Survival benefit of extended D2 resection for proximal gastric cancer

The long-term survival of patients with adenocarcinoma of the proximal stomach remains dismal. Despite its increasing frequency and poor prognosis, a general consensus has not been reached on the extent of surgical resection. The significance of extended lymph node dissection (D2 gastrectomy) for the surgical treatment of patients with proximal gastric cancer was evaluated.

Sixty-two patients who underwent a potentially curative total or proximal gastric resection were retrospectively divided by extent of lymphadenectomy into two groups: the extended resection group (D2,D2.5) and limited resection group (D1,D1.5). Survival rates were estimated by the method of Kaplan and Meier [J Am Stat Assoc 53:457-486, 1958] and the differences compared by the log rank test. Multivariate analysis of prognostic parameters was performed using the Cox proportional hazard model.

The median overall survival time for the extended resection group (D2,D2.5) was 34 months compared to 18 months for patients treated by a more limited resection (D1,D1.5). Patients treated with extended resection had an estimated 5-year overall survival rate of 37% compared to 21% for patients treated with limited resection. This difference was statistically significant with a P value of 0.04. The median disease-free interval for the extended resection group was 31 months compared to 17.6 months for patients in the limited resection group. The 5 year disease-free survival rate for both groups was 37% and 17%, respectively (P = 0.09). Extent of lymphadenectomy and stage of disease were found to be independent predictors of overall and cancer-free survival.

Patients treated with an extended lymph node dissection (D2 gastrectomy) were more likely to survive 5 years, had longer disease-free intervals, and prolonged median survival times (particularly patients with T1-3,N0-1,M0 cancers) as compared to those patients treated with a more limited lymph node dissection (D1,D1.5). These differences reached or approached statistical significance.