Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2343031084
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, Z. J.
Right arrow Articles by Cohen, A. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, Z. J.
Right arrow Articles by Cohen, A. R.
Related Collections
Right arrowRelated Article
(Radiology 2005;234:749-755.)
© RSNA, 2005

Experimental Studies

1/T2 and Magnetic Susceptibility Measurements in a Gerbil Cardiac Iron Overload Model1

Zhiyue J. Wang, PhD, Lurong Lian, MD, Qiukan Chen, MD, Huaqing Zhao, MA, Toshio Asakura, MD and Alan R. Cohen, MD

1 From the Department of Radiology (Z.J.W.), Divisions of Hematology (L.L., Q.C., T.A., A.R.C.) and Biostatistics and Epidemiology (H.Z.), Department of Pediatrics, Children’s Hospital of Philadelphia, Pa. Received July 10, 2003; revision requested September 25; final revision received June 3, 2004; accepted June 23. Supported by NIH grant R01 HL61182. Address correspondence to Z.J.W., Department of Diagnostic Imaging, Texas Children’s Hospital, Department of Radiology, Baylor College of Medicine, 6621 Fannin St, MC2 2521, Houston, TX 77030 (e-mail: zjwang@bcm.tmc.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To measure the transverse relaxation rate (1/T2) and magnetic susceptibility of the heart in conditions of iron overload by using magnetic resonance (MR) imaging and to correlate these with the tissue iron concentration in a gerbil model.

MATERIALS AND METHODS: With prior approval by the institutional animal care and use committee, iron overload was induced with one to 15 weekly subcutaneous injections of iron dextran. Nine gerbils had one to five injections, 10 had six to 10, and eight had 13–15. T2 of the whole heart was measured ex vivo (n = 27), and the magnetic susceptibility of the tissue was estimated through measurement of the tissue lysate (n = 25). The iron level was measured (in milligrams of iron per gram of wet tissue) with chemical analysis after MR imaging. While 1/T2 and magnetic susceptibility are not equivalent measures of the chemically determined tissue iron level, correlations were expected and were identified by using linear regression models.

RESULTS: Iron concentration range was 0.28–1.95 mg/g wet tissue. Iron concentration was strongly correlated with 1/T2 (r = 0.92, P < .001, and the root of the mean squares error of the linear prediction, {epsilon}RMS, was 0.17 mg Fe/g wet tissue with a repetition time of 700 msec). Iron concentration also was strongly correlated with magnetic susceptibility (r = 0.90, P < .001, {epsilon}RMS = 0.19 mg Fe/g wet tissue). Multiple regression analysis with combined 1/T2 (with repetition time of 700 msec) and magnetic susceptibility data led to a slight increase in r and decrease in {epsilon}RMS (r = 0.93, P < .001, {epsilon}RMS = 0.16 mg Fe/g wet tissue).

CONCLUSION: The results of this animal model study demonstrate that 1/T2 and magnetic susceptibility values can be used for estimation of the iron level in the heart.

© RSNA, 2005


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
In the absence of chelating therapy, heart disease is a life-limiting complication of transfusional iron overload. Patients receiving regular red blood cell transfusions for many years frequently develop left ventricular hypertrophy, conduction disturbances, ventricular arrhythmia, and refractory congestive failure (1) associated with cardiac iron deposits (2). Iron-chelating therapy has beneficial effects on survival and cardiac disease in patients with thalassemia who receive multiple blood transfusions (3). It is known that when the liver iron level is high, the risk for cardiac disease increases, and it is believed to be beneficial to the heart to keep the liver iron at a low level. In certain cases, however, patients with relatively low liver iron levels develop cardiac disease. Although the total body iron store is well correlated with liver iron (4), the cardiac iron may not be well correlated with the liver iron, as suggested by the lack of correlation between the signal intensity in the liver and that in the heart on magnetic resonance (MR) images. The factors that determine the cardiac iron level in transfusional iron overload have not been identified. Although liver iron concentration can be measured reliably with biopsy or by measuring the tissue magnetic susceptibility by using a superconducting quantum interference device (SQUID) (5), reliable measurement of the iron level in the heart is difficult. Cardiac biopsy is usually not performed, and an endomyocardial biopsy will have very limited value because the highest iron levels are found in the outer one-third of the myocardium (6,7). Furthermore, to date, the SQUID technique has not been successfully used for measuring cardiac iron. In current clinical practice at most medical centers, little information is known about the cardiac iron level in individual patients. Therefore, the development of noninvasive methods for quantification of the iron concentration in the heart is very important (8).

