RÉSUMÉ
BACKGROUND: Skeletal muscle indices have been associated with improved peri-operative outcomes after surgical resection of non-small-cell lung cancer (NSCLC). However, it is unclear if these indices can predict long term cancer specific outcomes. METHODS: NSCLC patients undergoing lobectomy at our institute between 2009-2015 were included in this analysis (N = 492). Preoperative CT scans were used to quantify skeletal muscle index (SMI) at L4 using sliceOmatic software. Cox proportional modelling was performed for overall (OS) and recurrence free survival (RFS). RESULTS: For all patients, median SMI was 45.7 cm2/m2 (IQR, 40-53.8). SMI was negatively associated with age (R = -0.2; p < 0.05) and positively associated with BMI (R = 0.46; P < 0.05). No association with either OS or RFS was seen with univariate cox modelling. However, multivariable modelling for SMI with patient age, gender, race, smoking status, DLCO and FEV1 (% predicted), American Society of Anesthesiology (ASA) score, tumor histology and stage, and postoperative neoadjuvant therapy showed improved OS (HR = 0.97; P = 0.0005) and RFS (HR = 0.97; P = 0.01) with SMI. Using sex specific median SMI as cutoff, a lower SMI was associated with poor OS (HR = 1.65, P = 0.001) and RFS (HR = 1.47, P = 0.03). CONCLUSIONS: SMI is associated with improved outcomes after resection of NSCLC. Further studies are needed to understand the biological basis of this observation. This study provides additional rationale for designing and implementation of rehabilitation trials after surgical resection, to gain durable oncologic benefit.