Abstract
BACKGROUND: Current understanding of the fluid biomarker profile in early-onset Alzheimer’s disease (EOAD) associated with pathogenic APP variants remains limited. We characterized four EOAD pedigrees carrying pathogenic APP variants using clinical, neuroimaging, and biofluid biomarkers, comparing them with contemporaneous sporadic AD (SAD) to inform early detection. METHODS: Between December 2022 and December 2024, we enrolled 206 individuals from four APP-associated families harboring the D678G, V717I, or M722K variants; 92 individuals underwent genetic testing. Based on genotype and clinical status, participants were classified into three groups: symptomatic APP variant carriers (SMC; n = 11), asymptomatic APP variant carriers (AMC; n = 19), and cognitively normal non-carriers (NC; n = 62). In addition, 60 patients with sporadic Alzheimer’s disease and an age at onset of ≤ 65 years (SAD; sporadic EOAD) were included as a comparison group for fluid biomarker analyses. All participants included in the plasma analyses underwent clinical assessments, and plasma levels of P-tau217, P-tau181, Aβ42, Aβ40, GFAP, NfL, and the Aβ42/Aβ40 ratio were quantified. Structural MRI, PET, and cerebrospinal fluid biomarkers were obtained in a subset of participants, where available. Group comparisons were performed using the Kruskal-Wallis and Mann-Whitney U tests, with false discovery rate adjustment for multiple comparisons using the Benjamini-Hochberg procedure (P_FDR). Additionally, two patients treated with lecanemab underwent longitudinal follow-up during the first 15 infusions. RESULTS: Based on family informant reports and clinical observations, the median age at onset (AAO) among affected individuals in these pedigrees was 48.50 (44.0, 51.8) years. Plasma biomarker analyses showed that P-tau181, P-tau217, and GFAP levels were significantly higher in the SMC and SAD groups than in the AMC and NC groups (P_FDR < 0.05). Compared with the SMC group, the SAD group exhibited higher Aβ40 levels and a lower Aβ42/Aβ40 ratio (P_FDR < 0.05). Among V717I carriers, multiple biomarkers differed significantly between SMC and AMC (P_FDR < 0.05), whereas no significant differences were detected in the D678G or M722K subgroups. In addition, given the current sample size, APOE ε4 status was not associated with additional modification of plasma biomarker levels or AAO. No amyloid-related imaging abnormalities (ARIA) were observed during the lecanemab treatment period. From the pre-infusion baseline to immediately before the 15th infusion, one patient showed decreases in P-tau181 (− 5.70%), P-tau217 (− 12.23%), and GFAP (− 42.93%), accompanied by a mild improvement in cognitive performance; the other patient showed a reduction in P-tau181 (− 28.30%) but increases in GFAP (+ 36.59%) and NfL (+ 174.01%), with overall stable cognition. CONCLUSIONS: Plasma biomarkers provide critical evidence to advance our understanding of APP variant-associated EOAD. Both APP variant-associated EOAD and sporadic EOAD showed a similar biomarker profile, with higher plasma P-tau181, P-tau217, and GFAP levels than those in AMC and NC. In contrast, the two groups exhibited divergent patterns in Aβ40 levels and the Aβ42/Aβ40 ratio. In addition, given the current sample size, APOE ε4 status was not found to have an additional modifying effect on plasma biomarkers among APP variant carriers. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13195-026-02016-5.