Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210

Contents lists available at ScienceDirect

Techniques in Gastrointestinal Endoscopy

0049-01http://d

☆Theinterest

n CorrE-m

journal homepage: www.techgiendoscopy.com/locate/tgie

Therapeutic flexible endoscopy replacing surgery:Part 2—Gastroesophageal reflux disease and its complications☆

Todd A. Worley, MDn, Rohan Joseph, MDHouston Methodist Texas Medical Center, Houston Methodist Surgical Associates Willowbrook, Houston, Texas

a r t i c l e i n f o

Article history:Received 18 June 2013Accepted 29 July 2013

Keywords:EndolumenalGERDBarrett'sTIFEsophyx2StrettaHALORFAEMRESD

72/$ - see front matter & 2013 Elsevier Inc. Ax.doi.org/10.1016/j.tgie.2013.07.003

author reports no direct financial interests tin connection with the submitted manuscripesponding author.ail address: TAWorley@HoustonMethodist.org

a b s t r a c t

Gastroesophageal reflux disease (GERD) is a complex entity whereby gastric contents reflux into theesophagus owing to either a lack of a natural barrier between the stomach and esophagus or adysfunction of foregut motility. In many cases, elements of both are present. GERD presents with eithertypical or atypical symptoms and can lead to metaplasia of the esophageal epithelium, a precancerouscondition called Barrett's esophagus (BE). GERD is a chronic disease, and like other chronic diseases, itrequires a thorough evaluation to deduce the exact etiology and also necessitates a dynamic and evolvingspectrum of therapy. Until relatively recently, the spectrum of treatment for chronic GERD and BEincluded lifestyle modification, pharmacotherapy, and laparoscopic fundoplication. Considering theeffect of chronic GERD on health and quality of life, and with further progress in understanding thedisease process, newer and less invasive techniques for treatment have emerged and developed over thepast decade. The 2 current FDA-approved therapies for the endolumenal treatment of GERD are transoralincisionless fundoplication with the Esophyx2 device and the Stretta system. If chronic GERD has lead toBE, endolumenal therapies include radiofrequency ablation with the HALO system or endoscopicresection of the metaplastic changes. Data regarding the outcomes of each of these procedures areaccumulating rapidly; however, further data and evaluation are necessary to determine the mostappropriate place for these procedures in the spectrum of therapy for chronic GERD and BE.

& 2013 Elsevier Inc. All rights reserved.

1. Introduction

Gastroesophageal reflux disease (GERD) is a complex entitywhereby gastric contents reflux into the esophagus owing toeither a lack of a natural barrier between the stomach andesophagus or a dysfunction of foregut motility. In many patients,both elements are present. Proper function is dependent upon theanatomy and physiology of the esophagus, stomach, and dia-phragm, working in concert to promote antegrade flow andprevent retrograde flow of the intraluminal contents. The anatomyand interactions of these organs form the basis for understand-ing GERD and lead us toward a better understanding of itstreatment. This article briefly reviews the pathophysiology ofGERD and describes the role endolumenal therapies play intreating this disease and one of its complications—Barrett'sesophagus (BE).

ll rights reserved.

hat might pose a conflict oft.

(T.A. Worley).

2. Manifestations of chronic GERD

2.1. Typical symptoms

The symptoms of GERD can be classified as typical and atypical.For a list of the typical symptoms, please refer to Table 1.

2.2. Atypical symptoms

Atypical symptoms, which are classically the extraintestinalmanifestations of GERD and include laryngopharyngeal refluxsymptoms, are not as straightforward in their contribution to anaccurate diagnosis. Often, the pathway to diagnosis is made easierwhen typical symptoms are present in addition, but it can be quitecircuitous when they are the primary manifestations. Table 2 liststhe most common atypical symptoms.

2.3. Barrett's esophagus

BE is another manifestation of GERD and is considered absoluteproof of an ineffective antireflux barrier [2]. It is the condition inwhich metaplastic columnar epithelium replaces the stratifiedsquamous epithelium that normally lines the distal esophagus.The diagnosis strictly mandates the presence of intestinal

Table 1Typical symptoms of GERD. (Adapted from Bhanot and Soper [1].)

Typical symptoms

HeartburnRegurgitationWater brashDysphagiaOdynophagia

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210 201

metaplasia, because it is the only columnar epithelium that has aclear predisposition for esophageal adenocarcinoma [3]. BE iscaused by chronic GERD and may progress from metaplasiathrough low-grade dysplasia (LGD), high-grade dysplasia (HGD),and ultimately to adenocarcinoma.

2.4. GERD as a chronic disease

GERD is more commonly thought of as a chronic disease. Thesigns and symptoms generally develop over an extended period oftime, rather than acutely. Surgical intervention is based upon thereconstruction of the body's natural antireflux barrier, which, inpatients suffering from GERD, is very often due to an anatomicalderangement [4]. Carrying this principle forward, as with otherchronic diseases, GERD should then be thought of as a disease thatexists on a spectrum, thereby requiring a spectrum of therapy. Justas coronary artery disease may necessitate medication, percuta-neous intervention, or surgery, GERD may require similarlyplanned interventions, whether this is pharmacotherapy, endolu-menal therapy, or surgery.

3. Etiologies of Chronic GERD

The exact etiologies that result in chronic GERD provideguidance as to the best antireflux treatment for each patient. Forthe purposes of this article, we focus on the etiologies as theyrelate to endolumenal treatment, rather than medication orsurgery. The principle causes of reflux are hiatal or paraesophagealhernia or a poorly functioning or weak lower esophageal sphincter(LES) [5]. It is these findings that constitute the basis and strategyfor antireflux surgery, and thus, endolumenal therapy. Otherdiagnoses contributing to chronic GERD are global esophagealand gastric dysmotility [2]. These 2 diagnoses alone do not oftenlead to endolumenal or surgical intervention, but they certainlyplay a role in determining the best course of endolumenal therapyor surgical treatment for patients with chronic GERD. In ourdiscussion, they have only been considered when found in con-junction with a hiatal hernia or an abnormal LES or both.

Table 2Atypical symptoms of GERD. (Adapted from Bhanot and Soper [1].)

Atypical symptoms

AsthmaAspiration pneumoniaChronic bronchitisHoarsenessLaryngitisChronic sore throatChronic coughFrequent swallowingMouth soresChest painDental erosion

3.1. Diagnosis of a hiatal hernia

The diagnosis and measurement of a hiatal hernia is perhapsthe single-most important aspect in selecting patients for endolu-menal treatment of chronic GERD. The anatomical derangementthat occurs disrupts the antireflux barrier and often leads tochronic GERD. The modalities for measuring a hiatal herniaprimarily consist of an upper gastrointestinal contrast swallowstudy (UGI), esophagogastroduodenoscopy (EGD), and to a lesserextent, manometry. We discuss the first 2 modalities in thefollowing sections.

