Reply to Letter to the Editor: “Diagnostic performance of ultrasound attenuation imaging for assessing low-grade hepatic steatosis”
by Jong Keon Jang, So Yeon Kim (sykim.radiology@gmail.com)
Diagnostic performance of ultrasound attenuation imaging for assessing low-grade hepatic steatosisDear Editor in Chief,
We would like to thank Dr. Karen Cifuentes G and colleagues for their interest in our article, “Diagnostic performance of ultrasound attenuation imaging for assessing low‐grade hepatic steatosis” [1]. It is our great pleasure to respond to their comments.
The accurate identification and quantification of hepatic steatosis is important for patients at risk of nonalcoholic fatty liver disease. The controlled attenuation parameter (CAP) in Fibroscan (Echosens, Paris, France) and attenuation coefficient (AC) in attenuation imaging have been increasingly used for the identification and quantification of hepatic steatosis. However, as Dr. Karen Cifuentes G and colleagues pointed out, there are some concerns regarding the use of these US-based parameters to quantify hepatic fat. Although the quantification of hepatic steatosis using both these US-based parameters is mainly determined by histopathologic hepatic fat quantification, other liver pathologies, such as hepatitis, fibrosis, and hemochromatosis, can affect the propagation and attenuation of sonic beams [2, 3]. Notably, our study population did not include a sufficient number of patients with other liver pathologies, which could confound the results of US-based parameters, as there was only one patient with portal fibrosis without septa (F1) and none of the patients had evidence of iron deposition or hepatitis. Therefore, we could not evaluate the effects of these potential confounding factors on the diagnostic accuracy of CAP and AC for quantifying hepatic steatosis. Future studies are needed to analyze the impact of other liver pathologies in patients with low grade hepatic steatosis. In addition, our study found that both CAP and AC showed a sensitivity of 61.5% and values of 0.808 and 0.829 for the area under the receiver operating characteristics (AUROC), respectively, which were slightly lower than those observed in previous studies [1]. This can be attributed to the fact that participants in our study had a lower prevalence (3.5%) of advanced hepatic steatosis than those of previous studies (20–50%) [4-6]. Our results suggest that these US-based parameters may have some limitations to evaluate low-grade hepatic steatosis.
In accordance with the comments of Dr. Karen Cifuentes G and colleagues, the proton density fat fraction (PDFF) derived from magnetic resonance imaging is arguably the most accurate non-invasive method to quantify hepatic steatosis [7, 8]. Thus, we analyzed our data using PDFF as a reference standard. Of a total of 57 participants in our study, 51 had available PDFF values measured by high-speed T2-corrected multi-echo spectroscopic sequences on the same day of US-based fat quantification. The median PDFF value was 3.1%, ranging from 0.9 to 16.6%. AC was positively correlated with PDFF (ρ = 0.613; 95% confidence interval [CI], 0.406–0.760; p < 0.001). Eleven patients (21.6%) had any grade of hepatic steatosis with the reference value of PDFF ≥5% [9]. The AUROC of AC for the identification of hepatic steatosis was 0.837 (95% CI: 0.708–0.926). The sensitivity and specificity of AC were 81.8% (9/11) and 80.0% (32/40), respectively, with an estimated cut-off AC value (0.62 dB/cm/MHz) with the highest Youden index. For the identification of low-grade hepatic steatosis, sensitivity of AC values seemed to be higher when PDFF was used as a standard reference than when the histologic fat fraction was used, which can be attributed to the differences in the fat quantification methods between those two reference standards [10].
Once again, we deeply appreciate the insightful comments of Dr. Karen Cifuentes G and colleagues, which have provided a different point of view in interpreting our results. Through their comments, we thoroughly considered the advantages and disadvantages of various methods of hepatic fat quantification.