Is It Possible to Estimate the Depth of Submucosal Invasion of Early Gastric Cancer By Chromoendoscopy?

Abstract

Background/Aim: Recently, indications for endoscopic resection(ER) of early gastric cancer (EGC) have been expanded according to clinicopathological studies and development of endoscopic procedure, such as endoscopic submucosal dissection (ESD). Differentiated gastric cancers with a size less than 3 cm, without lymphovascular involvement, and submucosal (sm) invasion of less than 500 μm (classified as SM1) have been able to define further the risk of lymph node metastasis and become candidates for ER. It is, therefore, important to estimate the depth of sm invasion, prior to ER. In this study, we evaluated positive predictive value (PPV) in estimating the depth of sm invasion and the specific findings to sm invasion in the endoscopic and pathological features. Methods: Since February 2007 to October 2008, a total of 166 patients with 200 differentiated EGCs who underwent ESD were enrolled in this retrospective study. Patients with recurrent cancer after ER were excluded. We reviewed recorded endoscopic images and evaluated PPV in estimating the depth of sm invasion by chromoendoscopy. And we investigated, 1) Endoscopic findings specific to SM1 for example, nodule in the depressed area and surrounding elevation, 2) Average width and form of sm invasion pathologically. The form of sm invasion was assigned to two groups, destructive and non-destructive pattern defined as muscularis mucosae (mm) being broken by tumor invasion and focal infiltration into mm, respectively. Results: PPV in estimating the extent of EGCs was 90.4% in M(mucosal lesion), 21.1% in SM1, 100% in SM2(sm invasion > 500 μm) and 92.9% in M-SM1. In endoscopic finding, the lesions with nodule in the depressed area were observed in 17.6% of SM1, 7.1% of SM2 and surrounding elevation were detected 17.6% in SM1, 57.1% in SM2. Among 35 lesions with sm invasion pathologically, endoscopic “M”, “SM1” and “SM2” was 17, 14 and 4 lesions and their Average width of sm invasion were 1885, 3700 and 7050 (μm), respectively. With the progression of sm invasion, the incidence of surrounding elevation and destructive pattern were increased, resulting in being excluded SM2. Conclusion: Although accurate estimation of SM1 was difficult by chromoendoscopy, PPV of endoscopic diagnosis for M-SM1, possible candidate for ER, was satisfactory with excluding surrounding elevation specific to SM2 lesions. And estimating cancer with focal infiltration might be difficult endoscopically since endoscopic findings depended on width and pattern of the progression.