Quantitative molecular imaging of angiogenesis-targeted fluorinated nanoparticles: new approaches for B₁-mapping compensation for ¹⁹F-MRI

Background

Quantitative MR molecular imaging allows for the detection of targeted contrast agents to diagnose disease states and monitor response to therapy, such as anti-angiogenic therapy in atherosclerosis and cancer with α_νβ₃-integrin targeted perfluorocarbon (PFC) nanoparticles. Recently, ¹⁹F MR using a ¹⁹F/¹H dual-tuned RF coil has been utilized to directly image and quantify the fluorinated core of these PFC nanoparticle (NP) emulsions. However, low concentrations of these fluorine agents in the body, in conjunction with varying RF coil sensitivity profiles (B₁-field inhomogeneities) raise obstacles to accurate quantification. This study presents a strategy to more accurately quantify the sparse ¹⁹F signal from PFC NP emulsions with a ¹H image-based Actual Flip-angle Imaging (AFI) B₁-mapping correction to the ¹⁹F and ¹H images.

Methods

New Zealand White Rabbits (2 kg) were implanted with a VX2 adenocarcinoma tumor (2-3 cm) in the hind leg. Angiogenesis imaging was performed 2 weeks post implantation, under ketamine/xylazine anesthesia. An α_νβ₃-integrin targeted perfluoro-octyl bromide (PFOB) nanoparticle emulsion was prepared, and injected intravenously 3 hours before imaging. MR data were acquired on a 3T clinical whole-body scanner (Achieva, Philips Healthcare) with a dual ¹⁹F/¹H spectrometer system and a dual-tuned transmit/receive single loop surface RF coil (7×12 cm). A simultaneous ¹⁹F/¹H gradient echo (GRE) imaging sequence was used with: ¹⁹F offset frequency on the center of the PFOB CF₂ peak, 15 4-mm slices, 140 mm FOV, 48³ matrix, α = 60°, TE/TR = 2.2/8.5 ms, 21 min scan time. The B₁ field was mapped using an AFI sequence with matching geometry. Using the flip angle map and a model of the GRE signal, a spatially-dependent calibration mask was calculated in MATLAB (MathWorks) and used to compensate the ¹H and ¹⁹F signal intensities for the GRE sequence.

Results

PFC NP targeted the tumor neovasculature, and provided localized ¹⁹F signal as expected. Figure 1 displays the uncorrected (top) and corrected (bottom) ¹H images with the ¹⁹F signal superimposed, using the AFI (middle) B₁-mapping correction technique. After correction, the ¹H signal intensity profile as a function of distance from the surface coil (located at right) is improved. After the same correction to the ¹⁹F signal, the measured concentration of nanoparticles when compared to a standard was 10.2 ± 1.0 mM_19F, versus 9.0 ± 2.2 mM_19F before correction.

Figure 1

Top: uncorrected ¹H image of rabbit model with angiogenesis-targeted ¹⁹F imaging (blue overlay). Middle: AFI B₁ map (% actual/requested flip angle). Bottom: ¹H and ¹⁹F superimposed images using AFI B₁-mapping correction.

Conclusions

An image-based B₁-mapping correction acquired with ¹H can be used to correct signal intensities for ¹⁹F and ¹H images of angiogenesis in an in vivo rabbit model. The correction results in a more homogeneous ¹H image of the anatomy and facilitates accurate measurement of bound α_νβ₃-integrin targeted nanoparticles with ¹⁹F imaging.

Funding

AHA 11PRE7530046; NIH R01 HL073646.