MR imaging has an important role in assessment of iron overload in the heart. Several studies demonstrated a correlation between the signal intensity on spin-echo images or T2 of the heart and the serum ferritin levels in iron overload (9–11). MR imaging studies also demonstrated that decreased cardiac function is correlated with a high 1/T2* relaxation rate (12). However, the interpretation of the data is not straightforward, and T2* may reflect tissue properties other than the iron level. Because biopsy samples from patients are not routinely available, some investigators applied a modified 1/T2-sensitive signal intensity calibration from the liver to the heart (11,13). Animal model studies may shed light on the interpretation of the human MR images, but there are only very limited data (14). Thus, the purpose of our study was to measure the transverse relaxation rate (1/T2) and magnetic susceptibility of the heart in conditions of iron overload by using MR imaging and to correlate these with the tissue iron concentration in a gerbil model (15).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Animals and Iron Loading
Mongolian gerbils were prepared by using the method described by Carthew et al (15). This model is frequently used to study cardiac iron overload (16,17) because the resultant pathologic conditions are similar to those found in human cardiac iron overload. Mongolian gerbils were purchased either from Charles River Laboratories (Wilmington, Mass) or, during a temporary supply shortage at this vendor, from Harlan Sprague Dawley (Indianapolis, Ind). Thirty-six (18 male, 18 female) gerbils were purchased for the study. The ages of the gerbils ranged from 7 to 12 weeks, and the weight of the animals ranged from 40 to 80 g. Iron dextran (100 mg of iron per milliliter; Sigma-Aldrich, St Louis, Mo) was injected subcutaneously once each week through the back of the neck. The iron dose was 0.5 mg per gram of body weight, but twice this dose was used for the first injection, according to the method reported by Carthew et al (15). Between one and 15 weekly injections were used to obtain a range of iron levels. Three animals had one injection; two had three, four, five, six, or seven injections each; two had eight; one had nine; three had 10; four had 13; and four had 15. After the final injection, there was a waiting period of 0–7 weeks before the animals were sacrificed and examined. The study was approved under category A (no pain or minimal pain for the animals) by the institutional animal care and use committee of the Children’s Hospital of Philadelphia, in Philadelphia, Pa. If a gerbil was found to be ill because of the treatment, it was euthanized within 24 hours to stop the animal’s discomfort.

After iron loading, the gerbil was euthanized by using carbon dioxide gas, and the heart was harvested. The whole heart was rinsed carefully to remove residual blood from the tissue. Each gerbil heart was submitted to three examinations: MR imaging for measurement of T2 in the whole heart, MR imaging for measurement of magnetic susceptibility of tissue lysate, and chemical analysis for measurement of the average iron concentration of the entire heart.

T2 Measurement
MR imaging for T2 measurement was conducted with a 1.5-T whole-body clinical imager (Magnetom Vision; Siemens Medical Systems, Iselin, NJ). The whole-body coil was used for radiofrequency transmission, and the surface coil with a diameter of 5 cm (loop small; Siemens Medical Systems) was used for signal reception. The diameter of the heart was less than 1 cm, and the samples were positioned within 1.5 cm from the center of the coil. An apparatus was constructed to keep the tissue temperature constant during MR measurement. The sample and the radiofrequency reception coil were contained in a polystyrene plastic foam box. A heat bath circulator (Neslab Instruments, Newington, NH) circulated warm water at a stable temperature through a coil of copper tubing, in essence serving as a radiator, inside the box. A section of plastic tubing was wound around the copper tube and reached a steady temperature. Air was blown through the plastic tube into the box to establish an environment with a uniform and stable temperature. A thermometer was inserted into the box to allow monitoring of the temperature.

The whole heart was measured by using MR imaging within 11/2 hours of sacrifice. Each heart was weighed and wrapped in a self-sealing thermoplastic sheet (Parafilm M; Pechiney Plastic Packaging, Menasha, Wis) to reduce the tissue water loss, which is known to have the effect of decreasing T2 (18). The weight of the hearts was 0.27–0.41 g. One to three gerbil hearts were sealed in each of two plastic vials, and T2 was measured simultaneously in one section in the transverse plane through the hearts in each vial. The temperature in the samples was kept at 37°C during T2 measurement. The T2 was measured by using a spin-echo sequence with 90° and 180° pulses and multiple interleaved echo times (TEs) (5, 10, 20, 30, 40, 50, 60, and 80 msec). In this scheme, the data for one k-space line were acquired together for images with different TEs before the next k-space line. Interleaving of TEs in the measurement has the benefit that possible instabilities of the system are less likely to introduce errors in the measured T2. Repeated measurements were performed at three different TRs (700, 1000, and 2000 msec, in that order). Other measurement parameters were as follows: section thickness, 10 mm; matrix size, 32 x 64; field of view, 80 mm; and rectangular field of view, 50%. The total data acquisition time for the three T2 measurements was approximately 20 minutes. Data were processed by using specially written software in Interactive Data Language (IDL, version 4; Research Systems, Boulder, Colo). There was no noticeable image intensity variation associated with the use of a surface coil, because of the small organ size relative to the coil diameter and because the samples were positioned in regions where the receiving sensitivity was relatively uniform. The total signal from each whole heart was summed for the T2 analysis. All T2 measurements were performed by one author (Z.J.W.).

