T1-weighted contrast-enhanced magnetic resonance angiography (CEMRA) is a clinically established approach for the diagnosis and management of vascular diseases¹. However, the spatial resolution of the images is limited by the short half-life of gadolinium contrast agents in the blood, which can be problematic for imaging of arteriovenous malformations (AVMs). Ultrasmall superparamagnetic iron oxide particles (USPIOs) have been shown to stay for longer times in the blood pool and have been used to study vascular networks in preclinical studies². These compounds are considered negative contrast materials due to their strong T2* effects, which offer great sensitivity to small vascular structures but suffer from blurring effects and susceptibility artifacts. Less appreciated are their intrinsic T1 shortening properties that can produce positive signal using appropriate pulse sequences^3-4. In this study, we probed the potential of USPIOs to provide high-resolution angiographic images of the human brain using both T2* and T1 contrasts. We also explored the influence of USPIOs concentration on these two contrast mechanisms.

With IRB approval under waived consent and in HIPAA compliant fashion, 10 paediatric patients suspected of AVMs were scanned pre- and post-USPIO injection as part of their clinical exam. Ferumoxytol (Feraheme; AMAG Pharmaceuticals Inc, Cambridge, MA), a USPIO compound developed for iron replacement therapy primarily in patients with chronic kidney disease, was used ‘off-label’ as the MRI contrast agent. All studies were conducted on a 3T GE scanner (MR750, GE Healthcare Systems, Waukesha, WI) equipped with an 8-channel head-coil. Two sequences were added to the regular clinical paediatric brain protocol and were acquired both before and after the intravenous injection of a single dose of ferumoxytol (0.1 mL Fe/kg):

-For T2* weighted CEMRA: a flow-compensated 3D GRAPPA-accelerated multi-echo GRE sequence was used (resolution = 0.57x0.86x2.5mm³, 66 z-partitions, acceleration factor = 2, 8 echoes ranging from TE = 4.3ms – 37.5ms with 4.7ms increments, TR = 40.8ms, scan time = 5:44min). Data were reconstructed using MATLAB (MathWorks Inc., Natick, MA, USA) code. In addition to the outputting the T2* weighted images of the individual echoes, the following contrasts were computed: R2* (1/T2*) maps (mono-exponential fit of echoes), Quantitative Susceptibility Maps (QSM) (using the MEDI algorithm⁵), and Susceptibility Weighted Images (SWI)⁶.

-For T1 weighted CEMRA: a flow-compensated 3D GRAPPA-accelerated GRE sequence was used (resolution = 0.5x0.5x1mm³, 328 z-partitions, acceleration factor = 2, TE = 1.04ms, TR = 4.34ms, FA=15deg, scan time = 5:00min).

Additionally, to test the influence of USPIOs concentration on both T1 and T2* contrasts, a healthy volunteer was scanned with 7 incrementally increasing doses of ferumoxytol (0 to 7ml Fe/kg). The craniocaudal spatial coverage of the T2*- and T1-weighted sequences was reduced to 28mm.

Figure 1 shows data obtained after contrast injection. T1w images are displayed as Maximum Intensity Projection (MIP) over 20mm while T2*w images are displayed as Minimum Intensity Projection (mIP). Fine details can be observed on both contrasts. However susceptibility effects limit vascular visualization at the skull base (red arrows, left column). On the other hand, small vascular structures in the deep white matter can only be seen on the T2*w images red arrows, right column). The fusion of both contrasts also indicates that T2* contrast tends to artificially accentuate the size of blood vessels. Figure 2 shows MIPs of T1w images in 4 patients pre and post injection of USPIOs. Exquisite detail can be seen on the post contrast images with clear depiction of the AVM on Patient 4. Interestingly, the subtraction of pre and post contrasts images allows excellent suppression of both background signal and haemorrhage adjacent to the AVM (Patient 3, red arrows). The presence of a large concentration of blood vessels in the AVMs prevented adequate vascular delineation on SWI or T2* related images (Fig.3), while T1w images still provided excellent details. Figure 4 illustrates the evolution of T1w and the different T2*w related contrasts as a function of increasing contrast agent dosage. One can clearly see the competing effect of R1 versus R2* relaxation – whereby T2* contrast increases with increased in USPIO concentration while T1 contrast increases initially but decreases after the fourth injection.This study suggests that both T1w and T2*w USPIO-CEMRA can provide high-resolution images of the human brain, and offer complementary information on brain vasculature. According to signal evolutions with dosage, a single MR sequence can be designed to provide both contrasts simultaneously.