Biomarkers

Our objective of this study is to identify suitable local cut-off values of plasma-based biomarkers in Alzheimer disease (AD) in the discrimination of AD and non-AD subgroups in Thai population.

We conducted a prospective study in 51 individuals with clinically diagnosed AD (15 had amyloid PET positive (≥ 30 centiloid, CL), 16 had CSF Aβ₄₂ ≤547 pg/mL and CSF pTau 181 >57 pg/mL regarded as amyloid positive AD), 34 individuals with clinically diagnosed non-AD dementia (9 had amyloid PET negative, 25 had CSF Aβ₄₂ and CSF pTau181 negative), and 12 individuals with mild cognitive impairment (MCI) or normal controls (6 had amyloid PET negative and 6 had CSF Aβ₄₂ and CSF pTau181 negative). Aβ-PET using 18F-florbetapir and plasma biomarkers (Aβ₄₀, Aβ₄₂, p-tau181, pTau 217, Glial fibrillary acidic protein (GFAP)) were obtained in all subjects within 6 months before or after the PET study. The quantitative analysis of Aβ-PET to obtain Centiloid (CL) followed the standard method using the SPM8 pipeline. Blood biomarker analysis utilized Simoa® Quanterix immunoassay. CSF Aβ42, t-Tau and p-Tau were measured separately, in duplicate, by commercially available sandwich ELISA kits (Innotest; Innogenetics/Fujirebio, Ghent, Belgium). Participants underwent a standardized diagnostic dementia evaluation.

Among 97 individuals, 54 (55.7%) amyloid positive and 43(44.3%) amyloid negative individuals were included in this blood biomarkers validation study. Mean age was 65.86(8.01) years, mean TMSE was 19.57(7.44) and mean MOCA was 15.48(7.27). Mean(SD) plasma Aβ 42/40, pTau181, pTau 217, NFL, and GFAP were 0.054(0.013) pg/mL, 37.591(18.975) pg/mL, 1.088(0.949) pg/mL, 102.703(412.308) pg/mL, and 162.459(107.223) pg/mL respectively. Utilizing single cut off from ROC analysis: the cut off points are as followed; for plasma Aβ42/40≤0.054pg/mL (AUC(SE)=0.716(0.064), p <0.0001, accuracy 72.63(62.52-81.28)); plasma pTau181>32.7pg/mL (AUC(SE)=0.812(0.058), p <0.0001, accuracy73.20(63.24-81.68)); plasma pTau217>0.522pg/mL (AUC(SE)=0.856(0.056), p <0.0001, accuracy87.14(76.99-93.95)); GFAP>119pg/mL (AUC(SE)=0.796(0.058), p <0.0001, accuracy77.89(68.22-85.77)). For high and low cut- points, the results were shown in table 1.

Plasma biomarkers showed promising diagnostic performances and potential clinical usefulness in diagnosing dementia. Plasma p-tau217 was best differentiates AD positive from AD negative groups, followed by plasma GFAP, pTau181, and Aβ42/40 ratio. Thank you Health Systems Research Institute (Thailand) grant support for this study.