Volumetric Analysis of the Basal Ganglia and Cerebellar Structures in Patients with Phelan-McDermid Syndrome

Abstract

Objective: Phelan-McDermid syndrome is caused by haploinsufficiency of SHANK3 on terminal chromosome 22. Knowledge about altered neuroanatomic circuitry in Phelan-McDermid syndrome comes from mouse models showing striatal hypertrophy in the basal ganglia, and from humans with evidence of cerebellar atrophy. To date, no studies have performed volumetric analysis on Phelan-McDermid syndrome patients.

Methods: We performed volumetric analysis of baseline brain MRIs of Phelan-McDermid syndrome patients (ages three to 21 years) enrolled in a prospective natural history study (ClinicalTrials.gov NCT02461420). Using MRI segmentations carried out with PSTAPLE algorithm, we measured relative volumes (volume of the structure divided by the volume of the brain parenchyma) of basal ganglia and cerebellar structures. We compared these measurements to those of age- and sex-matched healthy controls part of another study. Among the patients, we performed linear regression of each relative volume using Repetitive Behavior Scale-Revised total score and Aberrant Behavior Checklist stereotypy score. Eleven patients with Phelan-McDermid syndrome (six females, five males) and 11 healthy controls were in this analysis.

Results: At time of MRI, the mean age of the patients and controls was 9.24 (5.29) years and 9.00 (4.49) years, respectively (P = 0.66). Compared to controls, patients had decreased caudate (P ≤ 0.013), putamen (P ≤ 0.026), and left pallidum (P = 0.033) relative volumes. Relative volume of cerebellar vermal lobules I to V (beta coefficient = -17119, P = 0.017) decreased with increasing Repetitive Behavior Scale-Revised total score.

Conclusions: The volumes of the striatum and left pallidum are decreased in individuals with Phelan-McDermid syndrome. Cerebellar vermis volume may predict repetitive behavior severity in Phelan-McDermid syndrome. These findings warrant further investigation in larger samples.

Keywords: 22q13.3 Deletion; Autism; MRI; Repetitive behaviors; SHANK3.

Conflict of interest statement

CONFLICTS OF INTEREST

None.

Copyright © 2019. Published by Elsevier Inc.

Figures

Figure 1.

Illustration of multi-template parcellation. All the templates are first non-linearly aligned to a target patient. A consensus map is then computed for each patient, providing a fully-automatic, robust to inter-individual variability parcellation of each patient’s brain.

Figure 2.

Scatter plots and fitted lines of RBS-R total score versus relative volumes of cerebellar vermal lobules I-V showing an inverse relationship between the relative volume and RBS-R total score.