Computed tomography compared with standard clinical measurements to assess body composition, facilitating the identification of sarcopenia and cachexia in colorectal malignancy

Background Malnutrition, sarcopenia and cachexia have been shown to adversely affect clinical outcome in patients with colorectal cancer (CRC). These conditions can be identified by assessing body composition, specifically skeletal muscle, in combination with measurements of physical function (PF), weight loss and inflammation. Computed tomography (CT) provides a novel method for assessing body composition. This study aimed to determine the role of CT scans in assessing reduced skeletal muscle and use of CT as a criterion measure in comparison to other clinical assessment techniques. Methods CT scans were obtained for people with CRC. Body composition was analysed from CT images (Slice-O-Matic software). This was compared to body composition measured by: bioelectrical impedance analysis (BIA), mid arm muscle circumference (MAMC); patient generated subjective global assessment (PG-SGA); and 4-site skinfolds. Handgrip strength and short form 36 (SF36) were used to measure PF. C-reactive protein (CRP) was measured to assess inflammation. Results CT scans were obtained for 100 people. CT scans identified low skeletal muscle in 29% of participants. BIA and MAMC showed good levels of agreement (Area under the curve [AUC] = 0.619, p < 0.001 and AUC = 0.625, p<0.005 respectively). Using handgrip as PF, CT scans identified 14.1% and 5.2% of participants as having sarcopenia and cachexia respectively. BIA showed good levels of agreement (AUC = 0.738, p<0.001 and AUC = 0.723, p<0.001 respectively). Conclusions These data highlight the importance of correct classification of low muscle mass. BIA and MAMC are valid to assess muscle mass, compared to CT as a criterion measure. Handgrip strength rather than SF-36 provides a good assessment of PF in relation to identifying reduced muscle strength in sarcopenia. Clinical data from this study will be useful in altering practise for correct identification of sarcopenia and cachexia, particularly in the absence of CT scans.