Abstract
BACKGROUND: We explored the prognostic value of the up-regulated carbohydrate antigen (CA19-9) in node-negative patients with gastric cancer as a surrogate marker for micrometastases. PATIENTS AND METHODS: Micrometastases were determined using reverse transcription quantitative polymerase chain reaction (RT-qPCR) for a subgroup of 30 node-negative patients. This group was used to determine the cut-off for preoperative CA19-9 serum levels as a surrogate marker for micrometastases. Then 187 node-negative T1 to T4 patients were selected to validate the predictive value of this CA19-9 threshold. RESULTS: Patients with micrometastases had significantly higher preoperative CA19-9 serum levels compared to patients without micrometastases (p = 0.046). CA19-9 serum levels were significantly correlated with tumour site, tumour diameter, and perineural invasion. Although not reaching significance, subgroup analysis showed better five-year survival rates for patients with CA19-9 serum levels below the threshold, compared to patients with CA19-9 serum levels above the cut-off. The cumulative survival for T2 to T4 node-negative patients was significantly better with CA19-9 serum levels below the cut-off (p = 0.04). CONCLUSIONS: Preoperative CA19-9 serum levels can be used to predict higher risk for haematogenous spread and micrometastases in node-negative patients. However, CA19-9 serum levels lack the necessary sensitivity and specificity to reliably predict micrometastases.