Magnetic nanoparticles (MNPs) can be used as carriers for magnetic drug targeting ¹. When assembled with oncolytic viruses they are a promising tool for cancer treatment. The contrast enhancing properties of these particles in MRI allow for non-invasive therapy monitoring. Our goal is to investigate how MRI relaxometry can be used to assess the spatial distribution of the particle concentration and the aggregation state of the MNP-virus-complexes in the tissue of interest.

For this purpose, we studied the influence of the assembling of MNPs with oncolytic viruses and their uptake into cancer cells on the MRI relaxivities r₁, r₂ and r₂^*. In addition, all samples were magnetically characterized using magnetic particle spectroscopy (MPS). MPS has been proven a straightforward technique for quantification and characterization of MNPs and allows for the validation of MRI relaxation measurement results.

PEI-Mag2 (PM) and SO‐Mag5 (SO) magnetic nanoparticles were assembled with oncolytic adenovirus (Ad) as described previously ². Multi drug resistant human pancreatic carcinoma (RDB) cells were labelled with PM and SO nanoparticles and infected with magnetic Ad complexes.

Both particles were also assembled with oncolytic vesicular stomatitis virus (VSV). Rat Morris hepatocellular carcinoma (McA) cells were labelled with PM and SO nanoparticles and infected with magnetic VSV complexes.

Dilution series of all sample types were prepared and embedded in agarose. R₁, R₂ and R₂^* relaxation rates were measured on a 3 T whole body imaging system (GE Discovery MR750w) using spin echo inversion recovery, (single) spin echo and multi-echo gradient echo sequences. ROI signal intensities were least-squares fitted for each voxel with three-parameter fits. A commercial magnetic particle spectrometer (Bruker BioSpin) was used to measure the spectral response to an oscillating magnetic field (B_excit=25 mT, f₀=25 kHz, T=37°C) of the samples.

For clarity, only exemplary data is shown here. Figures 1-3 show the relaxation rates of the dilution series of PEI-Mag2 nanoparticles homogeneously suspended in agarose, assembled with adenovirus, after internalization in RDB cells and assemblies with adenovirus after internalization in RDB cells. Relaxivities are summarized in Table 1. r₁ and r₂ relaxivities do not change significantly when the complexes are formed but drop when the particles are absorbed by the cells. Within the error margins, r₂^* relaxivites are very similar in all samples.

The MPS results of the dilution series demonstrate a highly linear correlation between the signal amplitude A₃ and the iron amount m_Fe for PEI-Mag2 and SO-Mag5 samples. The shape of the MPS spectra, expressed as the A₅/A₃ ratio, was not affected by dilution within uncertainty limits for samples with more than 0.5 μg iron. As the A₅/A₃ ratio is a strong indicator for a changing magnetic response, the results shown in Figure 4 indicate that the structure of the aggregates (and therefore the magnetic interaction between the particles in the aggregates) is largely the same for all samples. This can explain the constant r₁ and r₂ relaxivities for the free particles and the magnetic viral complexes. The distribution of the aggregates in the sample (homogeneous in agarose but more localized in the cells) has obviously no influence on the magnetic interaction between the particles and therefore on their MPS spectra, but does have an influence on the local interaction between the particles and the water protons. This can explain the drop in r₂ for particles/complexes internalized by the cells (partial refocusing ³).

However, lower concentrated samples show higher deviations in their MPS spectra including some trends, which is not reflected in the relaxation rates. Possible reason might be a changed aggregate structure due to dilution accompanied by an altered dipole-dipole interaction between MNPs within the MNP virus complex.

Since r₂^* relaxivites are very similar in all samples, R₂^* measurements seem most suitable for particle quantification after a calibration step. r₂ relaxivities, however, change significantly when the particles or the magnetic viral complexes are internalized by the cells. Measurements of R₂ in addition to R₂^* could thus be used to detect whether the particles are inside or outside of the cells. r₁ relaxivities are very low for all samples, which is why they are expected to give too little contrast in tissue for a quantitative interpretation. Our results also show that additional MPS measurements are important for the accurate interpretation of MRI relaxometry results because MPS detects MNPs and their interaction directly whereas MRI can only observe the interaction between particles and water protons.