A UGI study is a dynamic study that is best evaluated in realtime. A variety of maneuvers are used including abdominalpressure and Trendelenburg position to induce reflux. If possible,the amount of gastric cardia or fundus herniating through thediaphragm is measured longitudinally from the gastroesophagealjunction to the level of the diaphragm. A measurement should alsobe taken transversely to account for the diaphragmatic hiatus.Usually, the larger of the 2 is used as the measurement forplanning an endolumenal approach. As this is a dynamic study, asmall sliding-type hiatal hernia may not be visualized if it is in areduced state during the study [6].

EGD measurements are variable, and unless the hernia ismoderate to large, the size is often underestimated or missedentirely. This is especially true if the diaphragmatic fascia remainsintact, despite a hernia in the underlying muscle. Measurementsshould be taken as the endoscope first enters the distal third of theesophagus, before gastric insufflation, as this often reduces smallhiatal hernias. The size is measured from the Z line to the area ofdiaphragmatic narrowing, sometimes called the diaphragmaticpinch; this is a longitudinal measurement [6].

The next measurement occurs in the retroflexed view. Thestomach is completely insufflated, with no rugal folds remain-ing, so that it completely “layers out” on the inferior surface ofthe diaphragm. This gives the most accurate view of the hiatus.Moderate to large defects often do not hold the insufflation, butthis technique still allows for a fairly accurate measurement.Assuming that most upper endoscopes are almost 1 cm indiameter, this provides a relative scale in which the endo-scopist can calculate the transverse measurement. Again, thelarger of the 2 measurements should be taken, as this is thegreatest amount of potential herniation possible. This viewfacilitates appropriate evaluation of the Hill grade, which isused by many endoscopists to measure a hiatal hernia [7](Figure 1).

Fig. 1. Evaluation of a hiatal hernia using the Hill classification as given by Kahrilasand Pandolfino [7]. Grade I—the GEJ flap valve has a muscular ridge of tissue, whichis closely approximated to the retroflexed endoscope. Grade II—the GEJ flap valvehas a ridge of tissue that is less well defined than in Grade I, and it may open withrespiration but then quickly closes. Grade III—the ridge of tissue is effaced, there isoften failure of the GEJ to close around the endoscope, and the hiatus is patulous.Grade IV—there is no muscular ridge at all. The hiatus is continuously wide open,and a hiatus hernia is always present. (Color version of the figure is availableonline.)

Table 3Contraindications to TIF.

Contraindications

Severe reflux esophagitis, LA class DBMI 4 35 kg/m2

Esophageal ulcerEsophageal motility disordersEsophageal strictureHiatal hernia greater than 2 cm or Hill grade IV

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210202

After these measurements, the endoscopist has an accurateidea of the size and type of hiatal hernia (Type I, II, III, or IV). Thesemeasurements and classifications should be incorporated intoaccurate language for the endoscopic report, rather than lessspecific descriptors such as small, medium, or large, and used todetermine the best available endoscopic or surgical options.

3.2. Gastric dysmotility

The diagnosis of gastric dysmotility, particularly gastroparesis,which can cause or contribute to chronic GERD, is best made by agastric emptying study (GES). A GES is a nuclear medicine study inwhich a patient is made to eat a radiolabeled substance, such aseggs or oatmeal, and then the rate of gastric emptying is calcu-lated. This study reveals delayed emptying, which may lead togastric bloating and distension, thus potentially causing reflux. In amore crude approach, one may have somewhat of an indication ofthe gastric emptying based on the UGI study, but this is lessprecise. Preoperative GES may also have a role in predicting gasbloat syndrome after an antireflux procedure.

3.3. Esophageal dysmotility

Esophageal dysmotility is most precisely measured by esoph-ageal manometry. In small, type I hiatal hernias, there may be arole for high-resolution manometry to more precisely determinethe physiological elements of the GEJ, including hiatal herniameasurement [6]. At present, the major purpose of esophagealmanometry is to evaluate the efficiency and effectiveness ofesophageal peristalsis with swallowing and to measure the pres-sure and function of the LES. Equal in significance to a hiatalhernia, a poorly functioning LES is another finding thought tocontribute, or be a manifestation of, chronic GERD. Global esoph-ageal dysmotility may mimic the signs and symptoms of chronicGERD, or may contribute to the signs and symptoms when anotheretiology is present. A crude evaluation of esophageal motility canbe gained from an UGI study, but this is not as precise.

Esophageal manometry can also be helpful in predicting thelikelihood of de novo dysphagia in patients considering a 3601(Nissen) fundoplication procedure by determining if the peristalsisis adequate to propel food and fluid from the esophagus to thestomach even after surgical intervention. Esophageal manometryis not always used in patients undergoing a partial fundoplicationor radiofrequency treatment for LES (Stretta therapy), but it maybe useful if dysphagia or odynophagia is discovered duringpreoperative evaluation.

3.4. Esophageal pH testing

Esophageal pH testing is done for the following 2 reasonsbefore an endolumenal procedure to treat chronic GERD: (1) toprovide objective evidence of chronic GERD in the absence ofphysical findings and (2) for comparison of esophageal pH beforeand after a procedure to more objectively assess its effectiveness.

Fig. 2. The Esophyx2 device inserted over the endoscope. (Color version of thefigure is available online.)

4. FDA-approved endolumenal treatment for chronic GERD

Laparoscopic fundoplication has been the standard surgicaltreatment for chronic GERD. However, owing to concerns overcomplications such as dysphagia, gas bloat syndrome, and inher-ent surgical risks, interest remains high in developing less invasiveand even safer, yet equally efficacious procedures.

Currently, there are 2 FDA-approved endolumenal proceduresfor the treatment of chronic GERD on the market—transoralincisionless fundoplication (TIF) using the Esophyx2 device and

radiofrequency treatment of the area of the LES using the Strettasystem. Both procedures have been shown to be safe and effica-cious for the treatment of chronic GERD.

5. TIF (2.0) with the Esophyx2 device

The goals of the TIF procedure are to recreate the normalanatomy of the GEJ and provide a physiological barrier to reflux(compared with a supraphysiological response provided by aNissen fundoplication). This is done by accomplishing the follow-ing: (1) elongating the intra-abdominal esophageal segment,(2) reducing a hiatal hernia (less than 2 cm) if present, (3) creatinga minimum 2701 and 3 cm flap valve, (4) recreating the dynamicsof the angle of His, and (5) restoring the physiological pressure atthe GEJ. Contraindications to TIF are listed in Table 3.

