Dual-Energy CT for Patients Suspected of Having Liver Iron Overload: Can Virtual Iron Content Imaging Accurately Quantify Liver Iron Content?

Purpose To prospectively assess the feasibility of using virtual iron content (VIC) imaging at dual-energy computed tomography (CT) to evaluate the liver iron content (LIC) in patients suspected of having liver iron overload and to compare the LIC grading performance of VIC imaging and magnetic resonance (MR) imaging. Materials and Methods This study was approved by the institutional review board, and informed consent was obtained from all patients. Fifty-six patients suspected of having liver iron overload (serum ferritin concentrations >500 μg/L) underwent unenhanced dual-energy CT and MR imaging of the liver. MR imaging–measured LICs were obtained in 34 of the 56 patients. VIC images were generated with dual-energy analysis. R2* and MR-measured LIC were obtained with gradient-echo and spin-echo sequences, respectively. Correlations between CT and MR measurements were analyzed. The diagnostic performance of VIC and R2* in the differentiation of different LIC thresholds were evaluated with receiver operating characteristic (ROC) analysis. Results Hepatic VIC showed significant correlation with R2* and MR-measured LIC (r = 0.885 and 0.871, respectively; P < .0001). To differentiate among different LIC thresholds of 1.8, 3.2, 7.0, and 15.0 mg of iron per gram of dry tissue, the corresponding optimal cutoff values for VIC were 2.50, 5.13, 8.93, and 17.97 HU, respectively. At a LIC threshold of 7.0 mg of iron per gram of dry tissue or higher, 100% sensitivity (15 of 15 patients) and 100% specificity (19 of 19 patients) were obtained for VIC. There was no significant difference between VIC and R2* (area under the ROC curve, 0.964 vs 0.993, respectively; P = .299) in grading LIC levels at a LIC threshold of 3.2 mg of iron per gram of dry tissue or higher. Conclusion Hepatic VIC is a potential index for accurately evaluating and grading clinically significant liver iron accumulation, with a diagnostic performance similar to that of MR imaging. © RSNA, 2015 Online supplemental material is available for this article.