Motor Impairment and Adaptation in a Novel Nonhuman Primate Model of Internal Capsule Infarct.

BACKGROUND

Loss of hand control is a major source of poststroke disability, particularly when infarcts involve subcortical white matter. Yet, most preclinical models target the cortex and infer recovery from task performance without quantifying whether improvements reflect restoration or adaptation. To address this, we developed a novel nonhuman primate model of focal internal capsule infarct. This study is exploratory and hypothesis-testing, including prespecified hypotheses regarding motor impairment and adaptation.

METHODS

Three adult male rhesus macaques (Macaca mulatta, 5-10 years) received stereotactically guided ET-1 (endothelin-1) injections (0.5 µg/µL; 9-10 µL total) into the left posterior limb of the internal capsule in a within-subject pre-infarct versus post-infarct design. Infarcts were characterized by magnetic resonance imaging and postmortem histology. Motor function was tested with a center-out joystick task and a Klüver board task using standard behavioral metrics; hand posture was additionally scored by blinded raters to quantify adaptation. Pre-infarct versus post-infarct comparisons used generalized linear mixed models with per-session random effects; adaptation covariates were evaluated using likelihood-ratio tests.

RESULTS

Magnetic resonance imaging and histology confirmed focal infarcts centered on the posterior limb of the internal capsule, with variable volumes. All animals significantly reduced contralesional hand use post-infarct (Fisher exact test; all P<0.05). Generalized linear mixed models showed reduced center-out success rates in 2 animals (P<0.01) and increased path length in one (P=0.0015). Pooled generalized linear mixed models showed significant post-infarct increases in retrieval time (estimate=0.73 s; P<0.0001) and digit flexions (estimate=0.727; P<0.0001). Two animals developed a compensatory wrist-extended, digit-flexed posture by 4 weeks; augmenting generalized linear mixed models with these postural features significantly improved fit (likelihood-ratio tests; P<0.0001).

CONCLUSIONS

This model recapitulates key features of human subcortical stroke, including persistent distal motor deficits, shifts in hand preference, systematic motor adaptation, and interindividual variability, establishing a translationally oriented platform for studying stroke mechanisms and evaluating recovery-promoting interventions.