Evaluation der diagnostischen Genauigkeit der B-Bild Sonographie zur Graduierung der Lebersteatose bei Patienten mit chronischer Lebererkrankung

Fatty liver disease is characterised by excessive storage of fat. The prevalence in Germany is about 30%. The worldwide prevalence is approximately 25%. This means that about 1.8 billion people are affected. Steatosis hepatis is diagnosed when 5% or more of the liver parenchyma are fatty. Based on the intracellular fat content, four stages are differentiated.

The spectrum of manifestations of steatosis is very diverse. Also the clinical presentation of patients with the disease is quite non-specific.

The aim of this work was to investigate the diagnostic accuracy of B-Mode sonography in determining the degree of steatosis in patients with hepatopathy. A retrospective analysis of data from a total of 231 patients between 03/2016 and 05/2019 was performed. 157 patients were included in the study after applying the exclusion criteria. B-Mode ultrasound of the liver was performed with high-end ultrasound devices such as the Logiq E9 (GE Medical Systems, Wauwatosa, USA; software R1.0.6) or Hitachi ALOKA (Pro Sound Alpha 7).

B-Mode ultrasonography showed a sensitivity of 75.6% and a specificity of 76% in discrimination between steatosis and no steatosis (AUROC 0.758). This resulted in a positive predictive value of 0.775 and a negative predictive value of 0.74. The diagnostic accuracy was 0.758 (0.680 - 0.836). For higher grades of steatosis (>II°), the specificity for the presence of histological steatosis was >98.7%. The sensitivity was 90.9%. The agreement between the two investigators was 89.2% (κ=0.783) for the presence of steatosis and 80.9% (κ=0.704) for the individual grades of steatosis. We were able to show that B-Mode ultrasonography has excellent diagnostic accuracy in the diagnosis of moderate and massive steatosis. In the presence of mild steatosis, the sensitivity was somewhat lower. As a result, high-end equipment may have increased sensitivity for steatosis detection, but also decreased specificity, resulting in more false positives compared to older sonography equipment. However, higher grade steatosis (grade 2 or higher) is detected very reliably, as shown in our work. The results were also detectable in patients with aetiology other than NAFLD.

Furthermore, we were able to show that the presence of fibrosis has no decisive influence on the sonographic diagnosis of fatty liver disease. Also, we were able to record very good results in the interobserver agreement. Overall, the results of this work show that B-scan sonography is a very good method to detect higher grade steatosis, independent of the degree of fibrosis or concomitant liver disease. Thus, sonography can be considered the diagnostic tool of first choice for the detection and graduation of steatosis hepatis.