Christel Vangestel

Early prediction of tumor response to treatment: pre-clinical validation of 99mTc-duramycin

Noninvasive imaging of cell death can provide an early indication of the efficacy of tumor treatment, aiding clinicians in distinguishing responding patients from nonresponding patients early on. 99mTc-duramycin is a SPECT tracer for cell death imaging. In this study, our aim was to validate the use of 99mTc-duramycin for imaging the early response of tumors to treatment. Methods: An in vitro binding assay was performed on COLO205 cells treated with 5-fluorouracil (3.1, 31, or 310 μM) and oxaliplatin (0.7 or 7 μM) or radiation (2 or 4.5 Gy). 99mTc-duramycin cell binding and the levels of cell death were evaluated after treatment. In vivo imaging was performed on 4 groups of CD1-deficient mice bearing COLO205 human colorectal cancer tumors. Each group included 6 tumors. The first group was given irinotecan (100 mg/kg), the second oxaliplatin (5 mg/kg), the third irinotecan (80 mg/kg) plus oxaliplatin (5 mg/kg), and the fourth vehicle (0.9% NaCl and 5% glucose). For radiotherapy studies, COLO205 tumors received 4.5 Gy, 2 fractions of 4.5 Gy in a 24-h interval, pretreatment with an 80 mg/kg dose of irinotecan combined with 2 fractions of 4.5 Gy in a 24-h interval, or no treatment (n = 5–6/group). Therapy response was evaluated by 99mTc-duramycin SPECT 24 h after the last dose of therapy. Blocking was used to confirm tracer specificity. Radiotracer uptake in the tumors was validated ex vivo using γ-counting, cleaved caspase-3, and terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL) histology. Results: Chemotherapy and radiotherapy increased 99mTc-duramycin binding to COLO205 cells in a concentration/dose- and time-dependent manner, which correlated well with cell death levels (P < 0.05) as analyzed by annexin V and caspase 3/7 activity. In vivo, 99mTc-duramycin uptake in COLO205 xenografts was increased 2.3- and 2.8-fold (P < 0.001) in mice treated with irinotecan and combination therapy, respectively. Blocking with unlabeled duramycin demonstrated specific binding of the radiotracer. After tumor irradiation with 4.5 Gy, 99mTc-duramycin uptake in tumors increased significantly (1.24 ± 0.07 vs. 0.57 ± 0.08 percentage injected dose per gram in the unirradiated tumors; P < 0.001). γ-counting of radioactivity in the tumors positively correlated with cleaved caspase-3 (r = 0.85, P < 0.001) and TUNEL (r = 0.81, P < 0.001) staining. Conclusion: We demonstrated that 99mTc-duramycin can be used to image induction of cell death early after chemotherapy and radiotherapy. It holds potential to be translated into clinical use for early assessment of treatment response.

99mTc-Duramycin SPECT imaging of early tumor response to targeted therapy: a comparison with 18F-FDG PET.

Molecular imaging of cell death may provide a detailed readout of the cellular response to novel therapies and prognostic information on tumor treatment efficacy, assisting in the design of individualized therapy. We compared the predictive power of cell death imaging using 99mTc-duramycin with the current gold standard 18F-FDG for treatment response evaluation after targeted therapy. Methods: Early therapy response evaluation was assessed by 99mTc-duramycin SPECT and 18F-FDG PET imaging in treatment-sensitive COLO205 and treatment-resistant HT29 human colorectal cancer xenografts 24 h after a single dose of conatumumab or IgG1 control. The specificity of 99mTc-duramycin for apoptosis was assessed using 99mTc-linear duramycin control radiotracer. Radiotracer uptake was validated ex vivo by γ-counting and autoradiography and compared with cleaved caspase-3 (CC3) activation and DNA fragmentation (TdT-mediated dUTP nick-end labeling [TUNEL]). Data were analyzed with the Student t test and Pearson correlation. All statistical tests were 2-sided. Results: COLO205 tumor uptake of 99mTc-duramycin was increased 7-fold from baseline in conatumumab- versus IgG1-treated control mice (P < 0.001), in good correlation with histologic analysis of apoptosis (CC3, r = 0.842, and TUNEL, r = 0.894; P < 0.001). No response was detected in HT29 tumors. No change in 99mTc-linear duramycin uptake could be detected in COLO205 tumors after treatment, indicating specificity of the 99mTc-duramycin tumor signal. 18F-FDG uptake was not significantly increased from baseline in conatumumab- versus IgG1-treated COLO205 and HT29 tumor–bearing mice (P = 0.104 and 0.779, respectively) and did not correlate with immunohistochemical evidence of apoptosis. Conclusion: We have demonstrated that 99mTc-duramycin specifically accumulates in apoptotic tumors in which 18F-FDG was not able to differentiate responding from nonresponding tumors early after treatment. 99mTc-duramycin holds promise as a noninvasive imaging radiotracer for early treatment evaluation in the clinic.

