Radiation-induced inflammatory response in tumor-bearing immune-compromised mice by SPIO-enhanced T2-MRI
Natalie Julie Serkova1, Kendra M Huber1, Barbara Frederick2, Elizabeth R Kessler3, Thomas W Flaig3, and Brian D Kabanagh2

1Anesthesiology, University of Colorado Denver, Aurora, CO, United States, 2Radiation Oncology, University of Colorado Denver, Aurora, CO, United States, 3Medical Oncology, University of Colorado Denver, Aurora, CO, United States

Synopsis

Clinically, the radiation treatment (RT) is know to trigger an inflammatory response which can be beneficial for overall anti-cancer treatment efficacy. However, in pre-clinical mouse models, the tumor response to the RT is rather heterogenous. Our hypothesis is that tumor-associated macrophages which drive the pro-inflammatory response to the RT, are expressed differently in various mouse strains based on their genetic make-up. The goal of this study was to non-invasively assess the tumor inflammatory response to the RT based on iron oxide-induced changes in T2-MRI after injection of SPIO nanoparticles in two different mouse models with severely (NOD SCID) and moderately (nu/nu athymic) compromised immune system.

Purpose

The goal of this study was to develop quantitative T2-MRI to measure the inflammatory response triggered by the radiation treatment (RT) in tumor-bearing mice with a various degree of immune system deficiency. We hypothesize that, by taking advantage of macrophage-driven iron metabolism, it is possible to assess the level of RT-induced inflammation using superparamagnetic iron-oxide (SPIO) based T2-contrast in various models of immune-compromised animals due to their remaining ability to produce tumor-associated macrophages.

Methods

For this study, two immune-deficient mouse strains were chosen: (i) NOD SCID (non-obese diabetic/ severe combined immunodeficiency) female animals (no T-lymphocyte/ no macrophage model); and (ii) nu/nu athymic male and female mice (no T-lymphocyte/ low macrophage model). The animals were implanted with human tumor cells in the lateral hind leg based on four groups: (1) NOD SCID female with human breast cancer MDA-231 cells; (2) nu/nu female with human breast cancer MDA-231; (3) nu/nu female with human head-and-neck cancer UMSCC10; and (4) nu/nu male with human prostate cancer PC3 (Table 1). After tumors achieved a minimum volume of 150 mm3, the first quantitative MRI-map session was performed, consisting of a pre-contrast T2/ T2*-MRI and a post-contrast T2/T2*-MRI 24 hr after a tail vein injection of 15mg/kg of ferumoxytol (FDA-approved SPIO nanoparticles). Two weeks after the first MRI session, (allowing for a complete wash-out of SPIO nanoparticles), all animals underwent RT (10Gy as a single dose) using a 160 kV source (RS-2000); only the tumor-bearing hindquarter was irradiated. The second MRI session (pre- and post-) was performed using the same dose of ferumoxytol. All MRI scans were obtained on anesthetized animals using a 4.7 Tesla Bruker PharmaScan MRI with a 31mm-diamter Bruker volume coil and acquired using Bruker Paravision 4.0.1 software. After obtaining tri-pilots for ROI localization, a conventional rapid acquisition with relaxation enhancement (RARE) proton density (pd) scan was obtained for the volumetric assessment, followed by a multi-slice multi-echo (MSME, 16 echoes) T2-MRI map and a multi-gradient echo (MGRE, 12 gradient echoes) T2*-MRI map for the precise calculation of T2 and T2* relaxation times, respectively. T2-histograms were also reconstructed. Changes in relaxation times were calculated as ΔT2 = [T2post – T2pre]. At the end of post-contrast MRI scans the mice were sacrificed and tumors harvested for flow cytometry on macrophages and other immunocompetent cells, colorimetric iron quantification and F480/Cd11/Prussian blue immunohistochemistry staining for ex vivo correlates.

Results

All three human cell lines are well characterized as primary pro-inflammatory tumors. However, no drop in T2 after injection of ferumoxytol was seen at the Baseline (prior to RT) in any of animals (Table 1, Figure 1). After RT treatment, the severely immune-deficient (SCID) mice still failed to accumulate iron in their MDA-231 tumors as evident from ΔT2 of <-3ms, indicating no inflammatory response to RT due to their host impaired macrophage production (Figure 2). On the other hand, all three cancer xenografts (MDA-231, UMSC110, and PC3) hosted by partially immune-deficient nu/nu mice, revealed significantly reduced T2 (and T2*) relaxation times after injection of ferumoxytol following the RT (Figure 2). Immunohistochemistry showed no macrophage/ Prussian Blue iron staining in MDA-231 tumors from NOD SCID mice, while a moderate co-localization was seen for Cd11/ Prussian blue staining in MDA-231 xenografts from nu/nu hosts.

Discussion

Most recently, we have shown that highly inflamed pregnancy-associated breast cancer murine models show a significant drop in T2 relaxation times after ferumoxytol injection, due to the biological SPIO uptake by macrophages. Iron is well known T2-relaxing agent. However, our previous studies were performed in immune-competent animals. The RT is a well-known inflammatory trigger, however pre-clinical data show inconsistent response to the RT in various murine models. Here we demonstrate, using non- invasive SPIO-enhanced T2-MRI, that the RT-induced inflammatory response (which might contribute to RT therapeutic efficacy) is absent in a severe immune-deficient SCID model, due to the impaired host ability to produce macrophages. The athymic nu/nu mice are lacking the pro-inflammatory milieu prior to irradiation, but are able to produce tumor-associate macrophages in response to the RT.

Conclusions

Our study shows that NOD SCID mice are not an appropriate model to study the inflammatory microenvironment as a murine model of cancers. Additional studies will be conducted to correlate T2-MRI changes and tumor-associated macrophage levels with tumor growth after the RT. Clinically, the decreased T2-relaxation times after injection of ferumoxytol can be non-invasively observed in human tumors as a quantitative end-point for the inflammatory response to the RT and, most probably, to other novel immune-based anti-cancer therapies.

Acknowledgements

No acknowledgement found.

References

No reference found.

Figures

Table 1: Changes in T2 relaxation time 24hr after injection of 15 Femg/kg ferumoxytol

T2-MRI on NOD SCID MDA-231 xenograft at the baseline (no RT treatment)

T2-MRI on NOD SCID and nu/nu mice after Radiation Treatment



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0082