- Email: [email protected]
PD-0528 GERMLINE MRE11 VARIANTS PREDICTIVE OF RADIOTHERAPY OUTCOMES IN BLADDER CANCER BY NEXT-GENERATION SEQUENCING
S211
ESTRO 31
between EGFR inhibition, radiotherapy and chemotherapy and strong predictive markers for this combined treatment. SP-0524 HYPOXIA SIGNATURES PREDICT BENEFIT FROM RADIOSENSITIZERS J. Alsner1 1 Aarhus University Hospital, Department of Experimental Clinical Oncology, Aarhus C, Denmark Hypoxic tumours are associated with increased resistance to radiotherapy. In head and neck squamous cell carcinomas (HNSCC), this resistance can be counteracted via addition of hypoxic modification to the radiotherapy (1). Tumours are heterogeneous with respect to the degree and extend of hypoxia, and several methods have been developed to detect relevant tumour hypoxia. These include direct measurements in the tumour by an oxygen electrode, infusion and detection of exogenous hypoxia tracers, or quantification of endogenous hypoxia markers expressed by the tumour cells under hypoxic conditions. While there are extensive studies in relation to prognostic impact, less is known about the relevance of hypoxia assessing methods in relation to the prediction of response to hypoxic modification of radiotherapy. A recently developed 15-gene hypoxia classifier (2) has been shown to have predictive impact for identification of patients responding to hypoxic modification of radiotherapy with Nimorazole (3). Tumours from 323 HNSCC patients included in the randomized, double blinded DAHANCA 5 trial (4) were classified using the 15-gene hypoxia as being “more” or “less” hypoxic. 114 (35%) were classified as “more” hypoxic. The prognostic impact of the classifier was demonstrated in the placebo arm of the trial in terms of loco-regional tumour control, LRC (5-year actuarial values 44% vs 18%, p = 0.004), and disease specific survival, DSS (51% vs 30%, p = 0.04). Importantly, only patients with “more” hypoxic tumours had a significant benefit of hypoxic modification with Nimorazole compared with placebo (LRC: 49% vs 18%, p = 0.001; DSS: 48% vs 30%, p = 0.04). “Less” hypoxic tumours had no significant effect of hypoxic modification and the outcome was similar to “more” hypoxic tumours treated with Nimorazole. When stratified for HPV-status, the benefit of Nimorazole was restricted to HPV-negative and “more” hypoxic patients (LRC: 43% vs 9%; p = 0.002). Patients with HPV-positive tumours had a generally improved outcome, irrespective of their hypoxic status and whether or not the patients received Nimorazole. Studies are currently ongoing to implement the hypoxia classifier in clinical practice in Denmark, and to include the hypoxia classifier in international clinical trials on Nimorazole. Other ongoing studies are addressing the relevance of hypoxia in HPV-positive tumours and the correlations between the gene classifier and other hypoxia assessing methods (non-invasive imaging and circulating protein markers). Finally, the hypoxia classifier is being evaluated for other tumour sites. 1. Overgaard J. Radiother Oncol 100:22-32, 2011 2. Toustrup K, Sørensen BS, Nordsmark M et al. Cancer Res 71:59235931, 2011 3. Toustrup K, Sørensen BS, Lassen P et al. Radiother Oncol 102:122129, 2012 4. Overgaard J, Sand Hansen H, Overgaard M et al. Radiother Oncol 46:135-146, 1998
SYMPOSIUM: MODIFIERS
RADIOBIOLOGY
PRINCIPLES:
DOSE
SP-0525 IS THERE A NEED FOR INTERVAL BETWEEN DRUG AND RADIOTHERAPY DELIVERY? F. Herrera Centre Hospitalier Univ. Vaudois, Lausanne, Switzerland Abstract not received SP-0526 OPPORTUNITIES OF DOSE MODIFIERS: ILLUSTRATING THE CONCEPTS AND OUTLINING THE FUTURE C. Vens1 1 The Netherlands Cancer Institute - Antoni van Leeuwenhoek Hospital, Experimental Therapy, Amsterdam, The Netherlands One opportunity to improve radiotherapy (RT) is to combine RT with drugs that modify the radiation response of tumors. In the past few
