Optimal extent of lymph node dissection in patients with gastric cancer who underwent non-curative endoscopic submucosal dissection with a positive vertical margin

• Published in: European journal of surgical oncology 2020-12, Vol.46 (12), p.2229-2235
• Main Authors: Furukawa, Kenichiro, Kamiya, Satoshi, Sugino, Takashi, Aizawa, Daisuke, Kawabata, Takanori, Notsu, Akifumi, Hikage, Makoto, Tanizawa, Yutaka, Bando, Etsuro, Takizawa, Kohei, Ono, Hiroyuki, Terashima, Masanori
• Format: Article
• Language: English
• Subjects: Aged; Carcinoma - pathology; Carcinoma - surgery; Endoscopic Mucosal Resection; Endoscopic submucosal dissection; Female; Gastrectomy; Horizontal exposure length; Humans; Lymph Node Excision - methods; Lymph Nodes - pathology; Lymphatic Metastasis; Male; Margins of Excision; Middle Aged; Neoplasm Invasiveness; Neoplasm, Residual; Obvious tumor cell exposure; Positive vertical margin; Reoperation; Stomach Neoplasms - pathology; Stomach Neoplasms - surgery
• ISSN: 0748-7983; 1532-2157
• DOI: 10.1016/j.ejso.2020.07.002

The optimal extent of lymph node dissection in patients receiving non-curative endoscopic submucosal dissection (ESD) and diagnosed with a positive vertical margin is unclear. This study attempted to identify optimal candidates for D2 lymph node dissection among these patients. This study included patients who underwent gastrectomy for primary gastric cancer following non-curative ESD with a positive vertical margin between January 2002 and December 2018. We classified the patients according to the positive vertical margin pattern into an obvious exposure group and a non-obvious exposure group. We developed a score model for predicting lymph node metastasis (LNM) using factors selected by multivariate analyses and beta regression coefficients, and the incidence of LNM was evaluated. This study included 110 patients. LNM was detected in 17 patients (15%). We developed a predictive scoring system as follows: tumor size >30 mm (0, No; 1, Yes) + undifferentiated type tumor in the invasive front (0, No; 2, Yes) + depth of submucosal invasion > 1500 μm (0, No; 1, Yes) + obvious tumor exposure at the vertical margin (0, No; 1, Yes). In patients with 5 points, the incidence rates of all and group 2 LNM were as high as 60% and 40%, respectively. Conversely, in patients with fewer than 5 points, the incidence rates of all and group 2 LNM were just 11% and 5%, respectively. In patients with 5 points according to our score model for predicting LNM, gastrectomy with D2 lymph node dissection is recommended.