LAUSR

Quantifying liver iron concentration (LIC) is important in a variety of diseases, including thalassemia, sickle cell disease, and chronic liver diseases such as nonalcoholic fatty liver disease or viral hepatitis, in which iron overload may be present. LIC is considered representative of total body iron content and chronically elevated LIC may lead to cardiomyopathy, diabetes, and liver fibrosis and cirrhosis. Although LIC has historically been measured via biopsy, biopsies are severely limited by their small sample size. Hence, noninvasive magnetic resonance imaging (MRI)-based methods have been developed to facilitate treatment follow-up and to evaluate the liver. MRI methods of measuring LIC are based on either T2 (FerriScan; Resonance Health, Perth, Australia) or T2* relaxation times. Both methods provide reproducible estimates of LIC. However, methods based on T2* relaxation times, or equivalently R2* relaxation rates, have been adopted due to lower cost, reduced scan time, and the greater availability of processing software. Published R2* vs. biopsy-LIC calibration curves have expedited the adoption of the R2* methods. R2* measures, however, can be confounded by the presence of fat in the liver. Notably, the majority of MRI vendors have developed 3D multiecho Dixon gradient recalled echo (GRE) sequences for the simultaneous estimation of R2* and proton density fat fraction to assess iron overload and steatosis. These 3D techniques provide full liver coverage in a single breath hold. The associated advanced in line processing, which incorporates multipeak modeling of the fat signal, mitigates the influence of fat on R2* estimates. However, these methods will provide inaccurate results in those patients who cannot hold their breath for the requisite time (eg, young children, those with respiratory complications, or patients who must be sedated for the imaging examination). In this issue, Rohani et al compare a free-breathing 3D radial GRE sequence to a biopsy-calibrated 2D breath-hold GRE and to a 3D breath-hold Cartesian GRE acquisition to assess the ability of the 3D radial sequence to accurately estimate T2* relaxation (and thus LIC) and fat fraction. Studies were conducted in a group of 29 pediatric and young adult patients with hepatic iron overload using all three methods. Analysis of the results demonstrated that the 3D radial GRE sequence produced comparable results to both the 2D and 3D Cartesian GRE sequences, but with minimal motion artifacts in those patients who were unable to breath-hold. While promising, validation of the technique requires additional work. LIC in this study were below 20 mg iron/g. Higher LICs have been reported and may require the use of ultrashort echo times (eg less than 1 msec) for accurate quantification. Fat content can confound the R2* measurement. The 2D GRE method used as the standard was not fatsuppressed, as that can cause underestimation of R2*. Because subjects with a fat fraction greater than 10% were excluded from the study, it is unknown whether very high levels of fat will affect LIC measured with the proposed 3D radial technique. Studies with larger sample sizes across a wide range of LIC and fat fractions would address these issues. Furthermore, the sequence is available currently on scanners from only one major MRI vendor and requires the use of custom software for processing, which may limit the overall applicability of the method. If the technique can be validated in patients with higher LIC and fat fraction on scanners from multiple vendors, it would provide a viable alternative for those subjects who are unable to comply with breath-holding instructions. However, as the authors themselves stated, the 3D radial sequence is much longer than the conventional 2D and 3D breathholding sequences, so it would not replace breath-hold