Efficacy of Radiofrequency Ablation Combined With Endoscopic Resection for Barrett s Esophagus With Early Neoplasia

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2010;8:23 29 ENDOSCOPY CORNER Efficacy of Radiofrequency Ablation Combined With Endoscopic Resection for Barrett s Esophagus With Early Neoplasia ROOS E. POUW,* KATJA WIRTHS, PIERRE EISENDRATH, CARINE M. SONDERMEIJER,* FIEBO J. TEN KATE, PAUL FOCKENS,* JACQUES DEVIÈRE, HORST NEUHAUS, and JACQUES J. BERGMAN* *Department of Gastroenterology and Hepatology, Department of Pathology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands; Department of Internal Medicine, Evangelisches Krankenhaus, Düsseldorf, Germany; and the Department of Gastroenterology and Hepatopancreatology, Erasme University Hospital, Brussels, Belgium See Editorial on page 7. BACKGROUND & AIMS: Radiofrequency ablation (RFA) is safe and effective for eradicating intestinal metaplasia and neoplasia in patients with Barrett s esophagus. We sought to assess the safety and efficacy of RFA in conjunction with baseline endoscopic resection for high-grade intraepithelial neoplasia (HGIN) and early cancer. METHODS: This multicenter, prospective cohort study included 24 patients (mean age, 65 years; median Barrett s esophagus, 8 cm), with Barrett s esophagus of 12 cm containing HGIN or early cancer, from 3 European tertiary-care medical centers. Visible lesions were endoscopically resected, followed by serial RFA. Focal escape endoscopic resection was used if Barrett tissue persisted despite RFA. Complete response, defined as all biopsies negative for intestinal metaplasia and neoplasia, was assessed during endoscopy with 4-quadrant biopsies taken every 1 cm of the original Barrett s segment 2 months after the patient was last treated. RESULTS: Twenty-three patients underwent pre-rfa endoscopic resection for visible lesions; 16 patients had early cancer and 7 patients had HGIN. The worst residual histology results, pre-rfa (after any endoscopic resection) were: HGIN (10 patients), low-grade intraepithelial neoplasia (11 patients), and intestinal metaplasia (3 patients). Neoplasia and intestinal metaplasia were eradicated in 95% and 88% of patients, respectively; after escape endoscopic resection in 2 patients, rates improved to 100% and 96%, respectively. Complications after RFA included melena (n 1) and dysphagia (n 1). After additional follow-up (median, 22 months; interquartile range, 17.2 23.8 months) no neoplasia recurred. CONCLU- SIONS: This European multicenter study to show that early neoplasia in Barrett s esophagus can be effectively and safely treated with RFA, in combination with prior endoscopic resection of visible lesions. Barrett s esophagus (BE), defined as a columnar-lined esophagus and biopsy showing specialized intestinal metaplasia (IM), is the most important risk factor for the development of esophageal adenocarcinoma. 1 Patients with known BE undergo endoscopic surveillance to detect neoplasia (ie, highgrade intraepithelial neoplasia [HGIN] or early cancer [EC]) at a curable stage. 2 Patients with HGIN or with EC confined to the mucosa (T1m) may be treated endoscopically owing to low risk for lymph node metastasis at the time of diagnosis, but more advanced cancers are surgical indications. 3 8 The cornerstone of endoscopic treatment of early BE neoplasia is endoscopic resection (ER), which allows for removal of visible lesions and assessment of tumor infiltration depth and differentiation. After focal ER, however, the residual Barrett s mucosa remains at risk for malignant transformation, and cancer recurrences are found in 30% of patients during follow-up evaluation. 3,4 To prevent such metachronous lesions, endoscopic approaches have been studied in an attempt to eradicate the residual Barrett s mucosa (eg, radical ER, 9,10 photodynamic therapy, 11 13 and argon plasma coagulation). 14 16 These treatment options, however, do not always result in complete clearance of IM and associated intraepithelial neoplasia and are limited by other shortcomings such as esophageal stenosis, photosensitivity, and subsquamous foci of IM (also known as buried Barrett s). 