Efficacy and safety of stem cell therapy in ischemic stroke patients: a systematic review and meta-analysis of randomised controlled trials

Background Stem cell (SC) transplantation is a promising therapeutic approach for ischemic stroke (IS). However, the current literature lacks robust evidence substantiating its efficacy and safety. This systematic review aims to evaluate the efficacy and safety profile of SC therapy in patients who had an IS. Methods References up to November 2025 were sourced from OVID Medline, EMBASE, Scopus, Cochrane Central Register of Controlled Trials, Web of Science and ClinicalTrials.gov. Eligibility criteria followed a Population, Intervention, Comparator, Outcome framework: adult patients with image-confirmed IS, any human SC therapy (non-neural or neural), placebo or standard medical care and functional or neurological outcomes (modified Rankin Scale (mRS), Barthel Index (BI), National Institutes of Health Stroke Scale (NIHSS)), mortality or adverse events at ≥6 months. Only randomised controlled trials were considered. Random-effects meta-analyses compared efficacy and safety between SC and control groups. Study heterogeneity was measured using the I² statistic, and risk of bias was assessed using the Cochrane Risk-of-Bias V.2 tool. Results 17 RCTs involving 999 patients (SC therapy: 495; control: 504) were included. A pooled analysis showed that SC therapy was associated with significant improvements in mRS scores (mean difference (MD)=−0.27, 95% CI −0.51 to −0.03, p=0.027, I 2 =42.2%), BI scores (MD=7.78, 95% CI 0.50 to 15.06, p=0.036, I 2 =53.0%) as well as reduced mortality rates (relative risk=0.64 (95% CI 0.42 to 0.98), p=0.040, I 2 =0%). No significant effect was observed for NIHSS scores. The incidence of AE was comparable between groups. The included trials exhibited moderate heterogeneity and a moderate risk of bias. Conclusion SC therapy shows potential to improve functional outcomes and survival in patients who had an IS without significant safety concerns. However, the observed effect on disability is fragile, as statistical significance was lost in leave-one-out analyses, and the substantial heterogeneity and moderate methodological quality limit definitive conclusions. PROSPERO registration number CRD42024567397.