MRI of tumor-associated macrophages with clinically applicable iron oxide nanoparticles

Abstract

Purpose: The presence of tumor-associated macrophages (TAM) in breast cancer correlates strongly with poor outcome. The purpose of this study was to develop a clinically applicable, noninvasive diagnostic assay for selective targeting and visualization of TAMs in breast cancer, based on magnetic resonanceI and clinically applicable iron oxide nanoparticles.

Experimental design: F4/80-negative mammary carcinoma cells and F4/80-positive TAMs were incubated with iron oxide nanoparticles and were compared with respect to magnetic resonance signal changes and iron uptake. MMTV-PyMT transgenic mice harboring mammary carcinomas underwent nanoparticle-enhanced magnetic resonance imaging (MRI) up to 1 hour and 24 hours after injection. The tumor enhancement on MRIs was correlated with the presence and location of TAMs and nanoparticles by confocal microscopy.

Results: In vitro studies revealed that iron oxide nanoparticles are preferentially phagocytosed by TAMs but not by malignant tumor cells. In vivo, all tumors showed an initial contrast agent perfusion on immediate postcontrast MRIs with gradual transendothelial leakage into the tumor interstitium. Twenty-four hours after injection, all tumors showed a persistent signal decline on MRIs. TAM depletion via αCSF1 monoclonal antibodies led to significant inhibition of tumor nanoparticle enhancement. Detection of iron using 3,3'-diaminobenzidine-enhanced Prussian Blue staining, combined with immunodetection of CD68, localized iron oxide nanoparticles to TAMs, showing that the signal effects on delayed MRIs were largely due to TAM-mediated uptake of contrast agent.

Conclusion: These data indicate that tumor enhancement with clinically applicable iron oxide nanoparticles may serve as a new biomarker for long-term prognosis, related treatment decisions, and the evaluation of new immune-targeted therapies.

Figure 1. In vitro MR scans of iron oxide nanoparticle-labeled cells with corresponding quantitative MR signal enhancement and spectrometry data

(A) Axial T2-weighted MR images through test tubes containing F4/80⁺ versus F4/80⁻ cells labeled overnight with Ferumoxytol, P904, P1133 alone or P1133 with free folic acid (FFA). Cells were kept in suspension in ficoll solution and test tubes were placed in a water bath to avoid artifacts by surrounding air (which would cause a dark MR signal). Image parameters: 3 Tesla, SE 2000/60 (TR/TE in ms). (B) Corresponding R2 relaxation rates, i.e. quantitative measures of the MR signal effect, of iron oxide nanoparticle labeled and unlabeled F4 F4/80⁺ versus F4/80⁻ cells, displayed as mean +/− standard deviation from duplicate experiments. (C) Iron content of the same cell samples as shown in B, as determined by mass spectrometry.

Figure 2. In vivo MR imaging of iron oxide nanoparticles

(A) T2-weighted SE images of representative mammary tumors in MMTV-PyMT mice prior to (precontrast), 1 hour (h) and 24 h after administration of 0.5 mmol [Fe]/kg of ferumoxytol, P904 or P1133. The iron oxide nanoparticle-based contrast agents cause a negative (dark) signal effect in the tumor tissue on these scans (arrows point to tumors). (B) Quantitation of MR signal enhancement (delta R2 measurement),of mammary tumors in MMTV-PyMT mice before and after iron oxide-nanoparticle administration, displayed as means +/− standard deviation (n=7 mice/group, except P1133+FFA which contained 3 mice). Note that all tumors show a nanoparticle retention at 24 hours, which is most pronounced for the folate-linked nanoparticle P1133.

Figure 3. Uptake of ferumoxytol by TAMs in vivo

(A) Localization within OCT-embedded mammary tumors of ferumoxytol (iron; black contrast) to CD68⁺ macrophages (green) using phase contrast of DAB staining and confocal microscopy. (B) Localization of ferumoxytol-FITC (green) to CD68⁺ macrophages (red) but not Keratin 18⁺ carcinoma cells (red) within mammary tumors. Scale bars are shown in images.

Figure 4. Folate receptor expression and folate-targeted uptake of nanoparticles

(A) Staining for folate receptor α (FRα; red) and CD68⁺ macrophages (green) demonstrates that expression of FRα is localized to carcinoma cells within mammary tumors. (B) A subpopulation of CD68⁺ macrophages (green) express folate receptor β (FRβ;̣ red staining). (C) Prussian Blue staining for iron with DAB enhancement within mammary tumors from mice injected with ferumoxytol, P904 or P1133. Scale bars are shown in images.

Figure 5. Ferumoxytol-MRI detects TAM-depletion non-invasively in vivo

(A) Quantitative MR signal enhancement (delta R2 measurement) of MMTV-PyMT mammary before and after iron oxide-nanoparticle administration, displayed as means +/− standard deviation of three mice treated with anti-CSF1 monoclonal antibody and three controls. An additional control mouse underwent serial MRI without any contrast agent injection in order to confirm, that the evaluated MMTV-PyMT tumors do not show any intrinsic changes in MR signal within a two-day observation period. Note that mice treated with anti-CSF1 mAb demonstrated significantly smaller ΔR2 enhancement data compared to untreated controls. (B) Corresponding confocal microscopy evaluations confirmed TAM-depletion of anti-CSF1 mAb treated tumors. The detected quantity of CD68⁺ macrophages was markedly higher in anti-CSF1 mAb treated tumors compared to untreated control tumors.