Healing of mucosal defects after endoscopic mucosal resection in the oesophagus is prone to result in varying degrees of stenosis, especially if the resected area is large and/or circumferential. A new technique was described where autologous cells from the patients own oral mucosa are harvested and cultures on proprietary membranes (coated, temperature sensitive). These membranes, coined cell sheets, are used to cover the mucosal defects.

Aim Our aim is to evaluate if mucosal wound healing after endoscopic resection of early cancers or pre-cancerous lesions in the esophagus can be enhanced by transplantation of autologous cell sheets, thus reducing the risk of esophageal stenosis.

Background Adenocarcinoma and Barrett's esophagus The incidence of esophageal adenocarcinoma (EAC) has increased rapidly in the western world during the last decades. The reason for this increase is unknown, however, there is a strong association between the risk of developing EAC and gastro esophageal reflux. Chronic reflux may lead to Barrett's esophagus (BE), a condition where the squamous epithelium of the esophagus is replaced by columnar epithelium(1). BE is a premalignant condition with a lifetime risk of progressing to EAC of 5% in men and 3% in women (2).

The rate of progression is dependent on the grade of dysplasia in BE. One meta-analysis showed that low-grade dysplasia had a cancer incidence of 17/1,000 patient-years while high-grade dysplasia had an incidence of 66/1,000 patient-years (3). Medical treatment can reduce reflux symptoms, but there is no proof that it can decrease the risk of progression.

Treatment High-grade dysplasia The American Gastroenterological Association guidelines recommend that patients with high-grade dysplasia should be treated with surgical resection, endoscopic therapy or be offered surveillance with endoscopy every 3 months (4).

The Japanese Society for Gastrointestinal Endoscopy (JGSE) has described that endoscopic mucosal resection (EMR) is suitable for BE with high-grade dysplasia given that the diameter is ≤2 cm, < 1/3 of the esophageal circumference and limited to the mucosa. One of several advantages with EMR is that the resected mucosa can be examined histologically, compared to ablative therapies. Briefly, EMR is performed by submucosal injection to separate esophageal wall layers and then resection of the mucosa with a snare. A similar technique is endoscopic submucosal dissection (ESD) which also includes a submucosal injection, but instead of using a snare for resecting, the lesion is dissected with cutting instruments, such as the insulation-tipped (IT) knife.

Another approach is radiofrequency ablation using a cylindrical balloon with electrodes attached. One study showed complete eradication in 81.0% of high-grade dysplasia patients. Using ablation techniques, no tissue-samples are possible to collect for histology and there might also be a risk for buried Barrett's metaplasia or dysplasia which can progress. Complications to radiofrequency ablation include chest pain, upper gastrointestinal bleeding and stenosis formation (4). Moreover, patients undergoing ablation therapy require life-long surveillance endoscopy (5).

Early esophageal cancer The best curative treatment for early adenocarcinoma of the esophagus is considered to be esophagectomy with extended lymphadenectomy (3). Esophagectomy is an extensive surgical procedure with a mortality rate ranging between 4.9- 18.0% and risk for serious post-operative complications ranging between 26-41% (6). There are also frequently occurring post-operative complaints such as reflux, diarrhea and dysphagia (3).

Endoscopic therapy modalities that initially were reserved for patients with HGD have been used as alternatives to open surgery for early esophageal cancer. Although the practice was pioneered in Asia, it is increasingly practiced in Europe and the US (7). Early cancers classified as M1 (limited to epithelial layer) and M2 (invading lamina propria) may be good candidates for endoscopic therapy as they are not associated with lymph node metastasis (4). Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) are the two resective techniques for removing mucosal tissue from the esophagus (8).

EMR, during which resection of the mucosa is performed with a snare after separation of the mucosa and the muscularis propria by submucosal injection of saline or hyaloronic acid injection is reserved for lesions < 15 mm in diameter (8). Larger lesions can be removed by so called piece meal resection, but the risk for an incomplete resection and subsequent local recurrence is substantial (9).

With ESD, where a submucosal dissection is performed after lifting and delineation of the lesion with a full-thickness mucosal circumferential cut much larger lesions can be removed en bloc, reducing the risk for an incomplete resection (8, 9).

Complications to EMR and ESD A common complication after both EMR and ESD is stenosis of the esophagus that can result in dysphagia. Mucosal defects larger than ¾ of the esophageal circumference are strongly associated with the development of stenosis. This can be treated by repetitive balloon dilatations. Takahashi et al retrospectively analyzed 437 patients treated with EMR/ESD. 90 patients had mucosal defects > ¾ of the esophageal circumference, of which 70 developed strictures requiring dilatation (1-45 dilatations per patient with a dilatation treatment time of 1-51 months). During dilatation therapy 7 perforations and 1 bleeding occurred (9).

