Concordance between [18F] GTP1 PET and fluid tau biomarkers in the Tauriel phase II Study

Abstract

AbstractBackgroundThe Tauriel Study (NCT03289143) was a Phase 2 randomized, double‐blind, placebo‐controlled, parallel‐group clinical trial that assessed the safety and efficacy of semorinemab in patients with prodromal‐to‐mild Alzheimer’s disease. We examined the concordance of fluid tau biomarkers with Tau PET for the identification of patients with high tau burden at enrollment.MethodTau at baseline was assessed by [18F]GTP1 PET imaging and assays for multiple CSF and plasma tau indices, including pTau181 and mid‐domain tTau (Elecsys assays), tau peptides 2‐24 and 93‐105 (anti‐peptide capture mass‐spec assay), and phosphotau/unphosphorylated tau217 ratio (mass‐spec assay). The association between imaging and fluid tau levels were assessed with Spearman correlations. AUC, sensitivity and specificity are reported in the context of predicting tau burden with a single fluid biomarker, where ‘high tau’ is defined as GTP1 temporal meta‐ROI SUVR ≥1.3. Cutoffs were derived using Youden’s index.Result354 participants had both PET and plasma samples; 68 subjects had paired PET and CSF data. GTP1 significantly correlated with each fluid marker, highly with pT217 (r = 0.78), moderately with other CSF assays (r = 0.44 ‐ 0.56) and weakly with plasma tTau (r = 0.19) (Table 1). Discrimination was highest for CSF pTau217 ( AUC = 0.909). Specificity was 0.862 and 0.84 for pTau181 and pTau217 respectively, while sensitivity was 0.564 and 0.868 (Figure 1). Empirically defined cutoffs were 36.88 pg/mL for pTau181 and 0.052 for pT217.ConclusionBaseline Tau PET is shown to be prognostic for future cognitive decline and thus could be included in enrollment criteria for future clinical studies in AD. Alternative approaches include the use of fluid biomarkers for tau. While pTau181 and n‐terminal tau indices can only be optimized for either high specificity or high sensitivity, both can be concurrently optimized using pTau217, thus positioning it as a potential alternative marker for Tau pathology. Additional CSF and plasma tau indices are pending similar analyses.