Synthesis and in vivo preclinical evaluation of an 18F labeled uPA inhibitor as a potential PET imaging agent

INTRODUCTION The urokinase plasminogen activator (uPA) system is a proteolytic cascade involved in tumor invasion and metastasis. uPA and its inhibitor PAI-1 are described as biomarkers for breast cancer with the highest level of evidence. The present study describes the synthesis and first in vivo application of an activity based uPA PET probe. METHODS Based on the design of a small irreversible and selective uPA inhibitor we developed an (18)F-labeled activity based probe for uPA imaging. Human uPA expressing MDA-MB-231-luc2-GFP cells were inoculated in the mammary fat pads of nude mice and treated with the probe once tumors reached a volume of 150mm(3). Scans were performed at 0.25, 0.75, 1.5, 4 and 6h post injection. To evaluate tumor uptake in vivo and ex vivo data were gathered. Biodistribution data of the organs and tissues of interest were collected at all time points. Due to a relatively low tumor uptake, probe stability was further evaluated. RESULTS The uPA targeting PET tracer was produced in high purity and with good specific radioactivity. In vivo PET data showed a maximum tumor uptake of 2,51±0,32 %ID/g at 4h p.i. A significant correlation between in vivo and ex vivo tumor uptake calculation was found (R=0.75; p<0.01). Due to a high blood signal at all time points, probe stability was further examined revealing high plasma protein binding and low plasma stability. CONCLUSIONS In vivo and ex vivo results clearly demonstrate that uPA expressing tumors can be detected with non-invasive PET imaging. Stability tests suggest that further optimization is needed to provide a better tumor-to-background contrast.

ATG4B inhibitors with a benzotropolone core structure block autophagy and augment efficiency of chemotherapy in mice

ABSTRACT Autophagy is a cell survival mechanism hijacked by advanced tumors to endure a rough microenvironment. Late autophagy inhibitors such as (hydroxy)chloroquine have been used clinically to halt tumor progression with modest success. However, given the toxic nature of these compounds and their lack of specificity, novel targets should be considered. We recently identified a benzotropolone derivative that significantly inhibited the essential autophagy protein ATG4B. Therefore, we synthesized and tested additional benzotropolone compounds to identify a promising ATG4B inhibitor that impairs autophagy both in vitro and in vivo. A compound library containing 27 molecules with a benzotropolone backbone was synthesized and screened for inhibition of recombinant ATG4B. Depending on the benzotropolone compound, inhibition of recombinant ATG4B ranged from 3 to 82%. Active compounds were evaluated in cellular assays to confirm inhibition of ATG4B and suppression of autophagy. Seven compounds inhibited processing of the autophagy protein LC3 and autophagosome formation. Compound UAMC‐2526 was selected for further in vivo use because of its fair plasma stability. This compound abolished autophagy both in nutrient‐deprived GFP‐LC3 mice and in CD1−/− Foxn1nu mice bearing HT29 colorectal tumor xenografts. Moreover, addition of UAMC‐2526 to the chemotherapy drug oxaliplatin significantly improved inhibition of tumor growth. Our data indicate that suppression of autophagy via ATG4B inhibition is a feasible strategy to augment existing chemotherapy efficacy and to halt tumor progression.

Preclinical evaluation of [111In]MICA-401, an activity-based probe for SPECT imaging of in vivo uPA activity

Urokinase-type plasminogen activator (uPA) and its inhibitor PAI-1 are key players in cancer invasion and metastasis. Both uPA and PAI-1 have been described as prognostic biomarkers; however, non-invasive methods measuring uPA activity are lacking. We developed an indium-111 (111 In)-labelled activity-based probe to image uPA activity in vivo by single photon emission computed tomography (SPECT). A DOTA-conjugated uPA inhibitor was synthesized and radiolabelled with 111 In ([111 In]MICA-401), together with its inactive, hydrolysed form ([111 In]MICA-402). A biodistribution study was performed in mice (healthy and tumour-bearing), and tumour-targeting properties were evaluated in two different cancer xenografts (MDA-MB-231 and HT29) with respectively high and low levels of uPA expression in vitro, with either the active or hydrolysed radiotracer. MicroSPECT was performed 95 h post injection followed by ex vivo biodistribution. Tumour uptake was correlated with human and murine uPA expression determined by ELISA and immunohistochemistry (IHC). Biodistribution data with the hydrolysed probe [111 In]MICA-402 showed almost complete clearance 95 h post injection. The ex vivo biodistribution and SPECT data with [111 In]MICA-401 demonstrated similar tumour uptakes in the two models: ex vivo 5.68 ± 1.41%ID/g versus 5.43 ± 1.29%ID/g and in vivo 4.33 ± 0.80 versus 4.86 ± 1.18 for MDA-MB-231 and HT-29 respectively. Human uPA ELISA and IHC showed significantly higher uPA expression in the MDA-MB-231 tumours, while mouse uPA staining revealed similar staining intensities of the two tumours. Our data demonstrate non-invasive imaging of uPA activity in vivo, although the moderate tumour uptake and hence potential clinical translation of the radiotracer warrants further investigation. Copyright