9 16 A newer endoscopic ablation technique is radiofrequency ablation (RFA), 17 19 which has promising safety and efficacy results published for nondysplastic BE, 20,21 low-grade intraepithelial neoplasia (LGIN), 22,23 and HGIN. 23 25 Single-center studies effectively have combined focal ER of visible lesions with RFA for residual Barrett s mucosa in patients with HGIN and EC. 24,25 The aim of the present study was to evaluate further the safety and efficacy of this combined modality approach in BE patients with HGIN or EC in a European multicenter setting. Materials and Methods Study Design This was a prospective cohort trial conducted at 3 tertiary-care medical centers in Europe. The ethical committee of each institution reviewed and approved the protocol and the Abbreviations used in this paper: BE, Barrett s esophagus; CR-IM, complete response for intestinal metaplasia; CR-neoplasia, complete response for neoplasia; EC, early cancer; ER, endoscopic resection; HGIN, high-grade intraepithelial neoplasia; IM, intestinal metaplasia; IQR, interquartile range; LGIN, low-grade intraepithelial neoplasia; NBI, narrow-band imaging; RFA, radiofrequency ablation. 2010 by the AGA Institute 1542-3565/10/$36.00 doi:10.1016/j.cgh.2009.07.003

24 POUW ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 1 patient informed consent form. The trial was registered at www.trialregister.nl (NTR1434). A central study coordinator monitored all procedures and entered data into a central database. Patients were eligible if they met all inclusion criteria as follows: age 18 to 85 years, BE length of 12 cm or less, HGIN or EC in BE segment on 2 endoscopies in the prior 6 months, visible lesions removed with ER before RFA, and no signs of metastasis on endoscopic ultrasonography or computed tomography scan. Patients were excluded if they met any exclusion criterion: pre-rfa ER with cancer at the vertical resection margin, greater than T1sm1 invasion, poor differentiation or worse, or angiolymphatic invasion; esophageal stenosis preventing passage of an 11.3-mm endoscope; persistent visible lesions after ER and pre-rfa; invasive cancer on biopsies after ER and pre-rfa. To ascertain eligibility, patients underwent 2 high-resolution endoscopies with documentation of BE landmarks according to the Prague classification system. 26 In the case of visible lesions, these were removed with ER followed by 2 additional endoscopies with 4-quadrant 1-cm biopsies to exclude residual cancer and residual nonflat lesions. Endoscopic ultrasonography was performed in all patients to rule out exclusionary lesions. In the case of cancer on biopsy or ER, a computed tomography scan of the thorax and upper one third of the abdomen was performed to rule out metastatic disease. Endoscopic Interventions All procedures were performed on an outpatient basis using intravenous midazolam, fentanyl, pethidine, propofol, or a combination thereof. All patients were prescribed high-dose proton pump inhibitor therapy (esomeprazole 40 mg twice a day) during the entire study period, supplemented with sucralfate suspension 5 ml (200 mg/ml) 4 times a day and ranitidine 300 mg before bedtime for 2 weeks after any therapeutic endoscopy. ER of any nonflat lesions was performed using the ER-cap technique, the multiband mucosectomy technique, or endoscopic submucosal dissection, at the discretion of the physician. The HALO 360 and HALO 90 ablation systems used in this trial have 510(k) clearance by the Food and Drug Administration in the United States and the CE Mark for Europe for the treatment of BE (BÂRRX Medical Inc, Sunnyvale, CA). For circumferential RFA, a sizing catheter was introduced over a guidewire to measure the inner-esophageal diameter of the esophagus. A HALO 360 ablation catheter of appropriate outer diameter was introduced over the guidewire followed by the endoscope in a side-by-side manner. The electrode was positioned 1 cm above the proximal extent of the BE, the balloon was inflated, and energy was delivered (12 J/cm 2,40 W/cm 2 ), resulting in circumferential ablation of a 3-cm segment. The ablation catheter was repositioned distal to the prior ablation zone, allowing minimal overlap, and ablation was repeated until the entire BE was ablated. The ablation catheter was removed to clean the electrode surface, while a soft distal attachment cap (model MB-046; Olympus, Tokyo, Japan) was mounted on the endoscope and used to clean the ablation zone. Residual debris