TIF is in its second iteration, known as TIF 2.0, created to moreclosely resemble a laparoscopic fundoplication versus that foundin TIF 1.0. For the remainder of this article, all references to TIF,including outcomes, refer to TIF 2.0.

The evaluation of a patient undergoing TIF is the same for anypatient with chronic GERD who is considering an antirefluxprocedure. All patients undergo an EGD, with or without pHtesting, and most undergo a GES, UGI or manometry, or acombination of these.

5.1. Technique

General endotracheal anesthesia is administered to achieve fulldiaphragmatic paralysis. Preprocedure endoscopy is done undergeneral anesthesia to reassess the esophagus for pathology andmeasure the level of the Z line, diaphragmatic narrowing, and sizeof the hiatal hernia (if present) to ensure the suitability of TIF forthe patient.

The diagnostic gastroscope is then withdrawn and inserted intothe Esophyx2 device. The procedure requires an assistant toperform the endoscopic component of the procedure while thesurgeon operates the device. With the device over the endoscope,

Fig. 3. Visualizing the Esophyx2 device in the retroflexed view. (Color version ofthe figure is available online.)

Fig. 5. Key placement of the polypropylene fasteners during the TIF procedure.(Color version of the figure is available online.)

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210 203

it is inserted through the mouth and, under direct vision, passedinto the gastric lumen (Figure 2).

Once the device and endoscope (retroflexed) are positioned inthe completely insufflated gastric lumen, the procedure is begun(Figure 3).

The distal GEJ in the retroflexed view is best visualized as aclockface for the purposes of the TIF procedure. The 12-o'clockposition is the lesser curve of the stomach and 6-o'clock positioncorrelates with the angle of His toward the greater curve of thestomach. Thus the crural fibers of the anterior esophageal hiatusare at 1-2-'clock and the posterior portion is at 7-8-o'clockpositions (Figure 4).

To begin the fundoplication, a helical retractor is secured intothe gastric tissue just distal to the Z line at 11-o'clock position anda portion of the stomach and esophagus are pulled into the devicewhile wrapping a portion of the stomach clockwise to create theposterior portion of the fundoplication (Figure 5).

After this, 4 polypropylene (nonabsorbable), H-shaped fasten-ers are deployed to fixate the gastric wall to the esophageal wall.This creates the esophagogastric apposition of gastric serosa toesophageal adventitia. The device deploys 1 set of 2 fasteners andis then reloaded and repositioned for a second set of fasteners,creating a staggered pair of plications. Similarly, using the techni-que described above but at the 1-o'clock position, 4 more fastenersare placed after wrapping a portion of the stomach counterclock-wise to create the anterior portion of the fundoplication (Figure 5).

Lastly, the length of the angle of His, and hence the length ofthe antireflux flap valve, is augmented by placing up to 4 fastenersin both the 5- and 7-o'clock positions each (Figures 6 and 7).

Fig. 4. The distal GEJ viewed as a clockface in the retroflexed view. (Color version ofthe figure is available online.)

Once the TIF is complete, the device and scope are removedtogether. Postprocedure endoscopy is performed to assess andmeasure the restored antireflux valve (Figures 8 and 9).

5.2. Outcomes

Outcome data for TIF are accumulating rapidly, and the mostrecent presentation is of the unpublished TEMPO trial datapresented at the Digestive Disease Week in May 2013. TEMPO isa randomized controlled trial comparing TIF with maximum-doseproton pump inhibitor (PPI) therapy in 63 patients. At short-term,6-month follow-up, 90% of patients in the TIF group discontinuedPPI use and 62% experienced elimination of all daily chronic GERDsymptoms (vs 5% in the PPI group). TIF normalized esophageal acidexposure (off PPIs) similarly to maximum-dose PPI therapy, 54%and 52%, respectively. However, 90% of the cases of esophagitishealed in the TIF patients vs 38% in the patients on maximum-dosePPIs. There were no serious intraoperative complications in the TIFgroup, and there were no cases of de novo dysphagia, gas bloat, orincreased flatulence. Longer-term data are still needed for TIF, andthe TEMPO trial will continue up to 12 months [9].

The longest period of evaluation for TIF is 3 years. The results ofa study by Muls et al in 2012 demonstrated significantly improvedGERD health-related quality of life (GERD-HRQL) scores at 3 years.Discontinuation of daily PPI use was maintained in 74% of patientsat 3 years. Only 11 patients had esophageal pH data at 3 years, and9 (82%) remained normal. There were no adverse events. Recently,2-year data from the TIF US Registry, a multicenter, prospective,

Fig. 6. Building length in the antireflux flap valve. Used with permission from Belland Cadiere [8]. (Color version of the figure is available online.)

Fig. 9. Postprocedure TIF. Used with permission from Bell and Cadiere [8].Fig. 7. Building length in the antireflux flap valve. Used with permission from Belland Cadiere [8].

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210204

open-label, and postmarket registry of approximately 500patients, was presented at the Society of American Gastrointestinaland Endoscopic Surgeons 2013 Annual meeting. These datarevealed that TIF is safe and effective in improving chronic GERDsymptoms in up to 81% of patients, and esophagitis was signifi-cantly or fully healed in 81% of them [10]. In 2011, Testoni et al.reported that 69% of patients sustained their elimination orreduction of PPI therapy at 24 months [11].

In 2011, 4 other peer-reviewed articles revealed greater than 50%improvement in GERD-HRQL scores in 79%-86% of patients, withoverall health satisfaction as per GERD-HRQL scores ranging between64% and 83% satisfaction. Daily PPI therapy had been eliminated by76%-96% of patients [3,12-14]. In the study by Bell and Freeman [12],esophageal pH normalized in 61% of patients undergoing TIF.

Interestingly, in the study by Trad et al [13] they also reportedthat atypical laryngopharyngeal reflux symptoms were eliminatedin 63% of patients.

Despite the need for more level 1 evidence and results of long-term follow-up, the complication rates for TIF remain low in allstudies. There are no reported cases of long-term dysphagia, gasbloat syndrome, or case-related mortality, and serious adverseeffects, such as gastrointestinal perforation, esophageal leak,bleeding, pleural effusion, or mediastinal abscess, occur in lessthan 1% of cases [3,8,15,16]

6. The Stretta system for the radiofrequency treatment ofGERD

The goal of Stretta therapy is to remodel the musculature of theLES and gastric cardia by the application of radiofrequency (RF)

Fig. 8. Postprocedure TIF. Used with permission from Bell and Cadiere [8].

energy. As the LES musculature is remodeled, tissue complianceand transient LES relaxations are reduced, thereby restoring theantireflux function of the LES. The contraindications to Strettatherapy are listed in Table 4.