decades, radiation science has yielded numerous potential dose modifiers for RT, mainly drugs that act as radio-sensitizers. Historically, conventional chemotherapeutic agents have been the most clinically used RT dose modifiers. Currently, modifiers with the greatest potential include those that target tumor-specific or radiation response specific elements. These agents emerged from studies that identified molecular targets in cellular and molecular pathways that are crucial for tumor survival and/or RT response. This lecture will provide an introduction to the main classes of classical and novel RT dose modifiers and illustrate their mode of action with a focus on the consequences for combination schedules and/or demand for biomarkers. Opportunities and limitations for the clinic, as well as future developments in the field will be discussed briefly. SP-0527 RADIOPROTECTORS: GENERAL OVERVIEW H. Langendijk1 1 University Medical Center Groningen / University of Groningen, Radiation Oncology, Groningen, The Netherlands Radiotherapy plays a pivotal role in the treatment of many malignancies. Despite its beneficial effects on locoregional tumor control, inevitable exposure of radiation to the normal tissues may result in acute and late side effects, which negatively affect quality of life. The best strategy to prevent side effects is to reduce the dose to normal tissues. However, in many cases, adequate sparing may hamper adequate dose delivery to the target and subsequently adversely affect locoregional tumor control. An alternative approach could be the use of radioprotectors, that are compounds designed to reduce damage to the normal tissues and thus to prevent side effects when given prior to or during radiation treatment. Radiation mitigators are compounds delivered at the time of irradiation or after completion of radiotherapy, but prior to the clinical manifestation of side effects. Radioprotectors and mitigators can only be used safely in clinical practice if their protective properties are limited to the normal tissues rather than to the tumor itself. Second, administration should be easy and their toxicity should be limited. Finally, their safety and efficacy in reducing side effects should be confirmed in properly designed trials. Radiation-induced normal tissue damage is mainly caused indirectly by the formation of free radicals or reactive oxygen species. Many radical scavengers and antioxidants exist that can be used as radioprotector, such as amifostine. Amifostine is a synthetic prodrug aminothiol that reduces these radicals through its free sulfhydryl moiety and that is selective as it will not be taken into cells until it is dephosphorylated by alkaline phosphatase, which is mainly present at normal cells rather than tumor cells. Amifostine appears to be effective in reducing xerostomia after head and neck irradiation. In a recent meta-analysis, no negative effects were observed on progression free survival and overall survival. There are several other antioxidants that showed some efficacy in preclinical studies, which however have not been tested properly yet in clinical trials. Radiation mitigators are compounds that intervene with ongoing mitotic cell death after radiation exposure and subsequent cytokine cascades that may lead to normal tissue damage. These agents can be used to prevent the clinical expression of toxicity (e.g. mucositis). There are numerous cytokines and growth factors that can be used as radiation mitigators that stimulate the differentiation of stem cells in normal tissues, such as in bone marrow or mucosa. Recently, the efficacy of palifermin, a recombinant human keratinocytes growth factor, was tested to reduce severe oral mucositis during definitive chemoradiation in head and neck cancer, in which a significant reduction of orals mucositis was observed. Although some new agents show promising results, further research is needed to further explore to what extend these compounds add to increasing the therapeutic ratio.