then was removed by forceful spraying of water through a spraying catheter with a pressure pistol (Alliance; Boston Scientific, Limerick, Ireland). After cleaning, the ablation catheter was re-introduced and the BE segment was treated a second time. For secondary focal ablation, each island or tongue of residual Barrett s mucosa was aligned with the focal HALO 90 catheter, the endoscope was deflected to bring the electrode into contact, and RF energy was delivered twice in succession (15 J/cm 2, 40 W/cm 2 ). All areas were cleaned by pushing the coagulum off with the leading edge of the device. The electrode was removed to clean its surface, and then re-introduced to re-treat all areas twice more in succession. In all patients, the gastroesophageal junction was treated circumferentially with the HALO 90 catheter at least once (2 2, 15 J/cm 2 ), to ensure eradication of IM at this level. Patient Flow Baseline visible lesions were removed by ER, followed by primary RFA at least 6 weeks after the ER and within 3 months after the last biopsy session. Additional RFA sessions were scheduled every 8 weeks thereafter until complete endoscopic eradication was achieved. A maximum of 2 circumferential and 3 focal RFA sessions was allowed. Any residual Barrett s epithelium persisting after the maximum number of allowable RFA sessions was removed with a focal ER as escape therapy. Two months after the last treatment session the original extent of BE was biopsied every 1 cm, including immediately ( 5 mm) distal to the neosquamocolumnar junction. If patients had reached complete resolution of intestinal metaplasia and complete resolution of neoplasia they were scheduled for follow-up endoscopy and 4-quadrant biopsies at 6 and 12 months after the last treatment session, and annually thereafter (Figure 1). Outcome Variables The primary end points were histology-based (biopsy specimens obtained 2 months after the last therapeutic intervention). A complete response was defined as all biopsies negative for IM (complete response [CR]-IM) and neoplasia (CRneoplasia), reported separately. Secondary end points were as follows: disease progression, adverse events, and durability of CR-IM and CR-neoplasia at the last biopsy available. Histologic Analysis All ER specimens and biopsy specimens from baseline and follow-up evaluation were evaluated by the central study pathologist (F.J.t.K.), an expert in gastrointestinal pathology. ER specimens were evaluated for neoplasia according to the World Health Organization classification, 27 tumor infiltration depth, differentiation, presence of lymphatic or vascular infiltration, and completeness of resection at the vertical margin. Biopsy specimens were evaluated for the presence of IM, LGIN, HGIN, and EC, and follow-up biopsy specimens from the neosquamous epithelium were evaluated for the presence of subsquamous areas of IM. Statistical Analysis Statistical analysis was performed with SPSS 16.0.2 software for Windows (SPSS Inc, Chicago, IL). For descriptive statistics, mean ( standard deviation) was used in case of a normal distribution of variables, and median (interquartile range [IQR]) was used for variables with a skewed distribution.

January 2010 ABLATION OF NEOPLASTIC BARRETT S ESOPHAGUS 25 Figure 1. Flow-chart of the study protocol. ND-BE, nondysplastic Barrett s esophagus; EGD, esophagogastroduodenoscopy. Where appropriate, the Student t test and the Mann Whitney test were used. Results Enrollment and Baseline Characteristics Twenty-four patients were included (20 men; mean age, 65 9.8 y; median BE length, C6M8; IQR, C2 9, M4 10). Twenty patients had a hiatal hernia (median, 2 cm; IQR, 2 3). Sixteen patients had a baseline diagnosis of cancer; 8 patients had HGIN as the worst histologic finding at 2 work-up endoscopies. Baseline Endoscopic Resection Twenty-three patients (96%) underwent a total of 25 ER sessions before RFA (2 patients had 2 ER sessions). Of the 25