The evaluation of a patient undergoing Stretta is the same as forany patient with chronic GERD considering an antirefluxprocedure.

6.1. Technique

The procedure can be done under moderate sedation ormonitored anesthetic care (MAC) in a typical endoscopy center.Preprocedure endoscopy is done to reassess the esophagus forpathology and measure the level of the Z line, diaphragmaticnarrowing, and size of the hiatal hernia (if present) to ensure thesuitability of Stretta therapy for the patient.

Next, a guidewire is inserted through the endoscope into thegastric lumen and the scope is withdrawn. The Stretta RF gen-erator and catheter are prepared, and the Stretta catheter isinserted over the guidewire to the level of the Z line, as premarkedon the catheter (Figure 10).

RF energy treatments are then delivered at 6 levels starting1 cm proximal to the Z line to approximately 1.5 cm distal to the Zline to treat the LES and gastric cardia (Figure 11).

In total, needles directed into the LES and gastric cardiamusculature precisely deliver more than 50 directed RF treatmentsthroughout the entire procedure (Figure 12).

The mucosa is protected from damage by a continuous, con-trolled flow of chilled sterile water coursing from the catheterduring the procedure. The generator and catheter have presetmechanisms to automatically shut off if a specific delivery needle

Fig. 10. The Stretta catheter inserted over a guidewire.

Fig. 12. Needle-directed RF energy treatment with Stretta therapy. (Color version ofthe figure is available online.)

Table 4Contraindications to Stretta therapy.

Contraindications

Less than 18 years oldPregnancyHiatal hernia greater than 2 cmIncomplete LES relaxation when swallowingErosive or ulcerative esophagitis

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210 205

should not be in proper position or if it were to approach unsafetemperatures.

Postprocedure endoscopy is performed to inspect the restoredLES Figure 13.

6.2. Outcomes

The outcome data for Stretta are robust. In a 2012 systematicreview and meta-analysis, Perry and colleagues reviewed 1441patients who underwent Stretta therapy for chronic GERD andfound significant improvement in heartburn scores and GERD-HRQL scores. Interestingly, esophageal acid exposure andDeMeester scores decreased significantly from 44.4 to 28.5,whereas the LES pressure numerically improved, though notsignificantly, after Stretta therapy [17].

There have been several randomized controlled trials. In 2012,Arts et al performed a double-blind, randomized controlled studyof 22 patients comparing Stretta therapy with a sham procedure.The Stretta procedure significantly decreased the tissue compli-ance at the LES and improved symptom scores compared with thesham procedure at 3 and 6 months. However the esophageal acidexposure did not change in either group in this study [18] In 2009,Aziz and colleagues performed a prospective, randomized, andsham-controlled multicenter study to compare a single session ofStretta with a repeat Stretta procedure (if GERD-HRQL scores hadnot improved in 4 months) and a sham procedure. At 12 months,50% of the Stretta patients discontinued PPI therapy and theremaining experienced a significant decline in PPI use. None ofthe patients in the sham group were able to discontinue PPItherapy. Other secondary outcomes including LES basal pressure(11.6-16.2 mm Hg in the single Stretta and 12.2-19.6 mm Hg in thedouble Stretta group), grade of esophagitis (improved at least1 grade), and esophageal acid exposure significantly improvedfrom baseline. No serious complications occurred [19]. In 2008,Coron et al performed a prospective, randomized controlled,multicenter study on 43 patients to compare Stretta therapy vsPPI therapy. At 12 months, 56% of patients in the Stretta armcompletely discontinued or decreased their PPI usage, whereasonly 35% of patients in the PPI arm reported decreased PPI usage.There was no difference in either the REFLUX-QUAL or esophageal

Fig. 11. Levels of multiple RF energy treatments with Stretta therapy. (Color versionof the figure is available online.)

acid exposure. No patients in the PPI arm completely discontinuedPPI therapy [20].

The longest period of evaluation for the Stretta procedure is8 years. Dughera and colleagues presented data at the OESO, andmore than 60% of the 19 patients available for 8-year follow-upremained off PPIs. Heartburn and GERD-HRQL scores remainedconsistent with their previously published data [21]. In the samegroup's 2011 study, 56 patients were evaluated, and at 48 months,72.3% of patients were completely off of PPI medications and 92.8%of patients reported significant improvement in heartburn andGERD-HRQL scores. One major adverse event occurred, which wasprolonged gastroparesis, and it required a 3-week hospitalization[22]. Separately, Noar and Lotfi-Emran as well as Reymunde andSantiago presented 48-month follow-up data that reported sig-nificant improvement in GERD-related quality of life and heart-burn scores, and 75%-86% of patients were able to eliminate orreduce daily PPI usage [23,24].

Lastly, Gao et al [25] reported significant improvement in refluxsymptoms in 505 chronic GERD patients with primarily atypicallaryngopharyngeal symptoms. PPI therapy doses were reducedsignificantly in 71.7% of these patients.

Complication rates for Stretta therapy remain low in all studies.The major complication rate is estimated to be 0.24%, and there areno reported cases of long-term dysphagia [24].

7. Endolumenal Treatment for BE

Although there remains a great deal of controversy in thedefinition of BE, once the diagnosis is made, it is known that thesepatients have a premalignant condition with a higher risk ofesophageal adenocarcinoma. As defined previously, BE is foundin 10%-15% of patients with chronic GERD, and it has an annual

Fig. 13. Postprocedure Stretta therapy. (Color version of the figure is availableonline.)

Fig. 14. Microscopic description of esophageal and Barrett's epithelium and the treatment depth of the HALO system. (Color version of the figure is available online.)

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210206

transformation rate to esophageal adenocarcinoma of 0.25%-0.5%per patient [26]. BE is histologically classified as nondysplastic,LGD, or HGD, and currently patients with non-nodular, dysplasticBE are appropriate candidates for radiofrequency ablation (RFA) astreatment for BE [27]. Patients with nodular BE, BE with high-grade dysplasia, and even intramucosal carcinoma are beingincreasingly treated and staged with endoscopic mucosal resection(EMR) or endoscopic submucosal dissection (ESD).28

8. RFA for BE

The goal of RFA with the HALO system is to completelyeradicate Barrett's epithelium without causing significant injuryto the underlying tissue. Ablative energy is delivered in less than1 second, and consistent application of bipolar energy uniformlyremoves the esophageal epithelium. The treatment depth iscontrolled and is between 500 and 1000 mm, which will includethe muscularis mucosae but not ablate through it. Barrett'sepithelium is approximately 500 mm thick (Figure 14). Table 5 liststhe contraindications for this technique.

