Abstract/Video opens by clicking at the talk title.
Abstract/Video opens by clicking at the talk title.
Single chain-variable fragments and humanized antibody fragments for targeting malondialdehyde-low density lipoprotein in atherosclerosis (#298)
Samata S. Pandey1, 2, Michael Mullin3, Alan Lewis4, Yasmin Morris1, Cleo Kontoravdi2, Dorian Haskard1, Ramzi Khamis1
1 Imperial College London, Vascular Sciences, London, United Kingdom
Despite the significant socioeconomic impact of atherosclerosis, its clinical characterization is challenging often relying on the measure of the degree of stenosis (1). Oxidation specific-epitopes (OSE) on lipoproteins have been identified as potential key features of rupture-prone atherosclerotic plaques (2). Here, we present the development and characterization of single-chain variable fragments (scFv) and antigen-binding fragments (Fab) for the non-invasive imaging of oxLDL in atherosclerotic lesions.
The murine monoclonal antibody LO1 was identified and characterized for binding against malondialdehyde low-density lipoprotein (MDA-LDL). An array of single-chain variable fragments (LO1-scFv) and humanized antibody fragments (LO1-Fab) were recombinantly expressed in mammalian cells with a cysteine tag allowing for site-specific conjugation with a fluorophore. In vitro characterization was performed through ELISA and immunohistochemistry for binding to antigen. Pharmacokinetics, biodistribution, and plaque specificity studies were performed in LDLR-/- mice labeled with VivoTag-750-MAL using fluorescence molecular tomography (FMT). Ex-vivo imaging was performed on whole tissue using Odyssey LiCor.
LO1-Fab and scFv bound specifically to MDA-LDL. Mutations to lysines in complementarity-determining region-H3 abrogated binding to antigen, indicating key residues for binding. ELISA revealed competition between full mAb and huFab for epitope binding. Site-specific conjugation resulted in only slightly reduced binding of labeled Fab to MDA-LDL. The half-life of LO1-scFv demonstrated a significant reduction in circulation compared to full Fab (12.7 minutes to 21 hours). Immunohistochemistry of aortic arch lesions demonstrated punctuate staining within the necrotic core, with an absence of staining with Control-Fab. In vivo FMT imaging revealed localization to the aortic arch and clearance of fragments allowing only imaging four-hours after injection due to a faster clearance time and improvement in signal-to-noise ratio.
We demonstrated that LO1-scFv and humanized Fab bound MDA-LDL in vitro and within an experimental mouse model of atherosclerosis. These antibody fragments have allowed the non-invasive imaging of atherosclerosis with the potential for their further use for therapeutic targeting of oxidized LDL.
 Dweck M, et al. Circulation Research. 2016;119:330–340
 Falk E, et al. Eur Heart J 2013;34:719–728
Keywords: near-infrared imaging, molecular imaging, atherosclerosis, oxidized low-density lipoprotein
Clinical Imaging of Carotid Atherosclerosis Using Multi-Spectral Optoacoustic Tomography (#435)
Angelos Karlas1, 3, 4, Michael Kallmayer3, Michael Bariotakis1, 2, Nikolina-Alexia Fasoula1, 2, Evangelos Liapis1, 2, Nikolaos Kosmas Chlis1, 2, Moritz Wildgruber5, Hans-Henning Eckstein3, 4, Vasilis Ntziachristos1, 2, 4
1 Helmholtz Zentrum München, Institute of Biological and Medical Imaging, Neuherberg, Germany
Imaging plays a key role in the early diagnosis and management of carotid atherosclerosis. Multi-Spectral Optoacoustic Tomography (MSOT) is a hybrid imaging technique that provides molecular tissue characterizations by sensing ultrasound waves generated upon tissue illumination by ultrafast laser pulses. MSOT can resolve hemoglobin and lipids, two of the main constituents of the atherosclerotic plaque, demonstrating great potential in the non-invasive and label-free detection and characterization of carotid atherosclerosis in the clinical setting.
