[Stroke Phase 3] Five-Year Safety and Functional Outcomes of Intravenous Administration of Autologous Bone Marrow-Derived Mesenchymal Stem Cells in Patients with Ischemic Stroke (STARTING Study)

Study Summary

Study Results

Patients who had suffered a severe cerebral infarction in the middle cerebral artery territory received two intravenous infusions of mesenchymal stem cells cultured from their own bone marrow. The primary endpoints were death and serious adverse events, while the secondary endpoint was the modified Rankin Scale (mRS, a score indicating dependence in daily living) at the time of the last visit.

During the follow-up period, 4 of the 16 patients in the treatment group and 21 of the 36 patients in the control group died; the cumulative survival rate at week 260 was 0.72 in the treatment group and 0.34 in the control group (P = 0.058). In functional assessments, 11 patients in the treatment group improved and 4 worsened (P = 0.051), and the proportion of patients with an mRS score of 0–3 increased only in the treatment group (P = 0.046).

No malignant tumors developed in the treatment group, and brain MRI scans 12 months after treatment showed no changes; however, one patient underwent resection of a benign tumor that developed in the ankle 18 months later. Seizures occurred in 5 patients in the control group and 3 in the treatment group; recurrent vascular events occurred in 3 patients in the control group and 4 in the treatment group; and there were no cases of arrhythmia. One patient was unable to receive the second dose due to fever during treatment.

Figure. Survival curves for the treatment and control groups over time following randomization.

Study Design

This was a randomized, open-label clinical trial conducted at a tertiary general hospital in Korea from July 2003 to December 2005. The study included patients aged 30–75 who were treated within 7 days of onset and still had severe neurological deficits on the 7th day of hospitalization. A total of 85 patients were randomized; after excluding 11 who died or were discharged within 4 weeks due to a lack of prospect for recovery, and 22 who declined to participate, 52 patients (16 in the treatment group and 36 in the control group) were included in the final analysis.

The control group received neither a bone marrow aspiration nor a placebo injection. Although the procedure was open-label after randomization, functional assessments were conducted by investigators who were blinded to patient information. Follow-up evaluations were conducted 1 week, 4–5 weeks, 7–9 weeks, and 14 weeks after hospitalization, and subsequently every 3 months.

Cells Used and Administration Method

Autologous bone marrow-derived mesenchymal stem cells obtained from the patients’ own bone marrow were used. Within one week of randomization, 5 mL of bone marrow was collected from the posterior iliac crest and cultured in a medium containing 10% fetal bovine serum until the cell count reached 1×10⁸ per patient. Manufacturing took place at Pamcell’s GMP facility, and the cell viability immediately prior to administration exceeded 95% on both occasions.

The dose was 5×10⁷ cells per administration, mixed with 100 mL of normal saline and administered intravenously. The first administration took place approximately 4 weeks after bone marrow collection (median 32.5 days), and the second administration occurred approximately 2 weeks later.

Discussion

This study reports the results of intravenous administration of autologous bone marrow-derived mesenchymal stem cells to patients with ischemic stroke, with follow-up lasting up to 5 years. Mortality was lower in the treatment group than in the control group, and the incidence of new comorbidities was similar between the two groups.

The researchers reported that functional improvement was more pronounced in patients with higher blood SDF-1α levels or less damage to the periventricular neural stem cell zone.

There are also clear limitations. Since the treatment is still in the experimental stage, the treatment group was small, consisting of only 16 patients, and enrollment was halted due to concerns regarding the use of bovine serum; only follow-up was continued. Because the study was not double-blind, the placebo effect cannot be ruled out, and the control group did not undergo bone marrow collection or receive a placebo injection. SDF-1α analysis was conducted on 9 patients in the treatment group. The research team also identified the long time interval between onset of symptoms and administration—due to the approximately 4 weeks required for cell culture—as an area for improvement. This study was conducted with support from the Ministry of Health and Welfare, among others, and the administered cells were manufactured at Pamicell’s GMP facility.