Abstract Book

S253

ESTRO 37

efficacy in cell line based and PDX mice models by tumor volume measurement and by MRI. Results We found that suppression of different DNA repair pathways showed differences in RT response in HPV- and HPV+ HNSCC cells. Inhibition of base excision repair (BER) led to a better RT response in HPV+ HNSCC, whereas inhibition of nucleotide excision repair (NER) had a more profound effect in HPV- HNSCC. Inhibition of mismatch repair (MMR), Fanconi Anemia (FA) repair, and non- homologous end joining (NHEJ) showed a strong radiosensitization in both groups. Among the BER genes, PARPs are the most promising hits as PARP inhibitors are currently at advanced clinical development and were therefore validated in HNSCC cells. In line with the screen, PARP inhibition (ABT-888) in combination with RT resulted in a decreased clonogenic survival of HPV+ cells, but not HPV- cells. The analysis of double-strands break repair showed that the combination treatment slowed down DNA repair in HPV+ cells, whereas only slight delay in DNA repair was seen in HPV- cells. Expression analysis of HNSCC in TCGA database revealed significant differences in BER genes between both groups with HPV+ HNSCC showing higher expression of BER genes compared to HPV- HNSCC. Most of the hits overlapping between HPV- and HPV+ cells were involved in the NHEJ pathway and inhibition of this pathway by the DNAPK inhibitor NU7741 radiosensitized both groups of HNSCC in vitro and in both recurrent and primary HNSCC PDX models. Conclusion Our results show the preclinical potential of several DNA repair inhibitors in HPV- and HPV+ HNSCC. With multiple DNA repair inhibitors now in the clinical pipeline it is important to translate preclinical results into clinical trial designs which will guide clinicians to better patient selection and ultimately to improvements in treatment response OC-0492 A pooled analysis of radiotherapy dose & targeting strategies for stage 3 non-small cell lung cancer S. Schild 1 , W. Fan 2 , T. Stinchcombe 3 , E. Vokes 4 , S. Ramalingam 5 , J. Bradley 6 , K. Kelly 7 , H. Pang 8 , X. Wang 2 1 Mayo Clinic, Radiation Oncology, phoenix, USA 2 Duke University, Alliance Statistics and Data Center, Durham, USA 3 Duke University, Medical Oncology, Durham, USA 4 University of Chicago, Medical Oncology, Chicago, USA 5 Emory University, Medical Oncology, Atlanta, USA 6 Washington University, Radiation Oncology, St. Louis, USA 7 University of California, Medical Oncology, Davis, USA 8 Duke University, Biostatistics and Bioinformation, Durham, USA Purpose or Objective Concurrent chemoradiotherapy (CRT) is standard therapy for locally-advanced non–small-cell lung cancer (LA- NSCLC) patients. This study was performed to examine thoracic radiotherapy (TRT) parameters and their impact on patient survival. Material and Methods We collected Individual patient data (IPD) from 3600 LA- NSCLC patients participating in 16 cooperative group trials of concurrent CRT. The primary TRT parameters examined included field design strategy (elective nodal irradiation (ENI) compared to involved field TRT (IF- TRT)), total dose, and biologically effective dose (BED). Hazard ratios (HRs) for survival were estimated with Proffered Papers: CL 9: Lung

univariable and multivariable Cox models adjusted for significant risk factors at baseline. Results TRT doses ranged from 60 to 74 Gy with most treatments administered once-daily. ENI was associated with poorer survival than IF-TRT (univariable HR,1.37; 95%CI,1.24- 1.51, p<0.0001; multivariable HR,1.31; 95%CI,1.08-1.59, p=0.002). The median survival of the IF and ENI patients were 24 and 16 months, respectively. Patients were divided into 3 dose groups: low total dose (60 Gy), medium total dose (>60Gy-66Gy) and high total dose (>66Gy-74 Gy). The multivariable HRs relative to the low dose group were 1.08 for the medium dose group (CI=0.93-1.25), and 1.12 for the high dose group(CI=0.97- 1.30). The p-value for dose-group effect (2-df) was 0.17 from the multivariable model adjusted for field design strategy and significant risk factors. BED was grouped as follows: low (<55.5Gy 10 ), medium (=55.5-Gy 10 ), or high (>55.5 Gy 10 ). The multivariable HR relative to the low BED group were 1.00 (CI=0.85-1.18) for the medium BED group and 1.10 (CI=0.93-1.31) for the high BED group. The p-value for BED group effect (2-df) was 0.16 from the multivariable model. Conclusion IF-TRT was associated with significantly better survival than ENI-TRT in the context of concurrent CRT for LA- NSCLC. This is likely the result of irradiating less volume of adjacent critical organs such as the heart and lungs. Within the range of doses explored, the TRT dose levels were not significantly associated with patient survival. Future progress will require research focusing on better systemic therapy and TRT. OC-0493 Pretreatment metabolic volume in PET: Prognostic in lung cancer? German ESPATUE randomised phase3trial M. Guberina 1 , W.E.E. Eberhardt 2 , M. Stuschke 1 , T. Gauler 1 , D. Cheufou 3 , M. Schuler 2 , G. Stamatis 3 , C. Pöttgen 1 1 Universitätsklinikum Essen, Department of Radiotherapy, Essen, Germany 2 Universitätsklinikum Essen, Department of Medical Oncology- West German Cancer Center, Essen, Germany 3 Ruhrlandklinik- Essen, Department of Thoracic Surgery, Essen, Germany Purpose or Objective This secondary analysis of the ESPATUE study data (site University Hospital Essen) aimed to validate pretreatment metabolic tumour volume (MTV) from PET/CT in patients with locally advanced non-small cell lung cancer (NSCLC) treated by induction chemotherapy followed by neoadjuvant radiochemotherapy(RCT) and surgery or definitive RCT.This analysis was aimed to validate results from ACRIN6668/RTOG 0235 that pretreatment MTV is prognostic factor for overall survival (OS). Material and Methods All patients recruited to the ESPATUE trial were included, who were treated at University Hospital Essen and underwent FDG-PET/CT staging prior to treatment. Patients received induction chemotherapy, 3 cycles of cisplatin 50mg/m(2) on days 1/8 and paclitaxel 175mg/m(2) on day1 every 21days, as well as concurrent RCT to 45Gy given as 1.5Gy twice daily, concurrent cisplatin 50mg/m(2) and vinorelbine 20mg/m(2) on days 2/9. Patients whose tumours were deemed resectable in the last week of radiotherapy were randomly assigned to receive RCT boost that was risk adapted to between 65 and 71Gy or to undergo surgery. The MTV of primary tumour, and lymph nodes as well as maximum and mean tracer activity within the tumour volume [Bq/ml] were documented for all patients. Cox proportional hazard analysis was used to evaluate the pretreatment MTV as prognostic parameter for OS. Kaplan-Meier survival curves were generated to depict the association between MTV