key: cord-0007020-4f1oyx9f
authors: nan
title: Proceedings of the World Molecular Imaging Congress 2019, Montreal Quebec, Canada September 4-7, 2019: Late-Breaking Abstracts
date: 2019-11-19
journal: Mol Imaging Biol
DOI: 10.1007/s11307-019-01453-z
sha: 3ea004eb44335f54797b9667af6e12e655aeb31a
doc_id: 7020
cord_uid: 4f1oyx9f

nan

In-vivo preclinical imaging of tumor hypoxia using EPR, DCE-MRI, and PET-CT with 18F-Miso to improve radiotherapy ID: LB37 Detection and characterization of hydroxyapatite as induced microcalcification in breast tumor by rhBMP-2 Asghar Hajibeigi, UT Southwestern Medical Ctr, asghar.hajibeigi@utsouthwestern.edu

Abstract Body : Background-The detection of microcalcification on mammography is considered an early indicator of a diagnosis of breast cancer1. There are two types of radiologically non-distinguishable calcium deposits in breast tissue, calcium oxalate, which typically occurs in benign cancer and hydroxyapatite which is a key diagnostic marker for malignancies,2,3. Although a limited number of research studies have been carried out to determine the molecular mechanism of microcalcification of breast tumor, recently we showed the role of rhBMP-2 as a physiological mineralization factor caused to induce microcalcification of syngeneic human breast tumor in nude mice by PET/CT imaging using 18F-NaF radiotracer4,5. The aim of present study is to identify and confirm the chemical composition of microcalcification of human breast tumor model induced by rhBMP-2 by both radiographic and microscopic imaging. Materials and Method-Syngeneic breast tumors in nude mouse were generated by subcutaneous injection of 2.5 x 106 MDA-MB231 cells with equal volume of matrigel in PBS into the murine mammary fat pad. After 4 days, each animal received 15ug rhBMP-2 protein by IP or direct injection into the tumor on a weekly basis for 3 weeks before CT scanning. Control animals received PBS as vehicle. After CT, the control tumors with undetectable microcalcification received an intratumoral injection with either a slurry suspension of calcium oxalate or hydroxyapatite in PBS. All animals were intravenously injected with 100uCi of 18F-NaF via tail vein, then scanned with 15 minutes list-mode PET centering at the tumor site and a medium magnification CT was performed centering on the breast tumor sites. After PET/CT imaging, the tumors were removed and fixed with 10% neutral buffered formalin for histological analysis using von Kossa staining. Results-In the present study, the CT images showed no distinction between micro-calcification induced by rhBMP-2 in breast tumor or control tumors injected with slurry suspensions of calcium oxalate or hydroxyapatite. Our findings in both PET/CT imaging and microscopic detection by von Kossa staining demonstrated the type of microcalcification induced by rhBMP-2 was hydroxyapatite similar to control tumors injected with the slurry suspension of hydroxyapatite. In contrast, the calcium oxalate had no detectable uptake of 18F-NaF. Conclusion-In present work, we suggest that rhBMP-2 participates in the transformation of mesenchymal cells into osteoblast-like phenotypes to contribute to the production of mammary hydroxyapatite in malignant breast tumor. Demonstrating the specific form of microcalcification as hydroxyapatite should be considered a helpful tool for the detection and diagnosis of breast cancer. Abstract Body : Background, Motivation and Objective: Ultrasound (US) imaging is routinely used to guide biopsies of the prostate gland in patients with suspected prostate cancer (PCa) (1) . However, US delineation of tumors within the prostate is extremely challenging, and contrast-enhanced US with microbubbles only marginally improves US detection of the tumor mass (2) . A more precise identification of PCa will fulfill an unmet clinical need and enable real-time US-guided tumor biopsy and treatment. Reducing the size of microbubbles to the nanoscale may allow them to extravasate from the leaky tumor vasculature but not in normal tissue, thus enabling improved tumor specificity. Building upon the extravasation concept, in this study, we explored a new dual-dose contrast enhancement technique using NBs to improve contrast in orthotopic prostate tumors in mice. Methods: Human prostate cancer cells were orthotopically implanted into the prostate gland of balbC (nu/nu) mice. For the imaging exam, tumor and liver were localized in the same section of view in B-mode using a clinical US scanner (Toshiba Aplio, 12 MHz). Next, 30 µl of NBs (1.2 x 1010 NBs total) were injected through the tail vein and imaged in the nonlinear contrast harmonic imaging mode (MI 0.1). After 3 minutes, an additional 70 µl of NBs (2.8 x 1010 NBs total) were injected and images collected for 30 min. Time intensity curves (TIC) were constructed from this data. A normalized TIC was calculated using the following equation: Intensityafter0mins -Intensityat0mins. An enhanced signal ratio was calculated using the following equation: (Intensityafter3mins -Intensityat3mins)/ Intensityat3mins. Results and Discussion: As seen in Fig. 1 , following the initial 30 µl injection, a rapid enhancement was observed in the liver. Tumor enhancement was comparatively lower. After the second injection, change in tumor enhancement was significantly higher than in liver. Fig. b1 shows the original TIC for both regions. Fig. b2 shows the enhanced signal ratio, with significant improvement in the difference between tumor and liver tissue signals. While the underlying mechanism for these differences in enhancement is not understood, the initial extravasation of NBs into tumors could explain the increased secondary enhancement. If properly validated, this approach could be a simple method for amplifying differences between tumors and normal tissue and aid in detection of solid tumors. Abstract Body : With mean survival rate of 5 years (and most cases are fatal) lysomal storage diseases (LSD) are among the most dismal of prognosis in all of medicine. LSD's represent a large number of monogenetic diseases and while rare the prevalence is to hemophilia. As monogenetic diseases with clearly defined genotype-phenotype relations, lysosomal storage diseases are excellent candidates for gene therapy. The transformative results documented in an adeno-associated virus (AAV) gene therapy clinical trial in infants affected by spinal muscular atrophy demonstrated unequivocally the potential of in vivo gene transfer to treat monogenic neurological disorders. To date there is a lack of non-invasive ways to determine biodistribution or activity levels of these AAV therapies in patients. This is a significant hinderance, leaving investigators guessing which organs or structures are effectively treated and, due to the lag time associated with clinical disease progression, this limitation ultimately impacts the evolution of treatment modalities. In order to overcome these limitations, we are developing of a new class of bioresponsive magnetic resonance imaging probes to track enzymatic activity in any organ, peripheral nervous system (PNS), or central nervous system (CNS) over time. Magnetic resonance imaging is an ideal technique for the study of neurological disorders. It has become a gold standard in diagnostic radiology as a result of the absence of ionizing radiation and is capable of true 3D imaging and has been in use for several decades We pioneered the development of bioresponsive (bio-activated) MR contrast agents and since then the library of this class of probes has expanded from enzyme activated agents to include pH sensitive, the detection of ions such as Zn(II) and Ca(II), and redox activated. Here, we describe the development of a platform where a substrate (blocking access of water to a Gd(III) ion) is removed by an enzyme that can be varied to accommodate a number of gene therapy targets. As a result, gene therapy can be noninvasively monitored at both where and when the gene of interest is activated. Abstract Body : Introduction: Inflammatory bowel disease (IBD) is characterized by inflammation and ulceration of bowel segments within the gastrointestinal tract1. It is a life-long chronic disease with no cure, where the primary goal of therapy is maintenance of remission1. Screening patients for regions of active or recurrent IBD and monitoring regions over time present different problems for molecular imaging. Screening is generally a one-time imaging session where increased costs (MRI, CT, SPECT, PET) or radiation (CT, SPECT, PET) is acceptable. However, IBD patients require monitoring of diseased bowel over their entire lives, where use of expensive modalities and contrast agents is not financially sustainable. We lack a low-cost contrast agent and molecular imaging approach that can be used to monitor regions of active IBD over time. Microbubbles (MBs) and ultrasound (US) are an ideal contrast agent/modality for repeated imaging of patients with active IBD due to its low construction and imaging costs, wide availability, and the ability to add targeting agents to the MB shell for molecular imaging of active inflammation2,3. However, addition of conventional targeting agents (antibodies, recombinant proteins) to the shell of MBs dramatically increases their construction cost, negating the cost benefits of this imaging approach. We require a low-cost labelling approach that can reduce the price of conjugating targeting agents to MB shells. Our goal is to make use of the SpyTag-SpyCatcher protein ligase system4 for conjugating a targeting agent (antibody Fab to P-selectin) to the MB surface and test its effectiveness for US MI in a murine model of acute inflammation. Methods: Microbubbles (MBs) were produced by sonicating perfluorobutane (C4F10)-sparged solution containing 90:5:5 molar ratio of DSPC, PE-PEG2000, and PE-PEG2000-Maleimide (Avanti Polar Lipids). PE-PEG2000-Maleimide was pre-labelled with SpyTag-Cys (13 aa; IDT Technologies) using a conventional maleimide-thiol reaction. A Fab specific for P-selectin was created using a phage display library, fused to SpyCatcher (116 aa) using a Gibson assembly, and expressed using a bacterial system4. SpyTag-MBs were incubated with SpyCatcher-Fab to form isopeptide bonds (reaction half-life ~1 min; referred to as SpyC-MBs). SpyC-MBs were washed and concentrated using centrifugation (600 RPM 10 min). Successful ligation of SpyCatcher-Fab to SpyTag-MBs was verified using FITC-Protein L and visualized using confocal microscopy. In vitro binding of SpyC-MBs to activated (TNFa; 10 ng/mL) MS-1 (mouse endothelial) cells was used to verify successful adhesion to activated target cells. For in vivo US molecular imaging, acute inflammation was induced through rectal administration of 1% 2.4.6-trinitrobenzenesulfonic acid (TNBS) in one of two female FVB mice (7-8 weeks). Mice were imaged using non-linear contrast mode on a small animal ultrasound system (Vevo3100; Visual Sonics; MV250 Transducer) following i.v. injection of 3x108 MBs. The resulting acoustic signal was analyzed using VEVOCQ. Mice were euthanized and ex vivo histology of the bowels then performed. Results: SpyCatcher-Fab was determined to have a dissociation constant (KD) of 10 nM. Successful ligation of SpyCatcher-Fab to SpyTag on the MB surface was verified using confocal microscopy after labelling with Protein-L-FITC (Fig.1) . Activated MS-1 cells bound 3.4-times as many SpyC-MBs in comparison to SpyTag-MBs (Persuasive Data). Induction of acute inflammation resulted in an increase in US molecular imaging signal of 2.5 l.a.u. compared to control 0.8 l.a.u (Fig.1) . We validated acute inflammation phenotypes using H&E histology (Fig.1) . Conclusion: In conclusion, these preliminary data show the SpyTag-SpyCatcher-Fab system can be used as an alternative, low-cost approach to add targeting agents to MB shells, and is suitable for imaging acute inflammation in a murine model of IBD. This cost-effective approach for creating a targeted MB may provide clinicians with a tool for identifying and monitoring disease progression in IBD patients. Abstract Body : Purpose: Immune cells bearing the cluster of differentiation 4 (CD4) cell membrane glycoprotein comprise an important subset of the immune system that includes monocytes, helper T cells, macrophages, and dendritic cells. We report the evaluation of a novel PET imaging agent that is potentially specific for imaging CD4 positive cells in vivo. Methods: The FDA approved human monoclonal antibody ibalizumab that is used for the treatment of HIV infection was selected as the targeting molecule on the basis of its low risk of immunogenicity, high specificity for CD4 and established clinical characterization. Ibalizumab (Creative Biolabs, Shirley, NY) was conjugated to desferrioxamine (DFO) and radiolabeled with zirconium-89 by InviCRO, Inc (Boston, MA) to produce [89Zr] -DFO-Ibalizumab. The affinity of ZrDFO-ibalizumab to rhCD4 was assessed by ELISA. [89Zr] ZrDFO-ibalizumab binding to CD4+ T-cells from human blood was evaluated by flow cytometry coupled with radioactivity measurements. Cynomolgus monkeys were chosen for in vivo investigation on the basis of cross-reactivity studies indicating that Ibalizumab has a similar binding affinity to rhesus, cynomologous and human tissues. Results: ELISA and flow cytometry results confirmed unchanged affinity for CD4 of ibalizumab after Zr-labelling. In vivo PET images were obtained at 4hr, 24hr, 72hr, and 168hr following injection of [89Zr] -DFO-Ibalizumab and the images illustrate a specific pattern of localization of the tracer to the lymphatic system, thymus as well as the expected spleen and liver reticuloendothelial clearance. Discussion: [89Zr] -DFO-Ibalizumab represents a new modality that may be useful for imaging CD4+ T-cells in vivo. Following administration in a healthy non-human primate (NHP), SUV from multiple lymph nodes increased over the 168hr time course while clearance was observed from the blood and peripheral organs apart from liver and spleen suggesting that the images may reflect the specific localization of CD4+ T-cells. Abstract Body : Photoacoustic imaging (PAI) is an emerging modality that non-invasively captures images by detecting high frequency pressure waves (ultrasound waves) that are emitted during the non-radiative relaxation of light-excited contrast agents.1 Although limited by light penetration and acoustic attenuation, PAI offers the capability to take real-time, high resolution images, superior to fluorescence imaging and without the use of ionizing radiation.2 With only a small number of clinically used optical imaging/photoacoustic contrast agents in use today, there remains a need for the development of new molecularly-targeted PAI agents. The objective was to develop, evaluate, and refine a general purpose strategy for creating targeted PAI agents using the bioorthogonal reaction between tetrazine (Tz) and trans-cyclooctene (TCO). Not only does this reaction display fast kinetics, biocompatibility and orthogonality, but it is also currently one of the only bioorthogonal-type reactions to be deemed permissible for use in humans.3 Through the use of this platform, a second generation Tz-derived PAI agent, IR-783-Tz, was designed with the intention to be used alongside any TCO-derived targeting or therapeutic agents. The albumin-binding ability of IR-783-Tz was evaluated in vitro through a 'cut-on fluorescence' binding assay with results that were consistent with the notion that near-infrared cyanine dyes possess the ability to bind proteins in the blood.4 As well, an in vivo PAI study using a 143B human osteosarcoma xenograft mouse model, was performed and showed a statistically significant increase in the signal within the tumor compared to background at 1 hour post-injection. Finally, using the well-established bone-targeting agent, alendronate (BP) tethered to a TCO moiety (TCO-BP), the use of IR-783-Tz as a targetable PAI agent was demonstrated. The results of both an in vivo PAI study as well as a bio-distribution study, to determine the degree of uptake of IR-783-Tz-TCO-BP at the bones/joints of healthy Balb/C mice, will be presented. Abstract Body : Background: Hyperpolarized (HP) 13C MR is a powerful technology for monitoring real-time metabolism in tumors. Elevated production of [1-13C] lactate from [1-13C] pyruvate is near-universally observed in many tumors, including those in the liver (1) . Recently, the hepatocyte-specific gadolinium-based contrast agent gadoxetate has been proposed as a method to selectively suppress hyperpolarized signal arising from normal liver cells, while isolating the signals arising from tumor cells(2). Although this strategy has been successfully implemented in vivo, little is known about the detailed interactions of hyperpolarized metabolites with intracellular and extracellular gadolinium. Aim: The aim of this study is to use cell culture and an NMR-compatible perfusion bioreactor to investigate the effects of intracellular gadoxetic acid on the observed signals arising from intracellular and extracellular 13C metabolites. Methods: Embryonic kidney HEK293 cells were stably transfected with human organic anion-transporting polypepetide 1B1(OATP1B1), which is the transporter that is chiefly responsible for importing gadoxetate into cells (3, 4) . These cells were prepared for bioreactor NMR experiments by electrostatic encapsulation into alginate microspheres(5). The cell-laden microspheres were maintained at physiological conditions by circulating DMEM media at 37 with continuing oxygen in the 5mm NMR tubes used in bioreactor. Experiments were conducted on a 500 MHz Varian Inova with a 5mm probe at 37. 7.0μL of [1-13C] pyruvic acid was hyperpolarized using a HyperSense DNP polarizer operating at 3T and subsequently dissolved and neutralized for a final concentration of 50mM. 900 uL of this solution was infused into the bioreactor system and NMR experiments were performed by application of a 30° degree pulse every 3 s and for a total acquisition time of 300 s, using a sweep width of 100ppm. The area under the peak for each metabolite was computed at every time point. Metabolite signals were expressed as a fraction of the total observed HP carbon signal. Intracellular and extracellular lactate signals were separated based on chemical shift as described previously. In order to determine the effects of intracellular gadoxetate on the HP metabolite signal, HP 13C experiments were performed prior to and after infusion of 0.5 mM gadoxetate (figure 1a). Control experiments were performed using the same protocol with gadopentetate dimeglumine, which is a strictly extracellular agent. Cell viability was dynamically monitored by measuring the peaks of nucleoside triose phosphate produced by the ATP molecule using 31P NMR. Results: For all experiments, extracellular lactate was higher than intracellular lactate, suggesting a rapid lactate export in these cell lines. Following infusion with gadoxetate, total production of intracellular and extracellular lactate was reduced by 41% and 43.2%, respectively. Alanine and bicarbonate signals were completely suppressed. This reflects the expected selective suppression of intracellular metabolite polarization due to gadoxetate residing in cells. Because all extracellular lactate was generated within the cells, decreased intracellular lactate signal lead to a similar decrease in the extracellular lactate signal. By contrast, when the strictly extracellular gadolinium agent (gadopentetate) was used, total production of intracellular and extracellular lactate was only reduced by 5.6% and 6.4%, respectively. Alanine and bicarbonate signals were preserved (Figure 1b-c) . Conclusion: Using a perfusion bioreactor system and isolated cells, we were able to observe the selective effect of the intracellular gadolinium agent, gadoxetate, on the hyperpolarized signal generated within cells. In the future, this system will enable better understanding of how targeted relaxation agents affect observed HP 13C signals, expanding the way these agents can be used to isolate specific cell types in vivo. Abstract Body : Although breast tumors are characterized by inherent heterogeneity, the evolving cellular organization through neoplastic progression is poorly understood. Here we report a rapid 3D imaging protocol termed Large-scale Single-cell Resolution 3D imaging (LSR-3D), which allows visualization of the architecture of normal tissue and entire tumors at cellular resolution. This technique incorporates a one-step clearing agent termed FUnGI that can be applied to endogenous fluorescence or immunofluorescence labeling of tissue. In concert with multicolor-lineage tracing to analyze the contribution of basal (K5+) and luminal progenitor (Elf5+)-derived cells to oncogenesis, imaging revealed profound clonal restriction during tumor progression. Expression profiling of tumor clones arising in Pten/p53-deficient glands identified distinct molecular signatures. Strikingly, most clones harbored cells that had undergone an epithelial-mesenchymal transition, indicative of widespread and inherent plasticity. Hence, an integrative pipeline that combines lineage tracing, LSR-3D imaging and clonal RNA-seq technologies offers a comprehensive path for studying mechanisms underlying cellular heterogeneity within entire tumors. 

