EMIM 2019
To search for a specific ID please enter the hash sign followed by the ID number (e.g. #123).

Onco-Immunology Imaging

Session chair: Nicolas Beziere (Tübingen, Germany); Filippo Galli (Terni, Italy)
Shortcut: PW07
Date: Wednesday, 20 March, 2019, 4:00 p.m.
Room: ALSH | level 0,BOISDALE | level 0,CARRON | level +1,DOCHART | level +1
Session type: Poster


Click on an contribution to preview the abstract content.


Imaging of therapy-induced immune activation in glioma by [18F]DPA-714 (#393)

Claudia Foray1, 4, 11, Silvia Valtorta7, 8, 9, Cristina Barca1, 4, 11, Oliver Grauer10, 11, Michael Schäfers1, 5, 6, Rosa M. Moresco7, 8, 9, Andreas Jacobs1, 3, 4, Bastian Zinnhardt1, 2, 4

1 European Institute for Molecular Imaging - EIMI, Münster , Germany
2 Imaging Neuroinflammation in Neurodegenerative Diseases (INMIND) EU FP7 consortium, Münster , Germany
3 Department of Geriatrics, Johanniter Hospital, Evangelische Kliniken, Bonn, Germany
4 PET Imaging in Drug Design and Development (PET3D), Münster , Germany
5 Cells in Motion (CiM) Cluster of Excellence, Münster , Germany
6 University Hospital Muenster, Department of Nuclear Medicine, Münster , Germany
7 Tecnomed Foundation and Medicine and Surgery Department, University of Milano-Bicocca, Milan, Italy
8 Experimental Imaging Center, IRCCS San Raffaele Scientific Institute, Milan, Germany
9 SYSBIO.IT, Centre of Systems Biology, Milan, Germany
10 University Hospital Muenster, Departmen of Neurology, Münster, Germany
11 Westfälische Wilhelms-Universität Münster, Münster, Germany


Glioblastoma is the primary malignant brain tumor in adults with a limited overall survival of only 14-15 months. The heterogeneous nature of gliomas and the dynamic interplay of tumor cells and immune cells in the disease course make the development and application of novel treatments challenging. To improve the understanding of the complex interaction of the tumor and the immune component of gliomas, we employed non-invasive PET/MRI using a combination of [18F]FET (aa-metabolism) and [18F]DPA-714 (TSPO) in order to monitor temozolomide (TMZ) treatment response.


N=12 female NMRI nude mice underwent MRI and PET/CT scans after intra-striatal implantation of 2x10^5 Gli36ΔEGFR-LITG glioma cells. 10 days p.i. mice were treated with 50 mg/kg of TMZ or DMSO vehicle intraperitoneally for 5 days. Mice were subjected to PET imaging at d0 (before) and at d6 after treatment with [18F]DPA-714 (TSPO; 20 MBq; 60-80 min p.i.) and [18F]FET (amino acid transport; 10 MBq, 20-30 min p.i.). Images were co-registered. Tumor-to-background (striatum) uptake ratios (T/B) and unique area of tracer uptake were calculated using an atlas-based volumetric approach. ROIs defining the area of interest were placed using the brain atlas as reference and a threshold based on 3.5xSD of the contralateral striatum area was applied. Brains were harvested for histological analysis.


TMZ treatment significantly reduced [18F]FET-derived tumor volume (0.009±0.008; p<0.05), while in DMSO treated animal it was significantly increased at day 6 (0.022±0.007; p<0.05). The tumor volume was also significantly reduced (p<0.0001) comparing treatment outcomes at day 6 between TMZ- and DMSO-treated groups (Fig 1A). TMZ treatment significantly reduced T/B ratio [18F]FET signal overtime (0.58±0.16; p<0.05) as well as [18F]DPA-714 signal (0.74±0.3; p<0.05). Comparison between DMSO- and TMZ-treated animals at day 6 showed also a significant decrease of [18F]FET signal (p<0.0001) (Fig 1B). The volumetric analysis of unique tracer areas indicated a reduction of [18F]FET uptake, between day 0 and day 6 in TMZ treated mice, as well as in DMSO control group, but the results did not reach significance (Fig. 2A). The analysis of unique [18F]DPA-714 tracer areas has shown a significant increase in TMZ treated mice (30.6±24.9; p<0.05) while in DMSO group the uptake remained stable(Fig 2B).


[18F]FET- and [18F]DPA-714-PET with MRI allow the non-invasive assessment of glioma growth and its associated inflammation. TMZ therapy effects can be detected by [18F]FET, while [18F]DPA-714 does not seem to be suitable for this purpose. However, [18F]DPA-714 seems to shift in response to therapy to areas which are not detected by [18F]FET. Additional analysis are currently ongoing to understand the therapy effects of the inflammatory component.





This work was supported by the EU 7th Framework Programme (FP7/2007-2013) under grant agreement n° 278850 (INMiND) and Horizon2020 Programme under grant agreement n° 675417 (PET3D).

Quantitative analysis of [18F]FET and [18F]DPA-714 T/B uptake ratios and tumor volume

Fig.1 A) Quantitative analysis of the changes in tumor volume (cm3/ml) after 6 days of therapy with TMZ. B) Quantitative analysis of changes in [18F]FET and [18F]DPA-714 T/B uptake ratio after 6 days from the beginning of the therapy with DMSO (vehicle) and 50 mg/kg TMZ. Differences between the treated (TMZ) and control (DMSO) groups were tested for significance using t-Test.

Volumetric analysis of unique tracers uptake areas
Fig.2 Quantitative analysis of changes in [18F]FET and [18F]DPA-714 tracers uptake volumes in DMSO and TMZ treated groups. Differences were tested for significance using t-Test. Data are represented as mean±SD.
Keywords: glioma, imaging, therapy-response, animal model, PET

In vitro assessment of radiolabelling effects on two types of cellular immunotherapy (#497)

Cameron Lang1, Francis Man1, Alessia Volpe1, Candice Ashmore-Harris1, 3, Lindsay Lim1, Ewelina Kurtys1, Rafael Torres Martin de Rosales1, Gilbert Fruhwirth1, 2

1 King's College London, School of Biomedical Engineering and Imaging Sciences, London, United Kingdom
2 KCL & UCL Comprehensive Cancer Imaging Centre, London, United Kingdom
3 King's College London, Centre for Stem Cells & Regenerative Medicine, London, United Kingdom


Chimeric antigen receptor (CAR) T cells and γδ-T cells are emerging anti-cancer cellular immunotherapies, but their fate after administration is unclear. Non-invasive tracking of T cell therapies by nuclear imaging, either through direct or reporter gene-based (indirect) radiolabelling, can shed light over the in vivo fate of adoptively transferred cells and potentially explain therapeutic outcomes. However, the effect of radiolabelling on T cell function and survival must be evaluated. Here we assessed the effect of radiolabelling with imaging radionuclides on CAR-T and γδ-T cells.


