Preclinical pharmacology of patient-derived extracellular vesicles for the intraoperative imaging of tumors

Abstract

Extracellular vesicles (EVs) derived from the plasma of oncological patients exhibit significant tumor-targeting properties, unlike those from healthy individuals. We have previously shown the feasibility of formulating the near-infrared (NIR) fluorescent dye indocyanine green (ICG) with patient-derived extracellular vesicles (PDEVs) for selective delivery to neoplastic tissue. This staining protocol holds promise for clinical application in intraoperative tumor margin imaging, enabling precise neoplastic tissue resection. To this end, we propose the ONCOGREEN protocol, involving PDEV isolation, ICG loading, and reinfusion into the same patients. Methods: By in vivo studies on mice, we outlined key pharmacological parameters of PDEVs-ICG for intraoperative tumor imaging, PDEV biodistribution kinetics, and potential treatment-related toxicological effects. Additionally, we established a plasmapheresis-based protocol for isolating autologous PDEVs, ensuring the necessary large-scale dosage for human treatment. A potential lyophilization-based preservation method was also explored to facilitate the storage and transport of PDEVs. Results: The study identified the effective dose of PDEVs-ICG necessary for clear intraoperative tumor margin imaging. The biodistribution kinetics of PDEVs showed favorable targeting to neoplastic tissues, without off-target distribution. Toxicological assessments revealed no significant adverse effects associated with the treatment. The plasmapheresis-based isolation protocol successfully yielded a sufficient quantity of autologous PDEVs, and the lyophilization preservation method maintained the functional integrity of PDEVs for subsequent clinical application. Conclusions: Our research lays the groundwork for the direct clinical application of autologous PDEVs, initially focusing on intraoperative imaging. Utilizing autologous PDEVs has the potential to accelerate the integration of EVs as a targeted delivery tool for anti-neoplastic agents to cancerous tissue. This approach promises to enhance the precision of neoplastic tissue resection and improve overall surgical outcomes for oncological patients.

Keywords: Bench-to-bedside Translation; EV Biodistribution Kinetics; Intraoperative Imaging; Toxicology.

Figure 1

Determination of the Minimum Effective Dose. Representative images of ICG fluorescence in vivo (A) and ex vivo in the tumor, lung, and liver (B), obtained with the IVIS Spectrum Imaging System and the SPY Elite intraoperative imaging device. The tumor margins in the in vivo pictures are highlighted by the green dotted line. In the color scale, blue represents the minimum fluorescence signal, whereas red represents the maximum. Additional ex vivo images from other replicates are presented in Supplementary Figure 3. Mice bearing tumors were administered three dosages of nanoparticles - 3.3E09 EVs/Kg, 3.3E08 EVs/Kg, or 3.3E07 EVs/Kg - of the ONCOGREEN formulation. C) Quantification of the fluorescent signals in the tumors 24 hours after injection is presented in the graph; bars in the graph represent the average +/- S.E.M values of eight animals, *** p < 0.001, ** p < 0.01 calculated by one-way ANOVA followed by Bonferroni's test.

Figure 2

Independence of the average fluorescent signal intensity from for the tumor size. Three groups of mice bearing tumors of different sizes (small 0.1-0.25 cm³, medium 0.3-0.5 cm³, and large 0.7-1.0 cm³) were intravenously treated with 3.3E09 EVs/Kg of the ONCOGREEN formulation (MED). Mice were sacrificed 24 hours after treatment. A) Representative ex vivo images of ICG fluorescence. In big and medium tumors, fluorescence is emitted from the entire surface of the tumor, as indicated by the blue coloration representing the presence of a near-infrared fluorescent signal. The samples reported in the 'Small'-labelled column include the tumor collected together with surrounding healthy tissue: the tumor margins are highlighted by the green dotted line. Each individual picture in the panel represents a different mouse. In the color scale, blue represents the minimum fluorescence signal, whereas red represents the maximum. B) Quantification of the tumor fluorescent signal divided by the total tumor area.

Figure 3

Kinetics of the biodistribution of EV-formulated ICG in tumor-bearing mice. Three groups of 4 mice each carrying syngeneic 0.4 cm³ MC38 tumors were administered a single MED of ONCOGREEN (3.3E09 EVs/Kg) or an equal dose of the standard formulation of free ICG (4 mg free ICG/Kg). The fluorescence biodistribution was analyzed ex vivo at 2, 24, and 96 hours using IVIS Spectrum Imaging. Other organs are depicted in Supplementary Figure 5. **:pVal < 0.01 with ANOVA test.

Figure 4

Characterization of EVs from colorectal cancer patients using the plasmapheresis protocol. The PDEVs exhibited size distribution, shape, EV-specific marker expression comparable to the PDEVs used in previous experiments. A) NTA of particle size distribution of PDEVs. The lines represent the mean of 5 readings. Details on the size distribution and concentration are provided in Table 4. B) Immunoblot analysis of TSG101 (47 KDa) expression in plasma-derived EVs from patients (lane 1: patient 1; lane 2: patient 2; lane 3: patient 3; lane 4: patient 4). C) Representative EV morphology and size obtained by cryo electron microscopy. Scale bar: 100 nm. D) Flow cytometry analysis of CFSE-labeled PDEVs showing the percentage (>80%) of CFSE-positive (CFSE+) vesicles. The cytogram depicts the side scatter (SSC)-A vs B3-A (green fluorescence triggering) used to trace the CFSE+ gate. CFSE, carboxyfluorescein succinimidyl ester; PDEVs, patient-derived extracellular vesicles.

Figure 5

Characterization of lyophilized EVs. EVs derived from the MCF7 tumor cell line were freeze-dried and were subsequently stored at room temperature (RT) or 4°C for a duration of 4 months. A) Comparison of nanoparticle tracking analysis of MCF7-EVs and lyophilized MCF7-EVs kept at 4 °C and RT. The lyophilized vesicles showed similarities in terms of shape and size distribution when compared to non-lyophilized EVs (CTRL). B) Immunoblot analysis of EV marker proteins alpha-tubulin (50 KDa) and TSG101 (47 KDa) in both lyophilized and non-lyophilized MCF7 EVs. The graphs show the band density of proteins normalized for ponceau staining. C) Representative pseudocolored images display ex vivo ICG fluorescence in MC38 tumor-bearing mice. These images were captured 24h after the intravenous injection of EVs loaded with ICG, comparing lyophilized or non-lyophilized EVs. The color scale represents the fluorescence signal, with black indicating the minimum intensity and yellow indicating the maximum. D) The graphs show the ex vivo ICG fluorescence, 24h after treatment with lyophilized or non-lyophilized EVs. Additional In vivo imaging pictures for the RT group are reported in Supplementary Figure 12.