Motor and sensory neurophysiology in relation to [18F]FDG PET imaging in children with dystonia.

Abstract

The network model of dystonia reconciles many of the neuroanatomical and electrophysiological abnormalities identified as potential pathophysiological factors. Recent [¹⁸F]fluorodeoxyglucose-PET brain imaging findings support this concept, revealing distinct patterns of abnormal brain metabolism in specific dystonia aetiologies. However, it is unclear how changes in specific neural pathways alter brain glucose metabolism. This observational study investigates patterns of brain metabolism using [¹⁸F]fluorodeoxyglucose-PET imaging in children with dystonia, awake-resting during the uptake period, in relation to measures of brain function using standard neurophysiological tests of motor and sensory pathway integrity. Central motor conduction times, somatosensory evoked potentials and [¹⁸F] fluorodeoxyglucose-PET scans were obtained in children with dystonia or dystonic-dyskinetic cerebral palsy undergoing standard clinical assessment for bilateral pallidal deep brain stimulation between 2007 and 2018 in Evelina London Children's Hospital. Data from 109 children aged 2.8-18.8 years were analysed retrospectively. Patients were divided into groups based on their neurophysiology results as follows: both tests normal (NN; 67), both abnormal (AA; 11), normal central motor conduction times/abnormal somatosensory evoked potentials (NA; 20), abnormal central motor conduction times/normal somatosensory evoked potentials (AN; 11). Groups were compared with a control group comprising [¹⁸F]fluorodeoxyglucose-PET scans from 39 healthy adults using Statistical Parametric Mapping 12 with age and groupwise global means as covariates. Taking into account groupwise global uptake, all four groups shared relative hypermetabolism in parietal areas, postcentral and precentral gyri. In addition, mild peri-insular hypometabolism was seen in the NN group. The NA group showed marked regional hypometabolism bilaterally in the thalami, globi pallidi, putamina, heads of the caudate nuclei and areas of peri-sylvian cortex. The AN group had hypometabolism in the thalami and posterior globi pallidi, the posterior putamina and areas of peri-sylvian cortex. The AA group also exhibited hypometabolism in the medial thalami and some areas of frontal and peri-insular cortex. Across the whole cohort, abnormal somatosensory evoked potentials were strongly associated with thalamic hypometabolism, with no marked differences for abnormal central motor conduction times. Brain metabolism patterns in dystonia relate to neurophysiological abnormalities in our study. Relative parietal hypermetabolism is more common than recognized previously, while thalamic hypometabolism is prominent in those with abnormal sensory pathway function. The findings support the network model of dystonia and emphasize the importance of multi-modal assessment in providing detailed phenotyping, which could inform individualized management strategies.