Patients with Barrett's esophagus (BE) have a change in the lining of the esophagus. The normal one the lining of the esophagus changes to a lining similar to that of the intestine. The new mucosa has increased the risk of developing cancer. Usually this type of cancer is detected in a late phase and the patients' survival is low (less than 25% at 5 years). In daily practice, we strive to detect early cancerous lesions in order to treat them and cure them the patients. It has been widely demonstrated in BE patients that if cancer or precursor lesions are detected in an early phase, patients can be cured with endoscopic treatment. Endoscopic treatment of BE is based on endoscopic resection of the lesions / early cancer. After resection, patients have a 20-47% risk of developing cancer later in the remaining Barrett's esophagus. So there is a need to remove the remaining Barrett's mucosa that has not been resected. Several techniques can be used for removal of remaining BE: radiofrequency ablation, argon plasma, cryotherapy or endoscopic resection. The goal is to after resection of cancer and removal of residual Barrett's mucosa, a normal esophageal epithelium will cover the esophagus and dramatically reduce the risk for cancer development. The most widely used strategy for removal of residual Barrett's mucosa is radiofrequency ablation.

It is an easy technique to perform, but it is hindered by some factors: 1) it requires several treatment sessions; 2) is associated with complications in 11% of patients, such as severe pain, bleeding, stricture and perforation 3) Barrett's mucosal glands may grow under the new epithelium after treatment; 4) there is no histological assessment of what is ablated; 5) there is a need for continuous follow-up; 6) it there are high costs associated with this strategy; 7) This approach may cause physical and physiological burdens on patients due to continuous follow-up and lack of complete histological assessment. Endoscopic submucosal dissection (ESD) is an advanced endoscopic technique that enables resection of lesions or cancer in one piece and has been used extensively along the gastrointestinal tract. Studies have showed good effect of ESD for neoplastic BE. Karolinska has a lot of experience with ESD and has one of the largest the cohorts of ESD on BE patients. ESD of BE can be associated with complications such as bleeding and perforation in 2-3% in most published studies and in less than 1% each in our series. Another complication that can occur is narrowing of the the esophagus during the healing process after ESD. That risk was historically high and increased with the increase in the size of the resected specimen. The high risk of crowding out was the main inhibiting factor the development of ESD in the esophagus. With the introduction of steroid therapy to prevent narrowing a paradigm shift was formed and the corresponding narrowing risk was drastically reduced to between 2-33% in according to the size of the resections. In our series of 132 ESDs on Barrett's esophagus, 103 cases corresponded resections up to 75% of the luminal(?) circumference of the esophagus, in these only 4/103 (3.9%) had strictures and all were successfully treated with endoscopic balloon dilatation. In the remaining 29 ESDs: included resection more than 75% of the luminal circumference. In these, there was narrowing in 10/29 cases, all patients was successfully treated with endoscopic treatment. So preventive measures and thorough follow-up are associated with good results and safety profile, even in large ESD on BE. Several years ago did not perform ESD for the treatment of BE, due to the need for skilled endoscopists and the potential the risks of this procedure such as bleeding, perforation and strictures. Full resection of the BE mucosa allows complete resection of all mucosa at risk, with complete histological assessment and virtually no risk of lesion presence in the margins or development of buried glands. It leads to complete removal of BE and may lead to the need for additional follow-up. With this study, we want to test the efficacy and safety of ESD for the removal of all Barrett's mucosa, instead of the more common approach of resection of Barrett's cancer followed by ablation of the remaining BE.