Primary human T cells were engineered using a lentiviral construct1 to co-express the anti-ErbB family CAR T1E28z2 and the radionuclide-fluorescence reporter human sodium iodide symporter (NIS)-TagRFP3 (T4:NT). T4:NT CAR‑T were selectively expanded via the co-expressed IL-4:IL-2/15 chimera4 and characterized in vitro for viability, CAR (tumour cell killing, IFNγ release) and reporter function after radiolabelling with the NIS radiotracers [18F]BF4- and 99mTcO4- (40 kBq/106 cells/mL). Human Vγ9Vδ2 T cells were isolated from peripheral blood, expanded in vitro, and directly radiolabelled with 89Zr(oxine)4.5 In addition, for both therapeutic cell types we assessed radiation-induced DNA damage by quantifying γH2AX expression by immunofluorescence staining.


T4:NT CAR-T were generated and confirmed fully functional. Radiolabelling with either 99mTcO4- or [18F]BF4- resulted in T4:NT containing 5.0±2.0 mBq 99mTcO4- or 3.6±1.7 mBq [18F]BF4- per cell, similar to what NIS reporter expressing cells experience upon in vivo imaging with these radiotracers. Neither CAR-T viability (5d post radiolabelling) nor tumour cell killing nor IFNγ were impaired by radiolabelling (Fig.1A-D). γH2AX foci (DNA double-strand breaks) were elevated upon external beam radiation (positive control, Fig.2A) and radiolabelling with 99mTcO4- and 89Zr(oxine)4 (Fig. 2B). γδ-T cells were directly radiolabelled to reach cellular 89Zr levels of [6-20], [50-90] and [150-450] mBq/cells5. Cell viability was impaired by radiolabelling higher than [6-20] mBq/cell, while tumour cell killing capacity did not change significantly5 (Fig. 1E-F). γH2AX foci increased in a radiation-dependent manner5 (Fig. 2C-D).


Therapeutic T-cells were radiolabelled and characterised. Radiolabelling of reporter gene-expressing CAR-T was confirmed to be non-detrimental to cells. Direct radiolabelling conditions were identified that enabled γδ-T cells to remain functional5, and are therefore suitable for in vivo tracking. We identified conditions allowing safe in vivo tracking of both cell types, which can serve as guidelines for other cellular T-cell immunotherapies.


1 Volpe et al., J Vis Exp. 2018; 133; doi: 10.3791/57088.

2 Davies et al., Mol Med 2012; 18:565-76; doi: 10.2119/molmed.2011.00493.

3 Diocou et al., Sci Rep 2017; 7:946; doi: 10.1038/s41598-017-01044-4.

4 Wilkie et al., J Biol Chem 2010; 285:25538-44; doi: 10.1074/jbc.M110.127951.

5 Man et al., Mol Ther 2018 (in press); doi: 10.1016/j.ymthe.2018.10.006.


We thank Dr John Maher for the T1E28z CAR construct and Cancer Research UK, Worldwide Cancer Research, and King’s Health Partners for financial support. We declare no conflicts of interest.

In vitro characterisation of NIS-RFP:CAR-T cells and radiolabelled γδ-T cells

A Viability of 18F- and 99mTc-labelled CAR-T cells. B Radiotracer uptake of CAR-T cells with (T4:NT) or without (T4) NIS expression. C Viability of cancer cells incubated with labelled CAR-T cells. D IFNγ production by labelled CAR-T cells. E Proliferation of 89Zr-labelled γδ-T cells. F Viability of MDA-MB-231 cancer cells after incubation with 89Zr-labelled γδ-T cells. Panels E,F from ref.5.

DNA damage in radiolabelled NIS-RFP:CAR-T and γδ-T cells

A γH2AX foci quantification in CAR-T cells after external beam irradiation. B Confocal microscopy images of γH2AX foci (green) and nuclei (blue) in CAR-T cells radiolabelled with 89Zr(oxine)4 or 99mTcO4-. C γH2AX foci quantification in 89Zr-labelled γδ-T cells. D Confocal images of γH2AX foci in 89Zr-labelled γδ-T cells. Scale bars = 10 μm. Panels C,D from Man et al. (ref. 5).

Keywords: CAR-T, Immunotherapy, Cell tracking, Reporter gene, Radiobiology

Investigating the Effect of Tumour Size and Fibrosis on Trafficking of 89Zr-labelled Gamma-delta T-cells and 111In-labelled PEGylated Liposomes to Xenograft Breast Tumours (#144)

Francis Man1, Alessia Volpe1, Alberto Gabizon2, Philip Blower1, Gilbert Fruhwirth1, Rafael TM de Rosales1

1 King's College London, School of Biomedical Engineering and Imaging Sciences, London, United Kingdom
2 Hebrew University, School of Medicine, Oncology Institute, Shaare Zedek Medical Center, Jerusalem, Israel


In vivo cell tracking by nuclear imaging could improve the development and clinical efficacy of anti-cancer T-cell therapies. Gamma-delta (γδ) T cells are highly cytotoxic immune cells, used successfully in several clinical trials in cancer immunotherapy (1,2). Using the tracers 89Zr(oxine)4 for PET (3,4) and 111In(oxine)3 for SPECT, we tracked human γδ-T cells and the tumour-sensitizing liposomal alendronate (4) in a xenograft breast cancer model to investigate the effect of tumour size on cell and liposome uptake. The effect of the antifibrotic drug tranilast was also investigated (5).


Human γδ-T cells were isolated from peripheral blood and expanded in vitro using zoledronate and IL-2. MBA-MB-231.hNIS-GFP cells were injected subcutaneously in the mammary fat pad of female SCID/beige mice. For tumour sensitisation, animals were treated with liposomal alendronate (5 mg/kg) 3 days before γδ-T cell injection. Liposomes were radiolabelled with 111In(oxine)3 (7 MBq/mouse). γδ-T cells were radiolabelled with 89Zr(oxine)4 (40 kBq/106 cells) and injected in the tail vein (11x106 cells/animal). To reduce tumour fibrosis, animals were administered tranilast (200 mg/kg, oral) daily from day 7 post-implantation. Animals were imaged by SPECT or PET/CT 48 h after administration of liposomes or γδ-T cells. Biodistribution studies were performed on day 7 after γδ-T cell administration.


Tumour uptake of 89Zr-labelled γδ-T cells and 111In-labelled liposomal alendronate was visible by PET and SPECT respectively after 48 h (Fig. 1) and measured by image-based quantification. Uptake after 7 days was measured by gamma-counting (Fig. 2). Tumour uptake of 111In-labelled liposomes was 22.8±3.6 %ID/mL (n=9) after 48 h and did not appear to correlate with tumour size. After 7 days, uptake was 26.1+11.1 %ID/g and was higher in smaller tumours, with the exception of very small, poorly vascularized tumours. Uptake of 89Zr-γδ-T cells was 1.5±1.1 %ID/mL (n=20) after 48 h and correlated with tumour size (Spearman r=0.5263, p=0.0171). After 7 days, uptake increased to 3.3±1.8 %ID/g and was independent of tumour size. No significant effect of tranilast was observed on the uptake of liposomes or γδ-T cells (p≥0.05, Mann-Whitney test).


