Multimodal imaging approach to track theranostic nanoparticle accumulation in glioblastoma with magnetic resonance imaging and intravital microscopy

Abstract

Theranostic nanoparticles (NPs) have been designed for simultaneous therapeutic and diagnostic purposes, thereby enabling personalized cancer therapy and in vivo drug tracking. However, studies thus far have focused on imaging NP tumor accumulation at the macroscopic level and correlating results with ex vivo histology. Limited evidence exists on whether in vivo NP tumor contrast enhancement on magnetic resonance imaging (MRI) correlates with in vivo NP tumor accumulation at the microscopic level. To address this gap, the purpose of our study was to correlate quantitative MRI estimates of NP accumulation with in vivo NP signal quantification as measured through two-photon intravital microscopy (IVM) in an orthotopic murine glioblastoma multiforme model (GBM). To enable multimodal imaging, we designed dual-mode NPs, composed of a carbohydrate-coated magnetic core (Ferumoxytol) as an MRI contrast agent, and a conjugated fluorophore (FITC) for IVM detection. We administered these NPs with or without a conjugated vascular disrupting agent (VDA) to assess its effect on NP delivery to GBM. We correlated in vivo MRI contrast enhancement in tumors, quantified as T₂ relaxation time, with IVM fluorescence spatial decay rate. Results demonstrated a significantly lower tumor T₂ relaxation time and spatial decay rate in tumors targeted with VDA-conjugated NPs compared to unconjugated NPs. Postmortem histological analyses validated the in vivo observations. The presented multimodal imaging approach enabled a quantitative correlation between MRI contrast enhancement at the macroscopic level and NP accumulation in the tumor microenvironment. These studies lay the groundwork for the precise evaluation of the tumor targeting of theranostic NPs.

Fig. 1. Electron microscopy and optical fluorescence. (a) TEM micrograph of Ferumoxytol-FITC showing the dispersed iron oxide cores. Scale bars, 50 nm. (b) Optical fluorescence spectra of Ferumoxytol-FITC and Ferumoxytol-FITC-VDA, employing an excitation wavelength of 480 nm, and yielding emission peaks at 521 and 517 nm, respectively.

Fig. 2. Dual-mode nanoparticle properties. (a) Representative axial T₂-weighted scan of phantoms with increasing concentrations of nanoparticles and corresponding T₂ multi-slice multi-echo (MSME) map. (b) Transverse relaxation rate (R₂) as a function of iron concentration for Ferumoxytol (in grey), Ferumoxytol-FITC (in black), and Ferumoxytol-FITC-VDA (in blue). (c) Optical fluorescence map demonstrating increasing fluorescence of FITC-conjugated nanoparticles with increasing iron concentration. (d) Fluorescence intensity as a function of iron concentration for Ferumoxytol-FITC (in black), and Ferumoxytol-FITC-VDA (in blue).

Fig. 3. In vivo magnetic resonance imaging before and after NP injection. (a) Representative axial T₂-weighted MR images of the brain and corresponding color T₂ maps of mice before (0 h) and after (24 h) after intravenous injection of either Ferumoxytol-FITC (50 mgFe per kg) or Ferumoxytol-FITC-VDA (50 mgFe per kg). (b) Mean T₂ relaxation time violin plots of the tumor area from mice injected with either Ferumoxytol-FITC (in black) or Ferumoxytol-FITC-VDA (in blue), before (t = 0 h, dashed lines) and after (t = 24 h, solid lines) NP administration. Significant difference between the two groups was indicated when *P < 0.05 (n = 4 per group).

Fig. 4. In vivo two-photon intravital microscopy of GBM after NP injection. (a) Schematic representation of the excitation laser and detection windows for conjugated iron oxide nanoparticles (SPIONs), in green) and implanted glioblastoma multiforme (GBM, in red), for imaging through two-photon intravital microscopy (IVM). (b) Representative IVM images of implanted brain tumors in mice injected with Ferumoxytol-FITC and Ferumoxytol-FITC-VDA at 24 h after NP administration, showing GBM (in red) and FITC (in green) signals in axial planes and corresponding sagittal planes. Scale bars, 100 μm.

Fig. 5. Nanoparticle extravasation and multiscale imaging correlation. (a) Boxplot of fluorescence spatial decay rate of Ferumoxytol-FITC (in grey) and Ferumoxytol-FITC-VDA (in blue). Significant difference between the two groups was indicated when *P < 0.05 (n = 4 per group). (b) Correlation scatter plot between spatial decay rate and T₂ relaxation time for mice administered with Ferumoxytol-FITC (in black) and Ferumoxytol-FITC-VDA (in blue). Experimental points (±SD) were fitted with a linear regression (dashed red line, R² = 0.84).

Fig. 6. Histopathology analysis. (a) CD31 staining (in brown) of mouse brain tumor tissue following injection with Ferumoxytol-FITC or Ferumoxytol-FITC-VDA, highlighting signs of vessel collapse (black arrows). Scale bars, 200 μm. (b) Optical fluorescence analysis of FITC signal from tumor tissue slides of mice injected with Ferumoxytol-FITC or Ferumoxytol-FITC-VDA. Scale bars, 200 μm. (c) Normalized intensity (±SD) bar plot comparing Ferumoxytol-FITC and Ferumoxytol-FITC-VDA. Significant difference between the two groups was indicated when **P < 0.005.

Fig. 7. Study concept. (a) Stereotactic implantation of C6 glioblastoma cells into the right brain hemisphere and surgical implantation of a cranial glass window above the tumor implantation site. (b) Baseline pre-contrast imaging with magnetic resonance imaging (MRI). (c) Nanoparticle administration via intravenous injection. (d) In vivo post-contrast MRI. (e) Post-contrast two-photon intravital microscopy.