Abstract
Background: In recent years pre-clinical stroke research has shown increased interest in the development of biomaterial-based therapies to promote tissue repair and functional recovery. Such strategies utilize biomaterials as structural support for tissue regeneration or as delivery vehicles for therapeutic agents. While a range of biomaterials have been tested in stroke models, currently no overview is available for evaluating the benefit of these approaches. We therefore performed a systematic review and meta-analysis of studies investigating the use of biomaterials for the treatment of stroke in experimental animal models. Methods: Studies were identified by searching electronic databases (PubMed, Web of Science) and reference lists of relevant review articles. Studies reporting lesion volume and/or neurological score were included. Standardized mean difference (SMD) and 95% confidence intervals were calculated using DerSimonian and Laird random effects. Study quality and risk of bias was assessed using the CAMARADES checklist. Publication bias was visualized by funnel plots followed by trim and fill analysis of missing publications. Results: A total of 66 publications were included in the systematic review, of which 44 (86 comparisons) were assessed in the meta-analysis. Overall, biomaterial-based interventions improved both lesion volume (SMD: -2.98, 95% CI: -3.48, -2.48) and neurological score (SMD: -2.3, 95% CI: -2.85, -1.76). The median score on the CAMARADES checklist was 5.5/10 (IQR 4.25-6). Funnel plots of lesion volume and neurological score data revealed pronounced asymmetry and publication bias. Additionally, trim and fill analysis estimated 19 "missing" studies for the lesion volume outcome adjusting the effect size to -1.91 (95% CI: -2.44, -1.38). Conclusions: Biomaterials including scaffolds and particles exerted a positive effect on histological and neurological outcomes in pre-clinical stroke models. However, heterogeneity in the field, publication bias and study quality scores which may be another source of bias call for standardization of outcome measures and improved study reporting.