Effect and Reporting Bias of RhoA/ROCK-Blockade Intervention on Locomotor Recovery After Spinal Cord Injury: A Systematic Review and Meta-analysis.
Watzlawick R, Sena ES, Dirnagl U, Brommer B, Kopp MA, Macleod MR, Howells DW, Schwab JM
IMPORTANCE Blockade of small GTPase-RhoA signaling pathway is considered a candidate translational strategy to improve functional outcome after spinal cord injury (SCI) in humans. Pooling preclinical evidence by orthodox meta-analysis is confounded by missing data (publication bias). OBJECTIVE To conduct a systematic review and meta-analysis of RhoA/Rho-associated coiled-coil containing protein kinase (ROCK) blocking approaches to (1) analyze the impact of bias that may lead to inflated effect sizes and (2) determine the normalized effect size of functional locomotor recovery after experimental thoracic SCI. EVIDENCE REVIEW We conducted a systematic search of PubMed, EMBASE, and Web of Science and hand searched related references. Studies were selected if they reported the effect of RhoA/ROCK inhibitors (C3-exoenzmye, fasudil, Y-27632, ibuprofen, siRhoA, and p21) in experimental spinal cord hemisection, contusion, or transection on locomotor recovery measured by the Basso, Beattie, and Bresnahan score or the Basso Mouse Scale for Locomotion. Two investigators independently assessed the identified studies. Details of individual study characteristics from each publication were extracted and effect sizes pooled using a random effects model. We assessed risk for bias using a 9-point-item quality checklist and calculated publication bias with Egger regression and the trim and fill method. A stratified meta-analysis was used to assess the impact of study characteristics on locomotor recovery. FINDINGS Thirty studies (725 animals) were identified. RhoA/ROCK inhibition was found to improve locomotor outcome by 21% (95% CI, 16.0-26.6). Assessment of publication bias by the trim and fill method suggested that 30% of experiments remain unpublished. Inclusion of these theoretical missing studies suggested a 27% overestimation of efficacy, reducing the overall efficacy to a 15% improvement in locomotor recovery. Low study quality was associated with larger estimates of neurobehavioral outcome. CONCLUSIONS AND RELEVANCE Taking into account publication bias, RhoA/ROCK inhibition improves functional outcome in experimental SCI by 15%. This is a plausible strategy for the pharmacological augmentation of neurorehabilitation after human SCI. These findings support the necessity of a systematic analysis to identify preclinical bias before embarking on a clinical trial.