• NAFLD
• steatohepatitis
• steatoscore
• ultrasound score

Non alcoholic fatty liver disease (NAFLD) is divided in non alcoholic fatty liver and non alcoholic steatohepatitis. The intrahepatic deposition of adipose tissue initially determines a pattern of hepatic inflammation that, with time, results in hepatocyte damage. The main risk factors for the onset of NAFLD are obesity, diabetes and dyslipidemia and its most serious complication is represented by its evolution into cirrhosis with the possible onset of hepatocarcinoma. Considering the significant worldwide prevalence of NAFLD (in Europe it is estimated around 25% of the general population), it has become essential an early diagnosis and follow-up through imaging methods that are easily available and low cost but, at the same time, reliable.

Currently the gold standard methods which can accurately diagnose NAFLD are the biopsy primarily and Magnetic Resonance (MR) secondarily. However, the first one is an invasive method and the second one is expensive and not always available. Abdominal ultrasound (US), vice-versa, is easily available, low cost and does not expose the patient to ionizing radiation. For this reason, US has been recommended by "European Association for the Study of the Liver" with the "European Association for the Study of Diabetes" and the "European Association for the Study of Obesity" as the first line imaging method for early diagnosis and follow up of NAFLD. However, US shows limitations due to the fact that is operator-dependent and provides only qualitative information. For this reason, in order to overcome these problems, US scores have recently been developed in order to change US in a more reproducible and quantitative imaging method for the quantification of fatty involvement of liver parenchyma.

On 2018, a group of italian Authors validated a new score by using US in comparison with MR, the so called SteatoScore. This score is obtained by the acquisition of three standardized abdominal US clips, that provide five parameters that are put into a mathematical formula for steatosis quantification. These parameters are calculated with a custom-made software tool created by using Matlab and they are: attenuation rate (AR), hepatic renal ratio (HR), diaphragm visualization (DV), portal vein wall visualization (PVW) and hepatic/portal vein ratio (HPV). The multi-parametric score obtained was called “UMB”, an acronym for “Ultrasound in Medicine and Biology”, the journal where the original paper was submitted. However, the mathematical model was more recently updated to “UMB2” and “NEW” by the same Authors on a larger group of patients.

The objectives of this study were an intra-operator and an inter-operator analysis of SteatoScore and its comparison with baseline CT scans in 51 consecutive oncologic patients (31 males and 20 females, mean age 64 years old). Patients, enrolled between January 2019 and September 2019, underwent US examination as part of their follow-up protocol. All patients had to have a CT examination performed within the previous 12 months. Patients with documented chronic hepatopathy and cirrhosis were excluded from the study.

In the first part of the study, the operator 1 performed the same procedure of US parameters collection twice with one year in between by using Matlab. Subsequently the operator 2 performed the same procedure for the third time. All US parameters collected were put in multi-parametric scores UMB, UMB 2 and NEW. In the second part of the study, the three groups of parameters measurement of the SteatoScore were compared with those obtained by the previous baseline CT performed. In particular, CT parameters calculated by the standardized placement of a region of interest (ROI) within liver parenchyma were hepatic attenuation (HA), liver spleen ratio (LS ratio) and liver spleen difference (LS difference).

Analyzing each single parameter (HR, AR, DV, HPV, PVW) and the multi-parametric scores (UMB, UMB 2 and NEW), according to coefficient variation analysis (CV), intraclass correlation coefficient (ICC), Bland Altaman analysis, intra-operator variability was less than inter-operator variability, and SteatoScore NEW was the best one in terms of reproducibility.
By the comparison of US parameters and CT parameters, no case exceeding 50% of agreement was observed. On the other hand, the correlation between multi-parametric scores (UMB, UMB2 e NEW) and CT parameters was higher, with 64% of correlation between SteatoScore NEW and LS difference. The multi-parametric scores (UMB, UMB2 e NEW) showed low sensitivity to classify patients with moderate/severe grade of steatosis, classifying them as low steatosis classes (S0 and S1).

In conclusion, the SteatoScore NEW has a high reproducibility and shows the best correlation with CT, even if it showed a low sensitivity to classify patients with a moderate/severe grade of steatosis in this small group of patients.