Enhancing atypical Alzheimer's disease phenotyping through the oculomotor window in posterior cortical atrophy

Abstract

Background

Although Alzheimer's disease (AD) is biologically well characterized, early and precise phenotypic diagnosis remains challenging, especially for atypical variants. Posterior cortical atrophy (PCA) is a rare form of AD with progressive neurovisual impairment related to degeneration in visual processing areas.

Objective

To investigate how alterations in eye movement metrics, measured through video-oculography (VOG), reflect dysfunction across distinct brain networks in various AD phenotypes, with particular emphasis on PCA.

Methods

This study compared oculomotor parameters derived from VOG saccade analysis in early AD patients exhibiting two clinical phenotypes, PCA-AD (n = 21) and amnestic mild cognitive impairment (aMCI-AD, n = 11), along with 27 age-matched controls. Parameters analyzed included saccade latency, gain, velocity, intrusions and antisaccade error rates. All patients exhibited cerebrospinal fluid biomarkers consistent with AD pathology.

Results

As expected, neuropsychological testing revealed more severe neurovisual and executive deficits in PCA-AD versus aMCI-AD, and greater memory storage impairment in aMCI. Oculomotor data showed that PCA-AD patients exhibited prolonged saccade latencies, reduced gain, slower vertical saccades, and increased antisaccade errors compared to controls and aMCI-AD. Receiver operating characteristic analysis combining key saccadic metrics demonstrated up to 90% sensitivity and specificity in distinguishing PCA from controls and aMCI.

Conclusions

These findings support the use of VOG oculomotor metrics as phenotypic biomarkers in differentiating AD clinical forms. In PCA, they reflect the dysfunction of visuo-spatial attentional networks and their interaction with subcortical eye movement control circuits.