26 POUW ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 1 ER sessions, 12 were performed with the ER-cap technique, 12 with the multiband mucosectomy technique, and 1 with endoscopic submucosal dissection. The worst histologic grade based on ER was EC in 16 patients (T1m2, n 8; T1m3, n 6; and T1sm1, n 2) and HGIN in 7 patients. The worst grade of residual intraepithelial neoplasia in biopsy specimens obtained from the remaining BE during at least 2 high-resolution endoscopies after any ER was HGIN in 10 patients (in the absence of any visible lesions), LGIN in 11 patients, and 3 patients had IM without neoplasia. Number of Treatment Sessions Patients underwent a median of 1 (IQR, 1 1) primary circumferential and 1 (IQR, 1 2) secondary focal RFA session. Escape ER was necessary in 1 patient to remove a resistant 8-mm Barrett s island (LGIN). Escape ER was necessary in a second patient to remove 4 small BE foci (IM, no neoplasia) because of difficulty in introducing the focal RFA device. Overall, patients required a median number of 3 (IQR, 3 4) therapeutic interventions (including any ER before and after RFA), during a median period of 6.4 months (IQR, 5.5 11.3 mo). Primary Outcome Variables Eradication of intraepithelial neoplasia and intestinal metaplasia. In 20 of 21 patients with residual LGIN/ HGIN in their BE after ER and before ablation, CR-neoplasia was achieved with RFA (95%). Escape ER for an 8-mm island of LGIN resulted in CR-neoplasia in all 21 patients (100%). CR-IM was achieved with RFA in 21 of 24 patients (88%). After escape ER in 2 patients (see earlier), CR-IM was reached in 23 of 24 patients (96%). The single patient in whom CR-IM was not achieved had a C10M10 nonneoplastic Barrett s segment after ER of a T1m2 carcinoma at baseline and did not regenerate neosquamous epithelium readily after RFA. Thus, he was removed from the trial. Secondary Outcome Variables Progression. During the follow-up period, there were no new cancers and no histologic progression of disease. Adverse events. Endoscopic resection related complications. One severe complication, an esophageal perforation, occurred after baseline ER (cap technique), and was treated nonsurgically with clips and a covered stent. This patient underwent a second ER session 3 months after the perforation and went on to have RFA 2 months after the second ER, without further complication, and is now CR-IM at 2, 6, and 12 months. Radiofrequency ablation related complications. No severe complications occurred after RFA. Two complications (8%) were graded as being of moderate severity because they required additional endoscopic procedures: 1 patient presented with melena 2 weeks after focal ablation. At endoscopic inspection no active bleeding was observed, but 2 visible vessels in the ablated area were clipped preventively. A second patient developed dysphagia after the first RFA. He had undergone widespread ER before RFA and had a mucosal laceration after circumferential RFA. The stenosis was resolved with 5 endoscopic dilatation sessions, after which ablation treatment could be resumed. Mucosal laceration was observed in 5 patients (21%) after circumferential RFA, all occurring within the mucosal scar region of a prior ER (Figure 2). All patients remained asymptomatic and no therapeutic interventions were required. These lacerations therefore were graded as mild complications. Durability of complete response. After a median follow-up period of 22 months (IQR, 17.2 23.8 mo) after the last treatment, and a median of 3 follow-up endoscopies (IQR, 3 4) per patient, no recurrence of neoplasia was observed. Long-term complete response for IM was maintained in 20 of 24 included patients (83%). In one patient with baseline C9M10 and CR-neoplasia and CR-IM at 2 months, a tiny (0.5 3 mm) glandular island with IM at biopsy was found at the 6-month follow-up endoscopy. The island was located distal to a reflux stenosis and was detected only after inspection with narrow-band imaging (NBI) (Figure 3). In 2 patients at 6 months, focal nonneoplastic IM was found in one biopsy each, obtained distal to the gastroesophageal junction. In neither patient was this finding reproduced at 12 months. No buried Barrett s was found in a total of 1201 neosquamous biopsy specimens obtained during the follow-up evaluation. Discussion There is no generally accepted management strategy for patients with early neoplasia in BE. However, studies in which focal ER for neoplasia was followed by surveillance of residual BE have reported a 20% to 30% rate of metachronous lesions, 3 5 whereas studies in which the whole BE was eradicated after focal ER only reported recurrence in up to 6% of patients. 