8.1. Technique

RFA is accomplished by the HALO 360þ ablation catheter forcircumferential BE, typically for segments greater than or equal to3 cm. HALO 90þ and HALO 60þ ablation catheters are used forfocal BE ablation, typically for treating smaller primary areas suchas islands and tongues or secondary areas found in follow-upevaluation after a primary treatment session. The case is accom-plished under moderate sedation.

The first step is to assess and measure the area of BE in relationto the Z line. A HALO 360þ guidewire is inserted through theendoscope, and the endoscope is exchanged for a HALO 360þ

Table 5Contraindications to RFA with the HALO system.

Contraindications

PregnancyPrevious radiation to the esophagusEsophageal varicesHeller myotomyEosinophilic esophagitis

sizing balloon. Once the inner diameter of the esophagus ismeasured, the appropriate ablation catheter is selected.

Next, the HALO 360þ ablation catheter is inserted over theguidewire, followed by the endoscope. Under direct visualization,the balloon is positioned just proximal to the superior edge of theBE. The balloon is automatically inflated and energy is applied. Thecatheter is moved distally by 3 cm to the edge of the previouslyablated segment. The procedure is repeated until the gastric foldsare reached (Figure 15).

The coagulum is removed from the ablation zone with theHALO EMR device, and the catheter is removed and cleaned withwater. The catheter is then reinserted and the process repeatedunder direct endoscopic visualization for a total of 2 treatmentapplications to the area of BE.

The process for HALO 90þ and HALO 60þ treatment is done bythe same sequence of steps as for HALO 360þ . The difference isthat each focal area of BE receives a total of 4 treatment applica-tions, as opposed to 2 with the HALO 360þ (Figure 16).

RFA is performed at approximately 2-month intervals until allvisible BE has been eradicated, but no more than 4 treatmentsessions should be performed. Multiple treatment sessions areusually necessary. The primary treatment session is typically themost extensive, and follow-up secondary treatment sessions aredone for the remaining focal areas of BE [26].

8.2. Outcomes

There are several randomized controlled trials to support theuse of endoscopic RFA to treat dysplastic BE. Its role in treatingnondysplastic BE is less clear.

In 2011, the American Gastroenterological Association per-formed a technical review on the management of BE, and thus

Fig. 15. HALO 360þ ablation catheter. (Color version of the figure is availableonline.)

Fig. 17. Initial marking of the planned resection margins during EMR.

Fig. 18. Lift-and-cut technique for EMR.

Fig. 16. HALO 90þ ablation catheter. (Color version of the figure is availableonline.)

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210 207

put forth a set of evidence-based guidelines for its treatment.These guidelines state that ablation is a reasonable option forthe treatment of BE with HGD if there are no visible nodules,especially if the patient is a poor surgical candidate. The issuessurrounding the durability of RFA, the importance of buriedmetaplasia, and the long-term efficacy in cancer preventionremain unsettled. As for nondysplastic BE and BE with LGD,they can be successfully ablated, but the benefits of eradication,in light of the difficulties in diagnosis of LGD and the risks andcost of RFA procedures, remain in question. Currently, theconclusion is that RFA should be an option in BE with confirmedLGD after a thorough discussion of the risks and benefits withthe patient [28].

In a randomized, sham-controlled, U.S multicenter trial of 127patients with dysplastic BE, 81% of patients with HGD achievedcomplete eradication vs only 19% in the control group. In patientswith LGD, 90.5% of patients achieved complete eradication vs22.7% in the control group. Overall, disease progression was seenin 3.6% of the ablation group vs 16.3% in the control group. Only3 serious adverse events occurred, which included bleeding in1 patient and chest pain in 2 others. An esophageal stricturedeveloped in 6% of patients, and they were all treated withsuccessful endoscopic dilatation [27].

RFA outcomes have also been reported from the U.S. Barrett'sRFA Registry data. It is a web-based, prospective, multicenterregistry with more than 5000 patients, and is the largest pro-spective trial in Barrett's endotherapy ever conducted. Overall,complete eradication of BE occurred in 73% of patients andcomplete eradication of dysplasia occurred in 87%. When strati-fied, complete eradication of LGD was 90%, HGD was 82%, andintramucosal carcinoma was 85% [29]. When accounting formultiple procedures in the same patient, the overall complicationrate was 2.9% per patient and 1.1% per procedure. The stricturerate was 2.1% per patient and 0.8% per procedure, bleedingoccurred at a rate of 0.4% per patient and 0.1% per procedure,and perforation occurred in 0.036% of patients and in 0.014% ofprocedures [30].

The AIM-II trial evaluated patients with nondysplastic BE. Thisis a prospective, multicenter U.S. trial revealed complete erad-ication of nondysplastic BE in 98.4% of patients after 2.5 years.Fifty patients were evaluated at 5 years and 92% of themmaintained complete eradication, whereas the remaining 8%had developed focal nondysplastic BE. Focal RFA was performedin all of these patients, and complete eradication was achieved[31]. Questions still remain regarding the benefits of eradicationin nondysplastic BE.

RF ablation of BE has transformed the management of thisdisease and stands as an excellent example of an endoscopictherapy replacing a highly morbid surgery— esophagectomy.In specialized centers that have expertise in treating patientswith BE, most patients are no longer subjected to esophagealresection.

9. Endoscopic Resection of BE

EMR and ESD are 2 distinct endolumenal therapeutic modal-ities that are gaining increasing popularity in the management ofBE. These procedures are currently being used in Japan for thedefinitive management of early gastric cancer and are increasinglyfinding a role in the treatment of BE in patients presenting withhigh-grade dysplasia or intramucosal carcinoma. The rationale forendoscopic resection is that the risk of lymph node metastasis isnegligible in patients with isolated mucosal disease, thus offeringthe chance for a curative endolumenal strategy [32,33]. Anotheradvantage of EMR and ESD is that it offers a substantially greateramount of tissue for biopsy, thereby allowing more accuratestaging of disease [33]. Based on the staging results, othertherapeutic modalities such as esophagectomy may be indicated.Finally, both EMR and ESD may be combined with RFA as part ofthe definitive treatment strategy to cure select patients who havenodular disease in association with BE.

