Optimizing Tissue Iron Quantification at 3 Tesla

摘要

ABSTRACT Iron overload is a surprisingly common clinical complication, resulting from hyperabsorption, as in hereditary hemochromatosis and thalassemia intermedia, or from recurrent blood transfusions in patients with hemoglobinopathies or bone-marrow failure. Iron accumulates silently for years but ultimately poisons the liver, endocrine glands and heart. Our laboratory has pioneered the use of 1.5 Tesla (1.5T) MRI to quantify the iron burden in the heart, liver, pancreas, and pituitary gland, stratifying clinical risk according to the MRI parameters, R2 and R2*. As a result, MRI-derived estimates of liver and heart iron have become the standard of care in hemoglobinopathy centers and are accepted surrogates for clinical trials of iron chelation therapy. To date, iron quantification has been limited to 1.5T magnets; however, newly developed 3 Tesla (3T) magnets potentially offer improved recognition of end-organ toxicity and insight into the size and species of tissue iron deposits, but require new approaches to imaging the high liver iron concentrations (LIC) observed in some patients. Our first specific aim is to cross-calibrate and clinically validate R2 and R2* estimates at 3T. Patients who undergo 1.5T scanning for clinical indications (approximately 4 per week) will be invited to undergo a combined research 3T examination and serologic assessment of endocrine/hepatic function. Patients will be stratified to provide a broad sampling of organ iron burdens and underlying disease states; examinations will be performed for heart, liver, and pancreas assessment (n=100) and for pituitary measurements (n=60). We hypothesize that R2 and R2* at 3T will scale linearly with respect to measurements at 1.5T but at different slopes. We further hypothesize that 3T assessments of organ volume and fat concentration will better discriminate target organ toxicity than iron assessment alone. Our second aim is to develop and validate new imaging tools to overcome current dynamic range limitations of liver iron quantification at 3T and to exploit the improved tissue-characterization produced by high field measurements. Rapid signal loss current prevents measurement of high LIC values at 3T. We will use novel approaches, including ultrashort echo time techniques and coil-localized free induction decay and spin- echo acquisitions, to accurately measure high LIC at 3T. Coil-localized multi-echo spin-echo acquisitions will also be used to measure differential signal decay properties of hemosiderin aggregates and cytosolic ferritin in liver. We postulate that withholding iron chelation therapy for one week will detectably increase the cytosolic ferritin pool in the liver while leaving the hemosiderin pool unchanged; resumption of therapy should reverse the observation. The ability to track changes in liver iron store on such a short-time scale may prove valuable for rapidly evaluating response to therapy as well as for studying the mechanisms and dynamics of hepatic iron uptake and clearance. This work will broaden patient access to noninvasive iron estimation, improve detection of preclinical iron toxicity, and offer new insights in dynamic changes of iron storage pools.