• Feasibility and Accuracy of Dual-Source Dual-Energy CT for Noninvasive Determination of Hepatic Iron Accumulation

    Radiology: Volume 262: Number 1 -January 2012

    Eugene Joe, MD Se Hyung Kim, MD Kyoung Bun Lee, MD Ja-June Jang, MD Jae Young Lee, MD Jeong Min Lee, MD Joon Koo Han, MD Byung Ihn Choi, MD

    Purpose: To retrospectively assess feasibility of dual-source dual-energy (DSDE) computed tomography (CT) for evaluation of hepatic iron accumulation in a liver phantom and liver transplantation candidates and to compare its accuracy with that of 3-T magnetic resonance (MR) imaging.

    Materials and Methods: This study was approved by the institutional review board; informed consent was waived. A liver agar phantom con-taining six tubes of iron (concentrations, 100-5000 mg of iron per 100 mL of solution) was scanned at 80 and 140 kVp with both DSDE mode and single-source dual-energy (SSDE) CT with sequential scanning mode. Difference of averaged attenuation between 80 and 140 kVp at CT (AH) was measured and correlated with iron concentration. Thirty-two liver transplant recipients and 55 donors who underwent DSDE CT at 80 and 140 kVp were included. Twenty-three underwent 3-T liver MR with dual-echo in-phase and opposed-phase Tl-weighted and spin-echo T2-weighted imaging. Hepatic AH was measured at CT. On Tl- and T2-weighted MR images, iron indexes were calculated. Degree of iron accumulation and macrosteato-sis were determined at histologic examination (reference). Diagnostic performance of AH at CT and iron indexes at MR for diagnosing clinically important iron accumulation was evaluated (receiver operating characteristic [ROC] analysis).

    Results: For phantom study, AH obtained from both DSDE mode and SSDE CT with sequential scanning mode was correlated with iron concentration (correlation coefficient, 1.00 and 0.943, respectively; P = .173). AH (13.53) in 10 patients with clinically important (≥10%) iron accumulation was significantly higher than that (7.39) in 77 patients with normal or mild iron deposition (P < .001). AH was significantly correlated with degree of iron accumulation (correlation coefficient, 0.430; P < .001) but not with degree of hepatic macroste-atosis (P = .216). Area under the ROC curve for diagnosing clinically important iron accumulation was 0.881 and 0.897 with CT and MR, respectively (P = .851).

    Conclusion: DSDE CT is accurate for diagnosing clinically important hepatic iron accumulation without confounding influence of hepatic steatosis, with diagnostic performance on par with MR.