Non-invasive characterization of NAFLD has been a clinical requirement. A demonstration of key factors to identify NASH is a considerable issue in the creation of a novel imaging tool which enables early and definite diagnosis. To the best of our knowledge, this is the first study to report the characteristics of the acoustic properties in NASH livers. The NAFLD liver samples in both mouse and human subjects are related with lower acoustic impedance, and it tended to decrease according to the disease progression, from SS to NASH. The authors stress that this unique feature should represent the potential to develop a radiological imaging alternative to liver biopsy.
Although a previous study has shown that the impedance differs depending on the kind of FFAs, the data in the study were obtained with only five FFAs, linoleate acid, α-linolenic acid, oleate acid, palmitate acid, and palmitoleic acid [13]. The present study, with the use of 35 FFAs, further demonstrated that the impedance varies according to the kind of FFA. Moreover, by the interpretation of intrahepatic percentage composition of FFAs, the impedance of FFAs may account for the pathophysiology of lower impedance in NASH. As previously reported, there is a difference in the plasma phospholipid and FFA composition between SS and NASH [18, 19]. However, there is no relationship in the FFA composition between liver tissue and serum [18, 20], and this fact may enhance the application of radiological imaging based on intrahepatic FFAs. Identification of factors that specify the acoustic properties of FFAs may be challenging in the future.
The intrahepatic FFA composition differs between control, SS, and NASH [20]. In the present study, pentadecanoic acid, palmitic acid, palmitoleic acid, elaidic acid, and linoleic acid in the low-impedance group were elevated, and stearic acid in the high-impedance group was decreased in NASH livers. The data may be compatible with the previous report [20] and may explain the lower acoustic impedance in NASH. However, contrary to our results, the remaining six FFAs in the low-impedance group (α-linolenic acid, arachidonic acid, nervonic acid, cis-4,7,10,13,16,19-docosahexanoic acid, γ-linolenic acid, and cis-5,8,11,14,17-eicosapentaenoic acid) were reported to be decreased, and two FFAs in the high-impedance group (oleic acid and linolelaidic acid) were reported to be increased in NASH livers [20]. The differences in race between Asian and Western countries may be one of the reasons for the results. Moreover, interaction of different kinds of FFAs in the liver may affect the mutual acoustic characteristics, which needs to be determined in the future.
One of the typical imaging modalities based on acoustic parameters is ultrasound elastography, an assessment tool using propagation velocity, which has attracted interest worldwide. A recent study has shown increased velocity with the degree of hepatic fibrosis and decreased velocity with the accumulation of fat [21]. However, the early stage of NASH shows less fibrosis, and the assessment of fat deposition may not be effective to differentiate between SS and NASH, suggesting the difficulty in the early diagnosis of NASH. In fact, a study performed in 164 biopsy-proven NAFLD patients [22] has shown that vibration-controlled TE could rule out advanced fibrosis and avoid the need for biopsy in at least 45% patients with NAFLD in the USA. A more well-designed prospective study reported that the model with both the liver stiffness measurement and controlled attenuation parameter had an area under the receiver operating characteristic of 0.71 in diagnosing NASH, which appears unsatisfactory [23]. Taken together, current US-based quantitative tools do not seem sufficiently sensitive to identify steatohepatitis without advanced fibrosis in patients with NAFLD. It is expected that an FFA-based impedance technique may overcome this problem because composition of FFAs shows characteristic features even in pre-cirrhotic NASH livers.
A recent study [21] reported the higher impedance in fibrotic livers than in normal livers. This finding may be reasonable because the presence of fibrosis may resist sound wave propagation. Our study demonstrated that the difference of the acoustic impedance between SS and NASH was not statistically significant (p = 0.113) in mice, and that was marginal (p = 0.050) in human. The data may be explained by the influence of the presence of fibrosis on the acoustic impedance as a confounding factor. It should be further investigated whether the interrelationship between fibrosis and fat may affect the acoustic data.
The major limitation of our study is that the data are based on measurements using an 80-MHz transducer using much higher frequencies than those typically used in the clinical setting. Second, the observation setting may also be far from that in the human body, which shows much greater attenuation affected by the physical size and intervening tissues. Thirdly, small sample size, particularly in human subjects, may limit the value of the data. The reason is that the acoustic measurement required surgically resected specimens because percutaneous biopsy samples were too small to perform the measurement. Such obstacles must be overcome for the impedance-based system to be equipped in the actual US machine.
In conclusion, this study identified lower acoustic impedance in NAFLD, with the reduction appearing dominant in NASH livers. The acoustic property may be based on the intrahepatic composition of FFAs showing characteristic impedance. These data strongly encourage the practical application of this technique to identify NASH in the near future.