Emerging understanding of brain plasticity has opened new avenues for the treatment of stroke. The promising preclinical evidence with neuroprotective drugs has not been confirmed in clinical trials, thus nowadays, researchers, pharmaceutical companies, and funding bodies hesitate to initiate these expensive trials with restorative therapies. Since many of the previous failures can be traced to low study quality, a number of guidelines such as STAIR and STEPS were introduced to rectify these shortcomings. However, these guidelines stem from the study design for neuroprotective drugs and one may question whether they are appropriate for restorative approaches, which rely heavily on behavioral testing. Most of the recovery studies conducted in stroke patients have been small-scale, proof-of-concept trials. Consequently, the overall effect sizes of pooled phase II trials have proved unreliable and unstable in most meta-analyses. Although the methodological quality of trials in humans is improving, most studies still suffer from methodological flaws and do not meet even the minimum of evidence-based standards for reporting randomized controlled trials. The power problem of most phase II trials is mostly attributable to a lack of proper stratification with robust prognostic factors at baseline as well as the incorrect assumption that all patients will exhibit the same proportional amount of spontaneous neurological recovery poststroke. In addition, most trials suffer from insufficient treatment contrasts between the experimental and control arm and the outcomes have not been sufficiently responsive to detect small but clinically relevant changes in neurological impairments and activities. This narrative review describes the main factors that bias recovery studies, both in experimental animals and stroke patients.
Keywords: Functional recovery; Neuroprotection; Plasticity; Restorative therapies; Stroke; Translational research.