15th European Molecular Imaging Meeting
supported by:

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

Late-breaking Pitches | Cancer

Session chair: Margy Koffa (Evros, Greece); Harry Tsoumpas (Leeds, UK)
Shortcut: PS 03
Date: Wednesday, 26 August, 2020, 10:00 a.m. - 11:30 a.m.
Session type: Parallel Session


Abstract/Video opens by clicking at the talk title.

10:00 a.m. PS 03-01

In vivo tracking of endogenous CD4+ T cells for guidance and monitoring of cancer immunotherapy

Stefania Pezzana1, Bredi Tako1, Simone Blaess1, Dominik Seyfried1, Natalie Mucha1, Linda Schramm1, Andreas Maurer1, 2, Alessandro Mascioni5, Bernd J. Pichler1, 2, Manfred Kneilling1, 2, 3, Dominik Sonanini1, 4

1 Eberhard Karls University, Department of Preclinical Imaging and Radiopharmacy, Werner Siemens Imaging Center, Tübingen, Germany
2 Eberhard Karls University, Cluster of Excellence iFIT (EXC 2180), Tübingen, Germany
3 Eberhard Karls University, Department of Dermatology, Tübingen, Germany
4 Eberhard Karls University, Department of Internal Medicine VIII, Tübingen, Germany
5 ImaginAb, Inglewood, United States of America


Checkpoint Inhibitor Therapy (CIT) has become an important weapon against cancer, even though only 20-40% of patients benefit from its therapeutic effects. Beyond CD8+ cytotoxic T cells, increasing importance has been attributed to CD4+ T helper cells in antitumoral immunity. CD4+ T cells display a highly diverse group with divergent effector functions. Here, we report on the in vitro and in vivo validation of a novel Zirconium-89 (89Zr) labeled CD4 targeting minibody for Positron Emission Tomography (PET) imaging of CD4+ cells and its relevance in cancer immunotherapy.


We conducted in vitro binding and internalization assays of the αCD4 minibody 89Zr-dfo-IAB46 (ImaginAb, Inglewood) using freshly isolated CD4+ T cells from C57BL/6 mice. For in vivo studies we employed two subcutaneous syngeneic tumor mouse models: MC38 colon adenocarcinoma and OVA-B16 melanoma. Tumor-bearing mice were treated intraperitoneally (i.p.) with αPD-L1/αLag-3 monoclonal antibodies (mAbs), cyclophosphamide + αPD-L1/αLag-3 mAbs or isotype control mAbs. 7 days after treatment onset, we injected 89Zr-dfo-IAB46 intravenously (i.v.) and performed simultaneous PET/MR studies after 6, 24 and 48h. Finally, to cross-validate in vivo data, we harvested tumors and organs of interest for biodistribution analysis and CD4 immunohistochemistry (IHC). All data are presented as mean ± SEM.


In vitro CD4+ T cell binding assays revealed an excellent immunoreactivity (93.9±3.9%) and internalization capability (40.0±0.6% 24h post-incubation at 37°C) of 89Zr-dfo-IAB46. We observed a significant higher 89Zr-dfo-IAB46 uptake in αPD-L1/αLag-3 mAbs responding MC38 tumors (4.4±0.2 %ID/ml) when compared to non-responding (2.78±0.4 %ID/ml) or isotype control mAbs treated mice (3.0±0.1 %ID/ml). To note, 89Zr-dfo-IAB46 tumor uptake strongly correlated with the anti-tumoral immune response (r2= 0.98; p<0.001). We further investigated CD4+ T cell homing into OVA-B16 melanomas of mice treated with a combination of cyclophosphamide + αPD-L1/αLag-3 mAbs. PET/MR imaging revealed an increase in 89Zr-dfo-IAB46 αCD4 uptake in OVA-B16 melanomas of mice efficiently treated with the combined therapy (4.9±0.6 %ID/ml) when compared to isotype control mAbs treatment (2.9±0.3 %ID/ml). Ex vivo 89Zr-dfo-IAB46 biodistribution analysis, as well as CD4 IHC of tumors, confirmed our in vivo data.


CD4+ T cells are underestimated key players that elicit and catalyze efficient anti-tumoral immune responses. Our in vitro and in vivo results point out the feasibility of 89Zr-dfo-IAB46 PET/MR imaging to monitor CIT induced CD4+ T cell infiltration. The correlation between 89Zr-dfo-IAB46 tumor uptake and therapy response qualifies CD4-PET imaging as a promising novel tool to guide CIT.

