ESTRO 2021 Abstract Book

S40

ESTRO 2021

Research Center, Heidelberg; National Center for Tumor Diseases (NCT), partner site Dresden; German Cancer Consortium (DKTK), core center Heidelberg; German Cancer Research Center (DKFZ), Dresden, Germany Purpose or Objective Tumor hypoxia is an established important factor contributing to radioresistance. A novel hypoxia-activated prodrug CP-506 has been proven to selectively target these hypoxic tumor cells and to cause anti-tumor activity in a panel of tumor xenografts. The current study was designed to investigate whether CP-506 improves outcome of radiotherapy in vivo. Materials and Methods Two human HNSCC lines, FaDu and UT-SCC-5, were transplanted subcutaneously into the hind-leg of NMRI nude mice. At 7 mm tumor diameter, the animals were randomized to receive 5 daily i.p. injections (QD5) of CP-506 (600 mg/kg) or vehicle followed by single dose irradiation (FaDu: 26 Gy, n=17-22; UT-SCC-5: 31 Gy, n=21-24, respectively). In parallel, untreated tumors and tumors treated with CP-506 or vehicle (QD5) were isolated after injection of pimonidazole hypoxia marker for histological analysis (n=10-11 per group). The relative hypoxic area within the viable tumor compartment (pimonidazole hypoxic fraction, pHF) was determined in two serial cross-sections per tumor by fluorescence microscopy and computerized image analysis. Animals were followed-up for at least 120 days to score all recurrences. Results CP-506 treatment significantly increased local tumor control rate after irradiation in FaDu, i.e. 62% vs. 27% (p=0.019), resulting in enhancement ratio of 2.3. The therapeutic effect of CP-506 combined with radiation in UT-SCC-5 was not curative albeit significant, i.e. risk of recurrence was significantly lower if tumors were targeted to CP-506 prior to RT (HR=0.44 [0.2; 0.95], p=0.034). Pimonidazole labelling confirmed the hypoxic status of the tumors prior to CP-506 treatment (FaDu: pHF of 4.5% [2.7]; UT-SCC-5: 9.8% [3.2]). Pimonidazole marker was not able to differentiate tumors treated with CP-506 from tumors treated with vehicle in both tumor models as pHF was not different across different treatment groups. Conclusion This study confirms that CP-506 improves outcome of radiotherapy with large single doses. Although pimonidazole hypoxia marker was not able to detect the decrease in hypoxic fraction upon CP-506 at the selected time point, some FaDu tumors had uniquely low pHF, highlighting the need for non-invasive tumor hypoxia imaging in each individual patient in order to accurately predict tumor response to CP-506. Currently, the DNA adducts formed by CP-506 and crosslink repair deficiency are being investigated in these tumors as potential predictors of tumor response. Because tumors reoxygenate during fractionated irradiation, a preclinical trial testing the effect of CP-506 in combination with fractionated radiotherapy with 30 fractions over 6 weeks on local control and hypoxic fraction in FaDu has been initiated and results will be presented. OC-0066 Heterogeneity of a hypoxia gene signature within the dominant tumor lesion in prostate cancer U.B. Salberg 1 , V. Eide Skingen 1 , C. Sæten Fjeldbo 1 , T. Hompland 1 , L. Vlatkovic 1 , K.H. Hole 1 , T. Seierstad 1 , H. Lyng 1 1 Oslo University Hospital, The Norwegian Radium Hospital, Oslo, Norway Purpose or Objective Gene expression signatures measured in tumor biopsies have shown promise as hypoxia biomarkers. However, biopsy-based biomarkers originate from a small part of the tumor and may fail to reflect aggressive features of the disease due to intratumor heterogeneity. For treatment decision in prostate cancer, collection of biopsies from the dominant intra-prostatic tumor lesion is aimed for. As a step to evaluate how well our previously published gene signature reflects hypoxia-related aggressiveness in prostate cancer, we here assessed its heterogeneity within the dominant tumor lesion. Materials and Methods The study included 114 prostate cancer patients, where 102 patients received the hypoxia marker pimonidazole before prostatectomy. Totally 141 punch biopsies were collected from the dominant lesion, including two biopsies from the same lesion in 47 patients. Hypoxic fraction was calculated for the entire lesion in whole-mount sections (HF W-m ) and for the individual biopsies (HF Biopsy ) by digital analysis of pimonidazole immunostaining. Gene expression profiles were generated from all biopsies and used to determine the signature of 32 genes. A gene score was calculated from the signature and applied in the analyses. For 41 biopsy pairs with both signature and pimonidazole data, total variance (T) and variance within pairs (W) were determined by variance component analysis, and the ratio W/T was used as a measure of the heterogeneity within the dominant lesion. Publically available gene expression data based on a single biopsy from each patient of three independent prostatectomy cohorts were used to assess the prognostic impact of the gene signature. Results High HF W-m was associated with signs of aggressive disease, including positive lymph node status (P=0.002) and Gleason grade above 4+3 (P=0.042). Although HF Biopsy and gene score showed heterogeneity within the dominant lesion, they both were positively correlated to HF W-m (P=< 0.001) when based on a single biopsy from each patient. The heterogeneity was lower for the gene score (W/T=0.10) than for HF Biopsy (W/T=0.22). The gene score also showed a lower heterogeneity than all but two genes in the signature. A significant association was found between the gene score and signs of aggressiveness (lymph node status, P=0.003; Gleason grade, P=0.017). Moreover, in two independent cohorts, significant correlation between the gene score and outcome was achieved when stratifying 33% of the patients to the high score group, consistent with the reported frequency of relapse for patients treated with radical prostatectomy. In the third independent cohort, significant association was achieved when 15% or less of the patients were in the high score group. Conclusion Our hypoxia gene signature shows a low heterogeneity within the dominant intra-prostatic tumor lesion. When based on a single biopsy from each patient, the signature reflects hypoxia-related aggressiveness of prostate