EMIM 2018 ControlCenter

Online Program Overview Session: PS-21

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

Understand Tumour Biology

Session chair: Tim Witney - London, UK; Benoit Busser - Grenoble, France
Shortcut: PS-21
Date: Friday, 23 March, 2018, 8:30 AM
Room: Lecture Room 01 | level -1
Session type: Parallel Session


Click on an contribution preview the abstract content.

8:30 AM PS-21-1

Introductory Talk by David Lewis - Glasgow, UK

This talk provides an overview of state-of-the-art research and refers to the following presentations selected from abstract submissions.

8:50 AM PS-21-2

Imaging the intracellular redox environment using [18F]FSPG positron emission tomography (#74)

P. N. McCormick1, H. E. Greenwood1, M. Glaser2, K. Sander2, T. Gendron2, G. Gowrishankar3, A. Hoehne3, A. Shuhendler3, D. Lewis3, M. Berndt4, N. Koglin4, M. F. Lythgoe1, S. S. Gambhir3, E. Årstad2, T. H. Witney1

1 University College London, Centre for Advanced Biomedical Imaging, London, United Kingdom
2 University College London, Department of Chemistry, London, United Kingdom
3 Stanford University, Department of Radiology, Stanford, United States of America
4 Piramal Imaging GmbH, Berlin, Germany


Dysregulation of cellular redox biochemistry contributes to numerous pathologies, including cancer1. Currently, no non-invasive method exists to monitor intracellular redox status in vivo. The transporter xCT is crucial to antioxidant biochemistry, being the main supplier of intracellular cysteine, the rate-limiting substrate in glutathione (GSH) biosynthesis. The PET radiotracer [18F]FSPG is transported into the cell by xCT, thereby providing a direct means to investigate the function of this transporter2. Here, we evaluate [18F]FSPG PET as a sensitive marker of the tumour redox environment.


In a redox model using A2780 ovarian cancer cells, redox manipulations were made using the oxidant tert-butyl hydroperoxide (TBHP; 1h, 200 µM), the antioxidant N-acetylcysteine (NAC; 2h, 5 mM), or both in combination. Key redox indices were measured, including reactive oxygen species (ROS), GSH, and the expression of crucial proteins, with corresponding changes in [18F]FSPG uptake assessed. A panel of mechanistically diverse oxidant compounds was also used to increase ROS and further evaluate [18F]FSPG’s redox sensitivity. Finally, mice bearing subcutaneous A2780 xenografts were treated with the oxidant chemotherapeutic Doxil. Tumour volume and GSH were measured 24 h and 6 d post-treatment. At the same time points tumour [18F]FSPG uptake was measured using PET.


In our cellular redox model, treatment with TBHP caused a 5.6-fold increase in ROS (P < 0.001) and doubled the ratio of oxidised to reduced GSH (P < 0.0001; Fig. 1a); effects that were prevented by NAC pre-treatment. TBHP also increased the antioxidant transcription factor Nrf2 but did not induce apoptosis as measured by caspase-3 cleavage. Mirroring intracellular redox changes, TBHP decreased [18F]FSPG uptake by 46% relative to untreated cells (P < 0.001), while NAC doubled [18F]FSPG uptake (P < 0.0001; Fig.1b). Five other oxidant compounds caused decreases in [18F]FSPG uptake that were highly correlated with the degree of ROS induction (R2 = 0.85, P < 0.0085; Fig. 1c). In living subjects, doxil treatment caused a reduction of tumour volume that was preceded by changes in total and oxidised GSH (Fig. 2a). A 48% decrease in [18F]FSPG tumour uptake (P < 0.01) was seen early in treatment (24 h; Fig. 2b), coinciding with the changes in tumour GSH that occurred prior to tumour shrinkage.


These data demonstrate that [18F]FSPG is a sensitive intracellular redox marker. In response to mechanistically diverse oxidative stress-inducing compounds, the decrease in [18F]FSPG tumour cell uptake was highly correlated with levels of drug-induced ROS. In A2780 xenografts, [18F]FSPG PET was responsive to the early redox changes that precede tumour shrinkage, demonstrating the ability of this radiotracer to monitor tumour response to therapy in vivo. The redox sensitivity of [18F]FSPG may also have utility in other pathophysiological conditions involving dysregulated redox biochemistry.


