J Korean Surg Soc.
2003 Aug;65(2):145-149.
Clinical Usefulness of Preoperative and Postoperative CA-19-9 Antigen in Pancreatic and Bile Duct Neoplasms
- Affiliations
-
- 1Department of Surgery, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, Korea.jsheo@smc.samsung.co.kr
Abstract
- PURPOSE
The CA-19-9 antigen is a commonly used tumor marker for pancreatic and bile duct neoplasms. It is well known that the CA-19-9 antigen is a good predictor of resectability, prognosis and recurrence. The aim of this study was to evaluate the efficacy of the CA-19-9 antigen as preoperative resectability and postoperative recurrence markers in pancreatic and bile duct neoplasms. METHODS: Between February 1995 and June 2001, 234 patients, with pancreatic and bile duct neoplasms, at the Department of Surgery, Samsung Medical Center, were followed up with the CA-19-9 antigen, both preoperatively and postoperatively. The ROC curve was used to analyze the relationship between a radical resection and the preoperative CA-19-9 antigen. During the postoperative follow- up, 124 patients that had radical resections, and postoperative follow-up for the CA-19-9 antigen, were studied for this relationship using Chi-square tests. RESULTS: The cut-off values in the pancreatic (P=0.527) and bile duct (P=0.688) neoplasms were 325.50 U/ml and 58.35 U/ml, respectively, using the ROC curve for the relationship between a radical resection and the preoperative CA-19-9 antigen, although the result was statistically insignificant. From the 124 patients used for the statistical analysis, 64 had a recurrence, of which 57 showed an increase in the CA-19-9 antigen, with the 7 showing no elevation. 60 patients had no recurrence, of which 28 showed an increase in the CA-19-9 antigen. The elevation of the CA-19-9 antigen was analyzed between the recurrence and no recurrence groups using Chi-square tests, which revealed a statistical significance (P<0.001). CONCLUSION: In patients with pancreatic and bile duct neoplasms, the CA-19-9 antigen could not predict the preoperatie resectability, but was a clinically effective follow-up marker for a recurrence.