9 11 Based on these data, we believe that combining ER with complete ablation of residual Barrett s mucosa is the preferable treatment approach, and in this multicenter European trial we therefore evaluated RFA in conjunction with ER for the treatment of BE containing neoplasia and early cancer. The primary end point of complete eradication of all residual neoplasia after ER was achieved in 21 of 21 patients (CR-neoplasia, 100%). The primary end point of complete eradication of all residual IM after ER was achieved in 23 of 24 patients (CR-IM, 96%). These results comport with those from 2 recently published, singlecenter studies in which RFA was used in conjunction with ER for HGIN and EC to achieve 100% CR-IM and 100% CRneoplasia. 24,25 These favorable outcomes collectively endorse a central role for RFA in the treatment of patients with flat-type neoplasia in a BE. All but one patient in this study underwent baseline ER for visible lesions. This emphasizes that in our opinion ER should be performed not only for nodules, but for any mucosal irregularity or area with suspicious glandular patterns no matter how subtle, even if prior biopsies do not show cancer but HGIN or lower. Not only is baseline ER important to render the mucosa flat for subsequent ablation, but it provides an adequate specimen for histopathologic analysis and has been shown to change the histologic diagnosis on the basis of prior biopsies in 49% of patients. 28 Such a rigorous baseline evaluation is of the utmost importance to enable selection of patients who are eligible for further endoscopic treatment with RFA, that is, patients without deep submucosal invading cancer ( T1sm1), or poorly differentiated cancer, versus those who require surgical intervention. In the case of minimal submucosal infiltration (T1sm1), recent studies have suggested that the risk of lymph node metastasis is very low. 8,29 Two patients had minimal submucosal invasion diagnosed in the ER specimen. Because of the absence of other risk factors for lymph node

January 2010 ABLATION OF NEOPLASTIC BARRETT S ESOPHAGUS 27 Figure 2. Endoscopic images of circumferential RFA, complicated by nontransmural mucosal laceration. (A) C11M12 BE. (B) At 32 cm from the incisors a suspicious lesion was observed. (C) The lesion was removed in 2 pieces using multiband mucosectomy. Histology showed HGIN. (D) Before circumferential RFA, 6 weeks after ER, esophageal scarring was observed at the resection site. (E and F) During circumferential RFA, inflation of the balloon catheter caused nontransmural mucosal laceration, probably because of overstretching at the level of the ER scar. (G) Three months after primary circumferential RFA, 2 residual islands of Barrett s mucosa were detected with NBI. (H) The islands were treated with focal RFA. (I L) Two months after the last ablation, complete endoscopic eradication of neoplasia and IM was reached, and the esophagus was covered with normal-appearing neosquamous epithelium, as seen with white-light endoscopy and NBI. metastasis (poor differentiation grade, irradical resection, lymphatic/vascular infiltration), surgical as well as endoscopic treatment options were discussed elaborately with the patients, and both decided to be treated endoscopically within the study protocol. In a prior clinical study, 24 we allowed only single-piece baseline ER before RFA. In a subsequent study, 25 we relaxed this restriction to allow multiple-piece ER (median, 2 specimens; IQR, 2 3 specimens) before RFA. No mucosal lacerations occurred and a stricture occurred in only one patient after 2 ER sessions and RFA. In the present study, based partially on prior results, we did not initially restrict the extent of baseline ER. During the first half of this study, we noted 5 mucosal laceration events within the ER scar zone during circumferential RFA and one stenosis; and we related this to extensive baseline ER and scarring (Figure 2). None of the lacerations required intervention or caused complaints, therefore they were regarded as mild complications. However, because lacerations may provoke severe bleeding or esophageal perforation, the investigator group added a restriction to the baseline extent of ER (maximum 50% of circumference; 2-cm length, and 1 session). We also used a more conservative approach