OS08.8.A Integrin-specific CAR T cells for the treatment of glioblastoma

Background

Current standard of care for glioblastoma patients has limited therapeutic efficacy and novel innovative treatment strategies are urgently needed. One such strategy is chimeric antigen receptor (CAR) T cell therapy that has shown great success in hematological malignancies. αv integrins are overexpressed in several neoplasms and have already been used as therapeutic targets for small molecule inhibitors and antibodies, which did not cause major toxicities. Herein, we propose αv integrins as an ideal target for CAR T cell therapy in glioblastoma.

Material and Methods

CAR T cells targeting specific heterodimers (αvβ3, αvβ5 or αvβ8) were generated by transducing primary human T cells from healthy donors with a lentiviral vector expressing a second-generation CAR. Activity and specificity of CAR T cells was determined by co-culture assays with different glioma cells. Efficacy of CAR T cells to control tumor growth in vivo was investigated in clinically relevant orthotopic xenograft glioma mouse models. Additionally, we generated CAR T cells from T cells from a glioblastoma patient and measured their activity against the patient’s autologous tumor cells.

Results

All newly generated integrin-targeting CAR T cells exerted strong anti-glioma activity in vitro. Long-term and repetitive killing assays as well as cytokine-release measurements demonstrated highest activity of αvβ5 and αvβ8 integrin-specific CAR T cells. Antigen specificity of these cells was confirmed, as glioma cells with a CRISPR/Cas9-mediated knockout of the target antigen were resistant to CAR T cell-mediated cytotoxicity. Intratumoral injection of αvβ5 or αvβ8 CAR T cells significantly prolonged the survival and cured a substantial fraction of glioma-bearing mice in two different xenograft models. When used in a patient-derived setting, matched CAR T cells exerted strong anti-glioma activity.

Conclusion

We show strong and integrin-specific anti-glioma activity of CAR T cells developed from healthy donor T cells and glioblastoma-patient-derived T cells in vitro. αvβ5- and αvβ8-specific CAR T cells exerted the best therapeutic activity in two different xenograft glioma models in vivo. These data support the evaluation of integrin-specific CAR T cells as a therapeutic strategy in clinical neuro-oncology.

OS08.7.A Lomustine and the immunocytokine L19TNF are a promising treatment combination for recurrent glioblastoma

Background

Treatment options for recurrent glioblastoma are limited and with the possible exception of regorafenib, no agent has demonstrated superior activity to lomustine. Therefore, there is an urgent need for more effective treatment strategies for recurrent glioblastoma. Here, we investigated different treatment combinations based on the tumor-stroma targeting antibody-cytokine fusion protein L19TNF in preclinical glioma models and translated the most effective treatment combination to patients with recurrent glioblastoma.

Material and Methods

Orthotopic immunocompetent mouse glioma models were used to study the anti-glioma activity of L19TNF in combination with anti-PD1, bevacizumab or lomustine. Tumor growth was monitored by MRI. Flow cytometry and microscopy were used to characterize tumor-infiltrating-immune cells. MHC immunoaffinity purification and mass spectrometry were used to characterize the MHC immunopeptidome. Genetic mouse models were used to study immune-dependent effects. Subsequently, we translated the most efficient treatment combination to patients with recurrent glioblastoma within a phase I/II clinical trial (NCT04573192).

Results

The combination of L19TNF and lomustine demonstrated strong synergistic anti-tumor activity in two immunocompetent orthotopic glioma models and cured a majority of tumor-bearing mice. In contrast, combinations with anti-PD-1 or bevacizumab had only limited anti-glioma activity. Furthermore, compared to the monotherapies, the combination of L19TNF and lomustine led to the strongest increase in tumor-infiltrating lymphoid cells as demonstrated by flow cytometry and microsopy and to the highest number of peptides presented in the context of MHC-I. The treatment effect was abrograted in different genetic immunodeficient mouse models. The treatment combination of L19TNF and lomustine was well tolerated in the first patients treated within a phase I/II clinical trial and we observed partial tumor responses also in patients with an unmethylated MGMT promoter.

Conclusion

The combination of L19TNF and lomustine demonstrated promising anti-glioma activity and patients are currently recruited within a phase I/II clinical trial for patients with recurrent glioblastoma.

P10.20.A Mechanisms of synergistic glioma growth suppression by radiotherapy and MET inhibition

Background

Glioblastoma remains to be one of the most lethal solid cancers. Despite multi-modal therapy including surgery as safely feasible, radiotherapy and chemotherapy with the alkylating agent temozolomide, the median survival of affected patients is still limited to approximately one year on a population level. Thus, novel therapies are urgently needed. There is increasing interest in the role of the HGF/MET pathway in the response of glioblastoma to radiotherapy since MET may be involved in radioresistance via proinvasive and DNA damage response pathways.

Material and Methods

Here we assessed the role of the MET pathway in the response to radiotherapy in vitro and in vivo in syngeneic mouse glioma models and explored potential modes of action responsible for the synergistic effects of MET pathway inhibition and irradiation on tumor growth in vivo.

Results

Murine glioma cells express HGF and MET and show increased MET phosphorylation upon exposure to exogenous HGF. In vitro, glioma cell viability and proliferation are not affected by pharmacological MET inhibition using tepotinib or genetic MET inhibition using CRISPR/Cas9-engineered Met gene knockout and sensitization to irradiation by MET inhibition is not seen. In vivo, the combination of MET inhibition with focal radiotherapy mediates prolonged survival of syngeneic orthotopic glioma-bearing mice compared with either treatment alone. Complementary studies demonstrate that synergy is lost when gliomas are established and treated in immunodeficient mice, but also if MET gene expression is disrupted in the tumor of wildtype mice. Combination therapy suppresses a set of pro-inflammatory mediators that are upregulated by radiotherapy alone and which are positively regulated by transforming growth factor (TGF)-β. In line with this data, ex vivo analysis of mouse brains reveal increased TGF-β pathway activity upon irradiation alone that is counteracted by concomitant MET inhibition.

