P2‐259: OCULOMOTOR RESPONSES LINKED TO COGNITIVE MARKERS FOR ALZHEIMER'S DISEASE CAN ENHANCE RISK PROFILING IN PATIENTS WITH MILD COGNITIVE IMPAIRMENT

Abstract

Background: Combining eye-tracking methodologies with a cognitive marker for Alzheimer’s disease, namely the Short-Term Memory Binding Test (STMBT), has enhanced the effectiveness of the assessment increasing its sensitivity and specificity to 100% (1). Whether such classification accuracy would help identify patients with Mild Cognitive Impairment (MCI) who may hold a higher risk of progressing to dementia remains unexplored. Methods: We measured the pupil size in 14 patients with MCI and 9 healthy controls while they performed the STMBT. The STMBT assessed the ability to temporarily hold colours presented in bicoloured objects either as individual features (baseline) or integrated within object representations (binding). Patients were also assessed with standard cognitive screening tests (MMSE, ACE-R, IFS). We applied a ROC-derived cut-off score recently obtained by (2) which achieved 100% classification between patients with AD dementia and healthy controls to our MCI patients. We aimed to identify an oculomotor profile in MCI patients compatible with that seen in patients with AD dementia. Results: Of the 14 patients with MCI, 5 fell below cut-off. They were compared with those who fell above cut-off, and with healthy controls. MCI patients above and below cut-off significantly differed from controls on all the cognitive screening tests used, but not between them. Patients above cut-off were significantly different from controls on both conditions of the STMB test (baseline and binding). However, patients below cut-off significantly differed from controls only on the binding condition. The pupil responses of above cut-off MCI patients and controls was indistinguishable (p=1.0). However, that one of below cut-off MCI patients was significantly different from both controls (p=0.007) and above cut-off MCI patients (p= 0.027). Conclusion: Pupil behaviours during performance on tests known to be markers for AD help identify MCI patients who show a profile compatible to that seen in patients with AD dementia. Such a profile was only driven by oculomotor responses as none of the neuropsychological tests used distinguished between MCI patients. Future studies are needed to explore the presence of AD pathology in those positive to the oculomotor-marker here presented or their likelihood to progress to AD dementia in longitudinal assessments.