© Springer Science+Business Media New York 2015
Norio Fukami (ed.)Endoscopic Submucosal Dissection10.1007/978-1-4939-2041-9_1717. Management of Gastrointestinal EMR and ESD Perforation: From Lab to Practice
(1)
Department of Gastroenterology, Hepatology, and Nutrition, The University of Texas MD Anderson Cancer Center, 1400 Pressler Street—Unit 1466, Houston, TX 77030, USA
Keywords
Endoscopic submucosal dissectionGastrointestinal perforationOver-the-scope clipsThrough-the-scope clipsEndoscopic suturingIntroduction
At the beginning of the twenty-first century, endoscopists started exploring curative resection of early stage gastrointestinal neoplasms (defined as localized disease without lymph node or distant metastases) as an alternative to surgery. Although polypoid neoplasms can be removed with a biopsy or snare resection, these techniques are unsuitable for complete and safe resection of large non-polypoid lesions [1]. Endoscopic submucosal dissection (ESD) and endoscopic mucosal resection (EMR) allow complete resection of such flat lesions, as they utilize submucosal fluid injection to lift the lesion and make it accessible for complete resection. However, perforation is a serious risk as these procedures involve cutting the mucosa and submucosa while sparing the deep submucosa and muscularis propria of the rather thin gut wall. Advances in clinical practice as well as investigative work in the animal laboratory have helped us to develop techniques for successful endoluminal closure of perforations. A detailed review of the literature will be presented to help the reader who is interested in EMR and ESD to gain further insight on this topic.
ESD Versus EMR: An Overview
Credit goes to the Japanese endoscopists for their enthusiasm to develop endoluminal resection techniques to avoid gastrectomy in patients with early gastric cancer. En bloc resection of larger lesions can be accomplished with ESD, while the same lesion could only be removed in a piecemeal fashion, instead of en bloc, with EMR. However, piecemeal resection poses a risk of recurrence as shown in a Japanese study, where gastric cancer recurred in 2.8 % of patients after piecemeal resection of >15 mm early gastric cancer lesions compared to no recurrences in patients treated with en bloc resection [2]. Other studies have described even higher rates of cancer recurrence of lesions treated with EMR; for instance, Saito et al. reported higher local cancer recurrence rate of after colorectal EMR compared to ESD (14 % vs 2 %; p < 0.0001) [3]. Still others have quoted a range of 2 % up to 35 %, from a variety of studies on endoscopic resection of gastric cancer [4]. In a meta-analysis of 15 published studies, ESD was associated with a higher rate of en bloc resection (OR 13.87, 95 % CI 10.12–18.99) and curative resection (OR 3.52, 95 % CI 2.57–4.84) compared to EMR [5].
Although EMR and ESD offer the advantage of complete resection of early neoplasms and a less invasive alternative to surgery, both are associated with risk of complications such as bleeding, perforation, and stricture. The ESD perforation rate is significantly higher than that of EMR (4–10 % vs. 0.3–0.5 %) [6]. A single-center, single-operator Japanese study of 1635 early gastrointestinal neoplasms of the esophagus, stomach, and colon reported ESD perforation rates of 0 %, 1.8 %, and 1.9 % respectively [7].
Identification of Perforation and Tension Pneumoperitoneum
Perforation is a serious complication of ESD. Perforations can be classified by size: microperforations and macroperforations, and by time of onset: immediate and delayed [8].
Microperforations are detected after the procedure as free air on routine post ESD imaging due to the escape of air through invisible perforations in a wall thinned by cautery and dissection.
Macroperforations are obvious to the endoscopist during the procedure and result from inadvertent deep cautery during the incision or dissection phase of ESD or entrapment of muscularis propria during snare resection of EMR. These perforations can be closed with clips, and the resection completed during the same session.