Magnetic Susceptibility Measurement
MR imaging for measurement of the tissue magnetic susceptibility was based on a phase mapping method with a gradient-echo sequence (19). The phase discontinuity at the interface of two compartments was quantified to find the susceptibility difference between the two sides, with the assumption being made that the phase signal had been properly unwrapped as expressed in the following equation:

{r05mr14e01}
where {Delta}{phi} is the difference in the phase of the signal between the two compartments, f0 is the transmit frequency of the MR imager, TE is the gradient echo time, {theta} is the angle between B0 and the normal direction of the interface, Shf is the hyperfine contact shift, and {Delta}{chi} is the susceptibility difference between the two sides. If {Delta}{phi} is measured for more than two {theta} values, {Delta}{chi} and Shf can be determined from Equation (1). If Shf is known, {Delta}{chi} can be determined with a least squares fit. The Système International d’Unités unit was used to quantify magnetic susceptibility.

Although it is desirable to quantify the susceptibility of the intact cardiac tissue, our initial experience showed that this was difficult because of the tissue heterogeneity, presumably caused by fibrosis, that is known to exist in the iron-loaded gerbil heart (15). From the tissue lysate measurement, it was feasible to obtain the average magnetic susceptibility of the whole heart. After T2 measurements, the cardiac tissue lysate was prepared. The heart was cut into small pieces, and 2 mL of distilled deionized water were added. Tissue lysate was obtained by using an ultrasonic cell disruptor (Sonifier Cell Disruptor 200; Branson Ultrasonic, Danbury, Conn). The lysate was used to fill custom-made disk-shaped containers consisting of two thin and flat walls made of plastic. The diameter of the disk was 20 mm and the thickness was 4 mm, allowing the container to accommodate a total lysate volume of 1.27 mL. One of the two flat plastic walls extended beyond the container and was positioned on a holder to keep the container in place. The container and the holder were then submerged in water in a beaker, and the whole system was imaged at room temperature by using a knee coil. A multisection two-dimensional fast low-angle shot gradient-echo sequence was used to acquire the MR images with the following parameters: TR msec/TE msec, 400/30; seven sections, each with a thickness of 3 mm; field of view, 100 mm; matrix size, 256 x 256; and number of signals acquired, 10. Two measurements were performed, with {theta} values of approximately 0° and 90°, respectively. Data analysis was performed by using a custom-written software program (IDL; Research Systems). The susceptibility difference between tissue lysate and water was obtained by using Equation (1). The analysis was complicated by the presence of Shf in Equation (1). If we assume that the iron in all hearts has similar properties, then we expect Shf to be a constant, independent of the iron concentration. The value of Shf, however, was unknown initially, and there would be errors in the determination of Shf for individual specimens. To address this issue, a two-pass data analysis strategy was used. In the first pass, Shf was treated as unknown, and both Shf and {Delta}{chi} were determined from two different values of the pair ({theta}, {Delta}{phi}) by using Equation (1). Based on the results of the first pass, the mean value of Shf was obtained. In the second pass, Shf was assumed to have a fixed value, the mean value, and {Delta}{chi} was determined again from two different values of the pair ({theta}, {Delta}{phi}) by using Equation (1). Because water was added to the intact tissue to obtain the lysate, a correction factor was needed to obtain the average magnetic susceptibility of the whole heart. Because water was used as reference, a volume correction was sufficient and was performed as follows:

{r05mr14e02}
where Vti is the volume of the intact tissue used for lysis and Vlys is the resulting lysate volume. The volume ratio was calculated with the following equations:

{r05mr14e03}
and

{r05mr14e04}
where Wti is the weight of the heart measured immediately after sacrifice; {rho} is the density of the tissue, assumed to be 1.06 cm3/g; and Vwa is the added water volume, which is 2 mL. All magnetic susceptibility measurements were done by one author (Z.J.W.).

