Magnitude of contrast-enhanced ultrasonography as a noninvasive predictor for hepatic fibrosis: comparison with liver stiffness measurement and serum-based models.

PURPOSE

To elucidate the efficiency of contrast-enhanced ultrasonography alone and in combination with other noninvasive models for grading hepatic fibrosis.

METHODS

This prospective study included 74 patients with four grades (F1-F4) of chronic liver disease (17, 20, 18, and 19 patients, respectively). Diagnostic performances of the contrast parameter (time to the maximum intensity ratio between the right portal vein and liver parenchyma from the onset of contrast enhancement in the right portal vein) assessed by ultrasonography, liver stiffness measurement (LSM), FIB-4 test, and type IV collagen 7s were compared with histological findings.

RESULTS

Greatest areas under the receiver operating characteristics curve (Az) with the single model were 0.83 (95 % confidence interval 0.71-0.91) for marked fibrosis (≥F2) by FIB-4 test; 0.85 (0.73-0.92) for advanced fibrosis (≥F3) by LSM, and 0.92 (0.83-0.96) by type IV collagen 7s for cirrhosis (F4). When combined, Az for marked fibrosis was ≥0.82; the best Az value was 0.87 (0.74-0.94) for the combination of contrast parameter with FIB-4. Similarly, the Az for advanced fibrosis was ≥0.82, and the best Az value was 0.89 (0.78-0.94) for the combination of contrast parameter with LSM. The Az for cirrhosis was ≥0.95, and the best Az was 0.99 (0.97-1.00) for the combination of contrast parameter with LSM.

CONCLUSIONS

The contrast parameter is a promising predictor for grading hepatic fibrosis when combined with LSM or FIB-4.