In a retrospective study of 1,711 patients undergoing ESD for early gastric cancers, Jeon et al. described an overall perforation rate of 2.1–3.2 % [9]. A total of 26 macroperforations and 13 microperforations occurred over 5 years. All patients with microperforations recovered with conservative management, but one patient with a macroperforation required emergent surgery. Though ESD was successfully completed in all patients after clip closure of macroperforations, the rate of incomplete resection and recurrence of tumor was greater among this group of patients, though not statistically significant. Perforation is an inherent risk in performing ESD, and the identification and rapid management of perforation is of utmost importance while maintaining a high quality resection.
Tension Pneumoperitoneum results from the rapid escape of air through a perforation. Although tension pneumoperitoneum from escaped luminal gases can occur, it has become infrequent since the use of CO2 for insufflation during EMR and ESD has become routine. Immediate abdominal decompression using an 18-gauge needle puncture is required to relieve tension pneumoperitoneum. Subsequently, patients should be kept nil by mouth, adequately hydrated with intravenous fluids, and started on broad-spectrum intravenous antibiotics. Patients who do not improve with these clinical interventions may require emergency surgery for repair [8, 9].
Risk Factors for ESD Perforation
Risk factors for ESD perforation include some related to the operator and others related to the lesion itself.
Operator-related factors include utilization of a precise endoscopic technique, adequate experience with ESD, and the volume of ESD performed at a particular center.
Lesion-related factors include the size and the luminal distribution of the lesion and its location, whether esophagus, stomach, duodenum, or colon. Large lesions and lesions that are difficult to access are at a higher risk for complications.
Endoscopic Closure Options
Binmoeller and colleagues were the first to describe the use of endoclips to close a gastric perforation after EMR of a leiomyoma in the early 1990s [10]. The ability to immediately close perforations endoscopically often spares the patient a surgical intervention and all the intrinsic risks and costs associated with it. Several models of through-the-scope clips, over-the-scope clips, and endoscopic suturing device prototypes are available on the market for endoscopic closure of perforations.
Through-The-Scope Clips
Through-the-scope clips (TTSC) were initially introduced in Japan in the 1970s to mark lesions and achieve hemostasis [10]. Currently available TTSC products include: Quick Clip (Olympus America Inc., Center Valley PA), which is free 360° bidirectionally rotatable; Resolution Clip (Boston Scientific Inc., Natick MA), which allows reopening prior to final clip deployment; and Instinct Clip (Cook Medical Inc., Bloomington IN), which has 360° bidirectional rotation as well as reopening capability.
TTSCs have been used in a variety of clinical situations, including closure of gastrointestinal fistulae and leaks as well as spontaneous and iatrogenic perforations. TTSCs can be used to successfully manage both fresh perforations and chronic fistulae and, with appropriate technical expertise, even close defects up to 25 mm in size [11]. Limitations of these TTS endoclips include their limited wing-span when open and their lower closure force that can lead to sub-optimal tissue apposition and often necessitate the placement of multiple endoclips [12]. The larger endoclips offer greater tissue grasp and a better leak proof seal, but may interfere with continuation of the ESD procedure.
Over-The-Scope Clips
The introduction of over-the-scope clips (OTSC) in 2007 helped to overcome some of the limitations of through-the-scope endoclips. OTSC are composed of nitinol (nickel titanium alloy) and are housed within a metal applicator cap that is fitted over the distal tip of the scope (Ovesco Endoscopy AG, Tubingen Germany). These clips are deployed by the release of a wire attached to a hand wheel mounted on the scope at the biopsy valve, in a mechanism similar to esophageal variceal rubber band ligation. Caps and clips are available in three sizes to accommodate different endoscope diameters. OTSC are offered in two depths (3 mm and 6 mm), to vary the amount of tissue that can be grasped within the clip [13]. The clips have different shapes of teeth, rounded, pointed, and longer pointed, and the clinical indication for using OTSC will determine which clip configuration is used. The rounded teeth OTSC are used for hemostasis and are especially helpful in the esophagus and colon where the organ wall is thinner. The pointed teeth clip has the ability to grasp tissue with minimal slippage and can be useful for closing perforations and fistulae. The longer pointed teeth clip is effective in the thicker-walled stomach [13]. Since OTSC are able to grasp more tissue, they produce a more durable closure than TTSCs and can apply eight to nine Newtons of permanent closing force [14]. The U.S. Food and Drug Administration approved the use of OTSCs in 2010.
