Because they possess immunomodulatory, anti-inflammatory, and paracrine signaling features, umbilical cord mesenchymal stromal cells (UC-MSCs) are being investigated in autism spectrum disorder (ASD). The field remains investigational, however, in 2026. A previous review found nine published early-phase MSC trials registered for ASD, as of March 2026, with ongoing issues that include trial heterogeneity, small sample sizes, little randomized data and continued lack of dose optimization. So the most defensible conclusion at this point is that UC-MSC therapy for autism should be characterized as an experimental approach under clinical investigation, not a standard treatment.
- Introduction
Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by substantial clinical heterogeneity and although mainstream care remains focused on behavioral, developmental, educational and symptom-focused medical interventions the areas of biological therapy remain couched in non-cure based modalities. In part, interest in UC-MSCs has increased because ASD has been associated with immune dysregulation, neuroinflammation and oxidative stress and altered interactions between the immune system and brain for at least some subgroups, establishing a theoretical basis for biologically active cell-based interventions. Still, theory is not clinical proof, and the current evidence base is still early.
- Why UC-MSCs Are Being Studied in Autism
MSCs are considered appealing candidates since they can modulate inflammatory signaling, microglial activation, trophic support, and tissue microenvironments via their secretome rather than through direct structural replacement And that mechanistic reasoning is one of the reasons UC-MSCs keep showing up and getting included in ASD trial pipelines. Nevertheless, the same review from 2026 argues that important translational questions have not yet been answered by the field, particularly regarding which patients might respond to cell therapy, what product attributes are most relevant for each patient phenotype, and how dose should be selected for different routes of administration and protocols.
- Current Human Evidence
Duke’s open-label phase I trial in children with ASD utilizing hCT-MSCs represents the most robust publication of a human UC-MSC case series for indicating amelioration of autism-related symptoms. In that study, children were administered a single intravenous dose (1 × 10^6 cells/kg) or two (2 × 10^6 cells/kg) at 1-month intervals. The study said infusion was feasible and that some children improved on clinical measures, but the design was small and open-label so it cannot establish efficacy.
This led to the selection of a single intravenous dose of 6 × 10^6 hCT-MSC/kg for escalation in a Duke phase II protocol, and ultimately as the chosen dose to test in the randomized trial. Evaluation of the safety and tolerability of this single 6 × 10^6/kg dose in children with ASD is included as an explicit research endpoint under ‘Study endpoints’, demonstrating that selection of this dose, while established here, is still being actively tested. This is helpful for writing articles because it gives an up-to-date, evidence-based dose without suggesting that the figure represents some traditional standard of care.
Earlier studies also report a 2013 experiment in which four weekly intravenous and intrathecal stem cell infusions using cord blood mononuclear cells only or combined with UC-MSCs were delivered separately, along with rehabilitation. Other historical studies cite UC-MSCs, but they were done with mixed products and mixing routes and therefore do not establish a clean modern UC-MSC dosage structure in autism.
One further note of caution: a 2019 paper describing multiple fixed-dose infusions of UC-MSC in children with autism was subsequently retracted, after the journal revisited concerns relating to whether participants were charged for taking part. As a result of that retraction, it’s not the best primary anchor for any efficacy claims and I’d be wary of leaning on it as your principal dosage reference.
- What Can Be Said About Dosage
The most precise scientific formulation is that no standardized dosage of UC-MSC for autismis currently established at a universal level. We are currently left with trial-specific investigational regimens, not a consensus protocol. That said, clinically the non-retracted dose signals most clear in the autism literature are: first, 2 × 10^6 cells/kg IV per infusion repeated at one- to three-month intervals in Duke’s phase I study; and second, a single dose of 6 × 10^6 cells/kg IVin the Duke phase II trial protocol. Such numbers make for good ivory tower talk, but ought to be understood as experimental micro-doses under study.
In grimoire, this journal-style sentence could be: “To date, there is no standardized human UC-MSC dose for autism spectrum disorder. Published dosing remains investigational, with early Duke studies employing 2 × 10^6 cells/kg IV per infusion in phase I and a single 6 × 106 cells/kg dose administered IV selected for evaluation in phase II.” That wording sticks to the best available evidence and avoids sounding overconfident.
Conclusion
There is a hope behind UC-MSC therapy for autism, but the research field remains scientifically intriguing yet clinically unproven. The best current interpretation is not that we have solved an autism stem cell dose, but rather that dose optimization remains active area of research. In the context of a journal article, we would recommend that UC-MSC dosing be presented as investigational in nature with citations to the non-retracted Duke phase I and phase II dosing frameworks being included along with acknowledgement that no regulatory authority has approved any regenerative medicine therapies for autism.