Original Research MRI Quantification of Splenic Iron Concentration in Mouse Eric Hitti, PhD, 1–3 * Pierre-Antoine Eliat, PhD, 3,4 Emmanuelle Abgueguen, PhD, 5,6 Martine Ropert, MD, 5,7 Patricia Leroyer, 5,6 Pierre Brissot, MD, 4–6,8 Yves Gandon, MD, 4–6,9 Herve Saint-Jalmes, PhD, 1–4 and Olivier Lore ´al, PhD, MD 4–6 Purpose: To quantify hepatic and splenic iron load, which is a critical issue for iron overload disease diagnosis. MRI is useful to noninvasively determine liver iron concentration, but not proven to be adequate for robust evaluation of sple- nic iron load. We evaluated the usefulness of MRI-derived parameters to determine splenic iron concentration in mice. Materials and Methods: A mouse model of experimental iron load was used. Multi-echo spin-echo images of liver and spleen were acquired at 4.7 Tesla. The parameters were tested at all echoes with and without an external ref- erence. Splenic and hepatic iron concentrations were determined using biochemical assay as the gold standard. Results: Our results show that (i) use of an internal or external reference is essential; (ii) optimal echo times were TE ¼ 19.5 ms and TE ¼ 32.5 ms for the liver and spleen, respectively; (iii) in the liver, the relationship between biochemical and MRI iron concentration determi- nations is logarithmic; (iv) in the spleen, the best relation- ship is an inverse function. Conclusion: A single spin-echo sequence allows robust estimation of hepatic and splenic iron content. Parameters classically used for hepatic iron concentration cannot be applied to splenic iron determination, which requires both the specific sequence and the adapted fitting function. Key Words: MRI; iron overload; liver; spleen; mice; hemochromatosis J. Magn. Reson. Imaging 2010;32:639–646. V C 2010 Wiley-Liss, Inc. SYSTEMIC IRON OVERLOADS, especially during genetic hemochromatosis, may lead to the develop- ment of complications having a strong impact on life expectancy or on the quality of life (1). The positive diagnosis of an iron overload disease (2) requires the quantification of iron excess, which is mainly based on the direct evaluation of liver iron concen- tration (LIC). Historically, LIC was determined on liver biopsies using a biochemical method (3). Histo- logical evaluation performed by a trained patholo- gist, after Perls staining, may also be useful to quantify iron overload, especially during genetic hemochromatosis (4). Several groups have reported noninvasive quantification of hepatic iron overload using MRI methods (5). Most studies were done on humans (6–11), only a few having been performed on rodents (12–15). Recently identification of new genes, which when mutated are involved in the development of iron over- load diseases, has improved both the molecular understanding of iron overload diseases and the clas- sification of iron overload (1,16). Most of these genetic diseases are characterized by a low splenic iron con- tent, whereas other genetic entities, including ferro- portin disease, are associated with maintained or an increased splenic iron concentration (17). Moreover, iron overload can be related to hematological diseases or to dysmetabolic syndrome, in which splenic iron has been reported not decreased and/or increased. Therefore, the evaluation of splenic iron content may be useful for the diagnosis procedure of iron overload diseases. Due to the unavailability of splenic biopsies for obvious ethical reasons, the correlation between MRI parameters and biochemical splenic iron concentra- tion (SIC) determination has not yet been established. As a consequence, SIC is calculated using the same formulas as for the liver (18,19). Therefore, to explore the appropriateness of trans- ferring the parameters used for liver iron concentra- tion to splenic iron determination, our work aimed at (i) comparing the performance of different MRI param- eters for LIC determination and (ii) studying the valid- ity of MRI for SIC determination. For this purpose, we used a mouse model for iron overload. 1 Universite ´ de Rennes 1, LTSI, Rennes, France. 2 INSERM, U642, Rennes, France. 3 Universite ´ de Rennes 1, IFR 140 GFAS, PRISM, Rennes, France. 4 Universite ´ Europe ´enne de Bretagne, Rennes, France. 5 INSERM, U991, Rennes, France. 6 Universite ´ de Rennes 1, Rennes, France. 7 Laboratory of Biochemistry, CHU Rennes, Rennes, France. 8 Liver Diseases Unit, CHU Rennes, Rennes, France. 9 Imaging Department, CHU Rennes, Rennes, France. *Address reprint requests to: E.H., Faculte ´ de pharmacie, 2 avenue du Professeur Le ´on Bernard, 35043 Rennes Cedex, France. E-mail: eric.hitti@univ-rennes1.fr Received December 28, 2009; Accepted June 3, 2010. DOI 10.1002/jmri.22290 View this article online at wileyonlinelibrary.com. JOURNAL OF MAGNETIC RESONANCE IMAGING 32:639–646 (2010) V C 2010 Wiley-Liss, Inc. 639