Ultrasound muscle thickness for the bedside assessment of sarcopenia in patients with cirrhosis

Background and aims: Sarcopenia is an independent predictor of mortality and morbidity in patients with chronic liver disease (CLD).1 A major difficulty in evaluating sarcopenia in patients with decompensated CLD is accurately measuring skeletal muscle mass (SMM) due to sodium and fluid retention. Appendicular ultrasound muscle thickness (USMT) has been shown to correlate with total SMM in intensive care patients with fluid overload.2 We aim to assess the validity of this technique in patients with CLD.

Methods: We recruited 28 adult patients with CLD referred to the Queensland Liver Transplant Service. USMT was measured with a linear transducer 12-5MHz. Three muscle thickness measurements were taken at each of the following sites on the dominant side anteriorly: mid forearm, mid upper arm and mid thigh. The primary observer was an experienced sonographer (SK). Ten patients were assessed by a blinded, inexperienced observer (AW) to assess for inter-observer variability. Reference body composition analysis was performed using dual energy X-ray absorptiometry (DXA) appendicular lean mass (ALM), and the computed tomography (CT) skeletal muscle area (SMA) at the level of the 3rd lumbar vertebrae using ImageJ. The reference values were also corrected for height squared to calculate a skeletal muscle index (SMI).

Results: The median age of patients was 54 y (range 18–65) with 22 males. The most common aetiology of liver disease was hepatitis C (54%) and alcohol (14%). The mean Model for End Stage Liver Disease score was 13.4 ± 5 (6–25), and 13 (46%) had hepatocellular carcinoma. Four patients had moderate to severe ascites, and nine patients had mild or diuretic-controlled ascites.

USMT of the forearm correlated moderately with DXA ALM (rho = 0.606; p = 0.001) and with CT SMA (rho = 0.648; p <0.001). USMT of the upper arm correlated strongly with DXA ALM (r = 0.798; p <0.001) and moderately with CT SMA (r = 0.666; p < 0.001). There was a weak correlation of USMT of the thigh with DXA ALM (r = 0.374; p = 0.05) but no correlation with CT SMA (r = 0.323; p = 0.093).

Total USMT moderately correlated with DXA ALM (rho = 0.632; p <0.001) and CT SMA (r = 0.614; p = 0.001). Removing USMT of the thigh improved the correlation with DXA ALM (rho = 0.688; p <0.001) and CT SMA (rho = 0.695; p <0.001). Total USMT corrected for height squared moderately correlated with DXA SMI (r = 0.587; p = 0.001) and weakly with CT SMI (r = 0.490; p = 0.008).

Comparison with an inexperienced observer showed minimal inter-observer variability with an ICC of 0.98 (CI 0.94–0.99).

AUROC evaluation of USMT to diagnose sarcopenia in men was assessed using the DXA SMI cut-off value of 7.26 kg/m2. The AUROC for total USMT was 0.677 (CI 0.428–0.926; p = 0.210), which improved to 0.755 (CI 0.518–0.992; p = 0.071) by removing the thigh USMT. Correcting USMT of the arm for height-squared improved the AUROC to 0.802 (CI 0.568–1.00; p = 0.033*). Insufficient numbers were available to perform an AUROC analysis on the female population.

Conclusions: USMT correlated with DXA and CT reference measures of skeletal muscle mass. In male patients, USMT of the upper arm and forearm corrected for height had an AUROC of 0.802 for diagnosing sarcopenia. USMT can be utilised with good reliability by inexperienced sonographers with minimal training. USMT appears to be a clinically useful measure of muscle mass that could be used in an outpatients setting however further validation studies are required.