Research plan Title Complete resection of Barrett's esophagus harboring neoplasia with endoscopic submucosal dissection. Background Barrett esophagus (BE) is a stablished risk factor for the development of esophageal adenocarcinoma. Current guidelines recommend complete eradication of Barrett's mucosa when neoplasia is detected, due to the risk of synchronous or metachronous neoplasia in 20-47% of cases [1, 2]. Advocated treatment demands resection of visible lesions and ablation of remaining Barrett's mucosa. This strategy is hampered by some factors: 1) it demands several treatment sessions (1 for resection of lesions and 2 or more for ablation); 2) is associated with complications in 11% of patients, such as severe pain, bleeding, stricture and perforation in up to 1-2%, 1%, 7 % and 0.6 % of cases, respectively [3-5]; 3) buried glands can grow beneath the neosquamous epithelium after treatment; 4) there is no histological assessment of what is ablated; 5) there is the need of continuous follow-up; 6) there are high costs associated with this strategy; 7) this approach may cause physical and physiological burden to the patients due to continuous follow-up and lack of full histological assessment. Endoscopic submucosal dissection (ESD) enables en bloc resection of neoplastic lesions and has been widely used along the GI tract. Initial reports of ESD on BE were discouraging due to high rate of positive lateral margins and significant rate ofcomplications . The lack of lateral free margins in those studies can be attributed to suboptimal diagnostic and endoscopic techniques.. More recent studies have shown good efficacy of ESD for neoplastic BE, with lateral R0 rates of 75-90% [6]. In our early series, of neoplastic BE treated with ESD, the rates of en bloc, R0 and curative resection were 99%, 86% and 72%, respectively. ESD of Barrett's esophagus might be associated with complications such as bleeding and perforation in 2-3% in most published studies, and in less than 1% each in our series. Another complication that might occur is stricture formation in the esophagus during the healing process after the ESD. That risk was historically high, increasing with increase in size of the resected specimen and with increase in the percentage lumen circumference resected. Historically, the rates of strictures after esophageal ESD were of > 30% in small resections, > 50% when more than 75% of the lumen circumference was resected and virtually 100% in cases of circumferential resection.

That high rate of strictures was the major factor hampering the development of ESD in the esophagus. With the introduction of steroids treatment for the prevention of stricture formation, there was a paradigm shift, and the corresponded stricture rates turn to be 2-10% in small resections [6], 10-14% in resections as big as > 75% of lumen circumference and 27-33% in cases of circumferencial ESD. Most of these studies evaluated the role of subepithelial injection of triamcinolone and the use of oral steroids [7-9]. In our series of 132 ESDs on Barrett's esophagus, 103 cases corresponded to resections up to 75% of luminal circumference, in those, only 4/103 (3.9%) had strictures and all were successfully treated with endoscopic balloon dilation. In the remaining 29 ESD, resection engaged more than 75% of luminal circumference. In these there was formation of stricture in 10/29 cases, all patients were successfully treated with balloon dilation and one with a temporary esophageal stent. Considering the 9 patients with Barrett's neoplasia that were treated by our team with circumferential ESD, 3/9 did not have stricture and the remaining 6/9 were successfully treated with endoscopic treatment (balloon dilation or temporary stent). So, preventive measures and close follow-up are associated with good outcome and safety profile, even in large ESDs on BE. Full resection of BE mucosa enables complete resection of all mucosa at risk, with full histological assessment and virtually no risk for presence of lesion in the margins or development of buried glands. Aims To evaluate the role of full resection of neoplastic Barrett's esophagus in one endoscopic session as an alternative to conventional treatment with partial endoscopic resection in one session, followed by several sessions of ablation of Barrett's mucosa and further follow-up endoscopies. We want to test the efficacy and safety of this approach of complete resection of Barrett's esophagus in patients with Barrett's neoplasia. We aim to do so, through the evaluation of resected specimens and the rates of complications such as bleeding, perforation and stricture. Furthermore, we aim to evaluate the histological results after full endoscopically resection comparing it with the previous histological assessment with conventional endoscopy and biopsies (prior to treatment).We also want to study this strategy in terms of early recurrence of Barrett's metaplasia and dysplasia, patient quality of life and in costs reduction. The main endpoints are the rates of complications such as bleeding, perforation and stricture. Secondary endpoints include: 1) length of the procedures; 2) recurrence of intestinal metaplasia or and dysplasia; 3) HRQL status; 4) costs 5) rates of en bloc, R0 and curative resections. Material and methods Study design This is a pilot study that will be run by Karolinska University Hospital. Study population Patients with neoplastic long-segment Barrett's esophagus will be invited to participate in this study. The study will include 40 patients (interim analysis will be performed after inclusion of the first 10 and 20 patients). Inclusion criteria