Abstract Body : Gamma cameras are frequently used for SPECT (Single Photon Emission Computed Tomography) imaging both in preclinical research and in human medicine. SPECT visualizes biodistribution and accumulation of radiolabeled tracers in vivo. The main disadvantage of the standard SPECT examination is its low sensitivity (less than 1%) caused by losses of signal by use of collimators. Very high radiation activity of the tracer is needed for successful imaging which results in considerable irradiation of the imaged object. Compton camera can be used for increasing of SPECT sensitivity. Its principle is based on the use of Compton scattering, which makes it possible to create an image of a radiation source without using a heavy and signal weakening collimator. The Compton camera usually contains two layers of detectors. The first interacts with the primary gamma radiation and the second records secondary energy scattering. By capturing and evaluating a series of such scatterings, it is possible to reconstruct the shape and location of the original gamma radiation source. In our study, we used a commercially available photon counting pixel detector to construct the Compton camera. In the first part of the study, the camera consisted of two detectors -the first one had a CdTe sensor with a thickness of 1 mm and the second one had 1 mm thick Si sensor. The detectors were placed 8 mm apart and connected by a sync cable. The study determined that Compton camera-based cameras can display the position and shape of a gamma source with a resolution of about 6 times better than current available devices, in addition to a significantly wider display field. In the second part of the study, we managed to construct a single-layer Compton camera (with 2mm thick CdTe sensor) and we achieved similar results. The cost has been halved and the field of view was even wider. Based on experimental data, we designed (constructed) a prototype of SPECT camera using Compton scattering (Fig. 1 ). The advantages mentioned above (higher resolution, wider display area) are complemented by a significant reduction in the overall size and weight of the device. 

