Age and severity change what is plausible. Most positive trial results come from younger children with less severe involvement — which is rarely the group being quoted the result.

Therapy	Evidence here	Note
Stem cells	In clinical trials	Randomised trials exist; pooled results suggest small average gains on gross motor scales, larger at higher cell doses and younger ages.
Magnetic stimulation	In clinical trials	Studied as a primer for intensive hand therapy rather than as a stand-alone treatment.
Exosomes	Early research only	No controlled paediatric trial evidence. Biological rationale only.
Muse cells	Not supported by evidence	Not tested for this indication in children.
Photobiomodulation	Not supported by evidence	No controlled evidence for motor outcome in cerebral palsy.

Autism

A field where expectation runs far ahead of data, and where the outcome being measured deserves as much scrutiny as the treatment.

Autism is not a brain injury to be repaired, and treatments borrowed from injury models rest on a weaker premise from the start.

Most studies here measure parent-rated behaviour scales. Those scales move when a family is hopeful, when a child is in an intensive programme, and when the child simply gets older — which is why sham-controlled design matters so much and why uncontrolled reports should carry almost no weight.

The interventions with the strongest evidence in autism remain developmental and behavioural, alongside treatment of co-occurring problems such as sleep disturbance, epilepsy, constipation and anxiety. Those are frequently neglected while a family pursues something newer.

Therapy	Evidence here	Note
Magnetic stimulation	In clinical trials	More than a decade of controlled studies. Results inconsistent, effects modest; not established care.
Photobiomodulation	Early research only	Small sham-controlled studies report behavioural improvement. Short follow-up, differing devices.
Stem cells	Early research only	The largest randomised cord blood trial did not meet its primary endpoint.
Exosomes	Early research only	Marketed widely; controlled paediatric evidence essentially absent.
Muse cells	Not supported by evidence	No paediatric data for this indication.

Newborn brain injury

Hypoxic-ischaemic encephalopathy — the one setting where the timing of a regenerative treatment makes clear biological sense.

Injury here evolves over hours and days, so there is a genuine window in which an anti-inflammatory or protective intervention could matter. That is why most serious early-phase work in this field is done in newborns.

Therapeutic cooling is the established treatment and the comparator. Everything else is studied as an addition to it, within hours of birth, in a neonatal intensive care unit.

This has a practical consequence for families of older children: a cell therapy justified by neonatal trial data is being offered outside the window that justified it.

Therapy	Evidence here	Note
Stem cells	In clinical trials	Early-phase trials alongside cooling; feasibility and safety are the reported findings.
Muse cells	Early research only	Early-phase study in newborns; dose and safety, not outcome.
Exosomes	Early research only	Strong animal data in this exact model; human paediatric trials lacking.
Magnetic stimulation	Not supported by evidence	Not applicable in the acute newborn setting.

Genetic epilepsies and encephalopathies

Where a single gene is driving the disease, treatments that supply generic growth signals have the weakest rationale of all.

In conditions such as Dravet, CDKL5 deficiency or KCNQ2-related epilepsy, the problem is an ongoing fault in how neurons signal — not a one-off injury that has finished happening.

Infused cells or vesicles do not correct a gene. This is the group where a biological argument for cell therapy is hardest to make, and where families are nonetheless approached often.

The genuinely promising work in these conditions is gene-directed: antisense oligonucleotides, gene replacement and targeted small molecules, mostly in trials and tied to a confirmed genetic diagnosis. A precise diagnosis is therefore worth more than any of the therapies on this site.

Therapy	Evidence here	Note
Magnetic stimulation	Early research only	Low-frequency protocols explored for focal epilepsy; limited evidence, and parameters matter for safety.
Stem cells	Early research only	No controlled paediatric evidence.
Exosomes	Early research only	Preclinical only.
Photobiomodulation	Not supported by evidence	No evidence for seizure control in children.

One rule that applies to every row above

Whatever is added, the therapy programme, the medication plan and school should stay in place. The clearest sign that something has gone wrong is a family being told to pause the things that already work.