Visceral fat area, not body mass index, predicts postoperative 30-day morbidity in patients undergoing colon resection for cancer.

PURPOSE: Colectomy for cancer in obese patients is technically challenging and may be associated with worse outcomes. Whether visceral obesity, as measured on computed tomography, is a better predictor of complication than body mass index (BMI) or determines long-term oncologic outcomes has not been well characterized. This study examines the association between derived anthropometrics and postoperative complication and long-term oncologic outcomes.
METHODS: Retrospective review of patients undergoing elective colectomy for cancer at a single tertiary-care center from 2010 to 2016. Adipose tissue distribution measurements, including visceral fat area (VFA), were determined from preoperative imaging. The primary outcome was 30-day postoperative complication; secondary outcomes included overall and disease-free survival. Multivariable logistic regression was performed to determine association between obesity metrics and outcome.
RESULTS: Two hundred and sixty-four patients underwent 266 primary resections of colon cancer. Twenty-eight patients (10.5%) developed major morbidity (Clavien-Dindo grade ≥ III). VFA but not BMI was significantly associated with morbidity in multivariate analysis (p = 0.004, odds ratio 1.99, 95% confidence interval 1.25-3.19). No other imaging-derived anthropometric was associated with increased morbidity. In receiver operating characteristic analysis, VFA was predictive of major morbidity (area under curve 0.660). A cutoff value of VFA ≥ 191 cm2 was associated with 50% sensitivity and 76% specificity for predicting major morbidity. Patients with VFA ≥ 191cm2 had 19.4% risk of morbidity, whereas those with < 191 cm2 had 7.2% risk (relative risk ratio 2.69, unadjusted p = 0.004). Neither VFA nor BMI was associated with overall or disease-free survival.
CONCLUSION: VFA but not BMI predicts morbidity following elective surgery for colon cancer.