151751-Najiba-Chargi

33 Diagnostics: validation of skeletal muscle mass measurement INTRODUCTION Over the last decade, research into the specific body composition of cancer patients and its relationship with clinical outcomes has tremendously increased due to the use of diagnosti- cally performed imaging for quantification of different body compartments, including skeletal muscle mass (SMM) and adipose tissue mass. 1,2 Specifically a state of low SMM, sometimes termed sarcopenia, has gained interest as a novel risk factor for negative short- and long- termoutcomes. In breast, gastro-intestinal, hepato-pancreatic-biliary and respiratory cancer, amongst others, low SMM is associated with increased incidence of postoperative complica- tions, chemotherapy-related toxicity, prolonged hospital stay and shorter disease-free and overall survival. 3,4 SMM is most commonly assessed on a single CT slice at the level of the third lumbar vertebra (L3), which has shown to have excellent correlation with whole body skeletal muscle volumes as measured using whole body MRI. 5,6 The cross-sectional skeletal muscle area (SMA) at the level of L3 is then most commonly normalized for stature, to calculate the lumbar skeletal muscle index (lumbar SMI). 5 The lumbar SMI is used as a proxy for SMM as a whole, and several cut-offs have been published to identify patients with low SMM. 4 In head and neck cancer (HNC), abdominal CT imaging is not commonly performed as part of the routine diagnostic work-up. Therefore, abdominal CT imaging to quantify SMM is not routinely applicable in HNC patients. To overcome this, a measurement method for SMM at the level of the third cervical vertebra (C3), which is featured on standard CT imaging of the head and neck area, was published by Swartz et al. 7 A multivariate formula to calculate SMA at the level of L3 from SMA at the level of C3 was also published, to allow for comparison to other oncological research. 7 Wendrich et al. published a cut-off value for low SMM in HNC patients based on this method. 8 The measurement method for SMM at the level of C3 was used in several studies in HNC patients. The incidence of low SMMwas high in several studies; typically 50% of patients and sometimes up to 77% of patients had low SMM prior to start of treatment. 8–11 In HNC patients, low SMM was associated with negative short- and long-term outcome such as chemotherapy dose-limiting toxicity, postoperative complications and de- creased survival. 8,9,12,13 Only one previous study by Ufuk et al. has investigated the correlation between SMA measurement at the level of C3 and L3. They showed that SMA at the level of C3 was best associated with SMA at the level of L3, and that the correlation between SMA at the level of C3 and SMA at the level of L3 was excellent. 14 Ufuk et al. segmented the sterno - cleidomastoid(SCM) and paravertebral muscles (PVM) separately, Swartz et al. recommends using the SMA at C3 of both the SCM and PVM. Ufuk et al. also used cut-off values for low SMM based on the study of Prado et al. which did not include HNC patients and did not validate the formula proposed by Swartz et al. Our current study aimed to reevaluate the association between SMA at the level of C3 and the level of L3 in a larger cohort of treatment naïve HNC patients. It also aimed to investigate the accuracy of identifying patients with low SMM using a previously published cut-off value. 2