Abstract Body : Near-Infrared Photoimmunotherapy (NIR-PIT) is a newly-developed cell-specific cancer therapy which employs the combination of a monoclonal antibody-photoabsorber (IRDye700DX, IR700) conjugates (APCs) and therapeutic NIR light. When APCs are bound to antigen-expressing tumor cells and exposed to therapeutic NIR light, necrotic cell death is selectively induced on target cells within a few minutes post-exposure. Therefore, compared to other conventional treatments, NIR-PIT can destroy target cells with minimum side effects due to its highly selective cell killing. A Food and Drug Administration (FDA) designated fast-track global phase III clinical trial of NIR-PIT for patients with inoperable head and neck cancer has been underway worldwide. IR700 is not only a photo-activatable toxic agent but also a diagnostic NIR fluorescence dye. Hence, fluorescence imaging before and during NIR light exposure could be used for evaluating precise tumor location where NIR light should be exposed. Furthermore, as shown in previous studies, fluorescence signal of IR700 turns off after the photo-chemical reaction. Therefore, real-time monitoring of the fluorescence intensities decay is beneficial for quantitative evaluation of the therapeutic efficacy resulted in estimating appropriate therapeutic light dose. In mouse models, imaging with low-power NIR light before and after therapeutic exposure has been already performed. However, real-time fluorescence imaging for monitoring NIR-PIT using therapeutic NIR light have not been established. This technology allows us to perform easy and accurate NIR-PIT. In this study, we performed real-time IR700 fluorescence imaging during NIR-PIT treatment using a NIR fluorescence imaging system under co-development with Shimadzu Corporation. For tumor identification during EGFR-targeting NIR-PIT, A431 tumor-bearing mice were exposed to therapeutic NIR light (wavelength: 690 nm, output: 150 mW/cm2), and real-time fluorescence imaging was performed. Fluorescence on the tumor was successfully captured with the new system during exposure to the therapeutic NIR light. The obtained images were consistent with fluorescence images which were taken with commercially available small animal imaging devices before NIR-PIT. This result suggested that real-time fluorescence imaging of IR700 could be applied for accurate tumor localization that allowed us to easily expose NIR light to entire tumors. Next, we took videos of fluorescence images of the tumor-bearing mice during NIR-PIT and evaluated their temporal changes quantitatively. Consequently, these videos revealed that fluorescence intensity on the tumor decreased in a light-dose-dependent manner. These results indicated that real-time monitoring of fluorescence intensity can be used for estimating real time therapeutic efficacy, resulting in optimizing NIR light dose at the treatment. In conclusion, real-time fluorescence imaging during NIR-PIT could be used for determining precise tumor area and appropriate therapeutic NIR light dose that allows us to easily perform accurate NIR-PIT. Abstract Body : Background: Colon wall thickening is a feature of colon disease that can be detected by MRI in the clinical setting as well as in preclinical models of colon disease. Ordinarily, this measurement requires an expert reader to manually segment multiple slices by determining and marking the boundaries of the colon wall, a process that requires significant skill and time. Using curated preclinical training and validation image datasets previously segmented by an expert reviewer, we sought to develop an automated segmentation algorithm based on machine learning methods and to measure the performance of this algorithm in comparison to expert human readers. Methods: Previously segmented T2 weighted mri images of the colon in mice from studies of dextran sodium sulfate (DSS) induced colitis were used to train and validate a convolutional neural network (CNN) implemented with a UNET architecture able to estimate the probability of a pixel being within the colon wall in an unsupervised automated analysis. The training dataset consisted of approximately 850 slices from N=15 animals and the validation dataset consisted of approximately 1100 slices from N=16 animals. Following analysis of the validation dataset by the CNN, the largest connected group of voxels assigned a probability greater than or equal to 50% were then segmented as colon wall. We subsequently compared the expert and machine segmentations as well as metrics computed from the segmentations such as thickness and intensity. Results: Among high quality images, the thickness as measured from the automated segmentations of individual slices had a correlation of R2 = 0.69 when compared with thickness as measured from the expert segmentations. Inclusion of all images including those of lower quality reduces the R2 to 0.57. When the thickness measurements from the high quality images are averaged across slices within individual animal imaging timepoints, the R2 increases to 0.83. Mean image intensity across the colon wall also exhibited high correlation with expert reader measurements when similarly averaged across slices (R2 = 0.99). The similarity of all voxels with greater than 50% probability of being within the colon wall was was calculated using the Sørensen-Dice coefficient within the validation set was 0.74. Conclusion: Using a modest but practically acquirable training dataset, the UNET CNN was able to reproduce measurements made manually by an expert reader with similar results and reasonable correlation between derived measurements like thickness, area and average ROI intesity. As expected, Image quality impacts the ability of the UNET CNN to reliably segment the colon wall, and correlation with expert human readers drops with falling image quality. Further studies are warranted to understand the performance of this system under different experimental conditions. Automatic colon wall segmentation can reduce the turnaround time in preclinical studies while standardizing the output metrics without requiring immediate expert review. Abstract Body : Cyclosporine A (CsA) remains the gold standard immunosuppression therapy for kidney transplant recipients despite the high risk of allograft injury, but Rapamycin (RPY), considered to be significantly less nephrotoxic than calcineurin inhibitors, has been suggested as an alternative to CsA for long term maintenance. Due to the impact of drug toxicity on R&D costs, pharmaceutical companies are adopting earlier preclinical safety testing to minimize drugs with adverse side effects from advancing in their pipeline. In vivo fluorescence molecular imaging offers non-invasive, sensitive detection of a variety of biological changes associated with disease and treatment in preclinical animal models, and this approach has been shown to detect drug-induced liver injury even after single doses of drugs. To examine the potential utility of NIR fluorescence imaging for early assessment of acute drug-induced kidney injury in mice, we propose two methods, 1) direct kidney imaging using a cocktail of probes detecting three different biological changes and 2) use of a NIR fluorescent inulin to determine functional Glomerular Filtration Rate (GFR) changes. We assessed CsA and RPY as model immunosuppressive drugs due to their significantly different toxicology profiles and used sensitive in vivo imaging techniques to characterized very early drug-induced biological and functional changes in the kidneys and various organs. To generate dose response curves, we injected individual animals an IP bolus dose of each drug starting near the drug's TD50. For detection of drug-induced effects, each animal received a cocktail of probes, AMT-750, consisting of Annexin-Vivo 750 (cell death), MMPSense 750 FAST (inflammation), and Transferrin-Vivo 750 (iron uptake) at 2 or 24h post-drug to image kidney biological changes 24h later. This method for early risk assessment of drugs facilitated detection of toxicity within 24-48h post-drug, allowed single or multi-drug administration in a small n of animals, and enabled multi-organ analysis. Initial results showed elevated kidney signal, indicative of possible injury, observed by epifluorescence imaging for CsA mice at all doses compared to vehicle, with the kidney change lasting over 24h. RPY kidney signal did not increase significantly, an indication that it is less toxic even at high dose. Evaluation of excised kidneys at high dose CsA and RPY agreed with the in vivo results but more importantly revealed cortical localization of kidney damage compared to the pathology results which failed to distinguish between these drugs and was not sensitive enough to detect toxicity, inflammation, or necrosis by hematoxylin-eosin. GFR-Vivo 680 imaging, noninvasive measurement of probe blood clearance over time in heart blood for assessment of GFR, confirmed the AMT-750 results despite some degree of vehicle-induced GFR effects. Although vehicle effect can contribute to the early drug results, it is evident that multi-dosing of CsA induced a cumulative effect that appears to predict irreversible changes in kidney biology. These results indicate that RPY is significantly less nephrotoxic than CsA and suggest that our optical imaging approaches can effectively distinguish between two pharmacologically distinct drugs. The ability to detect both transient and prolonged drug-induced kidney changes and can be useful to the study of kidney disease and in monitoring of drug safety. Abstract Body : Since the introduction of PSA testing, significantly more men have been diagnosed and treated for prostate cancer. Patients with slow growing, early prostate cancer may opt for watchful waiting, but an individual with early prostate cancer that is aggressively growing might choose surgery and radiation therapy. Clinical radiation therapy is a noninvasive means of killing cancer cells and effectively reducing tumor burden. This method of treatment is prescribed for more than 50% of prostate cancer patients.1 Although radiation therapy is highly effective for the majority of cancer patients, the nonspecificity of irradiation can result in toxicity for surrounding tissues, which is especially problematic for patients with tumors that require high radiation doses or are difficult to target with image-guidance. Magnetic resonance imaging (MRI) is a powerful clinical imaging modality that provides high-resolution three-dimensional images of soft tissues.2 Magnetic resonance (MR) molecular imaging or molecular MRI of prostate cancer biomarkers, such as Prostate-specific membrane antigen (PSMA), can facilitate non-invasive prostate cancer detection3 and MRI-guided radiotherapy. The objective of this study is to develop a nanoparticle technology that will improve the success rate for prostate cancer diagnosis, precise cancer localization and radiation therapy. PSMA-targeted gold nanoparticles (AuNPs) were synthesized and conjugated with a 1,2 dithiolane modified Gd(III)-macrocycle, and the particle size was measured by dynamic light scattering (DLS) and TEM, and relaxivity was measured by a NMR minispec (1.4T). In vitro binding affinity, cell targeting, in vitro MR imaging, and radiotherapy were tested with PSMA-expressing PC3pip cells and PSMA-lacking PC3flu cells. In vivo tumor targeting and MR imaging was monitored by 7T preclinical MRI scanner. Radiation of tumor was performed and tumor growth was monitored over 18 days. PSMA-targeted Au-Gd NPs had an average hydrodynamic size of 7.8 nm, and conjugating Gd(III) complex to AuNPs significantly improved the r1 relaxivity to 20.6 mM−1 s−1, compared to only 5.5 mM−1 s−1 for the free Gd(III) complex. In vitro cellular uptake experiments demonstrated significantly higher Au-Gd NPs uptake in PC3pip cells than in PC3flu cells. Not surprisingly, the PC3pip cells demonstrated a greater (brighter) MR contrast at 7T than the PC3flu cells, and the signal-to-noise ratio increase for PC3pip cells was 4.2 times of that for PC3flu cells. In vivo, MRI results demonstrated higher accumulation and MR contrast of targeted Au-Gd NPs in PC3pip tumors than in PC3flu tumors, with a contrast-to-noise ratio (CNR) of 13.9 and 5.64, respectively. X-ray irradiation of tumors resulted in diminished PC3pip tumor. The nanoparticle system described herein is envisioned to provide early prostate cancer detection and permit MRI-guided precise radio-ablation of cancerous tissues with potentially less damage to surrounding tissues. Abstract Body : Neuroblastoma surgery is considered a big challenge for the pediatric surgeon1,2,3. During debulking surgery, it is difficult to distinguish between tumor-and healthytissue. Tumor-specific fluorescent imaging agents are upcoming and recently introduced for clinical use4. The disiaganglioside GD2 is overexpressed on neuroblastoma cells and currently being targeted in immunotherapy treatment5. This study describes the evaluation of anti-GD2, a novel fluorescent target, for intraoperative tumor imaging. Anti-GD2 (Qarziba®) was conjugated to the fluorophore IRDye800CW and binding capacities were determined and validated using Flowcytometry and 3D multiphoton microscopy. In vivo, dose escalation, blocking-, pharmacokinetic-, and biodistribution studies were performed. Images were obtained using both PEARL small animal imager and QUEST imaging system. The optimal dose for fluorescence tumor detection was 1nmol. This study describes the development and the evaluation of Qarziba-IRDye800CW and the feasibility to visualize neuroblastoma derived tumors in vivo. Abstract Body : Introdution: The pathogenesis of bacteria that cause infections in humans is highly associated with a biofilm-mode of growth that renders them resistant to antimicrobials and phagocytic cells (1) . Bacterial biofilms are especially problematic in prosthetic joints and other implanted devices as well as in the cystic fibrosis lung (2). Adding to the challenge of combating biofilm infections is the difficulty in detecting their presence. No clinical imaging modality can specifically detect biofilms and obtaining biopsies on deep-seated infections is highly invasive (3) . We have identified a peptide probe, (4Iph)(f)LPNSNHIKQGL (4Iphf-HN17), that we show targets biofilms of P. aeruginosa, making it a potentially suitable diagnostic probe for in vivo optical and/or PET application. The concept of our approach is innovative because the sensitive detection of biofilms has not been the focus of previous efforts related to imaging bacterial infection in the body. Additionally, the uptake of the probe should not be reliant on the metabolic state of bacterial cells, which is known to be quiescent for biofilm embedded cells, and the neutral charge of the probe may facilatate more robust biofilm penetration over cationic probes. We used in vitro and in vivo infection models to evaluate the ability of 4Iphf-HN17 to specifically target P. aeruginosa biofilms, used as a model biofilm forming pathogen in this study. Materials and Methods: 4Iphf-HN17 was synthesized by solid state methods and labeled at K with FITC or Cy5. Planktonic (free floating) cultures of P. aeruginosa cells (strain PAO1) were incubated with HN17-FITC (1 nM to 20 µM). Sixteen h after peptide addition the optical density of the cultures was measured as a surrogate for bacterial growth. Flow cytometry and confocal laser scanning microscopy (CLSM) were used to measure peptide association to planktonic P. aeruginosa cells and 24 h flow-cell grown biofilms, respectively. The probe was evaluated for bacteria specificity using human A549 epithelial cells infected with mCherry-expressing P. aeruginosa. 4Iphf-HN17-FITC as well as an irrelevant peptide of similar size and identical fluorescent label were then incubated for 1 h followed by washing. Image derived co-localization measurements were obtained to quantify probe distribution. Finally, 4Iphf-HN17-Cy5 was evaluated in a mouse infected wound model using real-time optical imaging and CLSM on isolated infected tissues. Results and Discussion: 4Iphf-HN17-FITC and 4Iphf-HN17-Cy5 were synthesized at a purity of >90%. 4Iphf-HN17-FITC labeled planktonic P. aeruginosa cells in a dose-dependent manner and did not inhibit growth over 16 h. CLSM analysis showed that 4Iphf-HN17-Cy5 labeled P. aeruginosa biofilms in a pattern consistent with biofilm matrix interaction. 4Iphf-HN17-Cy5 preferentially targeted aggregates of P. aeruginosa cells adhered to epithelial cells (64.9 ± 6.4 % colocalization with P. aeruginosa aggregates vs. 8.5 ± 2.6 % co-localization with epithelial cells; p < 0 .01). The irrelevant probe showed no labeling. Optical imaging of mice with P. aeruginosa-infected wounds imaged 18 h post injection of 4Iphf-HN17-Cy5 showed wound P. aeruginosa lux-signal that strongly colocalized with Cy5 signal (Figure 1 ). CLSM confirmed a punctate Cy5 signal pattern that co-localized with Tdtomato-expressing P. aeruginosa and anti-psl staining (a surrogate for the bacterial biofilm matrix). In contrast, Cy5 signals were minimal and more diffuse in wounds harvested from probe-injected but uninfected animals. Conclusions: 4Iphf-HN17 is a promising probe for imaging P. aeruginosa infections in vivo. Additional in vitro studies are warranted to understand targeting mechanisms to P. aeruginosa biofilms as well as to determine probe interaction with other relevant bacterial pathogens. Future in vivo studies will seek to determine if the probe can sensitively report on varying bacterial loads and treatment responses and further examine bacterial specificity under sterile inflammatory conditions. Abstract Body : Dysregulation of the innate immune system contributes to the pathophysiology of a variety of diseases, which impact the global inflammatory state of the body, leading to a variety of complications.1 These include obesity and diabetes, atherosclerotic cardiovascular disease (CVD), defined by arterial wall inflammation,2 and auto-immune inflammatory diseases, such as rheumatoid arthritis, colitis, and lupus, which all contribute to significant morbidity and mortality. [3] [4] [5] In oncology, the tumor microenvironment (TME) plays an important role in immune suppression, immunosurveillance, and immuno-editing.6 Critically, oxidizing species such as reactive oxygen species (ROS), for example, generated by NADPH oxidase 2 (Nox2) and Myeloperoxidase (MPO), are key effector and signaling molecules in all of these inflammatory disease states. These inflammatory states may occur at deep tissues or sites not amenable to repetitive biopsy, thereby limiting access to vital information to guide therapeutic choices. Robust clinical imaging of the innate immune system in vivo remains a challenge. Contrast MRI, [18F]FDG-PET, or [68/67Ga]Ga-Citrate-PET/SPECT lack the sensitivity and/or specificity to detect changes in the innate immunity system in many of these diseases,7-9 and novel approaches to detect inflammation are underway.10,11 Could the production of oxidizing species provide a mechanistic basis for improving the sensitivity of PET methods for detecting the activation of the innate immune system? Naphthol is known to be oxidized by Nox2-and MPO-derived ROS species in vitro and to bind to activated neutrophils in cellulo.12 Herein, we develop and test 4-[18F]Fluoronaphthol ([18F]4FN, Figure 1A ) as a novel radiopharmaceutical to detect active inflammation by PET. A robust and automated synthesis of [18F]4FN using copper-mediated radiofluorination was developed and validated on a commercial synthesizer (GE TracerLab-FX).13 To date, [18F]4FN has been successfully synthesized more than 20 times with an average activity yield of 6.8±2.5% (n=22), with >99% radiochemical purity, and up to 140 GBq/µmol molar activity. The tracer has at least a 4 h shelf-stability in a clinically translatable formulation (90% PBS, 10% EtOH), indicating its suitability for both on and offsite use. Further, [18F]4FN was also stable in mouse plasma for at least 1 h. Pilot in vitro and in vivo studies were promising. Using physiologically relevant concentrations of MPO, we can readily oxidize >90% of the reporter. When incubated with "neutrophil-like" human cells (all-trans retinoic acid-differentiated HL-60 ALL cells) for 30 min, [18F]4FN demonstrated a robust 4.2 +/-0.3 fold increase in retention in cells treated with the immunostimulant phorbol-12-myristate-13-acetate (PMA) ( Figure 1B , SEM, n=3 experiments, n=3-4 replicates per experiment). L-012 was utilized to validate ROS production both via bioluminescence macroscopy (Spectrum, Perkin Elmer) and microscopy (TiE, Nikon). A PMA model of mild contact dermatitis (earlobe) was utilized to conduct pilot in vivo experiments.14 Imaging of [18F]4FN at 1 hr post injection of the radiotracer in vivo yielded good contrast-to-noise ratios in two independent strains of adult female mice (Balb/c and C57Bl6) ( Figure 1D ). The reporter is sufficiently robust that both IP and IV injections yielded images with good contrast ratios and large effect sizes (Cohen's coefficient » 0.7) ( Figure 1C) . Further, in this model, [18F]4FN yields superior contrast to [18F]FDG (p=0.004). Finally, in an LPS model of arthritis, [18F]4FN correlated well with L-012, a validated bioluminescence reporter of ROS and activation of the innate immune system ( Figure 1E , r=0.9316 p=0.0008 n=4 mice, n=8 ankles). Broadly, [18F]4FN demonstrates mixed renal and hepatobiliary excretion similar to other clinically translatable PET agents. In summary, [18F]4FN could be readily synthesized with high molar activity, good yields, was stable in both buffer and mouse plasma and appeared to be a suitable PET agent for monitoring ROS produced by activation of the innate immune system in deep tissues.

