Progression of brain atrophy in PSP and CBS over 6 months and 1 year

Abstract

Objective: To examine the utility and reliability of volumetric MRI in measuring disease progression in the 4 repeat tauopathies, progressive supranuclear palsy (PSP) and corticobasal syndrome (CBS), to support clinical development of new tau-directed therapeutic agents.

Methods: Six- and 12-month changes in regional MRI volumes and PSP Rating Scale scores were examined in 55 patients with PSP and 33 patients with CBS (78% amyloid PET negative) compared to 30 normal controls from a multicenter natural history study. Longitudinal voxel-based morphometric analyses identified patterns of volume loss, and region-of-interest analyses examined rates of volume loss in brainstem (midbrain, pons, superior cerebellar peduncle), cortical, and subcortical regions based on previously validated atlases. Results were compared to those in a replication cohort of 226 patients with PSP with MRI data from the AL-108-231 clinical trial.

Results: Patients with CBS exhibited greater baseline atrophy and greater longitudinal atrophy rates in cortical and basal ganglia regions than patients with PSP; however, midbrain and pontine atrophy rates were similar. Voxel-wise analyses showed distinct patterns of regional longitudinal atrophy in each group as compared to normal controls. The midbrain/pons volumetric ratio differed between diagnoses but remained stable over time. In both patient groups, brainstem atrophy rates were correlated with disease progression measured using the PSP Rating Scale.

Conclusions: Volume loss is quantifiable over a period of 6 months in CBS and PSP. Future clinical trials may be able to combine CBS and PSP to measure therapeutic effects.

Figure 1. Cross-sectional gray and white matter voxel-based morphometry analysis at baseline

The t maps are results of 2-sample t tests at baseline with age, sex, and total intracranial volume included as covariates of no interest between normal controls and (A) patients with CBS, (B) patients with PSP, and (C) patients with PSP-DAV. All t maps are thresholded at cluster-level family-wise error–corrected p < 0.05. Slices displayed are as follows: coronal (−10, 10), sagittal (−23, 5, 23), and axial (−20, 6, 36). Minimum and maximum t scores are displayed on the color bars. CBS = corticobasal syndrome; DAV = davunetide; GM = gray matter; PSP = progressive supranuclear palsy; WM = white matter.

Figure 2. Profile plots of longitudinal clinical and volumetric change

Profile plots show percent change in (A, B) clinical scores and (C–J) volumes in select regions at baseline, 6 months, and 12 months. Red stars indicate significant difference between PSP and NC, green stars significant difference between CBS and NC, and purple stars significant difference between CBS and PSP. Mann–Whitney U tests with p < 0.05 were considered statistically significant. Error bars indicate standard error of the mean. CBS = corticobasal syndrome; DAV = davunetide; NC = normal control; PSP = progressive supranuclear palsy; PSPRS = Progressive Supranuclear Palsy Rating Scale; SEADL = Schwab and England Activities of Daily Living.

Figure 3. Longitudinal gray and white matter voxel-based morphometry analysis at 6 and 12 months

The t maps are results of 2-sample t tests at 6- and 12-month intervals with age, sex, and total intracranial volume included as covariates of no interest between normal controls and (A) patients with CBS, (B) patients with PSP, and (C) patients with PSP-DAV. All t maps are thresholded at cluster-level FWE-corrected p < 0.05, except for the following: PSP-DAV 12 months, displayed at p < 0.001 FWE-corrected; CBS 6 months, displayed at p < 0.001 uncorrected. Minimum t score is represented by dark blue for GM and orange for WM: CBS 6 months GM and WM, t = 3.2; CBS 12 months GM, t = 4.9; CBS 12 months WM, t = 4.7; PSP 6 months GM, t = 4.7; PSP 6 months WM, t = 4.5; PSP 12 months GM, t = 4.9; PSP 12 months WM, t = 4.6; PSP-DAV 12 months GM, t = 4.4; PSP-DAV 12 months WM, t = 4.2. Maximum t score (light blue for GM; yellow for WM) displayed for all groups is t = 6, except for PSP-DAV, which is t = 8. Slices displayed are as follows: coronal (−10, 17), sagittal (−38, 6, 51), axial (−20, 6, 36). CBS = corticobasal syndrome; DAV = davunetide; FWE = family-wise error; GM = gray matter; PSP = progressive supranuclear palsy; WM = white matter.

Figure 4. Clinical correlations

Scatterplots show correlations in a combined PSP and CBS group between annual change in PSPRS and annual change in (A) midbrain volume, (B) pontine volume, (C) left precentral gyrus volume, and (D) right precentral gyrus volume. In addition, correlations are shown in the PSP-DAV group between annual change in PSPRS and (E) midbrain volume and (F) pontine volume. Spearman ρ values and p values are the result of nonparametric Spearman rank order correlations, with p < 0.05 considered statistically significant. CBS = corticobasal syndrome; CI = confidence interval; DAV = davunetide; PSP = progressive supranuclear palsy; PSPRS = Progressive Supranuclear Palsy Rating Scale.