Biomarkers

FDG-PET has been a widely used imaging modality for evaluating hypometabolism in Alzheimer's disease (AD) for decades. This study investigates differences in biomarker profiles between patients with AD and non-AD patterns of FDG-PET. Additionally, we aim to explore whether the AD FDG-PET pattern complements the diagnostic performance of plasma p-tau217 for detecting beta-amyloid (Aβ) and tau pathology.

We included 51 participants diagnosed with amnestic mild cognitive impairment or mild dementia. The participants were classified by radiologist-reported FDG-PET patterns as either AD (n = 32) or non-AD (n = 19). AD patterns were defined by typical AD-related hypometabolism (involving posterior cingulate, precuneus, and posterior temporal and parietal lobes). Aβ-PET centiloid (AβCL), Aβ positivity (A+), tau standardized uptake value ratio (SUVR) in Braak III and/or IV, tau-PET status (T+), and plasma biomarkers, including p-tau217, p-tau181, glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL) levels, were determined. ROC analyses identified optimal cut-off values for p-tau217, which were further assessed alongside FDG-PET patterns for their diagnostic performance for A+ and A+T+.

Participants with AD FDG-PET patterns, compared to those with non-AD patterns, exhibited higher proportions of A+ (28/32 (87.5%) vs. 9/19 (47.4%), p = 0.002) and A+T+ (22/32 (68.8%) vs. 2/19 (10.5%), p <0.001). They also showed elevated AβCL levels and tau SUVRs across Braak III-IV regions (all p <0.001, Figure 1A-1D). Plasma biomarkers, including p-tau217, were markedly higher in the AD pattern group (all p <0.001, Figure 1E-1H). The AD FDG-PET pattern demonstrated high diagnostic accuracy, while plasma p-tau217 alone (at optimal cut-off values) showed excellent performance for predicting A+ and A+T+ (Table 1). Combining FDG-PET patterns with p-tau217 (cut-off ≥ 4.874 pg/mL) enhanced specificity for A+T+ prediction (from 0.556 to 0.741) while maintaining high sensitivity (from 1.0 to 0.917) (Table 1). The determination of p-tau217 cut-off values will be illustrated with ROC curves (Figure 2A-B).

FDG-PET patterns, particularly when combined with plasma p-tau217, improve diagnostic specificity for A+ and A+T+ pathology. These findings emphasize the complementary value of integrating imaging and fluid biomarkers for accurate AD diagnosis.