MRI with Magnetic Nanoparticles Monitors Downstream Anti-Angiogenic Effects of mTOR Inhibition

Abstract

Purpose

To study the effect of mammalian target of rapamycin (mTOR) inhibition on angiogenesis with magnetic resonance imaging (MRI) using magnetic iron oxide nanoparticles (MNP).

Procedures

One million CAK-1 renal cell carcinoma cells were subcutaneously implanted into each of 20 nude mice. When tumors reached ∼750 μl, four daily treatment arms began and continued for 4 weeks: rapamycin (mTOR inhibitor) 10 mg/kg/day; sorafenib (VEGF inhibitor) high dose (80 mg/kg/day) and low dose (30 mg/kg/day); and saline control. Weekly MRI (4.7 T Bruker Pharmascan) was performed before and after IV MION-48, a prototype MNP similar to MNP in clinical trials. Vascular volume fraction (VVF) was quantified as ΔR2 (from multi-contrast T2 sequences) and normalized to assumed muscle VVF of 3%. Linear regression compared VVF to microvascular density (MVD) as determined by histology.

Results

VVF correlated with MVD (R² = 0.95). VVF in all treatment arms differed from control (p < 0.05) and declined weekly with treatment. VVF changes with rapamycin were similar to high-dose sorafenib.

Conclusion

This study demonstrates noninvasive, in vivo anti-angiogenic monitoring using MRI of mTOR inhibition.

Appendix

The theory behind the use of superparamagnetic contrast agents with long vascular half lives to determine blood volumes by MRI, which is supported by detailed numerical simulations, is that the change in gradient-echo transverse relaxation rate (ΔR ₂*) relative to the pre-injection baseline is proportional to the local blood volume times some function of the plasma concentration of the agent or $\Delta {R_2} {*} = k \times f(P) \times V$. If a steady state concentration of the contrast agent is assumed in the blood plasma then there is a simple linear relationship between ΔR ₂* and blood volume at any time (t) and the formula reduces to $\Delta {R_2} {*} (t) = K \times V(t)$. Stated another way, $V(t) = {{{\left[ {{{\Delta}}{{\hbox{R}}_2} {*} (t)} \right]}} \left/ {K} \right.}$ where V(t) is the blood volume, ΔR ₂* (t) is the change in the transverse relaxation rate of the region, and the constant K includes the contrast agent blood pool concentration and is therefore dose dependent. While this technique allows for easy measurement of total blood volume in a given voxel, tissue slice, or entire organs, additional methods more sensitive to microvessels have also been developed.

These methods are based on the property that compartmentalization of these magnetic nanoparticles also induces long-range magnetic field perturbations that extend over many microns and increase both transverse relaxation rates, R ₂ and R ₂*, of the tissues. The enhancement of R ₂ and R ₂* caused by these agents can be expressed as follows: $\Delta {R_2} = { }{{{1}} \left/ {{T{{2}_{\rm{post}}}}} \right.}-{{{1}} \left/ {{T{{2}_{\rm{pre}}}}} \right.} \approx - {{{1}} \left/ {\hbox{TE}} \right.}{\, \ln }\left( {{{{{S_{\rm{post}}}}} \left/ {{{S_{\rm{pre}}}}} \right.}} \right)$ and $\Delta {R_2} {*} { } = {{{1}} \left/ {{{\hbox{T2}}{{*}_{\hbox{post}}}}} \right.}-{{{1}} \left/ {{{\hbox{T2}}{{*}_{\hbox{pre}}}}} \right.} \approx {{{ - {1}}} \left/ {\hbox{TE}} \right.}{ \ln }\left( {{{{{S_{\hbox{post}}}}} \left/ {{{S_{\hbox{pre}}}}} \right.}} \right)$ where S is the signal intensity, TE the echo time, and T2 is the transverse relaxation time. It has previously been shown that there exists a unique relationship between ∆R ₂ (spin echo), ∆R ₂* (gradient echo) as a function of vessel diameter and contrast agent concentration. The ∆R ₂ peaks for vessels 1-2 μm in diameter whereas ∆R ₂* is fairly independent of vessel size beyond 3-4 μm. These studies thus provide the rationale for spin-echo imaging being more sensitive to microvasculature, while gradient-echo imaging is more sensitive to total vasculature. Ultimately, gradient echo/spin ratio imaging could theoretically be utilized to measure vessel size noninvasively, as the ΔR ₂*/ΔR ₂ ratio increases nearly linearly with vessel size. Sequestered magnetic nanoparticles, as occurs with uptake by macrophages, may also be detected based on long-range magnetic field perturbations utilizing a similar approach.

Therefore, if the R ₂ is calculated prior to and following administration of MNP, then the ∆R ₂ can be determined accurately for both the tumor, over multiple slices, and the musculature. Following this conversion, a straightforward proportionality, utilizing a well-accepted muscular VVF of 3%, can be performed.