Abstract Book

S141

ESTRO 37

Germany 24 Clinical Cooperation Unit Radiation Oncology, University of Heidelberg Medical School and German Cancer Research Center DKFZ, Heidelberg, Germany 25 German Cancer Research Center DKFZ - Heidelberg - and German Cancer Consortium DKTK, partner site Munich, Munich, Germany 26 Department of Radiation Oncology, Ludwig- Maximilians-Universität, Munich, Germany 27 Clinical Cooperation Group - Personalized Radiotherapy in Head and Neck Cancer, Helmholtz Zentrum Munich, Munich, Germany 28 Department of Radiation Sciences DRS - Institut für Innovative Radiotherapie iRT, Helmholtz Zentrum Munich, Neuherberg, Germany 29 Department of Radiation Oncology, Technische Universität München, Munich, Germany 30 German Cancer Research Center DKFZ - Heidelberg - and German Cancer Consortium DKTK, partner site Tübingen, Tübingen, Germany 31 Department of Radiation Oncology, Faculty of Medicine and University Hospital Tübingen - Eberhard Karls Universität Tübingen, Tübingen, Germany 32 Tumour- and Normal Tissue Bank, University Cancer Centre UCC - University Hospital Carl Gustav Carus - Technische Universität Dresden, Dresden, Germany 33 Medical Systems Biology, University Cancer Centre UCC - University Hospital Carl Gustav Carus - Technische Universität Dresden, Dresden, Germany Purpose or Objective To compare and improve the performance of a hypothesis-driven 7-gene signature with a signature based on whole transcriptome analysis for the prognosis of loco- regional tumour control (LRC) in patients with HPV- negative locally advanced head and neck squamous cell carcinoma (HNSCC) after postoperative radiochemo- therapy (PORT-C). Material and Methods Gene expression analyses were performed on a multicentre retrospective cohort of 125 patients with HPV16 DNA negative HNSCC using the GeneChip® Human Transcriptome Array 2.0 (Affymetrix) for whole transcriptome analysis. To identify a gene signature prognostic for LRC from the whole transcriptome data, 3085 genes were considered, which previously have been related to radioresistance or response to radiotherapy [1- 4]. Internal cross validation was used to compare different signature sizes, feature selection algorithms and prognostic models and to identify the final gene signature. The performance of the whole transcriptome- based signature was compared to a previously identified 7-gene signature based on nanoString analysis of a hypothesis-driven gene set containing 171 genes, using the concordance index (ci). The signatures were applied independently to stratify patients into groups of low (LR) and high (HR) risk of recurrence. Finally, a combined high risk group was defined, including patients who were classified as high risk patients by both gene signatures. Results The identified gene signature based on whole transcriptome data showed improved performance (ci: 0.79-0.87) compared to the signature based on the hypothesis-driven gene set (ci: 0.72-0.78) and contained genes related to tumourigenesis, invasion, cell cycle regulation and immune response. Patient stratification into low and high risk groups was performed for both signatures, see figure A and B. The difference between LR and HR regarding LRC was highly significant (p<0.001) between both groups. Compared to the 7-gene nanoString signature, the LR group showed a slightly improved LRC for the Affymetrix-based signature, similar to that of HPV positive tumours. The combined high risk group showed a poor LRC of only about 45%, see figure C.

Conclusion We determined a gene signature predicting LRC in a patient cohort with HPV16 DNA negative HNSCC after PORT-C based on whole transcriptome analysis. The signature showed improved performance compared to the 7-gene signature based on a limited hypothesis-driven gene set, indicating that additional genes of high prognostic value were identified. More importantly, the combination of both models allowed for the identification of a subgroup of patients with HPV-negative HNSCC who are on a particularly high risk of developing a recurrence, and may be considered for dose-escalation trials in future. [1] Subramanian et al. Proc Natl Acad Sci U S A. 2005;102:15545-50. [2] Mootha et al. Nat Genet. 2003;34:267-73. [3] Ashburner et al. Nat Genet. 2000;25:25-9. [4] The Gene Ontology Consortium. Nucleic Acids Res. 2015;43:D1049-56. OC-0277 Development and validation of distant metastases risk group classification in oral cavity cancer A. Hosni 1 , S.H. Huang 1 , K. Chiu 1 , W. Xu 2 , J. Su 2 , L. Tong 1 , A. Bayley 1 , S. Bratman 1 , J. Cho 1 , M. Giuliani 1 , J. Kim 1 , B. O’Sullivan 1 , J. Ringash 1 , J. Waldron 1 , J. De Almeida 3 , D. Chepeha 3 , D. Goldstein 3 , A. Hope 1 1 Princess Margaret Cancer Centre, Department of radiation oncology, Toronto, Canada 2 Princess Margaret Cancer Centre, Department of biostatistics, Toronto, Canada 3 Princess Margaret Cancer Centre, Department of Otolaryngology-Head & Neck Surgery/Surgical Oncology, Toronto, Canada Purpose or Objective Distant metastasis (DM) is a major determinant of prognosis in oral cavity squamous cell carcinoma (OSCC). The objective of this study was to define different prognostic groups with regard to DM following postoperative intensity modulated radiotherapy (PO- IMRT). Material and Methods Retrospective review was conducted for OSCC patients treated with curative intent at our institution with PO- IMRT. All patients were staged according to the UICC/AJCC 7 th edition TNM. Two sequential cohorts of OSCC patients compromised the discovery (2005-2012) and validation cohorts (2013-2014). In the discovery cohort, a set of variables was evaluated by multivariable analysis as potential predictors of DM, including: pT- category, pN-category, resection margin status, extranodal extension, histological grade, lymphovascular invasion, perineural invasion, tumor thickness, and use of adjuvant concurrent chemotherapy. We used the competing risk regression method to derive risk-group classification and compared distant control (DC) and overall survival (OS) according to the derived risk-group in the discovery cohort and subsequently in the validation cohort.