1) Lewerenz et al. Antioxid Redox Signal 18(5):522

2) Koglin et al. Clin Cancer Res 17(18):6000

Figure 1. [18F]FSPG uptake in A2780 cancer cells mirrors intracellular redox environment.
a) Changes in reactive oxygen species (i) and oxidised glutathione (ii) with t-butyl hydroperoxide (TBHP), N-acetylcysteine (NAC) and the combined treatment. b) [18F]FSPG uptake mirrors intracellular redox changes. c) The decrease in [18F]FSPG uptake following diethylmaleate (DEM), rotenone (ROT), antimycin A (AMA), auranofin (AUR), and butein (BUT) correlates with ROS induction.
Figure 2. [18F]FSPG tumour uptake responds to Doxil-induced changes in redox environment.
a) Doxil treatment causes tumour volume reductions (i) which are preceded by changes in tumour glutathione (ii and iii) characteristic of oxidative stress. b) Representative [18F]FSPG PET images of Doxil treated mice (i), and quantification (ii) showing reduced [18F]FSPG tumour uptake as early as 24 h post-treatment, preceding tumour shrinkage and coinciding with changes in tumour glutathione.
9:00 AM PS-21-3

Pharmacological Targeting of the DNA Damage Response Pathway Improves 177Lu-PSMA617 Radioligand Therapy in a Prostate Cancer Mouse Model (#189)

A. D. Stuparu1, K. Lückerath1, S. L. Evans Axelsson1, J. R. Capri1, C. Mona1, F. Ceci1, S. Poddar1, T. M. Le1, E. Abt1, L. Wei1, M. Eiber1, W. P. Fendler1, J. Calais1, K. Herrmann1, C. G. Radu1, R. Slavik1, J. Czernin1

1 University of California, Los Angeles, Molecular and Medical Pharmacology, Los Angeles, California, United States of America


Prostate specific membrane antigen (PSMA) radioligand therapy (RLT) has emerged as a promising new treatment option for prostate cancer.1-4 However, disease relapses invariably. The DNA damage response (DDR) pathway has been implicated in mediating RLT cytotoxicity5-10, but a systematic evaluation of tumor cell stress responses is lacking. This study tests the hypotheses (i) that RLT selectively activates components of the DDR in tumors which modulate efficacy in murine models; and (ii) that pharmacological targeting of critical DDR effectors sensitizes prostate cancer xenografts to RLT.


NOD scid gamma (NSG) mice bearing bilateral C4-2 xenografts were treated with 30MBq or 120MBq (84GBq/µmol) 177Lu-PSMA617 or vehicle (formulation buffer). PSMA expression was confirmed before the start of treatment using 68Ga-PSMA11 PET/CT imaging. Tumors of a subset of mice were resected after 4h and 48h and analyzed for global proteomic and phosphoprotemic alterations using mass spectrometry (MS), as well as by IHC staining for DNA damage. In a second experiment, mice with C4-2 tumors were treated with vehicle, RLT alone or in combination with small molecule inhibitors of DDR kinases, or DDR kinase inhibitors alone. MS analyses were performed on a subset of tumors resected after 48h. In both experiments, primary endpoint for the remainder of the mice was tumor size and survival.


177Lu-PSMA617 RLT efficiently reduced tumor burden, but did not completely eradicate tumors in the C4-2 prostate cancer xenograft mouse model. Tumors eventually relapsed, which indicates the presence of molecular mechanisms that limit the efficacy of RLT. As early as 4h post-RLT, proteomic mass spectrometry analysis revealed significant alterations in 5% of the total identified proteome, with changes to cyclin D1 and p21 being consistent with established genotoxic stress response. Phosphorylation of known ATM and CDK substrates increased in RLT treated tumors. Combination of RLT with pharmacological inhibition of the DDR kinases ATM, ATR, and WEE1 improved RLT efficacy leading to reduced tumor burden and longer time to progression. However, all tumors relapsed eventually. Global and phosphoproetomic profiling of DDR kinase inhibitors and RLT treated tumors will reveal further liabilities of these tumors.


In summary, our findings indicate that 177Lu-PSMA617 RLT induces tumor cell stress responses that mediate repair of DNA damage in a prostate cancer xenograft mouse model. RLT efficacy can be improved by concomitantly inhibiting kinases in the DNA damage repair pathway. Further studies elucidating additional mechanisms of resistance to PSMA RLT in prostate cancer are highly warranted to identify vulnerabilities that can be exploited for rationally designed combination therapies.