POSTER DISCUSSION: 11: PREDICTIVE AND SNP PD-0528 GERMLINE MRE11 VARIANTS PREDICTIVE OF RADIOTHERAPY OUTCOMES IN BLADDER CANCER BY NEXT-GENERATION SEQUENCING M.T.W. Teo1, H. Snowden1, J. Nsengimana1, M. Harland1, J. Morgan1, I. Carr1, G. Taylor1, J. Barrett1, D.T. Bishop1, A.E. Kiltie2 1 University of Leeds, Leeds Institute of Molecular Medicine, Leeds, United Kingdom
S212
2 University of Oxford, Gray Institute of Radiation Oncology and Biology, Oxford, United Kingdom
Purpose/Objective: Radical radiotherapy is an effective curative bladder conserving treatment modality for muscle invasive bladder cancer (MIBC). Radiotherapy causes cancer cell death by inducing lethal DNA double strand breaks (DSB). MRE11 is crucial for the detection of DSBs and activating DNA repair pathways. We have previously shown that pre-treatment bladder tumour MRE11 expression is predictive of radiotherapy outcomes in MIBC patients. We undertook a pilot study to identify germline MRE11A variants predictive of radiotherapy outcomes by bar-coded multiplexed nextgeneration sequencing of germline DNA from 186 MIBC patients treated with radical radiotherapy at the Leeds Teaching Hospitals NHS Trust between 2002 and 2009. Materials and Methods: The MRE11A gene was amplified by long polymerase chain reaction in all 186 germline DNA samples individually and bar-coded sequencing libraries made using unique oligonucleotide tags for each individual sample. Twenty to twenty five indexed samples were pooled for sequencing per Illumina Genome Analyser flow cell lane with variant calling performed using in-house designed software, Illuminator. All variants identified were validated by conventional sequencing. Cox multivariate analyses were performed for all variants on their effect on cancer-specific survival (CSS). Results: Common intronic SNP rs1805363 (G minor allele frequency 0.11), sited close to an alternative splice site for MRE11A isoform 1 and 2, was significantly associated with worse CSS (Per-allele Hazard Ratio (HR) 2.10, 95% Confidence Interval (CI) 1.34 – 3.28, Ptrend = 0.001) with a gene-dosage effect with increasing number of G minor alleles carried with a 5-year CSS of 58.3% in wildtypes, 42.0% in heterozygotes and 0% in minor allele homozygotes. On collapsed analysis, carriers of at least one rare variant allele within the 3’untranslated region of the MRE11A gene had a worse 5-year CSS compared to non-carriers (42.9% vs 54.8% respectively; HR4.04 95% CI 1.42-11.51, P-value = 0.009). Conclusions: By using next-generation sequencing technology, we identified germline MRE11A variants predictive of radiotherapy outcomes in MIBC that could aid future clinical identification of MIBC patients best suited to receive bladder conserving treatment. PD-0529 AMONG 45 VARIANTS IN 11 GENES, HDM2 POLYMORPHISMS EMERGE AS NEW CANDIDATE SNPS ASSOCIATED WITH RADIATION TOXICITY G. Alsbeih1, N. Al-Harbi1, L.A. Venturina1, K. Al-Hadyan1, M. ElSebaie2, N. Al-Rajhi2 1 King Faisal Specialist Hospital & Research Centre, Biomedical Physics, Riyadh, Saudi Arabia 2 King Faisal Specialist Hospital & Research Centre, Radiation Oncology, Riyadh, Saudi Arabia Purpose/Objective: Single nucleotide polymorphisms (SNPs) are the largest types of inherited genetic variations in humans, and the challenge of radiobiology is to determine which SNPs influence radiation sensitivity and the risk to develop severe complications. The aim of this study was to investigate the association between candidate SNPs and late complications in Head and Neck cancer patients. Materials and Methods: SNPs in genes that mitigate the response to DNA damage are likely to affect radiotherapy outcome. These involve multiple pathways of which we selected 11 candidate genes (ATM, XRCC1, XRCC3, XRCC4, XRCC5, PRKDC, LIG4, TP53, HDM2, CDKN1A, TGFB1) for their presumed influence on radiosensitivity. Forty five (12 primary and 33 neighboring) SNPs were genotyped by direct sequencing. A total of 140 patients with nasopharyngeal carcinomas, treated with 3D conformal technique with (73) or without (67) chemotherapy, were included in the study. Normal tissue fibrosis was scored using RTOG/EORTC grading system. Blood samples were obtained following signed informed consent; DNA was extracted, target sequences were amplified and run on the MegaBase 1000 sequencer. SNPs were genotyped using SeqManII software. Results: Patients with severe fibrosis (radiosensitive cases, G3-4, n = 46) were compared to those with little or moderate reactions (controls G0-2, n = 94). Only 6/45 SNPs genotyped (ATM G/A rs1801516, HDM2 promoter T/G rs2279744 and T/A rs1196333, XRCC1 G/A rs25487, XRCC5 T/C rs1051677 and TGFB C/T rs1800469) were significantly associated with radiation complications (P = 0.020, 0.037, 0.041, 0.037, 0.021 and 0.041, respectively).
