OA03-RF01.06. Topographical Classification of Postesophagectomy Anastomotic Leakage Based on CT Findings

Abstract

Abstract Topic Esophageal Cancer: Surgical Treatment of Esophageal Cancer – early outcomes and complications Background Anastomotic leakage is a severe complication after esophagectomy. Postoperative imaging is essential for detecting it, with CT scans using both intravenous and oral contrast as one of the main diagnostic tool combined with gastroscopy. The aim of the study is to define a radiological classification of post esophagectomy leakages and how it correlates with outcome measures. Methods A retrospective study was conducted in a tertiary referral center for esophageal diseases. Data have been collected from 2015 to 2024. A total of 85 patients who developed anastomotic leaks after esophagectomy with intrathoracic anastomosis and underwent CT scan were included. Demographics, preoperative, intraoperative, postoperative variables as well as radiological characteristics of the fistula and postoperative outcomes were evalutated. In particular, two committed radiologists and surgeons reassessed CT scans in order to classify the fistula site as follows: right anterior (RA), left anterior (LA), right posterior (RP) and left posterior (LP). Jamovi software was used to perform descriptive, bivariate and multivariate analysis. Results Left anterior fistulas were most common (40%), followed by right anterior (20%), left posterior (14.1%), and right posterior (11.8%). 72% of leaks occurred by day 8, with a peak on days 7–8. On bivariate analysis, the LA fistula was most strongly associated with perianastomotic collections and air within the collection (p < 0.001). On multivariate analysis, all fistula locations were associated with collection development; the strongest association was observed for RP, followed by LA, LP and RA fistulas (p < 0.001). Air in the collection were strongly associate in order with fistula RA, LA, LP and RP (p<0.001). LA fistulas were most strongly associated with respiratory and circulatory complications and, on multivariate analysis, with severe outcomes (Clavien–Dindo IVb–V; p < 0.001). We assessed the position of the gastroepiploic arcade relative to the fistula: it was ipsilateral in 38.9% of cases and contralateral in 61.1% (p = 0.045). Conclusion These findings suggest that fistula topography may represent a relevant prognostic factor and could support early risk stratification and eventually tailored postoperative management. Although, this is a preliminary analysis on small sample of patients, the results may encourage future studies with larger patient populations.