The initial clinical experience with OTSC was described by Kirschniak et al. when they used OTSC to achieve hemostasis in seven patients and successfully closed iatrogenic perforations in four patients [14]. Investigators have subsequently used OTSC to treat post-surgical leaks and chronic fistulae, and have diversified the clinical experience in closing iatrogenic perforation and bleeding with varying degrees of success [12, 15, 16]. Baron et al. described the clinical experience using OTSC in the first multi-center study in the United States [17]. OTSC were successful in closing 65 % of fistulae and post-surgical leaks and 75 % of iatrogenic perforations and achieved hemostasis in all seven patients. One clinical failure in this study was due to shifting of the OTSC and tissue away from the gastro-cutaneous fistula site. The clip was sectioned using argon plasma coagulation on the highest setting to remove it from the tissue [17]. In a recent systematic review, Weiland et al. reviewed 17 published clinical studies that involved closure of gastrointestinal perforations and leaks by OTSC [18]. The intervention was technically successful between 80 and 100 % of the time, and clinically successful (when no further intervention was necessary) 60–100 % of the time. The failure of OTSC was associated with necrotic tissue around the wound edges that prevented adequate clip deployment [18].
Endoscopic Suturing
The U.S. Food and Drug Administration approved an endoscopic suturing device in 2000 for gastroesophageal reflux disease [19]. Much of the experience with endoscopic suturing devices comes from experimental models that were developed for natural orifice translumenal endoscopic surgery (NOTES). In NOTES, the reliable closure of controlled perforations made in the wall of the gastrointestinal tract is essential.
Organ-Specific Endoscopic Closure
Esophagus
The application of the aforementioned techniques for managing esophageal, gastro-duodenal, and colonic perforation after EMR and ESD will now be discussed, as well as the animal work that laid foundation for development of endoscopic closure techniques in clinical practice.
Experimental Studies
Endoscopic suturing has been evaluated in experimental studies. Fritscher-Ravens et al. have investigated the role of endoscopic suturing to close iatrogenic defects created in the esophagus [20]. One such study was performed to compare different techniques of iatrogenic esophageal perforation closure. The investigators created a small esophageal full-thickness perforation in 18 pigs and randomized them to closure with TTSC, endoscopic suturing with a custom-made suturing system, or standard thoracoscopic closure. Each of the study animals had a technically successful closure. One animal in the suturing group had a mediastinal abscess and another animal in this group and in the surgical group expired prematurely. Closure with TTSC was statistically significantly faster than either of the other groups (9 min vs. 21 min in suturing group and 42 min for surgery, p = 0.006). On necropsy, peri-esophageal adhesions were most pronounced in the suturing group. In this small study, although TTSC, suturing and surgery were all comparable, the animals treated with endoclips had a better outcome.
Another NOTES study involving trans-esophageal mediastinoscopy by these investigators revealed no adhesions and on histological examination on necropsy; sutured closure had normal wall healing while there was disruption of the muscle layer but healing of the mucosal and submucosal layers after TTSC closure [21].
In another study, where a 2 cm full-thickness resection of the esophagus followed by closure with a prototype endoscopic suturing device, necropsy after 3 months revealed no complications related to incision, resection, or closure; there was no evidence of mediastinitis and all but one animal had well-healed scars [22]. This study demonstrates the feasibility of successful suture closure of full-thickness resection of a segment of the esophagus in an animal model.
Clinical Studies
Esophageal perforations can occur in a variety of other clinical settings, including spontaneous rupture, Boerhaave’s syndrome, iatrogenic rupture (after endoscopic dilation, EMR, or surgical intervention), or secondary to foreign body ingestion [23, 24]. The treatment modality of choice is usually TTSC as these clips are readily available in the endoscopy suite and most gastroenterologists have experience using them. In a multi-center European cohort study, Voermans et al. described five cases where iatrogenic esophageal perforations were treated with OTSC. All five lesions were successfully closed. Three patients required a combination of OTSC and TTSC placement, while the other patients each had two OTSC deployed to close the defect [25]. Nishiyama et al. described one elderly patient who experienced an iatrogenic perforation of the distal esophagus after stomach feeding tube placement that was successfully treated with a 9 mm OTSC [26].