ESTRO 31
Conclusions: These finding suggest that certain SNPs are associated with radiation toxicity in radiotherapy patients. Many of these SNPs were studied before with variable results. Interestingly, this is the 1st study to include SNPs in HDM2 gene where 2 SNPs in the promoter region were significantly associated with fibrotic reaction. HDM2 plays a critical role in regulating TP53. The SNP309 T/G (rs2279744) is common and might affect HDM2 activity and associates with altered risk to develop cancer. The variant G allele binds to Sp1 and increases HDM2 mRNA and protein, thereby suppressing TP53. Here, the G allele was protective (OR = 0.56, CI: 0.33-0.97) probably due to increased rate of cell division, required to maintain tissue function, despite risk of genomic instability. In addition, the radiosensitive patients harbored significantly higher number of risk alleles than the controls (Mann-Whitney test, P < 0.001). Moreover, Western blotting showed significant decrease in HDM2 protein levels in radiosensitive patients. Independent replication of these findings and genome wide association studies (GWAS) are required in order for SNPs to be used as biomarkers to individualize radiotherapy on genetic basis. Supported by KFSH&RC grants 2000 031 and 2040 025. PD-0530 PROTEOMIC IDENTIFICATION OF PUTATIVE BIOMARKERS OF RADIOTHERAPY RESISTANCE L. Scaife1, V.C. Hodgkinson1, D. ELFadl1, S. Mehmood1, I.A. Hunter2, G.P. Liney3, A.W. Beavis3, P.J. Drew4, M.J. Lind1, L. Cawkwell1 1 University of Hull, Oncology, Hull, United Kingdom 2 Castle Hill Hospital, Surgery, Hull, United Kingdom 3 Castle Hill Hospital, Oncology, Hull, United Kingdom 4 Hull York Medical School, Surgery, Hull, United Kingdom Purpose/Objective: The proteomic analysis of radioresistant cancer cell lines may identify proteins which could be used as predictive biomarkers in order to individualise treatment for cancer patients. Materials and Methods: We previously established 7 novel radioresistant (RR) cancer cell sublines from parental breast cancer cells (MCF-7, MDA-MB-231, T47D), oral cancer cells (PE/CA-PJ41, PE/CA-PJ49) and rectal cancer cells (SW837, HRA-19). These RR cell sublines were established by exposing parental cell lines to fractionated external beam radiation to a clinically relevant total dose. A colony counting assay was used to confirm significant increase in radioresistance in RR cell sublines. Two complementary comparative proteomics methods were used to identify differentially expressed proteins (DEPs) in RR cells when compared with the respective parental cells. An antibody microarray (AbMA) platform containing 725 antibodies was utilised in addition to 2 dimensional gel electrophoresis (2-DE) with MALDI TOF/TOF mass spectrometry (MS). The DEPs identified were mapped onto canonical signalling pathways using Ingenuity Pathway Analysis (IPA) and a selection of proteins investigated further using immunoblotting and immunohistochemistry (IHC). Results: Using AbMA analysis a total of 62 DEPs were identified in the 3 breast cancer RR sublines with 11 of these DEPs, including the death receptor DR4, seen in 2/3 sublines. A total of 31 DEPs were identified in the 2 oral RR cancer sublines with 6 of these DEPs, including BID and RAD17, seen in both sublines. A total of 63 DEPs, including DR4 and BCL2, were identified independently in 2/2 rectal cancer RR sublines. The 2-DE/MS analysis of oral and rectal cancer RR sublines identified a further 27 DEPs, which included members of the ubiquitin-proteasome pathway (UPP) as we have previously reported from the breast cancer RR sublines. The DEPs identified using proteomics in all 7 RR sublines were submitted for IPA analysis to reveal canonical pathways which were significantly associated with radioresistance. These pathways included p53 signaling, ATM signalling and the UPP. Immunoblotting was used to confirm the differential expression of selected proteins, including DR4, BID and RAD17 in RR sublines. A small pilot IHC study was performed to confirm the association of DR4 expression in radioresistant tumour samples. Conclusions: Two complementary proteomic methods have been used to discover putative biomarkers of radioresistance using novel cell line models. Further clinical validation of selected proteins within key canonical pathways, including the p53, ATM and UPP signalling pathways, is underway in order to identify a panel of predictive biomarkers of radioresistance.