Indications:

(1)

BE with visible mucosal changes consistent with high-gradedysplasia or intramucosal carcinoma (carcinoma in situ) withwell to moderately differentiated cytology.

(2)

EMR is recommended for lesions up to 3 cm in size. (3) ESD is recommended for lesions greater than 3 cm.

9.1. Technique

9.1.1. EMREMR is performed in 1 of 2 ways: nonsuction (“lift-and-cut”)

and suction (“suck-and-cut”) techniques. Both techniques begin bymarking the planned resection margins with brief bursts ofcautery using an endoscopic snare with the wire deployed mini-mally (Figure 17).

9.1.1.1. Lift-and-cut techniques. The original strip-off biopsytechnique described by Tada et al [34] in 1984 used a lift-and-cut technique. All these techniques begin with the creation of asubmucosal bleb to elevate the lesion from the submucosa anddecrease the risk of perforation. This injection was originally donewith saline, but other solutions have been used to gain betterresults. Epinephrine can be added to decrease bleeding.

Fig. 19. Suck-and-cut technique using an endoscope cap for EMR.

Fig. 21. (A-E) ESD technique. (Color version of the figure is available online.)

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210208

Hypertonic saline (3.75% NaCl), 20% dextrose, or sodiumhyaluronate are used to improve the development andmaintenance of the bleb [35], and Indigo carmine (0.004%) isoften added to the injectate to stain the submucosa and provide abetter evaluation of the depth of resection. The strip-off biopsytechnique is performed with a double-channel gastroscope. Aftercreating the submucosal elevation, the lesion is grasped with a rat-tooth forceps that has been passed through an open polypectomysnare. The forceps lifts the lesion and the snare is pushed downaround its base and resection ensues. This procedure has beensimplified so that it can be accomplished with a standard single-channel gastroscope by simply elevating with submucosalinjection and then snaring freehand with the polypectomy snare(Figure 18).

9.1.1.2. Suck-and-cut technique. In this method, the lesion isaspirated into a cap attached to the tip of the endoscope toelevate it off the muscularis propria, with or without submucosalinjection, and then resected. The most common methods useeither a cap (EMR-C) or a band ligating device (EMR-L) fitted onthe end of the endoscope. For EMR-C, a specialized cap is fitted tothe end of the endoscope providing a chamber for aspiration ofmucosa. Typically, the lesion is elevated with a submucosalinjection and a crescent-shaped snare is mounted into the distalinner rim of the cap. The lesion is then aspirated into the cap andthe snare tightened around its base. The snared “pseudopolyp” isthen pushed outside the cap and excised with monopolar energy.The special EMR caps come in various sizes and shapes, the largestof which is 18 mm in diameter (Figure 19).

When performing EMR-L, submucosal injection may not berequired. This technique uses a variceal band ligator to ligate themucosal lesion and create a pseudopolyp. The banded tissue isthen simply snare excised either above or below the band(Figure 20).

For both EMR-C and EMR-L, the maximum diameter amenableto one-piece excision is approximately 20 mm. If multiple resec-tions are required, they should be accomplished at the initial

Fig. 20. Suck-and-cut technique using a band ligating device for EMR.

setting if possible to optimize the submucosal lift, which will notbe readily attainable once scar tissue has formed.

9.1.2. ESDESD allows for larger en bloc resections and proceeds in a

fashion completely different from EMR. The procedure begins bymarking the periphery of the lesion with small cautery burns usingthe tip of a snare. A margin of at least 5 mm is planned and marksare placed approximately every 2 mm Figure 21A. A submucosalinjection is then accomplished around the periphery of the lesionusing a solution such as sodium hyaluronate so that the elevationpersists throughout the procedure Figure 21B. The submucosalsolution is often stained with 0.004% indigo carmine. A circum-ferential incision is then made to isolate the lesion using an ESD

Fig. 22. (A-E) ESD knives: Needle knife: This knife has a fine tip and small contactarea that allows sharp incision. Because of its sharp nature, it can easily causeperforation if not controlled carefully. Mucosal incisions are usually begun with aneedle knife, but then a switch is made to a protected tip knife to minimize therisk of perforation. IT knife: The insulated tip knife is a needle knife with the tipcovered by a ceramic ball. This blunt, nonthermal tip reduces the risk ofperforation. A second-generation IT knife has a conducting surface on the bottomof the ball tip to allow for better tissue division when drawing back on the knife.Hook knife: This is a needle knife with the distal 1 mm of the tip bent at a rightangle. The knife also rotates for optimal positioning. Flex knife: This knife has arounded tip made of a twisted wire like a snare. Its length can be adjusted asneeded. The shaft of the catheter is flexible with a thickened tip that acts as atissue stop to minimize the chance of perforation. Triangle tip knife: Has atriangular conductive tip that facilitates cutting mucosa. This knife was designedfor use in all parts of the ESD procedure. (Color version of the figure is availableonline.)

Fig. 23. (A and B) Types of endoscopic electrosurgical hemostatic forceps.

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210 209

electrosurgical knife (Figure 21C-E). Figure 22 details the differentESD devices.

Once the circumferential incision is complete, additional sol-ution is injected into the submucosa in the center to obtain a morecomplete lift. One of the ESD knifes is then used to excise thelesion in the submucosal plane. Meticulous hemostasis is criticalto facilitate visualization, and the lesion is often positionedopposite the ground to use gravity to clear the field. An electro-surgical hemostatic forceps may be useful to maintain hemostasis(Figure 23A and B).

A transparent hood is often mounted on the end of theendoscope to facilitate visualization and dissection into thesubmucosal plane.

After EMR or ESD, patients are placed on short-term, double-dose PPIs and Carafate to aid in the healing of the resected site. Apostresection endoscopic surveillance program is important toensure complete resection.

9.2. Outcomes

Both EMR and ESD have been demonstrated to be safe andfeasible as definitive treatment in subgroups of patients with BE.This may avoid the need for a surgical operation in these selectpatients and may offer a similar outcome. Recurrence of diseaseafter endoscopic resection continues to be an area of study withresults varying from 0 to 22% [36-40]. To further improve theseoutcomes, endoscopic resection has been combined with ablativestrategies to treat any potential residual disease.

Bleeding, perforation, and stricture are the most commonserious complications of EMR and ESD. Bleeding is the mostfrequent complication and can occur at the time of the procedureor later (0-30 days afterward). Intraprocedural bleeding canusually be controlled using a combination of diathermy andendoscopic clips. Delayed bleeding, manifesting as hematemesisor melena, usually occurs within 12 hours of the procedure and isan indication for urgent repeat endoscopy. The incidence of majorbleeding after endoscopic resection is less than 1%, but this ratedepends on the definition of major bleeding to some extent[36,40].

