Background: Body composition measures may predict outcomes of cancer surgery. Whereas low muscle mass shown on preoperative computed tomography (CT) scans has been associated with worse outcomes after surgery for pancreatic cancer, less consideration has been given to low muscle attenuation, reflecting poor muscle quality. Studies relating muscle mass and muscle attenuation with outcomes for patients with periampullary, nonpancreatic cancer are lacking.
Methods: Skeletal muscle mass and attenuation were assessed in 166 consecutive patients undergoing pancreatoduodenectomy (PD) for periampullary, nonpancreatic cancer at a single center between 2000 and 2012. The skeletal muscle index (SMI) was calculated from cross-sectional muscle area on preoperative CT imaging at the third lumbar vertebra level (L3) and normalized for height. The skeletal muscle attenuation index (MAI) was calculated by measuring the average Hounsfield units of the total muscle area at the L3 level. Overall survival (OS) and the rate of major postoperative complications (Clavien-Dindo ≥3) were extracted from prospectively maintained databases.
Results: Low SMI was present in 78.3 % and low MAI in 48.8 % of the patients. The multivariate analysis showed lymph node metastasis [hazard ratio (HR) 1.8; 95 % confidence interval (CI) 1.1-2.9], microscopic radicality (HR 2.0; 95 % CI 1.2-3.4), and low MAI (HR 2.0; 95 % CI 1.2-3.3), but not low SMI to be significantly associated with decreased OS. Low MAI (HR 1.9; 95 % CI 1.0-3.8) was the only independent risk factor for major postoperative complications.
Conclusion: Skeletal muscle quality, but not muscle mass, predicted survival and major complications after PD for periampullary, nonpancreatic cancer. Preoperative CT-derived body composition measures may stratify patients into risk categories and support shared decision making.