ESTRO 31
between EGFR inhibition, radiotherapy and chemotherapy and strong predictive markers for this combined treatment. SP-0524 HYPOXIA SIGNATURES PREDICT BENEFIT FROM RADIOSENSITIZERS J. Alsner1 1 Aarhus University Hospital, Department of Experimental Clinical Oncology, Aarhus C, Denmark Hypoxic tumours are associated with increased resistance to radiotherapy. In head and neck squamous cell carcinomas (HNSCC), this resistance can be counteracted via addition of hypoxic modification to the radiotherapy (1). Tumours are heterogeneous with respect to the degree and extend of hypoxia, and several methods have been developed to detect relevant tumour hypoxia. These include direct measurements in the tumour by an oxygen electrode, infusion and detection of exogenous hypoxia tracers, or quantification of endogenous hypoxia markers expressed by the tumour cells under hypoxic conditions. While there are extensive studies in relation to prognostic impact, less is known about the relevance of hypoxia assessing methods in relation to the prediction of response to hypoxic modification of radiotherapy. A recently developed 15-gene hypoxia classifier (2) has been shown to have predictive impact for identification of patients responding to hypoxic modification of radiotherapy with Nimorazole (3). Tumours from 323 HNSCC patients included in the randomized, double blinded DAHANCA 5 trial (4) were classified using the 15-gene hypoxia as being “more” or “less” hypoxic. 114 (35%) were classified as “more” hypoxic. The prognostic impact of the classifier was demonstrated in the placebo arm of the trial in terms of loco-regional tumour control, LRC (5-year actuarial values 44% vs 18%, p = 0.004), and disease specific survival, DSS (51% vs 30%, p = 0.04). Importantly, only patients with “more” hypoxic tumours had a significant benefit of hypoxic modification with Nimorazole compared with placebo (LRC: 49% vs 18%, p = 0.001; DSS: 48% vs 30%, p = 0.04). “Less” hypoxic tumours had no significant effect of hypoxic modification and the outcome was similar to “more” hypoxic tumours treated with Nimorazole. When stratified for HPV-status, the benefit of Nimorazole was restricted to HPV-negative and “more” hypoxic patients (LRC: 43% vs 9%; p = 0.002). Patients with HPV-positive tumours had a generally improved outcome, irrespective of their hypoxic status and whether or not the patients received Nimorazole. Studies are currently ongoing to implement the hypoxia classifier in clinical practice in Denmark, and to include the hypoxia classifier in international clinical trials on Nimorazole. Other ongoing studies are addressing the relevance of hypoxia in HPV-positive tumours and the correlations between the gene classifier and other hypoxia assessing methods (non-invasive imaging and circulating protein markers). Finally, the hypoxia classifier is being evaluated for other tumour sites. 1. Overgaard J. Radiother Oncol 100:22-32, 2011 2. Toustrup K, Sørensen BS, Nordsmark M et al. Cancer Res 71:59235931, 2011 3. Toustrup K, Sørensen BS, Lassen P et al. Radiother Oncol 102:122129, 2012 4. Overgaard J, Sand Hansen H, Overgaard M et al. Radiother Oncol 46:135-146, 1998
SYMPOSIUM: MODIFIERS
RADIOBIOLOGY
PRINCIPLES:
DOSE
SP-0525 IS THERE A NEED FOR INTERVAL BETWEEN DRUG AND RADIOTHERAPY DELIVERY? F. Herrera Centre Hospitalier Univ. Vaudois, Lausanne, Switzerland Abstract not received SP-0526 OPPORTUNITIES OF DOSE MODIFIERS: ILLUSTRATING THE CONCEPTS AND OUTLINING THE FUTURE C. Vens1 1 The Netherlands Cancer Institute - Antoni van Leeuwenhoek Hospital, Experimental Therapy, Amsterdam, The Netherlands One opportunity to improve radiotherapy (RT) is to combine RT with drugs that modify the radiation response of tumors. In the past few
