225. PREDICTORS OF ANASTOMOTIC LEAK AMONG PATIENTS UNDERGOING ESOPHAGECTOMY: A RETROSPECTIVE COHORT OF 6669 PATIENTS
Uzair Jogiat, Kevin Verhoeff, Valentin Mocanu, Simon Turner, Eric Bédard- Gastroenterology
- General Medicine
Abstract
Background
Despite advances in operative techniques, anastomotic leak (AL) after esophagectomy remains a dreaded complication associated with a relatively high morbidity and mortality. Identification of patient and surgical factors associated with AL are useful for risk stratification to improve patient and procedure selection. Furthermore, an adequately powered analysis for the incidence of AL after minimally invasive esophagectomy is lacking.
Methods
The ACS-NSQIP database was used to extract patients who underwent esophagectomy between 2016 and 2021. Categorical variables were analyzed using chi-squared tests and continuous variables were evaluated using ANOVA tests. Using a hypothesis driven purposeful selection methodology, a multivariable logistic regression model was created with the primary outcome of interest being the odds of AL. Statistical analysis was completed with StataSE 17 (STATACorp, LP, College Station, TX). The primary outcome of interest was the effect of minimally invasive surgery on AL. Secondary outcomes were the effect of patient characteristics and surgeon specialty on AL.
Results
Of the 6669 esophagectomies, the leak rates were 14.6% (open; 360/2451), 13.4% (MIS-abdominal; 64/478), 21.1% (MIS-chest; 69/326), 14.3% (MIS-total; 369/2590), and 20.1% (Robotic; 166/824), respectively. On multivariate analysis, positive margin (OR 1.37, 95% CI 1.01–1.85, p = 0.043), MIS-chest (OR 1.40, 95% CI 1.01–1.96, p = 0.047), MIS-total (OR 0.95, 95% CI 0.78–1.15, p = 0.595), MIS-robotic (OR 1.37, 95% CI 1.07–1.77, p = 0.014), transhiatal esophagectomy (OR 1.28, 95% CI 1.01–1.62, p = 0.040), and three-hole esophagectomy (OR 1.38, 95% CI 1.12–1.77, p = 0.014) were surgical factors associated with increased odds of AL (Table 1).
Conclusion
Anastomoses that were performed by certain types of minimally invasive surgery and those performed in the cervical region were associated with increased odds of AL. Consideration of this increased risk needs to be weighed against the established benefits of minimally invasive surgery when selecting the appropriate procedure for each patient. Identifying patients who are high-risk for an AL pre-operatively may prevent delay in the diagnosis and treatment of ALs in this patient population.