Keywords: ImmunoPET, CD4 T cells, Cancer immunotherapy
10:10 a.m. PS 03-02

Human PD-L1 Nanobody For Immuno-PET Imaging: Radiolabeling Strategies

Jessica Bridoux1, Katrijn Broos2, Quentin Lecocq2, Charlotte Martin6, Frederik Cleeren7, Mikhail Kondrashov8, Steven Ballet6, Guy Bormans7, Geert Raes5, 3, Karine Breckpot2, Magnus Schou8, Chad Elmore9, Nick Devoogdt1, Vicky Caveliers4, 1, Marleen Keyaerts4, 1, Catarina Xavier1

1 Vrije Universiteit Brussel (VUB), In Vivo Cellular and Molecular Imaging (ICMI), Jette, Belgium
2 Vrije Universiteit Brussel (VUB), Laboratory of Molecular and Cellular Therapy (LMCT), Jette, Belgium
3 Vrije Universiteit Brussel (VUB), Cellular and Molecular Immunology (CMIM), Brussel, Belgium
4 UZ Brussel, Nuclear Medicine Department, Jette, Belgium
5 VIB Inflammation Research Center, Myeloid Cell Immunology Lab, Gent, Belgium
6 Vrije Universiteit Brussel (VUB), Research group of Organic Chemistry (ORGC), Brussel, Belgium
7 University of Leuven, Radiopharmaceutical Research, Department of Pharmacy and Pharmacology, Leuven, Germany
8 Karolinska Institute, Department of Clinical Neurosciences, AstraZeneca PET Science Centre, Solna, Sweden
9 AstraZeneca, Early Chemical Development, Pharmaceutical Sciences, Mölndal, Sweden


Immune checkpoints such as Programmed death-ligand 1 (PD-L1) limit the T-cell function, and tumor cells have developed this receptor to escape the anti-tumor immune response. Monoclonal antibody-based treatments have shown long-lasting responses, but only in a subset of patients. Therefore, there is a need to predict the response to treatments. This study aims to develop a Nanobody (Nb)-based probe to assess human PD-L1 (hPD-L1) expression using PET imaging. The Nb has been site-specifically modified for Gallium-68 (68Ga) or Fluorine-18 (18F) radiolabeling.


The hPD-L1 Nb with a sortag-motif at its C-terminus was site-specifically coupled to a bifunctional chelator or a tetrazine (Tz) via the Sortase A enzyme coupling reaction. Modified Nbs were purified by immobilized metal affinity chromatography (IMAC) and size-exclusion chromatography (SEC), characterized by Mass Spectrometry (ESI-Q-TOF), SDS-PAGE and Western Blot. NOTA-(hPD-L1) Nb was labeled with 68Ga, RESCA-(hPD-L1) Nb with [18F]AlF and Tz-(hPD-L1) Nb with [18F]F-PEG3-BCN. Radiochemical purity (RCP) was assayed by SEC and iTLC. In vivo tumor targeting of [68Ga]Ga-NOTA-(hPD-L1) Nb was assessed in xenografted-athymic nude mice bearing PD-L1 positive cells, or PD-L1 negative cells as a control. PET/CT imaging was performed with 68Ga-labeled NOTA-(hPD-L1) Nb 1h post-injection.


Site-specifically functionalized hPD-L1 Nbs with NOTA or RESCA were obtained with high purity (≥99%) in 52% and 59% yield (Figure 1). (hPD-L1)-Tz was obtained in 63% yield. Functionalization did not affect affinity.

Labelling of NOTA-(hPD-L1) with 68Ga was performed at room temperature (RT) for 10 min in a 80% decay corrected radiochemical yield (DC-RCY), ≥99% RCP and apparent molar specific activity of 85 GBq/μmol. Over 4 hours, the radiolabeled probe and metal complex were stable (≥99% RCP). In vivo tumor targeting studies revealed high tumor uptake of (3.2 ± 1.2) %IA/g organ, and no unspecific organ targeting, except in the kidneys and excretion to the bladder (route of excretion).

Labeling of RESCA-(hPD-L1) with [18F]AlF was performed at RT for 12 min at in a 29% DC-RCY and with a RCP ≥99%. Biodistribution in healthy animals showed higher bone uptake than for the 68Ga-labeled Nb.

Preliminary results with [18F]F-PEG3-BCN-Nb showed about 30% labeling and more tests will be performed.


A site-specific functionalized (hPD-L1) Nb was obtained using Sortase enzyme approach, which could be radiolabeled with 68Ga or 18F. [68Ga]Ga-NOTA-(hPD-L1) Nb proved to specifically target the hPD-L1 receptor in vivo. Further studies will be performed with the 18F-labeled Nb after optimization of the labelling conditions. The best probe in terms of ease of production and in vivo behavior will be selected for clinical translation.

AcknowledgmentThe authors would like to thank Cindy Peleman for animal handling and PET imaging.
[1] K. Broos, Q. Lecocq, C. Xavier et al., Evaluating a Single Domain Antibody Targeting Human PD-L1 as a Nuclear Imaging and Therapeutic Agent, Cancers 2019, 11(6), 872.
Figure 1
Site-specific functionalization of the hPD-L1 Nanobody with NOTA, RESCA or tetrazine (Tz) and radiolabelling with 68Ga, [18F]AlF or [18F]F-PEG3-BCN respectively.
Keywords: PD-L1, Nanobody, PET imaging, site-specific
10:20 a.m. PS 03-03

Contrast enhanced ultrasound alters tumor pathophysiology and therapy outcome in orthotopic triple negative breast cancer in mice