decades, radiation science has yielded numerous potential dose modifiers for RT, mainly drugs that act as radio-sensitizers. Historically, conventional chemotherapeutic agents have been the most clinically used RT dose modifiers. Currently, modifiers with the greatest potential include those that target tumor-specific or radiation response specific elements. These agents emerged from studies that identified molecular targets in cellular and molecular pathways that are crucial for tumor survival and/or RT response. This lecture will provide an introduction to the main classes of classical and novel RT dose modifiers and illustrate their mode of action with a focus on the consequences for combination schedules and/or demand for biomarkers. Opportunities and limitations for the clinic, as well as future developments in the field will be discussed briefly. SP-0527 RADIOPROTECTORS: GENERAL OVERVIEW H. Langendijk1 1 University Medical Center Groningen / University of Groningen, Radiation Oncology, Groningen, The Netherlands Radiotherapy plays a pivotal role in the treatment of many malignancies. Despite its beneficial effects on locoregional tumor control, inevitable exposure of radiation to the normal tissues may result in acute and late side effects, which negatively affect quality of life. The best strategy to prevent side effects is to reduce the dose to normal tissues. However, in many cases, adequate sparing may hamper adequate dose delivery to the target and subsequently adversely affect locoregional tumor control. An alternative approach could be the use of radioprotectors, that are compounds designed to reduce damage to the normal tissues and thus to prevent side effects when given prior to or during radiation treatment. Radiation mitigators are compounds delivered at the time of irradiation or after completion of radiotherapy, but prior to the clinical manifestation of side effects. Radioprotectors and mitigators can only be used safely in clinical practice if their protective properties are limited to the normal tissues rather than to the tumor itself. Second, administration should be easy and their toxicity should be limited. Finally, their safety and efficacy in reducing side effects should be confirmed in properly designed trials. Radiation-induced normal tissue damage is mainly caused indirectly by the formation of free radicals or reactive oxygen species. Many radical scavengers and antioxidants exist that can be used as radioprotector, such as amifostine. Amifostine is a synthetic prodrug aminothiol that reduces these radicals through its free sulfhydryl moiety and that is selective as it will not be taken into cells until it is dephosphorylated by alkaline phosphatase, which is mainly present at normal cells rather than tumor cells. Amifostine appears to be effective in reducing xerostomia after head and neck irradiation. In a recent meta-analysis, no negative effects were observed on progression free survival and overall survival. There are several other antioxidants that showed some efficacy in preclinical studies, which however have not been tested properly yet in clinical trials. Radiation mitigators are compounds that intervene with ongoing mitotic cell death after radiation exposure and subsequent cytokine cascades that may lead to normal tissue damage. These agents can be used to prevent the clinical expression of toxicity (e.g. mucositis). There are numerous cytokines and growth factors that can be used as radiation mitigators that stimulate the differentiation of stem cells in normal tissues, such as in bone marrow or mucosa. Recently, the efficacy of palifermin, a recombinant human keratinocytes growth factor, was tested to reduce severe oral mucositis during definitive chemoradiation in head and neck cancer, in which a significant reduction of orals mucositis was observed. Although some new agents show promising results, further research is needed to further explore to what extend these compounds add to increasing the therapeutic ratio.