Ping Wang, Michigan State University, wangpin4@msu.edu

Abstract Body : Type 1 diabetes (T1D) is an autoimmune disorder in which T cells destroy insulin-producing β-cells, leading to a lifelong dependence on exogenous insulin. Islet transplantation (Tx) is a promising approach to cure this disease. However, the long-term success of islet Tx remains frustratingly limited because of severe grafts loss post Tx [1, 2] . In this study we synthesized and tested iron oxide-based magnetic nanoparticles (MN) conjugated to siRNA against caspase-3 gene (MN-siCaspase3) for delivery to pancreatic islets prior to Tx [3] . We expected that MN-siCaspase3 nanoparticles would provide islet protection prior to Tx. At the same time, by the use of a magnetic reporter, labeled islets could be monitored after Tx by in vivo magnetic resonance imaging (MRI). Islets isolated from a healthy baboon donor were labeled with MN-siCaspase3 for 48 hrs prior to Tx. The anti-apoptotic protective effect of MN-siCaspase3 was tested with an apoptotic ladder assay in vitro. The results showed minimal loss of islets after 2-day culture compared to controls. For in vivo study, the labeled islets were transplanted into the diabetic recipient's portal circulation. The animals were maintained on a standard immunosuppressive protocol. In experimental group we observed a dramatic reduction in insulin requirement even when animals were transplanted with a marginal number of islets compared to the animals transplanted with the islets labeled with control nanoparticles (Fig. 1A) . In vivo assessment of the grafts was performed using a clinical MRI scanner (Fig. 1B) . In addition, we have shown histologic evidence of positive insulin staining in cells from the liver in recipients of direct portal infusion. Our results demonstrated for the first-time protective effect of siRNA-conjugated magnetic nanoparticles in transplanted islets that can be monitored by in vivo MRI in nonhuman primates. We expect that similar conjugates will further minimize the number of islets needed for Tx. 

Abstract Body : Pancreatic neuroendocrine tumor (PNET) are notoriously difficult to find and treat surgically due to their small size and ectopic locations, thus, the avoidance of normal pancreas represent a significant unmet clinical need. Substantial efforts have been focused on improving preoperative and intraoperative identification of tumors and their margins, however, it presents a major challenge to detect PNET with sufficient sensitivity. Therefore, bioengineering of a contrast agentto target such a tiny neoplasm is of significant importance with improved specificity and biodistribution. An ideal tumor-targeted contrast agent should have high uptake and prolonged retention in malignant tissue as well as minimum uptake and fast clearance from surrounding normal tissues. Bioconjugation of fluorophore(s) to a ligand has been mainly used to target overexpressed receptors on tumors. However, the size of the final targeted ligand can be large (>10 kDa) and cannot readily cross the microvasculature to meet the specific tissue, resulting in low targetability with a high background. The delivery and retention of a molecule in tumor tissue is dependent on the size, charge, hydrophobicity, pharmacodynamics, pharmacokinetics and its transport across the tumor vasculature. However, the physicochemical properties of cancer-targeting fluorophores are not well established. In the present study, a lead PNET-targeted, phenoxazine-based Near-infrared (NIR) fluorescent contrast agent has been developed by analyzing the biodistribution of a small fluorophore library with the potentials to accumulate into PNET using insulinoma-bearing transgenic mice. Thereby, we have determined the physicochemical properties and pharmacokinetics required to achieve the specific targeting for PNET that can provide image-guidance needed during the tumor resection. This bioengineered fluorophore permits sensitive detection of ultrasmall ( < 0 .5 mm) ectopic tumors within a few seconds after a single bolus injection, highlighting every tumor in the pancreas from the surrounding healthy tissues with reasonable half-life. Abstract Body : Objective: Cellulose is a naturally occurring, ubiquitous, amphiphilic polymer that displays remarkable physical and chemical properties with a wide variety of applications.1 Cellulose based nanoparticles (CNPs), such as cellulose nanofibers and cellulose nanocrystals, are characterized by their large surface area, high aspect ratio, high Young's modulus, biodegradability and abundance, leading to potential applications in packaging, sensing, electronics and biomedicine.2 Traditionally, CNPs have been produced from cellulose by acid hydrolysis, enzyme treatment or chemical oxidation that are cumbersome and long winded. 3 We hereby propose the synthesis of spherical CNPs using a facile oil in water (o/w) emulsification technique. This unique strategy can be used to employ the encapsulation of magnetic iron oxide nanocrystals and/or X-ray dense tanalum oxide nanocrystals (TaOx NCs) for Magnetic Resonance Imaging (MRI) and X-ray Computed Tomography (CT) applications respectively. Methods, Results and Discussion: Homogenously sized (7-10 nm) TaOx NCs were made using a sol-gel procedure.4 Rapid in situ surface functionalization using a variety of modified silanes afforded hydrophobic and fluorescently labeled NCs with highest reported Ta content till date. Similarly, hydrophobic iron oxide NCs (10 nm) were prepared by the high temperature thermal decomposition of an iron precursor.5 All NC types were characterized extensively using a variety of techniques such as TEM, SEM, IR, EDS, XPS and ICP-OES. For the synthesis of CNPs, firstly spherical Cellulose Triacetate nanoparticles (CTNPs) were formed. Next, these were subjected to base-catalyzed hydrolysis to re-generate spherical CNPs. Various conditions for the base-catalyzed hydrolysis were examined, and the resulting CNPs were analyzed for complete removal of acetate functionality as well as maintaining the structural integrity of the NPs. Based on these experiments, the regeneration of CNPs from precursor CTNPs using 4% sodium hydroxide in methanol was adopted. This process provides a major improvement pertaining to the size and structural integrity of final CNPs. Further, the strategy was used to encapsulate magnetic Iron Oxide NCs within the CTNPs. The resulting Iron Oxide CTNPs (IO-CTNPs) exhibit high in vitro cell viability. All NPs produced were extensively characterized using a variety of techniques, such as DLS, SEM, TEM, IR and ICP-OES. Subsequently, the base-catalyzed regeneration of IO-CNPs from the IO-CTNPs was successfully carried out. In this work, we will further discuss their detailed characterization, extensive in vitro biocompatibility and in vivo MRI. In addition, co-encapsulation of X-ray dense TaOx NCs within the CNPs and related CT evaluation of the resulting TaOx-CNPs will also be reported. Conclusion: We report the easy to scale-up synthesis of CNPs, regenerated using a base-catalyzed hydrolysis of corresponding CTNPs. The facile encapsulation of both magnetic iron oxide NCs and CT-dense TaOx NCs within the CNPs is also reported. Further, their extensive characterization, in vitro and in vivo performance is also discussed. of Information Engineering, I-Shou University, Kaohsiung, Taiwan Objective: Successful boron neutron capture therapy (BNCT) needs sufficient and specific delivery of boron atom to malignant lesions. Gold nanoparticles (AuNPs) have been considered as an effective delivery system to carry cytotoxic payloads to the rumors. This study aims to determine the potential of human serum albumin-coated boron cage-containing gold nanoparticles (HSA-B-AuNPs) for BNCT. Materials and methods: Human serum albumin (10 mg) was dissolved in 250 μL of distilled water and added dropwise to a solution containing 20 nm AuNPs (7.2x1011 particles/1 mL). The reaction mixture was stirred for 4 h in dark. The self-synthesized boron cage (5 mg) dissolved in ethanol (3 mL) was dropwise added to the AuNPs solution and the mixture was stirred at room temperature for 30 min. After reaction, 8% glutaraldehyde (4 μL) was added to the reaction mixture and kept stirring for 24 h. After centrifugation, the final product, HSA-B-AuNPs, were obtained. The diameter and boron content of HSA-B-AuNPs were determined by dynamic light scattering (DLS) system and transmission electron microscope (TEM), and ICP-MS, respectively. For radiolabeling, the chelate, DTPA, was modified on the HSA. 111In-HSA-B-AuNPs was afforded by adding 111InCl3 solution (40 μL in 0.1 M of sodium citrate buffer, pH=5.0) in to the vial containing DTPA-HSA-B-AuNPs (200 μg). The microSPECT imaging of 111In-HSA-B-AuNPs were performed to determine its feasibility as imaging agent and its pharmacokinetics in a BT474 human breast cancer xenograft-bearing mouse model. Results: The average diameter of HSA-B-AuNPs and DTPA-HSA-B-AuNPs assessed by DLS was 178.2±63.5 and 197.8±47.7 nm, respectively. The TEM images confirmed that most of AuNPs were coated with HSA. The encapsulation efficiency of boron atom in PLGA-Gd/B-AuNPs was around 1%, which may arise from the poor solubility of boron cage. The radiolabeling efficiency and yield of 111In-HSA-B-AuNPs was 23.9±5.3% and 13.6±3.3%, respectively. After purification, the radiochemical purity reached 92.8±1.3%. In vivo microSPECT/CT demonstrated that 111In-HSA-B-AuNPs would be trapped in liver at 12 h after intravenous injection. However, the radioactivity 111In-HSA-B-AuNPs was uniformly retained in the tumor at 12 and 36 h after intratumoral injection while no noticeable accumulation in healthy tissues, suggesting these nanoparticles would not enter the blood stream to cause systemic circulation. Besides, the radioactivity of In-111 can reflect the concentration of boron atom in tumor, which was obtained from ICP-MS. Conclusion: We successfully developed reliable surrogates of boron-containing AuNPs for noninvasive determination of the boron concentration in tumor by microSPECT imaging. Additional experiments are warrant to investigate the size effect and to improve the encapsulation efficiency of boron atoms in NPs. 

