Long-term outcomes after noncurative endoscopic resection of early gastric cancer: the optimal time for additional endoscopic treatment

Background and Aims: We aimed to evaluate long-term outcomes with noncurative endoscopic submucosal dissection (ESD) for early gastric cancer (EGC) and surveillance strategies such as the optimal time for additional endoscopic treatment in patients with noncurative ESD. Methods: Of 2527 patients who underwent gastric ESD for EGC, 512 (20.3%) patients with noncurative resection were reviewed. Noncurative resection is defined as positive resected margins on histology, lymphovascular infiltration, or beyond the expanded criteria for ESD. Results: The mean ± standard deviation follow-up duration was 79.0 ± 55.7 months. A total of 264 patients (51.6%) and 50 patients (9.8%) underwent surgery and endoscopic treatment after noncurative resection, respectively, whereas 198 patients (38.7%) were observed. Cancer-specific survival and disease-free survival rates were significantly different among the surgery, other endoscopic treatment, and observation groups (96.7%, 86.8%, and 86.2%, respectively; P =.030; and 92.5%, 73.6%, and 63.0%, respectively; P <.001). When patients who underwent surgery were excluded, the disease-free survival rate of recurrence was not significantly different between the endoscopic treatment and observation groups (73.6% vs 63.0%; P =.548). To exclude the potential for the presence of lymph node metastasis, we further analyzed disease-free survival of local recurrence by comparing the patients with only a positive lateral resection margin. The disease-free survival rate was higher in the endoscopic treatment group than in the observation group (89.2% vs 69.1%; P =.023). Moreover, additional endoscopic treatment within 3 months showed significant associations with lower risk of local recurrence on multivariate analysis (hazard ratio, 0.017; 95% confidence interval, 0.002-0.260; P =.003). Conclusions: In patients with noncurative ESD, additional surgery showed a better long-term outcome; moreover, when a positive lateral resection margin was the only noncurative factor, additional endoscopic treatment within 3 months could be considered to improve disease-free survival.