A novel non-invasive index for the prediction of liver fibrosis in chronic hepatitis B patients with concurrent nonalcoholic fatty liver disease

Background and aims: Controlled Attenuation Parameter (CAP) was first introduced on the FibroScan (FS) device (Echosens, France) in 2010 and has been shown to be well correlated with liver steatosis in patients with various chronic liver diseases [PMID:28039099].A new acquisition method for CAP (continuous CAP, CAPc) has recently been introduced in order to decrease the intra examination measurement variability (IEMV) [ILC2020;poster FRI073].The aim of this interim analysis of an ongoing study is to compare the diagnostic performances and the IEMV between CAP and CAPc.Method: Consecutive patients, all etiologies combined, underwent FS, Magnetic Resonance Imaging (MRI) within 21 days and CAPs obtained by both methods (standard and continuous).A central reading of the MRI images was performed to assess the Proton Density Fat Fraction (PDFF) which was used as the reference to assess the steatosis grades (S >0: PDFF ≥5% and ≤10%; S >1: PDFF ≥10%).The CAPs standard deviations (SD) were compared using the unilateral Wilcoxon signed-rank test.The agreement between CAP and CAPc was assessed with the average bias, the Pearson's coefficient of correlation (CC) and the absolute agreement intraclass coefficient (AA-ICC).The CAPs performances versus the PDFF were estimated with operating characteristics curves (ROC), and their areas under the curves (AUC).AUCs were compared with Delong's bilateral test.Results: 157 patients were included (41% female, mean BMI = 28.4 ± 5.5 kg/m 2 , mean age = 55.0 ± 13.4 years, Prevalence of S >0 and S >1 were 61% and 22%, respectively).The median CAPc SD (13.1 dB/m²) was significantly lower to median CAP SD (28.4 dB/m²) ( p < 0.0001).The median reduction in IEMV was 50% (Fig).The average bias between the CAP and CAPc was 7.9 dB/m.The CC and AA-ICC were 0.90 and 0.89, respectively, highlighting a strong correlation and agreement.AUCs for S >0 were 0.86 (0.79-0.92) and 0.85 (0.78-0.91) for CAP and CAPc, respectively.AUCs for S >1 were 0.85 (0.78-0.92) and 0.86 (0.79-0.93), for CAP and CAPc, respectively.AUCs of CAP and CAPc methods were not significantly different.