ESD has traditionally been used to treat superficial squamous cell carcinoma in the esophagus. ESD esophageal perforation rates vary among the published clinical studies but are generally below 10 % [27, 28]. Fujishiro et al. published some of the earliest esophageal ESD experience with their study of 58 squamous cell carcinoma lesions in 43 patients performed between 2002 and 2005. An R0 resection was achieved in 78 % of lesions, which ranged between 2 and 66 mm in size. The majority of the lesions were in the thoracic esophagus and occupied less than half of the luminal circumference. In this particular study, perforation occurred in 7 % of cases (4/57) and was detected during the course of ESD and immediately closed with endoscopic clips. All patients demonstrated pneumomediastinum on chest X-ray and were treated conservatively with good clinical improvement and resolution of pneumomediastinum within 1 week. Pneumomediastinum was not seen in any patients without perforation noted during endoscopy [27]. Others have reported pneumomediastinum on post-procedure X-ray, without detection of perforation during ESD [29, 30]. In another study, a perforation was detected and presumably treated (likely with TTSC, though this is not certain) with good result [31].
Yamashina et al. described their ESD experience on esophageal lesions greater than 50 mm in diameter [32]. In 39 patients, the en bloc resection rate was 100 %, with tumor-free margins achieved in 92 %. The procedure was curative in 70 % with a complication rate of 2.5 %. One patient experienced mediastinal emphysema without perforation (microperforation), and stricture developed in 11 of 39 patients, as would be expected after large ESDs in the esophagus [32].
Sato et al. summarized the clinical presentation and management of esophageal perforations that occurred during or after EMR/ESD [33]. They treated 472 esophageal neoplasms (171 EMR, 306 ESD), and seven patients (1.9 %) experienced esophageal perforation. Three perforations occurred intra-operatively, three during balloon dilation for stricture prevention, and one due to food impaction. All were treated endoscopically and did not require surgery.
Although a great deal of ESD work has been done to treat early squamous cell carcinoma of the esophagus, recently there is growing interest in defining its role in esophageal adenocarcinoma. This condition occurs more commonly in the West where ESD experience is often limited to fewer skilled gastroenterologists. Nonetheless, a few Japanese investigators have described their experience in the literature [34–36]. En bloc resection rates were high (97–100 %) and R0 resection rates ranged between 79 % and 100 %. In these three studies where 112 patients were treated, only one perforation occurred during ESD, and it was successfully treated with TTSC placement [34–36]. Further study is required in the treatment of esophageal adenocarcinoma with ESD.
Stomach
Experimental Studies
Many investigators have described the role of different devices to close gastric perforations in experimental studies. OTSC have the ability to grasp a larger volume of tissue than TTSC and lead to full-thickness, more durable closure due to the greater force of closure. The superiority of OTSC over TTSC in repairing gastrotomies was demonstrated by von Renteln et al. with their in vivo study of 20 female swine [37]. Swine were randomized to gastrostomy repair with either TTSC or OTSC. Four TTSC pigs had a positive leak test after closure, compared to none in the OTSC group. Peri-gastric abscesses were present in two of the OTSC pigs and three of the TTSC pigs, but peritonitis and adhesions were only present in the TTSC group. These differences were not statistically significant, given the small number of animals in the study. It is interesting to note that not all closures with OTSC were full-thickness. 70 % of the closures involved mucosa and submucosa, 20 % also included the muscularis mucosa, and only 10 % involved the serosa; in contrast, 20 % of the closures with TTSC incorporated the mucosa and submucosa while the remainder were mucosal closures alone [37].