We tracked γδ-T cells and liposomes in a xenograft cancer model to investigate parameters governing their uptake. Fibrosis appeared unaffected by tranilast, possibly explaining the absence of effect on cell/liposome uptake. Unlike liposomes, γδ-T cell uptake depended on tumour size at 48 h only, suggesting T-cell uptake is not solely governed by vascular permeability. Future studies will help elucidate the in vivo dynamics of γδ-T cell therapies.


(1) Silva-Santos B et al. Nat Rev Immunol (2015) 15:683-91

(2) Fisher JPH et al. Oncoimmunology (2014) 3(1):e27572

(3) Charoenphun P et al. Eur J Nucl Med Mol Imaging (2015) 42:278-87

(4) Man F et al. Mol Ther (2018) (in press). DOI: 10.1016/j.ymthe.2018.10.006

(5) Papageorgis P et al. Sci Rep (2017) 7:46140


This work was supported by Cancer Research UK (CRUK), the KCL/UCL Comprehensive Cancer Imaging Centre funded by CRUK and EPSRC in association with the MRC and DoH (England), the Wellcome EPSRC Centre for Medical Engineering at KCL, the Medical Research Council Confidence in Concepts scheme, the Experimental Cancer Medicine Centre at KCL, the KHP/KCL CRUK Cancer Centre and the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust and KCL. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health. The authors declare no conflict of interest.

PET and SPECT tracking of gamma-delta T cells and liposomes
(A) 111In in tumour-bearing mice 48 h post-injection of 111In-labelled liposomal alendronate. (B) 89Zr in tumour-bearing mice 48 h post-injection of 89Zr-labelled γδ-T cells. MIP: maximum intensity projection; Sagitt: sagittal view; Liv: liver; Sp: spleen; Tum: tumour. (C) Histology: CD3+ (γδ-T) cells in formalin-fixed, paraffin-embedded excised tumours. Arrows indicate representative cells.
Quantification of gammadelta-T cell and liposome uptake

A-D: Effect of tumour size on cell and liposome uptake. Image-based quantification (A,B) and gamma-counting (C,D,E,F) measurements of 111In-labelled liposomal alendronate (A,C,E) and 89Zr-labelled γδ-T cells (B,C,F) in MDA-MB-231.hNIS-GFP tumour-bearing mice at 48 h and 7 days post-injection. E,F: effect of tranilast on cell and liposomal uptake after 7 days. ns: p≥0.05 (Mann-Whitney test).

Keywords: T-cell therapy, Cell tracking, Zr-89, Gamma-delta T cells

Pharmacokinetics of 111In-anti-mPD-L1 in immune challenged tumor-bearing mice (#74)

Gerwin Sandker1, Peter Wierstra1, Janneke Molkenboer-Kuenen1, Martin Gotthardt1, Gosse Adema2, Johan Bussink2, Erik H. J. G. Aarntzen1, Sandra Heskamp1

1 Radboudumc, Department of Radiology and Nuclear Medicine, Nijmegen, Netherlands
2 Radboudumc, Department of Radiation Oncology, Nijmegen, Netherlands


Immune checkpoint inhibitors show impressive anti-tumor efficacy in cancer patients. However, mixed treatment responses and serious side effects call for predictive biomarkers. Preclinical studies show that microSPECT/CT using radiolabeled anti-programmed death-ligand 1 (PD-L1) antibodies (Abs) can be used to quantify PD-L1 expression in vivo. However, the activation status of PD-L1+ immune cells may influence anti-PD-L1 Ab pharmacokinetics (PK). Therefore, we investigated the effects of lipopolysaccharide-mediated (LPS) immune activation on the PK and tumor-targeting of 111In-anti-PD-L1.


The effect of immune activation on anti-PD-L1 in vivo biodistributions was evaluated in three conditions; healthy BALB/c mice, Renca tumor-bearing BALB/c mice, and LPS-challenged (0.6 mg/kg body weight) Renca tumor-bearing BALB/c mice. Mice were intravenously injected with 30 or 100 µg 111In-labeled anti-mouse PD-L1. PK was assessed by taking blood samples and biodistribution was quantified by microSPECT/CT and ex vivo biodistribution studies 72 h after tracer injection. PD-L1 expression in organs of interest was evaluated immunohistochemically.


There were no statistically significant differences in the in vivo biodistribution of 111In-anti-PD-L1 between tumor-bearing and non-tumor-bearing mice. However, in immune-challenged mice, splenic tracer uptake significantly increased compared with non-LPS challenged tumor-bearing mice (65.0±10.3 %ID/g vs. 35.2±4.9 %ID/g; p<0.001), resulting in accelerated blood clearance and reduced tumor targeting (Blood 24 hours pi: 4.2±0.3 %ID/g vs. 9.3±3.1 %ID/g; p<0.05, Tumor: 7.9±5.3 %ID/g vs. 25.9±11.6 %ID/g; p<0.05). Increasing the tracer dose to 100 µg resulted in reduced splenic uptake (27.3±4.9 %ID/g vs. 12.1±3.5 %ID/g; p<0.05) and slower blood clearance (Blood 24 hours pi:12.2±2.5 %ID/g vs. 14.3±2.6 %ID/g; ns), and restored tumor targeting (18.1±1.7 %ID/g vs. 15.3±4.3 %ID/g; ns).


This study shows that systemic inflammatory responses can significantly alter PK and tumor targeting of anti-PD-L1 Abs. Increasing the anti-PD-L1 Ab dose saturates splenic uptake and restores efficient tumor targeting. This information is essential to better understand alterations in in vivo anti-PD-L1 Ab biodistribution and to avoid suboptimal Ab-dosing.

