Skip to main content

Use of innovative strategies for beta-cell imaging in diabetes mellitus

Final Report Summary - BETAIMAGE (Use of innovative strategies for beta-cell imaging in diabetes mellitus)

Executive summary:

BetaImage aimed to develop new approaches for in vitro and in vivo life beta-cell imaging. BetaImage research aimed to accomplish: (1) successful clinical PET/SPECT beta-cell imaging in humans, (2) increase the number of beta-cell specific targets for in vivo imaging using a Systems Biology approach to characterise the global receptome of beta-cells, (3) increase the chance for successful tracer development by development of key technologies and methodologies to enhance the efficacy of tracer characterisation, (4) development of imaging methodology for research and clinical application of PET, SPECT, optical imaging, and MRI, and (5) setting up the basis for successful future European scientific and economic competition in the field of beta-cell imaging and diabetes by systematic dissemination of expertise.
In the runtime of the project, BetaImage scientists have identified several potential new targets, including beta cell specific splice variants of non-beta cell specific targets. Specific ligands for these targets have been developed and for others, development is continued in subsequent projects. The lead compound (serving as reference) and several analogues have been optimised and labelled with a range of suitable radionuclides for SPECT and PET imaging. Successful in vivo determination of the pancreatic beta cell mass with the lead compound has been achieved and the acquisition and reconstruction protocols have been optimised in rodent models. Furthermore, quantitative imaging of transplanted beta cells has been achieved with different tracers. A clinical study providing a proof-of-principle of in vivo imaging of pancreatic beta cell mass is performed.
In addition to the novel targets defined, two beta-cell specific targets have been identified with known ligands and their suitability for in vivo imaging of beta cells has been confirmed. Bioluminescences as well as fluorescence imaging have been used for in vivo imaging of pancreatic and transplanted beta cells. Mn++ was successfully used for imaging of pancreatic islets in vivo and ex vivo. xf-FDOCT technology has been optimised and repeated in vivo 3D real-time imaging of native and transplanted beta cells has been achieved with very high resolution in animal models, enabling to monitor the decrease of beta cell mass in the course of diabetes in rodent models as well as real time blood flow measurements and correlation of these parameters to tracer distribution (fluorescent and metal-labelled tracers)
Training courses and individual training have been performed, including courses on small animal imaging, labelling techniques, and beta cell transplantation. In addition, several individual trainings have been performed in several places addressing specific needs of BetaImage partners. The coordinator of BetaImage has participated as 3 co-chair/program committee of the “Imaging the Pancreatic Beta Cell” conference of the National Institutes of Health in Washington in April 2009 and in Bethesda in April 2013. International conferences have been held in association with the European Association for the Study of Diabetes (Stockholm 2010), the European Association for Nuclear Medicine (Birmingham 2011) and the World Molecular Imaging Society (Dublin 2012) as well as a workshop on beta cell imaging on the EASD meeting in 2010 with nearly 3000 attendants. In 2010, a theme issue of “Current Pharmaceutical Design” on beta cell imaging has been published with the coordinator of BetaImage as the guest editor and contributions from several BetaImage partners as well as a review article in Diabetologia in 2012 covering the topics of the meetings in Stockholm and Dublin.
Reliable, sensitive, specific, and non-invasive methods for a comprehensive structural and functional characterization of living pancreatic beta-cells in vivo and in vitro would enhance our understanding of the pathophysiology of both type 1 and type 2 diabetes and would help to assess novel therapeutic approaches to diabetes, whether based on pharmacology (e.g. GLP-1 derivatives and agonists etc.) or cell replacement (e.g. islet transplantation, implants of surrogate insulin-producing cells etc.). The technology developed in BetaImage would therefore have a major impact on diabetes research as well as for clinical management of the (at this point in time) estimated 200 million diabetics worldwide. Especially the clinical data obtained with an Exendin analogue for radiotracer imaging suggest that for the first time, the scientific community has an imaging tool available that provides information about beta cell mass independent of beta cell function. Therefore, the results obtained in BetaImage have a high potential impact for Europe, scientifically as well as economically.

