Prospective, comparative study for the evaluation of lymph node involvement in gastric cancer: Maruyama computer program versus sentinel lymph node biopsy.

BACKGROUND

Stage-adapted surgery guarantees the best outcome for patients with gastric cancer. Successful identification of lymph node involvement may help to reduce the number of extended lymphadenectomies. Preoperative diagnostic tools have low sensitivity and specificity for determining lymph node involvement. Evaluation of sentinel lymph nodes (SLNs) intraoperatively has good results, while the accuracy of the Maruyama computer program (MCP) is controversial.

METHODS

We investigated 40 patients by the Maruyama computer model and labeled lymph nodes with blue dye for SLN mapping. To compare the probability calculations by MCP and the results of SLN mapping, we had to define a cutoff level; we did this using receiver-operating characteristics analysis. Sentinel lymph nodes were examined in frozen sections intraoperatively and by standard hematoxylin and eosin staining postoperatively.

RESULTS

A total of 795 lymph nodes were removed and examined. The Maruyama computer model had a sensitivity of 91.3 %, specificity of 64 %, and accuracy of 80 % by the best cutoff point. The false-negative rate was 8.7 %. The sensitivity of SLN mapping was 95.7 %, the false-negative rate was 4.3 %, and the specificity was 100 %. The accuracy of SLN mapping was 97.4 %. Only the sensitivity of MCP and SLN biopsy was proven equivalent.

CONCLUSIONS

Our results suggest that intraoperative SLN examination is superior to preoperative estimation with the MCP. Correct definition of lymph node involvement helps in planning the best stage-adapted surgery in gastric cancer.