Cell sheet transplantation In 2006, Ohki et al published an article on how to prevent inflammation after aggressive endoscopic resection. Scarring and stenosis of esophagus is caused by severe inflammation. In a canine model they tried to improve wound healing and reduce constriction. They created transplantable cell sheets from autologous oral mucosal epithelial cells (OMEC). In six animals esophageal ulcers were created by ESD. In three of the animals, two cell sheets were transplanted. Neither of the groups showed stricture, probably owing to the thicker muscle layer in dogs. However, the study showed that the transplant group had an enhanced wound healing with less inflammation. The results imply that cell sheet transplantation could potentially prevent stenosis after extensive mucosal resections. The cells did not cover the whole ulcer, but secretion of growth factors and cytokines are believed to have accelerated healing from native tissue. The harvested cell sheets still had intact extra cellular matrix which may act as glue to the tissue, making sutures unnecessary (10).

With the background of the above mentioned study Ohki et al have performed a feasibility study in humans (unpublished data), that thus far has shown very promising results (by direct communication with Dr Ohki).

Project description Study design Part 1 A feasibility study in patients with early adenocarcinoma or HGD in Barrett's mucosa where endoscopic resection is clinically indicated.

Part 2 A randomized controlled trial comparing ESD/EMR without cell sheet transplantation and ESD/EMR with cell sheet transplantation with respect to mucosal wound healing time and stricture formation .

Endpoints

1. Primary endpoint: time to complete wound healing
2. Secondary endpoints: esophageal stricture occurrence, inflammatory response measured with CellVizio, quality of life questionnaire (QLQ-OES18) Patients Inclusion criteria: Patients with intramucosal adenocarcinoma or high grade dysplasia in Barrett's esophagus where endoscopic resection is indicated in accordance with clinical guidelines.

Exclusion criteria:

Inability to give informed consent due to mental disorder, language barrier etc.

Power calculation:

Study part 1: The feasibility study in Western patients with adenocarcinoma, 10 patients will be recruited. No formal power calculation is made for the feasibility study.

Study part 2: Based on the results of the Japanese feasibility study wound healing was x % more rapid among cell sheet transplanted patients compared to expected healing time without transplantation. Assuming a difference of healing time of 30%, based on these Japanese data 12 patients would be required in each group to obtain a power of 80% with alpha= 0.05.

Preparation and transplantation of cell sheets Isolation and culturing of oral mucosal epithelial cells (OMEC) The patients' oral cavity will be sterilized with iodine (??) and local anesthesia will be administered to the inside of cheek. A 5 mm punch biopsy will be taken and the wound sutured. The biopsy will be treated with dispase to separate the epithelial cells, and after that they will be exposed to trypsin to create a single-cell suspension. The single-cell suspension will be seeded to temperature-responsive cell dish inserts and cultured for two weeks. No animal or bacterial derived material will be used for the culturing.

Harvesting of cell sheets When the endoscopists approximate that there is 40 minutes left of the ESD the investigators will put the dishes in a 20°C incubator for 30 minutes. The cell sheets will be transferred to a support membrane and transplanted to the esophageal ulcer by endoscopic forceps. A light pressure will be applied to improve attachment. Depending on the size of the ulcer, several cell sheets will be transplanted in the described manner.

Follow-up Weekly gastroscopies with Cellvizio. Weekly blood samples - (CRP, interleukins, Blood status) Quality of Life questionnaire (EORTC C30 and OES18)

Outcome measures:

1. Healing will be measured endoscopically by visualising and photo-documenting a completely intact mucosa during the weekly gastroscopies. Along with these endoscopical pictures, a grading of healing based on confocal laser endomicroscopy (pCLE) will be developed, as the pilot study (first part of the study) demonstrated feasibility.
2. Post-operative stenosis will be assessed by the easy passage of a standard 9 mm endoscope through the area of resection at weeks 1-4 and then in three-monthly intervals. Additionally, standard questionnaires will be used. Dysphagia is part of the EORTC OES18 module.

Relevance The investigators are hoping that this study will help us identify a method of preventing stenosis of esophagus after EMR/ESD. The stenosis commonly causes dysphagia that not only seriously reduces the quality of life but also could pose a threat for malnourishment.

Inclusion Criteria:

• Barrets esophagus, histologically confirmed, or early oesophageal cancer (both squamous and adeno)

Exclusion Criteria:

• no consent can be given
• prior ESD/EMR