Prediction of recurrence after radical surgery for gastriccancer. A scoring system obtained from a prospective multicenter study on 536 patients

Background: The aim of this prospective multicenter study of the Italian Research Group for Gastric Cancer (IRGGC) was to define a scoring system for the prediction of tumor recurrence after potentially curative surgery for gastric cancer. Methods: Five hundred and thirty-six patients who underwent UICC R0 resection between 1988 and 1998 were considered. All patients were followed-up using a standard protocol after discharge from the hospital. The mean follow-up period was 55 ^ 45 months, and 91 ^ 34 months for surviving patients. Time and site of relapse were recorded for all recurrent cases. The scoring system was calculated on the basis of a logistic regression model, where the presence of the recurrence was the dependent variable, and age, gender, tumor location, tumour size, depth of invasion, nodal status, Lauren histotype, extent of resection and extent of lymphadenectomy were the covariates. A constant and the coefficients of significant variables were included in a formula for the estimation of the probability of recurrence for each patient. Results: Recurrence occurred in 272 patients on 536 (50.7%). The site of relapse was locoregional in 127 cases (23.7%), peritoneal in 87 cases (16.2%), and hematogenous in 92 cases (17.2%). In 35 patients (6.6%) multiple sites of recurrence were present. The scoring system for the prediction of the risk in individual cases gave values ranging from 0 to 100. The model distributed most of cases in the extremes of the range; 170 cases (31.7%) were classified below 20, and 171 (31.9%) over 80. The risk of recurrence increased remarkably with score values; it was only 3.4% in 59 patients with score below 5, 7.3% in 41 patients with score 5–10, 25.2% for score 10–50, 73.4% for score 50–90, and 95.4% in 108 patients with score 90–100. No recurrence was observed in 43 patients with score below 4, whereas all of the 56 patients with score over 97 recurred. The model correctly predicted recurrence in 227 on 272 patients (sensitivity: 83.4%), whereas the absence of recurrence was correctly predicted in 214 on 264 patients (specificity: 81.1%); the overall accuracy was 82.2%. Prognostic score was clearly superior to UICC tumor stage in predicting recurrence. The time of recurrence was lower in cases with high score values. Conclusions: The scoring system obtained with a regression model on the basis of our follow-up data is useful to define subgroup of patients at very low or very high risk of tumour recurrence after radical surgery for gastric cancer. The easy reproducibility of the formula for the computation of the score may allow a prospective validation of the model.