Csanad Varallyay1, Daniel Schwartz2, Joao Prola Netto1, Prakash Ambady2, Andrea Horvath2, and Edward Neuwelt2
1Diagnostic Radiology and Neurology, Oregon Health and Science University, Portland, OR, United States, 2Neurology, Oregon Health and Science University, Portland, OR, United States
Synopsis
Steady state blood volume (SS-CBV) mapping using the blood
pool agent ferumoxytol as an MRI contrast agent is feasible in brain tumors and
other intracranial pathologies. It allows high resolution, distortion
free blood volume maps, which can be a useful MRI tool to improve diagnosis and assessment of response to therapy. Ferumoxytol dose and MRI sequences may be optimized for various clinical applications. Background and Purpose
Ferumoxytol, an iron oxide nanoparticle has been marketed as
Feraheme® with the indication of iron replacement in patients with kidney
failure. Off label use, as an MR imaging agent has been tested in various parts
of the body. Brain imaging is one of the most investigated areas so far with
the largest number of cases (~700) in our institution, including subjects with
brain tumors and various other intracranial pathologies1.
Based on our previous data, ferumoxytol enhanced MRI may
help describe the most vascular/malignant tumor areas for surgical planning; differentiation
of progression from treatment related pseudoprogression in primary brain tumors2;
advanced visualization of pathologies of vascular origin; visualization of
abnormal blood vessels in the tumor, to improve radiological diagnosis;
differentiation of extra-axial benign lesions (meningioma) from intraaxial
malignant neoplasms (high grade glioma, metastases); assessment of inflammatory
components of brain lesions by increased cellular iron uptake.
Due to its large particle size and long plasma half life of
14-21hours, ferumoxytol is considered a blood pool contrast agent early after
injection. It uniquely allows high resolution steady state blood volume
(SS-CBV) mapping.3 This technique is not feasible with gadolinium
based small molecular weight contrast agents.
The purpose of this study was to assess the best
acquisition technique and ferumoxytol dose to image the macro and
microvasculature.
Methods
50
subjects with brain tumors or other intracranial pathologies underwent brain imaging with multiple
doses of ferumoxytol injected during the MRI scanning. SS-CBV maps were
calculated using T2* weighted sequences before and after 75mg, 225mg, and 510mg
Fe doses in protocol eIRB# 1562 (n=43) and 1mg/kg, 3mg/kg and 7mg/kg
(not exceeding 510mg total) in eIRB# 9846 (n=7). This latter protocol is a
newly launched multicenter study, including 5 US centers aiming to investigate
ferumoxytol SS-CBV mapping at various stages of glioblastoma.
Subjects were
scanned using 3T Philips scanners. Two types of sequences were used: In both protocols
magnitude images of modified susceptibility weighted imaging (SWI) were used to
create SS-CBV maps, (3D acquisition, TR/TE/FA: 26/20/15, FOV:210x200mm2, acq
matrix 300x300, 2mm thickness, overlap 1mm, scan time: 214s), in protocol eIRB#
9846, multi echo fast field echo (mFFE) were also acquired (2D acquisition, TR
920ms, five echo times between 6.9 and 29 ms were acquired, FA:18, FOV 230x184mm2 acq matrix 384x307, 4mm thickness, 1mm gap, scan time:227s.) The 5 echoes were analyzed separately as well as averaged. Given the intravascular properties
of ferumoxytol, SS-CBV maps were created by calculating ΔR2* between
coregistered early post and pre contrast scans for each ferumoxytol dose using FSL. T2*
weighted images and SS-CBV maps were visually tested and analyzed.
Results
Both
SWI and mFFE scans were anatomically correct without major distortions, but both sequences were prone to motion artifacts. (Figure 1) Less
susceptibility artifacts were present on the mFFE scans with shorter echo
times. Both 3mg/kg and 7mg/kg Fe (225mg and 510mg in study eIRB#1562) doses
visualized the abnormal vasculature (Figure 1), and provided clinically sufficient SS-CBV maps. (Figure 2) However, when analyzing lesions within the cortex, higher
doses were beneficial. The high spatial resolution and thin slices of SWI
allowed good sagittal and coronal reformations, whereas the mFFE sequences
allowed calculation of SS-CBV maps with higher signal to noise ratios. Unlike
the SWI sequence, the mFFE had no substantial T1 properties.
Conclusion
SS-CBV
mapping is feasible using various sequences and various ferumoxytol doses. For
surgery planning, SWI magnitude images can be used, with good multi-plane
reformations. mFFE may be more beneficial if SS-CBV mapping is used after
gadolinium enhanced MRI as an add-on scan, since it lacks T1 weighting. In most
cases, lower ferumoxytol dose may be sufficient (3mg/kg), however if thin slices
are needed for cortical lesions, 7mg/kg (510mg total) may be beneficial.
This preliminary analysis of data allowed the design of a
larger multicenter study, which prospectively analyses SS-CBV during
glioblastoma treatment, to further exploit the potentials of this new MRI application.
Acknowledgements
No acknowledgement found.References
1. Neuwelt, E.A., C.G.
Varallyay, S. Manninger, et al. The
potential of ferumoxytol nanoparticle magnetic resonance imaging, perfusion,
and angiography in central nervous system malignancy: a pilot study.
Neurosurgery, 2007. 60(4): p. 601-11
2. Gahramanov, S., C.
Varallyay, R.M. Tyson, et al. Diagnosis
of pseudoprogression using MRI perfusion in patients with glioblastoma
multiforme may predict improved survival. CNS Oncol, 2014. 3(6): p. 389-400.
3. Varallyay, C.G., E.
Nesbit, R. Fu, et al. High-resolution
steady-state cerebral blood volume maps in patients with central nervous system
neoplasms using ferumoxytol, a superparamagnetic iron oxide nanoparticle. J
Cereb Blood Flow Metab, 2013. 33(5):
p. 780-6.