In another study, Voermans et al. compared OTSC closure with hand surgical suturing of gastrostomies [38]. Gastrotomies were created with a needle knife puncture followed by dilation with an 18 mm balloon in 26 ex vivo porcine stomachs. Eleven specimens were closed with OTSC and 15 specimens had surgical suturing. All defects were successfully closed; specimens repaired with OTSC showed air leakage at pressures (233 mm Hg, SD 47) that were non-inferior to the gold standard of hand suturing (206 mm Hg, SD 59), thereby confirming the potential role for OTSC gastric perforation closures.
Matthes et al. conducted an ex vivo study to determine maximal closure capacity and pressure threshold for a single OTSC placed over gastrotomies [39]. Full-thickness gastric defects were created measuring 5, 10, 15, 20, and 25 mm and closed with a single OTSC. There was a linear inverse relationship between the size of the defect and the pressure associated with rupture of the closure. Optimal results were seen with the use of a single OTSC for defects up to 15 mm in diameter, and though adequate closure was achieved in 20 mm gastrotomies, the burst pressures were significantly lower, indicating the need for more than one OTSC for larger gastrotomies or supplemental endoclip placement [39].
Zhang et al. studied the feasibility of using OTSC to close perforations created in the gastric fundus of an in vivo canine model [40]. Seven canines underwent needle knife gastrostomy followed by placement of OTSC, which was technically difficult in two canines because of the retroflexed nature of the endoscope. Leak testing was performed with air insufflation and methylene blue solution instilled into the stomach and detected by laparoscopy. A minor leak, defined as a slight detection of methylene blue solution in the peritoneal cavity, was observed in one of the two canines in which OTSC placement was technically difficult. No OTSC-related complications were noted at necropsy. The feasibility and successful placement of OTSC over gastric fundal defects was demonstrated in this small study [40].
Suturing devices have also been used to successfully close gastric defects. Rajan et al. performed endoscopic suturing (Overstitch, Apollo Endosurgery Inc.) to close the mucosal defect after obtaining a full-thickness gastric biopsy using submucosal endoscopy with mucosal flap technique [41]. Defects were successfully closed in all 12 pigs with an uneventful clinical course. Endoscopy at 2 weeks showed stellate scarring without mucosal ulceration, and necropsy showed complete serosal healing in all study animals [41].
Park et al. conducted a study comparing closure of gastrotomies with a tissue apposition system comprised of tissue anchors against surgical repair in an in vivo porcine model [42]. After gastrotomy and peritoneoscopy, 32 pigs were randomized to endoscopic or surgical closure. All repairs were technically successful, recovery times were similar in both groups, and all closures remained secure at necropsy. Two pigs in the endoscopic group died, one from gastric distention and another was euthanized for rectal prolapse, while one pig was euthanized in the surgical group for wound dehiscence. One animal in the surgical group had a leak detected at necropsy and animals in both groups had signs of peritonitis, adhesions, and abdominal abscesses, though adhesions were significantly higher in the surgical group [42]. Since endoscopic treatment was comparable to surgery, this shows the potential use of endoscopic closure in NOTES and ESD.
Other endoscopic closure techniques include threaded tags (T tags) or tissue anchors. Dray et al. conducted an in vivo study comparing histologic healing after closure of gastrotomy using TTSC versus T tags in 12 pigs [43]. Closure was technically successful in all 12 animals followed by an uneventful 2-week follow up. Transmural healing was seen in 75 % (n = 3) of pigs treated with Resolution clips (Boston Scientific Corporation) and in 12 % (n = 1) of the T tag group. The authors concluded that although TTSC only perform mucosal closure, it consistently resulted in layer-to-layer closure and transmural healing. T tags form a tight plication and cause inversion of the gastric edges, which impairs layer-to-layer healing [43]. Sumiyama et al. performed a short-term animal survival study to evaluate the performance of tissue anchors in closure of gastric perforations [44]. These investigators created 12 large gastric perforations (median 3 cm) in 6 pigs and closed them using a flexible needle-catheter tissue-anchoring device. All closures were technically successful and all animals survived for 1 week without complications. At follow up endoscopy and necropsy, all perforations were secure and there were no findings of peritonitis; however, 12.5 % of the tissue anchors had penetrated surrounding organs [44].
