Local Magnetic Hyperthermia and Systemic Gemcitabine/Paclitaxel Chemotherapy Triggers Neo-Angiogenesis in Orthotopic Pancreatic Tumors without Involvement of Auto/Paracrine Tumor Cell VEGF Signaling and Hypoxia

Abstract

There is a growing interest in exploring the therapeutically mediated modulation of tumor vascularization of pancreatic cancer, which is known for its poorly perfused tumor microenvironment limiting the delivery of therapeutic agents to the tumor site. Here, we assessed how magnetic hyperthermia in combination with chemotherapy selectively affects growth, the vascular compartment of tumors, and the presence of tumor cells expressing key regulators of angiogenesis. To that purpose, a orthotopic PANC-1 (fluorescent human pancreatic adenocarcinoma) mouse tumor model (Rj:Athym-Foxn1nu/nu) was used. Magnetic hyperthermia was applied alone or in combination with systemic chemotherapy (gemcitabine 50 mg per kg body weight, nab-pacitaxel 30 mg/kg body weight) on days 1 and 7 following magnetic nanoparticle application (dose: 1 mg per 100 mm³ of tumor). We used ultrasound imaging, immunohistochemistry, multi-spectral optoacoustic tomography (MSOT), and hematology to assess the biological parameters mentioned above. We found that magnetic hyperthermia in combination with gemcitabine/paclitaxel chemotherapy was able to impact tumor growth (decreased volumes and Ki67 expression) and to trigger neo-angiogenesis (increased small vessel diameter) as a result of the therapeutically mediated cell damages/stress in tumors. The applied stressors activated specific pro-angiogenic mechanisms, which differed from those seen in hypoxic conditions involving HIF-1α, since (a) treated tumors showed a significant decrease of cells expressing VEGF, CD31, HIF-1α, and neuropilin-1; and (b) the relative tumor blood volume and oxygen level remained unchanged. Neo-angiogenesis seems to be the result of the activation of cell stress pathways, like MAPK pathways (high number of pERK-expressing tumor cells). In the long term, the combination of magnetic hyperthermia and chemotherapy could potentially be applied to transiently modulate tumor angiogenesis and to improve drug accessibility during oncologic therapies of pancreatic cancer.

Keywords: angiogenesis; chemotherapy; hypoxia; magnetic hyperthermia; orthotopic tumor model; pancreatic adenocarcinoma (PDAC).

Figure 1

Timeline for animal experiments with the orthotopic PANC-1 tumor model. Chemo: systemic chemotherapy; MH: magnetic hyperthermia; MNP/µCT: magnetic nanoparticle injection followed by NIRF and CT imaging of animals. MSOT: multispectral optoacoustic tomography.

Figure 2

Both bimodal therapy and chemotherapy significantly inhibited pancreatic tumor growth. (a) Mean of relative tumor volumes determined over time from ultrasound images for the various animal groups are presented as percentages in relation to the tumor volume on day 0 before hyperthermia (MH1) therapy. *** p < 0.001; **** p < 0.0001 for day 28 in hyperthermia alone, MNP, and non-treated animal group compared to day 0. n = 5 per group. Each bar represents mean ± SD. (b) Representative ultrasound images of the tumors in each group. (c) Representative light images of excised tumors between the spleen and stomach, demonstrating a decrease in tumor volume after bimodal therapy and chemotherapy (caliper 1 mm). Tu: tumor; Ki: kidney. Animal groups: MHsC = magnetic hyperthermia + chemotherapy, MH = magnetic hyperthermia, sC = systemic chemotherapy, M = MNP alone, and N = non-treated.

Figure 3

The different therapies induced varying degrees of cellular damage in the residual pancreatic tumor tissue. Tumors treated with MNPs only show intact cell-to cell connections and MNPs. (a) Cytoplasmatic vacuolization, (b) single cells (macrophages) with incorporated iron oxide nanoparticles. Animal groups: MHsC = magnetic hyperthermia + chemotherapy, MH = magnetic hyperthermia, sC = systemic chemotherapy, and M = MNPs alone. Scale: #: 5.0 µm, +: 2.5 µm.

Figure 4

Magnetic hyperthermia combined with chemotherapy decreased intratumoral angiogenesis in residual pancreatic tumors. Semi-quantitative evaluation of intratumoral markers: (a) VEGF, (b) VEGFR2, (c) Neuropilin 1, (d) Ki67, and (e) phosphorylated ERK (pERK). Each bar represents mean ± SEM. Eight images were quantified per treatment group (n = 2 mice/group). Rel. = relative: Positively stained cells computed relative to the nuclei number in a ROI stained with hematoxylin using the Image-J software. (f) Representative histological micrographs of tumor tissue slices, scale bar: 20 µm. Animal groups: MHsC = magnetic hyperthermia + chemotherapy, MH = magnetic hyperthermia, sC = systemic chemotherapy, M = MNP alone, and N = non-treated. * p < 0.05, ** p < 0.01, **** p < 0.0001.

Figure 5

Chemotherapy in combination with magnetic hyperthermia significantly decreased the number of endothelial cells in residual pancreatic tumors. (a) Semi-quantitative analysis of the tumor’s CD31 expression levels. (b) Representative micrographs showing the number of CD31-positive cells in tumor tissue slices of the various treatment groups. Each bar represents mean ± SEM. Eight images were quantified per treatment group (n = 2 mice/group for CD31 expression analysis). Positive stains were computed relative to the nuclei number in a ROI stained with hematoxylin. Scale bar: 20 µm. Animal groups: MHsC = magnetic hyperthermia + chemotherapy, MH = magnetic hyperthermia, sC = systemic chemotherapy, M = MNP alone, and N = non-treated. * p < 0.05, ** p < 0.01.

Figure 6

Bimodal therapy leads to a greater number of smaller blood vessels in residual pancreatic tumors than other treatments. (a) Histogram showing distribution of tumor blood vessels with regard to their diameters. The bimodal therapy group revealed a higher number of smaller tumor vessels of 0–5 µm diameter (area of analyzed ROI: 9.3 mm²) (MHsC, yellow arrow) as compared to those from the other treatment groups. (b) Semi-quantitative evaluation of CD31-positive tumor vessel diameter (0–120 µm). A bin is a bar whose height reflects the number of data points it contains; a bin center is the center value of the grouped data. Each box-plot represents mean ± SEM (n = 100 vessels/group). Animal groups: MHsC = magnetic hyperthermia + chemotherapy, MH = magnetic hyperthermia, sC = systemic chemotherapy, M = MNP alone, and N = non-treated. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 7

Bimodal therapy effectively abolished HIF-1α expression and hypoxic regions (necrotic areas with high HIF-1α concentrations) in residual pancreatic tumors. (a) Semi-quantitative evaluation of HIF-1α expression levels within the tumor slices computed in relation to the number of cell nuclei in a ROI (area: 9.3 mm²). Each bar represents mean ± SEM. Eight images were quantified per treatment group (n = 2 mice/group). Positive stains were computed relative to the cell (nuclei) number. (b) Representative histological micrographs of tumor tissue slices showing hypoxic regions. Animal groups: MHsC = magnetic hyperthermia + chemotherapy, MH = magnetic hyperthermia, sC = systemic chemotherapy, M = MNP alone, and N = non-treated. * p < 0.05, **** p < 0.0001. Yellow arrows = HIF-1α expression; black arrows: necrotic regions.