1.2 Summary description of project context and objectives
Background and Aims
According to the WHO ( there are currently 180 million people suffering from diabetes mellitus worldwide with an expected doubling of this number until 2030. In Europe, the number of diabetics is currently estimated to be 30 million. The WHO predicts that in upper- and middle-income countries (including Western Europe) the number of diabetes-related deaths will increase by 80% between 2006 and 2015. Despite these alarming numbers, the precise molecular mechanisms leading to the decrease in beta-cell mass responsible for the development of impaired glucose tolerance and diabetes still remain to be elucidated. Specifically, we do not know the natural history of beta-cell mass loss in human diabetes, and we are unable to detect both, the endogenous attempts of beta-cell neogenesis and the putative beneficial effects of interventions aiming to protect beta-cells and/or increase their replication during diabetes.
Reliable, sensitive, specific, and non-invasive methods for comprehensive structural and functional characterization of living pancreatic beta-cells in vivo (and in vitro) would enhance our understanding of the pathophysiology of both, type 1 and type 2 diabetes. Furthermore, the ongoing development of novel therapeutic approaches to diabetes, whether based on pharmacology (e.g. GLP-1 derivatives and agonists etc.) or cell replacement (e.g. islet transplantation, implants of surrogate insulin-producing cells etc.) also calls for rapid methods for the longitudinal in vivo assessment of the beta-cell mass and function. Apart from direct determination of beta-cell mass, such imaging techniques will also be used for imaging of the response of beta-cells to conditions leading to beta-cell dysfunction and eventually apoptosis.
This project has aimed to identify and apply innovative approaches for sophisticated in vivo and in vitro life beta-cell imaging, especially beta-cell mass regulation (loss and neogenesis) with the perspective of entering first clinical trials. Our overall concept was to: (1) focus on imaging technologies that offer the potential of entering first clinical trials in the runtime of the project, (2) evaluate new potential imaging strategies with the highest possibility of success, (3) focus on research based on European excellence, and (4) create independent European expertise in beta-cell imaging to succeed, in the long term, in a scientifically and economically highly competitive field.
1. Imaging technology with a perspective for clinical trials:
Beta-cells constitute less than 2% of the pancreatic mass and are scattered throughout the pancreas in tiny islets of Langerhans. Single islets cannot yet be spatially resolved non-invasively in vivo in humans. A highly sensitive method is thus required allowing to easily quantifying uptake of a tracer in the target organ, i.e. the endocrine pancreas, without necessarily requiring resolution of single beta-cells or islets. Therefore, new approaches for beta-cell mass imaging with a clear clinical perspective should primarily rely on chemical resolution with high sensitivity, most probably to be achieved by using tracers for positron emission tomography (PET) and single photon emission computed tomography (SPECT),. We will use novel radiotracers for GLP-1 receptor imaging) to succeed in clinical beta-cell imaging within the runtime of the project..
2. Evaluation of new imaging strategies:
Innovative strategies such as the construction of labelled “design” molecules/peptides targeting novel surface beta-cell antigens, and the parallel use of different molecules targeting diverse beta-cell surface proteins may increase the number of potential PET applications in beta-cell imaging. New targets will therefore be defined based on identification of novel surface beta-cell proteins identified by a functional genomics and Systems Biology approach. . Tracers will then be synthesised, characterised, and optimised in animal models. Extended focus Fourier domain optical coherence tomography (xf-FDOCT) will be optimised for in vivo real-time dynamic imaging of tracer distribution in animal models. This innovative imaging strategy complemented by PET/MRI imaging will serve as technological platform to improve efficiency of beta-cell tracer characterisation and optimisation. Finally, strategies for specific targeting of pancreatic and transplanted beta-cells with nanoparticles for magnetic resonance imaging (MRI) will be developed.
3. European excellence in research:
Development of tracers based on peptides and organic molecules for imaging has traditionally been a strength of European research in molecular imaging. In diabetes research, unique European expertise exists for analysis of beta-cell functional genomic data by a Systems Biology approach and for generation of models for beta-cell loss and neogenesis. Therefore, will (1) use a Systems Biology approach for definition of novel beta-cell targets, (2) focus on development of small organic molecules and peptides, and (3) characterise novel tracers in animal models before proceeding to clinical assessment.
BetaImage strives to disseminate expertise, experience, and technology to create a long-lasting European expertise in beta-cell imaging.
Objectives at a glance
1. Achieve the clinical proof-of-principle for non-invasive beta-cell mass imaging.
2. Develop new tracers with a clear clinical perspective.
3. Increase the number of potential beta-cell imaging strategies by identifying new targets using a Systems Biology approach
4. Develop new tracers for imaging of novel beta-cell targets as defined by the Systems Biology approach.
5. Establish xf-FDOCT as a highly efficient method for 3D in vitro and dynamic in vivo imaging of beta-cells.
6. Build up a strong and long-lasting European expertise in the field of molecular imaging of beta-cells.
In conclusion, this project aimed to develop new approaches for in vitro and in vivo life beta-cell imaging with a focus on beta-cell mass, loss, and neogenesis. BetaImage research has accomplished to: (1) successfully establish clinical beta-cell imaging in humans based on the use of a novel tracer for the GLP-1 receptor, (2) increase the number of beta-cell specific targets for in vivo imaging aiming to increase the number of different beta-cell imaging strategies, (3) increase the chance for successful tracer development by development of key technology and methodologies to enhance the efficacy of tracer characterisation for beta-cell imaging, (4) develop new tracers and optical imaging methodology for future research application, and (5) set up the basis for successful future European scientific and economic competition in the field of beta-cell imaging and diabetes by systematic dissemination of expertise.
Work strategy and general description
In BetaImage, we have utilized strong interdisciplinary synergy to identify novel beta cell specific targets and tracer molecules for beta cell imaging. Through thorough characterization and optimization steps, several promising new targets and ligands have been identified that have further been developed as radiotracers or contrast agents for MRI. xfDOCT has been established as a platform for efficient in vivo imaging of pancreatic islets, islet specific blood flow and dynamic real-time tracer kinetics, enabling improved characterization of tracer molecules. Finally, clinical GLP-1 receptor imaging with highly promising results has been achieved thought the efforts of the whole consortium, utilizing kit formulations produced according to full GMP standards.
The strength of this highly structured project has been this systematic step-by-step approach of target definition, tracer development, tracer characterisation and optimisation, finally leading to clinical assessment. This step-by-step approach to clinical beta-cell imaging, representing the backbone of the project, spanned from the definition of novel targets for beta-cell imaging via ligand development and testing up to clinical trials. The steps of this project systematically followed the line of tracer development in advanced molecular imaging (step 1 definition of targets; step 2 target characterisation; step 3 tracer development; step 4 tracer characterisation; step 5 tracer optimisation; step 6 preparation for clinical trials and toxicity testing; step 7 clinical assessment). These steps have been achieved through highly complementary work packages:
BetaImage has been organised in 6 work packages, including 4 research work packages, one management work package and one work package responsible for dissemination. These work packages had been chosen based on the lead expertise required: Systems Biology and functional genomics, tracer development, animal models for tracer characterisation, clinical assessment of beta-cell tracers, management, and expertise/knowledge dissemination. The research work packages were divided in tasks numbered 1-15, contributing to the steps of tracer development.
WP1 (Target Definition): this WP has used array analysis and a Systems Biology approach in in vitro and in vivo models of diabetes to define potential targets for life beta-cell imaging (task 1). These targets have been evaluated with respect to their biological properties and their suitability for in vitro and in vivo beta-cell imaging (task 2). In vitro models have been established to facilitate the choice of adequate ligands (task 3). The most suitable target proteins have then been chosen for further ligand development (task 4).
WP2 (Tracer Development): in this WP, ligands that are suitable to serve as tracers for beta-cell imaging have been defined and developed (task 4 and 5). These ligands have been modified for high metabolic stability and coupling of molecules for labelling (i.e. chelators for radiolabelling, optical dyes, nanoparticle coating, task 6). These tracers were characterised in close interaction with WP 3 in in vitro and in vivo models of diabetes and further optimised with respect to metabolic properties, binding affinity, chemical stability of the label etc. (task 7 in close cooperation with WP3; expertise in animal models of diabetes has been provided via WP 3).
WP3 (Tracer Characterisation): WP 2 and 3 have been highly interacting since tracer optimisation in WP 2 depends on tracer characterisation in WP 3. Both are translational WPs, but WP 2 focused more on basic tracer development while WP 3 has been bridging the gap between pre-clinical and clinical development of tracers. Tasks 7 and 8 were performed in close interaction between WP 2 and 3. WP 3 has also establish a standard methodology with xf-FDOCT/OCM as technological platform complementary to PET and MRI for tracer characterisation and development, focussing on beta-cell mass, islet mass, islet blood flow, and 3D dynamic in vivo tracer microdistribution. The aim was to develop a platform suitable for increasing efficacy of tracer development by quick characterisation of key parameters. The complete methodology has been established in cooperation of all partners and WPs (tasks 9 and 10). Finally, task 11 has been focused on ethical aspects of animal experiments.
WP4 (Clinical Assessment): the tracers considered suitable for clinical imaging have been prepared for clinical use including toxicity testing (task 12). In case further optimisation or other changes was required (as for example for Ga-labelled Exendin), this has been done in conjunction with WP 2. The main focus of WP 4 lay in preparation and implementation of clinical assessment, including preparation of clinical imaging protocols, coordination of ethical issues, distribution of clinical grade kit formulations for tracer labelling (task 13), and coordination of clinical studies (task 14). Finally, task 15 was focused on ethical aspects of studies in humans or human tissues.
WP5 (Management, led by GABO): The tasks of this WP were to (1) ensure proper scientific functioning of the project, (2) ensure that all contractual duties of the consortium are carried out, (3) set up and maintain an effective communication structure, and (4) manage and coordinate the processes related to capture and exploitation of intellectual property. The Coordinator has mainly carried out the scientific management while the administrative management mainly belonged to GABO’s duties.
WP6 (Dissemination, led by SKU): the aim of WP6 was to disseminate the expertise and knowledge already present among the partners as well as the new knowledge that will be generated by the project. WP 6 was split in 4 tasks aiming to: (1) transfer skills and expertise within and outside of the consortium, (2) build up a European expertise for successful scientific and economic competition, (3) inform the diabetes and imaging research community as well as the public about the progress and achievements of the project, and (4) to organise courses in close cooperation with tasks 11 and 15 of the research WPs.
Outside the consortium, the expertise and the results obtained have been disseminated in the scientific community via scientific publications, by several workshops, training programmes, and international conferences (EANM, EASD, WMIC). An EASD study group “Biomedical Imaging in Diabetes” has been established. Furthermore, achievements of the project have also be communicated to the general public via patient organisations and internet.
Management structure and procedures
The Project Coordinator Prof. Martin Gotthardt ensured the smooth operation of the project and guaranteed that all efforts were focused towards the objectives. He submitted all required progress reports, deliverables, financial statements to the European Commission, and, with the assistance of GABO:mi he was responsible for the proper use of funds and their transfers to participants. The BetaImage office was established by and based at the coordinating institution in Nijmegen, The Netherlands and at GABO:mi in Munich, Germany. The Project Office at the Coordinator was concerned with the scientific management and the co-ordination of all research activities. The Project Office at GABO:mi was responsible for administrative, financial and contractual management and the organisational co-ordination of the project activities.
The Project Governing Board (PGB) was in charge of the political and strategic orientation of the project and acted as the arbitration body. It met once a year unless the interest of the project required intermediate meetings. The Project Coordination Committee (PCC) consisted of all work package leaders and the Coordinator and was in charge of monitoring all activities towards the objective of the project in order to deliver as promised, in due time and in the budget. The Project Coordination Committee met every six months during the funding period. Furthermore, a scientific advisory board as well as an ethics advisory board were implemented to ensure a high standard of research and monitor the progress of the project. They were always invited to the annual Governing Board Meetings.
The scientific advisory board consisted of
Veikko Koivisto, MD PhD, Managing Director of the Lilly Research Laboratories in Hamburg, Germany
Donald F. Steiner, MD PhD, Professor, Biochemistry & Molecular Biology, Dept. Med., Howard Hughes Medical Inst. and University of Chicago
Heinz H. Coenen, PhD, Professor of radiochemistry and head of the Institute of Nuclear Chemistry at the Research Center Juelich, Germany
The ethics advisory board consisted of
Pirjo Nuutila, leader of WP4
Luc Bowens, leader of WP3
Veikko Koivisto from the scientific advisory board
Finally, BetaImage established a special task group for promotion of female scientists. This task group consisted of Pirjo Nuutila (Turku), Anne Grapin-Botton (Lausanne), Daisy Flamez (Brussels), and Paolo Meda (Geneva). It supervised and promoted the career advancement of female junior researchers in the consortium, in close co-operation with the Project Coordination Committee and the Coordinator.
Objectives of BetaImage:
1. To achieve the clinical proof-of-principle for non-invasive beta-cell mass imaging with PET for determination of beta-cell mass, mainly relying on an Exendin-based tracer for GLP-1 receptor imaging currently undergoing pre-clinical optimisation.
2. To develop new tracers with a clear clinical perspective within the runtime of BetaImage for known promising targets for which no optimal tracers exist so far (Nanobodies® for targeting the VMAT2 transporter; the concept is explained in 1.2.6-2).
3. To increase the number of potential beta-cell imaging strategies in vivo and in vitro by identifying new targets that could be used for PET as well as other imaging techniques by definition of new targets using a Systems Biology approach to characterise the global receptome of beta-cells.
4. To develop new tracers for imaging of novel beta-cell targets as defined by the Systems Biology approach using in vitro and in vivo models (PET, SPECT, optical imaging, MRI).
5. To establish xf-FDOCT as a highly efficient method for 3D in vitro and dynamic in vivo imaging of beta-cell mass, microscopic tracer distribution, islet size, and islets perfusion.
6. To establish a streamlined methodology to increase efficiency of high-throughput tracer characterisation based on in vitro models and in vivo complementary xf-FDOCT (microscopic) and PET/MRI (macroscopic) tracer imaging as technological platform.
7. To build up a strong and long-lasting European expertise in the field of molecular imaging of beta-cells by dissemination of expertise and knowledge acquired within BetaImage.

