Background: Body mass index (BMI) has proven unreliable in predicting survival following pancreaticoduodenectomy for cancer. While measures of intra-abdominal fat correlate with medical and postoperative complications of obesity, the impact of intra-abdominal fat on pancreatic cancer survival is uncertain. We hypothesized that the quantity of intra-abdominal fat would predict survival following resection of pancreatic cancer.
Methods: Preoperative CT imaging was used to measure intra-abdominal fat. Cox regression analyses were used to identify independent predictors of survival.
Results: Sixty-one patients from 2000-2009 underwent pancreaticoduodenectomy for exocrine pancreatic adenocarcinoma. After adjusting for age and perineural invasion status, preoperative BMI did not predict overall survival (p < 0.827). Unlike BMI, quartile of intra-abdominal fat predicted survival. Relative to patients with the least intra-abdominal fat (lowest quartile), those with more intra-abdominal fat demonstrated worse overall survival, but in a non-linear fashion. Individuals in the second quartile showed a fourfold increase in likelihood of death (HR 4.018, 95% CI 1.099-14.687, p < 0.035) relative to the lowest quartile. Patients in the third (HR 2.124, 95% CI 0.278-16.222, p < 0.468) and fourth quartile (HR 1.354, 95% CI 0.296-6.190, p < 0.696) also showed greater risk of death.
Conclusions: Measuring intra-abdominal fat identifies a subset of patients with worse prognosis in pancreatic cancer.