Conclusion

In summary, we demonstrate synergistic suppression of syngeneic glioma growth by irradiation and MET inhibition that requires MET expression in the tumor as well as an intact immune system. Clinical evaluation of this combined treatment approach in newly diagnosed glioblastoma is warranted.

P10.19.B An immunotoxin targeting CD317 for the treatment of glioblastoma

Background

CD317 is an interferon-inducible cell surface receptor expressed in several solid cancer types. HM1.24-ETA’ is a small immunotoxin with a CD317 single-chain variable fragment (svFv) antibody fused to a truncated version of Pseudomonas aeruginosa exotoxin A (ETA’) that is explored as a novel therapeutic approach in CD317-expressing tumors.

Material and Methods

CD317 mRNA expression in human gliomas and its association with survival was analyzed using the database of the Cancer Genome Atlas (TCGA). CD317 protein levels in human gliomas were assessed by immunohistochemistry. CD317 mRNA expression was assessed by RT-PCR and CD317 protein levels by flow cytometry in 13 human glioma cell lines in vitro. Efficacy of HM1.24-ETA’ was analyzed in acute cytotoxicity assays in vitro. Finally, HM1.24-ETA’ was evaluated in the intracranial human LN-229 glioma xenograft nude mouse model after intravenous injection.

Results

Interrogation of the TCGA database showed that increased CD317 mRNA expression correlated with grade of malignancy among isocitrate dehydrogenase (IDH) wildtype and IDH-mutant gliomas. Enhanced CD317 mRNA expression was associated with inferior survival in glioblastoma and astrocytoma, IDH-mutant, WHO grade 4. Immunohistochemistry confirmed CD317 overexpression in human glioblastoma compared to lower grade astrocytomas. CD317 was expressed heterogeneously on mRNA and protein levels in glioma cell lines in vitro. HM1.24-ETA’ induced acute cytotoxicity in CD317-positive glioma cells in vitro. CD317 expression and susceptibility to HM1.24-ETA’-induced cell death were enhanced by interferon-β. HM1.24-ETA’ prolonged survival in the LN-229 xenograft nude mouse model.

Conclusion

These data define CD317 as a novel target for treatment of human gliomas with immunoconjugates.

OS10.4.A A macrophage-based drug delivery platform for glioma treatment

Background

There is an urgent need for more effective treatment strategies against gliomas. At present, even though various drugs have potent anti-tumor activity in vitro, their application in vivo is limited by ineffective delivery and systemic toxicity. Therefore, novel strategies are needed to deliver these drugs effectively and safely to the tumor site. Here, we developed an adoptive transfer strategy against malignant brain tumors utilizing macrophages that are loaded with ferritin-protein cages containing drugs or other proteins and transfer these nanocarriers to cancer cells in vitro and in vivo.

Material and Methods

Live-time imaging, microscopy and flow cytometry were utilized to investigate the transfer of ferritin cages from loaded macrophages to human or mouse glioma cells. Co-cultures of glioma cells and macrophages loaded with ferritin-drug cages were used to study the anti-glioma activity in vitro and orthotopic immunocompetent mouse glioma models were used to study the anti-glioma activity in vivo. Affinity purification-mass spectrometry (AP-MS) was used to elucidate the mechanisms of transfer by characterizing the interactome of ferritin cages within macrophages and cancer cells.

Results

We observed a high transfer efficiency of ferritin-cages from loaded macrophages into human and mouse glioma cells in vitro in co-culture assays and confirmed the transfer from macrophages to glioma cells also in vivo upon intravenous or intratumoral treatment of GL-261 or CT-2A glioma-bearing mice. To study the anti-glioma activity with therapeutically active payloads, we loaded murine/human macrophages with ferritin cages carrying cytotoxic payloads. Co-culture of these loaded macrophages with murine or human glioma cells in vitro revealed a time- and concentration-dependent cytotoxicity to glioma cells. In vivo, intravenous or intratumoral administration of ferritin-drug protein cages was tolerated without toxicities and conferred a survival benefit in two orthotopic murine glioma models (GL-261 and CT-2A). Interactome studies of ferritin-cage-binding proteins revealed phagocytic and cytoskeleton re-arrangement pathways to be involved in uptake and transfer of ferritin-cages from macrophages to glioma cells.

Conclusion

This ‘Trojan Horse’ approach constitutes a promising platform to deliver cytotoxic drugs effectively and safely to gliomas and provides a rationale for clinical translation.

Rational Design of Highly Potent, Selective, and Bioavailable SGK1 Protein Kinase Inhibitors for the Treatment of Osteoarthritis

The serine/threonine kinase SGK1 is an activator of the β-catenin pathway and a powerful stimulator of cartilage degradation that is found to be upregulated under genomic control in diseased osteoarthritic cartilage. Today, no oral disease-modifying treatments are available and chronic treatment in this indication sets high requirements for the drug selectivity, pharmacokinetic, and safety profile. We describe the identification of a highly selective druglike 1H-pyrazolo[3,4-d]pyrimidine SGK1 inhibitor 17a that matches both safety and pharmacokinetic requirements for oral dosing. Rational compound design was facilitated by a novel hSGK1 co-crystal structure, and multiple ligand-based computer models were applied to guide the chemical optimization of the compound ADMET and selectivity profiles. Compounds were selected for subchronic proof of mechanism studies in the mouse femoral head cartilage explant model, and compound 17a emerged as a druglike SGK1 inhibitor, with a highly optimized profile suitable for oral dosing as a novel, potentially disease-modifying agent for osteoarthritis.

PL03.3.A Development and characterization of CD317-specific CAR T cells as an innovative immunotherapeutic strategy against glioblastoma

BACKGROUND

Due to the limited success of existing therapies for gliomas, innovative therapeutic options are urgently needed. Chimeric antigen receptor (CAR) T cell therapy has been successful in patients with hematological malignancies. However, using this treatment against solid tumors such as glioblastomas is more challenging. Here, we generated CAR T cells targeting the transmembrane protein CD317 (BST-2, HM1.24) which is overexpressed in glioma cells and may therefore serve as a novel target antigen for CAR T cell-based immunotherapy.

