Abstract
Splenic hilar (no.10) lymph node dissection during total gastrectomy is no longer recommended for advanced proximal gastric cancer. However, the treatment efficacy of no.10 lymph node dissection in Borrmann type 4 tumors remains unclear. We enrolled 539 patients who underwent total gastrectomy for Borrmann type 4 tumors between 2006 and 2016 in four major institutions in Korea. We compared the long-term survival of the no.10 lymph node dissection (n = 309) and no-dissection groups (n = 230) using the propensity score (inverse probability of treatment weighting). The treatment effects of no.10 lymph node dissection were estimated in the weighted sample using the Cox proportional hazards regression model with a robust sandwich-type variance estimator. After inverse probability of treatment weighting, there were 540.4 patients in the no.10 lymph node dissection group and 532.7 in the no-dissection group. The two groups showed well-balanced baseline characteristics, including tumor node metastasis stage. The 5-year survival rates in the no.10 lymph node dissection and no-dissection groups were 45.7% and 38.6%, respectively (log-rank p = 0.036, hazard ratio 0.786, 95% confidence interval 0.630-0.982). Multivariate analysis revealed that no.10 lymph node dissection was an independent favorable prognostic factor (adjusted hazard ratio 0.747, 95% confidence interval 0.593-0.940) after adjusting for other prognostic factors. Sensitivity analyses in other inverse probability of treatment weighting models and the propensity score matching model showed similar results. Patients undergoing no.10 lymph node dissection showed improved survival compared to those without. No.10 lymph node dissection is recommended during total gastrectomy for patients with Borrmann type 4 gastric cancer.