1.3 Description of the main S&T results/foregrounds of BetaImage
Workpackage 1
Task 1: Definition of the ligand-receptome of beta-cells in health and disease using a Systems Biology approach.
In a first system biology study MPSS analysis was performed on human islet samples under control condition and cytokine (IL-1β + IFN-γ) treatment and an Affymetrix transcriptome micrarrays analysis was performed on human islet samples, isolated rat beta cells and INS-1E cells. The data were compared with similar data for other tissues obtained from the LICR MPSS data set ( T1DBase and GEO. The IPA databank and the TMHMM software were used to select respectively proteins known to be located in the cell membrane or suspected to be transmembrane proteins. A list of 12 candidate genes was proposed and several of them are presently under study (First approach).
An Illumina RNA-seq analysis was performed on five human islet samples under control condition, cytokine treatment and palmitate treatment. The data were compared with RNA-seq data from the Illumina Human Body Map containing 5 tissues. A new list of candidate genes was obtained and a novel target was selected for further study (Second approach).
Task 2: Assessment of the specificity of candidate biomarkers for beta-cells in normal pancreas (rodent and human).
FXYD2 has 3 splicing isoforms: FXYD2-γa, - γb and -γc. PCR showed that all 3 are expressed in mouse kidney, but only the first 2 are expressed in mouse islets. Antibodies were generated against FXYD2-γa and -γb and used for immunohistochemistry ; isolated rat beta cells and INS-1E cells turned out to bind both antibodies but human islets bind only the antibody specific for FXYD2-γa, hence only this was used for further studies.
It was shown that cytokine treatment of human islets or rat islet dispersed cells does not change the expression of FXYD2-γa on transcript and protein level. Immunohistochemistry of human, macaque and rat pancreas showed that FXYD2-γa is only expressed in insulin producing cells.
Screening of a human tissue array using commercially available antibodies against the novel targets showed that those proteins are expressed in beta cells. Nanobodies against one of the targets have been generated and coupled with DTPA, DOTA and NODAGA for labeling with indium and gallium. Affinity control experiments will be done.
Task 3: Establishment of in vitro models to validate the specificity/selectivity and affinity of ligands with potential for tracer development.
We focused on creating stably transfected cell lines that express two of the novel targets, with the purpose of gratfting them in NUDE mice and using these mice as test model for imaging. As cell lines we chose a line of mouse fibroblasts that were stably transfected and 6 tumoral cell lines (HeLa, CAPAN-2, PANC-1, INS-1E, MIN6, CHO). First it was checked by RT-PCR that they do not express the targets under basal conditions. We selected Hela and CHO cells as suitable candidates and performed a stable transfection with those targets in human and mouse. Transfection was confirmed by RT-PCR, sequencing and Western blotting.
Task 4: Definition of the most suitable ligands by the partners.
For the identification of ligands of the new targets via IPA analysis and literature screening no new ligands were found. Nanobodies against two of the targets suitable for imaging have been produced. Final selection of better suited ligands than tested so far is still ongoing.
Workpackage 2:
Task 4: Ligand definition
In cooperation with WP01, targets have been defined that would be suitable for imaging of beta cells in vivo. Apart from definition of suitable targets, an alternative approach has been the identification of ligands known to bind to targets expressed on beta cells. Ligands (and respective targets) that have been identified as beta cell targeting compounds include KISS Peptide (GPR54), Maxadilan (PAC1), and GIP (GIP receptor).
Task 5 & 6: Characterisation of ligands in vivo & development and labelling of tracers for beta-cell imaging
The ligands chosen in task 4 have been characterised in vivo, including evaluation of different chelators and radionuclides (mainly DTPA, DOTA, NODAGA, HYNIC, and 125I, 131I, 111In, 99mTc, 68Ga, 64Cu, 18F). Alloxan treated mice and rats have been the most frequently used animal model. At later stages of the project, other animal models were also included, such as NOD mice or INS-DTR mice. Biodistribution experiments have been performed in animals and micro-distribution in the pancreas was evaluated by autoradiography / immunohistochemistry. Only ligands with promising characteristics with respect to stability, affinity, specificity, pharmacokinetics etc. were passed on to the next step of tracer development. Several ligands with sub-optimal characteristics with respect to metabolic stability, labelling, or pharmacokinetics that need to be optimised to meet the required criteria have only been passed on to task 6 after the modifications required appeared to be successful. The labelling protocols have been optimised to allow efficient labelling of the tracers with high purity and yield. Furthermore, the labels (e.g. chelator-radiometal complex) have also been tested for chemical stability to ensure that the “dye” (e.g. radionuclide) does not dissociate from the tracer. For MRI, small nanoparticles (<30 nm) have been coated with different polymers. This coat mediates the attachment of fluorochromes (typically Cy5.5) and a moiety (receptor-targeting polypeptides, for example) for targeting of beta cells. Different sizes and coatings have been compared in order to optimize the conditions that allow for both a sufficient persistence in the circulation and a sufficiently high binding to the targeted beta-cells of the islet.
Task 7: Evaluation of labelled compounds in the vitro and in vivo models provided by WP1 & 3
The labelled compounds have been evaluated in the in vitro and in vivo models provided by WP1 and 3. We have utilized different mouse models, including alloxan-treated C57Bl6 mice and Wistar rats, NOD mice, INS-DTR mice and nude mice in which subcutaneous tumor models were utilized for efficient characterization of tracer molecules. These animal models were only used after initial characterization of tracer molecules in cell culture using MIN6, INS and RINm5F cells and islets. The stability of the compounds (i.e. biological stability, chemical stability, labelling stability) will be tested in human serum, cell culture, and animal models. After initial characterization, the most suitable tracer molecules have been passed on to animal evaluation. In vivo imaging has been performed with the most suitable tracers using small animal PET, SPECT, optical imaging, and MRI. Within WP 2, the labelling procedures for the most promising compounds has been optimised in order to develop kit formulations for distribution within the consortium for further research. For MRI tracers, control mice and rats have been injected iv with the nanoparticle tracers and studied by in vivo fluorescence measurements and MRI to a assess the intensity and specificity of labelling of native islets, and for the eventual effects on beta-cell function as evaluated by the circulating levels of glucose, insulin and C-peptide. The same type of experiment has also been realised on transplanted islets.
Workpackage 3:
Task 8: Evaluation of beta-cell tracers in animal models of diabetes
Tracers selected for further characterization in previous tasks were evaluated. Task 7 and 8 were highly interacting tasks, so that many of what has been said for Task 7 is also applicable to Task 8. The characteristics the tracers have been tested for include tracer uptake in vivo, specificity for beta-cells, binding affinity, tracer retention in the target organ, labelling efficiency etc. Selected GLP-1R targeting tracers have been evaluated for their use in quantifying the beta-cell mass in animal models. Most tracer molecules evaluated were suitable for imaging of subcutaneous tumors, but for imaging of pancreatic islets they either lacked specificity or sufficiently high uptake in the target cells. Finally, most GLP-1R targeting trace molecules evaluated were suitable for imaging of pancreatic beta cells (although we have focused on those tracers only that provided a clear advantage over other GLP-1 analogs). Other tracer molecules have shown promising characteristics for imaging of tumor tissue, but could not fulfill the high demands for in vivo imaging of the islets of Langerhans distributed throughout the whole pancreas.
Task 9 and 10: Development and validation of xf-FDOCT as a platform for 3D real-time in vivo imaging of tracer distribution, pancreatic islets, blood flow, and beta-cell mass
xf-FDOCT (in the meantime renamed to OCM = optical coherence microscopy because of the cellular resolution reached in this project) has been established for in vivo imaging in animals. The imaging hardware has been optimised with respect to size and shape to make imaging more convenient (for example by reducing the necessary surgical intervention). At the same time, algorithms for image acquisition and reconstruction have been optimised. Iin vivo imaging protocols have been established. After establishing efficient in vivo imaging strategies for pancreatic islets, imaging protocols for determination of in vivo tracer distribution have been developed, including techniques for imaging of tracers based on their optical contrast, which will be enhanced by labelling tracers with metals as well as fluorochromes for simultaneous xf-FDOCT and fluorescence imaging. Imaging protocols have been developed to dynamically examine tracer distribution on a microscopic level. The last steps were development of protocols and technology for Doppler measurement of blood flow and a technique to determine mass of individual islets.
Task 11: Increase the awareness for ethical aspects of animal experiments.
Specific ethical issues for this task include (1) humane endpoints in diabetes models, (2) decrease distress of diabetic animals, (3) how to recognize inconvenience in diabetes models before serious distress occurs, and (4) raise awareness for choice of the animal models with the least inconvenience for the animals. The necessity of central animal ethics courses has been assessed in the consortium and it appeared that after the ethics course held in Geneva, BetaImage scientists have followed local ethics courses required for animal handling so that necessity for central ethics courses could be identified.
For all animal experiments conducted in BetaImage protocols have been collected including ethical approval documents. All specific information/standards about BetaImage animal models as well as standard protocols for pain reduction, anesthesia, and euthanasia are available via the home page of the project.
Workpackage 4:
Task 12: Preparation and assessment of clinical grade tracers: promising tracers as developed in WP1-3 will have to be optimised and prepared for use in humans.
After the successful development of a kit for the preparation of 111In-DTPA-exendin-4 as a tracer for SPECT we developed a kit for the preparation of 68Ga-labeled exendin-4 for PET. NODAGA was selected as a chelator for the labelling with 68Ga after evaluation of several chelators for 68Ga-labeling, since NODAGA exhibits excellent labelling characteristics. NODAGA-exendin-4 was synthesized (piCHEM Forschungs-und Entwicklungs-GmbH, Graz, Austria) and the labelling was optimized. NODAGA-exendin-4 could be labelled with 68Ga with a specific activity of 1.2 GBq/nmol (at time of synthesis) when HEPES (final concentration: 0.3 M, final pH 3.5) was used as a buffer. The new tracer has successfully been characterized and validated in vitro and in vivo as compared to the clinical grade tracer for SPECT imaging.
Since a “GMP-grade” 68Ge/68Ga generator is currently still unavailable (although the generator should have been available by 2012 according to the producer), the present generator is currently validated for approval for clinical use through the local pharmacy (at the RUNMC) (validation includes sterility, pyrogenicity, breakthrough of 68Ge and reproducibility of the elution).
Task 13: Set up methods/protocols for imaging beta-cells in vivo in humans: imaging in humans requires sophisticated protocols for imaging and dosimetric analysis.
We have successfully developed optimized standardized SPECT and PET acquisition and reconstruction protocols for human use. The high uptake into the kidneys causes challenges for image reconstruction not known from other tracer molecules. Therefore, a sophisticated phantom has been developed in order to optimize imaging data in the clinical study. The results of the phantom scans allowed us to define optimized settings for image acquisition. In addition, we have defined the optimal reconstruction software package for SPECT (Hybrid Recon, Hermes Medical Solutions) and optimized the reconstruction settings.
In addition, the SPECT scanner used in the study has been calibrated and a protocol has been developed that will allow to calibrate other scanners in order to achieve quantitative SPECT data for direct comparison in multi-centre trials. The PET scanner has been validated through the EANM (European Association for Nuclear Medicine) accreditation program.
Task 14: Implement studies with respect to beta-cell mass in T1D and T2D.
Studies in healthy volunteers and patients with type diabetes are currently conducted and the data evaluation is still ongoing. Completed results are highly promising and we expect that the evaluation of the remaining data will further support our findings obtained so far. Delays reported previously, mainly caused by changes in European legislation towards implementation of investigational medicinal products in clinical studies and difficulties faced in the production of the clinical grade tracer, have in part been compensated. In particular, the main hurdles for clinical studies (production of the radiotracer and to obtain the necessary approvals for clinical use including full GMP documentation) have been taken for SPECT and PET tracers now available as kit formulations, so that clinical studies building up upon the current cross-sectional small scale studies can comparably easily be commenced.
Task 15: Human ethics: increase the awareness for ethical aspects of studies with humans or experiments with human tissue samples or other material.
The need for additional ethics courses in the reporting period has been evaluated within the consortium. In view of the (now in large parts compensated) delay in the conduction of the clinical studies, there was no need for consortium-wide ethics courses. Instead, all investigators involved in the clinical studies have participated in additional local ethics courses and trainings.

Concise description of the most significant results of BetaImage
Target definition
For selecting alternatively spliced candidate biomarkers, first a system biology study MPSS analysis was performed on human islet samples under control condition and cytokine (IL-1β + IFN-γ) treatment and an Affymetrix transcriptome micrarrays analysis was performed on human islet samples, isolated rat beta cells and INS-1E cells (Bouckenooghe et al. 2010).
After this, an Illumina RNA-seq analysis was performed on five human islet samples under control condition, cytokine treatment and palmitate treatment. The data were compared with RNA-seq data from the Illumina Human Body Map containing 5 tissues (Eizirik et al. 2012).
Five human islet preparations were used for sequencing (Eizirik et al. 2012; Lists of differentially expressed transcripts were generated using Flux Capacitor. To create a library of potential candidate biomarkers, comparison was made with 16 tissues from the Illumina Body Map transcriptome project RNA-seq data (ENA accession ERP000546). The selected candidates were transcripts with a fold change of at least 100 between pancreatic islets and other tissues. Next, the location of the protein was screened using IPA and TMHMM prediction programs to search membrane protein as a good biomarker for imaging is a membrane protein.
Preclinical GLP-1 receptor imaging
We explored biodistribution and specificity of the 111In-exendin tracer in a rat model of alloxan-induced beta cell loss and correlated pancreatic uptake with BCM. Biodistribution in Brown Norway rats 1 h p.i. showed 111In-exendin accumulation in the pancreas to a concentration of 0.29 ± 0.04 %ID/g. Co-injection of an excess unlabeled exendin resulted in a marked decrease in pancreatic uptake (0.04 ± 0.01 %ID/g), indicating GLP-1R mediated uptake. Importantly, pancreatic uptake of 111In-exendin was reduced by >80% in severely diabetic rats with very low BCM, pointing towards beta cell specificity of the tracer. This specificity was confirmed by ex vivo autoradiography of pancreatic sections of healthy and alloxan-treated rats after in vivo 111In-exendin tracer injection. High activity concentrations of 111In-exendin were observed in hot spots throughout pancreatic sections in control rats representing the islets of Langerhans while very low uptake of the tracer was observed in the exocrine pancreas. Treatment with alloxan markedly reduced 111In-exendin uptake in islets. Immunohistochemical staining of the pancreatic sections with an anti-insulin mAb showed co-localization of the stained islets with the hot-spots in autoradiography, demonstrating specific accumulation of the tracer in the islets. These results were confirmed by microautoradiography combined with immunohistochemical staining of consecutive sections with an anti-insulin mAb. The uptake of 111In-exendin in the pancreas was plotted against the BCM, showing a very good correlation between these two parameters (Pearson r=0.89).
To visualize the uptake of 111In-exendin in the abdominal organs, SPECT scans of dissected abdominal organs including the pancreata of control and alloxan-treated Brown Norway rats were acquired. The pancreata of untreated Brown Norway rats were clearly visualized, whereas the pancreata of both alloxan-treated rats and rats co-injected with an excess of unlabeled exendin were barely visible by ex vivo SPECT. A strong signal was observed in the pancreas, whereas other abdominal organs showed low or non-detectable uptake of the tracer except for the proximal part of the duodenum and the distal part of the stomach that showed relatively high uptake.
The SPECT images of dissected tissues enabled accurate interpretation of the SPECT images of Brown Norway rats. In healthy Brown Norway rats, uptake in the pancreas was visible in in vivo SPECT scans after injection of 111In-exendin, with lower or negligible uptake observed in rats with reduced BCM. High uptake was also observed in the kidneys and lungs. The tracer uptake of 111In-exendin as determined by quantitative analysis of the SPECT images correlated in a linear manner with the BCM determined by morphometric analysis.
It was confirmed that the VOI used for quantification of pancreatic uptake was placed in the pancreas by simultaneous SPECT imaging after injection of [99mTc]Demobesin 3, specifically targeting the gastrin-releasing peptide receptor in the exocrine pancreas.
This well characterized rat model for beta cell loss enabled us to demonstrate the specificity of the tracer, the feasibility of in vivo visualization of the pancreas and the correlation of in vivo tracer uptake with BCM. Quantitative RT-PCR indicated that the GLP-1R expression in endocrine pancreas is substantially higher in humans than in rats, promoting clinical feasibility. In rats, there is a 13-fold higher GLP-1R mRNA expression in the endocrine pancreas as compared to the exocrine pancreas while in human tissues the endocrine-exocrine ratio reached a factor of 47. Of note, we cannot exclude some contamination of the exocrine pancreas with insulin producing cells, so that the actual endocrine – exocrine ratio of GLP-1R mRNA expression may even be higher.