MATERIAL AND METHODS

CAR T cells targeting CD317 were generated by lentiviral transduction of human T cells from healthy donors. The anti-glioma activity of CD317-CAR T cells was determined in lysis assays using different glioma target cell lines with varying CD317 expression levels. The efficiency of CD317-CAR T cells to control tumor growth in vivo was evaluated in clinically relevant orthotopic xenograft glioma mouse models.

RESULTS

We created a second-generation CAR construct targeting CD317 and observed strong anti-glioma activity of CD317-CAR T cells in vitro. Glioma cells with a CRISPR/Cas9-mediated CD317 knockout were resistant to CD317-specific CAR T cells, demonstrating their target antigen-specificity. Since CD317 is also expressed by T cells, transduction with a CD317-directed CAR resulted in fratricide of the transduced T cells. Silencing of CD317 in CAR T cells by integrating a specific shRNA into the CAR vector significantly increased the viability, proliferation and cytotoxic function of the CAR T cells. Importantly, intratumoral treatment with CD317-CAR T cells prolonged the survival and cured a significant fraction of glioma-bearing nude mice.

CONCLUSION

We demonstrate strong CD317-specific anti-tumor activity of CD317-CAR T cells against various glioma cell lines in vitro and in xenograft glioma models in vivo. These data lay a scientific basis for the subsequent evaluation of this therapeutic strategy in clinical neuro-oncology.

OS06.7A METPlatform identifies brain metastasis vulnerabilities and predicts patient response to therapy

BACKGROUND

The diagnosis of brain metastasis involves high morbidity and mortality and remains an unmet clinical need in spite of being the most common tumor in the brain. Exclusion of these cancer patients from clinical trials is a major cause of their limited therapeutic options.

MATERIAL AND METHODS

We report a novel drug-screening platform (METPlatform) based on organotypic cultures which allows identifying effective anti-metastasis agents in the presence of the organ microenvironment. We have applied this approach to clinically relevant stages of brain metastasis using both experimental models and human tumor tissue (by performing patient-derived organotypic cultures - PDOCs -). We have also used METPlatform to perform unbiased proteomics of brain metastases in situ to identify potential novel mediators of this disease and explore resistance mechanisms to targeted therapy. Finally, we have exploited METPlatform as “avatars” to predict response to therapy in patients with primary brain tumors.

RESULTS

We identified heat shock protein 90 (HSP90) as a promising therapeutic target for brain metastasis. DEBIO-0932, a blood-brain barrier permeable HSP90 inhibitor, shows high potency against mouse and human brain metastases from different primary origin and oncogenomic profile at clinically relevant stages of the disease, including a novel model of local relapse after neurosurgery. Furthermore, in situ proteomic analysis of brain metastases treated with the chaperone inhibitor revealed non-canonical clients of HSP90 as potential novel mediators of brain metastasis and actionable mechanisms of resistance driven by autophagy. Combined therapy using HSP90 and autophagy inhibitors showed synergistic effects compared to sublethal concentrations of each monotherapy, demonstrating the potential of METPlatform to design and test rationale combination therapies to target metastasis more effectively. Finally, we show that brain tumor PDOCs predict the response of the corresponding patient to standard of care, thus proving the potential of METPlatform for improving personalized care in cancer.

CONCLUSION

Our work validates METPlatform as a potent resource for metastasis research integrating drug-screening and unbiased omic approaches that is fully compatible with human samples and questions the rationale of excluding patients with brain metastasis from clinical trials. We envision that METPlatform will be established as a clinically relevant strategy to personalize the management of metastatic disease in the brain and elsewhere.

Large Quantum Delocalization of a Levitated Nanoparticle Using Optimal Control: Applications for Force Sensing and Entangling via Weak Forces

We propose to optimally control the harmonic potential of a levitated nanoparticle to quantum delocalize its center-of-mass motional state to a length scale orders of magnitude larger than the quantum zero-point motion. Using a bang-bang control of the harmonic potential, including the possibility of inverting it, the initial ground-state-cooled levitated nanoparticle coherently expands to large scales and then contracts to the initial state in a time-optimal way. We show that this fast loop protocol can be used to enhance force sensing as well as to dramatically boost the entangling rate of two weakly interacting nanoparticles. We parameterize the performance of the protocol, and therefore the macroscopic quantum regime that could be explored, as a function of displacement and frequency noise in the nanoparticle's center-of-mass motion. This noise analysis accounts for the sources of decoherence relevant to current experiments.

Outcomes of edoxaban-treated patients with atrial fibrillation and concomitant vascular disease in daily clinical practice: insights from ETNA-AF-Europe

Funding Acknowledgements

Type of funding sources: Private company. Main funding source(s): Daiichi Sankyo Europe

