Development and validation of equations for predicting appendicular skeletal muscle mass in male patients with head and neck cancer and normal hydration status

Objective: Muscle mass is typically assessed by abdominal computed tomography, magnetic resonance imaging, and dual-energy X-ray absorptiometry. However, these tests are not routinely performed in patients with head and neck cancer (HNC), making sarcopenia assessment difficult. This study aimed to develop and validate equations for predicting appendicular skeletal muscle (ASM) from data obtained in daily medical practice, with bioelectrical impedance analysis (BIA)-measured ASM (BIA-ASM) as a reference.
Research Methods & Procedures: This cross-sectional study included 103 male patients with HNC and randomly divided them into development and validation groups. The prediction equations for BIA-ASM were developed by multiple regression analysis and validated by Bland–Altman analyses. The estimated skeletal muscle mass index (eSMI) was also statistically evaluated to discriminate the cutoff value for BIA-measured SMI according to Asian Working Groups for Sarcopenia.
Results: Two practical equations, which include 24-hour urinary creatinine excretion volume (24hUCrV), handgrip strength (HGS), body weight (BW), and body height (BHt), were developed: ASM (kg) = −39.46 + (3.557 × 24hUCrV[g]) + (0.08872 × HGS[kg]) + (0.1263 × BW[kg]) + (0.2661 × BHt[cm]) if available for 24hUCrV (adjusted R2 = 0.8905), and ASM (kg) = −42.60 + (0.1643 × HGS[kg]) + (0.1589 × BW[kg]) + (0.2807 × BHt[cm]) if not (adjusted R2 = 0.8589). ASM estimated by these two equations showed a significantly strong correlation with BIA-ASM (R > 0.900). Bland–Altman analyses showed a good agreement, and eSMI accuracy was high (>80%) in both equations.
Conclusions: These two equations are a valid option for estimating ASM and diagnosing sarcopenia in patients with HNC in all facilities without special equipment.

論文本文は2024-08-06以降公開予定

© 2023. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/