Sampling variability of transient elastography according to probe location

Abstract

Heterogeneity of fibrosis throughout the liver has been reported. However, the need for several measurements when using transient elastography was not thoroughly investigated. The aim was to find out whether measurement of liver stiffness varies according to the probe location. Six hundred and twenty-five consecutive patients with chronic liver diseases referred for transient elastography were enrolled. All patients underwent successive liver stiffness measurements at three different sites. Representative measurements were acquisitions with a success rate greater than 60% and an interquartile range lower than 30% of the median. The sample included 371 eligible patients with three representative measurements. Comparing the three successive measurements categorized to fibrosis stages F0-F4, 68.2% of patients had agreement between all three sites. Discordance of one stage was noted in 28.3% of the patients, in 7% for two stages, and in 1.4% for three stages.The κ for comparing the maximal versus the minimal results was 0.43. There was no significant difference in the characteristics of patients with discordance and patients without discordance including age, sex, waist circumference, BMI, and etiology of liver disease. The stage of fibrosis was associated with discordance between measurements (P<0.001), demonstrating low discordance rate in patients with stages F0-F1 or F4 and high discordance rate in patients with stages F2 and F3. Sampling variability according to probe location is seen in transient elastography in approximately 30% of patients. Therefore, it may be suggested to perform transient elastography from various sites to minimize the sample error.