Clinical Factors Associated With Discordant Plasma P-Tau217 and Established Alzheimer Disease Biomarkers.

Background

AND

Objectives

Plasma phosphorylated tau (p-tau) 217 is increasingly used for Alzheimer disease (AD) diagnosis and treatment decisions but may be discordant with established AD biomarkers. False-positive (FP) plasma results may lead to overtreatment and false-negative (FN) results to missed or delayed disease-modifying therapies. We evaluated factors associated with discordant results.

Methods

This retrospective study included participants enrolled in memory and aging studies at Mayo Clinic with "positive" (>0.324 pg/mL) or "negative" (<0.186 pg/mL) plasma p-tau217 results and confirmatory AD biomarkers. Plasma results were classified as true positive (TP), true negative (TN), FP, or FN, referencing amyloid-PET and CSF p-tau181/β-amyloid (Aβ) 42. Clinical and laboratory variables were compared using univariate analyses; multivariable logistic regression evaluated independent associations with discordance.

Results

Among 706 participants (median 70.0 [interquartile range 15.8] years; 54% male), 60 (8.5%) had discordant plasma p-tau217 results. FN results (5.6%) were more frequent than FP results (2.8%; p = 0.007). Discordance was associated with older age (odds ratio [OR] 1.61 per decade, 95% CI 1.23-2.14) and obstructive sleep apnea (OR 1.94, 95% CI 1.08-3.49). FN (vs TP) results were associated with milder clinical stage (global Clinical Dementia Rating [CDR] ≤0.5 vs CDR ≥1; p < 0.001), higher Mini-Mental State Examination scores (OR 1.20 per point, 95% CI 1.05-1.36), obstructive sleep apnea (OR 2.01, 95% CI 1.01-4.3), higher body mass index (OR 1.11/kg/m2, 95% CI 1.03-1.20), and higher estimated glomerular filtration rate (eGFR: OR 1.33/10 mL/min/1.73 m2, 95% CI 1.05-1.69). In age-adjusted models, body mass index (OR 1.14, 95% CI 1.03-1.25), eGFR (OR 1.63/10 mL/min/1.73 m2, 95% CI 1.22-2.16), and earlier clinical stage (CDR ≤0.5) were independently associated with FN results. FP (vs TN) results were associated with older age (OR 2.37, 95% CI 1.36-4.09), thyroid disease (OR 4.41, 95% CI 1.36-14.3), lower eGFR (OR 0.48/10 mL/min/1.73 m2, 95% CI 0.31-0.66), and advanced disease stage (CDR ≥1). In age-adjusted models, eGFR (OR 0.54/10 mL/min/1.73 m2, 95% CI 0.35-0.82) and advanced disease stage (CDR ≥ 1 vs CDR 0: OR 11.11, 95% CI 2.0-50.0; vs CDR 0.5: OR 4.55, 95% CI 1.12-16.67) were independently associated with FP results.

Discussion

Plasma p-tau217 results and established AD biomarkers were discordant in 8.5% of participants. Clinical stage and factors related to brain amyloid clearance, renal clearance, and volume of distribution were associated with discordant results. These insights may inform mechanisms of discordance and identify patients requiring confirmatory biomarker measures.