PGE2 in Pancreatic Cyst Fluid Differentiates IPMN from Mucinous Cystadenoma and May Predict Degree of IPMN Dysplasia
C. Max Schmidt*1,5, Michele T. Yip-Schneider1,2, Matthew C. Ralstin3, Sarah Dutkevitch1, Sabrina C. Wentz1, John M. Dewitt4, Thomas J. Howard1, Lee Mchenry4, Keith D. Lillemoe1
1Surgery, Indiana University School of Medicine, Indianapolis, IN; 2Indiana University Cancer Center, Indianapolis, IN; 3Indiana University School of Medicine, Indianapolis, IN; 4Gastroenterology, Indiana University School of Medicine, Indianapolis, IN; 5Richard L. Roudebush Veterans Affairs Medical Center, Indianapolis, IN
Introduction: Management of pancreatic mucinous cyst may be contingent upon its identity as intraductal papillary mucinous neoplasm (IPMN) or mucinous cystadenoma. IPMNs, although unifocal on presentation, develop multifocality and cyst size is not a predictor of malignant potential. Conversely, mucinous cystadenomas are rarely multifocal and thus amenable to segmental resection/enucleation, and cyst size is a predictor of malignant potential. Distinguising between IPMN and mucinous cystadenoma is clinically important, necessitating novel biomarkers of degrees of IPMN dysplasia. Preliminary studies suggest a positive correlation of cyclooxygenase-2 (COX-2) expression with IPMN dysplasia. We hypothesized that prostaglandin E2 (PGE2), a measure of cyclooxygenase activity, distinguishes between IPMN and mucinous cystadenoma and may be a biomarker of IPMN dysplasia.
Methods: Fifty patients signed informed consent for collection of pancreatic cyst and/or ductal fluid at the time of endoscopy (EUS or ERCP) or operation (OR). Pancreatic fluid was analyzed by PGE2 ELISA. Secretin was used to enhance ductal fluid yield. Secretin decreased PGE2 level 5X (3±0.9 without vs 0.6±0.2 with; p<0.05) when the same patient was used as a control, which was adjusted in overall analyses. COX inhibitor medication use was recorded. COX inhibitor use did not influence PGE2 level when compared to the mean PGE2 of the category in question. PGE2 level was correlated with surgical pathologic diagnosis and stage.
Results: Mean PGE2 level (pg/μl) in IPMNs (2.3±0.6) was greater than in mucinous cystadenomas (0.2±0.1) (p<0.05). In comparison, serous cystadenomas had a mean PGE2 level of 0.2±0.1. Mean PGE2 level of IPMN by stage was 13.3±2.3 (invasive), 4.4±0.9 (carcinoma, i.e., carcinoma-in-situ), 1.3±0.4 (borderline), and 0.1±0.01 (adenoma). Thus, only IPMN adenoma grade attained the PGE2 range of mucinous cystadenomas. With progression of IPMN from adenoma to invasive, PGE2 showed a significant stepwise increase. By comparison, ductal fluid in patients with ductal adenocarcinoma not associated with IPMN had a mean PGE2 level of 4.8±1.4. In IPMN fluid samples, there was no difference in PGE2 level in duct fluid compared to cyst fluid, OR- vs. EUS-acquired cyst fluid or OR- vs. ERCP-acquired ductal fluid.
Conclusions: PGE2 level may be useful in predicting IPMN over mucinous cystadenoma in patients with suspected mucinous lesions based on cytology, mucin or elevated CEA. PGE2 level may also be a useful indicator of malignant progression of IPMN if followed longitudinally, and thus COX inhibitors deserve further study as chemopreventative agents in IPMN.
Back to 2007 Program and Abstracts