• Long segment Barrett's esophagus with circular component >= 3cm and maximum extent of 9cm.
• Presence of neoplastic Barrett's esophagus on previous biopsies confirmed by expert pathologist. Exclusion criteria
• Refractory esophagitis
• Presence of signs of deep submucosal invasion (>sm1)
• Anticoagulation therapy that cannot be discontinued (ASA allowed)
• Barrett's esophagus length > 9 cm (C and/or M)
• Immunosuppression that would contraindicate use of steroids.
• Diabetes Mellitus All patients will receive a detailed explanation regarding the procedure, its risks, benefits, and alternatives, and will provid written informed consent before inclusion (Appendix 1). Participants clinical, demographic, and procedural information will be obtained and recorded at enrollment (Appendix 2). Patients under antiplatelet drugs or anticoagulants will be handled according to published guidelines. OUTCOME PARAMETERS PRIMARY OUTCOME PARAMETERS Adverse events are defined as any complication in which ESD or ESD related procedures (such as anesthesia) are a contributing factor and include bleeding, perforation, stricture and aspiration pneumonitis. Perforation is defined as a full thickness breach in the GI wall with or without symptoms.

Intraprocedural bleeding is considered significant and a complication when:

• blood transfusion is needed

• premature termination of endoscopic occurs

• there is a drop in hemoglobin >= 2g/dL Delayed bleeding is defined as clinical evidence of bleeding manifested by melena or hematochezia from up to 14 days after the procedure that demanded presentation to the emergency department, hospital admission or medical intervention. Aspiration pneumonitis is defined as acute lung injury after the inhalation of regurgitated gastric contents, diagnosed by the development of new radiographic infiltrate and the presence of symptoms or signs suggestive of lower respiratory tract infection. Stricture: esophageal stricture is defined as

  • Inability to swallow solid food which results in the need for dilation, and/or
  • Inability to pass a standard diagnostic endoscope. ESD related mortality is considered as any death during or after the procedure in which the ESDs and all related procedures (such as anesthesia) were a contributing factor.

Adverse events are classified as:

• acute (during procedure)
• early (<48 h)
• late (>48 h).

In terms of severity they are characterized as:

• mild (unplanned medical assistance, hospital admission, hospitalization <3 days, hemoglobin drop <3 g, no transfusion)
• moderate (4-10 days hospitalization, <4 units blood transfusion, repeat endoscopic intervention, radiological intervention)
• severe (hospitalization >10 days, intensive care unit (ICU) admission, need for surgery, >4 units blood transfusion) or fatal (death attributable to procedure) [7, 8].

SECONDARY OUTCOME PARAMETERS En bloc resection is defined as resection of the aimed area in one fragment. R0 resection corresponds to en bloc resection with histologically free horizontal and vertical margins.

Curative resection corresponds to:

• intramucosal and R0 or
• superficial submucosal invasion up to SM1 and low histological risk criteria (highly differentiated andno lymphovascular invasion) and R0 Recurrence was defined by the presence of suspicious neoplastic tissue in the postESD scar under high-definition endoscopy confirmed by histology.

STATISTICAL ANALYSIS Demographic and baseline characteristics are summarized using descriptive statistics.

Mean (± SD) and t test used in case of normal distribution, or presented as median (interquartile range, IQR) and compared with the Mann-Whitney test for skewed distribution, in continuous variables. Categorical data are presented as percentages and were compared with the Chi-Square test. The significance threshold is 0.05 for all analyses. Multiple logistic regression with backward stepwise variable selection is used to identify the independent predictors of outcomes of interest. Statistical analyses were performed using a statistical software package (Statistical Package for the Social Sciences).

Significance With this study it will be possible to evaluate the strategy of full endoscopic resection for endoscopic treatment of neoplastic Barrett's esophagus. Contrary to current Barrett's esophagus management (partial resection followed by ablation), it will enable the full histological assessment of Barrett's esophagus.

In case of positive results: good curative resection rates (> 90%), very low recurrence rates (<10%) and treatable complications, this strategy may be further validated and might change the paradigm of Barrett's esophagus treatment.

Inclusion Criteria:- Long segment Barrett's esophagus with circular component >= 3cm and maximum extent of 9cm.

• Presence of neoplastic Barrett's esophagus on previous biopsies confirmed by expert pathologist.

Exclusion Criteria:Refractory esophagitis

• Presence of signs of deep submucosal invasion (>sm1)
• Anticoagulation therapy that cannot be discontinued (ASA allowed)
• Barrett's esophagus length > 9 cm (C and/or M)
• Immunosuppression that would contraindicate use of steroids.
• Diabetes Mellitus