Combining Dopaminergic Facilitation with Robot-Assisted Upper Limb Therapy in Stroke Survivors: A Focused Review

Abstract

Despite aggressive conventional therapy, lasting hemiplegia persists in a large percentage of stroke survivors. The aim of this article is to critically review the rationale behind targeting multiple sites along the motor learning network by combining robotic therapy with pharmacotherapy and virtual reality-based reward learning to alleviate upper extremity impairment in stroke survivors. Methods for personalizing pharmacologic facilitation to each individual's unique biology are also reviewed. At the molecular level, treatment with levodopa was shown to induce long-term potentiation-like and practice-dependent plasticity. Clinically, trials combining conventional therapy with levodopa in stroke survivors yielded statistically significant but clinically unconvincing outcomes because of limited personalization, standardization, and reproducibility. Robotic therapy can induce neuroplasticity by delivering intensive, reproducible, and functionally meaningful interventions that are objective enough for the rigors of research. Robotic therapy also provides an apt platform for virtual reality, which boosts learning by engaging reward circuits. The future of stroke rehabilitation should target distinct molecular, synaptic, and cortical sites through personalized multimodal treatments to maximize motor recovery.

FIGURE 1

Enhancing motor learning. A, The motor learning network: Effective motor therapy requires the formation of new motor memories, which is anatomically mediated by networks that connect the dorsolateral prefrontal cortex, the striatum, and the cerebellum. These structures produce motor drive, coordinate motor drive while exerting postural control, and coordinate movements respectively. New motor memories are formed and pruned by the processes of LTP and LTD, which require dopaminergic signaling between the substantia nigra and striatal medium spiny neurons. B, Multifaceted approach to the enhancement of motor learning: Robotic therapy is an efficient medium for the delivery of intensive motor therapy and has been shown to induce primary motor cortex neuroplasticity in patients with stroke. Treatment with dopaminergics such as levodopa enhances neuroplasticity by inducing LTP and LTD in the striatum, nucleus accumbens, hippocampus and cerebellum. Virtual reality can be coupled with robotic therapy to deliver rewards during training and therefore encourage learning through the dorsolateral prefrontal cortex, orbital frontal cortex, and nucleus accumbens in the ventral striatum. These structures weigh the magnitude of rewards, process abstract rewards, and manage motivation plus reinforcement respectively.

FIGURE 2

Individualized treatment for post-stoke rehabilitation. Every stroke population can theoretically be stratified into individuals who are either deficient in or have adequate dopaminergic signaling. Dopaminergic profiling could be objectively quantified by neuroimaging (lesion type, positron emission tomography), neuropsychiatric testing (Stroop color-word test, symbol digit modalities), physical examination (Halstead finger tapping), biomarkers (CSF dihydroxyphenylalanine/dihydroxyphenylacetate), dopaminergic gene polymorphisms (DAT, DRD1, DRD2, DRD3), and neuromodulation (TMS). There will likely be a gradient of baseline dopaminergic tone, and the most deficient patients would benefit from dopaminergic treatment. Patients who are profiled as non-dopamine-responders could be treated with nondopaminergic neurotransmitter treatment. The goal is to combine pharmacotherapy with robotic therapy to maximize outcomes. Eligibility for robotic therapy should include criteria such as Fugl-Meyer upper extremity scores, the modified Ashworth scale, the ability to follow commands, and the lack of limb pain.