[99mTc]duramycin for cell death imaging: Impact of kit formulation, purification and species difference

INTRODUCTION [99mTc]duramycin is a SPECT tracer for cell death imaging. We evaluated the impact of kit formulation, purification and species difference on the pharmacokinetic profile and cell death targeting properties of [99mTc]duramycin in order to define the optimal conditions for (pre-)clinical use. METHODS Three kits were prepared (A: traditional formulation, B: containing 1/3 of ingredients, C: containing HYNIC-PEG12-duramycin). Following labeling, the kits were used without purification, or with SPE or HPLC purification. The pharmacokinetic profile was evaluated in mice and rats at 24 h post tracer injection (p.i.). Non-specific accumulation of [99mTc]duramcyin was studied by μSPECT imaging in chemotherapy treated COLO205 tumor bearing mice pre-treated with cold duramycin (0.01-50 μg). Cell death targeting ability of the kits displaying the best pharmacokinetic profile was compared in a treatment response study in COLO205 tumor bearing mice treated with conatumumab (anti-DR5 antibody). RESULTS HPLC purification of kit prepared [99mTc]duramycin and reducing the amount of kit ingredients resulted in the best pharmacokinetic profile with low accumulation in liver, spleen and kidneys. The use of PEGylated [99mTc]duramycin required longer circulation times (> 4 h pi) to obtain good imaging characteristics. Pre-treatment with duramycin significantly decreased tracer uptake in chemotherapy treated tumors in a dose-dependent manner. A blocking dose of 50 μg significantly increased non-specific accumulation in liver and spleen. Non-specific accumulation of [99mTc]duramycin was however demonstrated to be species dependent. HPLC purified kit A (5.21±1.71 %ID/cc) and non-purified kit B (1.68±0.46 %ID/cc) demonstrated a significant increase in tumor uptake compared to baseline following conatumumab treatment. CONCLUSIONS To obtain [99mTc]duramycin with favorable imaging characteristics for cell death imaging in mice [99mTc]duramycin needs to be prepared with high specific activity by applying HPLC purification. The need for HPLC purification appears to be a species dependent phenomenon and might therefore not be required for clinical translation.

Coadministration of a Gloriosa superba extract improves the in vivo antitumoural activity of gemcitabine in a murine pancreatic tumour model

BACKGROUND Gloriosa superba L. (glory lily, Colchicaceae) contains colchicine, and related alkaloids such as 3-O-demethylcolchicine and its glycoside colchicoside. Previously the in vivo efficacy of a crude extract and a colchicine-poor / colchicoside-rich extract of G. superba seeds was shown in a murine model of pancreatic adenocarcinoma. HYPOTHESIS/PURPOSE The efficacy can be improved without obvious signs of toxicity by increasing the treatment dose; the efficacy of gemcitabine can be improved by coadministration of a Gloriosa superba extract. STUDY DESIGN A survival experiment was carried out in a murine model of pancreatic adenocarcinoma and the semi-long-term toxicity of both G. superba extracts was determined; a combination therapy with gemcitabine was evaluated. METHODS A crude ethanolic extract (GS) and a colchicine-poor / colchicoside-rich extract (GS2B) were prepared, containing 3.22% colchicine, 2.52% colchicoside and 1.52% 3-O-demethylcolchicine (GS), and 0.07%, 2.26% and 0.46% (m/m) (GS2B). They were evaluated in a murine model of pancreatic adenocarcinoma at a dose of 4.5mg/kg (p.o., daily) total content of colchicine and derivatives during 3 weeks, or at 3.0mg/kg (p.o., daily) combined with gemcitabine (60mg/kg, i.p., 3x/week) during 54 days. RESULTS A significant effect in tumour growth over time was observed for gemcitabine and the combination therapy compared to the control group. No significant difference was observed for the groups treated with colchicine and both extracts. However, combination therapy was significantly better than the monotherapy with gemcitabine. Moreover, survival analysis showed a significant prolongation of the survival of the groups treated with gemcitabine and the combination therapy. A slight difference in survival was observed between gemcitabine and the combination therapy, the latter one being slightly better. No significant prolongation of survival was observed for the extracts and colchicine compared to the control group. CONCLUSION Although a relevant tumour growth inhibition and a difference of relative tumour volume compared to the control group were observed on day 11, and a slightly longer survival was noticed for GS2B, the most important conclusion from this study is that the crude G. superba extract (GS) might have an added value combined with gemcitabine in the treatment of pancreatic tumours.