Improved visualization of the pancreas and correlation with beta cell mass
In order to evaluate future β-cell tracers in vivo, we aimed to develop a standardised non-invasive method allowing to quantify a prospective β-cell tracer within the pancreas. A known problem in small animal imaging is that soft tissues like pancreas are not anatomically identifiable with CT. To solve this problem we aimed at developing a dual isotope tracing method allowing to delineate the mouse pancreas, which represents our anatomical region of interest, and to subsequently analyze the uptake of a putative β-cell tracer in that region by SPECT. Following the observation that 123I-L-phenylalanine ([123I]-2-iodo-L-phenylalanine) has a very high uptake in the pancreas of rodents, 123I-L-phenylalanine was used to delineate the pancreas. 111In-exendin-3 (-Lys40(111In-DTPA)-exendin-3), that has previously been used to target insulinomas, was used as a β-cell tracer.
We used a controlled β-cell depletion model, i.e. RIP-DTR mice that was further validated by qRT-PCR and morphometric analysis. These mice express the diphteria toxin receptor on cells with an active insulin promotor. After administration of diphteria toxin there is an almost complete ablation of the β-cells. Because the transgene is located on the X-chromosome, heterozygous females only express the diphteria toxin receptor on 50% of their β-cells, and thus only 50% of their β-cells are ablated, while in mice that are homozygous for the transgene, nearly 100% of the β-cells are ablated.
Seven days after DT treatment, blood glucose concentrations of DT treated RIP-DTR+ mice were significantly elevated to 490.3 - 109.7 compared to non-treated RIP-DTR+ mice (135.3 - 13.28) (p=0.0325) and insulin mRNA levels dropped to 0.51 - 0.10 % of the control values (p < 0,0001). The relative insulin positive area was reduced to 63 and 7 % in respectively heterozygous and homozygous RIP-DTR mice.
Subsequently, 123I-L-phenylalanine and 111In-exendin-3 (-Lys40(111In-DTPA)-exendin-3) pancreatic uptake and biodistribution were evaluated using SPECT, autoradiography and an ex vivo biodistribution study in homozygous and heterozygous RIP-DTR mice.
Except from the kidneys, the uptake of 111In-Exendin-3 was highest in the pancreas and lungs. It was also significantly reduced in the pancreas of DT treated homozygous RIP-DTR mice.
Autoradiography showed only uptake of 111In-Exendin-3 within the pancreas in insulin expressing cells.
Quantification of 111In-Exendin-3 in the SPECT images based on the delineation of the pancreas with 123I-L-Phenylalanine showed a high correlation with the 111In-Exendin-3 uptake data of the pancreas obtained by dissection. The pancreas-to-blood ratios of the uptake of 111In-Exendin-3, measured in both ways, correlated highly with the relative insulin positive area.
These experiments show that 123I-L-phenylalanine is a promising tracer to delineate pancreatic tissue on SPECT images. It showed a high uptake in the pancreas as compared to other abdominal tissues. This study also demonstrates the feasibility and accuracy to measure the β-cell mass non-invasively in an animal model of diabetes.
Alternative Exendin tracers
In order to improve the image quality of 111In-exendin SPECT images we developed a strategy to increase the specific activity of 111In-exendin to administer higher activity doses. Exendin was modified by coupling of 5 additional Lysines (6Lys(DTPA)-exendin) to the already present C-terminal Lysine at position 40. These Lysines were then used to introduce multiple DTPA moieties to one exendin molecule. The reference compound ([Lys40(DTPA)]-exendin-3) was used as a control.
Biodistribution of 6Lys(DTPA(40+41))-exendin-3, 6Lys(DTPA(40-45))-exendin-3 and Lys(DTPA(40)-exendin-3 in BALB/c nude mice with subcutaneous INS-1 tumors showed high uptake in the INS-1 tumor, which could be blocked by an excess of unlabeled peptide. Accumulation in the pancreas was equal for all peptides. Blocking with an excess of unlabeled peptide showed receptor-mediated uptake in this organ. High kidney uptake was seen for all compounds, which could not be blocked. Uptake in other receptor-mediated organs (lung, stomach and duodenum) and non-target organs was similar for all three exendin-3 derivatives. Imaging studies in Brown Norway rats clearly showed higher accumulation of 6Lys(DTPA(40-45))-exendin-3 in the pancreas compared to Lys(DTPA(40)-exendin-3. In conclusion, the addition of 6 Lysines and multiple DTPA molecules had a positive influence on the in vivo imaging characteristics of exendin-3.
Comparison of 64Cu- and 68Ga-labelled NODAGA-exendin-4 for pancreatic beta-cell imaging in rats
Glucagon-like peptide 1 receptor (GLP-1R) is expressed in pancreatic beta cells and is considered a promising molecular target for imaging of intact pancreatic beta cells, transplanted islets and insulinomas. The aim of this study was to compare the labelling of native pancreatic islets by [64Cu]NODAGA-exendin-4 and [68Ga]NODAGA-exendin-4 in rats.
Biodistribution of tracers was assessed by ex vivo gamma counting and autoradiography. The whole-body distribution kinetics was studied by in vivo PET imaging. Stability, lipophilicity and affinity of the radiotracers for GLP-1R were tested and receptor specificity assessed using unlabelled exendin-3 as a competitor. Islet labelling was verified by combining autoradiography with immunohistochemistry. The radiation doses in humans were estimated based on the ex vivo biodistribution results of rats.
Both [64Cu]NODAGA-exendin-4 and [68Ga]NODAGA-exendin-4 specifically labelled the pancreatic islets. Uptake in pancreas was low with both tracers, [64Cu]NODAGA-exendin-4 having a higher islet-to-exocrine pancreas uptake ratio (106 ± 63) compared with [68Ga]NODAGA-exendin-4 (23 ± 3, P<0.01). Kidneys showed the highest uptake with both tracers followed by lung and stomach. Pancreas could not be visualised in the PET images. The IC50 values for [64Cu]- and [68Ga]NODAGA-exendin-4 were 2.7 ± 0.62 nM and 2.17 ± 0.42 nM, respectively. Autoradiographs of gut revealed high radioactivity uptake in proximal duodenum. Both of the tracers were stable in vitro. The estimated effective radiation dose for a 70-kg human adult was 0.074 mSv/MBq for [64Cu]NODAGA-exendin-4.
According to autoradiography, both copper-64 and gallium-68 labelled GLP-1R specific NODAGA-exendin-4 tracers labelled beta cells in rat pancreatic islets. Based on these results, [64Cu]NODAGA-exendin-4 is a better tracer for beta cell imaging. However, a high radiation burden to the kidneys limits the use of [64Cu]NODAGA-exendin-4 as a clinical tracer. (For additional details on dosimetry, please see section on clinical studies below)
In vivo evaluation of 18F-labelled silicon containing Exendin-4 peptide
Analogues of Exendin-4 have been radiolabelled for imaging the glucagon-like peptide type 1 receptors (GLP-1R) which are overexpressed in insulinoma. The aim of this research was to synthesize a 18F-labelled silicon containing Exendin-4 peptide ([18F]3) and to evaluate its in vivo behaviour in CHL-GLP-1 receptor positive tumour-bearing mice.
18F-labelled silicon containing Exendin-4 peptide ([18F]3) was prepared via one-step nucleophilic substitution of the silicon–hydride with 18F-fluoride. [18F]3 was then administered to tumor-bearing mice for PET imaging and ex vivo biodistribution experiments.
[18F]3 was produced with a radiochemical yield (decay corrected) of 1.5% and a specific radioactivity of 16 GBq/μmol. The tumours were clearly visualized by PET imaging. Biodistribution studies showed the high uptake of [18F]3 in the kidneys. The tumour uptake of the radiotracer was significantly specific and remained steady over 2 h.
Exendin-4 analogue [18F]3 can be successfully synthesized in a one-step procedure and used to image tumours in mice.
Longitudinal quantitative imaging of transplanted beta cells in a rat model with SPECT using In-111-DTPA Exendin-3
The feasibility of 111In-Exendin-3 as a BCM marker was evaluated in an intramuscular islet transplantation model in rats. We were able to demonstrate that the technique is highly sensitive, accurate and reproducible for the determination of the BCM in islet transplantation.
The radioactivity visualized by Microautoradiography (MAR) accumulated in the region of beta cells and corresponded closely with GLP-1 receptor staining as confirmed by immunohistochemistry performed on consecutive histological sections. Accumulation of radioactivity in the graft was quantified in vivo by SPECT and correlated linearly with the number of transplanted islets and qPCR.
In most of the animals followed for 6 weeks post transplantation, we have observed an initial increase in the accumulation of 111In-Exendin-3 in the grafts as determined by quantitative SPECT. The initial increase in tracer uptake is likely due to improving vascularisation of the engrafting islets, as it was reported that functional restoration of intra-islet blood supply takes place within first several weeks post transplantation. The subsequent decrease of the activity in some animals might be caused by ongoing apoptosis during the early post-transplantation period. Although in our experimental setting, the profile of radioactive uptake into the islet graft was highly reproducible during 14 weeks follow up and stable after 4th week, we realize that reproducibility of targeting a peptide receptor might depend on certain conditions (such as inflammation, glucotoxitiy, metabolic stress, rejection etc.), or the metabolic status of the individual islet graft which could influence the tracer uptake by changes in receptor expression and receptor metabolism. To address these questions, we will further characterize the method in ongoing experiments.
In conclusion, 111In-Exendin-3 can successfully be applied for non-invasive quantification of BCM in islet grafts in vivo. For clinical use, GLP-1R imaging has the advantage that no manipulation of the islets, such as virus infection, pre-labelling or genetic modifications is required prior to islet transplantation. Our method is highly sensitive, reproducible and robust. In addition, repeated imaging for longitudinal follow-up of an islet graft is possible. Therefore, our method holds great promise for in vivo determination of the BCM after islet transplantation in humans.
We evaluated a novel group of possible β-cell tracers, the nanobodies against two new targets.
We could only produce 4 nanobodies against target 1. They were evaluated with FACS on cells that express the target (Table 1) and with Biacore. None of the nanobodies had good affinities or even bound specifically to the target expressing cells. We therefore did not further evaluate these nanobodies.
Based on the FACS results (Table 1) on cells expressing human or mouse target 2, we found 2 very good binders to human target 2, 4hD-29 and 2hD-38 and 2 crossreactive nanobodies, 2hD-1 and 2hD-123. However, the affinities of these crossreactive nanobodies and 2hD-38 were rather low, so we decided to make some bivalent nanobodies of these constructs in the hope that they would show increased functional affinities.
The bivalent nanobodies indeed showed increased affinities for human target 2 (measured with Biacore) as compared to their monovalent ancestors. The most pronounced effect was seen with 2hD-123 and it’s bivalent form 123-123. The affinity of 123-123 was 8 times better than that of 2hD-123, mainly due to a better off-rate of the 123-123 nanobody.
Among other analyses, we tested these bivalent crossreactive nanobodies on 2 insulinoma cell lines, MIN6 and INS1E. The monovalent nanobodies almost didn’t bind to the MIN6, while some of the bivalent ones showed increased binding. This result was even more pronounced in the INS1E cell line. We did these analyses also for the other nanobodies, but they did not show any binding as expected.
We then used these 99Tc-labeled nanobodies in regular C57bl mice. Biodistribution studies showed a very high uptake in the kidneys and a very low uptake in the pancreas. The uptake of the monovalent nanobodies in the pancreas was not better as compared to a control nanobody. The uptake of the bivalent nanobodies in the pancreas showed a tendency to be higher as compared to the bivalent control nanobody, but only the 1-1 nanobody showed a significant higher uptake.
Autoradiography and insulin staining was performed on some consecutive cryosections of the pancreas. These sections showed a colocolisation of the radioactivity and insulin staining, which suggests a specificity of the nanobodies to the beta cells.
These bivalent nanobodies were then further evaluated in DT-treated and non-treated RIP-DTR mice. The DT-treated mice all were hyperglycemic. Due to some production problems, the 1-1 nanobody could not yet be evaluated. We also evaluated a new construct (123Cys-Cys123) that linked to 2hD-123 nanobodys with a Cys-tag together, rather than with a linker. There was no difference in any nanobody-uptake in the pancreas between the DT-treated mice and the control mice.
Because our best performing nanobody (4hD-29) only binds to human target 2, we also needed a human model for imaging and we explored the possibility of using tumor cell lines. We used the human neuroblastoma cell line Kelly, which also expresses target 2. Here, the bivalent nanobodies didn’t perform much better than their monovalent ancestors.This is probably because the monovalent nanobodies already showed quite good binding to the cells. We therefore continued with 4hD-29.
We used nude mice and implanted Kelly tumor cells subcutanously in the right leg. We then did imaging and dissection to measure the biodistribution. The uptake in the kidneys was rather high. This could later be reduced by some modifications to the nanobody. The uptake in most of the organs and tissues was higher for the control nanobody (BcII10) than for 4hD-29.This is due to different clearance rates of the 2 NBs. In the Kelly tumor, however, the uptake of our nanobody was higher than the control nanobody. Since the blood retention is significantly different between the 2 nanobodies, we looked at the tumor/blood ratio and this was a 4-fold higher for the 4hD-29 NB as compared to the control. You can also observe this in the SPECT/CT images that were made.
In conclusion, we generated and evaluated several Nanobodies against two different targets from WP1 and were able to use anti- target 2 Nanobodies to trace ß-cells. However, the lack of a suitable human model for evaluation forms a problem given the species-specificity of the Nanobodies; the latter could not entirely be resolved by the generation of bispecific Nanobodies.