1. Baum RP, Kulkarni HR, Schuchardt C, Singh A, Wirtz M, Wiessalla S, Schottelius M, Mueller D, Klette I,

Wester HJ. 177Lu-labeled prostate-specific membrane antigen radioligand therapy of metastatic

castration-resistant prostate cancer: Safety and efficacy. 2016; J Nucl Med 57(7) 1006-1013. doi:10.2967/


2. Rahbar K, Bode A, Weckesser M, Avramovic N, Claesener M, Stegger L, Bogemann M. Radioligand

therapy with 177Lu-PSMA-617 as a novel therapeutic option in patients with metastatic castration resistant

prostate cancer. 2016; Clin Nucl Med 41(7) 522-528. doi:10.1097/rlu.0000000000001240.

3. Rahbar K, Ahmadzadehfar H, Kratochwil C, Haberkorn U, Schäfers M, Essler M, Baum RP, Kulkarni HR,

Schmidt M, Drzezga A, Bartenstein P, Pfestroff A, Luster M, Lützen U, Marx M, Prasad V, Brenner W,

Heinzel A, Mottaghy FM, Ruf J, Meyer PT, Heuschkel M, Eveslage M, Bögemann M, Fendler WP, Krause

BJ. German multicenter study investigating 177Lu-PSMA-617 radioligand therapy in advanced prostate

cancer patients. 2017; J Nucl Med 58(1) 85-90. doi:10.2967/jnumed.116.183194.

4. Fendler WP, Reinhardt S, Ilhan H, Delker A, Böning G, Gildehaus FJ, Stief C, Bartenstein P, Gratzke C,

Lehner S, Rominger A. Preliminary experience with dosimetry, response and patient reported outcome

after 177Lu-PSMA-617 therapy for metastatic castration-resistant prostate cancer. 2017; Oncotarget 8(2)

3581-3590. doi:10.18632/oncotarget.12240. PMCID: PMC5356905.

5. Shiloh Y, Ziv Y. The atm protein kinase: Regulating the cellular response to genotoxic stress, and more.

Nat Rev Mol Cell Biol. 2013; 14(4) 197-210. doi:10.1038/nrm3546. PubMed PMID: 23486281.

6. Mazouzi A, Velimezi G, Loizou JI. DNA replication stress: Causes, resolution and disease. 2014; Exp Cell

Res 329(1) 85-93. doi:10.1016/j.yexcr.2014.09.030.

7. Branzei D, Foiani M. Maintaining genome stability at the replication fork. Nat Rev Mol Cell Biol. 2010;

11(3) 208-219. doi:10.1038/nrm2852. PubMed PMID: 20177396.

8. Maier P, Hartmann L, Wenz F, Herskind C. Cellular pathways in response to ionizing radiation and their

targetability for tumor radiosensitization. Int J Mol Sci. 2016; 17(1) doi:10.3390/ijms17010102. PubMed

PMID: 26784176. PMCID: PMC4730344.

9. Nonnekens J, Van Kranenburg M, Beerens CE, Suker M, Doukas M, Van Eijck CH, De Jong M, Van Gent

DC. Potentiation of peptide receptor radionuclide therapy by the parp inhibitor olaparib. Theranostics.

2016; 6(11) 1821-1832. doi:10.7150/thno.15311. PubMed PMID: 27570553. PMCID: PMC4997239.

10. Graf F, Fahrer J, Maus S, Morgenstern A, Bruchertseifer F, Venkatachalam S, Fottner C, Weber MM,

Huelsenbeck J, Schreckenberger M, Kaina B, Miederer M. DNA double strand breaks as predictor of

efficacy of the alpha-particle emitter Ac-225 and the electron emitter Lu-177 for somatostatin receptor

targeted radiotherapy. PLoS One. 2014; 9(2) e88239. doi:10.1371/journal.pone.0088239. PubMed PMID:

24516620. PMCID: PMC3917860.

RLT in a C4-2 prostate cancer xenograft mouse model

a. Experimental design. b. Tumor growth curves showing that both 30 and 120 MBq induced significant tumor growth inhibition. Data are mean ± SEM (n=12 tumors per group). ****P<0.001 c. Survival data (n=6 mice per group). + 120 MBq mice were sacrificed due to a vivarium contamination prior to reaching the end point. d. IHC analysis showing the formation of 53BP1 foci in tumors following RLT.

