Assessment of sarcopenia with ultrasound-based measurements in patients with liver cirrhosis and correlation with clinical outcomes

Abstract

Introduction

Sarcopenia is a common complication in patients with liver cirrhosis. In this context, its diagnosis is typically based on operational definitions, including the estimation of low muscle mass. Recently, muscle ultrasound-based measurements have drawn attention due to their improved feasibility and accessibility. However, only a limited number of studies evaluating this approach have been reported. Finally, the role of muscle strength respect to mass in identifying patients with the worst clinical outcomes has not been clearly elucidated.

Aim

In a cohort of cirrhotic patients, our primary aim was to investigate the correlation and agreement between ultrasound-derived measures of muscle mass and bioimpedance analysis (BIA) as the gold standard, as well as their discriminative power. In addition, as a secondary aim, we investigated the correlation of these techniques and muscle strength with clinical outcomes.

Materials and Methods

The study included consecutive adult outpatients attending the Hepatology Unit of the Fondazione Policlinico Campus Bio-Medico of Rome. Muscle mass was defined as appendicular skeletal mass (ASMM) according to the Sergi equation (EWGSOP 2019). Ultrasound was performed to measure muscle mass according to previously described standardized indices (quadriceps and iliopsoas muscles). Hand grip measurement was used to define muscle strength. Pearson's correlation coefficient and Bland-Altman plots were used to assess the correlation and agreement between ASMM and ultrasound indices. Predictive performance was estimated by calculating the area under the receiver operating characteristic curve (AUROC). Finally, crude and adjusted Cox regression analyses were performed to test the possible association between the different proxies of sarcopenia and liver decompensation or mortality within 24 months.

Results

88 patients were included [(mean age 73 years (7.07), 78% male, mean BMI 27 kg/m2 (10)]. The most common aetiology of cirrhosis was viral (40%) and the majority of patients (80%) had well preserved liver function. Average compression index (ACI) and average feather index (AFI) showed a good correlation with ASMM, while among the psoas indices, only psoas to height ratio (PHR)- but not ileopsoas index (IPI) - showed a correlation (Figure 1). Linear regression analysis confirmed that AFI [beta 0.64 (CI95% 0.37-0.92), p<0.001], ACI [0.5 (CI95% 0.21-0.78), p<0.001] and PHR [0.38 (CI95% 0.08-0.69), p=0.01] were significantly associated with ASMM, also independently of gender. In addition, Bland-Altman analyses showed good agreement for US with ASMM. Furthermore, these indices showed adequate discriminatory power, with AUROCs of 0.71 (0.57-0.854), 0.81 (0.69-0.931) and 0.75 (0.63-0.862) for ACI, AFI and PHR, respectively. Finally, in Cox regression analyses, only low muscle strength was associated with higher rates of mortality and liver decompensation [HR 1.62(1.06-2.47), p 0.026; HR 1.29(0.99-1.69), p 0.064].

Conclusions

Ultrasound measurements of quadriceps and psoas diameters showed good correlation and agreement with muscle mass defined by BIA in cirrhotic patients, displaying also an adequate discriminatory ability. At the same time, only low muscle strength exhibited a valuable predictive role for outcomes in our population. If these results are confirmed in larger external series, ultrasound can be proposed as a feasible and cost-effective tool for the assessment of muscle mass in patients with CLD, while dynopenia should be used to better identify patients with the worst outcomes.