PB 10 Identifying an imaging biomarker for progressive supranuclear palsy – a combined gray & white matter meta-analysis using ALE and SDM

Abstract

Background Progressive supranuclear palsy (PSP) is an atypical parkinsonian syndrome. PSP is described by clinical features of vertical gaze palsy and postural instability. Even though midbrain atrophy has been suggested as a hallmark of the disease, it has not been validated systematically so far. Methods To identify disease-specific regional atrophy in PSP, we conducted a systematic and quantitative meta-analysis on structural magnetic resonance imaging (MRI) data. Based on strict inclusion criteria 18 studies were identified in PubMed including 315 patients compared with 393 healthy controls. For analysis, the included studies were separated into studies investigating either gray or white matter atrophy. Ensuring consistent results, we used a unique approach by combining two widely used meta-analytical algorithms. The results of anatomical likelihood estimation (ALE) and Seed-basedDMapping (SDM) were superimposed in a conjunction analysis. Results The analysis of gray matter revealed that clusters in the bilateral thalamus, bilateral anterior insulae, midbrain and bilateral caudate nuclei were consistently identified as effected by both methods, SDM and ALE ( Fig. 1 ). White matter analyses revealed consistent atrophy in the bilateral superior cerebellar pedunculi, cerebral pedunculi, and midbrain. Discussion In light of previous meta-analyses investigating neurodegenerative diseases, our meta- analyses identify atrophy in the gray matter in the midbrain and in the white matter in the cerebral/cerebellar pedunculi and midbrain as disease-specific in PSP. Atrophy in these brain regions is suggested to be incorporated into new revised diagnostic criteria for PSP. The study design enabled high validity and statistical power. We combined two meta-analytic techniques, combined analyses in white and gray matter, and included the biggest cohort of PSP patients to date.