Abstract
Background/Objectives: Early detection of pancreatic ductal adenocarcinoma (PDAC) can improve patient survival and biomarkers to facilitate this are greatly needed. We recently reported a serum biomarker signature comprising tissue inhibitor of metalloproteinase 1 (TIMP1), intercellular adhesion molecule 1 (ICAM1), cathepsin D (CTSD), thrombospondin 1 (TSP1/THBS1), and carbohydrate antigen 19-9 (CA 19-9), that detected Stage I and II PDAC with high sensitivity and specificity. In this assay, CA 19-9 is measured with a commercial instrument and individual ELISAs were developed to measure TIMP1, ICAM1, CTSD, and THBS1. Here, we report the analytical performance of these four analytes in their ELISA formats. Methods: Biomarker precision, linearity, algorithm precision, matrix effects, hook effect, method comparison, interference, and analyte stability were evaluated against acceptance criteria per CLSI guidelines. Results: High, medium, and low concentrations of each biomarker met acceptance criteria for inter- and intra-day precision (%CVs < 14%) and for linearity (%CVs < 11%). Matrix effects did not impact quantitation of any analyte nor was hook effect present. All analytes met acceptance criteria for accuracy and stability (all biases < 11.2% and <16.5%, respectively). For interference, two CTSD measurements and one ICAM1 measurement in HAMA-spiked samples showed 20.7-29% biases, falling slightly outside of acceptance criteria (<20% bias). All other analyte concentrations met interference acceptance criteria. In total, 94.1% of all diagnostic calls were made with 100% certainty, indicating high precision of the assay's algorithm. Conclusions: All analytes demonstrated acceptable analytical precision, linearity, accuracy, and stability, showing high overall analytical performance of each analyte.