Perforation is the most feared complication of EMR and ESD,and the rate of occurrence is between 1 and 5% [41,42]. Perfo-rations recognized at the time of occurrence are often managedwith endoscopic clip closure. After clip closure, the patientsreceive nasogastric tube decompression, empiric broad-spectrumantibiotics, high-dose acid-suppression therapy, and frequentreevaluation. Signs of worsening overall condition or generalizedperitonitis are indications for operative exploration and repair.

A stricture in the distal esophagus at the site of the resection isa relatively underreported complication, but it can occur in nearly40% of patients [40]. It correlates with the extent of resection andis more common in patients who have undergone circumferentialresections. Fortunately, these strictures are amenable to endo-scopic dilation [43].

Both EMR and ESD are time-consuming, advanced endoscopictechniques requiring a high degree of skill and expertise. As theseprocedures become more common, EMR and ESD are additionaltools in the armamentarium of the surgical endoscopist.

10. Conclusion

Chronic GERD is complex and dynamic disease that exists on aspectrum. As such, appropriate treatment for this disease neces-sitates a spectrum of therapy. Chronic GERD significantly affects apatient's quality of life. However, once metaplastic changes occurin the esophagus, that is, BE, the ramifications of chronic GERDbecome far more serious in that there is now a premalignantcondition with an increased risk of esophageal adenocarcinoma.

Considering the effect of chronic GERD, and with furtherprogress in understanding the disease process, newer less invasivetechniques for treatment have emerged and developed over thepast decade. Owing to concern over surgical morbidity andmortality, along with the demand for less invasive treatmentmodalities that are safe and effective, the number of endolumenalprocedures performed each year to treat chronic GERD and BEcontinues to grow. Another common thread that enhances thevalue of these endolumenal procedures is that TIF, Stretta Therapy,RFA with HALO, EMR, and ESD are often repeatable procedures inthe event of recurrent GERD or BE. Should disease progress beyondthe realm of endolumenal therapy, surgical intervention remains aviable option for rescue treatment.

Further data and evaluation will help to determine the place ofeach procedure in the spectrum of therapy for chronic GERD.

References

[1] Bhanot P, Soper NJ. Gastroesophageal Reflux Disease. Current Surgical Ther-apy. 9th ed. Philadelphia, PA: Mosby Elsevier; 2008.

[2] Fanelli R, Richardson W, et al. Guidelines for Surgical Treatment of Gastro-esophageal Reflux Disease (GERD). Los Angeles, CA: SAGES; 2010 [Guidelinesfor clinical practice].

[3] Barnes WE, Hoddinott KM, Mundy S, Williams M. Transoral incisionlessfundoplication offers high patient satisfaction and relief of therapy-resistanttypical and atypical symptoms of GERD in community practice. Surg Innov2011;18:119–29.

[4] Jobe BA, O'Rourke RW, McMahon BP, Gravesen F, Lorenzo C, Hunter JG, et al.Transoral endoscopic fundoplication in the treatment of gastroesophagealreflux disease: the anatomic and physiologic basis for reconstruction of theesophagogastric junction using a novel device. Ann Surg 2008;248:69–76.

[5] Fein M, Ritter MP, DeMeester TR, Oberg S, Peters JH, Hagen JA, et al. Role of thelower esophageal sphincter and hiatal hernia in the pathogenesis of gastro-esophageal reflux disease. J Gastrointest Surg 1999;3:405–10.

[6] Kahrilas PJ, Kim HC, Pandolfino JE. Approaches to the diagnosis and grading ofhiatal hernia. Best Pract Res Clin Gastroenterol 2008;22:601–16.

[7] Hill LD, Kozarek RA, Kraemer SJ, Aye RW, Mercer CD, Low DE, et al. Thegastroesophageal flap valve: in vitro and in vivo observations. GastrointestEndosc 1996;44:541–7.

[8] Bell RC, Cadiere GB. Transoral rotational esophagogastric fundoplication:technical, anatomical, and safety considerations. Surg Endosc 2011;25:2387–99.

[9] Trad KS, Barnes, WE, Simoni, G, Shughoury, AB, Raza, M, Heise, J, et al.Transoral fundoplication vs. proton pump inhibitor therapy for treatment ofchronic gastroesophageal reflux: short-term results from a multicenter,randomized, open label clinical trial. Digestive Disease Week; May 2013;Orlando, FL: Gastroenterology; 2013. p. S163–S164.

[10] Bell RC, et al., ed. Long-term outcomes of endoscopic fundoplication: 2 yearresults from the prospective multicenter U.S. study (S041). SAGES scientificsession 6 therapeutic endoscopy; 2013; Baltimore, MD.

[11] Muls VEA, Marchese M, Bastens B, Buset M, Deviere J, Louis H, et al. Three-yearresults of a multicenter prospective study of transoral incisionless fundopli-cation. Surg Innov 2012 [PMCID: 22968006].

T.A. Worley, R. Joseph / Techniques in Gastrointestinal Endoscopy 15 (2013) 200–210210

[12] Bell RC, Freeman KD. Clinical and pH-metric outcomes of transoral esoph-agogastric fundoplication for the treatment of gastroesophageal reflux disease.Surg Endosc 2011;25:1975–84.

[13] Trad KS, Turgeon DG, Deljkich E. Long-term outcomes after transoral incision-less fundoplication in patients with GERD and LPR symptoms. Surg Endosc2012;26:650–60.

[14] Ihde GM, Besancon K, Deljkich E. Short-term safety and symptomatic out-comes of transoral incisionless fundoplication with or without hiatal herniarepair in patients with chronic gastroesophageal reflux disease. Am J Surg2011;202:740–6 [discussion 6-7].

[15] Demyttenaere SVBS, Pham T, Anderson J, Dettorre R, Melvin WS, Mikami DJ.Transoral incisionless fundoplication for gastroesophageal reflux disease in anunselected patient population. Surg Endosc 2010;24:854–8.

[16] Testoni PA, Vailati C, Testoni S, Corsetti M. Transoral incisionless fundoplica-tion (TIF 2.0) with EsophyX for gastroesophageal reflux disease: long-termresults and findings affecting outcome. Surg Endosc 2012;26:1425–35.

[17] Perry KA, Banerjee A, Melvin WS. Radiofrequency energy delivery to the loweresophageal sphincter reduces esophageal acid exposure and improves GERDsymptoms: a systematic review and meta-analysis. Surg Laparosc EndoscPercutan Tech 2012;22:283–8.