OnBehalf

ETNA-AF-Europe investigators

Background

Coronary (CAD) and peripheral artery disease (PAD) are common comorbidities in patients with atrial fibrillation (AF). Such concomitant vascular disease may affect the effectiveness and safety of treatment with edoxaban, but data from daily practice are limited. Purpose: To determine the incidence of ischaemic and bleeding events in edoxaban-treated AF patients with CAD/PAD, and to assess whether the effect of edoxaban might be influenced by concomitant vascular disease. Methods: With 1-year follow-up data from ETNA-AF-Europe, clinical characteristics and frequencies of adverse events were compared between AF-patients with and without CAD/PAD. Results: Of 13,089 patients, 2956 had vascular disease (n = 2738; 92.6% had CAD). Patients with CAD/PAD were older, had lower creatinine clearance, higher stroke and bleeding risk scores, and were more frequently prescribed the reduced dose of edoxaban than those without these comorbidities (Table). Rates of stroke/systemic embolism (0.99 vs 0.77%/year), and major bleeding (1.28 vs 0.98%/year) were numerically higher in patients with CAD/PAD than in those without. Myocardial infarction (1.31 vs 0.30%/year), cardiovascular deaths (2.33 vs 1.43%/year) as well as all-cause deaths (5.18 vs 3.01%/year) occurred significantly more often in the CAD/PAD subgroup (Figure). Conclusions: Rates of ischaemic and bleeding events are low in unselected edoxaban-treated AF patients, regardless of the presence/absence of concomitant vascular disease. The higher rates of myocardial infarction and (non-)cardiovascular death in the CAD/PAD subgroup are thought to be largely unrelated to differences in the effect of edoxaban, but attributable to the differences in baseline intrinsic risks instead. Nonetheless, these differences warrant further investigation. Baseline characteristics n (%) or Mean ± SDPatients with CAD and/or PADPatients without CAD and/or PADTotal (N = 2956)Edoxaban 60 mg (n = 2044)Edoxaban 30 mg (n = 915)Total (N = 10,133)Edoxaban 60 mg (n = 7947)Edoxaban 30 (n = 2186)Male2115 (71.5)1541 (75.4)574 (62.7)5315 (52.5)4520 (56.9)795 (36.4)Age, years75.3 ± 8.573.4 ± 8.279.4 ± 7.673.1 ± 9.771.4 ± 9.379.6 ± 8.1Body mass index, kg/m²28.1 ± 4.728.6 ± 4.727.0 ± 4.728.1 ± 5.228.6 ± 5.226.3 ± 5.1eGFR,* mL/min/1.73m²69.3 ± 28.378.2 ± 27.450.7 ± 19.875.7 ± 30.983.2 ± 29.650.1 ± 19.5(calc.)CHA2DS2-VASc3.5 ± 1.43.3 ± 1.44.0 ± 1.43.0 ± 1.42.8 ± 1.33.7 ± 1.2(calc.) mod. HAS-BLED†2.9 ± 1.12.8 ± 1.13.3 ± 1.12.4 ± 1.12.3 ± 1.12.8 ± 1.0Frailty‡391 (13.2)159 (7.8)232 (25.4)1001 (9.9)453 (5.7)548 (25.1)Data are presented as mean ± standard deviation for continuous variables or n (%) for categorical variables. *Glomerular filtration rate (GFR) was estimated by Cockcroft-Gault formula. †Not including labile international normalised ratio. ‡It was left to the physician"s discretion to categorise a patient as frail.Abstract Figure. One-year outcomes

Orders of magnitude loss reduction in photonic bandgap fibers by engineering the core surround

We demonstrate how to reduce the loss in photonic bandgap fibers by orders of magnitude by varying the radius of the corner strands in the core surround. As a fundamental working principle we find that changing the corner strand radius can lead to backscattering of light into the fiber core. Selecting an optimal corner strand radius can thus reduce the loss of the fundamental core mode in a specific wavelength range by almost two orders of magnitude when compared to an unmodified cladding structure. Using the optimal corner radius for each transmission window, we observe the low-loss behavior for the first and second bandgaps, with the losses in the second bandgap being even lower than that of the first one. Our approach of reducing the confinement loss is conceptually applicable to all kinds of photonic bandgap fibers including hollow core and all-glass fibers as well as on-chip light cages. Therefore, our concept paves the way to low-loss light guidance in such systems with substantially reduced fabrication complexity.

On the pole expansion of electromagnetic fields

In several publications, it has been shown how to calculate the near- or far-field properties for a given source or incident field using the resonant states, also known as quasi-normal modes. As previously noted, this pole expansion is not unique, and there exist many equivalent formulations with dispersive expansion coefficients. Here, we approach the pole expansion of the electromagnetic fields using the Mittag-Leffler theorem and obtain another set of formulations with constant weight factors for each pole. We compare the performance and applicability of these formulations using analytical and numerical examples. It turns out that the accuracy of all approaches is rather comparable with a slightly better global convergence of the approach based on a formulation with dispersive expansion coefficients. However, other expansions can be superior locally and are typically faster. Our work will help with selecting appropriate formulations for an efficient description of the electromagnetic response in terms of the resonant states.

First-order perturbation theory for changes in the surrounding of open optical resonators

The single-mode approximation of resonant state expansion has proven to give accurate first-order approximations of resonance shifts and linewidth changes when modifying the material properties inside open optical resonators. Here, we extend this first-order perturbation theory to modifications of the material properties in the surrounding medium. As a side product of our derivations, we retrieve the already known analytical normalization condition for resonant states. We apply our theory to two example systems: a metallic nanosphere and a periodic array of metallic nanoslits.

P1257ETNA-AF Europe: First 1-year follow-up snapshot analysis of more than 7,500 AF patients treated with edoxaban in routine clinical practice

Introduction

Edoxaban has been approved for stroke prevention in patients with atrial fibrillation based on its comparable efficacy and superior safety compared to warfarin in the pivotal ENGAGE AF-TIMI 48 trial. ETNA-AF Europe (NCT02944019) was initiated in agreement with the EMA to evaluate benefits and risks of edoxaban treatment in unselected patients in routine clinical practice.

Methods

13,980 patients from across 825 hospital and office-based physicians from 10 European countries (Austria, Belgium, Germany, Ireland, Italy, The Netherlands, Portugal, Spain, Switzerland and United Kingdom) were enrolled, and will be followed-up for 4 years. This snapshot analysis includes baseline and first outcome data of 7,672 patients (56.3% of all enrolled patients) that have completed their first 1-year follow-up visit (mean follow-up: 343.5 days).

Results

The average age of patients was 73.4 years, the mean weight was 81.9 kg (Table 1). Frequent comorbidities include hypertension (77.2%), valvular heart disease (17.4%), congestive heart failure (5.8%) and history of myocardial infarction (4.2%). Patients receiving the 30 mg dose (22.9%) were older, had a lower creatinine clearance and had a higher risk for both stroke and bleeding as compared to those on the 60 mg dose (77.1%). Overall, the incidence of clinical events was low: all-cause mortality: 3.56%/y, major bleeding 0.95%/y, intracranial haemorrhage 0.28%/y, any stroke or systemic embolic events 0.88%/y.