In this pilot human study, we included 5 patientswith diagnosed carotid atherosclerosis who underwent carotid endarterectomy (CEA) and 5 healthy volunteers. All participants signed an informed consent and were scanned with a clinical MSOT system with incorporated ultrasound (US) imaging. The carotid artery was scanned in the cervical region on both sides with 28 different wavelengths at the near-infrared range (NIR, 700-970nm at steps of 10nm). The carotid artery cross-section, the plaque area and the lumen area were manually segmented in the co-registered US images. For each segmented region the total hemoglobin (THb) and the lipid content were calculated based on the recorded MSOT images. All excised plaques were histologically analyzed.
MSOT achieved a preliminary characterization of the carotid plaque in vivo by measuring the signal corresponding to the total hemoglobin (oxygenated and deoxygenated) and lipids. Based purely on the MSOT images we were able to differentiate between healthy volunteers and patients with previously diagnosed carotid atherosclerosis by calculating the THb- and Lipid-content of the arterial cross-section. Patients showed significantly higher difference between Lipid- and THb-content in the whole cross-section compared to healthy volunteers (p < 0.01). Furthermore, our approach demonstrated statistically significant difference (p < 0.01) of the Lipid-THb signal between the plaque (high) and the lumen (low) area based solely on MSOT-extracted parameters. Thus, in contrast to the purely morphological signal of traditional US, MSOT demonstrated strong potential on extracting enhanced molecular information about the atherosclerotic plaque of the carotid artery.
Clinical MSOT imaging was capable of detecting and characterizing human carotid atherosclerosis in vivo with high optical contrast and acoustic resolution without the need for contrast agents. Further investigation and identification of correlations between MSOT, ultrasound and histology would enable the future stand-alone use of MSOT as a novel tool for molecular imaging of carotid atherosclerosis in our pursuit of the vulnerable plaque.
Keywords: carotid atherosclerosis, optoacoustic imaging, photoacoustics, cardiovascular disease
Helper T Cells Infiltrating The Aortic Media Are Associated With Aortic Dissection In Patients With A Bicuspid Aortic Valve (#220)
Alexander H. J. Staal1, Kimberly R. Cortenbach1, Mark A. J. Gorris1, Lambert J. Wisse3, Johannes Textor1, Guillaume S. C. Geuzebroek2, Mangala Srinivas1, Wim J. Morshuis2, I Jolanda M. de Vries2, Marco C. de Ruiter3, Roland R. J. Kimmenade4
1 Radboudumc, Tumor Immunology Lab, RIMLS, Nijmegen, Netherlands
Bicuspid aortic valve (BAV) is the most common congenital heart defect and is associated with ascending aorta aneurysms and dissections1. Presently, genetic factors2 and pathological flow patterns3 are considered responsible for aneurysm formation in BAV. Here we investigate whether the immune response is involved in aneurysm formation and aortic dissection4.
Dilated (n=8), non-dilated (n=14) and dissected (n=4) BAV ascending aortas were collected during surgery or from post-mortem donors. Tissue was stained with a novel immunohistochemistry technique allowing for simultaneous visualization of 8 markers on one slide. T cells, B cells and classic dendritic cells type 2 (cDC2s) were identified and comprehensively phenotyped using automated quantitative analysis.
Aneurysm formation was associated with an organized increase of lymphocytes in the adventitia. B cell follicles and helper T cell expansion were identified, suggestive of a targeted immune response (Fig. 1a,b). Dissected aortas were characterized by a statistically significant increase of helper T and cDC2s in the media, when compared to non-dilated and dilated samples (Fig. 1c,d). In the past, chronic inflammation of the dilatating ascending aorta of BAV patients has been considered absent, irrelevant or reactive to the underlying disease. With a novel sophisticated technique we show that the immune response is aimed and organized by the adaptive immune system. This could imply a more important role of the immune system than originally considered. Additional mechanistic study into this hypothesis in animal models is needed to further elucidate the importance if this finding.
