P1.17-34 Prognostic Significance of Preoperative CT-Determined Sarcopenia and Radiodensity in Resected Non-Small Cell Lung Cancer

Sarcopenia, reduced skeletal muscle mass, is associated with poor prognosis in advanced non-small cell lung cancer (NSCLC) patients. The purpose of this study was to evaluate predictive value of CT–determined skeletal muscle area on prognosis after curative resection of NSCLC. For a total 272 NSCLC patients (male=164, mean age=63±10 years) that underwent surgery between 2011 and 2016, skeletal muscle (CT Hounsfield unit: −29 to 150 HU) at the level of the third lumbar vertebra (L3) was assessed using preoperative CT images. Sarcopenia was defined as L3 muscle index of <55 cm2/m2 for men and of <39 cm2/m2 for women. Moreover, skeletal muscle was subsequently classified as abnormal muscle (-29 to <30HU) and normal muscle (30 to 150 HU) by radiodensity. Data on clinical characteristics, disease-free survival (DFS), and overall survival (OS) were retrospectively collected. The prevalence of sarcopenia was 22.4% for all study subjects, 32.9% for men, and 6.5% for women. Using a maximal chi-squared method determined optimal cut-off to determine unhealthy muscle (proportion of abnormal muscle >24.5%), the prevalence of patient with unhealthy muscle was 84.2% (78.7% for male and 92.6% for female). There was no significant difference in the prevalence of unhealthy muscle between sarcopenia and non-sarcopenia patients (86.9% vs. 83.4%, p = 0.690). No significant difference was observed between the 3-year DFS rate (77.0% vs 67.0%, p = 0.142) or 3-year OS rate (84.8% vs 87.9%, p = 0.576) between patients with and without sarcopenia. However, patients with unhealthy muscle tends to have shorter 3-year DFS rate (67.2% vs 79.4%, p = 0.094) and 3-year OS rate (86.1% vs 92.6%, p = 0.176). In the multivariable analysis, unhealthy muscle was one of independent indicator to predict a 3-year DFS rate (HR, 2.072, p=0.044), along with pathologic T stage (HR, 3.348, p<0.001), pathologic N stage (HR, 3.401, p<0.001), and R1 resection (HR, 3.804, p=0.033). Radiodensity based muscle quantification is associated with shorter DFS in resected lung cancer. Further research is warranted to establish whether muscle measures should be integrated into routine practice to improve prognostic accuracy.