BACKGROUND/AIMS: We hypothesized an advantage of using carbohydrate antigen 19-9 (CA19-9) over carcinoembryonic antigen (CEA) in postoperative surveillance of colorectal cancer (CRC) patients with elevated preoperative serum CA19-9 (PreCA19-9). METHODOLOGY: We analyzed 197 CRC patients with elevated PreCA19-9 (>=37U/mL). Serum levels of CEA and CA19-9 were monitored postoperatively, according to standard protocol. IndexCEA and IndexCA19-9 were defined as the levels of CEA and CA19-9, respectively, recorded at recurrence confirmation or at the last follow-up of non-recurrence. RESULTS: There were 56 recurrences during follow-up. IndexCA19-9 (cut-off, 38.6U/mL) had better sensitivity (0.80 vs. 0.68) and specificity (0.87 vs. 0.83) than did IndexCEA (cut-off. 2.8ng/mL). Area under the curve (AUC) analysis revealed that although IndexCA19-9 (0.90; 95% confidence interval [CI], 0.84-0.95) had better coverage than IndexCEA (0.84; 95% CI, 0.78-0.90), there was no significant difference (p=0.085). With the progressive rising of PreCA19-9 levels, we observed the trend of better AUC performance of IndexCA19-9 over IndexCEA. At a cut-off of 42U/mL for PreCA19-9, IndexCA19-9 began to have significantly better coverage than IndexCEA (p=0.027). CONCLUSION: The surveillance advantage of CA19-9 over CEA was shown in the group of patients with preoperative CA19-9 levels >42U/mL.
BACKGROUND/AIMS: We hypothesized an advantage of using carbohydrate antigen 19-9 (CA19-9) over carcinoembryonic antigen (CEA) in postoperative surveillance of colorectal cancer (CRC) patients with elevated preoperative serum CA19-9 (PreCA19-9). METHODOLOGY: We analyzed 197 CRCpatients with elevated PreCA19-9 (>=37U/mL). Serum levels of CEA and CA19-9 were monitored postoperatively, according to standard protocol. IndexCEA and IndexCA19-9 were defined as the levels of CEA and CA19-9, respectively, recorded at recurrence confirmation or at the last follow-up of non-recurrence. RESULTS: There were 56 recurrences during follow-up. IndexCA19-9 (cut-off, 38.6U/mL) had better sensitivity (0.80 vs. 0.68) and specificity (0.87 vs. 0.83) than did IndexCEA (cut-off. 2.8ng/mL). Area under the curve (AUC) analysis revealed that although IndexCA19-9 (0.90; 95% confidence interval [CI], 0.84-0.95) had better coverage than IndexCEA (0.84; 95% CI, 0.78-0.90), there was no significant difference (p=0.085). With the progressive rising of PreCA19-9 levels, we observed the trend of better AUC performance of IndexCA19-9 over IndexCEA. At a cut-off of 42U/mL for PreCA19-9, IndexCA19-9 began to have significantly better coverage than IndexCEA (p=0.027). CONCLUSION: The surveillance advantage of CA19-9 over CEA was shown in the group of patients with preoperative CA19-9 levels >42U/mL.