Thoracic skeletal muscle index is effective for CT-defined sarcopenia evaluation in patients with head and neck cancer

Abstract

Abstract Purpose Computed tomography (CT)-defined sarcopenia, as a measurement of low skeletal muscle (SM), is a poor prognostic indicator in patients with head and neck cancer (HNC), independent of weight or nutritional status. We used SM measures at the second thoracic vertebra (T2) to determine T2-SM index (SMI) thresholds for sarcopenia, and investigate the impact of low T2-SMI on overall survival (OS), and weight loss during radiotherapy (RT). Methods Adult patients with newly diagnosed HNC with a diagnostic PET–CT or RT planning CT scan were included. SM was analysed at T2 and a model applied to predict SM at L3. T2-SMI thresholds for sarcopenia were established with predicted measures, stratified by BMI and sex. Impact of sarcopenia and low T2-SMI on OS and weight loss during RT was investigated. Results A total of 361 scans were analysed (84% males, 54% oropharynx tumours). Sarcopenia was found in 49%, demonstrating worse OS (p = 0.037). T2-SMI cutoff values were: females—74 cm2/m2 [area under the curve (AUC): 0.89 (95%CI 0.80–0.98)], males (BMI < 25)—63 cm2/m2 [AUC 0.93 (95%CI 0.89–0.96)], males (BMI ≥ 25)—88cm2/m2 [AUC 0.86 (95%CI 0.78–0.93)]. No difference in OS with T2-SMI categories. Lowest T2-SMI quartile of < 63 cm2/m2 demonstrated worse OS (p = 0.017). Weight loss during RT was higher in patients; who were not sarcopenic (6.2% vs 4.9%, p = 0.023); with higher T2-SMI (6.3% vs 4.9%, p = 0.014) and; in the highest quartiles (3.6% vs 5.7% vs 7.2%, p < 0.001). Conclusions These T2-SMI thresholds are effective in assessing CT-defined sarcopenia in HNC. Further assessment of clinical application is warranted.