SPECT/CT images of 111In-anti-mPD-L1 in vehicle and LPS treated mice
Figure 1. MIP thresholded microSPECT/CT images at 72 hours post injection of 111In-anti-mPD-L1 in vehicle treated or LPS challenged tumor-bearing BALB/c mice. Increased splenic uptake and decreased tumor targeting can be observed in the LPS treated mice. Increasing anti-PD-L1 dose restores tumor targeting.
Keywords: Immune checkpoint, 111In-anti-mPD-L1 Ab, Immune challenge, Pharmacokinetics, Tumor targeting

In vivo live imaging of human T/B cell lymphoma cross-linking mediated by bispecific CD20-TCB antibody (#183)

Elena Menietti1, Dario Speziale1, Johannes Sam1, Stefano Sammicheli1, Stenford Chen1, Marine Richard1, Klein Christian1, 2, Pablo Umana1, 2, Marina Bacac1, 2, Mario Perro1, Sara Colombetti1

1 Roche , Pharmacology Department, pRED, Roche Innovation Center Zurich, , Schlieren, Switzerland
2 Roche , Cancer Immunotherapy Discovery, pRED, Roche Innovation Center Zurich, , Schlieren, Switzerland


Cancer Immune Therapies have shown unprecedented results in improving tumor control [1-3]. However, many patients are still refractory to treatment. A deeper understanding of the mode of action of the different CITs sub-classes may help improving therapeutic approaches to reach better anti-tumor response. For this reason, we developed a multi-photon intra-vital microscopy (MP-IVM) approach to study in vivo, at single cell level, the tumor microenvironment upon treatment with CD20-targeting T-cell bispecific antibodies (TCB) [4] in a preclinical model of diffuse large B cell lymphomas (DLBCL)


To selectively monitor clinical lead molecules in the context of human T cell responses, we developed a skinfold chamber model [5] in last generation humanized mice [6] that allows visualization, by MP-IVM, of labelled human T cells co-injected intra-dermally with WSU-DLCL2, a human DLBCL. We have used this model to investigate T cells recruitment to tumors upon CD20-TCB therapy: by intra-venously injecting labeled T cells in mice treated with selected blocking antibodies, we were able to identify dedicated pathways induced by CD20-TCB and regulating T cell influx into the tumor bed. Furthermore, we developed a user-independent quantification platform to assess changes in the dynamics of T cell motility and time of interaction with tumor cells


We have developed an experimental preclinical model that aims to reduce xenoreaction (human T cell reaction against mouse tissue) by utilizing T cells derived from humanized mice, educated within murine thymus. We demonstrate that such model is optimal to quantify human T cell dynamics in vivo. We show that CD20-TCB localizes in the tumor and acts on tumor-resident T cell motility within 1 hour post i.v. injection (defined as functional PK), causing a sharp reduction in their speed (from 4 to 2 µm/min) and an increase in tumor/T cell interaction time; those changes last up to 72h post-treatment. In addition, we prove how the initial tumor/T cell interaction mediated by CD20-TCB lead to peripheral T cells recruitment into the tumor. This mechanism is dependent on the presence of tumor-resident T cells and on IFNg-CXCL10 pathway. Inhibiting any of these two parameters resulted in reduced T cells infiltration from the periphery and reduced anti-tumor efficacy


We developed a reliable imaging and analysis pipeline to investigate in vivo T cell dynamics and recruitment and applied it to the study of CD20-TCB treatment of DLBCL model. Our approach has shed new lights into the MoA of this new class of immune-therapeutics, demonstrating that the IFNɣ-CXCL10 pathway is involved in T cell recruitment upon CD20-TCB treatment


  1. Yousefi, H., et al., Expert Rev Clin Immunol, 2017. 13(10): p. 1001-1015.
  2. Pishko, A. and S.D. Nasta, T Transl Cancer Res, 2017. 6(1): p. 93-103.
  3. Emens, L.A., et al., Eur J Cancer, 2017. 81: p. 116-129.
  4. Bacac, M., et al., Clin Cancer Res, 2018. 24(19): p. 4785-4797.
  5. Koehl, G.E., A. Gaumann, and E.K. Clin Exp Metastasis, 2009. 26(4): p. 329-44.
  6. Shultz, L.D., F. Ishikawa, and D.L. Greiner, Nat Rev Immunol, 2007. 7(2): p. 118-30.


We would like to acknowledge the oDTA of Roche pRED for the support of the project. We would like to thanks our colleagues of Large Molecule Research for providing the therapeutics, their  fluorescent conjugation as well as the genetically modified fluorescent cells. We would like to thanks the informatic teams for the help with analysis of the data and maintenance of the informatic infrastructure

Keywords: In vivo Imaging immunotherapy, T cell bi-specific, cancer immunotherapy, humanize mouse model, multi-photon

Comparison between PET and MRI techniques for tracking T-cells in an anti-PD1 immunotherapy animal model (#341)

Solenne Vaillant1, 2, Erwan Selingue2, Sébastien d’Heilly1, Françoise Geffroy2, Erwan Jouannot1, Sébastien Mériaux2

1 Sanofi, Bioimaging, Vitry-Sur-Seine, France
2 CEA, NeuroSpin, Gif-sur-Yvette, France


With the breakthrough of new innovative cell-based therapies, molecular imaging could be increasingly used in clinical practice to non-invasively predict the efficiency of treatments and in preclinical studies to better understand the mechanisms involved. For this purpose, the present study aims to implement specific methodologies to label immune cells for their detection by either nuclear imaging or MRI. The more appropriate labeling technique will be used to characterize an anti-PD1 immunotherapy animal model.


Human T cell line was labeled with 89Zr and USPIO. To assess radiosensitivity and early tissue distribution, 12 SCID mice were randomized in 4 groups and injected via the tail vein (IV) or peritoneal (IP) route with labeled cells. Images were acquired from 20 minutes to 11 days after injection on preclinical PET/CT (Inveon, Siemens) and 11.7 T MRI (BioSpec, Bruker) scanners. MRI sequences (T2-wieghted RARE and MGE) were acquired with a volume radiofrequency coil dedicated to mouse body.

To obtain an anti-PD1 animal model, 6 C57BL/6 mice were implanted with MC38 murine tumor and after 10 days, injected (IP) with murine T cells labeled with 89Zr and randomized in 2 groups (treated with anti-PD1 antibody or with control isotype). Mice were imaged during 3 days with PET/CT scanner.


For both labelings, in vitro studies showed viability and cell proliferation similar to the control group without labeling (Fig. 1A and 1B). However, different results have been obtained for the stability study: the MRI labeling was found stable (decrease of fluorescence signal inside cells correlated with cell proliferation), whereas an efflux of 50% in three days was observed for the nuclear labeling (Fig. 1C and 1D).

For in vivo experiments, as expected, both labelings revealed migration of cells first in the lungs, then in the liver after IV injections and in the spleen for IP injections (Fig. 1 C/D/E/F).

Finally, for the anti-PD1 animal model, radioactivity was found in the spleen for isotype group, but close to the noise in the tumor for both groups (Fig. 2). However, after quantification, radioactivity seems to increase in the tumor for anti-PD1 group (Fig. 2 A). Furthermore, tumor of anti-PD1 group presents more radioactivity after ex vivo counting than control group (Fig. 2 B).


This study presents two complementary methods to track immune cells in vivo: nuclear imaging enables to track cells with higher sensitivity, whereas MRI may enable to track cells more precisely. Finally, 89Zr labeling was applied to an anti-PD1 model and, if the images didn’t show significant increase of radioactivity in the tumor for treated mice, quantification and ex vivo analysis confirm that T cells are recruited by the tumor with treatment.


1. Bansal, A. et al. Novel 89Zr cell labeling approach for PET-based cell trafficking studies. EJNMMI Research 5, 19 (2015).