Tissue Iron Concentration
After all MR measurements had been made, the tissue specimens were subjected to chemical analysis to determine the level of elemental iron. Each tissue lysate specimen was digested in accordance with a protocol provided by an independent bioanalytic laboratory (National Medical Services, Willow Grove, Pa). The tissue lysate recovered after the MR measurement was first weighed and then digested in 2 mL concentrated (16N) trace metal–free nitric acid (Fisher Scientific, Pittsburgh, Pa). After the lysate was dissolved, 0.5 mL of concentrated hydrogen peroxide was added, and the sample was diluted to a 15-mL volume with deionized water. The solution was transferred to acid-washed bottles that were then carefully sealed, with the tissue weight and volume of the solution recorded. Two authors (L.L., Q.C.) performed the digestion procedure and recorded the data. The samples were then shipped to National Medical Services, where the iron level of the solutions was determined by using inductively coupled plasma atomic emission spectroscopy. The cardiac iron concentration for each gerbil heart was calculated as follows:

{r05mr14e05}
where Feti is the cardiac tissue iron concentration (in milligrams of iron per gram of wet tissue), Vsol is the volume (in milliliters) of the digested solution submitted to inductively coupled plasma atomic emission spectroscopy for iron level chemical analysis, Fesol is the iron concentration (in milligrams of iron per milliliter) at chemical analysis in the acid-digested solution, Wti is the weight of the newly harvested heart as specified in Equation (3), and Flys is the fraction of the original lysate that was recovered for chemical analysis after MR measurement.

Statistical Analysis
Pearson correlation coefficients were calculated for comparisons between the 1/T2 measurement and iron concentration, between susceptibility measurement and iron concentration, and between 1/T2 measurement and susceptibility. Simple linear regression models were fitted by using iron concentration as the outcome and the 1/T2 measurement or susceptibility measurement as a predictor. A power analysis was conducted to estimate the sample size required to detect a significant correlation.

Multiple linear regression analyses also were performed by using iron concentration as the outcome and both 1/T2 and susceptibility as predictors. A statistically significant difference was defined by P < .05. Analyses were performed by using statistical software (Stata release 7.0 for Windows; Stata, College Station, Tex). All statistical analyses were done by one author (H.Z.).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
The tissue iron concentration was determined from 27 gerbil hearts, with attrition of the animal or data on nine occasions. The highest average cardiac tissue iron concentration achieved was 1.95 mg Fe/g wet tissue. In vitro measurement of T2 of the whole heart was completed in all 27 gerbils. Magnetic susceptibility data were successfully acquired for 25 of the 27 gerbils. Overheating occurred in two specimens during lysis, and the resulting lysates were not uniform suspensions. These two samples still went through subsequent chemical analysis for iron level, but the MR susceptibility measurements were excluded from analysis. Power analysis showed that when the sample size is 10, the linear regression test for the Pearson correlation coefficient (null hypothesis, r = 0) for one normally distributed covariate will have 99% power to detect a correlation of 0.90 with a two-sided significance level of {alpha} = .01. A previous study in a mouse model by Liu et al (14) showed that iron content and T2 relaxivity in liver, heart, and peripheral muscle were linearly correlated (r = 0.96). On the basis of these data, we concluded that our sample size provided sufficient statistical power.

1/T2 Measurement and Correlation with Tissue Iron Concentration
The transverse decay of the signal is single exponential (Fig 1). Table 1 summarizes the results of linear regression analyses of 1/T2 versus the tissue iron concentration. The 1/T2 value was not sensitive to the TR used for the measurement. Figure 2 shows the correlation between 1/T2 (TR = 700 msec) and the cardiac tissue iron concentration.



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Curves for signal intensity versus TE from ex vivo MR imaging in two gerbil hearts with iron concentrations of 0.28 ({square}) and 1.72 ({triangleup}) mg Fe/g wet tissue. TR was 700 msec in both cases. Lines were fit with single-exponential decay functions.

 

View this table:
[in this window]
[in a new window]

  		TABLE 1. Simple Regression Analysis of MR Measurements versus Tissue Iron Concentration

 


View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Scatterplot shows correlation between 1/T2 (TR = 700 msec) and heart iron concentration. Solid line is linear regression line given by –0.834 + 0.0344/T2 (Pearson correlation, r = 0.92, P < .001).

