Network localization of hemichorea-hemiballismus

Objective: To determine whether neuroanatomically heterogeneous strokes causing hemichorea-hemiballismus localize to a common functional network. Methods: We identified 29 cases of lesion-induced hemichorea-hemiballismus from the literature and mapped each lesion volume onto a reference brain. Using a recently validated technique termed lesion network mapping, we tested whether these lesions belonged to the same functional network. To accomplish this, the network of brain regions functionally connected to each lesion was identified using a connectome dataset from healthy participants. Network maps were overlapped to identify any region functionally connected to our set of lesions. Specificity was evaluated using a case-control design; control cohorts included a group of similar lesions randomized to different brain locations and a second group of lesions causing a separate movement disorder, asterixis. Reproducibility was evaluated using an independent cohort of 10 additional hemichorea-hemiballismus cases. Results: Lesions showed heterogeneity in anatomical location, consistent with prior reports. However, at least 90% of these lesions showed network overlap in the posterolateral putamen. This result was specific to lesions causing hemichorea-hemiballismus and reproducible in an independent cohort. The putaminal overlap site was itself connected to a broader motor network that predicted the distribution of lesions causing hemichorea-hemiballismus. Conclusions: Strokes causing hemichorea-hemiballismus, while anatomically heterogeneous, localize to a common functional network. Specifically, lesions occur in regions functionally connected to the posterolateral putamen, a region previously implicated in hyperkinetic movement disorders. Lesion network mapping may be useful in identifying the neuroanatomical substrates of heterogeneous lesion-based disorders.