2. Krüger, K. & Mooren, F. C. T cell homing and exercise. Exerc Immunol Rev 13, 37–54 (2007).

3. Fischer, U. M. et al. Pulmonary Passage is a Major Obstacle for Intravenous Stem Cell Delivery: The Pulmonary First-Pass Effect. Stem Cells Dev 18, 683–691 (2009).


France Life Imaging is acknowledged for funding the 11.7 T preclinical MRI scanner of NeuroSpin.

Figure 1. Comparison between PET and MRI techniques for tracking T-cells

In vitro studies. Efflux for PET labeling (A) and MFI per cell for MRI labeling (B, red) compared with cell proliferation (B, blue).

In vivo studies. MRI: C (resp. D) shows RARE image of mouse abdomen 24h after IV (resp. IP) injection of labeled cells. Blue (resp. red) arrows show the spleen (resp. liver). PET: E (resp. F) shows image of mouse 24h after IP (resp. IV) injection of labeled cells.

Figure 2. Tracking T-cells with PET imaging in an anti-PD1 immunotherapy animal model
PET/CT images of mice bearing MC38 tumor (red arrows), injected with murine T cells labeled with 89Zr and treated with an anti-PD1 antibody or control isotype. Graphics represent the evolution of volume radioactivity in tumor (A) for the anti-PD1 (blue) and isotype (red) groups and the residual radioactivity (B) in organs after sacrifices.
Keywords: Immunotherapy, Cell tracking, 89Zr labeling, USPIO labeling, anti-PD1 treatment

Nanobody-based imaging of Macrophage Mannose Receptor expressing Tumor-Associated Macrophages by Fluorescence Molecular Tomography (#136)

Marco Erreni1, Evangelia Bolli4, 5, Roberta Avigni2, Francesca D'Autilia1, Andrea Doni1, Paola Allavena2, 3, Cecilia Garlanda2, 3, Alberto Mantovani2, 3, 7, Sophie Hernot6, Jo Van Ginderachter4, 5

1 Humanitas Clinical and Research Center, Unit of Advanced Optical Microscopy, Rozzano, Italy
2 Humanitas Clinical and Research Center, Immunology and Inflammation, Rozzano, Italy
3 Humanitas University, Rozzano, Italy
4 Vrije Universiteit Brussel, Lab of Cellular and Molecular Immunology, Brussels, Belgium
5 VIB Center for Inflammation Research, Lab of Myeloid Cell Immunology, Brussels, Belgium
6 Vrije Universiteit Brussel, Lab for in vivo cellular and molecular imaging, ICMI-BEFY/MIMA, Brussels, Belgium
7 The William Harvey Research Institute, Queen Mary University of London, London, United Kingdom


Tumor-associated macrophages (TAMs) have been shown to impact tumor progression by promoting angiogenesis, metastasis and tumor-immune suppression. In this context, targeting TAMs represents an effective prognostic and therapeutic anti-tumor approach.

Nanobodies (Nbs), the smallest available antigen-binding fragments derived from Camelid heavy-chain-only antibodies, are suitable for tumor targeting, due to their fast kinetics, low background and deep tissue penetration. 99mTc-labeled anti-mannose receptor (MMR)-Nbs have been successfully used to image pro-angiogenic MMR+-TAMs in vivo.


IRDye680RD-conjugated anti-MMR Nbs (IRDye-MMR-Nb) were used to image MMR+-TAMs in a mouse model of transplanted fibrosarcoma by Fluorescence Molecular Tomography (FMT). MN/MCA cells were intramuscularly injected in the mouse hind leg and MMRTAMs visualized every week, for 4 weeks. Hereto, IRDye-MMR-Nb was injected intravenously and animals imaged 1h post-injection. Intraperitoneal pre-injection with 20x molar excess of unconjugated-bivalent anti-MMR Nb (biMMR-Nb) was used to minimize the binding of IRDye-MMR-Nb in specific off-tumor sites, such as liver and spleen. Every week, a cohort of mice was sacrificed, organs were collected and imaged to analyze IRDye-MMR-Nb biodistribution. Localization of MMRTAMs by IRDye-MMR-Nb targeting in liver, tumor and lung was visualized by microscopy.


IRDye-MMR-Nb accumulation in the tumor-bearing mouse hind leg compared to healthy counterpart was already visible 1 week after cancer cell injection and the signal increased along with tumor progression. Importantly, pre-injection of unlabeled biMMR-Nb reduced the uptake of IRDye-MMR-Nb in extratumoral sites (mainly liver), without affecting tumor-specific uptake. Ex vivo IRDye-MMR-Nb biodistribution analysis confirmed the in vivo results, with tumor tissue and liver displaying the highest accumulation. MN/MCA-derived tumors develop lung metastasis at week 4 after inoculation. Accordingly, a weak IRDye-MMR-Nb accumulation could also be observed in vivo in the lungs of tumor-bearing mice. This was confirmed ex vivo, where signal in the lungs strongly increased at week 4 after tumor cell injection, compatibly with the occurrence of small metastatic foci. A colocalization between IRDye-MMR-Nb and a subpopulation of F4/80tumor-infiltrating macrophages was observed by confocal microscopy


TAMs are key components of the tumor microenvironment, orchestrating different aspect of cancer progression. We showed that IRDye-MMR-Nb can be used for efficient follow-up of disease progression through the in vivo imaging of TAMs by FMT. Since MMRTAMs are a major stromal component in many cancer types, IRDye-MMR-Nb can represent a tool to monitor TAM infiltration by optical in vivo imaging in a variety of unrelated preclinical tumor models.


Tumour-associated macrophages as treatment targets in oncology.

Mantovani A, Marchesi F, Malesci A, Laghi L, Allavena P.

Nat Rev Clin Oncol. 2017 Jul;14(7):399-416. doi: 10.1038/nrclinonc.2016.217


Nanobody-based targeting of the macrophage mannose receptor for effective in vivo imaging of tumor-associated macrophages.

Movahedi K, Schoonooghe S, Laoui D, Houbracken I, Waelput W, Breckpot K, Bouwens L, Lahoutte T, De Baetselier P, Raes G, Devoogdt N, Van Ginderachter JA.

Cancer Res. 2012 Aug 15;72(16):4165-77. doi: 10.1158/0008-5472.CAN-11-2994


Macrophage polarization in pathology.

Sica A, Erreni M, Allavena P, Porta C.

Cell Mol Life Sci. 2015 Nov;72(21):4111-26. doi: 10.1007/s00018-015-1995-y.


Immuno-imaging using nanobodies.

Vaneycken I, D'huyvetter M, Hernot S, De Vos J, Xavier C, Devoogdt N, Caveliers V, Lahoutte T.

Curr Opin Biotechnol. 2011 Dec;22(6):877-81. doi: 10.1016/j.copbio.2011.06.009


This work was supported by the Italian Association for Cancer Research (AIRC), Kom op tegen Kanker, Stichting tegen Kanker, FWO and EU-COST action Mye-EUNITER.