 
Susceptibility Measurement and Correlation with Tissue Iron Concentration
Figure 3 shows a phase map obtained during the magnetic susceptibility measurement. To determine whether the method showed a significant susceptibility difference when there was none, water was used both inside and outside the container. It was confirmed that there was no quantifiable susceptibility difference between water in one compartment and that in the other compartment.



View larger version (82K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Phase map from MR measurements in tissue lysate from two gerbils, with background water used as reference. B0 is along the vertical direction. Phase jumps at the boundaries (arrows) are quantified for magnetic susceptibility measurement.

 
After the analysis of the first-pass data, the mean value of Shf was determined to be –0.133 ± 0.121 (standard deviation). There was a poor correlation between Shf and {Delta}{chi} in the 25 samples studied (r = 0.15). With use of a fixed Shf of –0.133, {Delta}{chi} was recalculated in the second-pass data analysis. Correlations between tissue iron concentration and MR imaging–determined magnetic susceptibility for the 25 gerbils are shown in Figure 4. The result of the regression analysis is included in Table 1.



View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Scatterplot shows correlation between {Delta}{chi} (difference in magnetic susceptibility of cardiac tissue and water) and tissue iron concentration. Solid line is linear regression line given by 0.343 + 6.25 x 105 · {Delta}{chi} (Pearson correlation, r = 0.90, P < .001).

 
Multiple Regression Analysis with Both 1/T2 and Susceptibility Used as Predictors of Iron Level
When the 1/T2 and susceptibility data were combined for assessment of their value in predicting tissue iron concentration, the outcome was better than that with any single measurement alone. The results are summarized in Table 2, which shows decreased {epsilon}RMS values for the linear predictions and slightly increased r values compared with those in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Multiple Regression Analysis of Transverse Relaxation Rate and Magnetic Susceptibility versus Tissue Iron Concentration

 
Correlation between 1/T2 and Susceptibility
Regression analysis was performed also to study the correlation between average 1/T2 measured with the three TR values and {Delta}{chi} (Fig 5). The Pearson correlation coefficient r was 0.90 (P < .001). The linear fit is given by 36.6 + 1.55 x 107 · {Delta}{chi}.



View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Scatterplot shows correlation between 1/T2 (average for the three TRs) and {Delta}{chi} (difference in magnetic susceptibility between cardiac tissue and water). Solid line is linear regression line given by 36.6 + 1.55 x 107 · {Delta}{chi} (Pearson correlation, r = 0.90, P < .001).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
REFERENCES
 
Investigators in many MR imaging–based studies of cardiac iron overload have measured T2-weighted signal changes (13,20) or quantified T2 (9,10). However, unlike those who have studied iron overload in liver, in which MR imaging measurements could be calibrated by using the results of liver biopsy, investigators in studies of cardiac iron overload do not have patients’ biopsy results available for the calibration of in vivo MR measurements. In these circumstances, animal studies play an important role. There have been very few studies of this type. An early study in a guinea pig model of iron overload (21) demonstrated significantly increased iron level in the heart and a decrease in T1, but no quantitative relationship was reported. A study of a congenital heterozygous thalassemic mouse model also has been reported (14). However, that study included only a few data points for the heart, and data for several tissue types were grouped together for statistical analysis. To our knowledge, there has not been a previous study of the magnetic susceptibility of iron-overloaded cardiac tissue.

Although 1/T2 and magnetic susceptibility are both sensitive methods for measuring the tissue iron concentration, tissue iron level determination with chemical analysis is considered the reference standard. When the measurement results are compared with those from this standard, chemical analysis has zero variance, and MR imaging measurements have finite variance. Therefore, MR imaging–measured magnetic susceptibility and 1/T2 will not be equivalent to the measurements at chemical analysis. The aim of this study was to establish whether MR imaging measurements are correlated with those of chemical analysis. In addition, there are potential covariates, such as age, sex, and body mass index, that may affect the results. These factors were not considered in the statistical analysis, as was justified because of the small sample size and the relative homogeneity of the study population.

Our results show a strong linear correlation between 1/T2 and the cardiac tissue iron concentration. It is known that the tissue 1/T2 in iron overload depends on magnetic field strength and temperature. Therefore, we conducted this study at 37°C and at the field strength most commonly used for clinical patient studies. Because tissue water T2 decreases over time after resection (18), we kept the time between euthanasia and the MR imaging study short to minimize tissue changes that may affect the T2 value. Furthermore, the gerbil hearts were tightly wrapped and sealed in small plastic containers after harvest and during the MR measurements, to further prevent excessive loss of tissue water and resultant change in T2 (22). In spite of these measures, it is possible that T2 values declined slightly over the course of MR imaging. This would help to explain why the MR data acquired with a TR of 700 msec had the strongest correlation with iron level, because these data were acquired first.

