Arveux P, Faivre J, Boutron M C, Piard F, Dusserre-Guion L, Monnet E, Hillon P
Registre des Tumeurs Digestives (Equipe associée INSERM-DGS), Dijon, France.
Dig Dis Sci. 1992 May;37(5):757-63. doi: 10.1007/BF01296435.
A population-based series of 246 gastric cancer patients operated for cure and who survived the postoperative phase was reviewed to determine prognostic factors after potentially curative treatment. The overall five-year observed survival rate was 34.8%, and the relative survival rate was 43.9%. Previous history of gastric ulceration, tumor location, tumor size, gross appearance, extension within the gastric wall, and number of proximal lymph nodes involved were significantly related to both crude and relative survival rates. Age was a significant prognostic factor when considering crude survival rates, but it had no influence on relative survival rates. Multivariate analysis of crude and relative survival gave similar results except for age. The covariates retained in the final model were, by decreasing importance, extension within the gastric wall, lymph node involvement, gross appearance and tumor location. Combining the two major prognostic criteria, tumor extension through the gastric wall and lymph node involvement, four prognostic categories could be determined with five-year corrected survival rates ranging from 92% in patients with a carcinoma limited to the gastric wall to 17% in patients with more than two positive nodes whatever the extension in the gastric wall. Gross appearance had no influence on prognosis for carcinomas limited to the gastric wall, but had a significant impact on prognosis of more extended carcinomas. From these data, a simple staging system requiring only routinely available pathological data was proposed. This classification could be helpful for planning multicenter clinical trials on this disease where progress in therapy is needed.