Patient characteristics at Year 1 Patient characteristics All patients Edoxaban 60 mg Edoxaban 30 mg [7,672] [5,916 (77.1%)] [1,756 (22.9%)] Age [years] mean (SD) 73.4 (9.26) 71.8 (8.98) 79.1 (7.81) Body weight [kg] mean (SD) 81.9 (17.33) 84.1 (16.80) 74.3 (16.93) CrCl (CG) [mL/min] mean (SD) 75.0 (30.29) 82.5 (29.14) 51.2 (19.75) CHA2DS2-VASc mean (SD) 3.1 (1.38) 2.9 (1.34) 3.8 (1.28) HAS-BLED mean (SD) 2.5 (1.10) 2.4 (1.07) 2.9 (1.08) First occurrence of all-cause mortality (n, %/year) 257 (3.56%) 129 (2.31%) 128 (7.90%) First occurrence of intracranial haemorrhage (n, %/year) 20 (0.28%) 16 (0.29%) 4 (0.25%) First occurrence of major bleeding (n, %/year) 68 (0.95%) 49 (0.88%) 19 (1.18%) First occurrence of stroke/SEE (n, %/year) 63 (0.88%) 45 (0.81%) 18 (1.11%) CG, Cockcroft-Gault; CrCl, creatinine clearance; SD, standard deviation; SEE, systemic embolic events.

Conclusions

We found low bleeding and stroke rates in 7,672 unselected, mainly elderly AF patients treated with edoxaban in routine clinical practice. These findings were consistent across edoxaban doses and reinforce the effectiveness and safety of NOACs such as edoxaban in routine clinical care in Europe.

Acknowledgement/Funding

Daiichi Sankyo Europe GmbH, Munich, Germany

P4766Edoxaban Treatment in routiNe clinical prActice for patients with atrial fibrillation (AF) in Europe (ETNA-AF-Europe): 1-year follow-up according to body mass index

Background

Non-vitamin K antagonist (VKA) oral anticoagulants (NOACs) have substantially improved anticoagulation therapy for prevention of stroke and systemic embolism in patients with atrial fibrillation (AF), and available routine care data have so far broadly confirmed the safety of different NOACs in routine practice. However, such data for edoxaban are scarce, especially in extremely low and high body weight (BW). These extreme BWs may affect the bioavailability, distribution, and half-life of NOACs and, consequently, outcomes of treatment.

Methods

We analysed outcomes in normal-weight (BMI 18.5–25) vs overweight (BMI 25–30) and obese (BMI >30) patients enrolled into the ETNA-AF-Europe observational study (NCT02944019) collecting information on patients treated with edoxaban in 825 sites in 10 European countries. This snapshot analysis set includes data of 7,672 patients (56.3% of all enrolled patients) which have completed their 1-year follow-up visit (mean follow-up: 343.5 days).

Results

Median patient age was 74 years for all patients, 76 years for patients with a BMI 18.5–25 (group 1), 75 years for patients with BMI 25–30 (group 2), and 72 for patients with a BMI >30 (group 3). CrCl was 64 mL/min for patients with a BMI 18.5–25, 68 mL/min for patients with BMI 25–30, and 72 mL/min for patients with a BMI >30. The CHA2DS2-VASc (mean 3.1±1.38) and HAS-BLED (mean 2.5±1.10) score did not differ significantly between groups. As expected, diabetes and hypertension were significantly less prevalent in leaner patients and - accordingly - inversely correlated to age.

There was no correlation between body weight and life-threatening bleeding (group 1: 0.28%; group 2: 0.40%; group 3: 0.14%). Also, stroke rates (group 1: 0.74%; group 2: 0.81%; group 3: 0.76%) did not differ between groups.

Conclusion

BMI, within the range here assessed, does not affect 1-year outcomes in European AF patients treated with edoxaban.

Acknowledgement/Funding

Daiichi Sankyo Europe GmbH, Munich, Germany

P14.72 Multicentric glioblastoma - A retrospective study of imaging characteristics, treatment approach, pattern of relapse and survival

BACKGROUND

Multicentric glioblastoma (m-GBM), defined as well separated tumoral foci, is a rare GBM variant comprising 6–13% of all GBM cases. Published data regarding m-GBM is scarce and is largely reporting on multicentric enhancing foci. We performed a retrospective study to determine the incidence, imaging characteristics, treatment approach, pattern of relapse and prognosis of m-GBM.

MATERIAL AND METHODS

The neuropathological database of our institution was surveyed for histological diagnosis of adult GBM diagnosed between 01/01/2015 and 31/05/2018. All pre-operative MRI were reviewed to identify patients with m-GBM. We included in the definition of m-GBM well separated enhancing as well as non-enhancing tumor foci. The medical records and follow-up MRI studies were reviewed in order to retrieve the data.

RESULTS

Of the 170 patients with newly diagnosed GBM 14 (8%) presented with m-GBM. All of them had at least one enhancing lesion and 11 (78.5%) patients had additional well separated non-enhancing tumor foci. The total number of lesions was 37 (19 enhancing and 18 non-enhancing) with a median number of lesions per patient of 2 (range 2 to 4). Median age at diagnosis was 66 (range: 49–79) years. Nine of the patients (64%) underwent surgical resection of the enhancing component whereas 5 patients had only a biopsy. Median follow up was 14.3 (range: 2–30) months. All but one patients were treated by standard concurrent radiotherapy with temozolomide. Median progression free survival is 6.2 (range: 0–13.3) months. Five of the 18 non-enhancing tumor foci eventually displayed contrast enhancement during the course of the disease. At last follow up, 12 patients died, with an overall survival of 12.3 months. Information regarding radiation fields, pattern of disease progression and molecular profile will be presented at the meeting.

CONCLUSION

m-GBM presents therapeutic dilemas regarding the optimal surgical approach and radiation field planning. Better understanding of the disease course and pattern of progression may help to optimize the therapeutic approach implying particularly to non-enhancing tumor foci.

P12.08 Immunocytokines are a novel immunotherapeutic approach against glioblastoma

BACKGROUND

Glioblastoma is the most common malignant primary brain tumor in adults with an urgent need for novel treatment options. The administration of pro-inflammatory cytokines could be a potent immunotherapeutic approach to shift the balance between tumor-associated immunosuppression and immune activation. However, the systemic administration of therapeutically active doses of pro-inflammatory cytokines is not feasible due to toxic side effects and there is a need for strategies that enable a targeted delivery of pro-inflammatory cytokines to the tumor site.

