The normal iron content in the human body is 40 mg/kg in women by menopause and 50 mg/kg in men14,15. About 80% of total iron in the human body occurs in the red blood cells and in proteins other than hemoglobin. The remaining 20% is largely stored in the ferritin and hemosiderin16,17. Ferritin acts as a deposit storing excess iron and occurs primarily in the liver, spleen, and bone marrow17.
A smaller amount of iron is associated with transferrin in plasma18. In women, the amount of stored iron is lower and additionally depends on the severity of menstruation, pregnancy, lactation as well as the amount of iron supplements taken. Premenopausal women lose iron due to menstruation at a rate of about 0.5–1.0 mg/day.
There are no physiological mechanisms to eliminate iron from the body when there is excess2. If the supply of iron from the various sources is too high, then systemic iron overload occurs in the body. Initially, ferritin and hemosiderin (which is a partially denatured form of ferritin) store the excess iron. If the iron-binding capacity of the substances is exceeded, free iron is deposited in the cells of other organs.
Even if the level of ferritin in the blood can be used as a rough indicator of liver iron overload in a highly overloaded population19, it is considered that it does not allow the diagnosis of liver iron overload and should not be used solely for clinical decision making20,21,22.
In the past 20 more years, many MRI methods have been developed to measure hepatic iron concentration, including SIR methods at 0.5 T23, 1.5 T24,25, and 3 T26, R2 calculation at 1.5 T1 and R2* calculations at 1.5 T6,21,27,28,29 and at 3 T30.
Due to whole body iron metabolism31, MRI is the most practical and accurate method to monitor the concentration of iron in various human organs. The literature describes attempts to assess iron content in the body using R2* or T2*magnetization relaxation values (R2* = 1000/T2*). The majority of the studies showed a good correlation between iron concentration in the liver and T2 or T2* in patients with iron-overload diseases in the course of congenital hemochromatosis, thalassemia, and sickle cell disease1,6,19,20,23,25,32,33,34,35,36,37.
The MRI measurements hepatic R2* or T2* is a fast, widely available, and relatively inexpensive technique that some institutions are willing to use as an alternative to liver biopsies. However, for studies of T2*/R2* in healthy individuals, the literatures only reported a small number of participants6,35. Maris et al. measured the T2* value for the liver in 21 healthy volunteers35. In the Wood study, for comparative purposes, data were obtained in a 13-person control group6. The only more notable work assessing the levels of T2* in the liver and spleen in healthy volunteers and comparing these values with the level of ferritin in blood11 is of limited value due to the non-objective criterion of qualification: only a participant health declaration without basic corroborating laboratory tests.
The results from our study fill this existing gap in the literature. In our work, we included healthy volunteers with no history of liver disease and negative blood tests that excluded inflammatory processes in the body. The absence of fibrosis processes in the liver was additionally confirmed by MR elastography exam. The stiffness of the liver was in the range of 1.84–2.82 kPa with an average of 2.30 (0.23) kPa, which was within the normal range38.
In our study, the average value of R2* was 28.751 s−1 (T2* = 34.78 ms). For comparison, Anderson gives a T2* value of 33 ± 7 ms for the liver and 56 ± 22 ms for the spleen27. Maris et al. reported the average T2* for the liver as 24.2 ± 3.0 ms in 21 healthy volunteers35. The recommended upper limit of LIC is < 2 mg Fe/g dry liver tissue39. Another study by Nuttall et al., based on a series of 141 liver biopsies, suggests reference limit 1.8 mg Fe/g dry liver tissue in healthy adults40. The LIC increases with age, an average of 14 mcg Fe/g per year40. As expected, the LIC values in women are lower than in men41. The above-mentioned LIC value of 2 mg/g is equivalent to the R2* value of about 70 s−129. The R2* value measured in our healthy volunteers corresponds well with the proposition of classification of iron overload severity offered by Henninger et al.39. In the 13 healthy volunteers (9 male, 4 female) with ages of 29.3 ± 12.3 years (range, 12–50 years) in the Wood’s study, the normal R2* was 39.9 ± 2.8 s−1 (HIC = 1.17 mg/g dry weight)6.
In the presented series, there was no correlation between R2* and liver stiffness, neither for women nor for men. This is consistent with the results of other authors42.
Our study showed a clear difference in the mean liver R2* between men and women. Women had a statistically significantly lower R2*value (28.34 s−1) compared to men (29.57 s−1), p = 0.040. This was expected as premenopausal women lose iron due to menstruation. However, premenopausal women lose iron due to menstruation so our women-dominated population of young healthy volunteers should has a lower iron level than a more general population that includes more elderly people and men.
Also, in this study, we only used the R2* images on the scanner, generated by the manufacturer IDEAL-IQ complex-based fat-water R2* model. There are many other magnitude-based R2* models, for example, Yokoo et al.43 studied the effect of Rician and non-Rician noise on the accuracy of R2* estimation in 7 different magnitude-based models, and found, the agreement of the R2* estimation was excellent in patients with low ferritin but poor in patients with high ferritin, as expected. Tipirneni-Sajja et al.44 simulated 5 different magnitude-based R2* models (GRE-A, GRE-B, GRE-C, UTE-A, and UTE-A), and found, the constant offset monoexponential model overestimated the R2* at the lowest simulated SNR, and the monoexponential fitting with noise subtraction is the most robust and accurate R2* model for ultrashort TE (UTE) sequence.
A limitation of the presented work is the narrow age range of the studied group. Therefore, it was not possible to analyze the correlation of R2* with age. Further studies should include individuals within a wider age range to investigate such potential dependence. Equally interesting would be the analysis of LIC in menstruating and postmenopausal women.
The lack of liver biopsies as an alternative measure of iron content can also be considered a weakness of this study, but the design of the work a priori excluded such a reference point.
By measuring liver R2* during routine MR testing and referring it to the reference values, patients with mild elevation of ferritin blood levels and no iron overload can avoid invasive diagnostics (e.g., liver biopsy) or invasive blood-letting tests. It should be remembered that an increase in ferritin levels might occur in the course of other disease processes such as hepatitis, obesity, or alcohol consumption45.
Also, the MR method used in this study to assess the hepatic iron level has the additional advantage of being able to evaluate the total amount of iron in the whole liver, as opposed to the localized assessment from biopsy, which can cause a significant diagnostic error in the case of an inhomogeneous distribution of iron in the liver46.
In summary, MRI R2* has been developed to accurately measure hepatic iron concentration. Our study has shown a normative hepatic R2* value of less than 38 s−1 in a women-dominated (64%) and relatively large population of healthy young people, which is a valuable reference for the medical community (Fig. 6).