Objectives: The endoscopic features of gastric neuroendocrine tumors (G‐NETs) remain unclarified. The present study investigated the endoscopic features of G‐NETs in relation to the clinicopathological findings.Methods: This retrospective study analyzed consecutive patients with G‐NETs who received endoscopic or surgical treatment between January 2005 and December 2023. The endoscopic and clinicopathological findings of the lesions were analyzed to provide diagnostic information.Results: Among 29 patients, the characteristic endoscopic findings of G‐NETs on white‐light images were reddish color (66%), dilated vessels (83%), submucosal tumor‐like marginal elevation (59%), and central depression (48%). The gross appearance of G‐NETs was classified into two macroscopic subtypes: reddish polypoid lesions (n = 17) and submucosal tumor‐like lesions (n = 9). Magnifying narrow‐band imaging endoscopy revealed an absent microsurface pattern plus an irregular microvascular pattern in all cases of reddish polypoid lesions with central depressions (100%, 9/9). The findings of a reddish polypoid lesion and an absent microsurface pattern plus an irregular microvascular pattern corresponded to the subepithelial NET component close to the non‐neoplastic surface epithelium. Additionally, reddish polypoid lesions were significantly more frequent in type 1 G‐NETs than in type 3 G‐NETs (80% vs. 11%, p < 0.001), while submucosal tumor‐like lesions were significantly more frequent in type 3 G‐NETs than in type 1 G‐NETs (78% vs. 10%, p < 0.001).Conclusions: These endoscopic features should increase the index of suspicion and help clinicians to correctly diagnose G‐NETs through the pathological examination of biopsy specimens.
Background: Endoscopic full‐thickness resection (EFTR) is an effective treatment method for gastric submucosal tumors (SMTs). We aimed to perform a technical analysis of EFTR in gastric SMT and compare it with the outcome parameters. Method: Sixty‐one gastric SMTs from 60 patients were resected using EFTR. The indication criteria: size, 11–30 mm, connection to the muscularis propria on endoscopic ultrasonography, intraluminal growth type, no ulceration, and histologically evident or clinically suspicious gastrointestinal stromal tumors (GISTs). The following technical improvements were introduced during the study Periods 1–3: routine use of clip‐line traction (Periods 1–3); use of a plastic bag retriever (Periods 2–3); adaptation of the reopenable clip over‐the‐line method (ROLM, Period 3); implementation of no‐touch EFTR (Period 3); and elimination of submucosal injection (Period 3). Results: The endoscopic complete resection rate was 100%, with a similar tumor resection time (median, 50 min) throughout the periods. Specimen damage was less frequent after using the plastic bag retriever in Periods 2 and 3 ( p = 0.001). In Period 3, ROLM required longer full‐thickness defect closure time (39 min, p = 0.011), but it provided secure closure and shortened the fasting days ( p = 0.010). Histological diagnoses included 38 GISTs, 14 leiomyomas, and nine other pathologies. In Period 3, the implementation of no‐touch EFTR increased the resected specimen size (33 mm, p = 0.010) and improved the histological complete (R0) resection rate of the GISTs (13/13, 100%, p = 0.017). Conclusion: Several technical improvements significantly improved the outcomes of EFTR for gastric SMTs, warranting the external validation of this technique.