290. ANALYTIC MORPHOMIC INDEX AND QUANTITATIVE GASTRIC PERFUSION ASSESSMENT FOR RISK STRATIFICATION OF ANASTOMOTIC LEAKS FOLLOWING ESOPHAGECTOMY

Abstract

Abstract Sarcopenia, a reduction in muscle mass, has been considered a preoperative risk factor for complications after esophagectomy while impaired perfusion of the gastric conduit has been associated with anastomotic leaks. The aim of this study was to stratify patients by their risk of anastomotic leak following esophagectomy using morphomic factors and quantitative assessment of gastric conduit perfusion. Patients who underwent an esophagectomy with a cervical anastomosis and gastric conduit from 2015–2021 were included. Preoperative CT scans were processed using semiautomated algorithms (MATLAB; MathWorks, MA). Indocyanine green perfusion was assessed intraoperatively (SPY elite; Stryker, MI). Sarcopenia was defined using skeletal muscle area (Martin’s criteria). Decreased subcutaneous fat, decreased bone mineral density, and impaired conduit perfusion were defined as below the cutoff value from the ROC curve. The morphomic index was determined using the number of decreased morphomic factors present (0–3). The Fisher exact test, the Mann–Whitney U test, and multivariable logistic regression were used to make statistical comparisons. 140 patients with both perfusion and morphomic data were evaluated (Leak n = 27; No leak n = 113). Decreased subcutaneous fat, decreased bone mineral density, and impaired gastric conduit perfusion were associated with an anastomotic leak (p = 0.041, 0.017, and 0.010). Neoadjuvant chemoradiation was significantly more common in the leak group (p = 0.048). On multivariable analysis, a higher morphomic Index and impaired gastric perfusion were risk factors for anastomotic leak. The incidence of anastomotic leaks with a high morphomic index (2 or 3) and impaired gastric perfusion was 40.6% compared to 9.4% with a low morphomic index and without impaired gastric perfusion (p = 0.006). Impaired gastric conduit perfusion and decreased preoperative subcutaneous fat area, bone mineral density, and skeletal muscle area, likely surrogates for increased frailty and poor nutritional status, were associated with an increased risk of anastomotic leaks especially when multiple morphomic factors were simultaneously decreased. These findings may be useful in risk stratification and intraoperative decision making in patients undergoing esophagectomy.