Stapling devices have been studied to close gastrotomies in animal models with the goal of using this technology in closing access points in NOTES procedures. Magno et al. did an initial evaluation of the feasibility and effectiveness of a novel stapling device, which delivered a 6 cm long staple line (NOLC60, Power Medical Interventions, Langhorne, PA, USA) [45]. Leak-resistant gastric closure was achieved in all four pigs with an unremarkable clinical course. Necropsy demonstrated full-thickness healing confirmed by histologic examination. However, intra-mural micro abscesses were also seen in the two pigs that underwent the procedure using non-sterile technique [45]. Meireles et al. performed a follow-up study using this automated flexible stapling device (SurgASSIST, Power Medical Interventions, Langhorne, PA, USA), which delivered four rows of staples with each firing of the device and successfully created full-thickness closure of gastric defects in four pigs [46].
Surgical glues have been used with success in the treatment of variceal bleeding, embolization, and fistulae. Ersoy et al. conducted a study comparing suture, endoclip and combination of endoclip and topical cyanoacrylate glue to close gastric perforation in 60 rats [47]. Burst pressure levels were higher in the cyanoacrylate group along with improved histological healing indices (tissue granulation, chronic inflammation, and collagen deposition), however this was counterbalanced by significantly increased adhesion formation in this group [47]. The use of cyanoacrylate glue is a promising technique for closure of gastric perforations in ESD and NOTES.
Experiments using other bioprosthetics to close gastric perforations have been conducted. Agents that are successful in closing gastrointestinal perforations must form an effective seal, allow tissue ingrowth, and tolerate the local environment (for example, low gastric pH) [48]. The Gore Bioabsorbable Hernia Plug, which is made from the biodegradable polymer polyglycolide, trimethylene carbonate (PGA:TMC, W.L. Gore & Associates Inc, Flagstaff, AZ USA), was used to plug a surgically created 1 cm perforation on the anterior wall of the stomach in 12 canines. The bioprosthesis closed the perforation without leak in all animals when ex vivo testing was performed. Animals were sacrificed at different intervals to further examine the closure. At week one, the perforation was closed but there was no tissue ingrowth. However, at weeks four through twelve, the injury site was healed and the luminal portion of the plug had been absorbed. Bioprosthesis seem to offer a durable closure of gastric perforation with physiologic healing of the injury site [48].
Standard surgical treatment of gastro-duodenal perforations involves the placement of an omental patch over the defect. Hashiba et al. conducted one of the first published studies of experimental endoscopic repair with an omental patch in 2001 in ten pigs [49]. These investigators created a gastric perforation in two steps, first using a variceal band ligation cap fitted over the tip of the endoscope and a polypectomy snare to complete a mucosal resection. In the next step, they completed the perforation by repeating the first step with a smaller cap. To seal the perforation, omentum was aspirated into the defect and affixed to the muscularis propria layer with endoclips. Nine pigs had an uneventful clinical course, with an ulcer seen at the repair site on follow-up endoscopy. The omentum was completely adhered to the external gastric wall at necropsy without evidence of peritonitis. Histologic examination revealed complete healing without microabscess. One animal, in which the muscularis propria could not be clearly seen during omental patch placement, died prematurely, and postmortem examination revealed a persistent leak indicative of an unsuccessful repair with subsequent peritonitis [49].
More recently, Dray et al. have conducted experiments with omentoplasty [50]. Nine pigs were assigned to one of three groups: endoscopic full-thickness gastric resection using an EMR kit without closure, similar resection with omental closure (attaching the omentum to the intact gastric mucosa, away from the perforation site) or full-thickness resection using needle knife and balloon dilation followed by similar omental closure. All animals that did not have gastrotomy closure developed peritonitis. Complete healing was seen on endoscopy in all of the remaining animals except for a small ulceration seen in one animal from each of the omental closure groups. No complication was seen at necropsy or on histologic examination in these animals except for one animal in the group resected with the EMR kit that developed an abscess of the omental patch. As omentoplasty is the standard surgical repair for gastro-duodenal perforations and the greater omentum has known antibacterial, hemostatic, angiogenic, and adhesive properties, further studies for endoscopic placement of an omental patch are indicated [50]. As we have described, several novel techniques have been used to close gastric perforation in experimental models. Many of these have translated to clinical application in patients to address gastro-duodenal perforation and closure.
