Point shear wave elastography method for assessing liver stiffness

Abstract

Aim: To estimate the validity of the point shear-wave elastography method by evaluating its reproducibility and accuracy for assessing liver stiffness.

Methods: This was a single-center, cross-sectional study. Consecutive patients with chronic viral hepatitis scheduled for liver biopsy (LB) (Group 1) and healthy volunteers (Group 2) were studied. In each subject 10 consecutive point shear-wave elastography (PSWE) measurements were performed using the iU22 ultrasound system (Philips Medical Systems, Bothell, WA, United States). Patients in Group 1 underwent PSWE, transient elastography (TE) using FibroScan (Echosens, Paris, France) and ultrasound-assisted LB. For the assessment of PSWE reproducibility two expert raters (rater 1 and rater 2) independently performed the examinations. The performance of PSWE was compared to that of TE using LB as a reference standard. Fibrosis was staged according to the METAVIR scoring system. Receiver operating characteristic curve analyses were performed to calculate the area under the receiver operating characteristic curve (AUC) for F ≥ 2, F ≥ 3 and F = 4. The intraobserver and interobserver reproducibility of PSWE were assessed by calculating Lin's concordance correlation coefficient.

Results: To assess the performance of PSWE, 134 consecutive patients in Group 1 were studied. The median values of PSWE and TE (in kilopascals) were 4.7 (IQR = 3.8-5.4) and 5.5 (IQR = 4.7-6.5), respectively, in patients at the F0-F1 stage and 3.5 (IQR = 3.2-4.0) and 4.4 (IQR = 3.5-4.9), respectively, in the healthy volunteers in Group 2 (P < 10(-5)). In the univariate analysis, the PSWE and TE values showed a high correlation with the fibrosis stage; low correlations with the degree of necroinflammation, aspartate aminotransferase and gamma-glutamyl transferase (GGT); and a moderate negative correlation with the platelet count. A multiple regression analysis confirmed the correlations of both PSWE and TE with fibrosis stage and GGT but not with any other variables. The following AUC values were found: 0.80 (0.71-0.87) for PSWE and 0.82 (0.73-0.89) for TE (P = 0.42); 0.88 (0.80-0.94) for PSWE and 0.95 (0.88-0.98) for TE (P = 0.06); and 0.95 (0.89-0.99) for PSWE and 0.92 (0.85-0.97) for TE (P = 0.30) for F ≥ 2, F ≥ 3 and F = 4, respectively. To assess PSWE reproducibility, 116 subjects were studied, including 47 consecutive patients scheduled for LB (Group 1) and 69 consecutive healthy volunteers (Group 2). The intraobserver agreement ranged from 0.83 (95%CI: 0.79-0.88) to 0.96 (95%CI: 0.95-0.97) for rater 1 and from 0.84 (95%CI: 0.79-0.88) to 0.96 (95%CI: 0.95-0.97) for rater 2. The interobserver agreement yielded values from 0.83 (95%CI: 0.78-0.88) to 0.93 (95%CI: 0.91-0.95).

Conclusion: PSWE is a reproducible method for assessing liver stiffness, and it compares with TE. Compared with patients with nonsignificant fibrosis, healthy volunteers showed significantly lower values.

Keywords: Chronic viral hepatitis; Elastography; Liver biopsy; Liver fibrosis; Sensitivity and specificity; Ultrasound.

Figures

Figure 1

Box and whisker plots of stiffness values in patients at METAVIR stage F0-F1 and in healthy subjects. Box plots show interquartile range (box), median (line within box), range (whisker) and outliers (circle). Median values (IQR) and P values of differences are given. A: Point shear wave elastography (PSWE) values; B: Transient elastography (TE) values.

Figure 2

Distribution of stiffness values according to METAVIR fibrosis stage. Median values, interquartile ranges, ranges, numbers of outliers and P values are given for each fibrosis stage. A: Point shear wave elastography (PSWE); B: Transient elastography (TE).

Figure 3

Receiver-operating characteristic curves for different fibrosis thresholds. The performance of point shear wave elastography (PSWE) was compared to that of transient elastography (TE). P values for differences between areas under the curve are given. A: F0-F1 vs F2- F4 (F ≥ 2); B: F0-F2 vs F3-F4 (F ≥ 3); C: F0-F3 vs F4 (F = 4).