in selecting an ablation catheter balloon size, erring on the small size. After this modification, no further lacerations or stenoses were observed. One patient was removed from the trial and considered a failure because of poor healing of the esophagus after circumferential RFA. At baseline, the patient had an EC resected with ER, leaving a residual C10M10 Barrett s segment with no neoplasia. The patient previously had undergone a Nissen fundoplication and postoperative 24-hour ph-metry was normal off proton pump inhibitor treatment. Yet, despite a maximum acid-suppressant regimen (esomeprazole 80 mg twice a day, ranitidine 300 mg at bedtime, sucralfate 5 ml 4 times/d) the esophagus failed to regenerate readily with neosquamous epithelium. It is unclear what factors were at play in this particular patient s poor healing, but it highlights the fact that tailored management may be necessary in some cases. In 3 patients (83%) with complete eradication of IM and neoplasia, IM was found in biopsy specimens during the follow-up evaluation, which raises the question of whether endoscopic surveillance will remain necessary after complete eradication of IM and neoplasia has been reached. Currently, endoscopic follow-up evaluation after successful eradication of BE by RFA is recommended at regular intervals based on base-

28 POUW ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 8, No. 1 Figure 3. Endoscopic images of a small island of glandular mucosa detected during the follow-up evaluation. (A and B) C9M10 BE with proximal reflux stenosis, containing diffuse HGIN after ER for early cancer. (C and D) Residual Barrett s mucosa was converted completely to neosquamous mucosa by 1 circumferential and 1 focal RFA. No recurrence of neoplasia or IM was found at the 2-month follow-up evaluation. (E and F) At the 6-month follow-up evaluation, however, NBI revealed a tiny island of columnar epithelium just distal to the reflux stenosis and the biopsy showed IM. line histologic diagnosis because no long-term follow-up data are as yet available. In two thirds of patients with IM during the follow-up evaluation, focal nondysplastic IM was found in a single biopsy just distal to the neosquamocolumnar junction at a single follow-up endoscopy and in neither patient was this finding reproduced during following endoscopies. Because all patients had an initial diagnosis of HGIN/EC, one may argue that residual IM in the cardia reflects incomplete cure of the underlying disease. However, IM of the cardia can be detected in up to 25% of patients with a normal-appearing squamocolumnar junction and is not considered premalignant in those cases. 30 The clinical relevance of this finding is thus unclear, but in our opinion nondysplastic IM in biopsies distal to the neosquamocolumnar junction do not require additional treatment, whereas IM with LGIN or HGIN should be treated. However, long-term follow-up data of these patients may clarify how focal IM in the cardia will behave after RFA. One patient had a 0.5 3 mm island of columnar epithelium detected, with IM at biopsy, during the follow-up endoscopy at 6 months, despite being CR-IM at 2 months. The island was observed only after inspection with NBI (Figure 3). We hypothesize that we missed this island visually on prior follow-up endoscopy because of its limited size and position just distal to a reflux stenosis. It is in our opinion therefore recommended to use high-resolution endoscopy, NBI, Lugol s chromoendoscopy, or comparable techniques during follow-up evaluation after RFA to exclude the presence of residual BE tissue, especially small islands. As described in this article, the selection of patients for endoscopic treatment involves a thorough endoscopic work-up, the possibility to perform ER safely, and accurate histologic evaluation of tissue specimens for the presence of risk factors for lymph node metastasis. 