Abstract Body : Objectives: The oxytocin receptor (OTR) is a G protein-coupled receptor that regulates complex social behaviour and is a promising drug target for various mental disorders [1] . To advance our understanding of OTR's distribution, mechanisms of action and therapeutic potential, we have designed sensitive and OTR-specific PET tracers for in vivo imaging. Methods: Four tracer candidates and their labelling precursors were chemically synthesized and pharmacologically characterized at the OTR. Additionally, they were characterized for their polarity and formulation stability. The best candidate resulted from the modification of position 8 of native oxytocin with a lysine residue, and its labelling with [18F]SFB (Fig.1) . The conjugation reaction was performed using microfluidic technology [2] and the product purified and formulated using standard HPLC-SPE approach. Results: The selected PET tracer, [18F]dOTK8-SFB, was produced with 20-50% RCY and >95% RP in water for injection (>20MBq/µL, >100MBq/nmol). Healthy Sprague Dawley rats were injected i.v. with the radiolabelled tracer and competition experiments were performed using the non-radioactive isotopologue. PET imaging was performed for 60 min, highlighting a clear radioactive uptake (0.4-0.6 %ID) in the pituitary gland, which is known for being an OTR rich region (Fig.2) [3] . This uptake was abated in competition studies with the identical non-radioactive compound, confirming the specificity of the tracer. Conclusions: We have designed a series of OTR-specific PET tracers for in vivo imaging. The best candidate was radiolabelled with 18F and used in rats to visualise OTR. The radioactive uptake reflected the population of OTR and was successfully displaced using the non-radioactive isotopologue. Further investigation is ongoing to employ this PET tracer for human studies. Abstract Body : Introduction:Extracellular pH (pHe) is an important biomarker for cancer cell metabolism. Although latest well developing MRI CEST(Chemical Exchange Saturation Transfer)studies have shown spatial maps of pHe of tumor region with the relatively new contrast agent, detection of acid environment of in situ lung cancer cellsremains a formidable challenge. Modifiedperfluorooctylbromide nanoparticles (PFOB NPs)could be new potential CEST contrast agents that realize in situ lung cancer acid environment detection. Here we explored the potential use of PFOB NPs as pH responsive dual 19Fand CEST MRIcontrast agent. Method:Perfluorooctylbromide nanoparticles(PFOB NPs)wassynthesized and various concentration was prepared with varying pH valuesto get ratiomatric curve.19F-MRI was acquired using the CSSI sequence to determine fluorine signal intensity.1H-MRSwas acquired using the PRESS sequence to determine the chemical shift of the contrast agent while the CEST acquisition was acquired using the RAREsequence. Bruker 9.4T BioSpec MRI Scanner with a 1H-19F Double resonance volume coil wasused. PFOB NPs(20 vol%, 300 μL) was injected for in vivo studies. Result: The contrast agent has CEST signal which is >10% varying with concentration and Saturation Power (0.8μT -3.8μT) at varying Saturation time (1s-8s). MTRasmy results shows that PFOB NPshas pH sensitivity, and get higher CEST signal at weakly basic solution. In vivo studies shows the signal of tumor region weakened at the 0.68ppm after injection of 300μl PFOB NPs due to the acidic environment of the tumor. Conclusion: At present, we tried subcutaneous tumor bearing mice first and the relevant result indicated thatmodifiedPFOB NPscould be apotential candidate CEST agent for in situ lung cancer acid environment detection.Further studies are ongoing to determine the efficacy of this Perfluorooctylbromide nanoparticlesin CEST application insituNSCLC. Abstract Body : Lung cancer is the most frequently diagnosed malignancy and the primary reason of cancer death both in male and female. Nanoparticle-based integrate platforms represent a novel strategy to cancer theranostic. However, the diagnostic sensitivity, treatment efficacy as well as targeted tissue concentration pose huge challenges for clinical translation. Herein, a tumor pH-sensitive superparamagnetic iron oxide nanoclusters (SPIONC) were synthesized, used as a novel magnetic resonance imaging (MRI) molecular probe and a sensitizer for radiation therapy (RT) on rodent models. The SPIONC maintain the superparamagnetism of the single nanoparticles, meanwhile display sufficiently T2 relaxation enhancement offered by nanoclusters under the action of external magnetic field by self-assembling. By intratracheal delivery, the SPIONC significantly enhance the MRI signal of lung cancer in situ as a T2-MRI probe acquired by Bruker 9.4T MR scanner, demonstrating early stage diagnosis of lung cancer without targeting agents. In tumor microenvironment (TME), SPIONC disintegrate into smaller nanoparticles, resulting a greater specific surface area and tissue penetration in TME weak acid condition (pH=6.5), leading more iron ions released from larger parental nanoparticles. Under a single-dose X-ray irradiation, the hydrogen peroxide generated from mitochondria converted to ROS through Fenton reaction catalyzed by iron ions, making numerous of hydroxyl radicals accumulated inside the tumor, enhanced the intratumor permeability and radiation therapy efficacy, eventually inducing damage to DNA and obstructing tumor growth. In short, SPIONC can be used as a radiotherapy sensitizer to achieve lung cancer theranostic by intratracheal delivery, providing experimental foundation for clinical translate of novel theranostic nanoprobes. Abstract Body : p.p1 {margin: 0.0px 0.0px 6.0px 0.0px; font: 11.0px 'Trebuchet MS'; color: #000000; -webkit-text-stroke: #000000} span.s1 {font-kerning: none} In vivo imaging in the shortwave infrared (SWIR, 900-1700nm) is a relatively new field of research that is undergoing fast development in biomedical sciences. Advantages of the SWIR region are the significant decrease in light scattering and the very low tissue autofluorescence resulting in improved transparency of the biological sample and a significant gain in depth penetration and image resolution. In addition, the high speed of SWIR InGaAs cameras can support real-time data generation to aid in extrapolating signal intensities into concentration values for pharmacokinetic and tissue kinetic analysis. However, there is limited number of dyes/nanoparticles that have peak emission in this spectral region. METHODS: In this study, we generated real-time kinetic data from various anatomical compartments in the mouse using a SWIR preclinical imaging system (IR VIVO™, Photon Etc, Montreal, Canada) equipped with an InGaAs (ZephIR™ 1.7) detector, after intravenous (IV) injection of various molecules with different physicochemical properties and various spectral profiles. This includes ICG (Sigma), CF770 (Biotium) and PEGylated quantum dot nanoparticles (QDots, Creative Diagnostics) in both normal and subcutaneous tumor-bearing mice. We compared different SWIR spectral windows including 850nm LP, 980nm and 1250nm to determine the best combination of brightness with tissue transparency for drug biodistribution analysis. RESULTS: In vitro results indicated that Qdots, as expected, had a peak emission around 1275nm and lower emission at 850nm and 1050nm. In contrast, ICG (1kDa) and CF770 (3kDa), at equimolar concentrations, had a peak emission at 850nm, followed by a lower peak at 980nm LP and a long tale of low emission from 1250nm LP onward. However, while CF770 can be readily conjugated to proteins and has long-term stability in aqueous solutions, ICG exhibits auto-quenching upon conjugation and is less stable losing signal over days. Biodistribution analysis of IV injected Qdots into normal and tumor-bearing mice (figure 1A) by in vivo imaging at 1250nm showed rapid accumulation in the spleen, liver and unexpectedly to bone within the first 60mins. Qdots failed to penetrate readily into the tumor parenchyma and stayed circulating primarily within the tumor vasculature. IV injection of ICG showed significantly higher signal but much lower tissue transparency at 850nm compared to 980nm and 1250nm, with the most optimal signal/transparency ratio being achieved at 980 nm. Biodistribution of IV injected ICG in normal mice demonstrated a large accumulation in the liver over 60mins, most likely due to its ability to bind serum albumin. In contrast, IV injection of CF770 demonstrated faster clearance compared to ICG via renal filtration and bladder. From the in vivo imaging kinetic data (figure 1B), PK parameters were determined using Phoenix WinNonlin using a standard curve in same matrix and the results will be validated using concentration measurements from blood and tissue sampling at different time points. CONCLUSIONS: The SWIR imaging instruments allow better preclinical PK and biodistribution analysis of therapeutic/diagnostic drugs due to the increase tissue transparency in the spectral range they operate and the high speed imaging acquisition of their InGaAs cameras. The ability to image blood and tissue kinetics in real time and convert it into meaningful drug concentration PK/tissue kinetic data in a noninvasive manner can significantly aid in the characterization, optimization and development of new therapeutic/diagnostic agents. The combination of non-invasive SWIR imaging and PK modeling can reduce experimental animal use, time and cost while obtaining comprehensive PK and biodistribution profile data sets. Abstract Body : OBJECTIVES: The purpose of this pilot study was to evaluate the optimal injection time of indocyanine green (ICG) for detecting thymic tumor and the safety and feasibility of the intraoperative detection of thymic tumor with near-infrared (NIR) fluorescence imaging by low-dose ICG injection. METHODS: Thirty seven consecutive patients who were scheduled to undergo resection of thymic tumors were enrolled in this study. ICG (1, 2 and 5 mg/kg) was randomly administered (intravenous injection) at 3~24 hours before surgery and the retrieved surgical specimens were examined for fluorescence signal of tumor to normal ratio (TNR) by using NIR fluorescence imaging system on a back table. We analyzed the fluorescence intensity, pathology, size of the thymic tumors. RESULTS: There were no adverse effects related to the systemic injection of ICG, and no cases of major morbidity or mortality after surgery were noted. Thymic tumors were resected through thoracoscopic surgery in 30 patients, robot surgery in 5, and sternotomy in 2. Thymoma can be detected intraoperatively by intravenous injection of ICG 3 to 24 hours prior to surgery. The TNR of thymoma was highest at 3 hours with than other times. Based on TNR, there was no significant difference between 1, 2 and 5 mg/kg ICG injection for thymoma detection. Compared with thymoma, the intensity of the fluorescent signal detected in the thymic cyst showed lower than normal tissue. The mean TNR was 0.4 ± 0.1 in 10 patients with thymic cyst and 3.3 ± 0.8 in 27 with thymoma (1 with thymoma A, 9 with thymoma AB, 5 with thymoma B1, 3 with thymoma B2, 3 with thymoma B3 and 6 with thymoma C). The correlation between TNR with WHO pathologic classification (p>0.05) or tumor size (p>0.05) were not significant. CONCLUSIONS: NIR fluorescence imaging could successfully identify thymic tumors intraoperatively after the systemic injection of low dose ICG. Further study should be conducted to define intraoperatively the exact resection margin of thymic tumor. Abstract Body : Rationale: The ability to image tumors at depth with high selectivity and specificity is significant challenge in the field of biomedical optical imaging. Owing to their unique "fingerprint" like spectra, surface enhanced resonant Raman scattering (SERRS) nanoparticles (NPs) can be employed as image contrast agents and in addition, can specifically target cells in vivo.1 However, whilst the detection of SERRS NPs is extremely sensitive and specific, conventional approaches involving Raman spectroscopy are limited in their inability to probe through tissue depths of more than a few millimeters. Here, the use of spatially offset Raman spectroscopy (SORS),2 is combined with that of SERRS in a technique known as surface enhanced spatially offset resonance Raman spectroscopy (SESORRS),3 to image glioblastoma multiforme (GBM) tumors in vivo in mice through the intact skull. Methods: An in-house SORS imaging system was built and used for all imaging experiments. A PTFE-skull-tissue phantom was used to optimize the system and carry out proof of concept SORS imaging. Imaging of GBM in mice was achieved by using gold nanostars functionalized with a resonant Raman reporter to create SERRS nanostars. Such contrast agents were then encapsulated in a thin silica shell and functionalized with cyclic-RGDyK to generate integrin-targeting SERRS nanostars. Previous work has demonstrated the successful targeting of GBM using cyclic-RGD as a targeting moiety.4 Non-invasive in vivo SESORRS imaging of the integrin targeted nanostars was then performed in an RCAS/TV-a mouse model of GBM. Conventional Raman imaging was used as a direct comparison. Results: GBMs were imaged in vivo using SESORRS in mice (n = 5) and confirmed using MRI and histopathology. The results demonstrate the ability to acquire clear and distinct Raman spectra from deep-seated GBM in mice in vivo through the skull using the SORS approach. SESORRS images generated using classical least squares fit were used to demonstrate successful delineation of the tumor region as confirmed via MRI and histology. Conventional Raman (CR) imaging was carried out as a control, however in comparison to SESORRS, failed to outline the whole tumour region through the intact skull. Furthermore, fluorescent interferences typically associated with CR measurements in biological tissue were largely suppressed through the application of the SORS technique. Conclusion: These results demonstrate the first use of SESO(R)RS for in vivo imaging applications to enable the successful detection, and more importantly profiling, of GBM in vivo. It is reasoned that the use of SESO(R)RS for in vivo imaging will not be exclusive to brain cancer imaging but also to the monitoring of a wide range of diseases that require deep tissue penetration, such as internal organs in mice or relatively peripheral organs in humans ID: LB160 A Fluorogenic Trehalose Probe for Imaging Phagocytosed Mycobacterium Tuberculosis Tingting Dai, Stanford University, tingd@stanford.edu

Abstract Body : Tuberculosis (TB) is caused by the slow growing airborne pathogen Mycoabcterium tuberculosis (Mtb), which colonize resident macrophages as their natural habitat [1] . Bacterial mycomembrane plays an important role in the survival of Mtb, which has brought lots of interests as a drug target. Fluorescein-containing trehalose probes have been tested for labeling phagocytosed Mtb inside macrophages, however the extremely high working concentration (≥200 μM) brought nonspecific background and potential toxicity [2] . Here, a small molecular probe (L-TG-Tre) that enables labeling of mycobacteria in live macrophage at a concentration as low as 2 μM was described. Importantly, only activated by β-lactamase, BlaC, will the probe produce fluorescence signal and then be retained and trapped via trehalose mycolyltransesterase enzymes, Ag85s, mediated metobolic processes. This new fluorogenic probe was validated in different bacteria species and fungi. Super resolution structured illumination microscope further revealed selective labeling of poles and wall structure of growing Mtb within infected macrophages by L-TG-Tre. In comparison, dead bacilli could not be labeled under similar conditions. These studies suggest that cephalosporin may serve as a unique function moiety to increase the uptake of unnatural trehalose and potentially many other small molecules. A specific and sensitive detection of growing Mtb either extracellularly or inside macrophages may find many potential applications in pathogenesis study, drug screening, and most importantly, for detection of TB.