Aorta dilatation in patients with BAV is associated with an expansion of B and helper T cells in the adventitial compartment without changes in the media. Only dissections show an increase in helper T cells and cDC2s in the media. We hypothesize that helper T cells potentiate macrophages and are associated with the final step in the deterioration of the aorta towards a dissection.
AcknowledgmentThis work was funded by an ERC starting grant (ERC-2014-StG-336454-CoNQUeST), TTW-NWO open technology grant (STW-14716).
 Mathieu, P, et al., 2015 'The pathology and pathobiology of bicuspid aortic valve: State of the art and novel research perspectives', J Path: Clinc Res, 1 195-206.
 Andreassi, MG, et al., 2016, 'Genetics of bicuspid aortic valve aortopathy', Curr Opin Cardiol', 6, 585-592.
 Rodríguez-Palomares, JF, et al., 2018 'Aortic flow patterns and wall shear stress maps by 4D-flow cardiovascular magnetic resonance in the assessment of aortic dilatation in bicuspid aortic valve disease', J CarVasc Magn Res, 20, 28.
 Tobin, SW,. et al., 2018, 'Novel mediators of aneurysm progression in bicuspid aortic valve disease', J Mol Cell Card, 132, 71-83.
Multiplex immunohistochemistry and quantification of BAV aorta samples.
a Representative dilated ascending aorta of a patient with BAV. Adventitia shows organized activation of the immune system. Magnification shows the presence of B cell follicles (green) and the close interaction between helper T cells (red), B cells (green) and cDC2s (Magenta). b Quantification. c Representative high magnification images of the aortic media of non-dilated, dilated and dissected aortas of patients with BAV. Note the clear increase in helper T cells (red) and cDC2s (magenta) in the dissected aorta compared to non-dilated and dilated. d Quantification. (mean and SD, *p < 0.05)
Keywords: Bicuspid aortic valve, Ascending aorta aneurysms, Aortic dissection
Longitudinal In vivo micro-computed tomography monitors lung pathology development and potential treatment effects in Down syndrome mouse model (#479)
Birger Tielemans1, Sergi Llambrich Ferre1, Maxine De Corswarem1, Neus Martinez-Abadias2, Greetje Vande Velde1
1 KU Leuven, Department of Imaging and Pathology, Biomedical MRI unit/MoSAIC,, Leuven, Belgium
Although craniofacial malformations and cognitive deprivation are most prominent, patients with Down syndrome (DS) suffer from morbidity mainly due to cardiopulmonary problems1. Animal models of DS are crucial to study pulmonary pathogenesis but it is unknown if they recapitulate pulmonary pathology. We aim to characterize development of lung pathology in a mouse model of DS using longitudinal µCT from birth to adulthood and to assess potential modulatory effects of EGCG.
Ts65Dn mice and wild-type (WT) mice were left untreated or administered EGCG (epigallocatechin-3-gallate) starting prenatally (E9) via the drinking water. Mice were scanned under isoflurane anesthesia with whole body, in vivo µCT (Skyscan 1278, Bruker) on postnatal day (PD) 3, 14 and 29. Scan time was 3 min, with 50 µm reconstructed voxel size. We extract total, aerated- and non-aerated lung volumes, and mean lung density for these volumes.
Micro-CT can detect hyperdense areas in the lungs of Ts65Dn mice at PD3 (Figure 1). EGCG administration resulted in increased aerated lung and trachea volumes and in decreased aerated lung density. This effect is more prominent on WT compared to Ts65Dn mice at PD29.
We show that it is feasible to extract structural and functional lung biomarkers in a longitudinal manner starting at an age of 3 days. This technique has great potential in phenotyping transgenic mouse models and in other developmental research.
