Meta-analysis of the effects of clinically-effective therapeutics in the preclinical migraine model as a tool for design optimisation.

Abstract

The low reliability of the preclinical studys findings is of critical concern. The possible sources include poor experimental design and a lack of measures to reduce the risk of bias. In this study, we focused on anti-migraine drug discovery and a particular animal model with the aim to contribute to the elimination of these sources in future research. We performed a systematic search of controlled studies testing established migraine treatments in the model of trigeminovascular nociception (EMTVN) and meta-analysis for the main outcomes to estimate the overall effect sizes. In 13 studies reporting on 21 experiments, anti-migraine drugs significantly decreased trigeminovascular nociceptive traffic compared with a control intervention. Considering these effects biologically relevant, we used them in sample size calculation for future experiments. To refine the EMTVN and inform its users, we explored the impact of methodological features on the outcome and revealed several factors potentially impacting the results obtained in this model. We also assessed the internal validity of the included studies and found that the selection bias, particularly, the lack of randomisation, is likely a main source of bias. Based on our findings, we discuss the translational potential of the EMTVN and suggest what should be addressed for its improvement. We believe that this work highlights the importance of systematic reviews and meta-analyses as tools for design optimisation in animal research.