Introduction

Ferumoxytol, a superparamagnetic iron-oxide, is increasingly used for MR angiography due to its favorable MR properties and long intravascular half-life.^(1,2) Moreover, it has drawn increasing interest due to growing concerns over gadolinium-based contrast agent safety.^3,4 Brand name agents have historically been more expensive than generic agents, which may present an opportunity to reduce cost if there are no clinically relevant differences in MR properties.
Therefore, the purpose of this study was to compare the relaxivity and MR imaging performance of brand name ferumoxytol (Feraheme®, AMAG Pharmaceuticals) with generic ferumoxytol (Sandoz) through phantom and in-vivo experiments. The secondary purpose was to determine optimal flip angle and weight-based dose.

Methods

For all experiments both Feraheme®, AMAG Pharmaceuticals, Waltham, MA, USA and ferumoxytol, Sandoz Inc. Princeton, NJ, USA were used.

Phantom experiment
Each agent was diluted in 0.9% saline and also adult bovine whole blood at 5 concentrations (0.27; 0.54; 1.07; 1.61 and 2.15mM) within 30 ml vials, for a total of 20 vials. Vials were placed in an MR compatible water bath at 37°C (Fig 1) and imaged at 1.5T and 3.0T MRI systems (1.5T Artist, 3.0T Premier, GE Healthcare, Waukesha, WI) using a 32-channel phase array torso coil. The protocol consisted of coronal acquisitions 2D fast spin echo inversion recovery with multiple inversion times to quantify R1, 2D spin echo with multiple echoes to quantify R2, and 3D multi-echo spoiled gradient echo chemical shift-encoded imaging to quantify R2* (Fig 2). Relaxation rate constants, R1, R2, R2* were measured in a manually drawn region of interest (ROI) centrally located in each vial in three slices using Matlab. Regression lines were fit to estimate r1, r2, and r2* relaxivity values using first order polynomial fitting for saline and second order fitting for blood. Differences in relaxivity between the agents were statistically tested with Chi-square with 2 degrees of freedom for saline and 3 degrees of freedom for blood.

In-vivo study
Six healthy volunteers were imaged after IRB approval and informed written consent. All subjects were imaged twice using a cross-over design, in which one agent was administered, followed by a 1 month washout⁵, and then a second exam with the other agent, all in random order. Cumulative doses of 1, 3, 5 and 7 mg/kg diluted ferumoxytol were administered during the exam. MR angiography of the chest was performed using a sagittal MRA acquisition on a 3.0T clinical MRI system (Premier, GE Healthcare, Waukesha, WI) with a 32-channel phased-array body coil before and after each contrast dose administration. For each incremental dose, the flip angle was varied from 10 to 40 degrees in 10-degree increments over 5 breath-holds (Fig 2). No parallel imaging was used to facilitate ROI-based signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) measurements, including correction for noise rectification.⁶ For each MRA acquisition, ROIs were drawn in the aortic arch (1 cm2), the paraspinous muscles (1 cm2) and a noisy area outside of the patient free from obvious artifact (7.0 cm2), by a radiologist with 6 years of experience. Absolute SNR and CNR (aorta-muscle) values were calculated. The optimal flip angle was assessed using conglomerate data from all volunteer exams. Finally, the mean SNR and CNR of both ferumoxytol agents were compared at the optimal flip angle for each cumulative dose and plotted with corresponding 95% confidence intervals.

Results

Phantom experiment
The dependence of R1, R2 and R2* on ferumoxytol concentrations was linear in saline and non-linear in blood (Fig 3). There were no statistically significant differences between most relaxivities (Fig 4), and those with statistical significance were small and of doubtful clinical relevance (Fig 3).

In-vivo study
6 healthy volunteers (3 female, 3 male) were included with an age range of 25-58 years and a BMI range of 19.6-25.9 kg/m2. A flip angle of 20° exhibited the optimal SNR and CNR (Fig 5). There were no significant differences in mean SNR and CNR between agents at the optimal flip angle of 20° (Fig 5). Optimal SNR and CNR for both agents were reached at a dose of 5mg/kg.

Discussion and Limitations

Unfortunately, human blood was not available for the phantom experiments, but adult bovine whole blood has fairly similar properties. The small differences in relaxivity values observed between agents was very small, and likely due to small errors in the phantom preparation.

Conclusion

No clinically relevant differences were identified between brand name ferumoxytol (Feraheme®, AMAG Pharmaceuticals) and generic ferumoxytol (Sandoz) for MRI relaxation properties and MR angiography image quality. For both agents the optimal flip angle was 20° and optimal dose was 5mg/kg.

Acknowledgements

We wish to acknowledge support from GE Healthcare and Bracco Diagnostics who provide research support to the University of Wisconsin. Dr. Reeder is the John. H Juhl Endowed Chair of Radiology. This project was supported by the Departments of Radiology and Medical Physics, University of Wisconsin.