Efficacy of using robotics to characterize impaired upper-limb function in a non-human primate model of stroke
Non-human primate , Robotics , Stroke , Translational model
Rationale: Stroke is a leading cause of death and disability and has the largest socioeconomic burden of any disease in Canada. Unfortunately, thousands of stroke therapies that proved successful in animal preclinical trials failed in subsequent human populations. Non-human primates (NHPs) closely resemble humans and may represent an animal model that could help bridge the translational gap preceding clinical trials. Additionally, robotic technology has been proven to provide objective determination of outcomes in human stroke populations and it is possible that these outcomes are conserved across species. This study investigated the efficacy of using robotic tasks as a behavioural assessment tool in a NHP model of stroke. Methods: Stroke was induced in 2 cynomolgus macaques through transient 90-minute right middle cerebral artery occlusion. At 2.5 years post-stroke, neurobehavioural outcomes were assessed using a visually guided reaching (VGR) and a postural perturbation (PP) KINARM exoskeleton robotic task. Stroke NHP task parameters were compared to control performance (2 healthy, age matched controls) for both the affected and unaffected-arms to determine impairment. Results: In the VGR task, stroke animals made reaches with their affected-arm that were less accurate, had more corrective motions, travelled a greater distance, and took longer as compared to controls (p<0.01). In the PP task, responses of stroke animals to perturbations to the affected-arm were further displaced, took longer to stop, had more corrective motions, and took longer to return to centre as compared to controls (p<0.01). Several, specific unaffected-arm deficits were also identified for both stroke NHPs. Conclusions: The KINARM tasks were able to consistently quantify specific sensorimotor deficits in stroke NHPs in a way similar to that previously achieved in human populations. This study proves the efficacy of robotic assessment in a NHP model of stroke and supports the feasibility of this model in translating future stroke therapies from preclinical to clinical trials.