Liver iron quantification: studies in aqueous iron solutions, iron overloaded rats, and patients with hereditary hemochromatosis

Limitations of Magnetic Resonance Imaging in Measurement of Hepatic Iron

To evaluate the usefulness of magnetic resonance imaging for the quantitative determination of hepatic iron, we examined 43 patients with thalassemia major and assessed the influence of pathologic changes in the liver on the precision of estimates of the hepatic iron concentration. Tissue signal intensities were measured from magnetic resonance T1-weighted images derived from gradient-echo (GE) pulse sequences and the ratio of the signal intensity of liver to muscle calculated. By excluding patients (n = 9) having a signal intensity ratio (SIR) less than or equal to 0.2, a linear relationship with hepatic iron was found and subsequent analyses were limited to these 34 patients. In 27 patients with hepatic fibrosis, an overall correlation of −0.848 was found between hepatic iron and SIR. By contrast, in the seven patients with no fibrosis, the correlation coefficient (−0.993) was significantly greater (P < .0001). Despite the differences in correlation, the regression line between hepatic iron and SIR for the patients with no fibrosis did not differ significantly with respect to either slope or intercept from that of the patients with fibrosis. Thus, the presence of fibrosis did not seem to affect the pattern of the relationship between hepatic iron and the SIR, but rather to increase the variability of the relationship. Clinically, the presence of fibrosis makes estimates of hepatic iron derived from magnetic resonance imaging so variable as to be of little practical use in the management of transfusional iron overload.

Limitations of Magnetic Resonance Imaging in Measurement of Hepatic Iron

To evaluate the usefulness of magnetic resonance imaging for the quantitative determination of hepatic iron, we examined 43 patients with thalassemia major and assessed the influence of pathologic changes in the liver on the precision of estimates of the hepatic iron concentration. Tissue signal intensities were measured from magnetic resonance T1-weighted images derived from gradient-echo (GE) pulse sequences and the ratio of the signal intensity of liver to muscle calculated. By excluding patients (n = 9) having a signal intensity ratio (SIR) less than or equal to 0.2, a linear relationship with hepatic iron was found and subsequent analyses were limited to these 34 patients. In 27 patients with hepatic fibrosis, an overall correlation of −0.848 was found between hepatic iron and SIR. By contrast, in the seven patients with no fibrosis, the correlation coefficient (−0.993) was significantly greater (P < .0001). Despite the differences in correlation, the regression line between hepatic iron and SIR for the patients with no fibrosis did not differ significantly with respect to either slope or intercept from that of the patients with fibrosis. Thus, the presence of fibrosis did not seem to affect the pattern of the relationship between hepatic iron and the SIR, but rather to increase the variability of the relationship. Clinically, the presence of fibrosis makes estimates of hepatic iron derived from magnetic resonance imaging so variable as to be of little practical use in the management of transfusional iron overload.

T2 relaxation time of bone marrow water and lipid: correlation with serum ferritin in normal individuals

PURPOSE

This study investigated the correlation between the iron content of bone marrow and the transverse relaxation times (T2 values) of bone marrow water and lipid on MRI.

METHOD

The T2 of the water and lipid fractions of bone marrow was selectively measured in the L1-3 vertebral bodies using the chemical shift misregistration effect on MRI. Results were compared to the serum ferritin levels in 27 healthy subjects.

RESULTS

The 1/T2 of the water fraction ws strongly correlated with the serum ferritin, but there was no correlation between the 1/T2 of the lipid fraction and serum ferritin.

CONCLUSION

Selective T2 measurement of the water fraction in bone marrow allows precise estimation of the marrow iron content.

Hepatic hemosiderosis in non-human primates: quantification of liver iron using different field strengths

Using a non-human primate model of idiopathic hemochromatosis, hemosiderin-induced T2 shortening of the liver was assessed at nine different field strengths over a range of 0.05 to 1.5 Tesla. The 1/T2 values increased linearly with field strength, with all specimens having approximately the same zero-field intercept. The slope of the field increase, termed "field-dependent T2 proton relaxation enhancement (PRE)", appeared to be proportional to the chemically determined tissue iron content, viz. 10.8 s-1T-1(mg Fe/g wet tissue)-1. The correlation between iron content and field-dependent T2 PRE (r = 0.94) was better than the correlation between iron content and 1/T2 values obtained at single field strengths. For livers containing > or = 2 mg Fe/g wet weight, biexponential T2 relaxation behavior emerged at higher field strengths, with the short T2 component (intracellular water) exhibiting a linear dependence of 1/T2 on field, while T2 of the long component (extracellular/sinusoidal water) was nearly field-independent. After maceration of the specimens, all T2 relaxation curves became monoexponential, including those for high iron content at high field strengths. The present data suggest that the use of double-field MR imaging to assess the field-dependent T2 PRE has potential for specific quantification of (liver) tissue iron stores.

In situ measurement of iron overload in liver tissue by dual-energy methods

The errors for determining liver iron content by dual-energy computed tomography (dual-energy CT) are calculated for the ideal case where only monochromatic x-ray beams are used. Because of the strong influence of spatial resolution on the radiation dose needed to reduce the error to a given level, we have also calculated the error in dual-energy transmission measurements alone, where the spatial information along the beam path is lost. The prediction of error was tested by simulations and measurements using x-rays emitted by radioactive isotopes and synchrotron radiation. Good agreement between calculation, simulation and measurement was found. It is shown that concentrations of liver iron content (disregarding variation of tissue composition) can be studied with a skin dose of about 30 mGy using dual-energy CT and even with much lower dose using dual-energy transmission measurements. However, there are sources of error besides photon noise, especially errors caused by variation of tissue composition. For example dual-energy CT, although suggested to avoid artifacts caused by fat in the case of a fatty liver, still is affected by fat. The magnitude of these errors is discussed qualitatively, and possibilities for their reduction are suggested. For a definitive estimate of errors of iron content measurements with optimized apparatus more experimental data for well defined variations of body tissue, especially in the case of haemochromatosis, are needed.

Liver iron stores in patients with secondary haemosiderosis under iron chelation therapy with deferoxamine or deferiprone

Total body iron stores including liver and spleen iron were assessed by non-invasive SQUID biomagnetometry. The liver iron concentration was measured in groups of patients with beta-thalassaemia major or other posttransfusional siderosis under treatment with the oral iron chelator deferiprone (n = 19) and/or with parenteral deferoxamine (n = 33). An interquartile range for liver iron concentrations of 1680-4470 micrograms/g liver was found in these patients. In both groups a poor correlation between liver iron and serum ferritin values was observed. Repeated measurements of liver and spleen iron concentrations as well as determination of liver and spleen volume by sonography were performed in six patients under continuous deferiprone treatment for 3-15 months. In this group detailed information was obtained on the whole body iron store (5-36g) and the iron excretion rates (14-34 mg/d) for each patient. As indicated by decreasing liver iron concentrations, five out of six subjects showed a negative iron balance (2-13 mg/d). Conventional measurements of both serum ferritin and urine iron excretion gave fluctuating results, thus being only of limited use in the control of iron depletion therapy. The non-invasive biomagnetic liver iron quantification is a precise and clinically verified technique which offers more direct information on the long-term efficacy of an iron depletion therapy than the hitherto used methods. This technique may be of use in the clinical evaluation of new oral iron chelators.