METHODS

We investigated different antibody-cytokine fusion products that enable a targeted delivery of interleukin (IL)-2, IL-12 or tumor necrosis factor (TNF)-α to the tumor site by binding to a tumor-specific epitope of fibronectin. We investigated the expression of this tumor-specific epitope ex vivo in tumor-bearing mouse brains and human glioblastoma samples. Subsequently, we assessed the anti-tumor activity of IL-2, IL-12 or TNF-α fused to an antibody targeting this tumor-specific epitope in orthotopic syngeneic mouse glioma models.

RESULTS

The tumor-specfic extra domain B of fibronectin is expressed in murine glioma models and human glioblastoma samples. A fluorochrome-labeled antibody targeting this tumor-specific epitope accumulated at the tumor site in the brain in vivo upon systemic administration. IL-2, IL-12, or TNF-α fused to this antibody conferred a survival benefit in orthotopic tumor-bearing mice and cured a fraction of tumor-bearing mice. Mechanistically, antibody-fused TNF-α induced tumor necrosis and increased the activation of tumor-infiltrating natural killer (NK) cells, whereas antibody-fused IL-12 mainly boosted an anti-tumor immune response mediated by NK cells and T cells.

CONCLUSION

We demonstrate the expression of a tumor-specific epitope of fibronectin in glioblastoma and exploit this for the targeted delivery of IL-2, IL-12 or TNF-α to the tumor site. Our preclinical assessments indicate potent anti-tumor activity in orthotopic, syngneic glioma mouse models and reveal the mode of action for the different immunocytokines. Based on these findings, we initiated a phase I/II clinical trial in patients with recurrent glioma to investigate the targeted delivery of TNF-α (ClinicalTrials.gov identifier NCT03779230).

PL2.1 Exploiting the DNAM-1 system for chimeric antigen receptor (CAR) T cell therapy of glioblastoma

BACKGROUND

Cancer immunotherapy with genetically engineered T cells that express a chimeric antigen receptor (CAR) has led to impressive responses in extracranial malignancies and is also explored against glioblastoma. However, CAR T cell strategies that are currently being explored against glioblastoma target single tumor antigens, which are non-homogeneously expressed and are prone to antigen escape. Furthermore, the immunosuppressive brain tumor microenvironment hampers anti-tumor efficacy.

METHODS

By immunohistochemistry and flow cytometry, we investigated the expression of CD155 and CD112, which are ligands to the activating immune cell receptor DNAX accessory molecule-1 (DNAM-1), in human and mouse glioma cell lines as well as in human glioblastoma samples. To understand their functional role, we generated CD155 or CD112 knock-out glioma cell lines using CRISPR/Cas9 and studied proliferation, sensitivity to irradiation or temozolomide as well as migration. To exploit the promiscuous binding features of DNAM-1, we generated different first or second-generation CAR T cells that use DNAM-1 as a tumor-binding domain. Subsequently, we investigated their anti-tumor activity in vitro in co-culture assays and in vivo in syngeneic orthotopic murine glioma models.

RESULTS

CD155 and CD112 are homogenously expressed in human and mouse glioma cell lines and human glioblastoma tissues. Knock-out of these ligands affected the migration of tumor cells, but did not affect proliferation or sensitivity to irradition or temozolomide. DNAM-1-based CAR T cells demonstrated high cytolytic activity and effector cytokine secretion in vitro. In vivo, DNAM-1 based CAR T cells reached to the tumor site in the brain upon intravenous administration, prolonged survival of orthotopic glioma-bearing mice and led to a durable anti-tumor response in a fraction of mice. The treatment was tolerated without toxicities.

CONCLUSION

We elucidated the tumor-intrinisic role of CD155 and CD112 and provide the first systematical preclincal assessment of DNAM-1 CAR T cells against glioma. These findings provide a rationale to test this immunotherapeutic strategy also in human glioma patients.

Recruitment and inclusion procedures as "pain killers" in clinical trials?

Background

Recruitment and inclusion procedures in clinical trials are time critical. This holds particularly true for studies investigating patients with fluctuating symptom patterns, like those with chronic neck pain. In a feasibility study on neck pain, we found a clinically relevant decrease in pain ratings within the recruitment period. This paper analyses the phenomenon and gives recommendations for recruitment procedures in clinical trials on pain.

Methods

Changes in pain intensity scores of 44 chronic neck pain patients (6 males and 36 females; mean age: 45.3±13.2 years) between the first telephone contact and baseline assessment were analyzed. Inclusion criterion was a mean pain intensity of ≥40 on a 0-100 numerical rating scale during the last three months. Statistical analyses were performed using ANOVA and parametric/non-parametric correlation coefficients.

Results

Average pain intensity score decreased significantly from 60.3±13.3 at telephone interview to 38.1±21.7 at baseline assessment. This represents a relative change of 36.8%. A weak but significant negative correlation was found between number of days between assessments and pain rating differences. There was a positive correlation between change of pain intensity and the pain level at the first contact, indicating that the decreased pain ratings over time were also dependent on the initial pain rating.

Conclusions

The clinically significant changes in pain intensity were weakly related to waiting time and moderately dependent on initial pain intensity, suggesting regression to the mean. The natural course of the disease and the Hawthorne effect are also discussed as contributing factors.

Quasinormal mode solvers for resonators with dispersive materials

Optical resonators are widely used in modern photonics. Their spectral response and temporal dynamics are fundamentally driven by their natural resonances, the so-called quasinormal modes (QNMs), with complex frequencies. For optical resonators made of dispersive materials, the QNM computation requires solving a nonlinear eigenvalue problem. This raises a difficulty that is only scarcely documented in the literature. We review our recent efforts for implementing efficient and accurate QNM solvers for computing and normalizing the QNMs of micro- and nanoresonators made of highly dispersive materials. We benchmark several methods for three geometries, a two-dimensional plasmonic crystal, a two-dimensional metal grating, and a three-dimensional nanopatch antenna on a metal substrate, with the perspective to elaborate standards for the computation of resonance modes.

