Proposal of a risk score for recurrence in patients with curatively resected gastric cancer.

In this study, 168 patients who underwent curative resection for gastric cancer with prognostic serosal invasion [ps(+)] and 150 without prognostic serosal invasion [ps(-)] were analyzed separately to determine the prognostic importance of clinicopathological factors, and identify which patients were at high risk of recurrence. A multivariate analysis of survival time using Cox's proportional hazard model revealed the important prognostic factors to be: Lymph node involvement, the classification of gross appearance, macroscopic serosal invasion, and interstitial connective tissue in the ps(+) group; and lymph node involvement, macroscopic serosal invasion, and venous invasion in the ps(-) group. We proposed a risk score of recurrence based on the results of a further multivariate analysis called Hayashi's Quantification Analysis II, in which recurrence was chosen as an objective variable and the above prognostic factors were chosen as explanatory variables. Eighty-four percent of the patients with a score of 0 or higher in the ps(+) group and 83% of those with a score of +6 or higher in the ps(-) group showed recurrence. Thus, we believe that this score is useful for identifying those patients at high risk of recurrence, who should receive intensive chemotherapy even after curative resection.