Biglycan expression and clinical outcome in patients with pancreatic adenocarcinoma

• Published in: Tumor biology 2013-02, Vol.34 (1), p.131-137
• Main Authors: Aprile, Giuseppe, Avellini, Claudio, Reni, Michele, Mazzer, Micol, Foltran, Luisa, Rossi, Diego, Cereda, Stefano, Iaiza, Emiliana, Fasola, Gianpiero, Piga, Andrea
• Format: Article
• Language: English
• Subjects: Adenocarcinoma - metabolism; Adenocarcinoma - mortality; Adenocarcinoma - surgery; Biglycan - metabolism; Biomarkers; Biomarkers, Tumor - metabolism; Biomedical and Life Sciences; Biomedicine; Cancer Research; Clinical outcomes; Cyclooxygenase 2 - metabolism; Extracellular Matrix - metabolism; Female; Humans; Ki-67 Antigen; Male; Medical prognosis; Middle Aged; Pancreatic cancer; Pancreatic Neoplasms - metabolism; Pancreatic Neoplasms - mortality; Pancreatic Neoplasms - surgery; Proteins; Research Article; Survival; Treatment Outcome; Ubiquitin-Protein Ligases - metabolism
• ISSN: 1010-4283; 1423-0380
• DOI: 10.1007/s13277-012-0520-2

Conflicting results have been reported on the role of extracellular matrix (ECM) proteins in pancreatic cancer. Preclinical studies suggest that the overexpression of biglycan (proteoglycan-I, PG-I), a leucine-rich protein of the ECM, may induce growth arrest of pancreatic cancer cells. The aim of this study was to assess the prognostic role of biglycan expression in pancreatic cancer. We also evaluated MIB-1 and COX-2 expressions (as potential markers of growth and aggressiveness) to better characterize the biology of the tumors. The classical pathological parameters (grading, desmoplasia, perineural, or vascular invasion) as well as molecular determinants of prognosis were examined. MIB-1 (a proliferative index associated with prognosis in most tumors), COX-2, and PG-I expressions were detected by immunohistochemistry and immunofluorescence on tissue samples from 53 patients with pancreatic cancer and reviewed by two independent pathologists. To verify PG-I expression, three rabbit sera (LF104, LF112, and LF121 from NIH, Bethesda, MA, US) were tested. Logrank test and Cox’s model were applied for statistical analysis. Out of 53 patients, 40 had stage III and IV pancreatic cancer. Fourteen patients did not express any of the PG-I epitopes. The patients who expressed at least two PG-I epitopes had shorter survival compared to those with single epitope or lacking any expression (28 vs 44 weeks, P  = 0.0021). The MIB-1 higher expression predicted shorter survival (25 vs 41 weeks, P  = 0.0059). The other parameters were not associated with clinical outcome. Multivariate analyses confirmed PG-I expression and MIB-1 as independent negative prognostic factors. Patients who presented PG-I expression in the ECM had the worse prognosis compared to those who did not. Our results are not in contrast with the hypothesis that ECM proteins are a potential barrier to metastatic spread in localized pancreatic cancer. Rather, we underline the complexity of tumor–stroma interactions in the advanced stage of cancer and the need of further study.