Keywords: Nanobodies, Macrophage Mannose Receptor, Tumor-associated Macrophages, Fluorescence Molecular Tomography, Optical Imaging

In vivo imaging of CD8+ T cell tumor infiltration following radiotherapy (#78)

Peter Wierstra1, René Raavé1, Gerwin Sandker1, Janneke Molkenboer-Kuenen1, Milou Boswinkel1, Gosse Adema2, Johan Bussink2, Martin Gotthardt1, Erik H. J. G. Aarntzen1, Sandra Heskamp1

1 Radboud University Medical Centre, Radiology and Nuclear Medicine, Nijmegen, Netherlands
2 Radboud University Medical Centre, Radiation Oncology, Nijmegen, Netherlands


In recent years, cancer immunotherapy has proven shown clinical efficacy, but responses are heterogeneous across patients. Tumor characterization prior to treatment and early response monitoring are imperative for treatment personalization. One facet strongly associated with treatment response is increased infiltration of the tumor by CD8+ cytotoxic T cells. Quantitative imaging could allow response monitoring in a non-invasive manner. Presently, we investigated an 111In-labeled CD8 antibody for imaging influx of CD8+ T cells in irradiated tumors in vivo in mice bearing murine B16-F1 tumors.


C57Bl/6 mice of 12 weeks old (n=20) were injected with subcutaneous melanoma B16-F1 tumor cells on both hind legs. In half of the mice, right side tumors were irradiated with a single dose of 10 Gy. One day later, mice were intravenously injected with either: 8.1 MBq [111In]In-DTPA-anti-CD8 antibody (8.5 µg, n=10) or 8.7 MBq [111In]In-DTPA-Isotype rat IgG2A (8.5 µg, n=10), followed by SPECT/CT at day 2. After image acquisition, mice were dissected for ex vivo biodistribution, immunohistochemical analysis of CD8 in tumors, draining and distant lymph nodes, SPECT quantification and autoradiography.


Biodistribution data showed significant increased uptake of 111In-CD8 antibody in irradiated tumors compared to non irradiated tumors in different mice (10.19 ± 3.37 vs. 6.19 ± 1.36 %ID/g ± SD, p = 0.02). Uptake in left side non irradiated tumors of irradiated mice was slightly increased compared to tumors of non irradiated mice (7.76 ± 1.73 vs. 6.49 ± 1.28 %ID/g ± SD, p = 0.12 ). No statistically significantly differences in uptake were observed for the aspecific control IgG. Uptake of control IgG was significantly lower in all groups and tumors irradiated. Notably, irradiated right side tumor (10.19 ± 3.37 vs. 4.7 ± 1.04 %ID/g ± SD, p = 0.005). Tumor draining lymph nodes showed no significant difference between irradiated or non irradiated mice. Preliminary SPECT quantification shows a trend confirming the above findings but these analyses are ongoing.


In this study we demonstrate that 111In-CD8 antibody specifically targets tissues containing CD8+ cells and irradiation induces a significant increase in tracer uptake in B16-F1 tumors. This increase was not observed for the non specific IgG, suggesting a CD8 specific process instead of radiation induced changes in tumor perfusion. In the future, non invasive imaging of CD8+ T cells could assist in immunotherapy response monitoring.


This project is funded by the Netherlands Organization for Scientific Research (NWO, project number 91617039) and Dutch Cancer Society (KWF, project number 10099)

Biodistributions of CD8 antibody and control IgG in C57Bl/6 mice bearing two B16-F1 tumors
Top figure: Biodistribution of CD8 antibody. Bottom figure: Biodistribution of control IgG. All mice bearing two B16-F1 tumors on hind legs. In red: irradiated mice, in blue: non irradiated mice. Relevant organs shown.
Comparison of tumor uptake differences between CD8 antibody and control IgG
Significant difference in uptake between CD8 and its irrelevant control IgG in irradiated but not in non irradiated tumors.
Keywords: CD8, Radiotherapy, SPECT, Quantification, Monitoring

Development of a CD8 tracer for in vivo evaluation of CD8 T cell tumor infiltration during immunotherapy (#345)

René Raavé1, Milou Boswinkel1, Gerwin Sandker1, Peter Wierstra1, Erik H. J. G. Aarntzen1, Sandra Heskamp1

1 Radboud university medical center, Radboud Institute for Molecular Life Sciences, dept. of Radiology and Nuclear Medicine, Nijmegen, Netherlands


Immunotherapy is considered a hallmark in cancer treatment by its profound and durable clinical responses in patients with various types of cancer. However, only a subgroup of patients responds to immunotherapy and methods for accurate early response monitoring are lacking. Noninvasive quantitative imaging of CD8+ cytotoxic T cells can provide dynamic and spatial information of anti-tumor response. In the present study we characterized an 111In-labeled anti-mouse CD8 antibody for imaging of tumor infiltrating CD8+ cytotoxic T cells in vitro and in vivo in mice bearing murine CT26 colon tumors.


An anti-mouse CD8 antibody (clone: YTS 169.4) was randomly conjugated with a 30 times molar excess of NCS-DTPA and radiolabeled with 111In. Using CD8+ TK1 mice lymphoma cells, the immunoreactivity, IC50, internalization and affinity characteristics were determined.

CT26 tumor bearing BALB/c mice (10-12 weeks old) were intravenously injected with 8.5 µg (8.5 MBq) [111In]In-DTPA-anti-CD8. One group received an excess of non-radiolabeled CD8 antibody (250 µg). SPECT/CT imaging was performed and organs were collected to quantify tracer uptake at 4h, 24h, 48h and 72h after injection. Autoradiography and immunohistochemistry were performed on paraffin embedded  tissue sections of tumor, spleen, lymph nodes and duodenum.


In vitro assays demonstrated that the immunoreactive fraction was 44%, IC50 was 1.77 nM, Kd was 3.83 nM, and 6.5% internalization of the total membrane bound activity after 4.5 h of incubation. CD8+ T cell containing organs (lymph nodes, spleen and duodenum) were clearly visible on SPECT scans of mice injected with [111In]In-DTPA-CD8-antibody at all time points (Fig. 1A). Mice that received an excess of non-radiolabeled CD8-antibody showed most uptake in the spleen (Fig . 1B). Low to moderate tumor uptake was visible in all groups. Ex vivo biodistribution data confirmed results from SPECT imaging. In the lymph nodes, spleen, duodenum and tumor, an uptake of 38.6 ± 12.3% ID/g (±SD), 87.1 ± 18.0% ID/g, 31.7 ± 16.9% ID/g, and 12.9 ±2.9% ID/g at 24h after injection, respectively. The tumor-to-blood ratio increased from 0.48 at 4h after injection to 2.23 at 72h after injection. Autoradiography and immunohistochemistry confirmed these findings.


