Enhancing spontaneous recovery after stroke: a randomized controlled trial

This Phase IIa biomarker-guided stroke rehabilitation trial aimed to determine if a 3-week programme of high-dose, high-intensity (HDHI) virtual exploratory movement (VEM) using MindPod Dolphin could improve upper limb recovery and outcomes early after stroke. Sixty-four participants were randomized into VEM (n = 31) or conventional therapy (CoT) groups (n = 33) and began intervention within 2 weeks of stroke. Participants were given 90 min of therapist time per weekday for 3 weeks to undertake intensive upper limb therapy over and above their usual customary care. Outcomes were obtained immediately post-intervention, 3 and 6 months post-stroke. The primary endpoint was change in Action Research Arm Test (ARAT) score between baseline and 3 months post-stroke. Secondary outcomes were Fugl-Meyer assessment (FM-UE), hand dexterity, reaching kinematics and transcranial magnetic stimulation-derived measures across post-intervention time points up to 6 months. Data from all participants were analysed for intention to treat (ITT), while 54 met the weekly target of active therapy minutes, permitting inclusion in a per protocol (PP) analysis. The average weekly time on task increased over the intervention period, with progressively greater distances of arm movements (VEM) or task repetitions (CoT) made each week. Still, during the final week of intervention, participants analysed PP spent an average of only 50 (VEM)-66% (CoT) of the extra time available on task. For the ITT analysis, there was no effect of group for the primary endpoint. For the PP analysis, there was no effect of group or group by time interaction for any secondary outcome measure. An expected effect of time was observed indicative of recovery from impairment (mean ΔFM at 6 months = 23 points), increased activity capacity (mean ΔARAT at 6 months = 31 points) and improved secondary neuroscientific measures. PP participants from both groups were compared with an historical cohort matched for baseline age, stroke severity and impairment who only received usual customary care. Despite a 3-fold difference in active upper limb therapy minutes there were no differences in the 3-month ARAT or ΔFM. Indeed, the recovery in FM-UE at 3 months was proportional to baseline impairment, suggesting that the gains could be largely attributable to spontaneous biological recovery. While patient-related factors limited the dose of therapy achievable in this trial, clinician and service-related factors would also need to be overcome in routine clinical practice, to deliver equivalent doses at this early phase. It may be necessary to forego HDHI therapy until the late sub-acute phase.