Anne K. Rix1, Jasmin Baier1, Natascha I. Drude1, Maike Baues1, Jan-Niklas May1, Susanne Koletnik1, Milita Darguzyte1, Rupert Palme2, Rene H. Tolba3, Fabian Kiessling1

1 RWTH Aachen International University, Institute for Experimental Molecular Imaging, Aachen, Germany
2 University of Veterinary Medicine Vienna, Department of Biomedical Sciences, Vienna, Austria
3 RWTH AachenInternational University, Institute for Laboratory Animal Science and Experimental Surgery, Aachen, Germany


Contrast-enhanced ultrasound (CEUS) is an established imaging technique in preclinical routine for the characterization of tumor vascularization and expression of molecular markers on tumor endothelium. It is known that US contrast agents can permeabilize vessels during sonoporation1 or induce a vascular shutdown after inertial cavitation2, but no data are available to exclude an influence of diagnostic CEUS on tumor biology. Therefore, this study evaluated the possible influences of diagnostic CEUS on tumor pathophysiology and therapy outcome after antiangiogenic and antitumoral treatment.


Female orthotopic 4T1-tumor bearing BALB/c mice (n=100) were allocated randomly to the following groups: (i) therapy only, (ii) anesthesia only, (iii) destructive US, (iv) non-destructive CEUS or (iv) destructive CEUS. Animals were subdivided to either receive 10 mg/kg regorafenib or vehicle solution daily by oral gavage. US measurements were performed on days 7, 10 and 14 after tumor cell injection using phospholipid MB for non-destructive CEUS or VEGFR2-targeting phospholipid MB for destructive CEUS. Animal well-being (heart rate, motor coordination, fecal corticosterone metabolite concentration) and tumor size were evaluated daily. A histological characterization of tumors concerning vascularization (e.g. perfusion, angiogenesis) and immune cell infiltration was performed.


Longitudinal contrast enhanced ultrasound examinations had no impact on heartrate, motor coordination or fecal corticosterone metabolite levels of the animals. No influence of the different imaging protocols on the tumor size could be detected. However, histological characterization of tumors showed a significant increase in tumor vascularization, perfusion and angiogenesis in vehicle treated tumors exposed to non-destructive CEUS (Fig. 1A). Furthermore, macrophage infiltration was significantly increased in vehicle and regorafenib treated tumors after non-destructive CEUS (Fig. 1B). Surprisingly, both, non-destructive and destructive CEUS induced a systemic immune reaction reflected by decreased white blood cell counts and spleen sizes in 4T1-tumor bearing mice (Fig. 2).


Although US imaging did not increase the burden of the animals during the experiment, diagnostic CEUS influenced 4T1-tumor vascularization, perfusion, angiogenesis and induced a local and systemic immune response. Further investigations are ongoing to characterize the tumor microenvironment (e.g. immune cell polulations, collagen content, matrix metalloproteinases, etc.) and to unravel the biological mechanisms behind these unexpected findings.

AcknowledgmentThis work is supported by Deutsche Forschungsgemeinschaft (DFG) project number 321137804 and 331065168.
[1] Theek B, Baues M, Ojha T et. al. Sonoporation enhances liposome accumulation and penetration in tumors with low EPR J Control Release. 2016 Jun 10; 231: 77–85. 
[2] Goertz DE, Karshafian R, Hynynen K. IEEE. Antivascular effects of pulsed low intensity ultrasound and microbubbles in mouse tumors. IEEE Ultrason Symp 2008;1–4:670–3.
Figure 1: Influence of diagnostic CEUS on tumor pathophysiology
Figure 2: Influence of diagnostic CEUS on systemic immune reaction
Keywords: Cancer, Ultrasound, Imaging, Microbubbles
10:30 a.m. PS 03-04

Simultaneous DMI of [2H9]choline and [6,6 2H2]glucose uptake and conversion in tumors

Andor Veltien1, Nia Ravichandran1, Robin de Graaf2, Henk de Feyter2, Jeannette Oosterwijk3, Egbert Oosterwijk3, Arend Heerschap1

1 Radboudumc, Radiology, Nuclear medicine, Anatomy, Nijmegen, Netherlands
2 Yale University, Radiology and biomedical imaging, New Haven, United States of America
3 Radboudumc, Urology, Nijmegen, Netherlands


Imaging enhanced glucose uptake in tumors is traditionally done by 18FDG-PET [1], but is also possible by MR deuterium metabolic imaging (DMI) using non-radioactive [6,6 2H2]glucose [2]. DMI has several advantages, e.g. by tracking 2H it provides a quantitative biomarker for the Warburg effect [2]. Enhanced choline uptake is another hallmark of tumor growth and has been imaged by 18F or 11C choline PET[3,4], Here we demonstrate that DMI of [2H9]choline uptake in tumors is possible. And we show that DMI of [6,6 2H2]glucose and [2H9]choline can be performed simultaneously.


