Prediction of Biomarker Expression on Primary Pancreatic Ductal Adenocarcinoma Tissues Using Fine-Needle Biopsies: Paving the Way for a Patient-Tailored Molecular Imaging Approach.

Abstract

Targeted molecular imaging may improve tumor cell identification during diagnosis and resection of pancreatic ductal adenocarcinoma (PDAC). Although many molecular imaging biomarkers are (over)expressed in PDAC, intertumoral heterogeneity of biomarker expression hampers universal tracer administration. Preoperative, patient-specific screening and selection of the most optimal biomarker could therefore improve tumor delineation. This study evaluated whether fine-needle biopsy (FNB) specimens could be used to preoperatively predict biomarker expression in the corresponding primary PDAC specimen. Expression of previously identified PDAC biomarkers αvβ6, CEACAM5, EGFR, mesothelin, Lea/c/x, and sdi-Lea on FNB and corresponding primary tumor (PT) specimens (n=45) was evaluated using immunohistochemistry and quantified using a semi-automated image analysis workflow. Biomarker expression on FNB and PT tissues showed high concordance (∆H-score ≤ 50), i.e. was present in 62% of cases for αvβ6, 61% for CEACAM5, 85% for EGFR, 69% for mesothelin, 76% for Lea/c/x, and 79% for sdi-Lea, indicating high concordance. Except for αvβ6, biomarker expression on FNB tissues was positively correlated with PT expression for all biomarkers. Subgroup analyses showed that neoadjuvant therapy (NAT) had no major and/or significant effect on concordance, expression difference and, except for mesothelin, correlation of biomarker expression between FNB and PT tissues. This study demonstrated that biomarker expression in FNB tissues is predictive for PT expression, irrespective of the application of NAT. These findings thereby provide the foundation for the clinical application of an FNB-based biomarker-screening workflow, eventually facilitating a patient-specific approach of molecular imaging tracer administration in PDAC.