 Englund, A., et al. American Journal of Medical Genetics Part A 161, 642-9 (2013)
Keywords: Longitudinal In vivo micro-computed tomography, Respiratory system, Down syndrome, EGCG modulation
Site-specific bioconjugation of the hexahistidine tag (#592)
Terence Tshibangu1, Machteld Sillen2, Frederik Cleeren1, Nick Geukens2, Paul Declerck2, Guy Bormans1
1 University of Leuven, Laboratory for Radiopharmaceutical Research, Leuven, Belgium
Despite its great potential for site-specific bioconjugation (1,2), the functionalization of His-tag is barely explored by radiopharmaceutical scientists (3). To fill this gap, we report the characterization of an scFv fragment targeting plasminogen activator inhibitor-1, resulting from the specific alkylation on two imidazole rings of the poly-histidine tag by a bifunctional PEG-monosulfone (Figure 1). Owing to its high potential, the bioconjugation strategy on a His-tag could become a standard methodology for the radiolabeling of recombinant proteins.
Based on a literature procedure (1,2), hexahistidine peptide was used to explore the bioconjugation strategy. LC-MS allowed the evaluation of the bioconjugation procedure regarding the peptide-monosulfone ratio, temperature, pH, organic co-solvent and duration of the reaction. Stability of sulfo-Cy5-hexahistidine was assessed by LC-MS, after storage of the bioconjugate at 4°C for 5 months in solutions at pH5 to 7 as well as at room temperature for 10 days at pH 6.6. An scFv served as an example for complex proteins. The bioconjugation on scFv was monitored by LC-MS as well as by size-exclusion chromatography.
LC-MS monitoring of the reaction showed a fast bioconjugation of hexahistidine peptide, both at 4°C, room temperature and 40°C (70% conversion after 30 minutes at room temperature). The efficiency of the bioconjugation was highest at pH 7. However, after 17 hours and independently of the pH of the reaction mixture, the entire monosulfone reagent reacted with the hexahistidine peptide. A ratio 2:1 molar equivalent of hexahistidine peptide: PEG-monosulfone reagent results in exclusively mono alkylated hexahistidine. In addition, the bioconjugate is stable in solution, as the LC-MS analysis did not detect any release of the sulfo-Cy5 derivative after storage at 4 ° C for 5 months or at room temperature for 10 days. The work with the hexahistidine peptide served as a launching pad for the site-specific bioconjugation of recombinant proteins. LC-MS and size exclusion chromatography confirmed that bioconjugation of a hexahistidine-scFv fragment was also successful.
PEG-monosulfone reagent shows interesting properties for site-specific modification of proteins on the hexahistidine chain. The simplicity of the conjugation reaction makes this method suitable for the development of radio-biopharmaceuticals. Site-specific bioconjugation with PEG-monosulfone leads to better control of the number of groups conjugated, compared to the non-specific bioconjugation method on lysines.
 Cong, Yuehua, et al. "Site-specific PEGylation at histidine tags." Bioconjugate chemistry 23.2 (2012): 248-263.
 Peciak, Karolina, et al. "Site-selective protein conjugation at histidine." Chemical science 10.2 (2019): 427-439.
 Waibel, Robert, et al. "Stable one-step technetium-99m labeling of His-tagged recombinant proteins with a novel Tc (I)–carbonyl complex." Nature biotechnology 17.9 (1999): 897.
Keywords: poly-histidine, site-specific bioconjugation, recombinant proteins
Assessing abdominal aortic aneurysm response to imatinib treatment using [18F]fluorothymidine positron emission tomography / computed tomography (#258)
Richa Gandhi1, 2, John Wright1, Joanna Koch-Paszkowski1, Parkavi Kandavelu1, Christopher Cawthorne3, Stephen Archibald4, Marc Bailey1, Charalampos Tsoumpas1
1 University of Leeds, Leeds Institute of Cardiovascular and Metabolic Medicine, School of Medicine, Leeds, United Kingdom
An [18F]fluorothymidine ([18F]FLT) positron emission tomography / computed tomography (PET/CT) study has recently revealed increased cell proliferation 14 days post-angiotensin II (AngII) infusion in mice with abdominal aortic aneurysm (AAA). This finding suggests that it may be possible to measure an early response to anti-proliferative therapy using [18F]FLT PET/CT. Herein, we tested this in the AngII AAA mouse model when treated with imatinib, a commonly used cancer drug that inhibits proliferation.
