Association of Dilated Perivascular Spaces and Disease Severity in Patients With Huntington Disease

Abstract

Objective: To quantify the percent volume of dilated perivascular space (PVS) in the subcortical forebrain in patients with early Huntington disease (HD) and to explore the relationship between PVS and disease severity.

Methods: MRI scans were performed on 25 patients with HD and 23 healthy age-matched controls at Massachusetts General Hospital. The imaging data were analyzed with a novel algorithm to determine regional PVS volume. A fractional logistic regression analysis was used to quantify the association between regional percent PVS volume and (1) disease designation (HD or control) and (2) disease severity as assessed by normalized caudate volume.

Results: Patients with HD had the greatest percent volume of dilated PVS in the putamen (left putamen: odds ratio 2.06 [95% confidence interval (CI) 1.62-2.62], HD 3.27% [95% CI 2.83-3.78] vs controls 1.62% [95% CI 1.32-1.97], p _fdr < 0.001; right putamen: odds ratio 1.66 [95% CI 1.33-2.08], HD 3.43% [95% CI 2.94-4.01] vs controls 2.09% [95% CI 1.79-2.45], p _fdr < 0.001) and several subcortical white matter regions compared to controls. Dilated PVS increased with disease severity.

Conclusions: The objective quantification of dilated PVS suggests that PVS burden is high, is associated with disease severity, and may affect the distribution and success of treatments administered either intrathecally such as antisense oligonucleotides or by intraparenchymal administration such as cell and gene therapies.

Classification of evidence: This study provides Class II evidence that increased dilated PVS is associated with worse HD severity. The study is rated Class II because of the cross-sectional design.

Figure 1. Quantification of Dilated PVS

(A) Schematic diagram showing the algorithm used to derive the volume of dilated perivascular space (PVS). The procedure included (1) bias correction, (2) spatial registration of T1-weighted (T1-w) and T21-weighted (T2-w) images, (3) mask definition of white matter (WM) and ventricles, (4) intensity standardization, (5) definition and binary mask generation of PVS, (6) registration of the PVS binary mask to FreeSurfer space, and (7) extraction of PVS volume in basal ganglia and WM. (B) Distribution of PVS extracted with our automated algorithm in the basal ganglia and subcortical WM overlaid on the averaged high-resolution T1-w images in the group with Huntington disease (HD) and control the group. T1-w images of the HD group show significant atrophy of the caudate relative to control group. Color in each voxel represents the percentage of patients in the group having dilated PVS in that particular voxel. Increased numbers of patients in the HD group have dilated PVS in the putamen, temporal, and parietal WM regions.

Figure 2. Regional Analysis of Dilated PVS

(A) Regional brain maps showing areas with significant differences in percent perivascular space (pPVS) between the group with Huntington disease (HD) and the control group after adjustment for age, corrected at p_fdr < 0.05. Odds ratios >1 in the maps represent a meaningful increase in pPVS in the HD group and vice versa when odds ratio is <1. (B) Regional brain maps showing areas with significant correlations between pPVS and normalized caudate volume in patients with HD, corrected at p_fdr < 0.05. Odds ratios >1 in the maps represent an increase in pPVS for each 0.01-unit decrease in normalized caudate volume and vice versa when odds ratio is <1.