POSTER DISCUSSION: 11: PREDICTIVE AND SNP PD-0528 GERMLINE MRE11 VARIANTS PREDICTIVE OF RADIOTHERAPY OUTCOMES IN BLADDER CANCER BY NEXT-GENERATION SEQUENCING M.T.W. Teo1, H. Snowden1, J. Nsengimana1, M. Harland1, J. Morgan1, I. Carr1, G. Taylor1, J. Barrett1, D.T. Bishop1, A.E. Kiltie2 1 University of Leeds, Leeds Institute of Molecular Medicine, Leeds, United Kingdom
S212
2 University of Oxford, Gray Institute of Radiation Oncology and Biology, Oxford, United Kingdom
Purpose/Objective: Radical radiotherapy is an effective curative bladder conserving treatment modality for muscle invasive bladder cancer (MIBC). Radiotherapy causes cancer cell death by inducing lethal DNA double strand breaks (DSB). MRE11 is crucial for the detection of DSBs and activating DNA repair pathways. We have previously shown that pre-treatment bladder tumour MRE11 expression is predictive of radiotherapy outcomes in MIBC patients. We undertook a pilot study to identify germline MRE11A variants predictive of radiotherapy outcomes by bar-coded multiplexed nextgeneration sequencing of germline DNA from 186 MIBC patients treated with radical radiotherapy at the Leeds Teaching Hospitals NHS Trust between 2002 and 2009. Materials and Methods: The MRE11A gene was amplified by long polymerase chain reaction in all 186 germline DNA samples individually and bar-coded sequencing libraries made using unique oligonucleotide tags for each individual sample. Twenty to twenty five indexed samples were pooled for sequencing per Illumina Genome Analyser flow cell lane with variant calling performed using in-house designed software, Illuminator. All variants identified were validated by conventional sequencing. Cox multivariate analyses were performed for all variants on their effect on cancer-specific survival (CSS). Results: Common intronic SNP rs1805363 (G minor allele frequency 0.11), sited close to an alternative splice site for MRE11A isoform 1 and 2, was significantly associated with worse CSS (Per-allele Hazard Ratio (HR) 2.10, 95% Confidence Interval (CI) 1.34 – 3.28, Ptrend = 0.001) with a gene-dosage effect with increasing number of G minor alleles carried with a 5-year CSS of 58.3% in wildtypes, 42.0% in heterozygotes and 0% in minor allele homozygotes. On collapsed analysis, carriers of at least one rare variant allele within the 3’untranslated region of the MRE11A gene had a worse 5-year CSS compared to non-carriers (42.9% vs 54.8% respectively; HR4.04 95% CI 1.42-11.51, P-value = 0.009). Conclusions: By using next-generation sequencing technology, we identified germline MRE11A variants predictive of radiotherapy outcomes in MIBC that could aid future clinical identification of MIBC patients best suited to receive bladder conserving treatment. PD-0529 AMONG 45 VARIANTS IN 11 GENES, HDM2 POLYMORPHISMS EMERGE AS NEW CANDIDATE SNPS ASSOCIATED WITH RADIATION TOXICITY G. Alsbeih1, N. Al-Harbi1, L.A. Venturina1, K. Al-Hadyan1, M. ElSebaie2, N. Al-Rajhi2 1 King Faisal Specialist Hospital & Research Centre, Biomedical Physics, Riyadh, Saudi Arabia 2 King Faisal Specialist Hospital & Research Centre, Radiation Oncology, Riyadh, Saudi Arabia Purpose/Objective: Single nucleotide polymorphisms (SNPs) are the largest types of inherited genetic variations in humans, and the challenge of radiobiology is to determine which SNPs influence radiation sensitivity and the risk to develop severe complications. The aim of this study was to investigate the association between candidate SNPs and late complications in Head and Neck cancer patients. Materials and Methods: SNPs in genes that mitigate the response to DNA damage are likely to affect radiotherapy outcome. These involve multiple pathways of which we selected 11 candidate genes (ATM, XRCC1, XRCC3, XRCC4, XRCC5, PRKDC, LIG4, TP53, HDM2, CDKN1A, TGFB1) for their presumed influence on radiosensitivity. Forty five (12 primary and 33 neighboring) SNPs were genotyped by direct sequencing. A total of 140 patients with nasopharyngeal carcinomas, treated with 3D conformal technique with (73) or without (67) chemotherapy, were included in the study. Normal tissue fibrosis was scored using RTOG/EORTC grading system. Blood samples were obtained following signed informed consent; DNA was extracted, target sequences were amplified and run on the MegaBase 1000 sequencer. SNPs were genotyped using SeqManII software. Results: Patients with severe fibrosis (radiosensitive cases, G3-4, n = 46) were compared to those with little or moderate reactions (controls G0-2, n = 94). Only 6/45 SNPs genotyped (ATM G/A rs1801516, HDM2 promoter T/G rs2279744 and T/A rs1196333, XRCC1 G/A rs25487, XRCC5 T/C rs1051677 and TGFB C/T rs1800469) were significantly associated with radiation complications (P = 0.020, 0.037, 0.041, 0.037, 0.021 and 0.041, respectively).