Total-Body Dynamic PET Imaging with 100-ms Temporal Resolution Xuezhu Zhang, University of California, Davis, zhang@ucdavis.edu

Abstract Body : PET is the most sensitive molecular imaging modality for tracing biomedical processes in vivo. However, compared to other imaging modalities, the limited sensitivity of earlier PET scanners results in data with low signal-to-noise-ratio (SNR). To overcome this limitation, the EXPLORER consortium has built the world's first 194-cm long total-body PET/CT scanner (uEXPLORER) [1] . With the extremely high sensitivity of the uEXPLORER, we aim to push the limit of the temporal resolution of dynamic PET by applying an innovative image reconstruction method to uEXPLORER data. Ultra-high temporal resolution PET has the potential to reduce motion blurring artifacts, improve the effective spatial resolution of reconstructed images, directly capture respiratory, cardiac and gross body motion, and facilitate new approaches to kinetic modeling. In this work we demonstrate 100 millisecond dynamic PET imaging using a human FDG scan on the uEXPLORER. We developed a methodology to perform ultra-high temporal resolution dynamic PET imaging by applying the kernel-regularized reconstruction paradigm to the uEXPLORER data [2] . The complete dynamic PET data are used to construct a kernel matrix, which is in turn used to regularize reconstructed PET images in each temporal frame. To demonstrate the effectiveness of the method, we tested it on a one-hour total-body dynamic PET scan acquired during and following an intravenous injection of 256 MBq of 18F-FDG. We divided the dataset into 100-millisecond temporal frames and analyzed the data in the first and last minute of the scan, which show the initial tracer transit in the vasculature and tracer uptake at the equilibrium state, respectively. Dynamic data were reconstructed using the kernel-regularized algorithm with quantitative corrections. The reconstructed images show good quality. Individual heart beats are clearly visible in the 100-millisecond temporal frames. The extracted time activity curves from major vascular regions-of-interest (ROIs) show the dynamic change of tracer distribution in the left ventricle and major arteries during cardiac contraction and expansion. Furthermore, the cardiac motion signal was extracted directly from the reconstructed images and used to perform cardiac gating. Reconstructed gated images show excellent delineation of the myocardium at different cardiac phases, which clearly demonstrates the capability of freezing subject motion without using any external monitor device (e.g. ECG, breathing belt, optical markers, etc.). In conclusion, we have successfully developed a method to perform total-body dynamic PET imaging with ultra-high temporal resolution. Our method showed that the kernel-regularized reconstruction with a total-body PET scanner can achieve superior image quality for motion-frozen quantitative studies. Our method has applications in the studies of blood flow and transport, as well as motion-frozen (heart beating / breathing) monitoring of cardiovascular and cerebrovascular function and respiratory system function. Abstract Body : Introduction Microbubbles (MBs) are clinical ultrasound contrast agents, typically composed of fluorocarbon gas cores stabilized with phospholipid shells. Beyond their applications in medical imaging, MBs have shown promise in ultrasound-mediated drug delivery. It is well-established that the insonation of intravenous drugs in the presence of MBs can enhance delivery through vascular permeabilization [1, 2] , which is currently being investigated in several clinical trials to reduce the size of tumours and metastic lesions [3, 4, 5] . However, these procedures do not prevent the systemic toxicity that limits the injected dosage for chemotherapeutics. To address this, hydrophobic drugs have been incorporated within a MB shell [6, 7] . However, this strategy still excludes an important subset of pre-existing, effective, water-soluble chemotherapeutics. Previously, hydrophilic drug-loaded ultrasound-sensitive agents such as drug nanoparticles conjugated to MB shells [7, 8, 9] and echogenic liposomes (ELIPs) [10, 11, 12] have been investigated. However, drug-conjugated MBs typically use biotinylation, which raises immunological concerns linked to the activation of the complement system within humans and mice [13, 14] , while only a small percentage (~20-35%) of ELIP populations contains gas and are capable of undergoing ultrasound-triggered release and enhanced delivery [11] . To address these shortcomings, a drug-loaded MB system is proposed that features a shell comprised of phospholipids used within commercial microbubble formulations and filled with a perfluorocarbon gas, making the majority of its population ultrasound responsive. In this proof-of-concept study, a hydrophilic cargo, fluorescent calcein dye, is encapsulated as surfactant-stabilized nanoparticles enclosed within the phospholipid-shelled MBs. Methods Calcein nanoemulsions stabilized with Span 80 and Tween 80 were encapsulated in decafluorobutane gas-filled, phospholipid-shelled (DPPC/DPPE-PEG5K) MBs to form calcein-loaded MBs. Calcein was chosen as a surrogate cargo because of its fluorescent property, allowing for easy confirmation of co-localization with the microbubble structure. To observe this, calcein-loaded MBs were diluted 2× in 0.9% saline and examined with a fluorescent microscope (Eclipse Ci, Nikon) under brightfield and FITC (EX:465-495/DM:505/BA:515-555).To confirm the echogenicity of loaded-MBs, a 200× diluted sample was flowed through a 2wt% agar tube phantom (1.14mm diameter) and imaged with a L9-3 transducer (nominal frequency -5 MHz) on an ultrasound system (IU22, Philips). The loaded-MBs were also sized with a Coulter counter (Multisizer 4e, Beckman Coulter) before and after sonication to verify that the agent was capable of ultrasound-induced cavitation. Results Encapsulating hydrophilic drugs in ultrasound-responsive carriers can be challenging because of the insolubility of these drugs within the hydrophobic surfactant tails and fluorinated gas core of the bubbles. Under fluorescent microscope, the calcein-loaded MBs demonstrated a ring of fluorescence within/near the bubble shell confirming co-localization of calcein nanoparticles within the MB structure ( Figure A) . The size distribution of bubbles (with a 3.9µm mean diameter, Figure C ) was measured to be within the appropriate size range ( Conclusion Proof-of-concept hydrophilic dye-loaded MBs were synthesized for the development of ultrasound-responsive carriers for local drug delivery to tumours. The agent was shown to demonstrate cargo co-localization, echogenicity and ultrasound-induced cavitation.

Alexander Klibanov, University of Virginia, sklib1@gmail.com

Abstract Body : Targeted microbubbles (MB) for molecular ultrasound imaging are often designed with protein ligands, e.g., antibodies (AB) against vascular endothelium biomarkers of disease. Protein is routinely attached to the shell of pre-formed MB by covalent coupling, e.g., to primary aminogroups (1), thiols (2), or via click-chemistry (3) . These methods are either chemically cumbersome, or offer low coupling yield, which is unacceptable for expensive protein targeting ligands. An alternative is to use biotinylated bubbles (4) and attach biotinylated proteins to pre-formulated microbubbles via a streptavidin link. The latter approach is feasible for early testing in animal models, but unsuitable for clinical translation, as streptavidin is a foreign protein, and could induce immune response in humans. To avoid these difficulties, and provide a simple and effective translatable microbubble molecular imaging formulation with protein ligands, we first synthesize the chemical conjugate of the anchor lipid with the protein, add MB shell lipids with cosurfactant, and prepare MBs using a clinical amalgamator, avoiding sonication and high temperature, that are known to denature proteins. Lipid-modified protein is transferred to gas-water interface and anchored to the MB lipid monolayer shell. We tested MBs with two proteins, anti-VCAM-1 IgG AB and tomato lectin (TL), a 71 KDa glycoprotein.

Protein was incubated with NHS-PEG-DSPE aqueous micellar solution, to couple the primary amine functional groups of protein with the active ester of carboxy-PEG-DSPE. Reaction ratio was selected so that most protein molecules would have a PEG-DSPE residue attached, yet to avoid protein over-modification and inactivation. Reaction mixture was subjected to quick 10 KDa centrifugal ultrafiltration (to remove azide and other small molecules), added to the vials containing aqueous saline -propylene glycol (cosurfactant) media, DSPC and PEG stearate (MB shell components), and sealed under decafluorobutane atmosphere. Precursor vials could be stored under refrigeration. MBs were prepared by 45 sec amalgamation (5), followed by vial inversion and 15 min incubation to allow flotation of large microbubbles to the top. Desired microbubbles were collected from the bottom of the inverted vial using short needle. Coulter counter was used to assess particle size and concentration. Quantification of free and bound protein was performed with fluorescence (FITC-anti-VCAM-1 AB or DyLight 488-TL) following centrifugal MB flotation (200 g, 10 min). In vivo targeting was performed for murine colon adenocarcinoma (MC38 cells, a generous gift of J. Schlom, NCI) subcutaneous tumor model in a hind limb of C57BL/6 mice. Sequoia C512 (15L8 probe, 7 MHz, CPS mode, MI 0.2) was used for ultrasound imaging; targeting was assessed five and ten minutes following intravenous MB bolus. Microbubble preparation was highly efficient, with several billion particles per ml generated. Short normal gravity flotation allowed removal of >99% MBs that exceeded 5um diameter. Typically, attachment of 20,000 TL and ~5,000 AB molecules per MB was achieved, depending on the initial protein and DSPC lipid concentrations, with ~45% and 68% transfer of protein to MB shell, respectively. Targeting of anti-VCAM-1-MB in murine tumor vasculature was observed by contrast ultrasound imaging, compared with control contralateral leg muscle (p 0.6). As centrifugal wash purification of MB from residual free antibody is unnecessary, the simplified formulation as described here is suitable for practical use translation. Likewise, TL-PEG-DSPE-microbubbles, but not control bubbles lacking TL, were selectively accumulating in the vasculature of the target tumor tissue, but not in the contralateral leg muscle (p < 0 .001). In conclusion, we describe a simplified translatable procedure for rapid and efficient preparation of protein-decorated microbubbles, contrast agents for ultrasound molecular imaging. Abstract Body : Introduction Hemorheology is the study of flow properties of blood and the components (plasma, red blood cells, white blood cells and platelets). Blood viscosity is determined by plasma viscosity, hematocrit and mechanical behaviour of red blood cells leading to the mechanics of these cells being the major determinant of flow properties of blood [1] . However, blood does not behave like tissue on MR scan due to physiologic feature of maximizing oxygen transport. There is a close correlation between the rheologic and MR properties of protein molecules. The high concentration of intracellular hemoglobin is actually at a physiologic limit. Despite this high concentration, the lack of organelles and other macromolecular features of nucleated cells ensure blood's low viscosity and apparent magnetization transfer rates (RAMT) which was achieved with the use of magnetization transfer studies [2] . This demonstrates the importance of developing a model for mapping the viscosity-dependent NMR signal for hemorheology. Methods Above the normal intracellular hemoglobin concentration, viscosity rises very steeply and thus interferes with blood circulation. In fact, increased viscosity has been found to be lead to decreased mobility of the hemoglobin molecule; manifesting in increased RAMT and T2 relaxation rate [2] . Hence, we shall develop a viscosity-dependent transverse magnetization due to blood spin dynamics based on the time-independent Bloch NMR flow equation [3] . If the spin dynamics is within a rotating frame, then resonance condition exists at Larmor frequency [3] . From the NMR Bloch flow equations (My is the transverse magnetization), equation of motion of the spins moving with a variable velocity v(x) is given by eqn (1) . If the RF field B1(x) is applied such that My is sampled at maximum magnitude, M0 ≈ 0; since rheology is concerned with fluids with variable viscosity and if we then take v(x) as the mean flow velocity with x being the characteristic length of the blood vessel, assuming eqn (2) is true and given that δ = pulse time, Re = Reynolds number, μ = dynamic viscosity, β = dimensionless parameter while ρ = density of blood. Eqn (2) becomes eqn (3) whose solution is given in eqn (4) (where β = RAMTT2 and C1 is a constant). Results We have applied the result obtained in equation (5) to unclotted blood samples with intact red blood cells. The NMR properties of these samples as related to blood rheology have been measured in earlier study [2] and are presented in Table 1 . We developed a Mathematica (version 9) computer code for mapping the transverse magnetization of the components and the associated flow velocity. Using experimental ranges of viscosity in human blood flow [4] and Table  1 , we have the maps in Figure 1 for low Reynolds number (Re = 10).

