The Usefulness of Endoscopic Ultrasonoraphy for Discriminating Gastric Mucosal Cancer from Submucosal Invasion / 대한소화기내시경학회지

BACKGROUND/AIMS: Endoscopic ultrasonography (EUS) has been used to discriminate gastric mucosal cancer (T1m) from submucosal invasion (T1sm). Thus the aims of this study are 1) to determine the accuracy of EUS for diagnosing tumor depth, 2) to compare the accuracy of EUS with the endoscopic impressions of variously experienced endoscopists and 3) to compare the accuracy of performing EUS by one doctor according to the experience. METHODS: The EUS and pathologic reports of early gastric cancer patients were analyzed. The same endoscopic images were reviewed again by 3 endoscopists, who had one-, three- and five-years experience, respectively. The accuracies of EUS and conventional endoscopy were analyzed. RESULTS: 77 patients were included from November 2003 to October 2005. The kappa of the EUS for actual examiner and conventional endoscopy for reviewer 1, reviewer 2 and reviewer 3 were 0.421, 0.134, 0.359 and 0.307, respectively and accuracies were 68.8%, 45.5%, 67.5% and 62.3%, respectively. Of the 52 T1m patients, 23 (44.2%) were overstaged as T1sm with performing EUS. But of 25 T1sm patients, only 1 (4.0%) was understaged as T1m with performing EUS. The accuracy and kappa of the EUS for one doctor during the first-year experience were 60.6% and 0.316, respectively, and they were 75.0% and 0.508, respectively during the second-year experience. CONCLUSIONS: EUS is useful for complementing the conventional endoscopic discrimination of gastric mucosal cancer from submucosal invasion. Yet physician should keep in mind the relatively common overstaging.

Diagnosis and Treatment of Depressed Colorectal Neoplastic Lesion

PURPOSE: Depressed colorectal cancer is a newly recognized colorectal cancer. It has the characteristics of rapid growth and early invasion of the submucosa. Accordingly, recognition of that lesion is important. However, it is still rarely detected in Korea. This study was designed to evaluate the characteristics of depressed colorectal neoplastic lesions. METHODS: We experienced 22 cases of depressed neoplastic lesions from January 1997 to December 1998. All of them were detected by performing colonoscopy. Among them, 6 were early colorectal cancers. The twenty-two cases accounted for 1.3% of all neoplastic lesions but advanced colorectal cancers encountered during the same period, and the six accounted for 6.6% of all early colorectal cancers during that period. We reviewed and analyzed those 22 lesions with respect to their clinicopathologic characteristics, especially size and histology. RESULTS: The most common age group was the 6th decade. The male-to-female ratio was 2.7 to 1. The predilection of sites were the descending colon, the transverse colon, and the sigmoid colon in that order. The most common size was 3~4 mm, 9 lesions (40.9%) and the next was 5~6 mm, 7 lesions (31.8%). Twenty lesions (90.9%) were 8 mm or smaller in size. The overall malignancy rate was 27.3% (6/22), comprising 9.1% (2/22) for mucosal cancers, and 18.2% (4/22) for submucosal ones. The two lesions which were larger than 10 mm were submucosal cancers. Endoscopic mucosal resection (EMR) was the most common type of treatment, accounting for 59.1%. Two submucosal cancers and one mucosal cancer were operated on without any endoscopic treatment. That one mucosal cancer had initially been suspected of being a submucosal one upon endoscopic examination. There were neither complications nor recurrences during the average 10-month follow-up. CONCLUSIONS: The target for detecting and treating depressed colorectal cancer should be lesions below 10 mm in size, and the treatment of choice should be EMR.