Abstract WMP37: Methodological Quality and Meta-Analysis of Animal Studies of Mesenchymal Stromal Cells for Ischemic Stroke

Abstract

INTRODUCTION: Mesenchymal stromal cells (MSC) are multipotent cells that support numerous restorative processes after stroke. The ease of isolation and immunoprivileged status of MSC have stimulated numerous preclinical stroke studies. We performed a meta-analysis to estimate study quality, size of behavioral effects, and the impact of variables such as timing of MSC administration in relation to stroke onset. METHODS: Studies of MSC and stroke were identified through PubMed and Web of Science. Studies of hemorrhage, not in English, or using modified MSC were excluded. A Quality Score was determined for each study, estimating methodological quality using 10 criteria derived from STAIR guidelines, with higher Quality Scores reflecting greater compliance with issues such as randomization and outcome blinding. Outcome data extracted for MSC and control arms were used to derive estimates of effect size using Cohen’s d. RESULTS: A total of 46 studies met criteria, with 39 studying rats, 6 mice, and 1 primates. There were 61 treatment groups, as some studies had >1 independent MSC treatment arms; MSC were introduced intravenously in 41, intracerebrally in 15, and intraarterially in 6. MSC source was rat in 24, human in 16, and mouse in 6. Time of MSC administration ranged from 5 weeks pre- to 1 month post-stroke. MSC dose ranged from 1x10^4 to 3.25x10^7. The median Quality Score was 6 (IQR 5-7). Quality Score was not related to time of MSC administration relative to stroke or to behavioral effect size. Median effect size was 2.05 for the Modified Neurological Severity Scale (n=23), 1.88 for Adhesive Removal Test (n=19), and 2.70 for the Rotarod Test (n=14). Effect sizes were substantial across all routes of administration and differed only for the mNSS (p<0.04), favoring the IC route. Effect size did not vary with time of MSC administration relative to stroke for any behavioral measure. CONCLUSIONS: The quality of preclinical MSC stroke studies has generally been good. MSC consistently provide very large behavioral benefits, across scales and routes of administration. The magnitude of behavioral effects was not related to the Quality Score or to the time of MSC administration relative to stroke. These findings support translation of MSC to human trials.