Discussion and Conclusion We have developed a model in which My is obtained in terms of blood rheological parameters. With this model, different flow regimes could be obtained for various levels of viscosity and hence, different shear forces. This is demonstrated in the unique patterns shown in Figure 1 . In conclusion, ease with which images of rheological flow could be obtained is the most interesting part of this study. The influence of turbulent flows can be easily obtained by simply changing the values of the Reynolds number in the computer program.

The Role of Copper in Neurodegenerative Disease

Imaging Copper Metabolism Imbalance in Atp7b−/− Knockout Mouse Model of Wilson's Disease with PET-CT and Orally Administered 64CuCl2

A Responsive MRI Contrast Agent for Detection of Excess copper(II) in the Liver in Vivo

A responsive MRI contrast agent for detection of excess copper(II) in the liver in vivo

Analysis of progress and challenges for various patterns of c-MET-targeted molecular imaging: a systematic review

References: 1. Vogelzang

Evaluation of Immune-Related Response Criteria and RECIST v1.1 in Patients With Advanced Melanoma Treated With Pembrolizumab

Jenna Last Name: Steiner Email: jenna.steiner@ucdenver

Subsurface Probing in Diffusely Scattering Media Using Spatially Offset Raman Spectroscopy

High Precision Imaging of Microscopic Spread of Glioblastoma with a Targeted Ultrasensitive SERRS Molecular Imaging Probe

Surface-enhanced Raman spectroscopy: concepts and chemical applications

A Macrophage Invasion Mechanism of Pathogenic Mycobacteria

Uptake of Unnatural Trehalose Analogs as a Reporter for Mycobacterium Tuberculosis

RWTH Aachen University. Identifier NCT03385200

University of Tours

St. Olavs Hospital. Identifier NCT03477019

Thiol-Reactive Bifunctional Chelators for the Creation of Sire-Selective Modified Radioimmunoconjugates with Improved Stability

GUILLAUME Last Name: DEWAELE-LE ROI Email: dewaeleg@mskcc.org Organization: Memorial Sloan Kettering Cancer Center Country

Method: FAP-α monoclonal antibody and its IgG isotype control were conjugated with a NIR phthalocyanine dye, IR700, to form FAP-α-IR700 and IgG-IR700. SDS-PAGE and UV spectroscopy were performed to determine the purity and composition of the conjugates. FAP-α overexpression was achieved by transducing MDA-MB-231 and HT-1080 number of HT-1080 or HT-1080-FAP cells were inoculated bilaterally in the flank. 50 μg of FAP-α-IR700 or IgG-IR700 was injected i.v., and fluorescence images of IR700 in mice were obtained over a 24h period (n = 3 per group). Fluorescence images were acquired with a Li-Cor Pearl® Impulse under identical experimental conditions. At 24h post injection, mice were euthanized, and tumors were isolated for imaging. For PIT, MDA-MB-231 tumor bearing mice received two i

Results and Discussion: Figure 1a and 1c demonstrate the preferential accumulation of FAP-α-IR700 in FAP-α-overexpressing 231-FAP and HT-1080confirm the ability of FAP-α-IR700 to target and eliminate FAP-α-overexpressing cell populations, providing novel opportunities to selectively deplete FAP-α high

Rational design of a chalcogenopyrylium-based surface-enhanced resonance Raman scattering nanoprobe with attomolar sensitivity

First Name: Moritz Last Name: Kircher Email: moritzkircher@gmail.com Organization: Dana-Farber Cancer Institute/Harvard Medical School Country

Panitumumab for Injection: Our experience with implementing clinical radiochemistry for human studies

Oncology Abstract Body : INTRODUCTION 89Zr-Radiolabeled Panitumumab

Automatic Needle Segmentation and Localization in MRI with 3D Convolutional Neural Networks: Application to MRI-targeted Prostate Biopsy

An in vitro study of magnetization transfer and relaxation rates of hematoma

Distribution of blood viscosity values and biochemical correlates in healthy adults

2. Marx, J. Inflammation and cancer: the link grows stronger: research into a long-suspected association between chronic inflammation and cancer reveals how the immune system may be abetting tumors

United States References: 1-a) Cerenkov, P. A. Comp Rendus DokladyAkademiiNauk SSSR, 1934, 2, 451; b) Jelley, J. V. Cerenkov radiation and its applications

Optical imaging of Cerenkov light generation from positron-emitting radiotracers

Molecular Optical Imaging with Radioactive Probes

Piwnica-Worms, D. Cerenkov radiation energy transfer (CRET) imaging: a novel method for optical imaging of PET isotopes in biological systems. PLoSone 2010, 5, e13300; b) Boschi, F.; Spinelli, A. Quantum dots excitation using pure beta minus radioisotopes emitting Cerenkov radiation

Automated Quantitative Plaque Burden from Coronary CT Angiography Noninvasively Predicts Hemodynamic Significance by using Fractional Flow Reserve in Intermediate Coronary Lesions

Modular low-light microscope for imaging cellular bioluminescence and radioluminescence

Abstract Body : Over the last thirty years, radioimmunoconjugates have become increasingly important tools in nuclear medicine. Yet these sophisticated constructs are still synthesized in a remarkably imprecise way: the indiscriminate bioconjugation of chelators to lysines within the macromolecule. This random approach to bioconjugation not only produces heterogeneous mixtures of immunoconjugates but also results in radiotracers with suboptimal in vivo performance. [1] To circumvent these problems, several "site-specific" bioconjugation strategies have been developed. The most popular is the use of maleimide-based bifunctional probes for conjugations to cysteine residues; however, this thioether-forming reaction is reversible in vivo, a process which can result in the release of the radioactive cargo, thereby reducing the target-to-background activity concentration ratios and increasing radiation doses to healthy tissues. In an effort to address this problem ¾ and inspired by the work of Barbas, et al.[2] ¾ PODS has been developed in our laboratory: a phenyloxadiazolyl methyl sulfone-based reagent that efficiently and irreversibly forms covalent linkages with thiols. Indeed, just last year, it has been reported that 89Zr-DFO-and 177Lu-CHX-A²-DTPA-labeled radioimmunoconjugates created using a PODS-based bioconjugation approach exhibited superior stability and in vivo performance compared to analogous constructs synthesized using a maleimide-based bioconjugation strategy. [1] In the work at hand, these initial investigations have been expanded through the development of a novel PODS-based bifunctional chelator ¾ PODS-DOTA ¾ for the synthesis of 177Lu-, 89Zr-, and 225Ac-labeled radioimmunoconjugates. [3] PODS-DOTA was synthesized in five steps and 36% overall yield from via the construction of PODS and its subsequent coupling with an isothiocyanate-bearing variant of DOTA (p-SCN-Bn-DOTA). Next, we performed the bioconjugation of PODS-DOTA to a model antibody (the HER2-targeting mAb trastuzumab), ultimately producing an immunoconjugate with ~2 DOTA/mAb. In parallel, several alternative bioconjugation protocols were also explored using a fluorophore-modified variant of PODS (PODS-FITC) in an effort to streamline the antibody modification process and enable the re-oxidation of the disulfide bridges within the antibody after the attachment of the bifunctional probe. Along these lines, the selective reduction of the antibody's disulfide bonds was achieved via partial reduction with tris(2-carboxyethyl)phosphine (TCEP) followed by reoxidation using dehydroascorbic acid (dhAA) and conjugation to PODS-FITC. Using this approach, the DOL of the immunoconjguate increased with the number of equivalents of TCEP, the number of equivalents of PODS-FITC, the reaction time, and the reaction temperature. Ultimately, the incubation of the antibody for 16 hours at 37°C with 10 equivalent of TCEP and 20 equivalent of PODS-FITC produced an 'optimized' immunoconjugate without free thiols and with a DOL of 1.8 ± 0.1 FITC/mAb. These reaction conditions were then employed with PODS-DOTA, and preliminary radiolabeling experiments and stability experiment with 177Lu have proven very promising, suggesting that PODS-DOTA ¾ like its progenitors ¾ will facilitate the construction of radioimmunoconjugates with high immunoreactivity, high stability, high specificity activity, and superior in vivo behavior. In the end, we believe that this project could have a transformative impact on the field of bioconjugation, enabling the synthesis of safer and more effective site-specifically modified radioimmunoconjugates for both the laboratory and the clinic.

Development of a Molecularly-Targeted Intraoperative Imaging Agent for High-Grade Serous Ovarian Cancer Kimberly Fung, Hunter College, Fungkimberly8@gmail.com

Abstract Body : Introduction: Ovarian cancer is the fifth leading cause of cancer deaths among women, accounting for more deaths than any other cancer of the female reproductive system. The main treatment for this disease is cytoreductive surgery (CRS), but it is difficult to completely remove the cancer, as surgeons rely solely on the visual and manual assessment of tumor tissue. The goal of our investigation is to develop a near-infrared fluorescence (NIRF) imaging agent for the intraoperative imaging of high-grade serous ovarian cancer (HGSOC) during CRS. Elevated levels of cancer antigen 125 (CA125) have proven to be a useful biomarker of HGSOC, and the CA125-targeting antibody B43.13 has shown potential as a platform for immunoPET imaging in murine models of ovarian cancer.1,2 Herein, we have developed a NIRF imaging agent based on the B43.13 antibody. Methods: The B43.13 antibody was site-specifically modified with the near-infrared dye IR800CW using a click chemistry-based chemoenzymatic approach developed in our laboratory.3,4,5 In brief, this methodology is predicated on three steps: (1) the use of the EndoS enzyme to cleave the heavy chain glycans; (2) the harnessing of a mutant, promiscuous β-galactosyltransferase [GalT-(Y289L)] to transfer an azide-modified galactose residue (GalNaz) onto the truncated glycans; and (3) the click ligation between a dibenzocyclooctyne (DBCO)-modified IR800CW dye and the azide-modified antibody. It has been demonstrated that this strategy produces more homogenous and well-defined constructs with improved in vivo performance compared to randomly labeled analogues. The site-specifically modified construct-ssB43.13-IR800 (50 µg; 0.33 nmol) was administered intravenously to Nod-Scid-Gamma (NSG) mice bearing OVCAR3 xenografts 72 h prior to NIRF imaging. As a negative control, ssmIgG-IR800 (50 µg; 0.33 nmol)-was also administered intravenously to an additional cohort of tumor-bearing mice. NIRF imaging was performed on live mice (n = 4 per construct), and upon necropsy, various tissues were harvested from the mice and imaged. Results: The site-specifically modified B43.13 antibody-ssB43.13-IR800-was synthesized, producing an immunoconjugate with an average of 0.74 ± 0.02 dyes per antibody. In NSG mice with subcutaneous OVCAR3 tumors, higher levels of tumoral uptake were observed with ssB43.13-IR800 (Fig. 1A ) compared to the non-specific uptake of the negative control, ssmIgG-IR800 (Fig. 1B) . This observation was supported by ex vivo imaging, which likewise showed higher uptake in the tumor for ssB43.13-IR800 compared to the negative control ( Fig. 1) . Curiously, however, high non-specific uptake of ssB43.13-IR800 was seen in the large intestine, which may be a result of the mouse strain. Conclusions: In this study, we have demonstrated that ssB43.13-IR800 can be used to image CA125-expressing high-grade serous ovarian cancer tumors. NIRF imaging studies in subcutaneous, orthotopic, and patient-derived xenograft mouse models of HGSOC are forthcoming. A non-site-specifically modified construct-nssB43.13-IR800-will also be included in the study to allow for a comparison between the non-site-specific and site-specific methods of conjugation. Acknowledgments: We are very grateful to the MSKCC Department of Radiology and Department of Surgery for their support. We also thank the MSKCC Antitumor Assessment Core and Small Animal Imaging Core Facility. Abstract Body : Rationale: Although "Surface-enhanced Raman scattering" (SERS) nanoparticles (NPs) offer high sensitivity and specificity in cancer imaging [1] [2] [3] [4] [5] [6] , the technique currently requires a time-intensive point-by-point acquisition of Raman spectra, precluding the real-time image acquisition desired for clinical applications. Therefore, an unmet need exists for a multimodal NP that combines fast NIR fluorescence-based intraoperative imaging with the highly specific Raman-based cross-validation and clean margin confirmation (Fluorescence-Raman NPs, FRNPs). Further, additional therapeutic capabilities would enable the NPs to mitigate any residual malignancies after surgery. Methods: We conducted five layers of design refinements: (i) Raman reporter selection, (ii) DNA-shell selection, (iii) plasmonic core selection, (iv) plasmonic shell selection, (v) surface passivation (Fig. 1 ). After these optimization steps, both Raman and fluorescence signals were enhanced by several orders of magnitude and the FRNPs could be detected in a low fM regime. The optimized structures consisted of a 40nm by 10 nm gold nanorod (AuNR) core, with three consecutive layers: DylightTM780 labeled phosphorothioate backbone modified A6 ((PS)A6) DNA, a thin silver shell, and a 15-20 nm silica shell (Fig. 2) . To verify whether the fluorophore attached to A6 sequences adsorbed at very close proximity to the Au surface, we carried out extensive molecular dynamics simulations. After in vitro cell studies demonstrated excellent serum and photo-stability and nonexistent cytotoxicity profile, we proceeded to in vivo dual-mode fluorescence and Raman cancer imaging and photothermal therapy using OFRNPs, administering 200 µl of 10 nM PEGylated OFRNPs via tail vein injection into ovarian cancer xenograft (n=5) and glioblastoma (RCAS/tv-a) (n=5 each) mouse tumor models. The accuracy of the imaging and photothermal (PTT) data was confirmed via histological correlation. Results: We observed marked fluorescence signal in tumors compared to normal tissue as early as 3 hours post-injection. Whole tumors could be sequentially and completely resected using fluorescence imaging as visual guidance. We verified that the tumor tissue had the Raman signature of DylightTM780 during Raman scanning while the normal tissue exhibited no such signature. Ex vivo fluorescence and Raman images showed excellent correlation with histology (Fig. 3) . After irradiating tumors (PTT) with a 680 nm laser (power density of 3 W/cm2) for 2 minutes we observed a significant increase in temperature in the tumor area of the OFRNP injected mice (Fig. 3) . We monitored the tumor growth and other vital symptoms for 2 weeks post-injection. The average tumor sizes of the OFRNP injected group with photothermal therapy shrunk by about one-third of the original size. However, the other three groups exhibited tumor growth by a factor of 3 over 10 days (Fig. 3) . Histological examination revealed that the OFRNP-PTT tumors exhibited extensive necrosis as opposed to the tumors from other treatment groups (Fig.  3) . Conclusion: We have carried out a rational design optimization enabled by DNA to work out the optimal design construct for fluorescence and Raman imaging. OFRNPs showed excellent imaging capabilities in two different cancer models. We also demonstrated higly efficient PTT in ovarian cancer xenograft mouse models. We foresee that the current OFRNPs design would open up the possibility of NIR fluorescence and Raman-based cancer imaging and surgery given that Raman and fluorescence endoscopes are already in clinical trials or clinically approved, respectively, and could potentially be integrated.

