Ferumoxytol, an ultra-small-superparamagnetic-iron-oxide (USPIO), has been increasingly used as a blood pool contrast agent due to its high T1 and T2 relaxivities and for its strong susceptibility properties. The magnetization of Ferumoxytol, however, becomes saturated at field strengths above 1T introducing a perceived inverse correlation between field and apparent susceptibility. The goal of this study was to quantify the molar susceptibility of Ferumoxytol and Gadolinium at 1.5T, 3T and 7T.Magnetic susceptibility “χ” is a quantitative measure of a material’s tendency to interact with and distort an applied magnetic field (1,2). The magnetization ($$$\overrightarrow{M}$$$) is usually determined by $$$\overrightarrow{M}=χ\overrightarrow{H}$$$, where $$$\overrightarrow{M}$$$ is the main field strength. If the material saturates then $$$\overrightarrow{M}$$$ will remain constant above that field strength (i.e., above 1T for Ferumoxytol (3)) and the change in local field variations with field will be invariant.

Measurement of susceptibility by QSM:

Susceptibility values were measured using quantitative susceptibility mapping (QSM) at 1.5T, 3.0T, and 7T MR and by fitting the difference of frequency during rotations of a cylinder. Two phantoms were made with 5 iron concentrations ranging from [0.11- 0.93] mmol.l^‑1 and for Gd-DTPA from [0.05- 0.5 %]. Blood samples were analyzed before and after systemic injection at 4mgFe/kg of Ferumoxytol using QSM. Gradient echo images (TR/TE/ΔTE/FA/BW: 50/5/5.4/20/450; resolution: 0.5x0.5x2mm³) were acquired at 1.5T, 3T and 7T in order to quantify susceptibility.

QSM processing:

Phase images were unwrapped using a three-dimensional best path phase unwrapping algorithm (4) followed by background field removal using SHARP (5) to estimate the local magnetic field containing the local susceptibility sources. The QSM map was computed by solving the inverse field to source problem using SWIM (6).

Susceptibility measurement by tube rotation:

To validate the QSM measurements, magnetic susceptibility was also measured directly from the frequency changes induced in a long cylinder as it was rotated with respect to the applied field. Three 25mL pipettes (28cm long, 1.4cm inner diameter) were filled with [Fe]=0.35, 0.64 and 1.68 mmol.l^-1 and placed sequentially in a circular water bath at 7 orientations relative to B0 between 0° and 90° and imaged with a 3D GRE sequence, 1mm isotropic resolution and 7 monopolar echoes (TE = 3 – 24.6ms). The frequency difference between the inside of the tube (from the lumen) and the water bath was measured for each orientation and fitted to the equation:

$$Δf = \frac {1}{2}.f0.Δχ.(\cos ^2(θ)-\frac {1}{3})$$

where f₀ is the Larmor frequency in Hz and θ the angle relative to the main field B0. After the susceptibilities were measured, the molar susceptibility was calculated with a linear fit of the susceptibility in ppb and the [Fe] in mmol.l^-1 (ie. mM).

A representative example of susceptibility maps for a tube filled with Feraheme (middle range, [Fe]=0.49 mmol.l-1) at increasing static field was displayed with the same grayscale on Fig.1 with average values of the iron oxide 1015, 749 and 261 ppb at 1.5T, 3T and 7T. The measurements of the susceptibility values were repeated 3 times on different acquisitions for 5 dilutions and results were plotted against iron concentration (Fig. 2). The χmolar displayed an inverse correlation with field with values of 2089(ε²=0.99), 1109(ε²=0.95) and 526(ε²=0.98) ppb. mmol^-1.l at 1.5T, 3T and 7T respectively. Measurement of magnetic susceptibility by tube rotation at 1.5T and 3T yielded values of χ_molar = 2464 and 1051 ppb.mmol^-1.l respectively, in good agreement with the QSM results(Fig. 3). No field dependence was observed for the susceptibility of GdDTPA (Fig. 4). The susceptibility of blood alone was independent of field strength, and equal to 678 ppb (± 36ppb), whereas the susceptibility of blood containing ferumoxytol was measured to be 4662, 2660 and 886 ppb at 1.5, 3 and 7T, respectively (Fig. 5).We have shown that the magnetization of ferumoxytol saturates above 1.5T giving a constant shift in local field independent of the static magnetic field. Our susceptibility measurements made with two independent methods and were in good agreement with each other and with the value of 1675ppm.mmol^-1.l previously reported at 3T (7). By contrast, the susceptibility of blood alone was found to be independent of field, with a value of 678 ppb (±36ppb). Although it is well documented that the magnetization of USPIO materials saturates below 1T, researchers often fail to consider this when deciding on the best field strength at which to perform MRI exams involving USPIO agents. Furthermore, when quoting susceptibility for USPIO materials, it should be remembered that, if the usual expression for magnetization is used, it will appear as if susceptibility decreases with increasing field.