A Network Imaging Biomarker of X‐Linked Dystonia‐Parkinsonism

Abstract

ObjectiveThe purpose of this study was to characterize a metabolic brain network associated with X‐linked dystonia‐parkinsonism (XDP).MethodsThirty right‐handed Filipino men with XDP (age = 44.4 ± 8.5 years) and 30 XDP‐causing mutation negative healthy men from the same population (age = 37.4 ± 10.5 years) underwent [18F]‐fluorodeoxyglucose positron emission tomography. Scans were analyzed using spatial covariance mapping to identify a significant XDP‐related metabolic pattern (XDPRP). Patients were rated clinically at the time of imaging according to the XDP‐Movement Disorder Society of the Philippines (MDSP) scale.ResultsWe identified a significant XDPRP topography from 15 randomly selected subjects with XDP and 15 control subjects. This pattern was characterized by bilateral metabolic reductions in caudate/putamen, frontal operculum, and cingulate cortex, with relative increases in the bilateral somatosensory cortex and cerebellar vermis. Age‐corrected expression of XDPRP was significantly elevated (p < 0.0001) in XDP compared to controls in the derivation set and in the remaining 15 patients (testing set). We validated the XDPRP topography by identifying a similar pattern in the original testing set (r = 0.90, p < 0.0001; voxel‐wise correlation between both patterns). Significant correlations between XDPRP expression and clinical ratings for parkinsonism—but not dystonia—were observed in both XDP groups. Further network analysis revealed abnormalities of information transfer through the XDPRP space, with loss of normal connectivity and gain of abnormal functional connections linking network nodes with outside brain regions.InterpretationXDP is associated with a characteristic metabolic network associated with abnormal functional connectivity among the basal ganglia, thalamus, motor regions, and cerebellum. Clinical signs may relate to faulty information transfer through the network to outside brain regions. ANN NEUROL 2023;94:684–695