Feasibility of electrocautery snaring as the final step of endoscopic submucosal dissection for stomach epithelial neoplasms
O. Goto,M. Fujishiro,S. Kodashima,N. Kakushima,S. Ono,N. Yahagi,M. Omata
DOI: https://doi.org/10.1016/j.dld.2008.02.023
IF: 5.165
2009-01-01
Digestive and Liver Disease
Abstract:Results En bloc resection rate was significantly lower and delayed bleeding rate was significantly higher in the snaring group than in the non-snaring group (91.1% [41/45] vs. 100% [154/154], 11.1% [5/45] vs. 1.9% [3/154], respectively), although complete resection rate (86.7% [39/45] vs. 92.9% [143/154]) and mean operation time (70.2 min vs. 75.8 min) were not significantly different between the two groups. Six perforation cases (3 [6.7%] in the snaring group, 3 [1.9%] in the non-snaring group) were observed, but snaring did not lead to perforation in any case. When the subjects were divided into small (≤2 cm) and large (>2 cm) tumours, en bloc resection rate in large tumours was still significantly different between the groups (76.9% [10/13] vs. 100% [67/67]), whereas in small tumours it was no longer significantly different (96.9% [31/32] vs. 100% [87/87]). Conclusions Snaring may facilitate successful ESD for smaller tumours, but multiple-piece resection should be taken into account especially for larger tumours. Keywords Endoscopic submucosal dissection Piecemeal resection Snaring resection Stomach neoplasm 1 Introduction Endoscopic submucosal dissection (ESD) is a recently developed endoscopic treatment used mainly for intramucosal neoplasms of the gastrointestinal tract. It is characterised by a circumferential mucosal incision and submucosal dissection beneath the lesion. ESD has an invaluable advantage over the other treatment modalities, because the target lesion can be endoluminally resected in one piece without organ resection, even if the lesion is large or associated with submucosal fibrosis [1,2] . However, ESD is a highly advanced technique among the various endoscopic treatments, and special expertise in addition to training under the supervision of experienced hands is preferable. Among the several steps involved in ESD, the most difficult step is considered to be submucosal dissection, that, if it does not go as planned, may result in active bleeding or perforation. In the clinical scene of ESD, snaring in a half way, where en bloc resection may be possible after circumferential mucosal incision and an appropriate amount of submucosal dissection, can be performed as a substitute for completion of submucosal dissection. Otherwise, in the case of difficulty in continuing the submucosal dissection due to technical problems, we are unwilling to use a snare after proceeding with submucosal dissection as much as possible. It is thought that using a snare may shorten operation time spent on submucosal dissection or rescue the patient from a difficult situation, but it may lead to uncertainty for en bloc resection or to the possibility of an increased tumour-positive margin. So, we retrospectively analysed differences in the outcomes of ESD for stomach epithelial neoplasms between the groups with and without electrocautery snaring as the final step of ESD, in order to find out whether snaring has been beneficial or not in our case series. 2 Patients and methods One hundred and ninety-nine consecutive lesions with a diagnosis of early gastric cancer (EGC) or gastric adenoma (GA) between January 2004 and May 2007 were retrospectively investigated. These cases had all been resected by four, very experienced ESD experts, each of whom had performed ESD on more than 30 cases of EGC or GA. Lesions in a remnant stomach after gastrectomy or in a gastric tube after oesophagectomy were excluded, because the number was small and the specific conditions might affect subsequent analysis. Cases whose medical records were insufficient for retrospective analysis were also excluded. Clinicopathological features of 199 lesions were divided into two groups, the snaring group ( n = 45) and the non-snaring group ( n = 154), according to the requirement of an electrocautery snare for tumour resection ( Table 1 ). Age, sex, location and circumference of the tumour, gross type, histology type, coexistence of submucosal fibrosis, depth of invasion, and mean tumour size (the greatest diameter of the lesion on actual measurement) were not significantly different between the two groups, although the cases in the non-snaring group tended to be located predominantly in the lower third and be larger in comparison with those in the snaring group. The indication for ESD was determined by preoperative prediction of GA or EGC without nodal metastases described by the previous study [3] from thorough endoscopic examinations with endoscopic biopsy. The ESD technique has been described elsewhere [4,5] . In brief, a flex-knife (KD-630L; Olympus, Tokyo, Japan) [6,7] was used as the main electrosurgical knife, and other knives, such as an insulation-tipped diathermic knife (IT knife) [8] and a hook-knife [9] , were used when required by the lesion. These knives were used for mucosal cutting to isolate the lesion from surrounding non-neoplastic mucosa, and for submucosal dissection to detach the lesion from the muscle layer. A mixture of 10% glycerin plus 5% fructose and 0.9% saline preparation (Glyceol, Chugai Pharmaceutical Co., Tokyo, Japan) containing 0.005% indigo carmine and 0.0005% epinephrine was used to make a submucosal fluid cushion [10] . Hyaluronic acid was added to the injection solution when the lesion had ulcerative findings or when it was in a difficult location, such as the cardia [11] . Haemostatic forceps (HDB2422W; Pentax, Tokyo, Japan) were used for reduction in bleeding during the procedure or for the treatment of visible vessels on the mucosal defect after resection. During submucosal dissection, an electrosurgical snare 15 mm (SD-210L-15, Olympus) or 25 mm (SD-210L-25, Olympus) in diameter was used when the operator required. The decision as to whether a snare was used or not was completely left to the judgment of the operator in consideration of several factors, e.g., operation time, technical difficulty, encountering complications, patient's comorbidity ( Table 2 ). We investigated the differences between the snaring and the non-snaring groups in rates of en bloc resection and complete resection (the rate of cases revealing no tumour on the edge of the one-piece resected specimen), mean operation time and complications in an overall analysis. Moreover, as shown in Tables 3 and 4 , all the subjects were divided into two groups according to tumour size (small [≤2 cm] or large [>2 cm]), to compare the outcomes of the two groups for a sub-analysis, because the tumour size might affect the results. This was of particular concern in the snaring group because a snare might limit the ability to resect in terms of its size. For the statistical analyses, Student's t -test for age, operation time and tumour size, Chi-square test for location, circumference, gross type and histology type and Fisher's exact probability test for the other variables were used. A p -value < 0.05 in each analysis was considered to be statistically significant. 3 Results The overall results for using a snare in the ESD procedure are summarised in Table 5 . En bloc resection rate was significantly lower in the snaring group than in the non-snaring group (91.1% vs. 100%). There were no significant differences in complete resection rate and mean operation time between the two groups. One of the major complications, delayed bleeding rate, was significantly higher in the snaring group than in the non-snaring group (11.1% vs. 1.9%). In all six cases, perforation was occurred during direct submucosal dissection and were not due to snaring. Table 6 shows the results of the sub-analysis between the snaring and the non-snaring groups according to tumour size. In large tumours, en bloc resection rate in the snaring group was still significantly lower (76.9%) than in the non-snaring group (100%), whereas in small tumours, it was not significantly different between the two groups. In terms of operation time, there was no significant difference between the snaring and the non-snaring groups regardless of small or large tumour, although large tumours needed significantly longer time than small tumours in both groups. It was no longer significantly different in major two complications between the snaring and the non-snaring groups in this sub-analysis. 4 Discussion As tips and tricks for ESD have accumulated, performing ESD has gradually permeated into the repertoire of surgical techniques, especially in Japan [12] . The technique has significant advantages over other endoscopic treatments in the areas of controlling the shape and size of the resected specimen and the high probability of en bloc resection of an entire lesion. However, ESD is sometimes very difficult to perform and the operator must overcome many hurdles during the procedure. The snaring technique is a basic one used in polypectomy or endoscopic mucosal resection (EMR) for pedunculated or small flat lesions [13] . Although snaring is easily performed and considered to save time, the disadvantage is uncertainty of en bloc resection [1,14] . When snaring is applied during ESD procedure, we should always consider this disadvantage; if snaring causes multiple-piece resection, the original advantage of ESD, the precise histopathological evaluation of an entire lesion, will be lost. Why, then, is a snare used as the final step in ESD? There are various reasons. The most frequent one may be its time-saving advantage. However, Table 5 shows that snaring did not shorten operation time in our case series. Then, considering these results, does snaring actually shorten the operation time? The answer should be absolutely “yes”. In some cases, a snare was used because it took an unexpectedly long time to dissect the submucosal layer beneath the tumour. In other cases, it was used because continuation of the submucosal dissection until complete detachment of the lesion from the muscle layer was impossible. So, when the results obtained from the overall analysis are considered in their clinical context, we are convinced that snaring absolutely shortens the procedure. If snaring had been planned beforehand for time-saving purposes, then normal operations that had not involved these extenuating circumstances would have been finished in much shorter time. Although the overall analysis shows that en bloc resection rate was significantly lower in the snaring group than in the non-snaring group, the sub-analysis according to tumour size yielded quite important evidence for us. The results indicated that almost all small tumours (2 cm as a cut-off line between small and large tumours) could be removed in en bloc fashion. On the other hand, the rate of en bloc resection of large tumours in the snaring group was significantly lower than that in the non-snaring group. Furthermore, the larger the tumour size is, the longer operation time is. A dilemma emerges in the tradeoff between operation time and possibility of en bloc resection, especially for large tumours. It is reported that piecemeal resection leads to non-curative treatment, causing non-evaluable histopathology of the tumour and potentially local