Keywords: radioligand therapy, prostate cancer, DNA damage response, 177Lu-PSMA617, PSMA
9:10 AM PS-21-4

Selective elimination of EGFR-positive glioblastoma cells with IR700DX-labeled affibody molecules (#7)

T. A. Burley1, J. Mączyńska1, A. Shah1, W. Szopa4, K. J. Harrington1, J. K. R. Boult1, A. Mrozek-Wilczkiewicz2, M. Vinci3, J. Bamber1, W. Kaspera4, G. Kramer-Marek1

1 Institute of Cancer Research, Division of Radiotherapy and Imaging, Sutton, Surrey, United Kingdom
2 University of Silesia, A. Chelkowski Institute of Physics, Katowice, Silesian Voivodeship, Poland
3 Bambino Gesù Children’s Hospital, Department of Onco-Hematology, Rome, Lazio, Italy
4 Medical University of Silesia, Department of Neurosurgery, Katowice, Silesian Voivodeship, Poland


Glioblastomas (GBM) carry a poor prognosis with a median survival of ~14 months following standard therapy.[1] Currently, standard-of-care treatment involves maximal tumour resection, however, residual glioblastoma tumour cells have been reported in ~65% of cases and are a major factor in GBM reccurrence.[2] As GBM frequently arises from mutations and overexpression of the epidermal growth factor receptor (EGFR), we therefore developed an EGFR-targeting photosensitive conjugate to not only guide fluorescence based surgery, but remove any residual cancer cells by photoimmunotherapy (PIT).


The photosensitiser, IR700DX, was conjugated to EGFR targeting affibody molecules via a maleimide group. Specificity of binding of the ZEGFR:03115-IR700DX to EGFR was confirmed by flow cytometry and confocal microscopy. The therapeutic potential of the ZEGFR:03115-IR700DX was assessed in U87-MGvIII (EGFR+++), in a human derived primary cell line WSz4 (EGFR++) and MCF-7 (EGFR+) cells and tumourspheres by the CellTiter-Glo® assay. In vivo specificity of the conjugate to EGFR was confirmed in both subcutaneous and orthotopic tumours. The response to PIT (100J/cm2, 3 doses) was evaluated in subcutaneous U87-MGvIII xenografts. Mice were imaged using the IVIS Spectrum/CT and tumour growth inhibition was monitored for 10 days, followed by tumour dissection for ex vivo analysis (IHC).


Binding of the conjugate was found to correlate with the level of EGFR expression in the GBM cell lines. The cell viability assay revealed a receptor dependent response between the investigated cell lines, with a substantial loss in U87-MGvIII cell, as well as 3D spheroid, viability when treated with nanomolar concentrations of the ZEGFR:03115-IR700DX and irradiated. High contrast images were acquired for the orthotopic GBM tumours confirming that the conjugate could cross the disrupted blood-brain-barrier (Fig1). Additionally, PIT on subcutaneous EGFRvIII+ve tumours produced a significant difference in tumour size between control and treated mice (Fig2). Immunohistochemistry allowed the visualisation of numerous regions of tissue necrosis following PIT.


The ZEGFR:03115-IR700DX enabled imaging of EGF receptor expression in vivo and displayed therapeutic efficacy in cells, that closely replicate the human GBM, and in glioma xenografts. These results are particularly important in light of recent findings demonstrating that EGFRvIII+ve GBM cells are responsible for the survival of non-EGFRvIII expressing tumour cells, indicating that targeting these residual cells could improve GBM prognosis.[3]


1. McNamara, M.G., et al., Conditional probability of survival and post-progression survival in patients with glioblastoma in the temozolomide treatment era. J Neurooncol, 2014. 117(1): p. 153-60.

2. Stummer, W., et al., Fluorescence-guided surgery with 5-aminolevulinic acid for resection of malignant glioma: a randomised controlled multicentre phase III trial. Lancet Oncol, 2006. 7(5): p. 392-401.

3. Zanca, C., et al., Glioblastoma cellular cross-talk converges on NF-kappaB to attenuate EGFR inhibitor sensitivity. Genes Dev, 2017.


This work was supported by the funding from The Institute of Cancer Research and Cancer Research UK-Cancer Imaging Centre (C1060/A16464). The authors gratefully thank AffibodyAB (Stockholm, Sweden) for supplying the affibody molecules. We owe special thanks to Chiara Da Pieve for technical support and ongoing advice, Daniela Ciobota for technical assistance and Richard Symonds-Tayler for building the electronics box which provides the digital power control for the LED (The Institute of Cancer Research, London, UK). Kevin Harrington acknowledges the support of the ICR/RM NIHR Biomedical Research Centre. We thankfully acknowledge Frank Furnari for donating the U87-MG and U87-MGvIII cell lines (Ludwig Cancer Research, San Diego, USA). We also thank Piotr Czekaj and Marcin Michalik (Medical University of Silesia, Katowice, Poland) for immunohistochemistry and technical support.


Figure 1: ZEGFR:03115-IR700DX accumulates in U87-MGvIII orthotopic glioma tumours

(a) T2-weighted MRI sequences of an intracranial model of brain tumour 11 days post cell implantation.