Renal tumors were grown subcutaneously in mice up to ~155 mm3 in size. Mice were aneasthesized with isoflurane. To prevent a cholinergic reaction they were IV injected with 12 μg atropine. We injected 0.05gr/kg [2H9]choline in a 0.16 ml saline solution IV in ~15 sec. Or a 0.05gr/kg [2H9]choline with 1.3gr/kg [6,6 2H2]glucose combination in a 0.3 ml saline solution IV in ~20 sec. After 1H reference imaging and shimming, 2H MR was performed on a Bruker Biospec at 11.7T with a home-built 2H coil. DMRS was performed with an excitation pulse angle of 900 and a TR = 500 ms. 3D DMI was performed with a nominal spatial resolution of 2x2x2 mm, TR = 400 ms, TE = 0.4 ms and total acquisition time of 36 minutes.

Data was processed in DMIWizard. Initial HOD tissue concentration was assumed to be 13.7mM.


After 2H9 choline injection a 2H signal for choline appeared at ~3.2 ppm, well separated from that of HOD. It increased to a maximum level of ~1 mM 2H9 choline at ~10 min post-injection and then slowly decreased to ~0.6 mM (Fig. 1). As expected the HOD signal remained almost constant. (Fig. 1). No other clear signals were observed except for a small signal adjacent to the choline signal, which may be betaine the primary breakdown product of choline [5]. A DMI obtained at 60 min showed a clear spot for choline in the tumor (Fig. 1). 2H NMR spectra of blood taken at ~100 min showed no choline signal indicating that in vivo the 3.2 ppm signal arises from the tumor and not blood.

DMI results of a 2H9 choline and [6,6 2H2]glucose phantom demonstrates that their signals can be separately observed (Fig. 2a).

After injecting 2H9 choline and [6,6 2H2]glucose combined in a mouse the 2H MR spectra of the tumor clearly show separated signals for these compounds and also for lactate (Fig. 2b).


We developed a successful protocol for following uptake and imaging of 2H9 choline in tumors after a bolus administration of this compound. In addition we were able to perform DMI simultaneously of [2H9]choline and of [6,6 2H2]glucose. As DMI of glucose uptake and its metabolic conversions has been shown to be feasible in patients [2], the simultaneous performance of choline DMI is an important complementary extension with clinical potential.

[1] Gambhir, S. Molecular imaging of cancer by positron emission tomography. Nat Rev Cancer, 2002, 2 (9), 683
[2] De Feyter, H, Behar, K, Corbin et al. Sci. Adv. 2018, (8)4, 7314
[3] Glunde K, Bhujwalla Z, Ronenn S. Choline metabolims in malignant transformation. Nat Rev Cancer 2011, 11, 835
[4] Evangelista L, Briganti A, Fanti S, Joniau S, Reske S, Schiavina R, Stief C, Thalmann GN, Picchio M. New Clinical Indications for 18F/11C-choline, New Tracers for Positron Emission Tomography and a Promising Hybrid Device for Prostate Cancer Staging: A Systematic Review of the Literature. Eur Urol. 2016 Jul;70(1):161-175.
[5] Katz-Brull R, Margalit R, Bendel P, DEgani H. Choline metabolism in breast cancer. 2H, 13C and 31P NMR satudies of cells and tumors. MAGMA 1998, 6, 44
Figure 1

a) 2H MR spectra of subcutaneous tumor after injection of 2H9 choline.

b)Time curves of HOD and 2H9 choline signal integrals

c) T2 weighted MRI of subcutaneous tumor. d) 2H9 choline heat map overlaid on T2 weighted MRI

Figure 2

a) DMI of phantom with two tubes, one filled with 2H9 choline and the other with [6,6 2H2]glucose. The glucose and choline signals are clearly separated in the 2H spectrum

b) Left: 2H spectrum of renal tumor after IV bolus injection of 2H9 choline injection and [6,6 2H2]glucose combined in mouse. Right DMI maps of deuterated glucose, choline and lactate overlaid on T2 MRI 

Keywords: Deuterium, metabolic, imaging, tumors
10:40 a.m. PS 03-05

Stratification of colorectal cancer subtypes using PET/MRI biomarkers and machine learning

Gaurav Malviya1, Matthew Neilson1, Emma R. Johnson1, Agata Mrowinska1, Rene-Filip Jackstadt1, Dmitry Solovyev1, Gavin Brown1, Colin Nixon1, Crispin Miller1, Owen J. Sansom1, David Y. Lewis1

1 Cancer Research UK Beatson Institute, Glasgow, United Kingdom


Colorectal cancer has been classified into 4 consensus molecular subtypes [1]. The aim of this molecular stratification is to support development of subtype-specific therapies for patient subgroups. Gene expression data is the current basis for molecular subtyping but provides limited insight into subtype dynamics and spatial heterogeneity. To overcome these limitations we developed non-invasive PET/MR imaging biomarkers for CRC stratification as surrogates of gene-expression based subtyping. These biomarkers could identify co-existing subtypes and post-therapy subtype switching.


