• Guillame Ulmann, Joanna Kaï, Jean-Philippe Durand, Nathalie Neveux, Anne Jouinot, Jean-Pascal De Bandt, Francois Goldwasser, and Luc Cynober.
• Clinical Chemistry Department, Hôpital Cochin, AP-HP Centre, Université de Paris, Paris, France; EA 4466 PRETRAM, Faculty of Pharmacy, Université de Paris, Paris, France. Electronic address: guillaume.ulmann@aphp.fr.
• Nutrition. 2021 Jan 1; 81: 110895.

ObjectivesLean body mass (LBM) is an important prognostic factor in patients with cancer. Although the L3-computed tomography (CT) scan is considered a reference method for assessment, a convenient and easily available method for longitudinal follow-up is required. Although bioelectrical impedance analysis (BIA) is widely used, its accuracy is questioned; plasma creatinine-to-cystatin C (CC) ratio could be an attractive alternative. The aim of this study was to evaluate the ability of the CC ratio and BIA to detect myopenia in patients with cancer compared with the use of the CT scan as a standard.MethodsPatients with any kind of cancer had body composition evaluation by CT scan, BIA, and CC. Statistical analysis included correlation test, Bland-Altman, and receiver operating characteristic curve analysis.ResultsForty-four patients (14 women) were included. Of the participants, 59% had myopenia on CT scan. Both BIA LBM and CC ratio were well correlated with CT scan LBM (r = 0.763 and 0.648, respectively) but concordance analysis revealed a 3-kg constant bias toward BIA compared with CT scan. In terms of ability to detect myopenia, areas under the curve (AUC) for BIA were 0.675 and 0.388 for men and women, respectively. For CC ratio, AUCs were 0.813 and 0.673.ConclusionThis study demonstrated that LBM assessed by the CC ratio or BIA is well correlated with that determined by L3-CT scan. The ability of the CC ratio to detect myopenia was better than that of BIA. Findings from the present study demonstrated that CC ratio can be conveniently used in patients with cancer as a reliable biomarker of muscularity.Copyright © 2020 Elsevier Inc. All rights reserved.

19 keywords...

hide…