Novel LDLR-targeting radiotracers for imaging glioblastoma (GBM) Izabela Tworowska, Radiomedix inc., itworowska@radiomedix.com

Abstract Body : Introduction. Glioblastoma (GBM) is an aggressive malignant brain tumor associated with poor overall survival of patients and the average life expectancy of 12-15 months from diagnosis. One of the factors that limit the efficacy of the drugs, especially targeting the primary brain tumor, is permeability of the blood-brain barrier (BBB). The low-density lipoprotein receptor (LDLR) expressed at the BBB mediates the transport of endogenous ligands through the BBB, a process referred to as receptor-mediated transcytosis. VECT-HORUS (VH) has identified and chemically optimized a family of peptide-vectors targeting both the human and murine LDLR and able i) to cross the BBB and ii) to target tumors such as glioblastoma that express high levels of the LDLR. The objective of this study was to determine the LDLR targeting properties of 68Ga/177Lu-radiolabeled peptide vectors developed based on the VH proprietary platform using a glioblastoma model that expresses the human LDLR (hLDLR) at high levels. Methods. The LDLR targeted DOTA-conjugates (VH-DO31, VH-DO33) and NODAGA conjugate (VH-NO31), (10-30ug, VECT-HORUS SAS, France) were labeled with 68Ga (1.5mCi) eluted from 68Ge/68Ga generator (100mCi, ITG GmBH) or with 177Lu n.c.a (1mCi, ITG GmBH, Germany). The U87MG cell line was shown to express high levels of the hLDLR. The LDLR targeting properties of these conjugates were thus determined in vitro in U87MG cellular uptake studies, as well as in vivo in U87MG xenografted mice. The PET/CT images of U87MG xenograft generated in athymic nude mice (10 weeks, n=3) were acquired using G4 PET/Xray camera (Sofie Biosciences; 10min/scan) at 1h, 2h, 3h, and 4h post-injection. Results. All 68Ga/177Lu-labeled conjugates were synthesized with radiochemical purity higher than 91 % as determined by radio-HPLC. Radiolytic stability of agents was increased by the addition of scavenger and C18 ethanol purification of the final products. 177Lu-VH-DO33 showed the highest retention of the agent in U87MG cell line at 1h (13.88± 1.6 %ID/mg) and 21h incubation time (8.7± 4 %ID/mg) compared to 177Lu-VH-DO31 (8.28 ±6.2 %ID/mg). The microPET imaging studies showed rapid accumulation and retention of all VH derivatives in the tumor as monitored up to 4h post-injection. All agents were eliminated through bladder and kidneys. There was no accumulation of agents in the bone marrow. The image-based biodistribution studies of 68Ga-VH-DO31, 68Ga-VH-DO33 and 68Ga-VH-NO31 showed that the tumor to muscle ratios (SUV ratio) after 30 min post-injection were 4.12, 5.07 and 3.88, respectively and remained at the same levels up to 3h post-injection. The SUV ratios of the tumor to kidneys were as follows: 68Ga-VH-DO31 (0.46), 68Ga-VH-DO33 (0.84) and 68Ga-VH-NO31 (0.46) confirming renal elimination of the agents. Conclusions. VH derivatives showed favorable hLDLR targeting properties in vitro and in vivo in U87MG xenografted mice models. These preliminary results suggest that hLDLR may serve as a target for the development of radiodiagnostics and radiotherapeutic drugs for glioblastoma. Abstract Body : Rapid automatic needle segmentation is a vital step towards improving the safety and efficiency of image-guided percutaneous procedures. Current techniques in the clinic employ intermittent imaging and visual inspection to assess the placement of the needle relative to a planned trajectory. Machine learning strategies such as neural networks offer the potential for fast, robust, automated localization of a needle in the image frame. This work reports on retrospective segmentation of T2-weighted clinical prostate images of patients undergoing high, with multiple needles in view, using a convolutional neural network. A 3D convolutional neural network, adapted from Mehrtash et al, was built using the Keras API with a tensorflow backend in Python3 (1). A sequential model was used with an Adam optimizer and binary cross-entropy loss function. The network consists of 14 3D convolution layers, 3 max pooling layers, 3 upsampling layers and 1 dense layer. A SeLU activation function was applied to each convolution layer and was followed by a dropout layer with rate 0.1. Images from 12 patients undergoing HDR prostate brachytherapy (1340 images i.e. 67 image volumes; slice thickness=3mm; matrix=256x256; no. of slices=20) were used for training and testing. Needles were manually segmented using 3D Slicer to obtain 3D binary label map masks. The network was trained on data from 10 patients (55 volumes) for 10 epochs with a batch size of 4 image volumes and a validation split of 0.2. The remaining 2 patients were used to test the performance of the model. The model achieved a training and validation binary cross entropy of 1.92 and 1.88 respectively after 10 epochs. Each training step took 34-52s leading to an overall training time of 4hrs and 22mins. The decreasing loss suggested that training for more epochs would be beneficial. Model performance may be improved by incorporating time-series image volumes along with known insertion points and a larger training sample size. Rapid localization of needles within an MR volume will facilitate the integration of robotic systems with imaging feedback and ensure the safety of the patient during operation. The developed algorithm may be applied to MR images acquired in near real-time to update and potentially predict needle deflection during insertion through inhomogeneous soft tissue for robotic control.

Hunter Snoderly, West Virginia University, htsnoderly@mix.wvu.edu

Abstract Body : Background: Though initially considered to be only biomarkers, current literature suggests that cancer associated extracellular vesicles (EVs) actively contribute to tumor progression; indeed, elevated EV blood concentration is known to be associated with lower treatment success and survival rates. However, the means by which EVs increase mortality remains unclear. One possibility is that tumor-associated EVs provoke a pro-thrombotic and pro-inflammatory state by activating platelets, promoting neutrophil recruitment, and causing extrusion of neutrophil extracellular traps (NETs). Released NETs contain neutrophil DNA, histones and granular content in a web-like structure that can lead to a greater risk of thromboembolism and enhance the formation of the pre-metastatic niche. Objective: The objective of this study was to determine whether exogenous delivery of tumor-associated EVs to healthy mice was sufficient to induce NET release and neutrophil-platelet aggregation similar to that observed in tumor burdened mice. Methods: EVs were isolated from both cultured 4T1 murine mammary carcinoma cells and from plasma harvested from 4T1 tumor-burdened SCID mice with advanced metastatic disease. Nanotracking analysis showed that the average EV concentration in diseased mice was 1.33*1010 ± 8.49*108 particles/ml. Healthy SCID mice were intravenously dosed with 10%, 50%, or 100% of this EV concentration using cell culture-derived EVs or with 10% or 100% of this concentration using plasma-derived EVs approximately 24 hours prior to imaging. Intravital imaging of the lungs of live SCID mice using spinning disk confocal microscopy was performed after application of a vacuum-stabilized lung window. Intravenous delivery of FITC dextran, SYTOX Orange, Alexa Fluor 647 CD49b mAbs, and Pacific Blue Ly6G mAbs enabled in vivo labeling of lung microvasculature, NETs, platelets, and neutrophils, respectively. Results: This study is the first to observe neutrophil-platelet interactions in vivo in real-time after tumor EV challenge. Both plasma-derived and culture-derived 4T1 EVs significantly increased neutrophil and platelet recruitment at all dosages compared to sham injection, and promoted the formation of neutrophil-platelet aggregates. For culture-derived EVs, platelet and neutrophil recruitment significantly increased between dosages of 10% and 50% EV concentration, but did not significantly increase between 50% and 100% EV concentration. For plasma-derived EVs, neutrophil recruitment significantly increased between dosages of 10% and 100% EV concentration, but platelet recruitment did not. Plasma-derived EVs provoked greater platelet recruitment than cell-culture EVs at similar dosages, increasing the number of platelets by 1.6 and 2 times for 10% and 100% dosages respectively. Cell culture 4T1 tumor EVs also led to a higher percentage of field of views (FOVs) containing NETs (63%, 60% and 71% FOVs with NETs for the 3 dosage levels) compared to sham injections (22-26% FOVs with NETs). Increased FOVs with NETs were also noted for plasma-derived EV dosages (67% and 45% of FOVs for the 2 dosage levels). Conclusion: These results highlight that tumor derived EVs may serve as a novel target to inhibit subsequent neutrophil and platelet activation, neutrophil-platelet aggregation, and NET release in cancer. Differences in response between plasma-derived and culture-derived EVs suggest the presence of tumor-induced alterations to global EV phenotype, as culture-derived EVs originating solely from tumor cells caused significantly less platelet recruitment than did plasma-derived EVs consisting of both specifically tumor-derived EVs and EVs originating from other cells. Future studies will investigate the impact of dosages over longer timeframes and on characterization of the molecular mechanism behind neutrophil, platelet and tumor-associated EV interactions.

First Name: Hunter Abstract Body : Cherenkov Radiation (CR), this blue glow seen in nuclear reactors, is an optical light originating from energetic β-emitter radionuclides. CR emitter 90Y triggers a cascade of energy transfers in the presence of a mixed population of fluorophores (which each other match their respective absorption and emission maxima): Cherenkov Radiation Energy Transfer (CRET) first, followed by multiple Förster Resonance Energy transfers (FRET): CRET ratios were calculated to give a rough estimate of the transfer efficiency. While CR is blue-weighted (300-500 nm), such cascades of Energy Transfers allowed to get a) fluorescence emission up to 710 nm, which is beyond the main CR window and within the near-infrared (NIR) window where biological tissues are most transparent, b) to amplify this emission and boost the radiance on that window: EMT6-tumor bearing mice injected with both a radionuclide and a mixture of fluorophores having a good spectral overlap, were shown to have nearly a two-fold radiance boost (measured on a NIR window centered on the emission wavelength of the last fluorophore in the Energy Transfer cascade) compared to a tumor injected with the radionuclide only. Some CR embarked light source could be converted into a near-infrared radiation, where biological tissues are most transparent.