Abstract Pancreatic cancer is the fourth leading cause of cancer death in the US. The lack of specific diagnostic and prognostic biomarkers is a major obstacle in patient management. Pathologically, chronic pancreatitis can be sometimes difficult to distinguish from pancreatic ductal adenocarcinoma (PDA). Our group was the first to demonstrate the loss of the f subunit of eukaryotic initiation factor 3 (eIF3f) in many types of cancer cells including PDA cells, suggesting the potential tumor suppressive role of eIF3f. Our further studies showed that knockdown of eIF3f expression in normal pancreatic ductal epithelial cells leads to their malignant transformation-confirming the tumor suppressive function of eIF3f. However, the extent of the loss of eIF3f expression in clinical PDA and its precursor lesions (pancreatic intraepithelial neoplasia [PanIN], intraductal papillary mucinous neoplasm [IPMN], mucinous cystic neoplasm [MCN]), and its diagnostic utility is not known. We hypothesize that eIF3f expression is decreased or lost in clinical PDA and its precursor lesions, and the loss of eIF3f expression can be diagnostically useful to differentiate neoplasms from benign reactive ducts. We evaluated the protein expression level of eIF3f using tissue microarrays (TMAs) composed of a total of 49 duplicated cases of human benign pancreatic tissues (n=11), PDA precursor lesions (PanIN, IPMN, MCN, n=18), and PDA (n=20) from Pathology archive, followed by immunohistochemistry study of eIF3f. We then assessed the subcellular localization, intensity of the IHC staining, and percentage of positive cells by a gastrointestinal pathologist using light microscopy and correlated the findings with the clinical diagnosis. We also compared the staining pattern among the 3 most common IPMN subtypes defined by epithelial cell lineage (gastric, intestinal and pancreatobiliary subtypes). Fisher’s exact tests were used to test the association. A generalized linear regression model with a logit link was used to compare different diagnoses. ANOVA models were used to compare the continuous scores (intensity*percentage) across different diagnoses. All analyses were conducted using SAS (version 9.4, SAS Institute, Cary, NC). Interestingly, we found that there is a significant difference between the eIF3f staining pattern in benign pancreatic tissue, PDA precursor lesions, and PDA. In benign pancreatic ducts, eIF3f was exclusively located in the nuclei and none were located in the cytoplasm (0%). However, significantly increasing percentages of PDA and PDA precursor cases have cytoplasmic eIF3f staining (28% precursor lesions and 45% PDA, p<0.001). Furthermore, there is a significant loss of eIF3f staining in PDA and PDA precursor lesions (5.6% precursor lesion and 40% PDA are negative for eIF3f) compared to benign pancreatic ductal epithelium (all retained eIF3f staining, p<0.01). The diagnostic sensitivity and specificity of eIF3f are 60.5% and 100% respectively. The other interesting finding is that eIF3f is most frequently located in the cytoplasm or lost in the gastric type IPMN which was thought to have the most benign behavior compared to intestinal and pancreatobiliary type IPMN. In summary, we found altered subcellular localization and expression of eIF3f protein in not only PDA, but also PDA precursor lesions, suggesting its important role in PDA development. eIF3f is a highly specific biomarker for the diagnosis of PDA and its precursor lesions. These findings are significant because eIF3f can be a useful diagnostic biomarker to distinguish benign reactive pancreatic ducts from dysplastic or neoplastic pancreatic ductal lesions and a potential prognostic marker in pancreatic neoplasms. Citation Format: Jiaqi Shi, Brittany Silverman, Lili Zhao.{Authors}. Altered eIF3f subcellular localization and expression in pancreatic ductal adenocarcinomas and its precursor lesions. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2016 May 12-15; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(24 Suppl):Abstract nr B26.
Read full abstract