(b) Photographic image of the brain and the corresponding fluorescent image of ZEGFR:03115-IR700DX.

(c) ZEGFR:03115-IR700DX clearly delineated tumour mass from the surrounding normal tissues which correlated well with H&E and EGFR staining of the consecutive sections.

Figure 2: In vivo ZEGFR:03115-IR700DX-mediated PIT studies
Tumour growth inhibition of the ZEGFR:03115-IR700DX-targeted PIT in U87-MGvIII subcutaneous tumours after administering three doses of 18 µg of the conjugate and irradiating with 100 J/cm2 at days 1, 3 and 5 in comparison to control groups. Data are presented as mean ± SD (n = 6 for each group, ** P ≤ 0.01 as assessed by the Kruskal-Wallis test)
Keywords: Affibody, Photoimmunotherapy, glioblastoma, EGFR
9:20 AM PS-21-5

3D printing of a biomimetic tumor angiogenesis model (#445)

B. Theek1, F. De Lorenzi1, J. Schöneberg2, N. Güvener1, H. Fischer2, F. Kiessling1

1 Uniklinik RWTH Aachen, Institute for Experimental Molecular Imaging, Aachen, Germany
2 Uniklinik RWTH Aachen, Department of Dental Materials and Biomaterials Research, Aachen, Germany


A realistic 3D in vitro tumor model can provide mechanistic insights into tumor-associated angiogenesis. Tumor angiogenesis is an essential step for tumor growth and metastasis1. Current in vitro models are either focusing on the development of multicellular tumor spheroids in static conditions or are using much simpler tumor models in dynamic conditions2. Here we want to dynamically cultivate tumor spheroids in close proximity to a biomimetic feeding vessel to generate an in vitro tumor angiogenesis model for angiogenesis research and drug screening applications.


To generate the biomimetic feeding vessel we developed a new printing strategy using a drop-on-demand bioprinter. An endothelial cell-laden gelatin is printed into a custom-designed PEEK bioreactor and subsequently coated with fibrin gel containing smooth muscle cells (SMC). The printed channels are surrounded by a fibrin/collagen hydrogel blend containing fibroblasts, in which tumor spheroids can be positioned. After gelatin liquefaction, the endothelial cells (EC) will sediment and grow on the channel wall. Finally, after the liquid gelatin is removed, the bioreactors are connected to a pump and dynamically cultivated under physiological flow conditions for up to 3 weeks. In addition, the CT and MRI compatibility of our reactor was tested.


Figure 1A shows a schematic section of the custom-designed bioreactor in which our tumor model is printed and cultivated. The feeding vessel can be designed either with a single layer of EC (Fig 1 B) or with an additional SMC coating (Fig 1C). In both cases, the CD31 staining highlights the formation of a continuous endothelium already after 4 days under dynamic cultivation, surrounded by elongated fibroblasts (Fig 1D) or a thick layer of SMC (Fig 1E). Further characterization of the EC monolayer revealed the expression of VE-cadherin and collagen IV. Fig 1F and G, respectively, display the homogenous distribution of HUVECs along the channel and sprouts of EC into the surrounding hydrogel. Furthermore, initial tests show that our reactor is CT and MRI compatible and suggest that our model can be used for imaging experiments as well. Whereas the CT measurements allow to assess the general shape of the feeding vessel, MRI measurements allowed to image FLUSPIO labelled EC (Figure 2).


Our new multilayer printing strategy enables the generation of a biomimetic feeding vessel, which can be dynamically cultivated for several weeks. The positioning of tumor spheroids close to the vessel will enable the analysis of tumor-associated angiogenesis. Furthermore, we are aiming to use this tumor model as a more reliable pre-screening platform for chemotherapeutic drugs in future.


1 Weis SM, Cheresh DA. Tumor angiogenesis: molecular pathways and therapeutic targets. Nat Med. 2011. 17(11):1359-70

2 Amann A, Zwierzina M, Koeck S, et al. Development of a 3D angiogenesis model to study tumour – endothelial cell interactions and the effects of anti-angiogenic drugs. Sci Rep. 2017. 7(1):2963


This work has been supported by the Federal Ministry for Education and Research (BMBF: 031A578).