Analytic Mode Normalization for the Kerr Nonlinearity Parameter: Prediction of Nonlinear Gain for Leaky Modes

Based on the resonant-state expansion with analytic mode normalization, we derive a general master equation for the nonlinear pulse propagation in waveguide geometries that is valid for bound and leaky modes. In the single-mode approximation, this equation transforms into the well-known nonlinear Schrödinger equation with a closed expression for the Kerr nonlinearity parameter. The expression for the Kerr nonlinearity parameter can be calculated on the minimal spatial domain that spans only across the regions of spatial inhomogeneities. It agrees with previous vectorial formulations for bound modes, while for leaky modes the Kerr nonlinearity parameter turns out to be a complex number with the imaginary part providing either nonlinear loss or even gain for the overall attenuating pulses. This nonlinear gain results in more intense pulse compression and stronger spectral broadening, which is demonstrated here on the example of liquid-filled capillary-type fibers.

Relieving phantom limb pain with multimodal sensory-motor training

OBJECTIVE

The causes for the disabling condition of phantom limb pain (PLP), affecting 85% of amputees, are so far unknown, with few effective treatments available. Sensory feedback based strategies to normalize the motor commands to control the phantom limb offer important targets for new effective treatments as the correlation between phantom limb motor control and sensory feedback from the motor intention has been identified as a possible mechanism for PLP development.

APPROACH

Ten upper-limb amputees, suffering from chronic PLP, underwent 16 days of intensive training on phantom-limb movement control. Visual and tactile feedback, driven by muscular activity at the stump, was provided with the aim of reducing PLP intensity.

MAIN RESULTS

A 32.1% reduction of PLP intensity was obtained at the follow-up (6 weeks after the end of the training, with an initial 21.6% reduction immediately at the end of the training) reaching clinical effectiveness for chronic pain reduction. Multimodal sensory-motor training on phantom-limb movements with visual and tactile feedback is a new method for PLP reduction.

SIGNIFICANCE

The study results revealed a substantial reduction in phantom limb pain intensity, obtained with a new training protocol focused on improving phantom limb motor output using visual and tactile feedback from the stump muscular activity executed to move the phantom limb.

Analytical mode normalization and resonant state expansion for bound and leaky modes in optical fibers - an efficient tool to model transverse disorder

We adapt the resonant state expansion to optical fibers such as capillary and photonic crystal fibers. As a key requirement of the resonant state expansion and any related perturbative approach, we derive the correct analytical normalization for all modes of these fiber structures, including leaky modes that radiate energy perpendicular to the direction of propagation and have fields that grow with distance from the fiber core. Based on the normalized fiber modes, an eigenvalue equation is derived that allows for calculating the influence of small and large perturbations such as structural disorder on the guiding properties. This is demonstrated for two test systems: a capillary fiber and a photonic crystal fiber.

Team Players or Opponents: Coadministration of Selective Glucagon and GLP-1 Receptor Agonists in Obese Diabetic Monkeys

We assessed the therapeutic contribution of the individual components of glucagon-like peptide-1 receptor (GLP-1R) and glucagon receptor (GCGR) agonists alone and in combination upon energy homeostasis and glycemic control in diet-induced obese, diabetic nonhuman primates. The pharmacological active dose ranges of selective agonists were established through a dose-finding study, followed by a 6-week chronic study. Repeated subcutaneous administration of a selective GCGR agonist (30 µg/kg once daily) did not affect food intake or body weight, whereas the selective GLP-1R agonist (3 µg/kg once daily) alone decreased energy intake by 18% and body weight by 3.8% ± 0.9%. Combination of both agonists reduced significantly cumulative food intake by 27% and body weight by 6.6% ± 0.9%. Fasting plasma glucose (FPG) was improved by GLP-1R agonist (baseline vs end of study, 176.7 ± 34.0 vs 115.9 ± 16.1 mg/dL). In contrast, groups exposed to GCGR agonist experienced nonsignificant elevations of FPG. More accurate assessment of therapeutic interventions on glucose homeostasis was tested by an IV glucose tolerance test. Glucose excursion was significantly elevated by chronic GCGR agonist administration, whereas it was significantly decreased in GLP-1R agonist-treated monkeys. In the combination group, a nonsignificant increase of glucose excursion was seen, concomitantly with significantly increased insulin secretion. We conclude that chronic glucagon agonism does not affect energy homeostasis in nonhuman primates. In combination with GLP-1R agonism, glucagon agonism synergistically enhances negative energy balance with resulting larger body weight loss. However, adding GCGR to GLP-1R agonism diminishes glycemic control in diabetic monkeys. Therefore, long-term therapeutic implications of using GLP-1R/GCGR coagonists for weight management in diabetes warrants further scrutiny.

Running on mixed fuel-dual agonistic approach of GLP-1 and GCG receptors leads to beneficial impact on body weight and blood glucose control: A comparative study between mice and non-human primates

AIM

We performed acute and chronic studies in healthy and diet-induced obese animals using mouse-specific or monkey-specific dual GLP-1R/GCGR agonists to investigate their effects on food intake, body weight, blood glucose control and insulin secretion. The selective GLP-1R agonist liraglutide was used as comparator.

METHODS

The mouse-specific dual agonist and liraglutide were tested in lean wild type, GLP-1R knockout and diet-induced obese mice at different doses. A chronic study was performed in DIO mice to investigate the effect on body weight, food consumption and total energy expenditure (TEE) in obese and diabetic monkeys with a focus on body weight and energy intake.

RESULTS

The mouse-specific dual agonist and liraglutide similarly affected glycaemic control. A higher loss in body weight was measured in dual agonist-treated obese mice. The dual agonist significantly enhanced plasma glucose excursion in overnight fed GLP-1R^-/- mice, probably reflecting a potent GCGR agonist activity. It increased TEE and enhanced fat and carbohydrate oxidation, while liraglutide produced no effect on TEE. In obese and diabetic monkeys, treatment with the monkey-specific dual agonist reduced total energy intake to 60%-70% of baseline TEI during chronic treatment. A decrease in body weight and significant improvement in glucose tolerance was observed.

