Usefulness of endoscopic ultrasonography in endoscopic submucosal dissection Jeong Hoon Lee and Hwoon-Yong Jung Department of Gastroenterology, University of Ulsan, Asan Medical Center, Seoul, Korea See article in J. Gastroenterol. Hepatol. 2010; 25: 1747–1753. Endoscopic mucosal resection (EMR) is now firmly established as a treatment approach for gastric neoplasms, particularly early gastric cancer (EGC). It is an organ-saving method that is less invasive than surgical resection. Moreover, it can provide a concise pathological diagnosis that allows prognosis to be predicted. With the aid of instrumental developments, such as an electrosurgical knife, a more precise endoscope, and high-frequency electrosur- gical current generator, endoscopic submucosal dissection (ESD) enables dissection of deeper tissue layers. Further, ESD has been reported to be superior to EMR for en bloc resection and local recurrence rates. 1 In fact, patients with EGC treated by ESD expe- rienced a 100%, 5-year disease-specific survival rate. 2 Despite the above-mentioned advantages of ESD, complications, such as bleeding and perforation, are more prevalent than for EMR. 1,3 Many endoscopists have advocated the expansion of indications for ESD. There are two approaches for this. One way is to maxi- mize the inclusion criteria. Beyond the well-known extended cri- teria for ESD by Gotoda, 4 signet-ring cell carcinoma and poorly- differentiated adenocarcinoma of the stomach remain a therapeutic challenge. 5 The other approach is the minimization of exclusion criteria. Bleeding and perforation are the main obstacles that need to be treated for the popularization of ESD. Although a recent Korean study reported rates of delayed bleeding, significant bleed- ing, perforation, and surgery related to a complication were 15.6%, 0.6%, 1.2%, and 0.2%, respectively, 6 and these complication rates were higher for inexperienced operators. For example, one report showed bleeding and perforation rates up to 57% and 65%, respec- tively, during gastric ESD in a swine model for beginners; clearly there is a steep ‘learning curve’. 7 Some intraoperative bleeding is usual during ESD; however, minor bleeding is not counted as a complication. Although minor bleeding is not essential for vitality, it is associated with poor visual field, delayed operation time, and sometimes, unexpected perforation. Understanding the anatomy of vessels, techniques to reach the deep submucosal layer, frequent precoagulation (which is tediously slow, but like the tortoise, an eventual winner), and proper hemostasis techniques are imperative to avoid unnecessary intraoperative bleeding. 8 Prior to the determining the suitability of gastric lesions for ESD, the depth of invasion is an important factor. Endoscopic ultrasonography (EUS) is a useful diagnostic modality in various parts of the gastrointestinal tract and neighboring organs, and the most accurate method for T and N staging of upper gastrointestinal malignancy. Numerous studies have demonstrated the superiority of EUS over other modalities. 9 Recently, clinicians have made great efforts to make diagnoses more precisely and to minimize ‘blind spots’. Hwang et al. demonstrated that the overall accuracy of EUS for T and N staging was 62% and 66%, respectively. 10 In their study, EUS showed poor accuracy (31%) for the ulcerative type of EGC. However, another study reported usefulness in the differential diagnosis between benign and malignant gastric ulcers. Thus, Zhang et al. reported that the sensitivity of EUS was 84%, the specificity was 63%, and the accuracy was 72% for the diagnosis of malignant ulcers. 11 In this issue of the Journal of Gastroenterology and Hepatol- ogy, Kuroki et al. introduces EUS to the field of ESD. 12 They performed EUS before ESD and evaluated the submucosal vascu- lar structures using objective and reproducible criteria that included an abundance of vasculature and large-diameter vessels. The outcomes of intraoperative bleeding, which were hardly expected, as they mentioned, were assessed by a median fall in hemoglobin, procedure time, the use of clips, and the restarting of food on the postoperative day. The patients with rich submucosal vascularity showed a higher hemoglobin reduction rate (5.8% vs 3.5%), longer procedure time (151 min vs 100 min), and a greater use of the clip (79% vs 32%). A multivariate analysis revealed that submucosal invasion and the use of the clip were independent factors. The authors concluded that identification of the submu- cosal vascular structure by EUS might help predict the risk of intraoperative bleeding and the safety of ESD. One logistic issue is the subjectivity of diagnostic criteria of EUS on the submucosal vascular structure. The authors stated that the use of color Doppler might be beneficial to confirm structures as vessels. The development of a mini probe with color Doppler would be useful for the differential diagnosis of lesions with small caliber. With this device, we anticipate concurrent EUS with color Doppler during ESD and a prediction of an abundance of submu- cosal vascularity. This method might expand the safety of ESD use by reducing intraoperative bleeding. Previously, two Korean studies 13,14 reported on the predictors of intraoperative bleeding during gastric ESD. Jang et al. reported that the only factor that correlated with an ‘increased risk’ of bleeding with ESD was the presence of gastric malignancy. 13 Jeon et al. demonstrated that older age and lesions located in the antrum were associated with a ‘lower frequency’ of bleeding. 14 These clinical findings might be associated with vascular factors; the vasculature of malignancies is more tortuous and abundant than that of benign lesions. Moreover, submucosal arteries of the upper third of the stomach are larger than in other areas. 8 Therefore, Kuroki et al. revealed this correlation as a model using EUS. 12 One of the limitations of ESD is its technical difficulty. Endo- scopists performing ESD need to develop the ability to diagnose margins of the lesion and to perform hemostasis perfectly. Many endoscopists will want to learn how to perform ESD; however, training in an apprentice system is required. Most beginners start ESD at the lower part of stomach, because this part has less Accepted for publication 3 August 2010. Correspondence Dr Hwoon-Yong Jung, Department of Gastroenterology, University of Ulsan college of Medicine, Asan Medical Center, 388-1 Pungnap 2-dong, Songpa-gu, Seoul, Korea 138-736. Email: hyjung@amc.seoul.kr JH Lee and H-Y Jung Editorials 1715 Journal of Gastroenterology and Hepatology 25 (2010) 1713–1717 © 2010 Journal of Gastroenterology and Hepatology Foundation and Blackwell Publishing Asia Pty Ltd