Multiple biomarkers improve diagnostic accuracy across Lewy body and Alzheimer's disease spectra

Abstract

Abstract Objective More than half of neurodegenerative disease patients have multiple pathologies at autopsy; however, most receive one diagnosis during life. We used the α‐synuclein seed amplification assay (αSyn‐SAA) and CSF biomarkers for amyloidosis and Alzheimer's disease (AD) neuropathological change (ADNC) to determine the frequency of co‐pathologies in participants clinically diagnosed with Lewy body (LB) disease or AD. Methods Using receiver operating characteristic analyses on retrospective CSF samples from 150 participants determined αSyn‐SAA accuracy, sensitivity, and specificity for identifying clinically defined LB disease and predicting future change in clinical diagnosis. CSF biomarkers helped determine the frequency of concomitant Lewy body pathology, ADNC, and/or amyloidosis in participants with LB disease and AD, across clinical spectra. Results Following a decade‐long follow‐up, the clinically or autopsy‐defined diagnosis changed for nine participants. αSyn‐SAA demonstrated improved accuracy (91.3%), sensitivity (89.3%), and specificity (93.3%) for identifying LB disease compared to all non‐LB disease, highlighting the limitations of clinical diagnosis alone. When examining biomarkers of co‐pathology, amyloidosis was present in 18%, 48%, and 71% (χ2(2) = 13.56, p = 0.001) and AD biomarkers were present in 0%, 8.7%, and 42.9% (χ2(2) = 18.44, p < 0.001) of LB disease participants with different stages of cognitive impairment respectively. Co‐occurring biomarkers for αSyn‐SAA and amyloidosis were present in 12% and 14% of AD compared to 43% and 57% LB disease participants with different stages of cognitive impairment (χ2(3) = 13.87, p = 0.003). Interpretation Our study shows that using a combination of αSyn‐SAA and AD biomarkers can identify people with αSyn, ADNC, and co‐pathology better and earlier than traditional clinical diagnostic criteria alone.