A sensitive biomarker panel to distinguish pancreaticobiliary malignancies from benign disease.

Abstract

238 Background: Biliary strictures present a diagnostic challenge to differentiate benign disease from hepatopancreaticobiliary (HPB) malignancies. Cytology from Endoscopic retrograde cholangiopancreatography (ERCP) or Endoscopic ultrasound with fine needle aspiration (EUS-FNA) have both been plagued by poor sensitivity and high false negative rates. In a review of over 80 adjunct molecular biology tests that have been attempted on ERCP and EUS-FNA samples, four immunohistochemistry biomarkers with sensitivities ranging 74-80% were identified; von Hippel Lindau loss of expression (VHL), over-expression of insulin-like growth factor 2 mRNA-binding Protein 3 (IMP3), and EF-hand Calcium 2+ binding S100 subfamily members A4 (S100A4) and P (S100P). We sought to determine if these results could be validated in a tumor explant model and furthermore if combining these tests into a biomarker panel could boost overall diagnostic sensitivity to 100%. Methods: Tumor tissue and normal surrounding pancreas from 27 pancreaticoduodenectomy specimens were selected by an experienced pathologist, subjected to immunohistochemistry staining with VHL, IMP3, S100A4, S100P, and the intensity and percent of cells staining was graded. Using ROC curve analysis, threshold criteria were chosen for each biomarker to differentiate between tumor and normal pancreas. Biomarkers were then evaluated as a panel for their ability to discriminate malignant from benign specimens. Results: Individual sensitivity of VHL, IMP3, S100A4, and S100P were found to be 75.0%, 79.2%, 45.8%, and 0%. When VHL, IMP3, and S100A4 were grouped into a panel, they were able to distinguish cancer from normal tissue with a sensitivity of 100% and a specificity of 96%. S100P was dispensable in our assay and only VHL, IMP3, and S100A4 staining were required to achieve 100% sensitivity. Conclusions: A panel of three biomarkers, VHL, IMP3, and S100A4, were able to distinguish pancreatic cancer from surrounding normal tissue with 100% sensitivity in our tumor explant model. Prospective studies on patient biopsy specimens are required to further validate the clinical use of this biomarker panel.