Calcium phosphate cement (CPC), a material used for bone defect restoration, has poor MRI contrast relative to bone, due to similar T₁ and T₂ relaxation properties.¹ It is thus problematic to use MRI to assess CPC and surrounding bone tissue. To enhance image contrast, CPCs have been tagged with conventional gadolinium- or iron-oxide-based contrast agents, but these have unfavorable effects on the properties of CPC and the MR evaluation of iron-oxide particles suffers from a blooming effect.¹ As the benefits of ¹⁹F MRI to follow fluorinated tracers are well known,^2,3 we explored if it was possible to obtain ¹⁹F MR images of a ¹⁹F loaded nanoparticle³ as an alternative contrast agent in ¹H/¹⁹F MRI of CPC. The aim of this study was to optimize ¹⁹F MRI of ¹⁹F labeled CPC. For this purpose we first determined the ¹⁹F relaxation properties of the ¹⁹F nanoparticle.

CPC consists of 59.1wt% alpha-tricalcium phosphate, 1.5wt% carboxymethyl cellulose and 39.4wt% cryo-grinded poly-lactic-co-glycolic acid (PLGA) particles. As ¹⁹F tracer we selected perfluoro-15crown-5-ether (PFCE)² embedded into PLGA nanoparticles through a single emulsion reaction³. This was incorporated into CPC at a concentration of 16.7%.

In vitro samples

A vertical hole (diameter:3mm; height:3mm) was drilled in the upper surface of a pig jaw block, which was filled with mixed material including CPC (~50mg) and PFCE&PLGA (~10mg). The prepared pig bone was then embedded in a falcon tube with 5% gelatine. Additionally, 20mg PFCE&PLGA particles were dissolved into 500μm distilled water in an eppendorf tube as a control sample for T₁ and T₂^* evaluation and quantification analysis.

Ex vivo sample

A cylindrical defect (diameter:2.5mm; depth:3mm) was drilled in each leg of a Wistar rat cadaver. Defects were then filled with injectable CPC in right leg and CPC+PFCE&PLGA in left leg.

Experiments

Measurements were performed on an 11.7T MR-system (Biospec, Bruker) using a ¹H-¹⁹F dual-channel volume coil with 40mm inner diameter. ¹⁹F 3D-ultra-short TE (3D-UTE; TR=4ms, image resolution=1.56mm³) was applied with six flip angles (FAs) from 2°-8° for T₁ estimation and with eleven TEs from 8-608µs with increments of 20-140µs for T₂^* evaluation.

¹H/¹⁹F ZTE images for the bone phantom were acquired with TR=2ms/4ms, image resolution=1.56mm³, 1average/16averages and FA=2°/4° in 27seconds/13mins 53seconds. Additionally, the ¹H/¹⁹F ZTE imaging for rat legs were performed with TR=2ms, image resolution=0.31mm³/1.25mm³, 1averages/32averages and FA=2°/4° in less than 7mins/14mins.

Data Analysis

T₁ values were calculated based on the slope $$$e^{\frac{-TR}{T_{1}}}$$$ from the equation $$$\frac{S\left(FA\right)}{\sin\left(FA\right)}=e^{-\frac{TR}{T_{1}}}\times \frac{S\left(FA\right)}{\\tan\left(FA\right)}+M_{0}\times\left(1-e^{\frac{-TR}{T_{1}}}\right) \times e^{\frac{-TE}{T_2^*}}$$$⁴. A mono-exponential model $$$ S\left(TE\right)=S_{0}\times e^{\frac{-TE}{T_2^*}}$$$ was used for T₂^* fitting. The ¹⁹F signals in the bone phantom and control were quantified in MRIcro software. Regions of interest (ROIs) were manually outlined. ¹⁹F signal per ROI was quantified through multiplying the mean pixel intensity by the pixel number of the ROI and slice thickness. The total ¹⁹F amount was then summed over slices and expressed in arbitrary units (a.u.).

From three measurements of the ¹⁹F nanoparticles, we determined a mean T₁ of 1269±38ms and a mean T₂^* relaxation time of 76.5±8μs (Fig.1; R-squared goodness of fit > 0.99). From this T₁ value, Ernst angle settings were determined for ¹⁹F MRI. Because of the short T₂^* of ¹⁹F spins, a ZTE sequence was selected for optimal SNR in ¹⁹F MRI.

On ¹H-ZTE images of the in vitro sample, CPC cannot be distinguished from pig bone (Fig.2A). However, in ¹⁹F-ZTE images the PFCE in CPC becomes visible and overlaying the ¹⁹F- on ¹H-ZTE images enables a proper sample evaluation (Fig.2B).

The ¹⁹F amount was quantified as 156911 a.u. for the bone phantom and 319422 a.u. for the reference sample. The ¹⁹F quantity ratio of bone phantom to reference is 49.1% (156911/319422). As the true ratio is 50% (10mg/20mg), this means that an excellent quantification accuracy (98%) can be achieved with ¹⁹F-ZTE imaging.

Similar to findings for the in vitro samples, CPC in rat legs cannot be differentiated on ¹H-ZTE images (Fig.3A,C). However, also applying ¹⁹F-ZTE imaging the PFCE in the CPC allows visualization of this material (Fig.3B,D).

In this study we have optimized the parameter settings for the combination of ¹H and ¹⁹F ZTE MR imaging of ¹⁹F labeled CPC material administered to bone defects. CPC cannot be distinguished from bone in ¹H MRI due to similar MR relaxation properties,¹ but in combination with ¹⁹F MRI of ¹⁹F labeled CPC it is possible to analyze the fact of this material qualitatively and quantitatively. Further studies are ongoing to demonstrate that the PLGA&PFCE particles do not have serious adverse effects on CPC.