Tu1952 Survival Analysis of Veteran Patients With Resectable Pancreatic Tumors With and Without Surgery

Abstract

INTRODUCTION: Treatment of pancreatic cancer depends on the stage of disease. For patients with tumors deemed resectable, the best chance for prolonged survival is with early curative resection. However, it is not clear that this is the best general approach because of the high morbidity and mortality of pancreatic resection, and because tumors often are found to be unresectable at the time of surgery. The aim of this study was to compare survivals of patients who had early curative resection of their pancreatic tumors with those who did not. METHODS: We reviewed medical records of all patients diagnosed with pancreatic cancer at our VA Medical Center from 2005 through 2010, and gathered data on multiple demographic and clinical variables. Survivals of patients with and without tumor resection were compared using Kaplan Meier analysis. RESULTS: We identified 116 patients with pancreatic cancer (all men, 66% white, 29% black, 90% adenocarcinoma, 8% neuroendocrine tumor). Tumor location was head (66%), body (10%), tail (16%) and body/tail of pancreas (4%). Stage at presentation was: I (3%); II (22%); III (12%); IV (63%). 28 patients had tumors deemed resectable (group R) (mean age 64±1.7 years [SEM], 64% white, 81% adenocarcinoma), and 88 had non-resectable (group NR) tumors (mean age 67.1±9.7 years, 66% white, 92% adenocarcinoma). Ultimate treatments overall included surgery (21%), neoadjuvant therapy (4%), adjuvant therapy (10%) and palliative chemotherapy (35%). Kaplan-Meier analysis demonstrated an overall median survival of 153 days (95% CI 93195) for the total group of patients with pancreatic cancer. A total of 17 (14.7%) patients underwent surgery with curative intent (group R+) including 13 Whipple procedures and 4 pancreatic tail resections. 11 patients initially deemed resectable did not undergo tumor resection (group R-) because they were later deemed unfit for surgery (n=4), declined surgery (n=1) or were found to have unresectable disease with intraoperative metastases during surgery (n=6). Both groups (R+ and R-) had similar mean ages (63.6±2.2 vs. 64.5±2.7 years, p=0.80), weight at diagnosis (85.7±5.4 vs. 81.2±3.9, p=0.56), race, stage, tumor size, Ca 19-9 and serum albumin levels. Median survival of the R+ group (401±95 days) was significantly longer than that of the Rgroup (162±70, p =0.03) and the NR group (112±33, p=0.0009). Survival for the 6 patients who had surgery revealing unresectable tumors was poorer than that for the 17 patients who had tumor resections with curative intent (R+), but did not reach significance (246.5 days vs. 401 days, p=0.18). CONCLUSIONS: Early surgery for patients with resectable pancreatic cancer is associated with improved survival. The finding of unresectable disease at surgery is a poor prognostic sign, but the surgery itself does not appear to be the factor influencing the shortened survival.