Editorial illustration of acute ischemic stroke imaging, IV infusion, and parallel placebo-controlled trial design. Credit: CellXperience generated editorial illustration.
Two new ClinicalTrials.gov records make acute ischemic stroke a good place to watch MSC development this month.
Neither trial has posted outcome data yet. What they do show is how sponsors are trying to test IV MSC products in a disease setting where recovery can be noisy and interpretation can drift quickly.
Stroke is one of the hardest places to read an early cell-therapy program. Baseline severity matters. Reperfusion status matters. Timing matters. Rehabilitation matters. Spontaneous recovery can be real and clinically meaningful. Small single-arm studies can make a treatment effect look cleaner than it is.
That is why these two records are worth reading together.
Two records, two products
The first record, NCT07628933, is sponsored by BOE Technology Group. It is a not-yet-recruiting Phase 1 study of an umbilical-cord mesenchymal stromal cell injection for acute ischemic stroke. ClinicalTrials.gov first posted the record on June 5, 2026. The estimated enrollment is 35.
The design has two parts. Phase Ia is a single-dose, dose-escalation study with low, medium, and high IV dose groups: 5.0 x 10^7 cells, 1.0 x 10^8 cells, and 2.0 x 10^8 cells. The protocol allows escalation only after a 28-day safety assessment in the prior group. Phase Ib then selects two dose groups based on the Phase Ia results and moves to repeated dosing on Day 0, Day 7, and Day 14.
That structure is more informative than a single "MSC in stroke" label. It separates first-dose safety from repeated-dose feasibility and keeps placebo present across both phases.
The second record, NCT07635758, is sponsored by Chinese PLA General Hospital. It is a recruiting Phase 1/2 study of human placenta-derived 3D mesenchymal stem cells, listed as Guojiangqingke, in acute ischemic stroke. ClinicalTrials.gov first posted the record on June 9, 2026. The estimated enrollment is 24.
This study also uses randomized parallel assignment and quadruple masking. The registry describes Phase I as a single-dose escalation stage to identify the maximum tolerated dose and recommended Phase II dose. Phase IIa then explores preliminary efficacy. The product is administered by IV infusion, and follow-up extends up to 720 days.
The two records are related, but they are asking different questions. One is an umbilical-cord MSC program with a single-dose-to-repeated-dose Phase 1 structure. The other is a placenta-derived 3D MSC program with a Phase I/IIa structure and a Day 90 functional primary endpoint.
The stroke setting makes placebo control central
The shared feature is masking and placebo control.
In the BOE study, the registry lists randomized allocation, parallel assignment, and quadruple masking of participants, care providers, investigators, and outcome assessors. The placebo is described as a cell-free product with packaging, storage conditions, expiration date, and administration method kept consistent with the investigational product.
That matters in stroke. If patients, clinicians, or assessors can infer assignment, functional outcomes can become harder to interpret. Placebo control is not just a formality here; it is part of how the study protects the endpoint from expectation, differential care, and biased assessment.
The PLA General Hospital study uses the same broad masking structure: participant, care provider, investigator, and outcome assessor are masked. Its placebo comparator is listed as cell culture medium, cell-free.
Together, the two records treat IV cell delivery as a clinical intervention that needs comparator control, not a treatment whose effect can be read from before-and-after recovery alone.
The endpoints point to different questions
The BOE Phase 1 record is safety-led. The primary outcomes include dose-limiting toxicity at Day 28, adverse events and serious adverse events through two years, and all-cause mortality at Day 90, Day 180, and Day 360. Secondary outcomes include the proportion of participants with modified Rankin Scale score 0-2 at Day 30, Day 90, Day 180, and Day 360; NIHSS improvement; Barthel Index; Fugl-Meyer motor assessment; and CT/MRI findings.
That is a sensible hierarchy for an early IV MSC program. First, can the product be given at the planned doses and schedule? Then, are there functional or imaging patterns worth taking into a larger study?
The PLA General Hospital study puts the functional question closer to the front. Its primary outcome is the proportion of participants with mRS 0-2 at Day 90 after treatment. Secondary outcomes track mRS 0-2 at Day 28, Day 180, Day 270, and Day 360. The registry describes the 90-day mRS endpoint as functional independence.
That makes the study more immediately interpretable for a clinical audience, but also more exposed to the usual stroke-trial problem: small numbers and functional endpoints can be unstable. A Day 90 mRS result is clinically meaningful only if baseline severity, timing, rescue care, rehabilitation, missing data, and blinded assessment are handled cleanly.
The records are therefore asking different versions of the same question. BOE is mainly mapping safety and dose behavior, with functional outcomes behind it. The placenta-derived 3D MSC study is moving earlier toward an efficacy-style functional endpoint while still sitting in an early Phase I/IIa frame.
IV MSCs are working from a different therapeutic logic
The stroke setting also helps clarify what these programs are trying to do.
These records describe IV MSC products in acute ischemic stroke, not direct neural implantation. The therapeutic logic is more likely to sit around systemic and paracrine effects: inflammation, vascular response, tissue repair environment, and recovery biology after ischemic injury.
That distinction matters because the clinical burden is different from a brain-implant program. IV administration avoids the procedural burden of stereotactic or open CNS delivery. It also creates a different problem: the administered cells must produce a meaningful biological effect without being placed directly into the injured brain tissue.
For an IV MSC stroke program, the trial has to make the route credible. Dose, timing after stroke, repeat dosing, safety, and functional trajectory all become part of the same question.
The BOE record is particularly useful on that point because it tests both single and repeated dosing. A one-time infusion and three infusions over two weeks are not the same intervention. If a later result appears, the dosing phase will matter.
The placenta-derived 3D MSC record is useful for a different reason. The product identity is more specific than a generic MSC label, and the primary endpoint is placed at a familiar stroke time point: Day 90. That gives readers a cleaner way to follow the program as it moves.
What to watch next
For both programs, the first meaningful update would be practical: recruitment moving as planned, no obvious safety imbalance, tolerable infusion logistics, and follow-up that preserves masking and endpoint quality.
After that, the details become sharper.
For NCT07628933, watch which Phase Ia doses pass the Day 28 safety review, which two doses are carried into the multiple-dose stage, and whether repeated dosing changes the adverse-event profile. The secondary outcomes will be interesting, but they should be read in the context of a Phase 1 safety and dose-finding design.
For NCT07635758, watch the Day 90 mRS 0-2 result, the distribution of baseline stroke severity, how missing outcomes are handled, and whether the Day 180 to Day 360 results move in the same direction. A small Day 90 separation that fades, depends on imbalanced baseline severity, or appears only in one analysis window would be a different story from a consistent functional pattern across follow-up.
The bigger point is that both records are moving MSC stroke work into designs where future results can be checked against randomization, masking, placebo, dose, route, and functional outcome. That is the right direction for a field where early uncontrolled recovery reports have often been hard to interpret.
The important development this week is that two acute-stroke MSC programs are asking the question in a way that could make the answer more readable.
Key sources
- ClinicalTrials.gov NCT07628933, "A Phase I Trial of Umbilical Cord Mesenchymal Stromal Cells for Acute Ischemic Stroke" - https://clinicaltrials.gov/study/NCT07628933
- ClinicalTrials.gov NCT07635758, "Human Placenta-derived 3D Mesenchymal Stem Cells(Guojianqingke)" - https://clinicaltrials.gov/study/NCT07635758