Figure 1
Fluorescence microscopy images of the biomimetic feeding vessel.
Figure 2
Non-invasive imaging of vessel endothelialisation and channel morphology under dynamic cultivation:
Keywords: bioprinting, tumor-on-a-chip, tumor angiogenesis, hydrogel, drug screening platform
9:30 AM PS-21-6

PET/MRI evaluation of tumor heterogeneity using K-means clustering of dynamic [18F]FET-PET and its relationship with diffusion and permeability in Glioblastoma (#188)

S. G. Castaneda1, 2, J. A. Disselhorst2, P. Katiyar2, B. Bender4, M. Rheimold1, J. - M. Hempel4, U. Ernemann4, G. Tabatabai3, B. J. Pichler2, C. la Fougère1

1 University Clinic Tuebingen, Department of Nuclear Medicine and clinical molecular imaging, Tübingen, Baden-Württemberg, Germany
2 University Clinic Tuebingen, Werner Siemens Imaging Center, Department for Preclinical Imaging and Radiopharmacy, Tübingen, Baden-Württemberg, Germany
3 University Clinic Tuebingen, Center for Neurooncology, Tübingen, Baden-Württemberg, Germany
4 University Clinic Tuebingen, Department of Neuroradiology, Tübingen, Baden-Württemberg, Germany


Understanding tumor heterogeneity can provide insight into tumor aggressiveness, timing of therapies and tumor recurrence. We have shown that [18F]FDG could be used to evaluate tumor heterogeneity in small animals[1]. [18F]FET is a more specific candidate for translation of this methodology. Initial non-invasive tumor grading, tumor delineation, risk stratification and therapy response assessments is already performed in the clinic using [18F]FET [2, 3]. In this preliminary evaluation, we have focused on the segmentation of tumors using dynamic [18F]FET through a K-means clustering algorithm.


Five patients with [18F]FET-PET/MRI-based diagnosis of glioma, who provided written informed consent, were used for analysis. Apparent diffusion coefficient (ADC), calculated permeability maps using Gadovist, T2- and T1-weighted images and 40-minute attenuation-corrected [18F]FET images were coregistered and corrected for motion artifacts. Tumors were delineated using an isocontour of T2WI hyperintensities. For every patient, k-means was used to segment three cluster derived ROIs in the tumor using the 16 frames of the 40-min [18F]FET-PET acquisitions. Clusters were sorted according to low, mid or high [18F]FET uptake in the period from 20-40 min p.i. (Late [18F]FET).  2-Compartmental model was performed. The cluster ROIs were overlaid on all the maps and average values were extracted.


A reference healthy brain region of interest (ROI) per image was used as control region for data normalization. The cluster average was divided by the healthy region average per patient to normalize the data and reduce inter-subject variability.

The clusters showed increased [18F]FET uptakes in the Late [18F]FET period, in comparison to the reference healthy brain region (p<0.05) (Supplemental). ADC normalized to the healthy brain showed that Cluster 2 and 3 presented similar diffusion increments in comparison to Cluster 1. K1 was higher in Cluster 2 (p<0.05) in comparison to Cluster 1. Flux calculation represented by K1/k3/(k2+k3) showed an interesting trend with non-significant increments (ANOVA, p=0.06), denoting specific 2n-compartment accumulations. Flux presented a strong negative correlation to ADC (-0.89) and a positive correlation to contrast enhanced changes (0.80).


In this preliminary report we perform an evaluation of tumor heterogeneity by clustering the dynamic [18F]FET datasets. The tumors appear to have regions of high and intermediate flow with increased retention, and areas of reduced flow and reduced retention. It appears as if one region presents efficient vascularity and permeability, while another portion of the tumor presents slower kinetics but still high uptake. We aim to increase patient number and increase the robustness of this analysis to gain a better understanding glioma heterogeneity.


  1. Katiyar P, Divine MR, Kohlhofer U, et al (2017) Spectral Clustering Predicts Tumor Tissue Heterogeneity Using Dynamic 18 F-FDG PET: A Complement to the Standard Compartmental Modeling Approach. J Nucl Med 58:651–657. doi: 10.2967/jnumed.116.181370
  2. Albert NL, Weller M, Suchorska B, et al (2016) Response Assessment in Neuro-Oncology working group and European Association for Neuro-Oncology recommendations for the clinical use of PET imaging in gliomas. Neuro Oncol 18:1199–1208. doi: 10.1093/neuonc/now058
  3. Piroth MD, Pinkawa M, Holy R, et al (2011) Prognostic value of early [ 18F]fluoroethyltyrosine positron emission tomography after radiochemotherapy in glioblastoma multiforme. Int J Radiat Oncol Biol Phys 80:176–184. doi: 10.1016/j.ijrobp.2010.01.055
Keywords: Glioma, PET/MRI, Clinical, Kinetic modelling, multiparametric, ADC, FET, Contrast Agent
9:40 AM PS-21-7

