Abstract Book

S1270

ESTRO 37

outcomes not only in prostate cancer treatment, but other cancer treatments as well.

and proliferation (EdU incorporation). Our data showed no difference in the induction of DSBs but significantly higher number of residual gH2AX/53BP1 foci in ERG- overexpressing patients compared to ERG low expressers. Conclusion Collectively, these data indicate that ERG-overexpression in PCa drives a switch to the PARP1-EJ which in turn renders cells radiosensitized by PARP inhibition. Clinically, this would rationalize the combination of PARP inhibitor olaparib with radiotherapy in ERG-positive PCa patients.

EP-2301 ERG overexpression in PCa drives a switch to the alt-EJ which render cells radiosensitive to PARPi S. Köcher 1 , J. Volquardsen 1 , L. Nordquist 1 , T. Lange 2 , T. Schlomm 3 , B. Beyer 3 , C. Petersen 4 , K. Rothkamm 1 , W. Mansour 1 1 University Medical Center Hamburg - Eppendorf UKE, Laboratory for Radiobiology, Hamburg, Germany 2 University Medical Center Hamburg - Eppendorf UKE, Anatomy, Hamburg, Germany 3 University Medical Center Hamburg - Eppendorf UKE, Martini-Klinik, Hamburg, Germany 4 University Medical Center Hamburg - Eppendorf UKE, Radiotherapy, Hamburg, Germany Purpose or Objective Prostate cancer (PCa) is the leading cause of death worldwide in men. PCa typically shows gross structural rearrangements, including translocations and deletions of chromosomal material. The most frequent translocation involves the androgen-responsive gene TMPRSS2 and the ETS-transcription factor ERG. Consequently, ERG comes under the control of the constitutively active TMPRSS2 promoter leading to strong overexpression of ERG. ERG overexpression is reported in about 50% of PCa patients and was found to correlate with worse prognosis, which shed light on the need to improve the survival of these patients. Material and Methods reporter plasmid assay, colony formation assay, DSB repair analysis using immunofluorescence (foci), functional ex-vivo assay to analyse DSB induction and repair in fresh tumor tissues (xenograft and Patient material) Results Here we show that ectopic ERG overexpression leads to deregulation in double strand break (DSB) repair. Using a reporter plasmid assay we could show that about 30-40% of DSBs are repaired by the alternative PARP1-dependent end joining (PARP1-EJ). These data were further confirmed by monitoring the induction and repair of DSB (gH2AX/53BP1 foci) after ionizing radiation (IR) upon treatment with the PARP inhibitor (PARPi) olaparib. Our results revealed that PARP inhibition significantly increased the number of residual gH2AX/53BP1 foci after IR exclusively in ERG overexpressing cells. Consequently, olaparib causes a selective radiosensitization in ERG overexpressing cells. In order to further validate our data, we established a functional ex-vivo assay, which allows the analysis of DSB induction and repair in fresh tumor tissues. According to this system, tumor slices can be kept under optimum growth conditions for up to 4 weeks. During this time, tumor slices were irradiated and the overall DSB repair measured using immunofluorescence staining of gH2AX/53BP1. Firstly, we validated our assay using xenograft tumor models of a responder and a non- responder cell line to PARPi as radiosensitizer. We could clearly confirm the deficiency in DSB repair in the responder cell line as evidenced by elevated residual gH2AX/53BP1 foci after PARP inhibition. In collaboration with the Martini-Clinic we collected tumor tissues from different PCa patients. Tissues were slices in 100-300µm to prevent hypoxia, and tissues were treated with or without the PARPi olaparib 2h-pre irradiation with 2Gy. Indeed, optimum growth conditions were validated by investigating hypoxia (Pimonidazol)

Electronic Poster: Radiobiology track: Radiobiology of breast cancer

EP-2302 Differences in the Radiosensitivty Index (RSI) Between Metastatic and Primary Breast Cancer C. Liveringhouse 1 , K. Ahmed 2 , J. Purcell 2 , M. Mills 1 , G.D. Grass 2 , A. Orman 2 , J. Torres-Roca 2 , R. Diaz 2 1 University of South Florida, Morsani College of Medicine, Tampa, USA 2 H. Lee Moffitt Cancer Center, Radiation Oncology, Tampa, USA Purpose or Objective We have previously identified and validated a multigene expression index for tumor radiosensitivity (RSI) in over 2200 samples. This model predicts a radiosensitivity index (RSI) that is directly proportional to tumor radioresistance, (RSI, high index = radioresistance). Analyses from our group have revealed RSI to be significant in predicting outcomes in three independent datasets of breast cancer patients treated with radiation therapy. In this study, we assessed the RSI of primary and metastatic breast samples from our institution’s tissue biorepository. Material and Methods Patients were identified from our institutional IRB- approved prospective observational protocol. A total of 22 metastatic and 244 primary lesions were obtained from a de-identified meta-data pool. Gene expression was from Affymetrix Hu-RSTA-2a520709. The previously tested RSI 10 gene assay was run on tissue samples ranked according to gene expression. RSI was calculated using the previously published ranked based algorithm. Results When assessing all primary lesions with metastatic lesions we found primaries to be more radioresistant. The median RSIs were 0.41 (Q1 0.32; Q3 0.50) compared to 0.23 (Q1 0.18; Q3 0.47) p=0.005, respectively. Utilizing a cutpoint of ≥ 0.3745 across all tissue samples in our database, primary breast lesions are radioresistant. A total of 214 samples were noted to be invasive ductal and 33 noted to be invasive lobular. The median RSI of invasive ductal lesions was 0.38 (Q1 0.3; Q3 0.49) and invasive lobular 0.48 (Q1 0.32; Q3 0.53), p=0.06. Conclusion In this analysis assessing radiosensitivity between primary and metastatic tissues of breast cancer histology, significant differences were noted. In addition, invasive lobular samples were noted to be more radioresistant than invasive ductal. These findings may help inform radiation dose selection.