ESTRO 2020 Abstract Book

S28 ESTRO 2020

radio(chemo)therapy and a distinct molecular pathogenesis. microRNAs (miRNAs) act as important post- transcriptionalregulators and were shown to be of prognostic relevance in HNSCC. The aim of the study was to characterize the molecular phenotype of HPV-positive and -negative HNSCC at the miRNA and transcriptome levels. Material and Methods The HPV characterized multicentre DKTK-ROG cohort (n=128), the monocentric LMU-KKG cohort (n=102) and The Cancer Genome Atlas (TCGA) HNSCC cohort (n=244) were included. HPV status of LMU-KKG and DKTK-ROG specimens was determined using p16 immunohistochemistry in combination with HPV DNA analysis. For DKTK-ROG and TCGA, HPV status was provided by the DKTK-ROG consortium and by the GDC data portal, respectively. All LMU-KKG and DKTK-ROG patients had undergone surgical resection followed by adjuvant radio(chemo)therapy. The TCGA cohort was unselected regarding treatment modalities. For DKTK-ROG and LMU-KKG Agilent miRNA microarray data were generated, for TCGA the data were downloaded. Using lasso machine learning, a generalized linear model predicting HPV status was trained in the DKTK-ROG miRNA data set and its performance assessed in the LMU-KKG and TCGA HNSCC validation data. The prognostic value of the miRNA signature was compared to clinically defined HPV status. A subset of samples with highest (n=10) and lowest (n=10) HPV prediction scores from the LMU-KKG cohort was subjected to whole RNA sequencing followed by Gene Set Enrichment Analysis The optimal prediction model contained 24 miRNAs and predicted HPV status in the DKTK-ROG cohort with a sensitivity of 89.5% and a specificity of 96.7% (AUC: 0.94). In LMU-KKG the signature predicted HPV with a sensitivity of 73.9% and a specificity of 88.6% (AUC: 0.79) and in TCGA with a sensitivity of 48.3% and a specificity of 91.2% (AUC: 0.68). The prognostic value of the predicted HPV status with regard to all survival and disease control endpoints was comparable to that of p16-based HPV status. Gene set enrichment analysis showed an up-regulation of JAK STAT signaling pathway, chemokine signaling pathway, cytokine-cytokine receptor interaction in HPV-positive tumors and up-regulation of epithelial mesenchymal transition in HPV-negative tissues. Conclusion In our study we identified a robust miRNA signature predicting HPV status, which could be validated in two independent cohorts with a similar prognostic value compared to that of clinically established HPV typing. Specific pathways were shown to be associated with miRNA-derived HPV prediction score. The data provide insights into the specific molecular pathogenesis of HPV- driven and classical HNSCC tumors. PD-0067 An ex vivo assay to detect radiosensitization by PARP-inhibitors in soft tissue sarcomas M. Mangoni 1 , G. Salvatore 1 , D. Greto 2 , M. Sottili 1 , C. Talamonti 1 , V. Lorenzetti 1 , M. Aquilano 1 , A. Peruzzi 1 , V. Salvestrini 1 , L. Visani 1 , P. Bonomo 1 , M. Scorianz 3 , F. Muratori 3 , G. Scoccianti 3 , D.A. Campanacci 4 , L. Livi 1 1 University of Florence, Biomedical Experimental and Clinical Sciences, Firenze, Italy ; 2 Careggi Hospital, Radiotherapy, Firenze, Italy ; 3 Careggi Hospital, Orthopedic Oncology, Firenze, Italy ; 4 University of Florence, Health Sciences, Firenze, Italy Purpose or Objective Soft tissue sarcomas (STSs) are aggressive tumors with a poor prognosis that take advantage of radiotherapy. Poly(ADP-ribose) polymerase (PARP)-1 inhibitors (PARPi) enhance the cytotoxic effects of radiation. One of the most sensitive methods for monitoring radiation-induced (GSEA). Results

DNA damage is detection of gamma-H2AX foci. In vitro studies in rhabdomyosarcoma showed that olaparib enhanced radiosensitivity on clonogenic assays and increased the formation of radiation-induced gamma-H2AX foci. Recently a novel method using residual gamma-H2AX foci in ex vivo irradiated tumor specimens has been developed for clinical application. The aim of the study was to assess the radiosensitizing effect of PARPi on human STSs specimens by adopting a modified ex vivo culture assays. Material and Methods Fresh tumor material was retrieved from surgical specimens from three patients with STSs. Tumors were manually cut and incubated in medium by considering 6 conditions: control, olaparib (1µM), irradiation with a single dose of 2 or 6 Gy, with or without olaparib (1µM). After irradiation, specimens were fixed in 4% formaldehyde and embedded into paraffin. In the order to evaluate the impact of olaparib inhibition on gamma-H2Ax foci formation immunofluorescence analyses were performed. Results Pathology reports revealed that samples derived from spindle cells sarcoma (2) and pleomorphic sarcoma (1). We observed that olaparib and irradiation (at both 2 and 6 Gy doses) significantly increased the mean immunofluorescence of gamma-H2AX tumor samples as compared to control in all the samples. Moreover, the combination of olaparib and radiation resulted in an increased number of gamma-H2AX foci as compared to irradiation alone, at the 6 Gy dose. Results were not different in the two histological subtypes analyzed. Ex vivo data confirmed our previous in vitro and in vivo data. Conclusion Ex vivo analysis have demonstrated that PARPi enhance radiosensitivity in STSs and have a potential role in association with radiotherapy for treatment of these aggressive tumors. We are currently recruiting more samples for analysis. Since STSs are characterized by genetic variability that overcame histological typing, we expect to find an heterogeneous response in different STSs patients. Thus, gamma H2AX analysis on ex vivo samples could represent a practical and valid test to predict response of combined treatments in order to personalize treatment. PD-0068 Streamlining the image-guided radiotherapy process for proton beam therapy: a service evaluation L. Davies 1 , L. McHugh 2 , M.C. Aznar 3 , J. Lindsay 3 , C.L. Eccles 2 1 The Christie NHS Foundation Trust, Proton Beam Therapy, Manchester, United Kingdom ; 2 The Christie NHS Foundation Trust, Radiotherapy, Manchester, United Kingdom ; 3 The University of Manchester, Division of Cancer Services, Manchester, United Kingdom Purpose or Objective Modern radiotherapy requires image-guidance to ensure precision and accuracy of treatment delivery. To implement image-guided proton beam therapy (IGPBT) at our Proton Beam Therapy (PBT) Centre, a 3-step process was developed. This study reports on an evaluation to assess the feasibility of reducing this to a 2-step imaging process, thereby reducing overall treatment time and unnecessary imaging doses. Material and Methods The first 20 consecutive patients treated at the Centre were included in the study. Patient characteristics are recorded in Table 1. The initial IGPBT workflow consisted of: (i) 2-dimensional kilo-voltage (2DkV) image pair acquisition for gross positioning assessment, (ii) 3- dimensional cone-beam computed tomography (CBCT) Poster discussion: RTT 1