Background: Patients with early gastric cancer (EGC) who have undergone noncurative endoscopic resection (ER) generally require additional surgery due to the possibility of lymph node metastasis (LNM). This study aimed to develop a reliable risk-stratification system to predict LNM after noncurative ER for EGC. Methods: A total of 2368 patients had a diagnosis of EGC and underwent ER. The study analyzed 321 patients who underwent additive gastrectomy and lymph node dissection after noncurative ER. Independent risk factors for LNM were identified and used to develop a risk-stratification system to estimate the relative risk of LNM. Results: Of the 321 patients, 23 (7.2%) had LNM. A logistic regression analysis showed that female sex, lymphovascular invasion (LVI), and a positive vertical margin were significantly associated with LNM. The authors established a risk-stratification system using sex, LVI, and positive vertical margin (area under the receiver-operating characteristic [AUROC] curve, 0.811). The high-risk LNM group (score, ≥ 2 points) showed a significantly higher risk of LNM than the low-risk LNM group (score, <2 points) (14.0 vs 1.2%). No LNM was found in patients with a risk score of zero. After internal and external validation, the AUROC curve for predicting LNM was 0.788 and 0.842, respectively. Conclusions: The risk-stratification system developed in this study will facilitate identification of patients who should undergo LN dissection after noncurative ER. Although additive surgery should be performed after noncurative ER for patients with a high risk of LNM, a close follow-up visit could be considered for low-risk patients with multiple comorbidities or high operative risks.
Bibliographical noteFunding Information:
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2012R1A1A1042417).
© 2017, Society of Surgical Oncology.
All Science Journal Classification (ASJC) codes