Prognostic significance of tumor deposits in radically resected gastric cancer: a retrospective study of a cohort of 1915 Chinese individuals

Abstract

Abstract Background Tumor deposits (TDs) have been identified as an independent prognostic factor in gastric cancer (GC). However, the associated clinicopathological factors and how to simply and reasonably incorporate TD into the TNM staging system remain undetermined. The aim of the current study was therefore to assess the significance of TD among radically resected GC patients. Methods We retrospectively reviewed 1915 patients undergoing radical resection between 2007 and 2012. The patients were classified into two groups according to TD status (absent vs. present), and the clinicopathologic characteristics, DFS, and OS were compared. Associations of TD presence with other clinicopathologic factors were evaluated by logistic regression analysis. Univariate and multivariate Cox regression analyses were performed to determine the prognostic factors for DFS and OS in the primary cohort. Propensity score matching (PSM) was performed to reduce the possibility of selection bias according to the presence of TD. External validation of previously proposed modified staging systems incorporating TD was conducted. Results The detection rate of TD was 10.5% (201/1915). The presence of TD was significantly related to unfavorable clinicopathologic variables, including advanced T and N categories. According to the multivariate Cox regression analysis, the presence of TD was identified as an independent prognostic factor for DFS and OS in the primary cohort (both P < 0.001). In the after-PSM cohort, TD presence also significantly shortened DFS and OS. In the external validation, one system that incorporated TD into the pTNM stage had the best performance. Conclusions The presence of TD was significantly associated with poor survival in radically resected GC patients. The incorporation of TD into the TNM staging system can further improve the predictive capability. A multicenter cohort with a large sample size is needed to determine the appropriate method of incorporation.