Fourteen-week-old apolipoprotein E-knockout mice received 14-day AngII infusions via subcutaneously implanted osmotic mini-pumps. On day 12, the mice underwent abdominal ultrasound scanning to assess aortic size. On days 13 and 15, 20-min static PET/CT was performed 90 min after [18F]FLT injections (mean 9.4±0.1 MBq) in the lateral tail vein. Between the scans, 2 mice received 3 imatinib doses via oral gavage on days 14–15; a matched control received tap water. PET/CT images were reconstructed using the maximum likelihood estimation maximisation algorithm (25 iterations); 3D isocontour regions of interest were manually drawn in all aortic regions to extract maximum standardised uptake values (SUVmax). Ex vivo gamma counting of aortae was finally performed following PET/CT on day 15.
Aortic volumes were increased on day 12 and similar to those reported previously for this model, indicating AAA development (n=3: mean ± standard deviation (SD), aortic volume 23.7 ± 1.2 mm2). [18F]FLT uptake was observed in the AAA region on day 13 (mean ± SD, SUVmax 0.2 ± 0.0). After administering 3 doses of imatinib, no [18F]FLT signal was observed in the AAA region (n=2: mean ± SD, SUVmax 0.02 ± 0.0). However, after administering 3 doses of tap water, the [18F]FLT signal in the AAA region remained unchanged (n=1: SUVmax 0.2). Abdominal aortae from the imatinib-treated mice revealed reduced ex vivo [18F]FLT counts (n=2: mean ± SD, 18.7% ± 13.7% injected dose/g relative to the heart) than that in the abdominal aorta from the vehicle-treated animal (n=1: 51.5% injected dose/g relative to the heart).
To our best knowledge, this is the first pilot study to demonstrate the potential utility of [18F]FLT PET/CT to assess an early therapeutic response to an antiproliferative agent in the AngII AAA mouse model. AAA uptake of [18F]FLT was shown to decrease after receiving only three doses of imatinib. The preliminary results are promising, and a follow-up study with a larger sample size is warranted.
Pre- and post-treatment [18F]FLT PET/CT images
Representative coronal-view static PET/CT images of 14-day AngII AAA model pre- and post-treatment with vehicle or imatinib. White arrows indicate AAA location. [18F]FLT, [18F]fluorothymidine; PET/CT, positron emission tomography/computed tomography; AngII AAA, angiotensin II abdominal aortic aneurysm; SUV, standardised uptake value
Keywords: [18F]fluorothymidine, imatinib, abdominal aortic aneurysm, ApoE-/-, angiotensin II
Automated Diagnosis in Large Vessel Vasculitis Using FDG PET-CT (#565)
Lisa Duff1, 2, Andrew Scarsbrook3, 4, Sarah Mackie5, Russell Frood3, 4, Marc Bailey1, 6, Ann Morgan5, Charalampos Tsoumpas1, 7, 8
1 University of Leeds, Leeds Institute of Cardiovascular and Metabolic Medicine, Leeds, United Kingdom
Large Vessel Vasculitis (LVV) occurs in a variety of syndromes that inflame the aorta and its major branches. Due to variation in symptoms, rarity of the diseases, and variations in treatments administered prior to the scans, it can be difficult to distinguish LVV from atherosclerosis, accurately diagnose the underlying syndromes and determine long-term prognosis. The aim of this study was to explore the diagnostic utility of [18F]FDG PET-CT (Fluorine-18 Fluorodeoxyglucose Positron Emission Tomography – Computed Tomography) (Figure 1) in LVV by discovering radiomic imaging biomarkers [1-2].