We classified five colon cancer organoid models derived from genetically engineered mice with multiple modified alleles representing a spectrum of human colon cancer: Apcfl/+ KrasG12D/+ (AK), Apcfl/+ KrasG12D/+ Trp53fl/fl Tgfbr1-/- (AKPT), Apcfl/+ KrasG12D/+ Trp53fl/flRosa26N1icd/+ (AKPN), KrasG12D/+ Trp53fl/fl Rosa26N1icd/+ (KPN) and liver-homing KrasG12D/+ Trp5fl/flRosa26N1icd/+ (KPNLIVMET). We used a complementary panel of metabolic PET tracers ([18F]FDG, [18F]FET, [18F]FLT and [11C]acetate (ACE)) and 126 PET/MRI radiomic features derived from multi-tracer PET/MR images. The method of Deeb et al. [2], which employs support vector machines (SVMs), was used to find a radiomic signature that can best-discriminate between the five organoid models.


This is the first CRC stratification using a panel of PET and MRI imaging biomarkers. [18F]FDG and [18F]FET are more effective than [18F]FLT and [11C]ACE for stratifying models of CRC subtypes. Additionally, significant differences in SUVmean uptake values of [18F]FDG (p = 0.028) and [18F]FET (p = 0.008) in primary KPN and KPNLIVMET organoids demonstrated the potential to differentiate metastatic form of CRC within syngeneic mouse models.

The SVM analysis identified an optimal set of 16 PET/MRI parameters to classify each organoid model, with leave-three-out cross-validation producing a mean precision of 58.8% (AKPN 30.5%, AKPT 49.5%, KPN 72.7%, KPNLIV METS 79.9%, AK 61.2%). We are in the process of acquiring a validation data set to assess the generalisability of our SVM classifier.


The stratification of CRC subtypes using PET/MRI biomarkers and machine learning methods in this study might provide a novel approach for future clinical CRC stratification and advance the development of subtype-based personalised treatment.


The authors are grateful to the CRUK Beatson Institute Central Services, and West of Scotland PET Centre at Gartnavel Hospital, Glasgow, for their help and support.

[1] Guinney, J, Dienstmann, R, Wang, X, de Reyniès, A, Schlicker, A, Soneson, C, et al 2015, 'The consensus molecular subtypes of colorectal cancer',  Nature Medicine, 21(11),1350-6
[2] Deeb, SJ, Tyanova, S, Hummel, M, Schmidt-Supprian, M, Cox, J, Mann, M, 2015, 'Machine Learning-based Classification of Diffuse Large B-cell Lymphoma Patients by Their Protein Expression Profiles' Molecular & Cellular Proteomics, 14(11), 2947-60
[3] Nichols, TE, Holmes, AP, 2001, 'Nonparametric Permutation Tests for Functional Neuroimaging: A Primer with Examples', Human Brain Mapping,15,1–25
Figure 1:

CRC mouse organoids derived from five genetically engineered mouse models and subcutaneously transplanted into the right shoulder blade of CD-1 nude mice, showed different uptake of three metabolic PET radiotracers. [18F]FDG and [18F]FET PET uptake (SUVaverage) differentiate CRC subtypes more readily than [18F]FLT and [11C]ACE (Mann Whitney test, *P < 0.05, **P < 0.01).

Figure 2:
(A) Support vector machine analysis for optimal feature selection [2], using p-values of standard analysis of variance tests to rank radiomic parameters. The maximum mean precision (58.8%) was achieved using 16 radiomic parameters. (B) Monte Carlo permutation test [3] for the classification precision of each genotype, using support vector machine hyperparameters associated with the maximum mean precision. For each genotype, the black curve denotes the distribution of precisions from 1,000 random permutations, whilst the red line denotes the precision achieved using the correctly-labelled data.
Keywords: Colorectal cancer subtypes, Machine learning, CRC Stratification, PET-MRI, CRC mouse models
10:50 a.m. PS 03-06

Synthesis and validation of a humanized Fab fragment targeting Galectin-3 for the diagnosis of thyroid cancer via PET

Francesco De Rose1, Emanuel Peplau2, Markus Mittelhäuser1, Sybille Reder1, Markus Schwaiger1, Wolfgang A. Weber1, Armando Bartolazzi3, 4, Arne Skerra2, Calogero D'Alessandria1

1 Klinikum rechts der Isar, Technical University of Munich, Nuklearmedizinische Klinik und Poliklinik, München, Germany
2 Technical University of Munich, Lehrstuhl für Biologische Chemie, Freising, Germany
3 Cancer Center Karolinska, Karolinska Hospital, Pathology Research Laboratory, Stockholm, Sweden
4 Pathology Research Laboratory, Sant'Andrea University Hospital, Rome, Italy


In the last years the increasing incidence of thyroid cancer (TC), combined with a significant prevalence of benign lesions, raised attention for the lack of a rapid and specific imaging diagnostic tool. Galectin-3 (Gal3), whose expression is restricted to malignant thyroid tissue, is proposed as target for immunodiagnostic methods1,2.
To take advantage of the better tissue penetration and reduced immunogenicity compared to rodent mAbs, we developed a humanized anti-Gal3 Fab fragment for immuno-PET imaging. The tracer was characterized in vitro and validated in vivo in orthotopic murine models.


