BACKGROUND: Early recurrence of invasive pancreatic cancer is common even after curative resection. To establish appropriate selection criteria for radical surgery, it is essential to identify the patient population at risk for early recurrence. STUDY DESIGN: One hundred thirty-three of 136 patients who underwent potentially curative pancreatectomy for invasive ductal adenocarcinoma of the pancreas between 1999 and 2003 were divided into two groups: patients whose recurrence developed within 1 year after operation and those whose recurrence did not develop within 1 year. Clinicopathologic factors were retrospectively analyzed between these groups using univariate and multivariable methods. RESULTS: One postoperative death occurred, yielding an in-hospital mortality rate of 0.7% (of 136 patients). Eighty-one patients (61% of 133) were identified as having recurrent pancreatic carcinoma within a year. Paraaortic nodal involvement and positive washing cytology were independent predictors of early recurrence. The median survival time in 26 patients with paraaortic nodal involvement was 13 months, versus 30 months in 106 patients without paraaortic nodal involvement (p < 0.001). Paraaortic lymph node involvement was notably associated with elevated CA19-9 a month after operation (p = 0.03), larger tumor size (p = 0.02), and a positive surgical margin (p = 0.04). CONCLUSIONS: Sampling of paraaortic lymph nodes is recommended as a routine examination at laparotomy. When positive nodes are confirmed by frozen section, early recurrence and poor survival are inevitable, even after radical operation including extended lymphadenectomy.
BACKGROUND: Early recurrence of invasive pancreatic cancer is common even after curative resection. To establish appropriate selection criteria for radical surgery, it is essential to identify the patient population at risk for early recurrence. STUDY DESIGN: One hundred thirty-three of 136 patients who underwent potentially curative pancreatectomy for invasive ductal adenocarcinoma of the pancreas between 1999 and 2003 were divided into two groups: patients whose recurrence developed within 1 year after operation and those whose recurrence did not develop within 1 year. Clinicopathologic factors were retrospectively analyzed between these groups using univariate and multivariable methods. RESULTS: One postoperative death occurred, yielding an in-hospital mortality rate of 0.7% (of 136 patients). Eighty-one patients (61% of 133) were identified as having recurrent pancreatic carcinoma within a year. Paraaortic nodal involvement and positive washing cytology were independent predictors of early recurrence. The median survival time in 26 patients with paraaortic nodal involvement was 13 months, versus 30 months in 106 patients without paraaortic nodal involvement (p < 0.001). Paraaortic lymph node involvement was notably associated with elevated CA19-9 a month after operation (p = 0.03), larger tumor size (p = 0.02), and a positive surgical margin (p = 0.04). CONCLUSIONS: Sampling of paraaortic lymph nodes is recommended as a routine examination at laparotomy. When positive nodes are confirmed by frozen section, early recurrence and poor survival are inevitable, even after radical operation including extended lymphadenectomy.