Genetic sequence variants (GSV) in relation to acute and late toxicities in head and neck cancer (HNC) patients treated with radiation therapy (RT).

Abstract

5507 Background: Genetic variation may partly explain differences in normal tissue toxicities after RT; yet no clear picture has emerged from RT toxicity studies of common GSV in candidate genes. In a large cohort of HNC patients treated with RT, we assessed whether 8 GSV, systematically selected from published sources, predicted acute or late RT toxicity,: TP53 rs1042522; ERCC4 rs1799801; ERCC2 rs1799793; ERCC2 rs13181; XRCC1 rs25487; XRCC3 rs861539; GSTM1 deletion; GSTT1 deletion. Methods: In a secondary analysis of a randomized trial of vitamin E supplementation in 540 stage I-II HNC patients treated with RT, RT toxicities were assessed using RTOG Acute Radiation Morbidity Criteria (during RT, 1 month after RT) and RTOG/EORTC Late Radiation Morbidity Scoring Scheme (at 6, 12 months after RT). The most severe time and tissue-specific RT toxicity, graded 0-4, was taken as an overall measure of acute/late toxicity. All 8 GSV were genotyped from blood leucocyte derived DNA. Ordinal logistic regression estimated odds ratios (OR) and 95% confidence intervals (CI) for toxicity, adjusted for known toxicity predictors in multivariate models. The false discovery rate (FDR) due to multiple testing was taken into account. Results: As crude and adjusted ORs associated with GSV were similar, we controlled for FDR by adjusting the p-values associated with the crude ORs. For late toxicity, no statistically significant association was observed with any of the 8 GSV. For acute toxicity, one statistically significant association was observed: The OR of acute toxicity was 1.43 (CI: 1.11-1.85) for the XRCC3 rs861539 (Thr241Met). XRCC3 (X-Ray repair, complementing defective, in Chinese Hamster 3) is part of a BRCA2-FANCD2-FANCG complex that promotes homologous recombination and interacts with rad51. Conclusions: In our study, the rare allele of XRCC3 rs861539 was associated with increased acute RT toxicity, but not late toxicity. Previous studies of this SNP only assessed late toxicity, fibrosis or telangiectasia, with no clear results. Though results are promising, our study also underlines the limitations of the candidate gene approach to identify GSV that could predict RT toxicities.