Journal of Magnetic Resonance Imaging | 2021
Editorial for “Limits of Fat Quantification With Chemical Shift Encoded”
Abstract
Liver steatosis is the hallmark of nonalcoholic fatty liver disease (NAFLD), the most common cause of diffuse liver disease in the Western countries, but it is also frequently found in other liver diseases, such as viral hepatitis, alcoholic hepatitis, metabolic, or genetic disorders. Hepatic iron overload is characteristic of hereditary hemochromatosis or secondary hemosiderosis. Nevertheless, abnormal iron deposition in the liver is also often recognized in patients with diffuse liver diseases. In this scenario, hepatic iron and steatosis frequently coexist. Proton density fat fraction (PDFF), measured with multi-echo chemical-shift encoded (CSE) magnetic resonance (MR) sequences, is now considered the best imaging biomarker for quantifying liver steatosis. These sequences take advantage of the different chemical-shift of water and fat protons, to separate the water and fat proton signals and thus calculating PDFF. To be reliable, these sequences must be corrected for the main confounders, such as T1 and T2* effect, noise related bias, and the spectral complexity of fat. Particularly, if uncorrected, the inevitable T2* decay with increasing echo time will underestimate the fat fraction. The correction of T2* effect is obtained by estimating the R2* relaxation rate, which is directly related with the liver iron content. Comprehensibly, R2* correction is of great importance for MR PDFF quantification in patients with coexisting liver steatosis and siderosis. Additionally, the liver R2* value can be used as an imaging biomarker of tissue iron deposition. In the last decade, multi-echo CSE MR sequences were validated for simultaneous quantification of fat and iron and are now being widely used in different clinical and research scenarios. Nevertheless, they have some limitations that must be recognized to avoid misinterpretation. Iron (R2*) quantification may be inaccurate in cases of high iron overload livers, mainly at 3 T MR. Furthermore, PDFF quantification is corrupted by the presence of high iron overload and, therefore, the estimated PDFF values will not be reliable in patients with high hepatic iron content and elevated R2*. In fact, as liver R2* increases, the signal decay will be faster, with linewidth broadening, and oscillation of fat and water peaks will be nearly indistinguishable, making it difficult or even impossible to separate water and fat signals and, consequently, hampering PDFF quantification. Nonetheless, the precise iron related-R2* threshold above which PDFF cannot be correctly estimated is yet to be defined. In this issue of JMRI, an article by Zhao et al adds important information toward this purpose. The authors aimed to determine the upper limit of liver R2*, above which it is no longer feasible to quantify PDFF reliably. Theoretical analyses and Monte-Carlo computer simulations were used to characterize the variability of PDFF estimates at increasing R2* values, for both 1.5 T and 3 T. The empirical limit for R2* was determined by the choice of echo combinations and the desired number of effective signal averages. Their results suggested that PDFF estimates would be unreliable above an R2* of 538/second and 779/second at 1.5 T and 3 T, respectively. This empirical formula was then validated with phantom experiments and in a prospective study including 57 patients with known or suspected liver iron overload. This work is relevant for different clinical or research scenarios. Firstly, due to the high prevalence of liver steatosis in the population, patients with iron overload diseases may also have liver steatosis. In fact, liver steatosis was found to be more frequent than previously thought in patients with hereditary hemochromatosis, and a common finding in patients with hereditary hemochromatosis under maintenance phlebotomies. Secondly, hepatic iron deposition is frequent in patients with diffuse liver diseases, including NAFLD patients. Particularly in the setting of clinical trials for new drugs tailored for NAFLD and nonalcoholic steatohepatitis, precise and reliable PDFF quantification is needed for the diagnosis and longitudinal treatment monitoring of liver steatosis. As multi-echo CSE MR sequences are increasingly being used worldwide, the radiologists should be aware of those patients with coexisting hepatic steatosis and iron overload, in whom PDFF quantification will be hampered by elevated R2*. In practice, for a given patient, if estimated liver R2* is higher than the threshold-value, the PDFF estimated value will not be reliable and it should not be reported. Although the proposed R2* threshold is contextand application-specific and focused only on complex-based R2*-