Imaging Metastasis-Associated Macrophages Using 19-Fluorine Based MRI Cell Tracking (#307)

P. J. Foster1, 2, A. V. Makela1, 2

1 Robarts Research Institute, London, Ontario, Canada
2 Western University, Medical Biophysics, London, Ontario, Canada


The presence of tumor associated macrophages (TAMs) correlates strongly with breast cancer progression and metastatic spread (1). Macrophages located within or near metastases are referred to as metastasis associated macrophages (MAMs) (2). There have been few studies investigating the role of MAMs in breast cancer and mechanisms for the formation and growth of metastases are still mostly unknown. In this study, we demonstrate that in vivo fluorine-19 (19F)-based MRI cell tracking can evaluate the density and distribution of MAMs in murine models of breast cancer. 


4T1 murine breast cancer cell lines were implanted into the mammary fat pad in balb/c mice. For MRI, mice were administered 200 µl of a red fluorescent-tagged perfluorocarbon (PFC) agent via the tail vein 24 hours prior to imaging, at 4 weeks after cancer cell implantation.  1H and 19F images were acquired with a 9.4T small animal MRI system using a custom built dual 1H/19F birdcage coil.  Reference tubes of known 19F concentration (3.33x1016 19F/μL) were placed alongside the mouse for quantification. 1H and 19F images of the whole mouse body were acquired with balanced steady state free precession (bSSFP) pulse sequences.  The the total number of 19F spins were determined for each tumor. Validation was performed with fluorescence microscopy and immunohistochemistry (IHC).


Lung metastases were identified with 1H MRI as hyperintense regions within the black lungs (Fig1A,B). Multiplanar slicing of the 1H/19F 3D images allows clear visualization of the location of the metastasis within the lung (Figure 1C-Axial, 1D-Coronal and 1E-Sagittal). The average number of lung metastases which contained 19F signal was 3.8 per mouse (range 0-12 per mouse). 8 of the 13 mice had lung metastases in both lobes. Lung metastases had diverse MAM densities. The average number of 19F spins per lung metastasis was 1.1E+17. Tthe small size of the lung metastases and the relatively low image resolution prevents the analysis of precisely defining where within the metastases the 19F signal is located. Only 1 of the 13 mice had no visible lung metastases. Not all lung metastases which were visible in 1H images contained 19F signal.  Fluorescence microscopy and macrophage IHC confirmed the presence of PFC-positive macrophages (Fig2).


This study shows, for the first time, proof of the ability to use 19F MRI cell tracking to visualize MAMs in the lungs. The ability to detect and monitor the number of macrophages in individual metastatic tumors with 19F MRI may allow for identification of those tumors which are heavily infiltrated with MAMs and could be used as a biomarker for decisions about how to best treat these patients and for monitoring responses to therapy.


1. Gwak JM, Jang MH, Kim D Il, Seo AN, Park SY. Prognostic value of tumor-associated macrophages according to histologic locations and hormone receptor status in breast cancer. PLoS One 2015;10:1–14.

2. Qian B, Deng Y, Im JH, Muschel RJ, Zou Y, Li J, Lang RA, Pollard JW. A distinct macrophage population mediates metastatic breast cancer cell extravasation, establishment and growth. PLoS One 2009;4:e6562.


We thank Dr. Jeffrey M Gaudet for initial assistance with development of 19F MRI and Dr. Yuanxin Chen for help with IHC.

Figure 1
(A,B) Whole mouse body (1H) imaging allows for anatomical reference of metastases. 19F/1H overlay of axial (C), coronal (D) and sagittal (E) planes display the same lung metastasis containing 19F signal within the yellow circle. 

Colour scale bar represents minimum/maximum values for C, D & E (15515/27438). Scale bars: 5 mm.

Figure 2

PFC labeled macrophages in lung metastasis. (A) Anti FITC F4/80 for macrophages (green), (B) Red fluorescent PFC agent, (C) H&E and (D) Overlay of A, B & C demonstrate PFC+ macrophages within lung metastasis. Area void of fluorescence and H&E stained in centre may be a vessel. Scale bars: 500 μm.