The CD8 antibody showed specific uptake in CD8+ T cell containing tissues in vivo, but uptake in the tumor was limited because of presence low number of CD8+ T cells. In the future, this tracer has potential for in vivo evaluation of CD8+ T cell infiltration in tumors and lymphoid tissues before and during immunotherapy.


This project receives funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 116106. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation program and EFPIA.

Figure 1. SPECT/CT images of [111In]In-DTPA-antimouse-CD8 antibody in BALB/c mice with CT26 tumors
CT26 tumor bearing mice were injected with 111In labeled CD8 antibody and imaged with SPECT/CT at 4, 24, 48, and 72 h after injection (A). One group was co-injected with 250 µg unlabeled antibody (B). Tumors are indicated with white arrows. Images (ventral view) are manually scaled.
Keywords: CD8 T-lympocytes, Tracer development, immunotherapy, Monitoring, in vivo

T-cell tracking using Cerenkov and radioluminescence imaging (#131)

Federico Boschi1, Francesco De Sanctis2, Stefano Ugel2, Antonello Spinelli3

1 University of Verona, Department of Computer Science, Verona, Italy
2 University of Verona, Department of Medicine, Verona, Italy
3 San Raffaele Scientific Institute, Experimental Imaging Centre, Milan, Italy


Cancer immunotherapy includes a wide group of treatments focused to restrict tumor growth and possibly induce cancer remission by manipulating and restoring a functional immune response in the host. The availability of imaging technologies that follow in vivo T-cell trafficking in tissues and functionality after infusion in the host are mandatory to predict the outcome of an adoptive cell transfer (ACT) based immunotherapy approach.

Here, we report the potentialities of optical detection of radiolabeled T-cells via Cerenkov luminescence imaging (CLI) and radio-luminescence imaging (RLI).


Mixed leukocyte peptide cultures (MLPCs): spleens from OT-I CD45.1 mice were collected and mashed to isolate splenocytes. Then, OT-I splenocytes were plated at a concentration of 5 × 10^6/well in 24-well plates in complete media in the presence of 1 μg/mL OVA257–264 peptide and 20 IU/mL interleukin-2 to stimulate the proliferation of OVA-specific CD8+ CD45.1+T lymphocytes. The culture was split every other day andreplenished with fresh media containing 20 IU/mL interleukin-2. After 5 days of culture, MLPC was washed twice and dead cell were removed by Ficoll under standard procedures. T-cells were labeled with 32P-ATP for 1 hour at 37C and then washed in phosphate buffered saline.

CLI and RLI acquisitions were performed by using the in vivo imaging systems (IVIS) Spectrum optical imager.


We found that T-cell biodistribution can be followed during in vivo acquisitions for hours after cells injection, in particular we described a new method to in vivo follow T lymphocyte trafficking without sacrificing the animals based on CLI and RLI (see figure 1).The light signal is well detectable directly with CLI without the use of any other device allowing to track the biodistribution of T-cells. Instead, by using scintillator materials, the light emission coming from the conversion of radioactivity in light increases.


We conclude that 32P-ATP T-cell labeling can provide several benefits: it allows the imaging of T-cell trafficking and supports the cytotoxic activity on tumor cells by the “cross fire” exerted by the high-penetrating beta particles of 32P-ATP.

Figure 1

Cerenkov luminescence images of mice treated with 32P-ATP labeled T lymphocytes at different time points (from 30 minutes to 18 hours) after i.v. (left) and i.p. (right) injection.

Keywords: optical imaging, Cerenkov luminescence imaging, adoptive cell transfer immunotherapy, cancer imaging

Fixed and live imaging of neutrophils in the lung pre-metastatic niche (#565)

Amanda McFarlane1, Judith Secklehner1, Ximena Raffo1, John Mackey1, Grant McGregor1, David Novo1, Louise Mitchell1, Jen Morton1, Jim Norman1, Leo Carlin1

1 The Beatson Institute for Cancer Research, Glasgow, United Kingdom


The establishment of the premetastatic niche (PMN) during cancer is highly complicated, involving many factors including the immune system and the extracellular matrix (ECM). The lung is a complex and highly vascularized organ prone to metastasis. Previous data has revealed changes in ECM deposition in the lung during early pancreatic cancer prior to metastasis (1). We investigated if such changes in ECM organisation might affect immune cell composition, localisation and trafficking in the lung prior to and during the formation of the PMN.


KPC (Pdx1-Cre, KrasG12D/+, p53R172H/+), KFC (Pdx1-Cre, KrasG12D/+, p53fl/+), KC (Pdx1-Cre, KrasG12D/+) and WT (Pdx1-Cre) mice as previously described (2) were used. When a small palpable pancreatic tumour was detected in experimental mice, they were humanely killed and the lungs inflated with agarose, before producing precision cut lung slices (PCLS) cut at 300μm thickness using a vibratome. For live PCLS, slices were stained and imaged directly using the Zeiss 880 Airyscan Fast confocal microscope to produce time-lapse images which were segmented and tracked. For fixed PCLS imaging a Perkin Elmer Opera Phenix was used to evaluate whole lung slices by high-speed image tiling and high content segmentation and analysis.


Increased leukocyte numbers were observed in the lungs of mutant p53 bearing mice, with an increase in the ratio of myeloid to lymphoid cells. Fixed PCLS from KPC (highly invasive & metastatic), KFC (non-metastatic), KC and WT (controls) were stained to localize neutrophils within the tissue in the context of the intact stroma. Imaging of live PCLS revealed many more neutrophils present in KPC lungs compared with controls as expected. Analysis of their movement and trafficking through the vasculature has revealed differences which may be as a result of the changes in the ECM of these mice. High content analysis of the neutrophil positions across whole lobe PCLS allowed us to observe and quantify heterogeneity in the localisation of leukocyte subsets.


Investigating the formation of the PMN aids our understanding of metastasis and potential targets for therapy. We have imaged fixed and live mouse PCLS to investigate both localisation and trafficking of neutrophils in the lung vasculature of pre-metastatic pancreatic cancer. Identifying the localisation and motility of different leukocyte subsets in the lung in the context of intact stroma may shed light on their function in metastasis.


(1) Hingorani, S. R. et al. Trp53R172H and KrasG12D cooperate to promote chromosomal instability and widely metastatic pancreatic ductal adenocarcinoma in mice. Cancer Cell 7, 469–483 (2005).

(2) Novo, D. et al. Mutant p53s generate pro-invasive niches by influencing exosome podocalyxin levels. Nature Communications 9, 5069 (2018).