The variable region of an anti-Gal3 mAb was amplified from hybridoma M3/38 cells and used to construct a humanized anti-Gal3 Fab (hFab) via CDR grafting, fused with a Pro, Ala and Ser chain (PAS200) to tune its half-life3,4. The recombinant protein was conjugated with DFO and labeled with 89Zr. The hFab-PAS200-DFO-89Zr tracer was characterized by HPLC and TLC for integrity and stability. Binding tests were performed on the TC cell line BcPAP. For in vivo imaging, mice were inoculated with BcPAP cells in the left thyroid lobe, monitoring the tumor growth via ultrasound and fluorescence molecular tomography (FMT) with a Cy5.5-labeled PAS200-hFab. Mice were injected i.v. with ~3.0MBq of the radiotracer and imaged after 24h via PET/CT, followed by biodistribution analysis and autoradiography.


The hFab-PAS200 showed high affinity towards human Gal3 (KD = 0.3 nM) as measured via real time surface plasmon resonance spectroscopy (Biacore). The radiolabeling with 89Zr was achieved with a yield of 62±6% and high radiochemical purity (>98%). The stability measured in 0.25M sodium acetate pH5.5 (formulation buffer) and in human serum was >90%. In vitro binding studies on BcPAP cells showed a high affinity (KD = 15±7nM) and immunoreactivity (73±5%), also in comparison with the M3/38 mAb. Tumor growth monitoring via FMT confirmed the findings from ultrasound imaging (Figure 1). In vivo PET/CT imaging provided high image contrast already 12 h p.i., with best tumor-to-background value (20±4) after 24 h (Figure 2). Images and biodistribution analysis showed specific tracer accumulation in the tumor (4.7 ± 0.3% ID/g) compared to healthy thyroid tissue (1.2 ± 1% ID/g). Autoradiography on tumor slices indicated penetration of the tracer in the inner layers with exclusion of necrotic areas.


The immuno-PET targeting of Gal3 using 89Zr-DFO-haGal3-Fab-PAS200 offers a specific and reliable diagnostic method for the in vivo detection and characterization of TC. Due to the high resolution of the PET technology this novel tracer can be used in combination with other diagnostic modalities (ultrasound, fine needle aspiration) to improve TC diagnostics, thus reducing unnecessary surgery and social costs.

[1] Bartolazzi, A, Orlandi, F, Saggiorato, E, Volante, M, Arecco, F, Rossetto, R, Palestini, N, Ghigo, E, Papotti, M, Bussolati, G, Martegani, MP, Pantellini, F, Carpi, A, Giovagnoli, MR, Monti, S, Toscano, V, Sciacchitano, S, Pennelli, GM, Mian C, Pelizzo, MR, Rugge, M, Troncone, G, Palombini, L, Chiappetta, G, Botti G, Vecchione, A, Bellocco, R 2008, 'Galectin-3-expression analysis in the surgical selection of follicular thyroid nodules with indeterminate fine-needle aspiration cytology: a prospective multicenter study.', Lancet Oncol., 9, 543-549.
[2] De Rose, F, Braeuer, M, Braesch-Andersen, S, Otto, AM, Steiger, K, Reder, S, Mall, S, Nekolla S, Schwaiger, M, Weber, WA, Bartolazzi, A, D'Alessandria, C 2019, 'Galectin-3 Targeting in Thyroid Orthotopic Tumors Opens New Ways to Characterize Thyroid Cancer.', J Nucl Med, 60(6), 770-776.
[3] Mendler, CT, Friedrich, L, Laitinen, I, Schlapschy, M, Schwaiger, M, Wester, H-J, Skerra, A 2015, 'High contrast tumor imaging with radio-labeled antibody Fab fragments tailored for optimized pharmacokinetics via PASylation.', mAbs, 7, 96-109.
[4] Mendler, CT, Gehring, T, Wester, HJ, Schwaiger, M, Skerra, A 2015, '89Zr-Labeled versus 124I-labeled αHER2 Fab with optimized plasma half-life for high-contrast tumor imaging in vivo.', J Nucl Med, 56, 1112-1118.
Figure 1. FMT scan on mice bearing orthotopic TC
FMT scan 24 h after the injection of 2nmol of PAS200-hFab-Cy5.5 (right panel), compared with the necropsy findings (left panel).
Figure 2. PET-CT imaging at 24 hours
Left: PET-CT scan performed 24 h after injection of ~3.0 MBq 89Zr-DFO-haGal3-Fab-PAS200 tracer. Right: specific signal in non-blocked animal is highlighted and compared with necropsy findings.
Keywords: Galectin-3, humanized Fab, orthotopic xenograft, PET/CT-imaging, thyroid cancer
11:00 a.m. PS 03-07

Peroperative lymph node mapping using fluorescently-labeled mannose receptor-specific nanobodies

Łukasz Mateusiak1, Pieterjan Debie1, Noemi B. Declerck1, Christian Cyuzozo1, Danny M. van Willigen2, Geert Raes3, 4, Alex Mottrie5, 6, Fijs W. B. van Leeuwen2, 6, Philippe De Sutter7, Sophie Hernot1