ESTRO 31
Conclusions: These finding suggest that certain SNPs are associated with radiation toxicity in radiotherapy patients. Many of these SNPs were studied before with variable results. Interestingly, this is the 1st study to include SNPs in HDM2 gene where 2 SNPs in the promoter region were significantly associated with fibrotic reaction. HDM2 plays a critical role in regulating TP53. The SNP309 T/G (rs2279744) is common and might affect HDM2 activity and associates with altered risk to develop cancer. The variant G allele binds to Sp1 and increases HDM2 mRNA and protein, thereby suppressing TP53. Here, the G allele was protective (OR = 0.56, CI: 0.33-0.97) probably due to increased rate of cell division, required to maintain tissue function, despite risk of genomic instability. In addition, the radiosensitive patients harbored significantly higher number of risk alleles than the controls (Mann-Whitney test, P < 0.001). Moreover, Western blotting showed significant decrease in HDM2 protein levels in radiosensitive patients. Independent replication of these findings and genome wide association studies (GWAS) are required in order for SNPs to be used as biomarkers to individualize radiotherapy on genetic basis. Supported by KFSH&RC grants 2000 031 and 2040 025. PD-0530 PROTEOMIC IDENTIFICATION OF PUTATIVE BIOMARKERS OF RADIOTHERAPY RESISTANCE L. Scaife1, V.C. Hodgkinson1, D. ELFadl1, S. Mehmood1, I.A. Hunter2, G.P. Liney3, A.W. Beavis3, P.J. Drew4, M.J. Lind1, L. Cawkwell1 1 University of Hull, Oncology, Hull, United Kingdom 2 Castle Hill Hospital, Surgery, Hull, United Kingdom 3 Castle Hill Hospital, Oncology, Hull, United Kingdom 4 Hull York Medical School, Surgery, Hull, United Kingdom Purpose/Objective: The proteomic analysis of radioresistant cancer cell lines may identify proteins which could be used as predictive biomarkers in order to individualise treatment for cancer patients. Materials and Methods: We previously established 7 novel radioresistant (RR) cancer cell sublines from parental breast cancer cells (MCF-7, MDA-MB-231, T47D), oral cancer cells (PE/CA-PJ41, PE/CA-PJ49) and rectal cancer cells (SW837, HRA-19). These RR cell sublines were established by exposing parental cell lines to fractionated external beam radiation to a clinically relevant total dose. A colony counting assay was used to confirm significant increase in radioresistance in RR cell sublines. Two complementary comparative proteomics methods were used to identify differentially expressed proteins (DEPs) in RR cells when compared with the respective parental cells. An antibody microarray (AbMA) platform containing 725 antibodies was utilised in addition to 2 dimensional gel electrophoresis (2-DE) with MALDI TOF/TOF mass spectrometry (MS). The DEPs identified were mapped onto canonical signalling pathways using Ingenuity Pathway Analysis (IPA) and a selection of proteins investigated further using immunoblotting and immunohistochemistry (IHC). Results: Using AbMA analysis a total of 62 DEPs were identified in the 3 breast cancer RR sublines with 11 of these DEPs, including the death receptor DR4, seen in 2/3 sublines. A total of 31 DEPs were identified in the 2 oral RR cancer sublines with 6 of these DEPs, including BID and RAD17, seen in both sublines. A total of 63 DEPs, including DR4 and BCL2, were identified independently in 2/2 rectal cancer RR sublines. The 2-DE/MS analysis of oral and rectal cancer RR sublines identified a further 27 DEPs, which included members of the ubiquitin-proteasome pathway (UPP) as we have previously reported from the breast cancer RR sublines. The DEPs identified using proteomics in all 7 RR sublines were submitted for IPA analysis to reveal canonical pathways which were significantly associated with radioresistance. These pathways included p53 signaling, ATM signalling and the UPP. Immunoblotting was used to confirm the differential expression of selected proteins, including DR4, BID and RAD17 in RR sublines. A small pilot IHC study was performed to confirm the association of DR4 expression in radioresistant tumour samples. Conclusions: Two complementary proteomic methods have been used to discover putative biomarkers of radioresistance using novel cell line models. Further clinical validation of selected proteins within key canonical pathways, including the p53, ATM and UPP signalling pathways, is underway in order to identify a panel of predictive biomarkers of radioresistance.