Keywords: MRI, fluorine, cancer, macrophage
9:50 AM PS-21-8

Developing In Vivo Imaging Tools to Investigate the Role of Actomyosin Contractility in Melanoma Metastasis (#68)

B. Fanshawe1, 2, I. Hernandez2, A. Volpe1, L. Lim1, E. Crosas Molist2, J. Luis Orgaz2, G. Cantelli2, F. Mardakheh3, V. Sanz-Moreno2, G. Otto Fruhwirth1

1 King's College London, School of Biomedical Engineering and Imaging Sciences, London, United Kingdom
2 King's College London, School of Basic and Translational Biology Sciences, London, United Kingdom
3 Queen Mary University of London, Centre for Molecular Oncology, London, United Kingdom


Metastasis is the major cause of cancer mortality. Melanoma is a form of skin cancer with metastasis incidence and poor 5-year survival. Melanoma metastasis is regulated by actomyosin contractility driven by Rho-associated protein kinases (ROCKs) [1, 2]. Preclinical models for tracking spontaneous melanoma metastasis in vivo have not yet been reported. Our aim was to develop a preclinical pipeline, whereby our in vivo traceable melanoma metastasis model was used as a tool for preclinical validation of ROCK inhibitors.


The mouse melanoma cell line 4599 BrafV600E [3] was genetically engineered to stably express a radionuclide-fluorescence fusion reporter gene: the human sodium iodide symporter (NIS) fused to either GFP or mCherry [4]. Stable melanoma cell lines were characterised for reporter expression, localisation and function. Orthotopic melanoma models were established with these new reporter cells in immunocompromised (NSG) mice, with 250,000 cells intradermally injected. Spontaneous melanoma metastasis was studied in randomised cohorts treated with two different ROCK inhibitors (Y27632 or GSK269962A) or vehicle. All mice were longitudinally in vivo imaged using nanoSPECT/CT and tissues analysed ex vivo (γ-counting and histology).


Stable and correct reporter expression and function was confirmed in vitro and in vivo (Fig.1A-E). We showed by in vivo cell tracking that melanoma tumours spontaneously metastasized. Spontaneous melanoma metastasis was detected by in vivo imaging, and upon treatment with ROCK inhibitors (Y27632 or GSK269962A) slower tumour growth and reduced metastasis was observed. Y27632-treated tumours (315±28mm3; N=6) were smaller than vehicle treated tumours (V=606±54mm3; N=6; P=0.0008) by calliper measurement. Tumours also had a smaller live tumour volume (by NIS imaging; Fig.2C; P=0.0106) and tumour uptake (Fig. 2D; P=0.0047). Lung metastasis volume (by SPECT/CT: P=0.0005) and uptake (by SPECT/CT: P=0.0270). Histology confirmed in vivo imaging results and showed a significant reduction in lung metastases per lung area (1.7x10‑6±3.3x10‑7μm-2 vs. 6.6x10‑7±2.6μm‑2; N=4).


For the first time in melanoma research, a model of orthotopic spontaneous metastasis that can be tracked non-invasively in vivo by imaging has been established. It successfully employs a multi-scale dual-mode radionuclide-fluorescence imaging strategy and offers reliable and sensitive quantification of tumour growth and metastatic spread. Its application validated in vivo the utility of ROCK inhibitors as a treatment for reducing tumour growth and metastatic spread.


[1]        Cantelli, G., et al., Current Biology, 2015

[2]        Orgaz, J.L., et al., Nat Commun, 2014

[3]        Dhomen, N., et al., Cancer Cell, 2009

[4]        Fruhwirth, G.O., et al., J Nucl Med, 2014


This work was supported by the King’s Bioscience Institute, the Guy’s and St Thomas’ Charity Prize PhD Program in Biomedical and Translational Science, the Comprehensive Cancer Imaging Centre King's College London & UCL, and Cancer Research-UK.

Characterisation of 4599 NIS-mCherry Expressing Cell Lines

(A)  Viral NIS-mCherry construct for generating stable cells. (B) Immunoblot probed for mCherry in parental and NIS-mCherry cells. (C) 99mTcO4 uptake assays; n=3 mean and SD shown. (D) Confocal micrograph of 4599 NIS-mCherry cells showing plasma membrane localisation of NIS-mCherry and wheat germ agglutinin (WGA). (E) Immunofluorescence of mCherry stained tumours. Scale=25 µm.

The ROCK Inhibitor Y27632 Reduces Primary Tumour and Metastatic Burden

(A) 99mTc-SPECT/CT maximum intensity projections of control and inhibitor. Background from intestines (Int), lachrymal glands (L) and stomach (St) are shown. Primary tumour (T), lymph node metastases (LN mets) and lung metastases are also shown. (B) 3D reconstructions. (C-D) Primary tumour quantifications (analysed on Vivoquant™). (E-F) Lung metastasis quantifications.

Keywords: Melanoma, Reporter gene imaging, NIS, Cell tracking, SPECT/CT