1 Vrije Universiteit Brussel, Department of Medical Imaging and Physics (BEFY/MIMA) - In Vivo Cellular and Molecular Imaging Laboratory (ICMI), Brussels, Belgium
2 Leiden University Medical Center, Interventional Molecular Imaging Laboratory, Department of Radiology, Leiden, Netherlands
3 Vrije Universiteit Brussel, Research Group of Cellular and Molecular Immunology, Brussels, Belgium
4 Vlaams Instituut voor Biotechnologie, Inflammation Research Center, Laboratory of Myeloid Cell Immunology, Ghent, Belgium
5 Onze-Lieve-Vrouw Hospital, Department of Urology, Aalst, Belgium
6 ORSI Academy, Melle, Belgium
7 Vrije Universiteit Brussel, Universitair Ziekenhuis Brussel, Department of Gynecology and Oncology, Brussels, Belgium


Lymphadenectomy, the surgical removal of lymph nodes (LNs), is performed in clinically node-negative patients who have a high chance of nodal metastases. For quality assurance, often a minimum number of LNs in the area of interest needs to be harvested. Intraoperative identification of LNs can be a demanding procedure. We propose to use fluorescence molecular imaging to highlight all LNs within the surgical field to facilitate node picking. To realize LN staining we fluorescently labeled a previously developed anti-mannose receptor (MR) nanobody (Nb) that yields significant uptake in LNs1.


The nanobody MMR3.49 with high affinity for MR as well as a control nanobody were fluorescently-labeled with Cy5 or IRDye800CW. One hour after intravenous injection of 2 nmol of labeled compound in healthy mice (n=3 per group), axillary, inguinal and popliteal LNs were imaged in situ and ex vivo with the KIS (Kaer Labs) for Cy5 signals or Fluobeam800® (Fluoptics) for IRDye800CW signals. The translational nature of the approach was studied by upscaling the experiment to a healthy porcine model (n=1; 35 kg). Here, LN mapping in the pelvic area, following intravenous injection of 73 nmol of MMR3.49-Cy5, was evaluated 90 min after injection. LNs, muscle and fat were procured for further ex vivo assessment. Fluorescent signals and target-to-background ratios (TBRs) were quantified using ImageJ.


Both Cy5 and IRDye800CW-labeled anti-MR nanobodies enabled clear and specific in situ visualization of the LNs in mice, with ex vivo TBRs of 2.64±0.63 and 4.62±0.50 respectively. No uptake in LNs was observed for the control nanobodies (TBR 1.40±0.18 and 1.14±0.32) (Figure 1). Comparably, in the pig, MMR3.49-Cy5 helped to realize distinct in vivo fluorescent staining of pelvic LNs (Figure 2).


Intravenous injection of fluorescently-labeled anti-MR nanobodies enables intraoperative localization of LNs within the whole surgical view. This could potentially increase the accuracy of extended nodal dissections. Early data suggests the imaging findings in mice are scalable to porcine models.

[1] Xavier, C, Blykers, A, Laoui, D, Bolli, E, Vaneyken, I, Bridoux, J, Baudhuin, H, Raes, G, Everaert, H, Movahedi, K, Van Ginderachter, JA, Devoogdt, N, Caveliers, V, Lahoutte, T, Keyaerts, M 2019, ‘Clinical Translation of [68Ga]Ga-NOTA-anti-MMR-sdAb for PET/CT Imaging of Protumorigenic Macrophages’, Mol Imaging Biol, 21, 898, Springer International Publishing
Figure 1.

(A) In situ fluorescence imaging of inguinal LNs in mice using MR-targeting fluorescent Nb or control Nb. (B) Comparison of quantified average TBRs (with muscle as a background) for MR-targeting fluorescent Nb and control Nb.

Figure 2.

In site highlighting of a LN in the pelvic region of a pig using Cy5-labeled MR-targeting Nb, 90 minutes after intravenous injection.

Keywords: intra-operative imaging, lymphadenectomy, fluorescence image-guided surgery, mannose receptor, nanobody
11:10 a.m. PS 03-08

CD24-targeted intraoperative fluorescence image-guided surgery leads to improved cytoreduction of ovarian cancer in a preclinical orthotopic surgical model

Katrin Kleinmanns1, Vibeke Fosse1, 2, Ben Davidson3, 4, Line Bjørge1, 5, Emmet McCormack1

1 University of Bergen, CCBIO, Department of Clinical Science, Bergen, Norway
2 Erasmus Medical Centre, Department of Radiology, Rotterdam, Netherlands
3 Oslo University Hospital, Department of Pathology, Oslo, Norway
4 University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, Oslo, Norway
5 Haukeland University Hospital, Department of Obstetrics and Gynaecology, Bergen, Norway


The completeness of resection is a key prognostic indicator in patients with ovarian cancer1. The application of tumour-targeted fluorescence image-guided surgery (FIGS) has led to improved detection of peritoneal metastases during cytoreductive surgery2,3. CD24 is highly expressed in ovarian cancer and we have previously demonstrated it to be a suitable biomarker for tumour-targeted fluorescent imaging4. In this study, we applied a near-infrared (NIR) labelled CD24-tracer during FIGS in high-grade serous ovarian cancer (HGSOC) models and investigated the effect on surgical efficiency.


