Background: Nodal status is a predictor of outcome for patients with resected pancreatic adenocarcinoma. This study further evaluated the significance of pathologic nodal assessment and the extent of nodal metastases on patient survival.
Methods: From a prospective pancreatic cancer database, 560 consecutive patients were identified who underwent an R0 resection for adenocarcinoma between 1995 and 2005. Overall survival was compared to lymph node (LN) status, absolute number of pathologically assessed LN, and LN ratio (number of positive LN to the total number of LN assessed). Comparisons were made using a Cox regression model.
Results: Of the 560 patients who underwent R0 resection, 477 (85%) had pancreaticoduodenectomy, and 83 (15%) had distal pancreatectomy. Median follow-up was 12 months (range, 0-118 mo), and estimated 5-year survival was 16%. A total of 211 (37%) patients were LN-negative (N0). Median survival for N0 patients was 29 months. The median number of pathologically assessed LNs for N0 patients was 12 (range, 0-41). As a continuous variable, the number of pathologically assessed LN did not correlate with survival for N0 patients. The median survival for the 349 patients with node-positive (N1) disease was 16 months. The median number of pathologically assessed LN in this group was 19 (range, 3-61), and the median number of positive LNs was three (range, 1-21). When analyzed as a continuous variable, the absolute number of positive LNs was a significant predictor of survival for N1 patients (see graph A) with a linear relationship up to 8 positive LNs. Two positive LNs represented the greatest separation in survival (p=0.01). LN ratio was also associated with survival (see graph B). Ratio = 0.18 was the best cut-off for median survival, p<0.01.
Conclusions: The number of positive LNs and LN ratio are strong predictors of survival for patients with node-positive pancreatic cancer.