Survival After Pancreaticoduodenectomy is not Improved by Extending Resections to Achieve Negative Margins

Objective: This study was undertaken to determine the survival benefit of extending resections to obtain microscopically negative margins after positive intraoperative frozen sections. Summary Background Data: The impact of residual microscopic disease after pancreaticoduodenectomy is currently a point of controversy. It is, however, generally believed that microscopically positive margins negatively impact survival and this may be improved by ultimately achieving negative margins. Methods: Since 1995, patients undergoing pancreaticoduodenectomy for pancreatic adenocarcinoma have been prospectively followed. Margin status has been codified as macro/microscopically negative (R0) or macroscopically negative/microscopically positive (R1). The impact of margin status on survival was evaluated utilizing Survival curve analysis. Data are presented as median, mean +/- SD where appropriate. Results: For pancreatic adenocarcinoma, 202 patients underwent pancreaticoduodenectomy. R0 resections were achieved in 158 patients, 17 of whom required extended resections to achieve complete tumor extirpation after an initially positive intraoperative frozen section (R1 -> R0). R1 resections were undertaken in 44 patients. Median survival for patients undergoing R0 resections was 21 months, 26 +/- 23.4 months versus 13 months, 17 +/- 21.0 months for patients undergoing RI resections (P = 0.02). Median survival for patients undergoing R1 -> R0 resections was 11 months, 16 +/- 17.3, (P = 0.001). Margin status had a significant correlation with "N" stage and AJCC stage but not "T" stage. Conclusion: Survival after pancreaticoduodenectomy is not improved by extending pancreatic resections to achieve negative margins after initially positive intraoperative frozen sections. Tumor-specific factors beyond the presence of disease at a surgical margin are responsible for the abbreviated survival seen in patients undergoing RI resections.