151751-Najiba-Chargi

32 CHAPTER 2 ABSTRACT Background Low skeletal muscle mass (SMM) is associated with adverse outcomes. SMM is often assessed at the third lumbar vertebra (L3) on abdominal imaging. Abdominal imaging is not routinely performed in patients with head and neck cancer (HNC). We aim to validate SMMmeasurement at the level of the third cervical vertebra (C3) on head and neck imaging. Material and methods Patients with pre-treatment whole-body computed tomography (CT) between 2010 and 2018 were included. Skeletal muscle mass area (SMA) was manually delineated at the level of C3 and L3. Correlation coefficients and intraclass correlation coefficients (ICCs) were calculated. Cohen’s kappa was used to assess the reliability of identifying a patient with low SMM. Results Two hundred patients were included. Correlation between SMA at the level of C3 and L3 was good ( r = 0.75, p<0.01). When using a multivariate formula to estimate SMA at L3, including gender, age, and weight, correlation improved ( r = 0.82, p<0.01). The agreement between es - timated and actual SMA at L3 was good (ICC 0.78, p<0.01). There was moderate agreement in the identification of patients with low SMM based on the estimated lumbar skeletal muscle mass index (LSMI) and actual LSMI (Cohen’s κ : 0.57, 95%CI 0.45-0.69). Conclusions SMA at C3 correlates well with SMA at L3. There is moderate agreement in the identification of patients with low SMM based on the estimated LSMI (based on measurement at C3) and actual LSMI.