Keywords: pre-metastatic niche, lung, leukocytes, high content imaging

Imaging the tumor microenvironment in vivo – origin and regulatory mechanisms of tumor promoting S100A9 (#550)

Anne Helfen1, Jan Rieß1, Olesja Fehler2, Annika Schnepel1, Miriam Stölting1, Mirjam Gerwing1, Max Masthoff1, Walter Heindel1, Moritz Wildgruber1, 3, Michel Eisenblätter1, 4

1 University of Muenster, Institute of Clinical Radiology, Muenster, Germany
2 University of Muenster, Institute of Immunology, Muenster, Germany
3 University of Muenster, DFG EXC 1003 Cluster of Excellence `Cells in Motion`, Muenster, Germany
4 King´s College London, Division of Imaging Sciences & Biomedical Engineering, London, United Kingdom


Tumor progression and metastasis depend on tumor-infiltrating immune cells, which form a characteristic inflammatory tumor microenvironment (TME). Within this microenvironment but also in premetastatic niches, the protein heterodimer S100A8/A9 is released by activated infiltrating monocytes [1]. As a promoter of tumor invasion and TME formation, it has been associated with poor prognosis [2]. Our aim was to develop an in vivo imaging tool serving as a biomarker for TME influence on tumor behavior.


From syngeneic murine breast cancer tumors 4T1 (highly malignant) and 67NR (low malignancy), wildtype (wt) and S100A9 knock out cells (CRISPR/cas9-method, ko) were created and implanted into either female BALB/c wildtype or S100A9-/--mice (n=10 each). At 4 mm tumor diameter, anti-S100A9-Cy5.5-driven flurorescence reflectance imaging has been performed 0 and 24h after injection (contrast-to-noise ratios of fluorescence intensities in arbitrary units, AU). An isotype IgG-Cy5.5 served as a control (n=5) for unspecific label distribution in 4T1 tumors. In vivo imaging was correlated with immunohistology, Western Blot and FACS analyses. Statistical analysis was performed using unpaired t test and one-way ANOVA with Bonferroni post test.


24h after injection, anti-S100A9-Cy5.5 showed significantly higher fluorescence signals as compared to IgG-Cy5.5.

In S100A9-/--mice, fluorescence signals were significantly reduced for wildtype (4T1: 62,9 vs. 43,02AU;p=0,048, 67NR: 52,73 vs. 27,26;p=0,033) and S100A9-/--tumors (4T1ko: 65,33 vs. 42,51AU;p=0,005; 67NRko: 54,61 vs. 34,19AU;p=0,006, Figure). However, no significant differences were detected for 4T1ko and 67NRko cells as compared with wildtype cells (4T1:p=0.543, 67NR:p=0.85). In all subgroups, the fluorescence signal was significantly higher in 4T1 as compared to 67NR tumors, reflecting their higher malignant potential.Imaging results were confirmed by ex-vivo analyses.

Furthermore, FACS of 4T1 tumors revealed a significantly lower amount of CD115+/F4/80+ cells (15,8 vs. 30,9%;p=0,001) and a higher of Ly6C+ cells (82,2 vs. 47,53%;p=0,001) in wildtype mice as compared to S100A9-/--mice, indicating an immune cell shift to immature monocytes within the S100A9-positive TME.


Our results in the 4T1/67NR breast cancer model system confirm a secretion of S100A8/A9 by components of the TME, while tumor cells do apparently not release S100A8/A9. S100A9-specific in vivo imaging reflects tumor malignancy and may serve as a biomarker for TME formation and activity.


1) Arai, K., et al. (2008) S100A8 and S100A9 overexpression is associated with poor pathological parameters in invasive ductal carcinoma of the breast. Curr Cancer Drug Targets.

2) Roth, J., et al. (2003) Phagocyte-specific S100 proteins: a novel group of proinflammatory molecules. Trends Immunol.


This work has been funded by DFG EXC 1003 (PP-2017-01).

S100A9-specific fluorescence signals in 4T1 and 67NR tumors

Calculated contrast-to-noise ratios of fluorescence intensities for 4T1 (highly malignant) and 67NR (low malignancy) wildtype as well as S100A9 knock out tumors in female BALB/c wildtype and S100A9-/--mice (n=10 each).

Keywords: tumor microenvironment, tumor immunology, S100A9, fluorescence reflectance imaging

Visualising pharmacodynamic changes in the tumour microenvironment with in vivo imaging (#523)

Stefanie R. Mullins1, Aimee Ruffle1, 2, Thomas Murray2, Judit Espana-Agusti3, Shannon Burke1, Nadia Luheshi1, Simon Dovedi1, Robert Wilkinson1

1 MedImmune, Research Oncology, Cambridge, United Kingdom
2 MedImmune, ADPE, Cambridge, United Kingdom
3 MedImmune, IVS, Cambridge, United Kingdom


Efficacy of cancer immunotherapy is associated with pharmacodynamic (PD) changes in the tumour microenvironment (TME). The presence of proinflammatory cytokines, including interferon-γ (IFNγ), are associated with positive therapeutic responses to immuno-oncology (IO) drugs (Higgs et al.). We have developed an imaging model that detects changes in IFNγ in the TME to measure the proinflammatory response to IO therapeutics, which will allow for measurement of real-time PD changes in the same animal and provide further insight into the differences between IO responsive and resistant tumours.


CT26 and B16-F10 AP3 cells lines responsive to IFNγ were generated using a pCDH-IFNγ-luc2P-PGK(-)-mKATE2 lentiviral vector containing a constitutive mKATE2 fluorescent reporter and IFNγ-inducible luc2P bioluminescent reporter. FACS sorted clones were characterised in vitro by flow cytometry (mKATE2) and IFNγ-induced luciferase expression (luc2P). IFNγ reporter tumour (CT26 or B16-F10 AP3) bearing mice were injected with an IFNγ neutralizing antibody. Fluorescence and bioluminescence images were taken using the IVIS Spectrum. Intratumoural IFNγ was calculated as (total flux [p/s]) divided by (total radiant efficiency [RFU]; CT26) or tumour volume [mm3] (B16-F10 AP3). IFNγ-responsiveness was confirmed ex vivo by incubating tumour tissue slice cultures with IFNγ (0-10ng/ml).


Luc2P bioluminescence was observed in all mice injected with IFNγ-reporter cells and B16-F10 AP3 transduced cells had lower baseline IFNγ response compared with CT26. Tissue slices from tumours grown in vivo, then treated with recombinant IFNγ, also demonstrated that both CT26 and B16-F10 AP3 transduced tumours were responsive to IFNγ in a dose dependent manner. Anti-IFNγ-mAb treatment in mice was associated with a significant decrease in intratumoural IFNγ in CT26 IFNγ-reporter 7 and B16-F10 AP3 IFNγ-reporter 22 tumours respectively from 3 days and 6 days post first dose.


Changes in IFNγ signalling were observed in two models using in vivo imaging. Selectivity for IFNγ was validated using in vivo imaging and ex vivo tissue slice culture. It also revealed differences in IFNγ signalling between IO-responsive (CT26) and IO-resistant (B16-F10 AP3) tumours. This TME in vivo imaging method has several merits, compared with standard PD approaches, as it is less invasive and has the potential to reduce animal numbers.


Higgs, B.W., et al., ESMO presentation, 2015

Keywords: IFNgamma, immuno-oncology, PD