recurrence [1] . Even though it takes longer time to resect, the first priority should be curative resection of the tumour for the cases of no complications. So, in conditions where long operation time does not become a disadvantage for patients, such as under general anaesthesia, we recommend to accomplish complete submucosal dissection by an electrosurgical knife instead of using a snare. However, this consideration should be withdrawn when poor conditions during the operation occurred, and snaring should be used as a salvage technique, considering risks and benefits of the patient. Another possible point at issue of increasing the number of incomplete resections due to a burning effect of snaring for margins of resected specimens was not observed in our case series, which may surely encourage us to use a snare when en bloc resection is undoubtedly possible, in order to shorten operation time. In terms of complications, delayed bleeding frequently occurred when a snare was used. In the “snaring” series, a snare was used in some cases of declared technical difficulties, presence of comorbidity or occurrence of complications. These variables might have determined the higher rate of delayed bleeding observed in this subgroup of patients, although the reason for this is still unclear. Another explanation is that snaring might cause insufficient coagulation of submucosal vessels, which implies that non-bleeding visible vessels should be more intensively treated after resection of the tumour with a snare, especially in the area detached by a snare. Although perforation occurred in six cases, the causes for perforation were not related to snaring, but were due to inappropriate submucosal dissection by electrosurgical knives in all cases. On the contrary, an appropriate use of a snare may decrease perforation rate owing to reduction in time spent on direct submucosal dissection. In summary, we suggest that lesions >2 cm in size should be resected in complete submucosal dissection style without snaring, in order to avoid multiple-piece resection, and that lesions ≤2 cm in size may be treatable through snaring in order to reduce operation time. However, the shortcoming of this report was a retrospective, non-controlled setting, even though the results were obtained from the consecutive data. A prospective, randomised controlled study about the feasibility of “planned” snaring is desirable to confirm our results, especially for smaller tumours. Practice points • Snaring resection may be acceptable for small lesions (≤2 cm) as the final step of endoscopic submucosal dissection (ESD) for stomach epithelial neoplasms to save operation time. • Complete submucosal dissection by an electrosurgical knife is recommendable for large tumours (>2 cm) to avoid multiple-piece resection. Research agenda • A prospective, randomised trial is favourable to elucidate the feasibility of snaring as the final step of ESD, especially for small lesions. Conflict of interest None declared. References [1] T. Eguchi T. Gotoda I. Oda H. Hamanaka N. Hasuike D. Saito Is endoscopic one-piece mucosal resection essential for early gastric cancer? Dig Endosc 15 2003 113 116 [2] M. Fujishiro O. Goto S. Kakushima S. Kodashima Y. Muraki M. Omata Endoscopic submucosal dissection of stomach neoplasms after unsuccessful endoscopic resection Dig Liver Dis 39 2007 566 571 [3] K. Gotoda A. Yanagisawa M. Sasako H. Ono Y. Nakanishi T. Shimoda Incidence of lymph node metastasis from early gastric cancer: estimation with a large number of cases at two large centers Gastric Cancer 3 2000 219 225 [4] H. Ono H. Kondo T. Gotoda K. Shirao H. Yamaguchi D. Saito Endoscopic mucosal resection for treatment of early gastric cancer Gut 48 2001 225 229 [5] T. Gotoda H. Yamamoto R.M. Soetikno Endoscopic submucosal dissection of early gastric cancer J Gastroenterol 41 2006 929 942 [6] N. Yahagi M. Fujishiro N. Kakushima A. Imagawa M. Iguchi A. Okawa Merits and demerits of endoscopic submucosal dissection using a flex-knife Stomach and Intestine 39 2004 39 43 [in Japanese with English abstract] [7] S. Kodashima M. Fujishiro N. Yahagi N. Kakushima M. Omata Endoscopic submucosal dissection using flexknife J Clin Gastroenterol 40 2006 378 384 [8] T. Gotoda H. Kondo H. Ono Y. Saito H. Yamaguchi D. Saito A new endoscopic mucosal resection procedure using an insulation-tipped electrosurgical knife for rectal flat lesions: report of two cases Gatrointest Endosc 50 1999 560 563 [9] T. Oyama A. Tomori K. Hotta S. Morita K. Kominato M. Tanaka Endoscopic submucosal dissection of early esophageal cancer Clin Gastroenterol Hepatol 3 2005 S67 S70 [10] M. Fujishiro N. Yahagi K. Kashimura Y. Mizushima M. Oka S. Enomoto Comparison of various submucosal injection solutions for maintaining mucosal elevation during endoscopic mucosal resection Endoscopy 36 2004 579 583 [11] M. Fujishiro N. Yahagi M. Nakamura N. Kakushima S. Kodashima S. Ono Successful outcomes of a novel endoscopic treatment for GI tumors: endoscopic submucosal dissection with a mixture of high-molecular-weight hyaluronic acid, glycerin, and sugar Gastrointest Endosc 63 2006 243 249 [12] A. Repici From EMR to ESD: a new challenge from Japanese endoscopists Dig Liver Dis 39 2007 572 574 [13] M. Hirao K. Masuda T. Asanuma H. Naka K. Noda K. Matsuura Endoscopic resection of early gastric cancer and other tumors with local injection of hypertonic saline-epinephrine Gastrointest Endosc 34 1988 264 269 [14] Y. Yamaguchi N. Katsumi K. Aoki M. Toki K. Nakamura N. Abe Resection area of 15 mm as dividing line for choosing strip biopsy or endoscopic submucosal dissection for mucosal gastric neoplasm J Clin Gastroenterol 41 2007 472 476