B-264 Clinical validation of a plasma-based p-Tau217 immunoassay

Background Plasma P-tau217 is a promising biomarker demonstrating diagnostic accuracy to detect amyloid beta (Aß) pathology, with emerging potential for early detection of Alzheimer’s disease. The Quanterix SP-X P-tau217 immunoassay was evaluated for clinical sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), as compared to Aß positron emission tomography (PET) using Florbetapir (FBP). This data was used to derive two quantitative cut-points affording positive, negative, and indeterminate outputs where clinical sensitivity and specificity are =90% while the indeterminate range represented <20% of results. These outputs are the basis for diagnostic interpretation of P-tau217 results. Methods Samples with FBP standardized uptake value ratio (SUVR) results were acquired from Eli Lilly and Company I5T-MC-AACI trial (TRAILBLAZER-ALZ 2). This cohort (n=524) was comprised of male (48%) and female (52%) patients presenting with cognitive decline between 60-86 years of age (median=74 years). Samples were analyzed across 3 reagent lots using the plasma P-tau217 immunoassay employing antibodies 4G10E2 (specific for brain tau) and IBA493 (specific for P-tau217) from Eli Lilly and Company. The chemiluminescent signal from the antigen-antibody complex was captured using the Quanterix SP-X Imager. Results were used only from runs with acceptable quality control and where duplicate sample CVs (well-to-well) were <20%. Data analysis and receiver operating characteristic (ROC) curve evaluation, including calculations of sensitivity, specificity, NPV, PPV, and % indeterminate rates of paired results were performed using the cutpointr package (version 1.1.2) in RStudio (2024.09.1 Build 394; "Cranberry Hibiscus" Release). Analysis in R was performed by first converting Aß PET (FBP SUVR) values into diagnostic negative, represented by “0” for SUVR <1.1, and diagnostic positive, represented by “1” for SUVR = 1.1. This SUVR threshold was determined to correspond to 24.1 centiloids. Results This cohort was determined to exhibit 64.9% PET positive prevalence. The cut point producing the greatest sum of sensitivity and specificity was 0.145 U/mL of P-tau217. However, this cut point results in sensitivity and specificity values below the desired target of =90%. Therefore, two cut points of <0.13 U/mL and =0.20 U/mL afforded a negative, indeterminate, and positive outputs providing sensitivity of 90%, specificity of 90%, NPV of 84%, PPV of 94%, and a percent indeterminate of 17%. Further, these two cut points resulted in 2.9% false positive and 5.2% false negative rate. The area under the curve (AUC) for the ROC analysis was 0.941. Conclusion Cut points for plasma P-tau217 were established based on FBP PET and are as follows: negative, <0.13 U/mL; indeterminate, =0.13 to <0.20 U/mL; and positive, =0.20 U/mL. These cut points were selected to optimize outputs for clinical sensitivity and specificity with the lowest indeterminate range. Measurement of plasma P-tau217 continues to demonstrate accurate diagnostic utility in the identification of Aß pathology.