CD24 expression was investigated in cell lines (OV-90, Skov-3, Caov-3) and heterogenous patient-derived xenograft (PDX) tumour samples of HGSOC. After conjugation of the anti- CD24 monoclonal antibody to the NIR dye Alexa Fluor 750, and the evaluation of the optimal pharmacological parameters (OV-90, n = 21), orthotopic HGSOC metastatic xenografts (OV-90, n = 16) underwent cytoreductive surgery with real-time feedback. The impact of intraoperative CD24-targeted fluorescence guidance was compared to white light and palpation alone, and the recurrence of disease was monitored post-operatively (OV-90, n =12). CD24-AF750 was further evaluated in four clinically annotated metastatic PDX models of metastatic HGSOC, to validate the translational potential for intraoperative guidance.


CD24-expression was high in the majority of the in vitro analyses, and the OV-90 cell line had more than 300 000 CD24 binding sites per cell. The optimal contrast conditions for intraoperative imaging with CD24-AF750 was determined in vivo as 48 hours post-injection of 3 μg/g. CD24-targeted intraoperative NIR FIGS (Figure 1a) significantly (47·3%) improved tumour detection and resection (Figure 1b) and reduced the post-operative tumour burden compared to standard white-light surgery in orthotopic HGSOC xenografts. The application of CD24-AF750 in four heterogenous HGSOC PDX allowed identification of metastatic tumour lesions (Figure 2 a), and CD24 expression was confirmed by immunohistochemistry (Figure 2 b), verifying the clinically relevant targeting ability of the tracer.


The addition of CD24 as a promising biomarker for the application of fluorescence-guided surgery can aid further delineation of metastases and be of greater benefit to ovarian cancer patients with poor prognosis. CD24 is overexpressed in many other solid tumours, highlighting the potential of CD24-FIGS as an aid during surgical resection of other solid cancers.


The authors acknowledge the support from the MARIE SKLODOSWKA-CURIE ACTION (proposal number 675743; acronym ISPIC) carried out within the H2020 program MSCA-ITN funded by the EU.

[1] 1. du Bois A, Reuss A, Pujade-Lauraine E, Harter P, Ray-Coquard I, Pfisterer J. Role of surgical outcome as prognostic factor in advanced epithelial ovarian cancer: a combined exploratory analysis of 3 prospectively randomized phase 3 multicenter trials: by the Arbeitsgemeinschaft Gynaekologische Onkologie Studiengruppe Ovarialkarzinom (AGO-OVAR) and the Groupe d'Investigateurs Nationaux Pour les Etudes des Cancers de l'Ovaire (GINECO). Cancer. 2009;115(6):1234-44.
[2] van Dam GM, Themelis G, Crane LM, et al. Intraoperative tumor-specific fluorescence imaging in ovarian cancer by folate receptor-alpha targeting: first in-human results. Nat Med. 2011;17(10):1315-9.
[3] Randall LM, Wenham RM, Low PS, Dowdy SC, Tanyi JL. A phase II, multicenter, open-label trial of OTL38 injection for the intra-operative imaging of folate receptor-alpha positive ovarian cancer. Gynecologic oncology. 2019;S0090-8258(19):31396-4.
[4] Kleinmanns K, Bischof K, Anandan S, Popa M, Akslen LM, Fosse V, Karlsen I, Gjertsen BT, Bjørge L, McCormack E. CD24-targeted fluorescence imaging in patient-derived xenograft models of high-grade serous ovarian carcinoma. 2020, submitted.
Figure 1: Survival surgery in OV-90luc+ orthoptic xenografts.

(a) Representative intraoperative white light (colour), near infrared (NIR 800) fluorescent and pseudo-coloured fluorescence intensity merge images from CD24-targeted fluorescence image-guided surgery (FIGS) of intra-abdominal metastases. (b) Comparison of the number and total weight of resected metastatic lesions between the CD24 FIGS cohort (n = 8) and white light control surgery (n = 8)

Figure 2: Intraoperative CD24-targeted fluorescence imaging of four heterogeneous HGSOC PDX
(a) Intraoperative white light (colour), near infrared (NIR 800) fluorescent and pseudo-coloured fluorescence intensity merge images show positive identification of primary tumour and small metastatic lesions in four orthotopically implanted PDX models of HGSOC. Ex vivo optical imaging confirmed fluorescence specificity. (b) CD24 expression was confirmed by IHC in fluorescence positive PDX lesions. CD24 staining was evaluated in all five resected lesions with an immunoreactivity score (4·4 ± 1·1, scale: low 1 – 7 high). CK8 staining was used to confirm human tumour origin. 
Keywords: Intraoperative imaging, biomarker, near-infrared fluorescence, HGSOC, patient-derived xenograft