For malignant lesions, the combination of a local temperature increase inflicted by High Intensity Focused Ultrasound (HIFU) and temperature sensitive liposomes (TSL) enables a triggered drug release confined to the disease site, thereby limiting side effects. Co-release of an MR contrast agent (CA) allows for indirect imaging of the drug release with MR, providing an indirect assessment of the HIFU treatment.¹ However, liposome encapsulation of commonly used Gd-based MR-CA leads to prolonged retention times in the liver and spleen, increasing the risk on nephrogenic systemic fibrosis.² In this study an Fe-deferoxamine derivative is proposed as a safe alternative T₁- CA for TSL encapsulation.

In vitro: TSLs were prepared using DPPC, DSPC, cholesterol, DPPE-PEG2000 (61:14:15:3 molar ratio). They were either actively loaded with doxorubicin (dox) or passively loaded with Fe(III) N-succinyl Deferoxamine (Fe-SDFO, synthetized by SyMO-Chem, the Netherlands)³. TSLs were characterized by Dynamic Light Scattering, Differential Scanning Calorimetry and Inductive Coupled Plasma-Optical Emission Spectroscopy. Release assays were carried out in Fetal Bovine Serum (FBS) at 37, 40 and 42˚C either fluorometrically (dox) or by measuring the longitudinal relaxivity r₁ at 1.41T (Fe-SDFO).

In vivo: 9L glioma tumors were inoculated subcutaneously on the hind limb of Fisher 344 rats. The tumor-bearing animals were placed into a rat MR-HIFU setup in a clinical Philips 3T Sonalleve® MR-HIFU system⁴. Animals were divided into a MR-HIFU treatment group (n = 4) and a non-treated control group (n = 3). All animals were injected i.v. with a mixture of TSLs encapsulating dox and Fe-SDFO (5 mg/kg and 79 μg/kg respectively) just prior to MR-HIFU or control treatment. The MR-HIFU treatment consisted of two sonication periods of 10-15 minutes each (acoustic power 9 – 15 W, acoustic frequency 1.44 MHz) to reach and maintain a tumor temperature of 42˚C. Tumor temperature changes were monitored by proton resonance frequency shift MR thermometry. T₁ maps were acquired before and after the two hyperthermia treatments or at corresponding time points for the control animals, using a single slice Look Locker sequence (Figure 1).

The characteristics of both dox TSLs and Fe-SDFO TSLs are summarized in Figure 2 and were considered comparable and suitable for in vivo application. Both TSLs showed a fast release at 42˚C (Figure 3). 90 ± 4 % of the dox was released within 2 min, compared to a release of 80 ± 4% of the Fe-SDFO. At 37 ˚C, the dox TSLs showed no release over 1.5 h, while the Fe-SDFO TSLs showed a slight release of 15 ± 11%. At 1.41T, unheated Fe-SDFO TSLs displayed an longitudinal relaxation rate r₁ of 0.80 ± 0.01 mM^-1s^-1. After 60 min at 42 ˚C, r₁ increased to 1.35 ± 0.02 mM^-1s^-1.

T₁ maps obtained before and at various time points after TSL injection are shown in Figure 4. All treated tumors showed a contrast change upon the heat treatment (Figure 4, ΔR₁ = 0.18 ± 0.10 s^-1). In the non-treated tumors, the average change in R₁ change was much smaller (Figure 4, ΔR₁ = 0.028 ± 0.007 s^-1). Strikingly, the R₁ change across the treated tumors was homogeneous in some (animal 2, animal 4, Figure 4), while very inhomogeneous in others (animals 1 and 3, Figure 4), suggesting untreated tumor areas in the latter.

Dox-loaded TSLs and Fe-SDFO loaded TSLs were characterized in vitro, showing the potential of this combined system for MR-guided triggered drug release, even though the r₁ of the proposed CA is around a factor 3 lower than that of more commonly used Gd-based CA.¹ An in vivo proof-of-concept study was conducted to assess the feasibility of monitoring drug release using the newly designed drug and CA loaded TSL systems. Treated tumors showed an increase in R₁ while the R₁ of the untreated control tumors remained essentially unchanged. Moreover, the pattern of R₁ change could elucidate the pattern of drug release across the tumor. Future work aims to correlate the dox delivery to the tumors with the observed R₁ changes.