[18] Arts J, Bisschops R, Blondeau K, Farre R, Vos R, Holvoet L, et al. A double-blindsham-controlled study of the effect of radiofrequency energy on symptomsand distensibility of the gastro-esophageal junction in GERD. Am J Gastro-enterol 2012;107:222–30.

[19] Abdel Aziz A, El-Khayat HR, et al. A prospective randomized trial of sham,single-dose Stretta, and double-dose Stretta for the treatment of gastro-esophageal reflux disease. Curr Opin Gastroenterol 2009;25:352–7.

[20] Coron E, Sebille V, Cadiot G, Zerbib F, Ducrotte P, Ducrot F, et al. Clinical trial:radiofrequency energy delivery in proton pump inhibitor-dependent gastro-oesophageal reflux disease patients. Aliment Pharmacol Ther 2008;28:1147–58.

[21] Dughera L, Cisaro, F, Cassolino, P, DeCento, P. Long-term results of radio-frequency energy delivery (Stretta) for the treatment of gastroesophagealreflux disease. Poster presented at OESO; 2012. Congress; Torino, Italy.

[22] Dughera L, Navino M, Cassolino P, De Cento M, Cacciotella L, Cisaro F, et al.Long-term results of radiofrequency energy delivery for the treatmentof GERD: results of a prospective 48-month study. Diagn Ther Endosc2011:507157.

[23] Noar MD, Lotfi-Emran S. Sustained improvement in symptoms of GERD andantisecretory drug use: 4-year follow-up of the Stretta procedure. GastrointestEndosc 2007;65:367–72.

[24] Reymunde A, Santiago N. Long-term results of radiofrequency energy deliveryfor the treatment of GERD: sustained improvements in symptoms, quality oflife, and drug use at 4-year follow-up. Gastrointest Endosc 2007;65:361–6.

[25] Gao X, Wang ZG, Wu JM, Ji F, Zhang CC, Ning YC, et al. Radiofrequencytreatment on respiratory symptoms due to gastroesophageal reflux disease.Chin Med J 2011;124:1006–9.

[26] Wolf DS, Dunkin BJ, Ertan A. Endoscopic radiofrequency ablation of Barrett'sesophagus. Surg Technol Int 2012 [XXII].

[27] Shaheen NJ, Sharma P, Overholt BF, Wolfsen HC, Sampliner RE, Wang KK, et al.Radiofrequency ablation in Barrett's esophagus with dysplasia. New Engl JMed 2009;360:2277–88.

[28] Spechler SJ, Sharma P, Souza RF, Inadomi JM, Shaheen NJ. American Gastro-enterological Association technical review on the management of Barrett'sesophagus. Gastroenterology 2011;140:e18–52 [quiz e13].

[29] Shaheen NJ, Bulsiewicz WJ, Rothstein RI, et al. Eradication rates of Barrett'sesophagus using radiofrequency ablation (RFA): results from the US RFAregistry. Gastroenterology 2012;142:A1600.

[30] Shaheen NJ, Bulsiewicz WJ, Rothstein RI, et al. Radiofrequency ablation (RFA)safely treats Barrett's esophagus in a nationwide, multicenter cohort: resultsfrom the US RFA registry. Gastroenterology 2012;142:A1121.

[31] Fleischer DE, Overholt BF, Sharma VK, Reymunde A, Kimmey MB, Chuttani R,et al. Endoscopic radiofrequency ablation for Barrett's esophagus: 5-yearoutcomes from a prospective multicenter trial. Endoscopy 2010;42:781–9.

[32] Conio M, Cameron AJ, Chak A, Blanchi S, Filiberti R. Endoscopic treatment ofhigh-grade dysplasia and early cancer in Barrett's oesophagus. Lancet Oncol2005;6:311–21.

[33] Peters FP, Kara MA, Rosmolen WD, Aalders MC, Ten Kate FJ, Bultje BC, et al.Endoscopic treatment of high-grade dysplasia and early stage cancer inBarrett's esophagus. Gastrointest Endosc 2005;61:506–14.

[34] Tada M, Shimada M, Murakami F, et al. Development of strip-off biopsy.Gastroenterol Endosc 1984;26:833–9 [in Japanese with English abstract].

[35] Fujishiro M, Yahagi N, Kashimura K, Mizushima Y, Oka M, Enomoto S, et al.Comparison of various submucosal injection solutions for maintaining mucosalelevation during endoscopic mucosal resection. Endoscopy 2004;36:579–83.

[36] Ell C, May A, Gossner L, Pech O, Gunter E, Mayer G, et al. Endoscopic mucosalresection of early cancer and high-grade dysplasia in Barrett's esophagus.Gastroenterology 2000;118:670–7.

[37] Esaki M, Matsumoto T, Hirakawa K, Nakamura S, Umeno J, Koga H, et al. Riskfactors for local recurrence of superficial esophageal cancer after treatment byendoscopic mucosal resection. Endoscopy 2007;39:41–5.

[38] May A, Gossner L, Pech O, Muller H, Vieth M, Stolte M, et al. Intraepithelialhigh-grade neoplasia and early adenocarcinoma in short-segment Barrett'sesophagus (SSBE): curative treatment using local endoscopic treatmenttechniques. Endoscopy 2002;34:604–10.

[39] Ell C, May A, Pech O, Gossner L, Guenter E, Behrens A, et al. Curativeendoscopic resection of early esophageal adenocarcinomas (Barrett's cancer).Gastrointest Endosc 2007;65:3–10.

[40] Chennat J, Konda VJ, Ross AS, de Tejada AH, Noffsinger A, Hart J, et al.Complete Barrett's eradication endoscopic mucosal resection: an effectivetreatment modality for high-grade dysplasia and intramucosal carcinoma—anAmerican single-center experience. Am J Gastroenterol 2009;104:2684–92.

[41] Gerke H, Siddiqui J, Nasr I, Van Handel DM, Jensen CS. Efficacy and safety ofEMR to completely remove Barrett's esophagus: experience in 41 patients.Gastrointest Endoscopy 2011;74:761–71.

[42] Soetikno RM, Gotoda T, Nakanishi Y, Soehendra N. Endoscopic mucosalresection. Gastrointest Endosc 2003;57:567–79.

[43] Katada C, Muto M, Manabe T, Boku N, Ohtsu A, Yoshida S. Esophageal stenosisafter endoscopic mucosal resection of superficial esophageal lesions. Gastro-intest Endosc 2003;57:165–9.