The aorta was manually segmented using open-source software (3D Slicer) on the baseline [18F]FDG PET-CT scans of 29 patients with confirmed LVV and 10 normal controls (Figure 2) . The PET scans and segmented masks were then resampled to a voxel size of 4x4x4mm. Pyradiomics was used to extract 106 radiomic features [4-5]. The only modification added to the parameter input file for Pyradiomics was to set the bin width to 0.4. No additional filters were used. The Mann Whitney U test was used to find radiomic features that were significantly different between patients with active LVV and normal controls. The p value was adjusted from p=0.05 to p= 0.00047 using the Bonferroni correction to prevent Type 1 errors due to multiple testing. The Cohen’s d value (effect size) was calculated.
Three radiomic features were significantly different between active LVV and control groups (p<0.00047) - GLCM (Gray Level Co-occurrence Matrix) Correlation, GLCM IMC2 (Informational Measure of Correlation 2) and GLDM (Gray Level Dependence Matrix) Dependence Entropy (Figure 2). The Cohen’s d value is used to express the standardised difference between two means and while all three showed differences, GLCM Correlation was highest. GLCM Correlation is of particular interest as it performed best in the Mann Whitney U test, had a large Cohen’s d value and also performed well in both tests when applied to a smaller sample size (10 LVV patients, 10 controls).
Three radiomic features show potential as imaging biomarkers for active LVV in this pilot study. In particular GLCM Correlation performed well and was consistent. These initial findings will be tested in a larger sample size and externally validated in a multicentre study.
We thank Pratik Adusumilli for his help in validating the segmentation method. This study was funded by the EPSRC Centre for Doctoral Training in Tissue Engineering and Regenerative Medicine; Innovation in Medical and Biological Engineering – grant number EP/L014823/1. Ann Morgan is principal investigator on the Medical Research Council TARGET (Treatment According to Response in Giant Cell Arteritis) Partnership Grant (MR/ N011775/1). Dr. Bailey is personally funded by the British Heart Foundation Intermediate Clinical Research Fellowship (FS/18/12/33270). Dr. Tsoumpas is funded by a Royal Society Industry Fellowship (IF170011) and a Horizon 2020 Marie SkłodowskaCurie Action Research & Innovation Staff Exchange (645757). Dr. Frood is funded by Innovate UK via the National Consortium for Intelligent Medical imaging (NCIMI).
 Blockmans, D, 2011, 'PET in vasculitis'. Annals of the New York Academy of Sciences, 1228(1), pp.64-70.
 Coath, F, Gillbert K, Griffiths, B, Hall, F, Kay, L, Lanyon, P, Luqmani, R, Mackie, SL, Mason, JC, Mills, J, Mollan, S, Morgan, AW, Mukhtyar, C, Quick, V, Watts, R, Dasgupta, B, 2019, 'Giant cell arteritis: new concepts, treatments and the unmet need that remains', Rheumatology, Volume 58, Issue 7, Pages 1123–1125,
 Slicer. 2019. 3D Slicer. [ONLINE] Available at: https://www.slicer.org/. [Accessed 5 February 2020].
 van Griethuysen, JJM, Fedorov, A, Parmar, C, Hosny, A, Aucoin, N, Narayan, V, Beets-Tan, RGH, Fillon-Robin, JC, Pieper, S, Aerts, HJWL, 2017, 'Computational Radiomics System to Decode the Radiographic Phenotype'. Cancer Research, 77(21), e104–e107.
 Pyradiomics. 2020. Pyradiomics. [Online]. Available at: https://pyradiomics.readthedocs.io/en/latest/index.html#welcome-to-pyradiomics-documentation [Accessed 5 February 2020]
Keywords: Radiomics, Large Vessel Vasculitis, PET