There is urgent need of imaging methods for the in vivo characterization of beta cell mass and function, both for research on animal models (which implies high spatial resolution), and for clinical and therapeutic monitoring of patients (which implies safety and penetration depth). Magnetic Resonance Imaging (MRI), with the help of suitable tracers, could be useful in this context. We wished to 1) evaluate the potential of manganese-enhanced MRI (MEMRI) for the repeated and quantitative in situ imaging of the native endocrine pancreas of mice and humans, 2) develop novel probes to surface receptors, which could specifically enhance the MRI signal of beta cells.
For imaging, mouse and human pancreas were imaged by MR using 1.5-14.1 T equipments, after a short iv infusion of manganese. The MEMRI signal of islets (in rodents) and pancreas (in both rodents and humans) was then quantitatively compared under conditions of normo- and hyper-glycemia. In rodents, the evaluation of the MR signal was further compared to that provided by standard morphometry of histological sections of the very same pancreas. Using a 14.1T MRI equipment, we found that MEMRI visualizes individual islets, and ex vivo quantitates their numerical and volumetric differences in control and diabetic mice. However, the amplitude of the changes which were quantified by MRI was smaller than that of the metabolic alterations (blood glucose, insulin content) which was measured in the very same animals, after injection of a single diabetogenic dosis of streptozotocin. Using a 1.5T MRI equipment, we found that males and females mice, whose beta cells specifically expressed the receptor for diphtheria toxin, featured a similar pancreas MEMRI signal. In contrast, one week after injection of a cytotoxic dose of diphtheria toxin (DT), this signal was much more decreased in males than females Ins-DTr transgenic mice, consistent with the different loss in beta cell mass and insulin content, which was also higher in males than females mice. In a retrospective analysis of patients who underwent whole-body MEMRI for the diagnostic evaluation of a variety of clinical conditions, we found that the pancreas MEMRI signal, but not the signal of other abdominal organs, was reduced by 33% in type 2 diabetics, compared to normoglycemic patients, consistent with the difference in beta cells mass, which was anticipated form previous autopsy studies in Caucasian populations.
For developing probes, multiple copies of ligands targeting the GLP1 (exendin 4 derivatives) and/or the SUR1 receptors (glibenclamide derivatives) were linked to either PAMAM biodegradable polymers or superparamagnetic iron oxide nanoparticles. The probes were characterized by fluorescence on MIN6 cells and non insulin producing cell types, after labelling ligands, PAMAM, PEG and nanoparticles with different fluorochromes. Selected candidates were injected in control mice for imaging pancreatic islets by fluorescence and 1.5T MRI. To improve the beta cell specificity of the MRI signal, we synthesized probes targeted to surface receptors, by linking fluorochrome-tagged exendin and/or sulphonylurea derivatives to either rhodaminated PAMAM polymers, PEG polymers-tagged iron oxide nanoparticles. In vitro, naked PAMAM polymers, PEG polymers and nanoparticle did not specifically labelled MIN6 cells. In contrast, the ligand-carrying PAMAM, PEG and nanoparticle probes specifically labelled the membrane of MIN6 cells, but not that of the non insulin-producing HeLa and Panc1 cells, which were used as negative controls. After iv injection in normoglycemic mice, the probes also selectively labelled the membrane of most beta cells within control islets, as evaluated by ex vivo fluorescence and confocal microscopy. No obvious labelling was observed, within the very same islets, in cells which did not contain immunostainable insulin. Under 1.5T MRI, the same probes significantly reduced the pancreas signal, as expected from the presence of iron oxide cores, in the absence of detectable changes in the blood supply to the islets.
In summary, MEMRI quantitatively evaluates graded changes in islet mass/function in both rodents and humans, still undervalues these changes, presumably due to the non specific uptake of manganese by various cell types. Probes targeting GLP1 and SUR1 receptors specifically label beta cells, and significantly modify the MRI signal in rodent pancreas. The combination of MRI with functional (manganese uptake is enhanced in secreting cells) and beta cell-specific probes (targeted to abundant surface receptors) could provide a useful approach for the in vivo imaging of beta cell mass and function.

xf-FDOCT for 3D real-time in vivo imaging of pancreatic islets, blood flow, and beta-cell mass
In a first step, we developed a microscope platform to visualize islets of Langerhans. The following publications cover this part:
[1] M. Villiger, J. Goulley, M. Friedrich, A. Grapin-Botton, P. Meda, T. Lasser, and R. A. Leitgeb, “In vivo imaging of murine endocrine islets of Langerhans with extended-focus optical coherence microscopy.,” Diabetologia, vol. 52, no. 8, pp. 1599–1607, Aug. 2009.
[2] C. Berclaz, J. Goulley, M. Villiger, A. Bouwens, E. Martin-Williams, D. Van De Ville, A. C. Davison, A. Grapin-Botton, and T. Lasser, “Diabetes imaging — quantitative assessment of islets of Langerhans distribution in murine pancreas using extended-focus optical coherence microscopy,” Biomedical Optics Express, vol. 3, no. 6, pp. 1365–1380, 2012.
NOD-Mice could be imaged multiple times in longitudinal studies (up to 15 weeks) during diabetes progression. Semi-automatic recognition of islets extracted from tomograms has been developed as well as algorithms for the quantification of number, volume and contrast of islets. A statistical analysis was conducted to investigate the conclusions that might be extracted from different sampling densities. We could detect changes in islet number and contrast after streptozotocin treatment and in NOD mice prior to diabetes development (Goulley et al., in prep). xf-FDOCT was performed at 8 weeks and 15 weeks in the same laparotomized mice in NOD and control cohorts and diabetes progression was assessed in parallel in the same animals by glucose tests and glucose tolerance tests. In summary, we have established xf-FDOCT as a new and powerful optical imaging method to monitor islets and diabetes progression in mouse models (MOD mice).
xfOCM allows imaging of vascular networks through the use of previously developed scan protocol and phase variance algorithm (Srinivasan et al., 2009, Vakoc et al., 2009) . To quantify blood flow in the pancreas, we implemented the joint Spectral and Time domain method (jSTD) (Szkulmowski et al., 2008). jSTD provides a quantitative measurement of the spatial distribution of flow velocities using a Doppler frequency spectrum analysis. After a small incision of 0.5 -1 cm through the flank of the anaesthetized mouse, 3D imaging (> 30 min) of the pancreas of adult female ICR mice was performed. The duodenum encircling the pancreas was stabilized by a small pillar during image acquisition.
In a first series of experiments, we imaged the structure of an islet together with the vascular network around and inside an islet. The interferometric principle of xfOCM allows assessment of the phase variance between individual tomograms. This results in a well resolved vessel meshwork image without the use of exogenous labels. The high resolution of these OCM tomograms allows following individual vessels inside the islet.
In view of quantitative time resolved blood flow, we applied a jSTD analysis to the acquired tomogram stack. This analysis allows extracting the directional blood flow inside the pancreas.