In clinical cardiac iron overload, the iron distribution is not uniform (2). The highest iron level is found in the epicardium. The lowest iron level is found in the endocardium, which has an iron concentration about one-third of that in the epicardium. It is not clear whether this variance in spatial distribution is present in the gerbil heart. However, the appearance of single-exponential decays in signal intensity versus TE data suggests that the iron distribution in the gerbil heart may not be very heterogeneous. In this study, the highest average iron level found in the gerbil heart is comparable to that found in human epicardium after autopsy (6).

In some studies, measurement of liver iron concentration with a SQUID has demonstrated excellent agreement with findings at liver biopsy (23). In a SQUID used for measuring the liver iron content by means of magnetic susceptibility measurement, one set of coils generates a magnetic field that is primarily localized in the liver and induces liver magnetization. A second set of coils coupled to the SQUID detects and measures the magnetic field created by liver magnetization. This technique may be used to directly measure the magnetic susceptibility of the tissue. The iron cores in ferritin and hemosiderin have the same chemical structure (similar to that in ferrihydrite) and the same magnetic susceptibility. Therefore, when the tissue magnetic susceptibility is dominated by contributions from ferritin and hemosiderin, measurement of magnetic susceptibility can be considered a direct measurement of iron concentration.

Our results show that the magnetic susceptibility of the cardiac tissue increases linearly with the tissue iron concentration. From the relationship between {Delta}{chi} and tissue iron concentration, we calculated the mass magnetic susceptibility of the iron in the gerbil heart to be 1.51 per kilogram of iron at room temperature (22°C); at body temperature, according to the Curie law, the magnetic susceptibility is decreased to 1.44 per kilogram, a value that is consistent with the mass magnetic susceptibility value reported for ferritin iron in rabbit liver at 25°C, which is 1.46 per kilogram (1.40 per kilogram at body temperature) (24). Therefore, the results of our MR susceptibility study have confirmed that the magnetic susceptibility of the iron in the heart is similar to that in the liver. The agreement between measured values demonstrates that the tissue magnetic susceptibility can be treated as a direct measure of the iron. Once the magnetic susceptibility is reliably measured at any practical magnetic field strength, the value can be directly converted into tissue iron concentration, without tissue-specific calibration. Use of a tissue lysate was required in this study partly because of the high spatial resolution needed for MR imaging of the gerbil heart because the cardiac wall is only 1–2 mm thick. Tissue heterogeneity became a problem with pixel sizes in microns when we attempted to reveal field variation across the gerbil ventricular wall. Much bigger pixel sizes may be used in human studies, and each pixel will be more likely to include both fibrosis and normal tissue. Therefore, our approach is likely to be applicable in clinical studies.

Errors involved in these studies are estimated as follows. The coefficient of variation of inductively coupled plasma atomic emission spectroscopy measurement was approximately 5%. The standard error in the susceptibility measurement was typically 5 x 10–8 because of the uncertainty in {Delta}{phi} in Equation (1). There were also errors introduced by the sample handling and water evaporation during the experimental procedures. These errors will cause errors in {rho} in Equation (3) and in the lysate volume in Equation (2). In Equation (2), errors could cause a 5% coefficient of variation in the susceptibility difference and in the tissue iron level determination. Even when all of these errors are taken into account, the data still contain significant variations that may presumably be due to biologic variations among the samples.

Although we hope to use 1/T2 and magnetic susceptibility measurements at MR imaging for noninvasive estimation of tissue iron levels, these measurements are not equivalent to the chemical analysis of tissue iron content. From our study, much can be learned about the reliability and sensitivity of using 1/T2 to estimate the cardiac iron level. However, it is not expected that the numeric relationship between 1/T2 and iron concentration obtained from this ex vivo study will be directly applicable to patients, because the relaxation rate also depends on other tissue properties and the morphology of the iron, especially whether the iron is incorporated in ferritin or hemosiderin. Furthermore, T2 depends on tissue oxygenation level and on microcirculation, which are drastically altered ex vivo. The magnetic susceptibility of the cardiac tissue was measured from tissue lysate because the intact tissue was very heterogeneous, presumably because of fibrosis. It is possible that we may have a similar problem in patient studies, even though larger voxel sizes in patient examinations may alleviate the effects of tissue heterogeneity caused by fibrosis. The current study did not address the main difficulties in extending our approach to patient studies: cardiac motion and identification of a reference tissue. Finally, we did not analyze the effects of other potential factors that may have affected the results, including age, sex, and body mass index.