CONCLUSIONS

In DIO mice and non-human primates, dual agonists elicited robust glycaemic control, similar to the marketed GLP-1R agonist, while eliciting greater effects on body weight. Results from DIO mice suggest that the increase in TEE is caused not only by increased fat oxidation but also by an increase in carbohydrate oxidation.

Modeling of second-harmonic generation in periodic nanostructures by the Fourier modal method with matched coordinates

We present an advanced formulation of the Fourier modal method for analyzing the second-harmonic generation in multilayers of periodic arrays of nanostructures. In our method, we solve Maxwell's equations in a curvilinear coordinate system, in which the interfaces are defined by surfaces of constant coordinates. Thus, we can apply the correct Fourier factorization rules as well as adaptive spatial resolution to nanostructures with complex cross sections. We extend here the factorization rules to the second-harmonic susceptibility tensor expressed in the curvilinear coordinates. The combination of adaptive curvilinear coordinates and factorization rules allows for efficient calculation of the second-harmonic intensity, as demonstrated for one- and two-dimensional periodic nanostructures.

Resonant-state expansion for open optical systems: generalization to magnetic, chiral, and bi-anisotropic materials

The resonant-state expansion, a recently developed powerful method in electrodynamics, is generalized here for open optical systems containing magnetic, chiral, or bi-anisotropic materials. It is shown that the key matrix eigenvalue equation of the method remains the same, but the matrix elements of the perturbation now contain variations of the permittivity, permeability, and bi-anisotropy tensors. A general normalization of resonant states in terms of the electric and magnetic fields is presented.

Chronic intensive care-related pain: Exploratory analysis on predictors and influence on health-related quality of life

BACKGROUND

There is growing evidence for the development of chronic pain after intensive care. Nonetheless, there is only limited knowledge about factors leading to chronic intensive care-related pain (CIRP). Thus, the primary objective was the identification of predictors of CIRP. Moreover, we aimed to assess the impact of CIRP on patients' health-related quality of Life (HRQOL).

METHODS

Comprehensive information on patients' pain before ICU admission and present pain was collected longitudinally by means of the German Pain Questionnaire 6 and 12 months after ICU discharge (ICU_DC ). In addition, a subsample of patients underwent Quantitative Sensory Testing (QST). We used Generalized Estimating Equations to identify predictors of CIRP with logistic regression models.

RESULTS

In total, 204 patients (197/159 at 6/12 months after ICU_DC ) were available for the analyses. In the multivariate models, moderate to severe average pain in the 4 weeks after ICU_DC , lower age, female sex, increased inflammation and chronic pain conditions and increased levels of anxiety before ICU admission were predictive for CIRP. In addition, small fibre deficits and lower disease severity were associated with CIRP in the QST subsample (81 patients, 77/55 at 6/12 months after ICU_DC ). Patients with CIRP reported significantly lower HRQOL than patients without CIRP.

CONCLUSIONS

Chronic intensive care-related pain is associated with specific decrements in HRQOL. Knowledge about the identified predictors is of clinical and scientific importance and might help to reduce the incidence of CIRP.

SIGNIFICANCE

Chronic intensive care-related pain is associated with specific decrements in health-related quality of life. While most of the identified predictors for CIRP can only be considered as risk factors, especially adequate (post-) acute pain management should be studied as preventive strategy.

From Dark to Bright: First-Order Perturbation Theory with Analytical Mode Normalization for Plasmonic Nanoantenna Arrays Applied to Refractive Index Sensing

We present a first-order perturbation theory to calculate the frequency shift and linewidth change of photonic resonances in one- and two-dimensional periodic structures under modifications of the surrounding refractive index. Our method is based on the resonant state expansion, for which we extend the analytical mode normalization to periodic structures. We apply this theory to calculate the sensitivity of bright dipolar and much darker quadrupolar plasmonic modes by determining the maximum shift and optimal sensing volume.

Discovery of N-[4-(1H-Pyrazolo[3,4-b]pyrazin-6-yl)-phenyl]-sulfonamides as Highly Active and Selective SGK1 Inhibitors

From a virtual screening starting point, inhibitors of the serum and glucocorticoid regulated kinase 1 were developed through a combination of classical medicinal chemistry and library approaches. This resulted in highly active small molecules with nanomolar activity and a good overall in vitro and ADME profile. Furthermore, the compounds exhibited unusually high kinase and off-target selectivity due to their rigid structure.

Surface- and tip-enhanced resonant Raman scattering from CdSe nanocrystals

Surface- and tip-enhanced resonant Raman scattering (resonant SERS and TERS) by optical phonons in a monolayer of CdSe quantum dots (QDs) is demonstrated. The SERS enhancement was achieved by employing plasmonically active substrates consisting of gold arrays with varying nanocluster diameters prepared by electron-beam lithography. The magnitude of the SERS enhancement depends on the localized surface plasmon resonance (LSPR) energy, which is determined by the structural parameters. The LSPR positions as a function of nanocluster diameter were experimentally determined from spectroscopic micro-ellipsometry, and compared to numerical simulations showing good qualitative agreement. The monolayer of CdSe QDs was deposited by the Langmuir-Blodgett-based technique on the SERS substrates. By tuning the excitation energy close to the band gap of the CdSe QDs and to the LSPR energy, resonant SERS by longitudinal optical (LO) phonons of CdSe QDs was realized. A SERS enhancement factor of 2 × 10(3) was achieved. This allowed the detection of higher order LO modes of CdSe QDs, evidencing the high crystalline quality of QDs. The dependence of LO phonon mode intensity on the size of Au nanoclusters reveals a resonant character, suggesting that the electromagnetic mechanism of the SERS enhancement is dominant. Finally, the resonant TERS spectrum from CdSe QDs was obtained using electrochemically etched gold tips providing an enhancement on the order of 10(4). This is an important step towards the detection of the phonon spectrum from a single QD.