31 Because all patients in this study underwent endoscopic work-up and RFA in specialized tertiarycare centers, the high reported safety and efficacy should be extrapolated to general practice with care. In our opinion, it may be advisable to centralize RFA of BE patients with HGIN/EC in centers with multidisciplinary experience in this field (ie, expertise with endoscopic imaging and ER, access to esophageal surgery, expert gastrointestinal histopathology) and that have participated in dedicated RFA training courses at expert centers. One strength of this study was the use of a single expert gastrointestinal pathologist for all specimen reviews. Other strengths included the multicenter design, use of baseline ER for histologic staging, post-er and pre-rfa biopsy regimen to elucidate the post-er diagnosis, observation of all RFA procedures and primary end point biopsy procedures by the central study monitor to ensure protocol adherence, and patient follow-up evaluation that extended a median of 22 months beyond the last therapeutic intervention. One weakness of this study was that many of the baseline ER procedures were performed before patient enrollment during confirmation of eligibility, and therefore were not supervised by the study coordinator. However, all centers used standardized techniques and data collection, and all ER specimens were reviewed by the study pathologist. Also, we allowed focal escape ER in certain rare circumstances, which could bias our outcomes favorably. Furthermore, midway through the study we necessarily adjusted the extent of permissible baseline ER. This restriction appears to be associated with a lower risk of complications (namely, mucosal laceration), but more patients are needed to make this determination. Lastly, a relatively small number of patients were included in this study and treated in 3 expert centers. We therefore initiated a second multicenter study in 12 European centers that uses the inclusion criteria defined in this study and aims at including a minimum of 100 patients. The results of this European multicenter study show that patients with early neoplasia arising in BE can be treated effectively and safely with RFA, in combination with prior ER of visible lesions. References 1. Hameeteman W, Tytgat GN, Houthoff HJ, et al. Barrett s esophagus: development of dysplasia and adenocarcinoma. Gastroenterology 1989;96:1249 1256. 2. Enzinger PC, Mayer RJ. Medical progress: esophageal cancer. N Engl J Med 2003;349:2241 2252. 3. Peters FP, Kara MA, Rosmolen WD, et al. Endoscopic treatment of high-grade dysplasia and early stage cancer in Barrett s esophagus. Gastrointest Endosc 2005;61:506 514. 4. May A, Gossner L, Pech O, et al. Local endoscopic therapy for intraepithelial high-grade neoplasia and early adenocarcinoma in Barrett s oesophagus: acute-phase and intermediate results of a

January 2010 ABLATION OF NEOPLASTIC BARRETT S ESOPHAGUS 29 new treatment approach. Eur J Gastroenterol Hepatol 2002;14: 1085 1091. 5. Pech O, Behrens A, May A, et al. Long-term results and risk factor analysis for recurrence after curative endoscopic therapy in 349 patients with high-grade intraepithelial neoplasia and mucosal adenocarcinoma in Barrett s oesophagus. Gut 2008;57:1200 1206. 6. Bollschweiler E, Baldus SE, Schröder W, et al. High rate of lymph-node metastasis in submucosal esophageal squamouscell carcinomas and adenocarcinomas. Endoscopy 2006;38: 149 156. 7. Buskens CJ, Westerterp M, Lagarde SM, et al. Prediction of appropriateness of local endoscopic treatment for high-grade dysplasia and early adenocarcinoma by EUS and histopathologic features. Gastrointest Endosc 2004;60:703 710. 8. Westerterp M, Koppert LB, Buskens CJ, et al. Outcome of surgical treatment for early adenocarcinoma of the esophagus or gastroesophageal junction. Virchows Arch 2005;446:497 504. 9. Peters FP, Kara MA, Rosmolen WD, et al. Stepwise radical endoscopic resection is effective for complete removal of Barrett s esophagus with early neoplasia: a prospective study. Am J Gastroenterol 2006;101:1449 1457. 10. Seewald S, Akaraviputh T, Seitz U, et al. Circumferential EMR and complete removal of Barrett s epithelium: a new approach to management of Barrett s esophagus containing high-grade intraepithelial neoplasia and intramucosal carcinoma. Gastrointest Endosc 2003;57:854 859. 11. Buttar NS, Wang KK, Lutzke LS, et al. Combined endoscopic mucosal resection and photodynamic therapy for esophageal neoplasia within Barrett s esophagus. Gastrointest Endosc 2001; 54:682 688. 12. Overholt BF, Panjehpour M, Halberg DL. Photodynamic therapy for Barrett s esophagus with dysplasia and/or early stage carcinoma: long-term results. Gastrointest Endosc 2003;58:183 188. 