Development of xfFDOCT into a platform for in vivo imaging of islets and tracer characterization
To develop xf-FDOCT as a platform to dynamically monitor tracer distribution within the pancreas of a living mouse, we investigated two ways of coupling specific β-cell tracers for islet and β-cell imaging:
Nanoparticles for optical and future MRI based β-cell imaging.
Nanoparticles (diameter < 6nm) may have interesting features while detecting the back-scatter light by poliOCM. Au-nanoparticles of 6-30 nm diameter enter cultured cells spontaneously and has been detected by using poli-OCM. This method is described in the following publication:
Pache et al., Fast three-dimensional imaging of gold nanoparticles in living cells with photothermal optical lock-in Optical Coherence Microscopy”, Opt Express, 20(19):21385-99 (2012).
Fluorescent β-cell tracer.
β-cell imaging based on specific β-cell tracer (exendin-3 linked to the fluorophore Cy5.5 (provided by RUNMC)) has been performed and fully characterized. To this end, we enhanced our current optical setups (dfOCM, xfOCM) by adding an additional confocal fluorescence channel. In a first attempt, tracer data have been obtained in vitro by using CHL cells transfected with the GLP-1R. In vivo characterization was performed by intravenous injection of 14ug, 1ug, and 0.1 ug of Cy5.5-exendin3 in ICR mice and imaged at different times post-injection. Blocking experiment confirmed the specificity of Cy5.5-exendin-3 for beta-cells.
As shown in C. Berclaz et al. OCM allows the localization and shape determination of islets down to a volume of 8000 µm3 (corresponding to 20 µm in diameter). The overlay between the OCM detection and fluorescence channel is ideally suited for detecting NIR fluorescent β-cell tracers inside the islets, as well as the accumulation of this tracer over time. As OCM and its intrinsic contrast is a label-free method, even small rodents can be easily imaged several times over an extended time period. In less than 5 sec a pancreas volume (800 x 800 x 400 µm3) containing several islets can be visualized. This underlines the ease and minimal invasiveness of OCM imaging and monitoring developed during this project. The distribution of the tracer inside the islets is complemented by the assessment of blood flow inside and around the islets which adds an additional information about the vitality of assessed islets
Overall, OCM and its high sensitivity paired with high spatial resolution over an extended depth proved to be ideally suited for tracer characterization in small animals. We have implemented OCM as a platform for simultaneous dynamic real-time imaging of tracer distribution and blood-flow, allowing easy characterization of tracer molecule kinetics linking this information to functional islet parameters.
Preparation of clinical studies
The preparation of the tracer for clinical studies includes:
1. GMP production of the tracer molecule (DTPA-coupled Lys40-Exendin 4)
2. Toxicity testing following EMA guidelines
3. Development and validation of a GMP labeling procedures including full documentation
4. Preparation of an IMPD (investigational medicinal product dossier) according to EMA guidelines and a dedicated production IMPD describing the production steps from the GMP tracer molecule (point 1) to the fully radiolabeled ready-for-injection final product
5. Preparation of an IB (investigator´s brochure)
6. Development and complete validation of a GMP kit formulation for radiolabeling and aliquotation
7. Preparation of ready-to-use kit formulation for radiolabeling
All these steps have been taken and we now have a full GMP kit preparation available according to EMA guidelines. Through approval of a qualified person, the kit can be distributed throughout Europe, either for on-site labeling or as a labeled ready to use product. The amount of tracer produced is sufficient for several thousand scans.
In addition, the IMPDs and the IB can easily be modified so that they form the basis for the PET tracers (except for the production IMPD that will undergo major changes due to the differences in labeling – however, the structure will still form the backbone of the new production IMPD).
For submission to the ethics committee, detailed clinical study protocols have to be prepared together with information material for patients and informed consent forms. In addition, the study has to be registered in the EMA system (EUDRA CT). All these documents have been prepared; for Dutch documents, English translations have been prepared with approval of the RUNMC ethics committee so that all documents are available in English.
The available documents have been completed in a manner that they can serve as basis for following studies so that a platform for studies with Exendin derivatives is available (for SPECT and PET).
The tracer for PET was developed to a kit formulation following the same steps as the SPECT tracer (for further information see last paragraph).

Development of scanning protocols for clinical imaging
Sophisticated protocols were developed based on studies with a phantom specifically designed for this purpose. Based on these results, acquisition and reconstruction standards have been developed optimized to the unusually high kidney-to-pancreas ratio allowing improvement of image quality for the clinical study.

Clinical studies
In WP3, 111In-labelled Exendin derivatives have been tested in animal models of diabetes in respect to their specificity and sensitivity. It could be shown that Exendin accumulates highly specifically in pancreatic islets with low background activity and that the uptake into the pancreas directly correlates with beta cell mass. Quantitative PCR demonstrated a more favourable ratio in GLP-1 receptor expression between islets and exocrine pancreas in humans as compared to rats. Therefore, our data clearly indicated that imaging with Exendin is a highly promising approach for the non-invasive determination of beta cell mass.
Therefore, clinical studies have been prepared and initiated using the GMP Exendin kit formulation for SPECT imaging. After preparation of the necessary documents and full validation of the kit formulation according to EMA guidelines and ethical approval, we have evaluated the uptake of the radiotracer into the pancreas in patients with long-standing type 1 diabetes as compared to healthy controls matching for BMI, age and gender.
Description of the study
Patients with type 1 diabetes (T1D), who met the inclusion criteria listed in the table above, were included in this study. Only T1D patients without insulin production as confirmed by undetectable C-peptide, were included. To measure non-stimulated C-peptide levels, a blood sample was taken after fasting for at least 4 hours. For stimulation of insulin production a bolus of 1 mg glucagon was administered intravenously, and six minutes after stimulation a blood sample was taken and C-peptide was measured. Both the non-stimulated and the stimulated C-peptide were below 0.03 nmol/l.
In order to create two comparable groups, healthy controls matched for BMI, age and sex were selected for participation. The BMI difference between a T1D patient and a matched healthy control was set to a maximum of 2 kg/m2, the difference in age to a maximum of 10 years and sex had to be identical.
Only healthy controls with HbA1C <7% and normal glucose tolerance were included, which was assessed by an oral glucose tolerance test. After fasting for at least 10 hours, subjects drank 75 gram of anhydrous glucose dissolved in 250 mL water within 5 minutes. The blood glucose 2 h after glucose intake, had to be below 7.8 mmol/l.
From all subjects written informed consent in accordance with provisions of the Declaration of Helsinki was obtained, and the study was approved by the Institutional Ethics Review Board of the Radboud University Nijmegen Medical Centre.
After at least 4 hours of fasting, 150 MBq 111In-exendin-4, corresponding with a peptide dose of 1 µg exendin, was injected as a slow bolus over 5 minutes. Blood glucose levels and blood pressure were closely monitored prior to and 5, 10, 15, 30, 60, 120, and 240 minutes after injection.
SPECT was acquired 24 h after 111In-exendin administration using an integrated SPECT/CT scanner (Symbia T16, Siemens Inc, Munich, Germany) equipped with parallel-hole medium-energy collimators. All subjects were scanned with a fixed radius of rotation (25 cm) and a fixed bed height (13 cm). 128 views were obtained in step-and-shoot mode (64 views per camera head) with an acquisition time of 40s per view.
Symmetric 15% windows over the 172 and the 247 keV photopeaks were used, with additional 7% lower and upper scatter windows for both photopeak windows.
CT images were acquired with the following settings: 110 keV, 45 mAs (using Siemens’ “care dose” technology for automatic dose reduction), 39 cm scanning length, with 5 mm slices and a pitch of 1.
SPECT reconstruction was performed with Flash3D software (Siemens Inc, Munich, Germany). Images were reconstructed with 6 iterations and 16 subsets and a matrix size of 128 × 128 (corresponding with an isotropic voxel size of 4.7952 mm). CT-based attenuation correction and triple energy window scatter correction was performed. Post-reconstruction filtering was applied with a 8.4 mm Gauss filter.
Image quantification was performed with Inveon Research Workplace software (Siemens Inc, Munich, Germany). On the SPECT images, volumes of interest (VOI) were drawn around the pancreas. The low dose CT was used for confirmation of the shape. A VOI around the kidneys was drawn with a cut-off value of 200 counts and increased by a rim of 3 voxels in width; this VOI was not taken into account for quantification of uptake even if pancreatic tissue was lying within this VOI in order to avoid spill-over effects from the kidneys that might influence quantification. As a result, in all patients a part of the pancreatic tail was excluded and in some patients a small part of the posterior rim of the head was excluded. Therefore, the uptake in the VOI as mean counts per voxel was corrected for the total pancreas volume based on delineation on CT images for reproducibility.
The uptake of the whole pancreas (U, in counts/ pancreas) was calculated with the counts (C) per mm3 within the SPECT VOI, multiplied by the volume of the whole pancreas (V, in mm3) as determined by CT and corrected for administered activity (AA, in MBq), the time after injection (Tscan-Tinjection) and the half-life of 111In (T1/2, 2.81 d).
To date, the data of 5 healthy volunteers as well as 5 diabetes patients have been fully analyzed. Analysis of the remaining participants of the study is still ongoing (quantification, pancreatic tracer distribution and dosimetry).
111In-exendin uptake is clearly visible in the human pancreas. The pancreatic uptake of 111In-exendin showed considerable inter-individual variation, in line with variations in beta cell mass BCM as reported in the literature (Ritzel et al., Diabetes care 29, 717, 2006; Rahier et al., Diabetes, Obesity & Metabolism 10 Suppl 4, 32, 2008)). The pancreatic uptake of 111In-exendin was clearly decreased in T1D patients with a maximum difference between a patient with T1D and a healthy volunteer of more than 10-fold.
To date, no other radiotracer tested for BCM determination has shown such large differences between healthy volunteers and T1D patients, nor has such a high inter-individual variability of the uptake been reported previously (Goland et al, J Nucl Med 50, 382, 2009). Of note, there was also a large variation in uptake in T1D patients, which is in line with autopsy data (Klinke et al; PloS one 3, e1374, 2008). On the other hand, all these patients have very low beta cell function, as indicated by stimulated C-peptide < 0.03 nmol/L. Previous observations in mouse models of autoimmune diabetes (Strandell et al; The Journal of Clinical Investigation 85, 1944, 1990) and in patients with T1D (Coppieters et al; The Journal of Experimental Medicine 209, 51, 2012) indicate that there remains an important population of functionally suppressed but still viable beta cells. This relevant concept seems to be supported by our results.
To the best of our knowledge, no comparable evidence has been created for any other radiotracer tested for in vivo imaging of beta cell mass with respect to beta cell-specificity and correlation of uptake to actual beta cell mass so far (preclinical experiments from WP03 in rodents and human tissue). Therefore, our data indicate that this technology could indeed be the first to enable non-invasive determination of BCM in humans, offering a great potential to better elucidate the pathophysiology of the different types and stages of diabetes.

