A meta-analysis of functional outcomes in rat sciatic nerve injury models

Introduction Rat sciatic nerve injury (PNR) is the most utilized model in studies on peripheral nerve regeneration. However, large animal models are increasingly favored based on the assumption that nerve regeneration in rodents achieves more favorable outcomes than in humans. The purpose of this meta-analysis was to investigate which rat PNR models are more stringent and should be used before utilizing large animal experimentation. Methods A PRISMA-guided meta-analysis of the English literature regarding functional outcomes in rat peripheral nerve injury models was conducted. Outcomes of five basic scenarios: (1) transected nerve/negative control, (2) transection with primary microsurgical repair, (3) isogenic/autologous grafts, (4) acellular-allogenic grafts, and (5) limb transplantation were compared to sciatic nerves without any intervention/positive control. Outcomes were compared using Sciatic Functional Index (SFI). Log-based projections were generated and evaluated using mean squared error (MSE), one-way-ANOVA, and Tukey-HSD post-hoc analysis. Results In total, 167 articles met the inclusion criteria. The earliest manifestations of motor recovery were encountered in the transection and primary repair group (p <.0005). There was a significant difference in recovery time and degree of recovery between all surgical models (p <.0005). At 24 weeks, the SFI in hindlimb transplantation group was significantly worse than all other groups (-74.07 +/- 2.74, p <.0005). Autografts smaller than 10 mm recovered sooner than autografts longer than 10 mm (p = .021) and autografts recovered faster than allografts. Conclusion This meta-analysis does not support the belief that neuro-regeneration is exceptional in transection models. These models remain adequate to provide translatable information and should initially be used in investigational studies.