Dual‐energy computed tomography for non‐invasive staging of liver fibrosis: Accuracy of iodine density measurements from contrast‐enhanced data

2018 
AIM: To investigate whether iodine density measurements from contrast-enhanced dual-energy computed tomography (CT) data can non-invasively stage liver fibrosis. METHODS: This single-center, prospective study was approved by our IRB with written informed consent. Forty-seven consecutive patients (26 men and 21 women; mean age, 63.1 years) with chronic liver disease underwent contrast-enhanced dual-energy CT of the liver (non-contrast, arterial, portal venous, and equilibrium phase images), followed by liver biopsy. Iodine density of liver and aorta were obtained by two independent observers. Iodine uptake of the liver (Δ Liver), representing the difference in iodine density between equilibrium phase and non-contrast images, was calculated and normalized by aorta (Δ Liver/Aorta). We accounted for contrast agent distribution volume by using hematocrit level. Accuracy of iodine density measurements for staging liver fibrosis was assessed by using receiver operating characteristic (ROC) curves. Multivariate linear regression analysis was used to assess the impact of independent variables (liver fibrosis stage and patient-related confounders) on iodine uptake. RESULTS: The Δ Liver/Aorta significantly increased and moderately correlated with METAVIR liver fibrosis stage (ρ = 0.645, P < 0.001). Areas under the ROC curve ranged from 0.795 to 0.855 for discriminating each liver fibrosis score (≥F1-F4). METAVIR fibrosis stage was the most significant independent factor associated with Δ Liver (P = 0.005) and Δ Liver/Aorta (P < 0.001). CONCLUSION: Hepatic extracellular volume fraction with contrast-enhanced dual-energy CT can non-invasively stage liver fibrosis in chronic liver diseases. This technique could prove useful for monitoring disease progression and treatment response, potentially reducing the need for liver biopsy.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    29
    References
    21
    Citations
    NaN
    KQI
    []