Survival in early gastric cancer: multivariate analysis on 72 consecutive cases

Few reports from the Western hemisphere have investigated the impact of pathological features and surgical modalities on the prognosis of patients affected by early gastric cancer (EGC). In particular, the extent of lymphadenectomy (limited vs. extended) and the type of gastric resection (subtotal vs. total) remain controversial issues in the management of EGC. The aim of this study was to identify factors influencing prognosis in patients affected by EGC.

Hospital records and pathological specimens of 72 patients with EGC undergoing resective surgery during the period 1981-1995 were retrospectively reviewed. Patient status was determined by follow-up examination or by telephone contact. Univariate and multivariate analysis was used to calculate the 5-year survival probabilities with respect to the following variables: age (< or = 65, > 65), sex, depth of invasion (mucosal, submucosal) tumor location (upper, middle and lower third), gross appearance (type I, type II and type III), size (< or = 1.5 cm, > 1.5 cm), presence or absence of lymph node metastasis, histological type (intestinal, diffuse), extent of lymphadenectomy (limited or extended), and type of gastrectomy (total or distal subtotal). Survival was the outcome variable studied.

Multivariate logistic regression analysis showed that age, nodal involvement and depth of invasion were independently associated with poor survival.

Results showed a significant dominance of host- and tumor-related factors over the type of surgical procedure on prognosis of EGC patients.