Rehabilitation and functional neuroimaging dose-response trajectories for clinical trials

Background. In clinical trials, behavioral outcomes and physiological measures of activity-dependent plasticity that evolve with task-oriented therapies may fail to reach statistical significance. When significant, clinical effectiveness may not be robust enough to alter professional practices. Objective. Provide the conceptual basis for a research design to optimize the effect of an experimental treatment. Methods. Literature review. Results. Research designs usually do not take into consideration the dynamic state of each subject's potential responsiveness to an intervention. Providing a rational, rather than convenient, intensity and duration of therapy may remedy this potential confounder for clinical trials. To determine whether a most effective dose of a therapy exists, investigators could assess subjects before the intervention, administer interim measures at planned intervals, and continue the intervention until the primary behavioral outcomes or functional imaging parameters or both reach a plateau for at least 15h of additional treatment. Conclusion. Promising interventions ought to be continued in phase II/III trials until subjects reach an asymptote in the primary outcome for behavioral gains. For neuroimaging studies that aim to correlate brain-behavior measures during rehabilitation, the specific intervention should also continue until behavioral gains and cerebral adaptations have attained a persistent plateau. Future trials can investigate whether functional neuroimaging performed in parallel with repeated behavioral assessments can better inform researchers about the optimal duration of an experimental therapy and a subject's maximal capacity for intervention-induced cerebral reorganization.