18F-fluorodeoxyglucose PET/computed tomography metabolic parameters and sarcopenia in pancreatic cancer

Abstract

Objective This study aimed to investigate the association between 18F-fluorodeoxyglucose (18F-FDG) PET/computed tomography (CT) and clinicopathological characteristics and sarcopenia in patients with pancreatic cancer and to determine their prognostic roles. Methods Clinicopathological factors and 18F-FDG PET/CT metabolic parameters of maximum standard uptake value, metabolic tumor volume, and total lesion glycolysis of the primary tumor (SUVmax_P, MTV_P, and TLG_P) and of whole-body lesions (MTV_T and TLG_T) were retrospectively reviewed in 113 pretreatment patients with pancreatic cancer. Sarcopenia was defined based on skeletal muscle index (SMI) calculated at the third lumbar vertebra (L3), and SUVmax of the psoas major muscle (SUVmax_M) was measured at L3 as well. The primary endpoint used was the overall survival (OS). Results Among 113 patients, 49 patients (43.4%) were diagnosed with sarcopenia. Compared with nonsarcopenia, sarcopenia more frequently occurred in the older population (P = 0.027), males (P = 0.014), and lower BMI (P < 0.001), and displayed lower SUVmax_M (P = 0.011). Age, sex, BMI, and SUVmax_M were independently predictive of sarcopenia. Multivariate Cox regression analysis revealed that tumor stage (P = 0.010) and TLG_T (P < 0.001) were independently predictive of OS. Conclusion Sarcopenia increased with declining SUVmax_M in pancreatic cancer. Compared with SMI, SUVmax_M offers a more straightforward prediction of sarcopenia, thus a promising measurement to be incorporated into the diagnostic algorithm. Tumor stage and TLG_T, but not sarcopenia, were independent prognostic factors of pancreatic cancer.