A multicenter retrospective study of the esophageal triamcinolone acetonide-filling method in patients with extensive esophageal endoscopic submucosal dissection.

OBJECTIVE: Endoscopic submucosal dissection (ESD) for extensive esophageal cancer is sometimes associated with post-ESD stenosis, despite preventative steroid therapy. In this retrospective multicenter analysis, we evaluated the factors associated with therapy resistance. MATERIALS AND METHODS: We enrolled 73 patients with 75 extensive esophageal cancers treated with ESD. Stenosis prevention was performed using two esophageal triamcinolone acetonide (TA)-fillings, and repeated if stenosis was found on follow-up. Therapy-resistance factors associated with incidence of severe stenosis requiring endoscopic balloon dilation (EBD) were evaluated, including age, gender, previous treatment history, tumor location, morphology, resection size, histologic type, invasion depth, and horizontal resection grade (HR-grade 1, ≥ 9/12 and <10/12 of the circumference; grade 2, ≥ 10/12 and <11/12; grade 3, ≥ 11/12 but not circumferential; and grade 4, entirely circumferential). RESULTS: Severe stenosis occurred in 17.3%(13/75) of cases, with a median of two EBDs (range, 1-6 times). Severe stenosis was significantly associated with HR-grade elevation and previous treatment history (p < .05); multivariate analysis showed both as independent therapy-resistance factors (p < .05). Patients without previous treatment history demonstrated severe stenosis at 12.9%(9/70): 0%(0/26) HR-grade 1, 18.8%(3/16) grade 2, 17.6%(3/17) grade 3, and 27.3%(3/11) grade 4, showing a risk of HR-grade 2 or more resection but an acceptable stenosis prevention even after entirely circumferential resection. Conversely, patients with previous treatment history demonstrated severe stenosis at a high frequency of 80%(4/5). CONCLUSIONS: Esophageal TA-filling is a promising stenosis-preventive steroid therapy, even in entirely circumferential ESD cases. However, HR-grade 2 or more elevation and previous treatment history were independently associated with therapy resistance.