Predictors of Failure after Pancreaticoduodenectomy for Ampullary Carcinoma

BACKGROUND: Complete resection offers the only potential cure for ampullary carcinoma. We analyzed factors that contribute to treatment failure and survival in patients who underwent pancreaticoduodenectomy for ampullary carcinoma. STUDY DESIGN: We retrospectively reviewed all patients who underwent pancreaticoduodenectomy between August 1994 and August 2003 for ampullary carcinoma. Demographic, clinical, and pathologic data were collected. Chi-square analysis was used for categorical data and the t-test was used for continuous variables. Kaplan-Meier analyses were compared using the log-rank test to examine patient survival. RESULTS: Forty-three patients (24 men) aged 63.7 +/- 11.4 years (standard deviation) were followed for a mean of 23.9 months (median 660 days, range 18 to 2,249 days). Jaundice (n = 33) and weight loss (n = 13) were the most common presenting symptoms. Stage (p < 0.01) and degree of differentiation (p < 0.029) were significant predictors of failure by univariate analysis. But only stage (p < 0.04) was a significant predictor by multivariate analysis. Further analysis revealed that nodal status (p < 0.001), but not tumor grade, was a significant predictor of treatment failure. Neither demographic nor clinical variables were significant predictors. Five-year overall and disease-free survival rates were 67.4% and 51.4%, respectively. Both metastases and disease recurrence had significant impact on patient survival. CONCLUSIONS: Tumor stage is associated with treatment failure after pancreaticoduodenectomy for ampullary carcinoma and may identify candidates for adjuvant therapy. Because an aggressive surgical approach can be adopted safely with the best chance for cure, we recommend that pancreaticoduodenectomy be offered to all patients with ampullary tumors when malignancy or dysplasia is in question.