Adipose tissue imaging as nutritional predictors in patients undergoing enterocutaneous fistula repair

Objectives: The management of enterocutaneous fistulae (ECF) is complex, challenging, and often associated with metabolic, septic, and nutritional complications. Radiographic quantification of body composition such as fat or lean body mass distribution is a potentially valuable preoperative assessment tool to optimize nutritional status. The aim of this study was to investigate the correlation between total adipose tissue (fat) area (TFA), assessed by computed tomography and magnetic resonance imaging radiology tests, with body weight, body mass index (BMI), various biochemical parameters, need for nutritional support, and survival in patients undergoing ECF repair. Methods: Biochemical and anthropometric parameters at the time of ECF surgery were retrospectively collected for adult patients undergoing ECF repair at University College London Hospital, UK. Visceral and subcutaneous adiposity was measured at the level of the third lumbar vertebra (Image J) at computed tomography or magnetic resonance imaging. Statistical analysis included descriptives, univariate and multivariate analysis between TFA and various parameters, and their influence on postoperative survival. Results: A complete set of data was available for 85 patients (51 women, 56.9 +/- 14.5 y of age) who underwent ECF repair. ECF originated mainly as a surgical complication (86%) while 14% were undergoing a second ECF repair. Median BMI was 22.8 kg/m(2) and mean TFA was 361 +/- 174.9 cm(2), with a higher visceral fat content in men than in women (183.8 +/- 99.2 versus 99 +/- 59.7 cm(2), P < 0.001). BMI, body weight, and creatinine were significantly positively correlated with TFA (rho = 0.77, 0.73, and 0.50, respectively, P < 0.001); no correlation was noted between TFA and preoperative albumin levels. Patients in the low TFA group had a higher use of parenteral nutrition (P = 0.049). Hospital length of stay was longer in patients receiving artificial nutrition support (70 versus 22 d, P < 0.001). A TFA cutoff point of 290 cm(2) discriminated patients who required artificial nutrition versus no nutritional support with moderate sensitivity (75%) but poor specificity (45%). At multivariate analysis, only >60 y of age (hazard ratio [HR], 2.69, P < 0.02) and use of parenteral nutrition (HR, 3.90, P < 0.02) were associated with worse overall survival. Conclusion: Abdominal adiposity was strongly correlated with anthropometric parameters at the time of surgery. Earlier identification of patients requiring artificial nutrition at standard preoperative imaging might allow integration of nutritional optimization into initial clinical management plans reducing length of stay and improving clinical outcomes. (C) 2020 Elsevier Inc. All rights reserved.