Practical applications.—Using a gerbil model, we demonstrated a strong linear correlation between 1/T2 and iron level in the iron-overloaded heart and confirmed that the magnetic susceptibility of cardiac tissue directly reflects the iron level in that tissue. The combination of 1/T2 with susceptibility data yields a better estimation than does either measurement alone. Evaluation of the cardiac iron level noninvasively with MR imaging will have an important effect on patient care because cardiac iron overload can be reversed through chelation (25,26). Measurement of cardiac iron is especially important and may be life saving for patients in whom liver iron levels are only slightly or moderately increased and give no indication of the potential risk of iron-induced cardiac disease. In addition, if MR imaging can be used to assess cardiac iron levels, it also can be used to evaluate the response to chelation therapy (13).


   FOOTNOTES
 
Abbreviations: SQUID = superconducting quantum interference device, TE = echo time, TR = repetition time

See also Science to Practice in this issue.

Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, Z.J.W.; study concepts, Z.J.W., A.R.C.; study design, Z.J.W., L.L., H.Z., T.A., A.R.C.; literature research, Z.J.W., A.R.C.; experimental studies, Z.J.W., L.L., Q.C., T.A.; data acquisition, Z.J.W., L.L., Q.C.; data analysis/interpretation, Z.J.W.; statistical analysis, H.Z.; manuscript preparation, Z.J.W., H.Z.; manuscript definition of intellectual content, Z.J.W., A.R.C.; manuscript editing, A.R.C.; manuscript revision/review, L.L., Q.C., H.Z., T.A., A.R.C.; manuscript final version approval, all authors


