Plasma phosphorylated tau species across neurodegenerative and cerebrovascular diseases: An ONDRI study

Elevated plasma phosphorylated tau (p-tau) species are proposed as specific biomarkers of Alzheimer's disease (AD) pathology but have also been reported in amyotrophic lateral sclerosis (ALS). Their profile across neurodegenerative and cerebrovascular diseases, as well as their discriminatory value when combined with other plasma biomarkers, remain uncertain. We studied 551 participants (ages 40.1 to 87.1 years; 207 women), comprising 506 patients from the Ontario Neurodegenerative Disease Research Initiative (ONDRI) multi-site cohort and 45 amyloid-β (Aβ) PET-negative cognitively unimpaired controls from the Brain-Eye Amyloid Memory study. Ten diagnostic groups were studied: mild cognitive impairment (MCI), AD, Parkinson's disease subdivided into cognitively unimpaired (CU), MCI, and dementia, frontotemporal dementia (FTD), cerebrovascular disease (CVD) subdivided into cognitively unimpaired (CU) and cognitively impaired (CI), ALS, and the control group. Baseline visit plasma measurements included p-tau217, p-tau181, Aβ42/40, neurofilament light chain (NfL), and glial fibrillary acidic protein (GFAP). Cross-sectional analyses compared biomarker levels as well as p-tau217-based inference of amyloid positivity status across diagnostic groups. Receiver operating characteristic-area under the curve (ROC-AUC) analyses ascertained the discriminatory capacity of these biomarkers and ratio combinations. Within the ALS group, baseline biomarker measures were also tested for association against baseline clinical features as well as longitudinal measures of neuropsychological change. P-tau217, p-tau181, GFAP, and NfL differed across groups (all P < 0.001). Compared to controls, p-tau217 was elevated in AD and ALS, p-tau181 and GFAP only in AD, and NfL in ALS and FTD. Inferred amyloid-β; positivity was more probable in MCI, AD, ALS, and cognitively impaired CVD groups compared to controls (all P < 0.05). The p-tau217 to NfL ratio discriminated ALS from AD with an ROC-AUC of 0.92 (95% CI 0.85-0.97) compared to p-tau217 alone at 0.65 (95% CI 0.52-0.76). No ALS clinical or longitudinal association survived correction for multiple comparisons. Elevated p-tau217 and inferred amyloid-β; positivity in ALS provide further evidence that the assay signal is not necessarily specific for AD pathology. However, the use of p-tau217 in combination with NfL provides improved discriminatory power between ALS and AD, supporting the use of plasma biomarker combinations rather than sole reliance on individual plasma biomarkers. Inferred amyloid-β; positivity in cognitively impaired CVD may indicate AD and/or cerebral amyloid angiopathy co-pathology. Currently available plasma biomarkers should therefore be used as diagnostic complements and interpreted within clinical context rather than as standalone molecular diagnoses.