ESTRO 36 Abstract Book

S227 ESTRO 36 2017 _______________________________________________________________________________________________

represses transcription of genes within its boundaries. Chromothripsis and kataegis is evident in one fifth of these tumours and can be associated with more aggressive disease. Using driver mutations, copy-number alterations and methylation, we were able to categorize patients into prognostic categories which has less than 5% or greater than 50% probability of relapse. The use of genomic markers as prognostic factors for local failure and for systemic disease are therefore novel risk-stratification tools which help to triage patients to existing treatment options, and potentially identification of molecular targets for therapy. However, our data also suggest that novel therapeutic approaches focus on recurrent non- mutation targets. This new approach could then prevent castrate-resistance by targeting genetic instability earlier on the natural history of the disease when fewer driver events are operational. OC-0436 Cytokine-dependent regulation of prostate cancer stem cell maintenance in response to irradiation C. Peitzsch 1,2 , M. Baumbach 1 , M. Cojoc 1 , L. Hein 1 , I. Kurth 1,2 , M. Baumann 1,2,3,4,5 , M. Krause 1,3,4,5 , A. Dubrovska 1,3,4 1 OncoRay - Center for Radiation Research in Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus- Technische Universität Dresden, Dresden, Germany 2 National Center for Tumor Diseases NCT, German Cancer Research Center DKFZ, Dresden, Germany 3 German Cancer Consortium DKTK, German Cancer Research Center DKFZ, Dresden, Germany 4 Institute of Radiation Oncology, Helmholtz-Zentrum Dresden-Rossendorf, Dresden, Germany 5 Department of Radiation Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus- Technische Universität Dresden, Dresden, Germany Purpose or Objective According to the cancer stem cell hypothesis prostate cancer is driven by a malignant subpopulation with stem- like properties. These cancer stem cells (CSC) contribute to tumor-initiation, metastasis, therapy-resistance and tumor relapse. In parallel, genetic mutations accumulate over time and CSC subclones evolve. Therapeutic interventions like radiotherapy provide selective pressure for the expansion of resistant subclones with genetic diversification. We hypothesize that the determination of CSC-related biomarker in prostate cancer biopsies is correlating with clinical parameter and can be used for patient stratification and treatment selection to improve personalized radiotherapy. Material and Methods We generated isogenic radioresistant prostate cancer cell lines with a high expression of CSC marker, a epithelial- to-mesenchymal transition (EMT) phenotype, higher self- renewal properties, higher tumorigenicity and enhanced DNA repair capacity. We applied comparative genomic, proteomic, metabolomic, epigenomic and secretome analysis to identify novel biomarker for prostate cancer radioresistance and to unravel contributing molecular mechanisms. Novel biomarkers were validated using the Cancer Genome Atlas (TCGA) database and correlated with the tumor-free survival of prostate cancer patients after anti-cancer therapy including radiotherapy using SUMO software calculation. Results Within our first proof-of-principle study, we could show that ALDH-positive CSCs are radioresistant and maintained directly by the Wnt/β-catenin signaling pathway. In addition, we found that irradiation is inducing CSC marker and CSC properties in a dose- and time-dependent manner. This irradiation-induced CSC-plasticity was attributed to the modulation of the histone methylation code. Within the present study we analyzed a panel of

secreted cytokines and their corresponding cytokine receptors in the radioresistant prostate cancer sublines, in a s.c. xenotransplantation model, in ex vivo irradiated primary prostate cancer biopsies and in blood samples of prostate cancer patients during the course of radiotherapy and found, for example, the CXCR4-CXCL12 signaling to be involved in the CSC maintenance and the induction of prostate cancer radioresistance. Conclusion Our studies suggest that the combination of irradiation with cytokine signaling modulation, especially the CXCR4- CXCL12 signaling, may increase the cytotoxic effects of irradiation in prostate cancer cells. The expression profiling of proteins involved in the cytokine signaling can be used to predict clinical outcome of prostate cancer patients after radiotherapy. OC-0437 Scatter imaging: promising modality for image guided ablation radiotherapy for lung cancer patients J. Chu 1 , G. Redler 1 , G. Cifter 1 , K. Jones 1 , J. Turian 1 1 Rush University Medical Center, Department of Radiation Oncology, Chicago IL, USA Purpose or Objective Early stage lung cancers can be effectively treated by stereotactic ablation radiation therapy (SABR). Successful treatment requires hypofractionation and large dose per fraction (up to 20 Gy) while maintaining a high level of accuracy (≤1.0mm). By imaging the photons that are Compton-scattered out of the treatment beam, real-time, non-invasive monitoring of the tumor location may be possible. To assess the potential of this modality, we have obtained scatter images of static and movable tumor phantoms, and calculated images from CT-based Monte Carlo simulations. Material and Methods Compton scatter is a natural by-product of external beam radiation therapy. The scattered radiation contains information about the patient anatomy and the transient tumor location. An embedded tumor in a Quasar respiratory motion phantom (Modus Medical Devices Inc.) was programmed to move linearly over 2.5cm. While irradiating the embedded tumor using a 6MV Varian TrueBeam linear accelerator, experimental scatter images were measured with a Varian PaxScan flat panel detector and a pinhole collimator. Tumor centroid locations were then measured from various scatter images and compared with the expected values. Monte Carlo N-Particle (MCNP) code was used to simulate scatter images from phantoms and patient CT images using 10 - 1000MU, or 0.5 – 50 second time scales. The quality of the images was assessed to determine their potential for tumor localization during treatment. Results The measured tumor centroid locations agreed with the expected values to within 1mm, which is adequately accurate for clinical tumor tracking. Lung tumor phantom images showed excellent signal and contrast. The contrast-to-noise ratio ranged from 3.4 to 15.1 for scatter images acquired with 0.5 to 50s. The attached figures below show CT and simulated scatter images (corresponding to the red shaded region in CT) for both inhale and exhale breathing phases. The scatter images clearly show variation of tumor and diaphragm locations for two breathing phases. Other pertinent anatomical structures, such as chest wall, heart, and lung are also clearly visible. Conclusion This study has demonstrated the feasibility of using scatter imaging to track lung tumor movement during Proffered Papers: New technologies for imaging and therapy