13. Peters F, Kara M, Rosmolen W, et al. Poor results of 5-aminolevulinic acid-photodynamic therapy for residual high-grade dysplasia and early cancer in Barrett esophagus after endoscopic resection. Endoscopy 2005;37:418 424. 14. Sharma P, Wani S, Weston AP, et al. A randomised controlled trial of ablation of Barrett s oesophagus with multipolar electrocoagulation versus argon plasma coagulation in combination with acid suppression: long term results. Gut 2006;55:1233 1239. 15. Van Laethem JL, Jagodzinski R, Peny MO, et al. Argon plasma coagulation in the treatment of Barrett s high-grade dysplasia and in situ adenocarcinoma. Endoscopy 2001;33:257 261. 16. Schulz H, Miehlke S, Antos D, et al. Ablation of Barrett s epithelium by endoscopic argon plasma coagulation in combination with high-dose omeprazole. Gastrointest Endosc 2000;51:659 663. 17. Ganz RA, Utley DS, Stern RA, et al. Complete ablation of esophageal epithelium with a balloon-based bipolar electrode: a phased evaluation in the porcine and in the human esophagus. Gastrointest Endosc 2004;60:1002 1010. 18. Dunkin BJ, Martinez J, Bejarano PA, et al. Thin-layer ablation of human esophageal epithelium using a bipolar radiofrequency balloon device. Surg Endosc 2006;20:125 130. 19. Smith CD, Bejarano PA, Melvin WS, et al. Endoscopic ablation of intestinal metaplasia containing high-grade dysplasia in esophagectomy patients using a balloon-based ablation system. Surg Endosc 2007;21:560 569. 20. Sharma VK, Wang KK, Overholt BF, et al. Balloon-based, circumferential, endoscopic radiofrequency ablation of Barrett s esophagus: 1-year follow-up of 100 patients. Gastrointest Endosc 2007;65:185 195. 21. Fleischer DE, Overholt BF, Sharma VK, et al. Endoscopic ablation of Barrett s esophagus: a multicenter study with 2.5-year followup. Gastrointest Endosc 2008;68:867 876. 22. Sharma VK, Kim HJ, Das A, et al. A prospective pilot trial of ablation of Barrett s esophagus with low-grade dysplasia using stepwise circumferential and focal ablation (HALO system). Endoscopy 2008;40:380 387. 23. Shaheen NJ, Sharma P, Overholt BF, et al. Radiofrequency ablation in Barrett s esophagus with dysplasia. N Engl J Med 2009; 360:2277 2288. 24. Gondrie JJ, Pouw RE, Sondermeijer CM, et al. Stepwise circumferential and focal ablation of Barrett s esophagus with highgrade dysplasia: results of the first prospective series of 11 patients. Endoscopy 2008;40:359 369. 25. Gondrie JJ, Pouw RE, Sondermeijer CM, et al. Effective treatment of early Barrett neoplasia with stepwise circumferential and focal ablation using the HALO system. Endoscopy 2008;40:370 379. 26. Sharma P, Dent J, Armstrong D, et al. The development and validation of an endoscopic grading system for Barrett s esophagus: the Prague C& M criteria. Gastroenterology 2006;131: 1392 1399. 27. Hamilton SR, Aaltonen LA, eds. WHO classification. Tumours of the digestive system. Lyon: IARC Press, 2000. 28. Peters FP, Brakenhoff KP, Curvers WL, et al. Histologic evaluation of resection specimens obtained at 293 endoscopic resections in Barrett s esophagus. Gastrointest Endosc 2008;67:604 609. 29. Manner H, May A, Pech O, et al. Early Barrett s carcinoma with low-risk submucosal invasion: long-term results of endoscopic resection with a curative intent. Am J Gastroenterol 2008;103: 2589 2597. 30. Morales TG, Camargo E, Bhattacharyya A, et al. Long-term follow-up of intestinal metaplasia of the gastric cardia. Am J Gastroenterol 2000;95:1677 1680. 31. Pouw RE, Sharma VK, Bergman JJ, et al. Radiofrequency ablation for total Barrett s eradication: a description of the endoscopic technique, its clinical results and future prospects. Endoscopy 2008;40:1033 1040. Reprint requests Address requests for reprints to: Jacques J. Bergman, MD, PhD, Department of Gastroenterology and Hepatology, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. e-mail: j.j.bergman@amc.uva.nl; fax: (31) 20-6917033. Conflicts of interest The authors disclose the following: Roos Pouw is supported by an unrestricted research grant from AstraZeneca BV (The Netherlands); Jacques Bergman has received research support in the form of grants and materials for conducting studies from BÂRRX Medical (Sunnyvale, CA), AstraZeneca BV (The Netherlands), and Olympus Endoscopy (Tokyo, Japan). The remaining authors disclose no conflicts.