Using data‐driven approaches to reveal distinct neurocognitive profiles in posterior cortical atrophy

Abstract

AbstractBackgroundPosterior cortical atrophy (PCA) is a clinically defined syndrome characterized by impairment in higher‐order visual processing due to neurodegeneration of posterior brain regions (occipito‐parietal and occipito‐temporal areas). Previous findings have reported broader cognitive impairment profiles, but their association with neural correlates remains unclear. In this project we aimed to use principal component analysis to explore the relationship between various cognitive dimensions and FDG‐PET hypometabolism in PCA patients.MethodWe included 59 patients with a clinical diagnosis of PCA (with supporting radiological evidence) from the Memory and Aging Center database at the University of California in San Francisco (Table 1). The sample consisted of 64.4% females, mean age was 63.0 ± 7.6, and mean MMSE was 20.7 ± 5.6 at the diagnostic visit. Participants underwent extensive neuropsychological assessment; these scores were subsequently entered into an orthogonally‐rotated principal component analysis. In a subset of patients with FDG‐PET (N = 24), associations between the emergent cognitive components and whole‐brain hypometabolism were examined to explore neural correlates of cognitive heterogeneity.ResultFive components explained 72% of overall performance variation in PCA patients: episodic memory, visuospatial function, working memory, verbal generation, and naming/depression (Figure 1). Labels were assigned base on the primary area of greatest cognitive dysfunction. Visualization of component scores suggested systematic performance variation within the PCA sample. The mean metabolism map showed, as expected, predominant hypometabolism in occipito‐parietal and occipito‐temporal regions (Figure 2A). Voxel‐wise correlations revealed that hypometabolism in distinct frontoparietal regions underpinned episodic memory and visuospatial (inferior/superior parietal lobes), working memory and verbal generation (superior/middle frontal gyrus) performance (Figure 2B).ConclusionPCA is a cognitively heterogeneous syndrome, where patients can show varying levels of cognitive deficits across episodic and working memory, executive functions, language and affect, over and above canonical visuospatial impairments.