Endoscopic submucosal dissection (ESD) is the main treatment for esophageal early cancer. But post-ESD stricture has a serious impact on quality of life.When the lesion range reaches up to 3/4 circumferential mucosal defect or the longitudinal length over 5cm, the esophageal stricture rate will be over 70% or more, and when the whole circumferential mucosal defect, the rate is up to 100%. Local triamcinolone (TA) injection has positive effect on post-ESD stricture.But there are no uniform standards and results of the TA usage. Investigators also found a disadvantage that TA injection was hard to be absorbed completely in the submucosal layer and spilled over easily, which may decrease the preventive effect. The investigators tried to make a better protocol that could adjust the time and frequency of triamcinolone injection to improve the medicine duration of stay and decrease the stricture rate. So, the aim of this study was to investigate the efficacy and safety of TA injection in a new way for preventing post-ESD stricture.

This is a single-center prospective intervention study, and single-arm trial too, which is carried out at Sun Yat-Sen University Cancer Center in Guangzhou, China. Participants recruited into the study must meet the standards as follows: (a). be staged as cT1aN0M0 or cT1bN0M0 before surgery via magnifying endoscopy, endosonography, CT, or PET/CT; (b). refuse conventional surgery and CRT and get ESD as first-line treatment; (c). mucosal defect measured by a circular protractor and reached more than three-quarters of the circumference of the esophageal lumen including whole circumference after ESD surgery; (d). sign the consent before ESD;(e). meet the minimum follow-up period of three months. In the protocol, TA injection combined with fibrin glue was used twice, immediately (D0) and 7days(D7) after ESD to increase TA's residual time and effect in the study group. Meanwhile, the stickiness of fibrin glue can protect the wound for preventing perforation or bleeding. The dosage of TA and fibrin glue were 100mg and 5-10ml in every session separately. The first target of this study was the frequency of stricture lasting at least three months follow-ups. And the second targets were the number of endoscopic bougie dilation or balloon dilation (EBD), stricture classification, stooler score and complications in the study group.

The expectational results are as follows: the first is that TA injection combined with firin glue spray can greatly decrease the esophgeal stricture rate after long-range ESD; the second is that TA injection combined with firin glue spray can also decrease the frequency of endoscopic bougie dilation or balloon dilation (EBD), complications, and the degree of stricture classification and stooler score compared with historical data. This study will show that TA injection combined with fibrin glue spray is a safe and effective treatment in preventing post-ESD stricture of esophageal early cancer.

Background Post-ESD stricture is becoming a more valued problem in multidisciplinary collaboration protocol.Most research about TA injection is single-center small sample retrospective research. It has a big difference in 3/4 circumferential mucosal defect or more. Hanaoka et al showed that TA injection could decrease the stricture rate to 10% in 3/4 circumferential mucosal defect or more. The control group without any preventive treatment had a 66% stricture rate.The scar healing lives through three stages after esophageal ESD, as follows: the first one is that submucosa or muscularis are exposed to food residuals, acid, bile reflux, and bacteria or fungus reproduction; the second is immune system is activated, mainly including granulation tissue hyperplasia, intraepithelial inflammatory cell infiltrating and neovascularization; the last is scar tissue formationThis phase mainly happens between the third day and the seventh after ESD. So, when is the appropriate time to intervene in suppressing scar formation? Methods ESD procedure Before ESD surgery, all participants signed written consent. Participants were under tracheal intubation anesthesia during ESD procedure. ESD was completed by three senior professors to decrease technology-related complications. During the resection, the investigators used Q260J endoscope (Olympus，Japan)with a water-jet system (Olympus，Japan) and transparent hood (Olympus，Japan). The other devices were as follows, 650-L needle-knife (Olympus，Japan), IT knife (Olympus，Japan), coagulation forceps (FD-410LR, Olympus，Japan), Submucosal injection needle (22Gauge，Olympus，Japan), and endocut machine (EREB，Germany). Standard operation procedures are as follows: (1). After delineating the tumor by 1.5% iodine staining, the investigator used a needle-knife to mark the boundary outside 0.5-1cm of tumor margin；(2). Then the operator blended the submucosal injection with mannitol, sodium hyaluronate, and methylene blue; (3). Needle-knife was used to dissect the elevated submucosa, and the procedure began from the anal side first, then came back oral side and dissected the whole lesion finally；(4) During the dissection, minor bleeding was stopped using needle-knife. But when middle or massive bleeding happened, the operator used coagulation forceps to coagulate the big vessels；(5) The specimen was stained again with iodine to ensure the lesion boundary and sent to the pathology room；(6) A pathology report must contain the size, differentiated grade, depth of invasion，margin and state of lymphatic vessel infiltration.

