Introduction

Excess accumulation of iron in hereditary hemochromatosis and transfusional hemosiderosis is an established risk factor for cirrhosis, hepatocellular carcinoma, liver failure, and heart failure among other complications.¹ Diagnosis and treatment of iron overload is often evaluated through measurement of liver iron concentration (LIC). MR-based R2 (=1/T2) mapping methods have been shown to have good correlation with LIC,² but the current standard, the spin-echo sequence, is prone to lengthy acquisition times, limited coverage of the liver, and respiratory artifacts.

Preliminary work on a novel phase-based R2 mapping method has shown promise to quantify whole-liver R2 within a single breath-hold.^3,4 It has been shown that the signal phase (θ) of an RF phase modulated 3D gradient echo (GRE) decreases monotonically with increasing R2 through the use of small RF phase increments. Using a Bloch equation simulation lookup table (requiring approximate knowledge of the T1), R2 can be estimated from the signal phase.

A modified version of the phase-based R2 method for high R2 mapping with a short TR has been presented.³ However, clinical demonstration of this technique in high R2 tissues seen in liver iron overload has not been shown. Therefore, the purpose of this work is to optimize and demonstrate the feasibility of the phase-based R2 method in liver iron overload patients.

Methods

Phantom Acquisition: Eleven vials of varying MnCl2 concentration (0.3 – 3.7 mM) in agarose gel (2% w/v) were constructed to achieve high R2 values, nominally 40 – 400 s^-1 (Figure 1a). Vials were imaged on a 3.0T MR system (Signa Premier, GE Healthcare, Waukesha, WI) using a standard head coil (AIR Coil, 48 channel, GE Healthcare). Reference R2 and T1 values were found using single-echo spin-echo (SE) and inversion-recovered spin-echo (SE-IR) acquisitions, respectively. Acquisition parameters are shown in Table 1.

Volunteer/Patient Enrollment: A volunteer/patient feasibility study was performed with approval from the local IRB and informed consent was obtained from all subjects. Five healthy volunteers were enrolled in the study. Two adults with known liver iron overload were also recruited. Inclusion criteria included patients with an estimated LIC between 5-15 mg/g.

Volunteer/Patient MR Acquisition: Liver imaging was performed on a 3.0T MR system (Signa Premier, GE Healthcare, Waukesha, WI) using a posterior and anterior receive array coil (AIR Coil, GE Healthcare) for the abdomen. 3D RF phase modulated GRE images of the whole-liver were acquired for phase-based R2 mapping. IDEAL IQ (GE Healthcare, Waukesha, WI) was acquired for R2* mapping.

Reference R2 values were determined using a single breath-hold, single-voxel, multi-TE STEAM-MRS⁵ on a 20x20x20 mm³ volume in the right lobe of the liver, avoiding blood vessels. Acquisition parameters are shown in Table 1.

Reconstruction/Analysis: An “average first, fit second” method³ was applied to avoid inaccurate R2 estimates due to negative phase measurements related to image noise at high iron concentrations. The average phase from an ROI in the liver was estimated, and then mapped to the corresponding R2 value from a lookup table.

Since the value of T1 (required for the lookup table) is unknown, a T1 correction³ was applied using an a priori relationship between T1 and T2. This relationship was calculated from a linear regression of SE T2 as a function of SE-IR T1 for the phantom and using results from a recent multi-center liver iron study⁶ for in vivo imaging.

Bland-Altman analysis was performed to compare phase-based R2 estimates with reference spin echo R2 estimates in phantoms and reference STEAM-MRS R2 estimates in volunteers/patients.

Results

Phantom Experiments: Phantom studies confirm that phase-based R2 quantification closely matches reference single-echo spin echo R2 measurements at high R2 values (Figure 1). Bland-Altman (Figure 1b) and linear regression (Figure 1c) analysis comparing phase-based R2 and reference single-echo SE R2 (Figure 1a) estimates is shown.

Healthy Volunteer Study: Images from the volunteer study demonstrate good agreement between the phase-based R2 estimate and the reference STEAM-MRS R2 for all five volunteers (Figure 2). An example volunteer R2 map is shown in Figure 2a. Results from Bland-Altman analysis comparing phase-based R2 and reference STEAM-MRS R2 is shown in Figure 2b.

Iron Overload Patient Study: ROI histogram analysis comparing signal phase in the iron overload patient and a healthy volunteer is shown (Figure 3). The histogram demonstrates the need for “average first, fit second” in iron overloaded tissue. Images from the iron overload patient study demonstrate good agreement between the phase-based R2 estimate and reference STEAM-MRS R2 for patient 1 (Figure 4). The MRS acquisition failed for patient 2.

Discussion and Conclusions

In this work, we have investigated the application and feasibility of the phase-based R2 mapping method for iron overload assessment in phantoms, healthy volunteers, and patients with known iron overload. Compared with standard spin echo based R2 mapping methods, phase-based R2 mapping has potential to quantify R2 in even severely iron overloaded patients with a significantly reduced acquisition time. This allows for whole-liver R2 quantification over a wide range of LIC within a single-breath hold. Limitations of this work include the limited patient population size and use of only 3.0T. Further studies in a larger cohort of patients with known or suspected iron overload at both 1.5T and 3.0T are planned.

Acknowledgements

The authors wish to acknowledge support from the NIH (R01-DK100651, UL1-TR00427, R01-DK117354), as well as GE Healthcare who provides research support to the University of Wisconsin. Ruvini Navaratna is supported by an NIH Metabolism and Nutrition Training Program (MANTP) T32 Fellowship (T32-DK007665). Finally, Dr. Reeder is a Romnes Faculty Fellow, and has received an award provided by the University of Wisconsin-Madison Office of the Vice Chancellor for Research and Graduate Education with funding from the Wisconsin Alumni Research Foundation.