Assessment of prognostic factors in pancreatic ductal adenocarcinoma: Focus on the retroperitoneal margin

e15670 Background: Pancreatic ductal adenocarcinoma (PDA) is a leading cause of cancer death in the West, with a nearly superimposable incidence and mortality. Resection is the only chance for cure, and various features in resection specimens correlate with outcome. While most consider the uncinate margin (UM) to be the true retroperitoneal margin, it has been suggested that the posterior pancreatic surface (PPS) may also be important. At another site with a retroperitoneal margin (i.e. rectum), 1 mm margins are significant. We thus evaluated margin status in various ways, focusing on the retroperitoneal region and emphasizing outcome. Methods: We identified all pancreaticoduodenectomies for PDA over a 6 year period in which the PPS was histologically evaluable. Tumors were assessed for the following: size, stage, grade, lymph node (LN) status, vascular and perineural invasion, and margin status. Margin status was evaluated in 3 ways: traditional margins (tumor at pancreatic neck, bile duct, and/or uncinate margins), 1 mm margins (traditional + tumor within 1 mm of UM), and PPS margins (traditional + tumor within 1 mm of PPS or UM). Kaplan-Meier survival curves were constructed with univariate factors compared by log rank analysis; multivariate analysis was done using the Cox proportional hazard model. Results: Fifty-one tumors exhibited the following features: size (mean 3.3 cm), stage (48 T3), grade (27 low, 24 high), LN status (11 neg, 40 pos), positive margins (13 traditional, 23 1 mm, and 32 PPS). Nearly all tumors exhibited at least focal vascular and perineural invasion. Grade influenced survival (p=0.0001), while size (p=0.417) and traditional and PPS margins did not (p=0.5 and 0.95). LN status and 1 mm margins trended toward significance (p=0.17 and 0.2). Conclusions: Use of a two-tiered grading system is highly correlated with survival. Neither tumor size, LN status, nor traditional or PPS margins are significant, while 1 mm margins trend toward significance. Although lack of significance of some features (i.e. size, LN status, and traditional margin status) may be attributable to modest sample size, lack of significance of the PPS may reflect its anatomic nature (i.e. not a true surgical margin). Additional study of 1 mm margins in a larger tumor set is warranted. No significant financial relationships to disclose.