Evaluation of focused sentinel lymph node RT-qPCR screening for micrometastases with the use of the Maruyama computer program.

Abstract

Background: In this preliminary study, we investigated the sensitivity and specificity of reverse transcriptase (RT)-qPCR lymph node (LN) metastases detection, the accuracy of intraoperative dye navigation, and the incidence of micrometastasis (MM) detection with this protocol, compared to other published studies.

Methods: A total of 23 patients were enrolled in the study. The first stained LN was analyzed using RT-qPCR for carcinoembryonic antigen (CEA) and cytokeratin 20 (CK-20) expression, as markers for MM involvement. The Maruyama computer program was used to determine the most likely first metastatic site. These results were compared with the actual staining patterns to evaluate whether the first draining LN was extracted. We analyzed the correlations between MM and tumor characteristics. The incidence of MM detected with the present method was compared to other studies, as markers of the accuracy of the present protocol.

Results: At 35 threshold cycles, the RT-qPCR had a negative predictive value of 100 % and a positive predictive value of 83.3 %. MM were detected in 4 out of 14 node-negative patients (28.6 %). The extracted sentinel LN coincided in 76.9 % of cases with the most probable first metastatic LN predicted by the Maruyama program. MM were found more frequently in these 'high-risk' LNs. Significant differences were found in the Lauren's histological type distribution and the age distribution among the MM-positive and MM-negative groups.

Conclusion: Our preliminary results confirm that RT-qPCR is an accurate method of MM detection, that the dye navigation enables the determination of the first draining LN, and that the incidence of MM detection with this focused sentinel LN protocol is comparable to other studies.