Abstract
1773
Objectives To assess the feasibility of a multiparametric approach of using routine 18F-FDG PET/CT conducted for the initial staging of soft-tissue sarcoma patients to derive volumetric imaging metrics of skeletal muscle quality and quantity.
Methods Whole-body 18F-FDG-PET/CT images of human patients (N=14, 8F 6M, age 62±13.8 yrs) with established soft-tissue sarcoma prior to receiving treatment were retrospectively analyzed. Skeletal muscle was manually segmented in 3D from CT (mean 104 axial slices) between two predetermined anatomical landmarks (the T12 pedicles and the ischial tuberosities) to create a mask. A range of metrics of skeletal muscle quality and quantity was computed from CT: (1) mean skeletal muscle density (SMD) in Hounsfield units (HU), (2) skeletal muscle volume (SMV), (3) skeletal muscle volume percent (SMVP) as the ratio of SMV to that of the entire body between the anatomical landmarks, and (4) normalized intramuscular fat volume (NIFV) as the ratio of fat volume in skeletal muscle (based on the fat HU window) to the total SMV. The masks were transferred to co-registered PET images and the following muscle measures were derived: (1) SUVmax and SUVmean (conventional and sex-specific lean body) for the segmented skeletal muscle, and (2) metabolic skeletal muscle volume (MSMV), as the volume of skeletal muscle with an SUV>30% of SUVmax.
Results Descriptive statistics of the imaging metrics are in Table 1. Preliminary analysis showed a moderate correlation of SUV measures with serum albumin and strong negative correlation of SMD with body mass index (BMI). SMV did not correlate with BMI.
Conclusions This pilot study demonstrates the feasibility of deriving imaging metrics associated with skeletal muscle quality and quantity. Our future work will evaluate the predictive power of these imaging metrics before the start of therapy for patient outcomes such as survival and major post-operative complications.
Research Support This work was funded by the NIH grants R03EB015099 and K12HD051958