Pancreatic tracer distribution
Distribution of the radiotracer in the pancreas was evaluated in the patients who’s results are shown above. The pancreas was delineated on SPECT/CT images to define a volume of interest (VOI), excluding the parts located too close to the high normal activity in the kidneys (i.e. pancreatic tail and in some subjects a small part of the head). The VOI was segmented in 4.8 mm slices, with the slice orientation depending on the shape of the pancreas. For all slices the 111In-Exendin uptake was determined and for each subject a profile of the uptake in the pancreas slices was made. All subjects showed a pronounced peak in the radiotracer uptake in the head of the pancreas. In most subjects, an additional peak in the body of the pancreas was identified in the uptake profile. Although T1D patients showed a lower pancreatic uptake in general, the shape of their uptake profile curve was similar to that of healthy controls. Our results show that both healthy volunteers and T1D patients have a distinct profile of 111In-exendin uptake in the pancreas,

Radiation exposure
The radiation exposure has been determined using OLINDA, a clinical dosimetry software package. The calculated whole body dose ranges from 8.76 – 12.3 mSv using 150MBq of the radiotracer. As the image quality is better than expected and also the pancreas is the only organ showing uptake apart from the kidneys, we expect that the injected activity can be reduced to 100MBq without impairment of image quality; this would result in a radiation exposure of 6-8mSv. A further reduction will be feasible with the 68Ga-labeled tracer.

Determination of optimal clinical tracer for PET
Dosimetry of [64Cu]- and [68Ga]NODAGA-exendin-4
Estimation of radiation doses for humans: Absorbed doses were calculated with the OLINDA/EXM 1.0 software which includes radionuclide information and a selection of human body phantoms. Rat ex vivo biodistribution results and PET imaging results were integrated as area under the time-activity curves. The obtained residence times were converted into corresponding human values by multiplication with organ-specific factors to scale organ and body weights: (WTB,rat/WOrgan,rat)×(WOrgan,human/WTB,human), where WTB,rat and WTB,human are the body weights of rat and human (70-kg male), respectively and WOrgan,rat and WOrgan,human are the organ weights of rat and human (70-kg male), respectively.
Extrapolated from the rat ex vivo results, the mean [64Cu]NODAGA-exendin-4 effective dose for a 70-kg human adult was 0.074 mSv/MBq. Estimated from the PET imaging results, the mean effective doses were 0.144 mSv/MBq for [64Cu]NODAGA-exendin-4 and 0.012 mSv/MBq for [68Ga]NODAGA-exendin-4. The kidney was the dose-limiting organ. Calculated from the ex vivo results, the absorbed kidney dose for [64Cu]NODAGA-exendin-4 was 1.48 mSv/MBq. Based on the PET results, the corresponding value for [64Cu]NODAGA-exendin-4 was 3.95 mSv/MBq and for [68Ga]NODAGA-exendin-4 0.523 mSv/MBq.
Based on these results, the decision was taken that for clinical studies, 68Ga should be preferred over 64Cu as the radiation exposure for 64Cu was considered too high for clinical studies with PET tracers.
Choice for the optimal tracer labeling for PET scanning
Since the specific activity of 68Ga-DOTA-exendin-3 (110 MBq/nmol) is not sufficient for clinical use, we evaluated the labelling of exendin compounds that could be labelled with 68Ga (NOTA-exendin, desferal-exendin and NODAGA-exendin). Of these compounds, NODAGA-exendin-4 has an extremely high specific activity: 1.2 GBq/nmol (at time of synthesis) and would be appropriate for clinical use. The in vivo binding characteristics of 68Ga-NODAGA-exendin-4 were compared to that of 68Ga-DOTA-exendin-3 in BALB/c nude mice with a subcutaneous INS-1 tumour. The tumour and pancreas uptake of 68Ga-NODAGA-exendin-4 was even higher than that of 68Ga-DOTA-exendin-3 and 68Ga-NODAGA-exendin-4 was selected as the tracer to be used for clinical application.

1.4 The potential impact
Socio-economic impact and the wider societal implications of the project
Contribution to Community and social objectives
Reinforcing competitiveness and solving societal problems
The successful development of imaging technologies for in vivo monitoring of beta-cell mass as achieved in BetaImage is of utmost strategic importance. The incidence of diabetes is on the rise in Europe (and worldwide), therefore the EU must face this problem. The kind of innovative research conducted in the BetaImage consortium has been actively stimulated and funded in the USA for more than a decade and it is only in the past 4-5 years that equivalent research efforts are developed in Europe. BetaImage research has helped and in the future will continue will help to answer fundamental questions about the pathophysiology of regulation of beta-cell mass in diabetes mellitus eventually leading to improved treatment and new strategies to avoid the occurrence of the disease.
European researchers have historically been strong in the development of specific ligands for molecular imaging. Partners of BetaImage had been involved in the development of innovative PET/SPECT tracers and new labelling strategies before the project. They had developed new imaging technology and improved existing technologies to open new fields for molecular imaging research. BetaImage partners had implemented databases of beta-cell specific antigens and established cell and animal models for diabetes research.
Within the project, BetaImage researchers have built upon their respective strength, expertise and experience. The highly sophisticated strategies developed in BetaImage for in vivo and in vitro beta cell imaging will critically support the European position in a very competitive field, where transformation of frontier scientific discoveries into exploitable products may have a profound impact on public health and European economy. By implementing an elaborated a work program comprising 4 highly complementary scientific work packages allowed to consolidate collaborations based on previous European research framework programmes and extended these fruitful collaborations to novel partners bringing into the field colleagues with relevant expertise in diabetes/beta-cell physiology and molecular imaging. This work program has finally led to the highly impressive scientific achievements described in this report.
Improving health of the European citizens, including improvement of health of the ageing population and children
BetaImage achievements will help to solve a major societal problem, which is the high (and still increasing) level of incidence of diabetes in the EU. Type 2 diabetes is a key health problem associated with an ageing population as well as with life style risk factors. Determining the beta-cell mass and improving diabetes therapy by beta-cell monitoring might help to both enrich the quality of life of patients and to reduce the costs of this disease to the European community. Type 2 diabetes is an age-related disease with an immense impact on the quality of life of older people. With an ageing population, this impact will grow even further. BetaImage has laid the basis to improve the diagnostic and therapeutic options for patients, and BetaImage research may have a strong impact on improving the health of the ageing population. BetaImage results are expected to not only help improve therapy, but they may also help to prevent the disease by identifying sub-populations at risk of developing diabetes so that specific prophylactic measures can be taken. Type 1 diabetes is the most common metabolic disease in the young and the most common chronic disease in children, with an increasing incidence. Imaging technologies developed and validated in BetaImage will help in the early diagnosis T1D and support research. We expect that this will improve the possibilities of tackling prevention of the disease. This will have a significant impact upon the general health status of European children and adolescents and improve their situation with respect to the psychosocial consequences of T1D (diet, multiple daily injections, blood glucose monitoring etc.) but also help to avoid long-term complications (including blindness, renal failure, limb amputation, cardiovascular disease, premature mortality etc.). BetaImage achievements will therefore also help to reduce the economic burden from treatment of the disease and associated complications, amounting to circa 75 billion Euros per year in the European community.
Benefiting from the European approach
The successful development of imaging strategies for in vivo beta-cell imaging, especially with respect to detection of small relative changes in beta-cell mass, is a formidable challenge. This challenge has been achieved by bringing together leading European centres of expertise in the fields of molecular imaging and diabetes research. The groups participating in BetaImage have provided all technical and scientific skills required for the innovative advances achieved in the field of beta-cell (mass) imaging. Working in this cooperative approach with a clear translational focus has enabled European scientists and biotechnology industry to achieve a strong position in beta-cell imaging based on results achieved in BetaImage.
The project has provided an excellent opportunity for training of young scientists in the fields of molecular imaging and diabetes
The project has also provided an excellent framework for educating the next generation of young scientists in the field of experimental and clinical imaging and diabetes research because of its multidisciplinary nature and the quality of the work proposed. It has facilitated exchange of personnel within European countries and fostered contacts between different laboratories at all levels. Since several of the senior investigators have been women, including half of the WP leaders, they have provided role models and support for younger women scientists. In doing so, BetaImage has also helped to fulfil the objectives of The International Dimension of the European Research Area as defined by the European Commission in COM 2001 346.
Innovation related activities
BetaImage has achieved highly innovative results through setting-up a general technological and methodological strategy based on leading state-of-the-art European expertise in diabetes/beta-cell physiology and molecular imaging. The key innovation-related activities achieved in BetaImage include:
• The utilization of array analysis to determine the global pattern of genes and proteins expressed in beta-cells, exposed in vitro to cytokine- or gluco/lipotoxic-induced damage, or to the diabetogenic environment in vivo.
• The utilization of a set of carefully tuned in vitro and in vivo models of diabetes to elucidate the potential of beta-cell specific targets for imaging beta-cell mass and neogenesis.
• Development of PET and SPECT tracers for in vivo GLP-1 receptor imaging using Exendin-analogues under translational conditions, ready for clinical use or under clinical evaluation within the duration of the project.
• Development of a novel class of clinical tracers based on nanobody molecules, combining the advantages of peptides and antibodies in one tracer molecule while largely avoiding their specific disadvantages. These tracers have demonstrated their ability to specifically target a large variety of beta-cell surface proteins and other targets with high affinity.
• The BetaImage consortium has developed a cutting edge technology for in vivo optical imaging of islets in the pancreas, combined with islet-individual Doppler blood flow measurement with microscopic resolution (OCM).
• A beta-cell imaging technology complementary to PET/SPECT/MRI has been developed with microscopic resolution for real-time in vivo dynamic imaging of islets based on xf-FDOCT in combination with fluorescence tracers and tracers for optical contrast imaging labeled with metals. This technology is extremely useful for research purposes as well as for optimisation of the tracer development process. Clinical implementation is anticipated for guided surgery in congenital hyperinsulinism.
• MRI tracers have been developed based on optimised nanoparticles for specific targeting of beta-cells and beta-cell related endothelial targets. In addition, ex vivo imaging of pancreatic islets with contrast enhanced MRI has for the first time been achieved.
• A critical part of the overall strategy of BetaImage, which could only be performed in this European collaboration, is the constant interaction and close cooperation between scientists from all participating partners working with in vitro and in vivo models of diabetes on one side and those working in the fields of tracer development and technology development on the other side.