Locoregional triamcinolone injection and fibrin glue spray TA injection and fibrin glue spray were used in two sessions to prevent esophageal stricture, after ESD immediately(D0) and 7 days later(D7). In the first session, triamcinolone (40mg/1ml/ampul, China) in 100mg total quantity was diluted with 10ml saline to make a 4mg/0.5ml solution and injected in the surface of post-wound symmetrically at 25 points(0.5ml/point) depending on the defective area. But TA injection must avoid muscular injury place. The operator checked the wound to make sure there was no bleeding or perforation after the TA injection, then sprayed 5-10ml fibrin glue (Porcine fibrin sealant kit, China) on the injection wound. The spray dosage was depending on the wound area. The key point was that the glue could cover all the injection wound. Seven days later, the second TA injection and fibrin glue spray were carried out again. Before injection, the operator would check the wound carefully to exclude delayed perforation. A full dose of triamcinolone (100mg/25points) is strictly injected to minimize drug leakage in the process. If excessive leakage happened in the procedure due to underdosage, residual triamcinolone in ampul was used for isodose rescue. After injection, 5ml fibrin glue spray was used to cover the injection points. Because of high price, if the wound healed well, and participants disagreed to use, fibrin glue may not be used in D7.

Follow up and target The Stooler score classification was used to measure the dysphagia level internationally, as follows: 0: no dysphagia symptom, normal diet; I: dysphagia happens sometimes and can swallow soft solid food; II: participant can have a semi-liquid diet, for example porridge, and noodles; III: only a liquid diet such as rice-water/ milk/soup; IV: participant can have nothing even water or saliva.

Esophageal stricture was defined as couldn't get through by Q260 gastroscope (Olympus, Japan, diameter 9.2mm). If the Q260 gastroscope couldn't be used at the review date, another size scope like HQ290 or H290(Olympus, Japan) would be used with the exact diameter marked. When the normal scope couldn't pass, a transnaso-gastroscope (GIF-XP290N, Olympus, Japan, diameter 5.8mm) for children was chosen to check the scar healing condition. So depending on the diameter(D), stricture was classified as follows: Class 0: D≥10mm；Class I：6≤ D <10mm；Class II：D≤5mm.

For the whole circumferential mucosal resection, the first gastroendoscopy check was one month approximately after ESD. If stricture with Class 0, participants came back two months later for the next follow-up. But if participants had dysphagia of stooler II or more, stricture was reassessed by gastroendoscopy to make sure whether dilation needed. For the sub-circumferential mucosal resection, the first follow-up check was evaluated by stooler score classification. If the participant didn't have dysphagia symptoms and had a solid diet (stooler score :0-I), one came back three months later. But if the participants had dysphagia symptoms as score II or more, one needed to come back for stricture assessment to make sure whether dilation is needed by gastroendoscopy examination. Bougie dilation or balloon dilation was the first choice. Investigators recorded the total dilation times. All the participants needed a follow time of no less than three months and endoscopy reexaminations once at least. If the endoscopy was rejected, the Barium Esophagogram (BE) was replaced, and diameter was marked.

Treatment principles Participants who were diagnosed as cT1aN0M0 or cT1bN0M0 before surgery and showed pathological results as pT1bN0M0 with Lymphatic Vessel Infiltration (LVI) or positive margin were defined as non-curative resection (NCR). To prevent the risk of recurrence or metastasis, surgery or chemoradiotherapy (CRT) became good choices, especially CRT for the bad physical condition participants.

Inclusion Criteria:

• staged as cT1aN0M0 or cT1bN0M0 before surgery via magnifying endoscopy, endosonography, CT, or PET/CT;
• refuse conventional surgery and CRT and got ESD as first-line treatment;
• mucosal defect measured by a circular protractor and reached more than three-quarters of the circumference of the esophageal lumen including whole circumference after ESD surgery;
• meet the minimum follow-up period of three months;
• signed the consent before ESD;
• If extra treatment is needed after ESD, it is up to current guidelines.

Exclusion Criteria:

• excessive bleeding and perforation happened during the ESD procedure;
• excessive bleeding and perforation happened during TA injection;
• loss to follow-up or not meet three months follow-up periods;
• allergy to triamcinolone or fibrin glue;
• esophageal recurrence in situ during follow-up period

Beginning 1years and ending 3 years after the publication of results