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 

  1. Engle MA, Erlandson M, Smith CH. Late cardiac complications of chronic, severe, refractory anemia with hemochromatosis. Circulation 1964; 30:698-705.[Abstract/Free Full Text]
  2. Buja LM, Roberts WC. Iron in the heart: etiology and clinical significance. Am J Med 1971; 51:209-221.[CrossRef][Medline]
  3. Olivieri NF, Brittenham GM. Iron-chelating therapy and the treatment of thalassemia. Blood 1997; 89:739-761.[Free Full Text]
  4. Angelucci E, Brittenham GM, McLaren CE, et al. Hepatic iron concentration and total body iron stores in thalassemia major. N Engl J Med 2000; 343:327-331.[Abstract/Free Full Text]
  5. Brittenham GM, Farrell DE, Harris JW, et al. Magnetic-susceptibility measurement of human iron stores. N Engl J Med 1982; 307:1671-1675.[Abstract]
  6. Sanyal SK, Johnson W, Jayalakshmamma B, Green AA. Fatal "iron heart" in an adolescent: biochemical and ultrastructural aspects of the heart. Pediatrics 1975; 55:336-341.[Abstract/Free Full Text]
  7. Fitchett DH, Coltart DJ, Littler WA, et al. Cardiac involvement in secondary haemochromatosis: a catheter biopsy study and analysis of myocardium. Cardiovasc Res 1980; 14:719-724.[Medline]
  8. Brittenham GM, Badman DG. Noninvasive measurement of iron: report of an NIDDK workshop. Blood 2003; 101:15-19.[Abstract/Free Full Text]
  9. Mavrogeni SI, Gotsis ED, Markussis V, et al. T2 relaxation time study of iron overload in b-thalassemia. MAGMA 1998; 6:7-12.
  10. Papanikolaou N, Ghiatas A, Kattamis A, Ladis C, Kritikos N, Kattamis C. Non-invasive myocardial iron assessment in thalassaemic patients: T2 relaxometry and magnetization transfer ratio measurements. Acta Radiol 2000; 41:348-351.[CrossRef][Medline]
  11. Jensen PD, Jensen FT, Christensen T, Heickendorff L, Jensen LG, Ellegaard J. Indirect evidence for the potential ability of magnetic resonance imaging to evaluate the myocardial iron content in patients with transfusional iron overload. MAGMA 2001; 12:153-166.
  12. Anderson LJ, Holden S, Davis B, et al. Cardiovascular T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload. Eur Heart J 2001; 22:2171-2179.[Abstract/Free Full Text]
  13. Jensen PD, Jensen FT, Christensen T, Eiskjaer H, Baandrup U, Nielsen JL. Evaluation of myocardial iron by magnetic resonance imaging during iron chelation therapy with deferrioxamine: indication of close relation between myocardial iron content and chelatable iron pool. Blood 2003; 101:4632-4639.[Abstract/Free Full Text]
  14. Liu P, Henkelman M, Joshi J, et al. Quantification of cardiac and tissue iron by nuclear magnetic resonance relaxometry in a novel murine thalassemia–cardiac iron overload model. Can J Cardiol 1996; 12:155-164.[Medline]
  15. Carthew P, Dorman BM, Edwards RE, Francis JE, Smith AG. A unique rodent model for both the cardiotoxic and hepatotoxic effects of prolonged iron overload. Lab Invest 1993; 69:217-222.[Medline]
  16. Hershko C, Link G, Konijn AM, Huerta M, Rosenmann E, Reinus C. The iron-loaded gerbil model revisited: effects of deferoxamine and deferiprone treatment. J Lab Clin Med 2002; 139:50-58.[CrossRef][Medline]
  17. Kuryshev YA, Brittenham GM, Fujioka H, et al. Decreased sodium and increased transient outward potassium currents in iron-loaded cardiac myocytes: implications for the arrhythmogenesis of human siderotic heart disease. Circulation 1999; 100:675-683.[Abstract/Free Full Text]
  18. Thickman DI, Kundel HL, Wolf G. Nuclear magnetic resonance characteristics of fresh and fixed tissue: the effect of elapsed time. Radiology 1983; 148:183-185.[Abstract/Free Full Text]
  19. Wang ZJ, Li S, Haselgrove JC. Magnetic resonance imaging measurement of volume magnetic susceptibility using a boundary condition. J Magn Reson 1999; 140:477-481.[CrossRef][Medline]
  20. Johnston DL, Rice L, Vick GW, 3rd, Hedrick TD, Rokey R. Assessment of tissue iron overload by nuclear magnetic resonance imaging. Am J Med 1989; 87:40-47.[Medline]
  21. Schwartz KA, Fisher J, Adams ET. Morphologic investigations of the guinea pig model of iron overload. Toxicol Pathol 1993; 21:311-320.[Medline]
  22. Liu P, Johnston DL, Brady TJ, Lutrario DM, Okada RD. The alterations of magnetic resonance relaxation parameters in excised myocardial tissue during NMR spectroscopy: the effects of time, environmental exposure and TTC staining. Magn Reson Imaging 1989; 7:109-113.[CrossRef][Medline]
  23. Brittenham GM, Sheth S, Allen CJ, Farrell DE. Noninvasive methods for quantitative assessment of transfusional iron overload in sickle cell disease. Semin Hematol 2001; 38(suppl 1):37-56.
  24. Bauman JH, Harris JW. Estimation of hepatic iron stores by in vivo measurement of magnetic susceptibility. J Lab Clin Med 1967; 70:246-257.[Medline]
  25. Rahko PS, Salerni R, Uretsky BF. Successful reversal by chelation therapy of congestive cardiomyopathy due to iron overload. J Am Coll Cardiol 1986; 8:436-440.[Abstract]
  26. Wacker P, Halperin DS, Balmer-Ruedin D, Oberhansli I, Wyss M. Regression of cardiac insufficiency after ambulatory intravenous deferoxamine in thalassemia major. Chest 1993; 103:1276-1278.[Abstract/Free Full Text]

Related Article

Can MR Imaging Provide a Noninvasive "Biopsy" of the Heart to Measure Iron Levels?
David A. Bluemke and Robert P. Liddell
Radiology 2005 234: 647-648. [Full Text] [PDF]



This article has been cited by other articles:


Home page
 BloodHome page
L. J. Noetzli, S. M. Carson, A. S. Nord, T. D. Coates, and J. C. Wood
Longitudinal analysis of heart and liver iron in thalassemia major
Blood, October 1, 2008; 112(7): 2973 - 2978.
[Abstract] [Full Text] [PDF]

Home page
 CirculationHome page
J. C. Wood, M. Otto-Duessel, M. Aguilar, H. Nick, M. D. Nelson, T. D. Coates, H. Pollack, and R. Moats
Cardiac Iron Determines Cardiac T2*, T2, and T1 in the Gerbil Model of Iron Cardiomyopathy
Circulation, July 26, 2005; 112(4): 535 - 543.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
D. A. Bluemke and R. P. Liddell
Can MR Imaging Provide a Noninvasive "Biopsy" of the Heart to Measure Iron Levels?
Radiology, March 1, 2005; 234(3): 647 - 648.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, Z. J.
Right arrow Articles by Cohen, A. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, Z. J.
Right arrow Articles by Cohen, A. R.
Related Collections
Right arrowRelated Article

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE