Surgical resection offers the only potential cure for pancreatic carcinoma. Although the overall prognosis remains a dismal, several recent series have reported an encouraging increase in 5-year survival after resection, exceeding 20%. As the reasons for this improvement are not clearly understood, numerous clinico-pathological parameters (demographic, intraoperative and histopathologic factors) have been investigated to evaluate their role in predicting long term survival. In this single-institution study, immediate and long-term outcome after pancreatic resection in patients with pancreatic adenocarcinoma was retrospectively evaluated, focusing attention on the possible impact of different clinico-pathologic factors on long-term survival.Sixty-six patients with a confirmed histologic diagnosis of adenocarcinoma of the pancreas, treated by pancreatic resection at the Department of Surgery of the Catholic University of Rome in the years 1988-1997, were retrospectively analyzed. Morbidity and survival data were reviewed and potential prognostic factors were compared statistically by univariate analysis.There was no postoperative mortality. Twenty-five patients (38%) developed major operative complications. Pancreatic fistula was the most common complication, and occurred in 7 patients (11%). The actuarial overall and disease-specific survival for all 66 patients were respectively 58% and 59% at 1 year, 27% and 31% at 3 years, and 13% and 20% at 5 years, with a median survival time of 13.4 months. Nodal status was the only single factor significantly affecting survival by univariate analysis. The 3-and 5-year survival rates were respectively 35% and 19% for node-negative patients and 7% and 0% for node-positive patients (P = .04). A positive correlation with improved survival, even if not of statistical significance, was shown for other pathologic or intraoperative factors. Among the former, 5-year survival rates were better for patients with negative resection margins as compared to patients with positive margins (12% vs 7%, P = ns). Among the latter, a better actuarial 5-year survival rate was shown for patients with shorter operative time ( 4 hours 5%, P = ns) and for patients that received fewer transfusions (0-2 blood units, 14% survival vs 3 or more blood units, 0%; P = ns). Age, gender, tumor diameter and tumor grading showed no influence on survival in this series.Our series confirmed that nodal status is the strongest independent predictor of survival. Limited intraoperative transfusion, reduced operative time and clear margins could also yeald a prognostic significance, and require further confirmation in larger series.