Scientific, economical and societal impact achieved through the efforts of the BetaImage consortium
Expected impacts listed in the work programme
(page numbers refer to FP7 Cooperation Work Programme: Health, published by the EC) Contribution of BETA IMAGE towards the expected impacts (Expected) Impact of BETA IMAGE
Overarching issues of strategic importance, p.8; Health of the aging population: Whenever appropriate, the projects funded under this Theme should take into consideration the research aspects related to prevention, diagnostics and treatment of age-related diseases and the impact on quality of life of older people. Type 2 diabetes is a typical age-related disease with increasing incidence and has an immense impact on the quality of life of older people. With an ageing population, this impact will grow even further. The results of the clinical studies carried out in BetaImage are expected to have a strong impact on improving the health and quality of life of the ageing population by helping to elucidate the pathophysiology mechanisms leading to diabetes, and thus contributing for therapy. In the near future, BetaImage achievements may therefore not only help to improve therapy, but may also help to prevent the disease by identifying sub-populations at risk of developing diabetes so that specific prophylactic measures can be taken to protect these sub-populations. High
Overarching issues of strategic importance, p.8; Gender aspects in research: Gender aspects in research have a particular relevance to this Theme as risk factors, biological mechanisms, causes, clinical manifestation, consequences and treatment of disease and disorders often differ between men and women. The possibility of gender/sex differences must therefore be considered in all areas of health research where appropriate. There is a slightly higher incidence of diabetes in women, which can partly be attributed to a higher average life span. Though BetaImage achievements will have an impact on diabetes patients in general and there will be more women profiting, the overall impact on Gender issues is low. Low
Overarching issues of strategic importance, p.8; SME relevant research: This Theme emphasises the importance of innovation and the integration of SMEs in health research projects in order to reach the Lisbon goal by ensuring that new knowledge is disseminated and translated into new therapies and clinical practice. Innovative and flexible SMEs continue to play a major role in the development of molecular imaging strategies. A vast part of innovations in molecular imaging in the past two decades has been developed in or with SMEs and later adopted or bought by large worldwide operating industrial companies such as Siemens, Philipps, General Electric etc.. BetaImage research is typical for development of highly specialised molecular imaging technologies that may later be used in a broader context. Spin-offs from partner VUB, UNIMAR and RUNMC will largely become involved in and profit from the achievements of BetaImage, namely nanobodies serving as novel beta cell-specific tracer molecules and Exendin-derivatives for clinical radiotracer imaging.
Overarching issues of strategic importance, p.11; Clinical research and clinical trials: As translation of basic discoveries into clinical applications is one of the main objectives of this theme, clinical research is expected to be a major tool used in the funded projects. The main focus of BetaImage has been the translation of novel, promising strategies for beta-cell imaging from pre-clinical development to the use in humans. This strategy was based on a thorough analysis of bottlenecks and previous failures in this process. To be successful, BetaImage researchers have collaborated in highly interlinked preclinical as well as clinical WPs, following a straight-forward step-by-step approach leading to rapid clinical assessment of suitable tracers. This has indeed led to successful clinical translation of BetaImage research with 2 novel radiotracers already in or ready for clinical use. These results can easily be translated to other diseases, namely beta cell-derived tumours but also other diseases. High
2. TRANSLATING RESEARCH FOR HUMAN HEALTH, p.18: This activity aims at increasing knowledge of biological processes and mechanisms involved in normal health and in specific disease situations, to transpose this knowledge into clinical applications including disease control and treatment, and to ensure that clinical (including epidemiological) data guide further research. To date, diabetes research is hampered by the fact that beta-cell mass cannot be determined non-invasively in vivo. Beta-cell mass, its loss and – using new therapeutic regimens – its restoration through stimulation of neogenesis are processes that are not completely understood. BetaImage results will help to better understand the underlying pathophysiological processes and mechanisms thus leading to improvement of therapy and prophylaxis of disease in the near future. High
2.4 TRANSLATIONAL RESEARCH IN OTHER MAJOR DISEASES; 2.4.3. Diabetes and obesity, p 38: Research will address breakthroughs in the diabetes/obesity treatments but also in prevention and treatment of complications. Considering the heavy toll taken on life expectancy of these diseases, particular attention should be given to paediatric aspects, whenever possible. Healthy life styles being a pre-requisite to any stabilisation of escalating costs of diabetes/obesity, projects should also examine how their results will contribute to the societal issues linked to the diseases. BetaImage strived to produce results that will help solve a major societal problem, which is the high (and still increasing) level of incidence of diabetes in the EU (and worldwide). Diabetes is a key health problem associated with an ageing population as well as with life style risk factors. Assessment and monitoring of the beta-cell mass with BetaImage imaging technologies in different subpopulations of diabetic patients as well as in animal models of diabetes will help to further improve our understanding of the underlying pathophysiological factors leading to beta-cell loss and decreased glucose tolerance. BetaImage results are therefore expected to lead to a major break-through in diabetes treatment and prevention, as they will help to enable / improve individually-tailored therapy. Furthermore, prevention of disease will be improved by identifying persons at risk for diabetes and development of individual prevention strategies based on the individual pathophysiology. These measures will help to both, enrich the quality of life of patients and to reduce the costs of this disease to the European community. With an ageing population, this impact will grow even further. By helping to improve the diagnostic and therapeutic options for patients, BetaImage results is expected to have a strong impact on improving the health of the European population High

Main dissemination activities and exploitation of results
The envisaged clinical /translational end points of BetaImage were to deliver clinical PET/SPECT tracers as well as to define at least 3-5 tracers with clinical potential for beta-cell imaging (based on new targets defined in the project). The envisaged research/pre-clinical end points were to develop optical imaging strategies for in vivo real-time dynamic imaging of islets and optical tracers with microscopic resolution and specifically targeting nanoparticles for small animal MRI. Furthermore, a new methodology for streamlined high-throughput tracer development was to be developed. It was expected that the results of BetaImage would not only provide the basis for new results in diabetes research improving diagnosis and therapy but also enhance European excellence in the field of beta-cell imaging and tracer development.
Promising results of clinical tracer development (pre-clinical results) and first clinical studies in man in BetaImage appear to have great promise for clinical beta-cell imaging and have been used to actively find industrial partners. As a consequence, European health care companies have shown interest in this research. A pre-clinical study in a rat diabetes model has been carried out together with Sanofi-Aventis and currently, this company as well as Boehringer Ingelheim show interest in conducting clinical studies with the human use SPECT/PET tracer developed in BetaImage.
The ultimate aim remains to be exploitation of patents from the consortium and to obtain financial support for studies aiming at clinical approval of tracers.
All new knowledge and information gathered through the scientific research work in BetaImage has been disseminated to target groups consisting of all parties involved in diabetology and molecular imaging, namely scientists, physicians, nurses, hospitals, medical authorities and the industrial companies of the sector. Providing they are not confidential, results of the studies have been published or publication is under way. Based on the number of publications from this project so far (close to 70), we expect that the final number of publications will reach 90-100 in the coming 2 years.
In addition to scientific publications, the results of BetaImage have been diffused through book chapters, communications at national and international scientific meetings, and presentations for the public in conjunction with diabetes patient associations (JDRF).
A focus of dissemination of knowledge and expertise of BetaImage has been dissemination via workshops and training within and outside of the project in order to increase the number of European scientist capable of utilizing beta-cell imaging strategies and techniques. In addition, 3 international workshops have been held with international experts and participants to make the results of BetaImage as well as the state-of-the-art of beta-cell imaging known to an international scientific community in a condensed form. The topics of these workshops have been covered by a special issue of “Current Pharmaceutical Design” with the coordinator of BetaImage as guest editor and a review article in Diabetologia.
BetaImage members have set up an EASD study group “Biomedical Imaging in Diabetes” in order to further disseminate the results obtained in BetaImage and to create a stable platform for future European research in the field of beta cell imaging.
Furthermore, free information has been provided to the scientific community on all genes expressed in human and rodent islet cells, under both physiological and pathological conditions, via the Beta-cell Gene Expression Bank, an open resource developed by ULB (
Through the BetaImage website, interactive exchange of data within the network (restricted access) and presentation of the general policy and results (open access) have been provided.

Outlook and future research
The achievements of BetaImage with the highest impact on future research are:
1. The first clinical tracer in man appearing to measure pancreatic beta cell mass non-invasively
2. Establishment of the most complete data set of beta cell specific proteome, including many new splice variants
3. A new class of ligands for imaging of beta cell-specific targets offering high flexibility: nanobodies
4. Establishment of a platform for simultaneous dynamic real time in vivo imaging of pancreatic islets, blood flow and tracer kinetics based on xfFDOCT.

1. Clinical tracer for beta cell imaging
The GLP-1 receptor targeting Exendin analogues developed in this project display highly promising characteristics for in vivo non invasive PET and SPECT imaging in animal models as well as in man. Two of these tracer molecules are available now as kit formulation for human use and can be delivered to European centres under full GMP conditions for clinical use (after approval of the qualified person according to EMA standards).
As the availability of GMP tracer for clinical use is the bottleneck for conducting clinical studies, the kits enable us to perform clinical studies with a limited amount of funding and the compound has been produced for several thousand injections. Therefore, the BetaImage consortium will set up clinical studies with these compounds in diabetes patients to further elucidate the potential of these highly promising tracers.
In addition, within the BetaTrain project also coordinated by the coordinator of BetaImage, this technology will be further evaluated and cross-validated with other imaging technologies developed in other EU FP7 projects: BetaTrain unites scientists from BetaImage, IMIDIA, VIBRANT, ENCITE and MADEIRA. Our joint effort is not only training of young scientist in the fields of diabetes and biomedical imaging, but also to cross-validate currently available imaging technology in the field of beta cell imaging (pre-clinically as well as clinically) aiming at creating a map of imaging techniques that can be used to complement each other in diabetes research (more information can be found on the website: ). BetaImage achievements in the fields of PET/SPECT, MRI and optical imaging form a pivotal contribution to the objectives of BetaTrain.
Characterization of Exendin derivatives for in vivo imaging is also continuing in a joint project between partners VUB and RUNMC financially supported by JDRF. BetaImage partners will continue to strive for grant-funded joint scientific projects based on BetaImage achievements.

2. Establishment of the most complete data set of beta cell specific proteome, including many new splice variants
The database of beta cell specific targets will enable the scientific community to carry on with the identification of suitable target/ligand combinations for beta cell imaging, may it be beta cell mass, beta cell function, inflammation etc..

3. A new class of ligands for imaging of beta cell-specific targets offering high flexibility: nanobodies
Nanobodies have proven to be of the highest value as carrier molecules for radioactive or optical tags, enabling in vivo and in vitro characterization of tissue. Although nanobodies have mostly been used for targeting of tumour tissue, BetaImage achievements have opened a window of opportunity for the use of these compounds in beta cell imaging. Nanobodies with their unsurpassed flexibility in respect to targeting of different cell-specific molecules in combination with their optimal size for use as labelled tracer molecule are expected to enable targeting of alternative beta cell-specific molecules for measuring of mass, function, metabolic state etc.. This work will be continued und the lead of partner VUB.

4. Establishment of a platform for simultaneous dynamic real time in vivo imaging of pancreatic islets, blood flow and tracer kinetics based on xfFDOCT
The xfFDOCT platform allows the simultaneous real time evaluation of blood flow and tracer kinetics together with anatomical information on a close-to-cellular level. Therefore, the platform can not only be used for the development of beta cell tracers and their evaluation in the pancreas of living animals, but it also offers multiple possibilities for evaluation of tracer behaviour and blood flow in diseased tissue, namely tumour tissue or inflamed tissues. Research in these fields will be continued in addition to beta cell imaging.

List of Websites:

Related documents