Background: Staging laparoscopy for patients with radiographically resectable pancreatic adenocarcinoma has been reported to yield an 8-15% finding of unresectable disease. Factors associated with the likelihood of subradiographic unresectable disease have not been clearly defined.
Methods: A prospectively maintained pancreatic database was reviewed and patients were identified who underwent staging laparoscopy for radiographically resectable pancreatic adenocarcinoma between January 2000 and December 2006. Preoperative carbohydrate antigen 19-9 (CA 19-9) values were assessed for their association with the presence of subradiographic unresectable disease.
Results: Four hundred ninety-one patients underwent staging laparoscopy. Resection was performed in 80% (n = 395). Of the 96 patients with unresectable disease, 75 (78%) had metastases either in the liver (n = 60) or peritoneum (n = 15). Preoperative CA 19-9 values were available for 262 of the 491 patients. Fifty-one of these patients had unresectable disease, of which 78% were due to distant disease. The median preoperative CA 19-9 value for patients who underwent resection was 131 U/ml versus 379 U/ml for those patients with unresectable disease (P = 0.003). A receiver operating characteristics (ROC) curve was developed for preoperative CA 19-9 value and tumor resectability. The statistically optimal cutoff value was determined to be 130 U/ml. Unresectable disease was identified in 38 of the 144 patients (26.4%) with a preoperative CA 19-9 >or= 130 U/ml, and in 13 of the 118 patients (11%) with a CA 19-9 < 130 U/ml (P = 0.003). CA 19-9 values greater than 130 U/ml remained a predictor of tumor unresectability on multivariate regression analysis [hazard ratio (HR) 2.70, 95% confidence interval (CI) 1.34-5.44; P = 0.005].
Conclusion: In this study, preoperative CA 19-9 values were strongly associated with the identification of subradiographic unresectable disease. Preoperative CA 19-9 values may allow surgeons to better select patients for staging laparoscopy.