Clinical Studies
ESD has a several-fold increased risk of perforation over EMR of the stomach and duodenum. Gastroenterologists who perform ESD should be aware of the risk factors associated with perforation in the gastro-duodenum. Several investigators have retrospectively evaluated the experience at their centers to determine predictive risk factors for perforation during ESD for these upper gastrointestinal lesions.
Toyokawa et al. performed 1,123 gastric ESD between 2003 and 2010 and described an overall perforation rate of 2.4 % [51]. Lesions located in the upper or proximal stomach had a significantly higher risk of perforation (OR 4.9, 95 % CI 2.2–10.7) than in other parts of the stomach. They also described a lower rate of en bloc resection and a lower curative resection rate among patients who experienced a perforation (74 % versus 94 %, and 48 % versus 85 %, respectively).
Ohta et al., in a different center in Japan, performed 1795 gastric ESDs over an 8 year period [52]. Their overall perforation rate was 2.8 %, and risk factors for perforation again included proximal location of the gastric lesion (OR 2.4, 95 % CI 1.3–4.6). Another significant risk factor was tumor diameter greater than 2 cm (OR 1.9, 95 % CI 1.0–3.5). These investigators noted an improvement in perforation rate over their study period as there were significantly fewer perforations that occurred over the second half of the study period than had happened earlier in the study (OR 4.5, 95 % CI 2.1–9.4) [52].
Kim et al. described their experience with 1289 gastric ESDs between 2003 and 2010 [53]. Their overall perforation rate was 2.7 % and they also found that location of the lesion within the stomach impacted the risk for perforation: lesions in the gastric body had a higher risk for perforation than antral lesions (OR 2.6, 95 % CI 1.3–5.2). Also, lesions that required piecemeal resection had a significantly higher risk of perforation (OR 2.6, 95 % CI 1.0–6.6). The antrum appears to be a safer location to perform ESD as the en face position allows greater access to the lesion. The antral wall is also thicker and has fewer and smaller caliber blood vessels compared to the gastric body. The authors attribute the increased risk of perforation with piecemeal resection to intrinsic features of such lesions, including anatomic location, difficult scope maneuvering, or deeper invasion of the tumor [53].
Watari et al. reported a perforation rate of 8.2 % in 98 consecutive gastric neoplasms that underwent ESD; all patients were managed conservatively. Procedure time greater than 115 min was associated with an increased risk of perforation (OR 9.15, 95 % CI 1.08–77.54, p = 0.04). Subgroup analysis revealed similar post-ESD clinical course regardless of whether perforation was detected endoscopically versus those diagnosed on post-ESD imaging [54].
Investigators have reported not only their rates of perforation, but have further subdivided this into rates of microperforation and macroperforation. The clinical outcomes for patients experiencing micro- or macroperforation are favorable and comparable [9]. Kim et al. report a rate of 2.1 % microperforation and 0.6 % macroperforation [53]. Yoo et al., another group from Korea, performed 823 gastric ESD from 2005 to 2010 [55]. Their rate of microperforation was 2.1 % and was associated with patient age over 81 years (OR 20), tumor depth of invasion into the muscularis mucosa (OR 5.4) and procedure time greater than 2 h (OR 5.9). Macroperforation rate was 7.5 % and was associated with proximal gastric location of lesion (OR 7.9), fibrosis seen during submucosal dissection (OR 3.0) and prolonged procedure time of over 2 h (OR 3.3) [55].
Related posts:
Advanced Endoscopic Imaging in the Lower GI Tract
Submucosal Endoscopy: From ESD to POEM and POET
Regenerative Medicine for Stricture Management: What does the Future Hold?
Advanced Endoscopic Imaging in the Upper Gastrointestinal Tract
Indications of ESD in the Upper Gastrointestinal Tract
Endoscopic Submucosal Dissection for Superficial Esophageal Cancer
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