How to Modulate Tumor Hypoxia for Preclinical In Vivo Imaging Research

Tumor hypoxia is related with tumor aggressiveness, chemo- and radiotherapy resistance, and thus a poor clinical outcome. Therefore, over the past decades, every effort has been made to develop strategies to battle the negative prognostic influence of tumor hypoxia. For appropriate patient selection and follow-up, noninvasive imaging biomarkers such as positron emission tomography (PET) radiolabeled ligands are unprecedentedly needed. Importantly, before being able to implement these new therapies and potential biomarkers into the clinical setting, preclinical in vivo validation in adequate animal models is indispensable. In this review, we provide an overview of the different attempts that have been made to create differential hypoxic in vivo cancer models with a particular focus on their applicability in PET imaging studies.

Abstract 3910: Targeting urokinase plasminogen activator: evaluation of activity-based imaging probes in an orthotopic breast cancer model.

The urokinase plasminogen activator (uPA) system is a proteolytic cascade involved in tumor invasion and metastasis. uPA and its inhibitor PAI-1 are described as biomarkers for breast cancer with the highest level of evidence. For the screening of uPA in breast tumors with commercially available ELISA kits that determine the total uPA content (active and inactive), fresh tumor tissue is required. Molecular imaging with activity-based uPA probes might overcome this limitation. Therefore we have developed some highly selective, non-peptidic and specific uPA activity-based probes based on the irreversible uPA inhibitor UAMC-00150. In this study we describe the evaluation of a fluorescent Cy5 labeled and a 18F labeled PET probe in the high uPA expressing MDA-MB-231 breast cancer mouse model to make the step towards clinical translation of this imaging biomarker. Orthotopic MDA-MB-231 tumors were established in nude mice and subjected to uPA imaging at a tumor volume of 150mm3. For the fluorescent imaging, three test groups were considered (n=5): mice from the first and the second experimental group received a single dose of equimolar amounts (0.23 μMol/kg) Cy5-uPA-probe or its inactivated hydrolyzed variant, respectively. The third group received an injection with unlabeled uPA inhibitor, 30 min before administration of Cy5-uPA probe, in a concentration that was 70-fold higher than the probe concentration. Mice were imaged at different time points for 48 h post injection (p.i.) and were subsequently sacrificed for ex vivo imaging and histology. PET imaging (n=15) was performed after i.v. injection of the radiotracer with a maximum of 0,2ml and/or 0,5mCi per injection, resulting in a range of 0,15-0,5mCi/injection. Animals were scanned at 15, 45, 90, 240 and 360 minutes p.i. At each time point, 3 mice were sacrificed to generate ex vivo biodistribution data. The Cy5-uPA-probe demonstrated good tumor-targeting properties in the high uPA-expressing breast tumor model, with fluorescent intensities reaching a maximum at 24 h post-probe administration in mice treated with Cy5-uPA-probe alone. The groups treated with the inactivated probe or the unlabeled inhibitor showed a significant decrease in the fluorescent tumor signal (p>0,016), which was also confirmed on histology. PET imaging with the 18F-uPA-probe showed a peak tumor uptake of 2.76 ± 0.37 %ID/g at 4 h p.i. Further in vitro PPB and ex vivo HPLC data revealed a high affinity of the tracer for blood proteins and a high metabolisation rate potentially explaining the rather moderate to low tumor uptake of the current PET uPa biomarker. In conclusion, fluorescent imaging data clearly indicate that the Cy5-uPA-probe enables non-invasive NIR-imaging of uPA expression in tumors in vivo. The first PET experiments indicate translational capabilities, for which we are now improving the pharmacokinetics of this and future uPa radiotracers. Citation Format: Johan Ides, Christel Vangestel, Jonas Messagie, Dieter Verzele, David Thomae, Sofie Thys, An Wouters, John-Paul Bogers, Sigrid Stroobants, Jurgen Joossens, Pieter Van der Veken, Marc Peeters, Filip Lardon, Steven Staelens, Koen Augustyns. Targeting urokinase plasminogen activator: evaluation of activity-based imaging probes in an orthotopic breast cancer model. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3910. doi:10.1158/1538-7445.AM2013-3910