• A deep learning approach validates genetic risk factors for late toxicity after prostate cancer radiotherapy in a REQUITE multi-national cohort

      Massi, M. C.; Gasperoni, F.; Ieva, F.; Paganoni, A. M.; Zunino, P.; Manzoni, A.; Franco, N. R.; Veldeman, L.; Ost, P.; Fonteyne, V.; et al. (2020)
      Background: REQUITE (validating pREdictive models and biomarkers of radiotherapy toxicity to reduce side effects and improve QUalITy of lifE in cancer survivors) is an international prospective cohort study. The purpose of this project was to analyse a cohort of patients recruited into REQUITE using a deep learning algorithm to identify patient-specific features associated with the development of toxicity, and test the approach by attempting to validate previously published genetic risk factors. Methods: The study involved REQUITE prostate cancer patients treated with external beam radiotherapy who had complete 2-year follow-up. We used five separate late toxicity endpoints: ≥grade 1 late rectal bleeding, ≥grade 2 urinary frequency, ≥grade 1 haematuria, ≥ grade 2 nocturia, ≥ grade 1 decreased urinary stream. Forty-three single nucleotide polymorphisms (SNPs) already reported in the literature to be associated with the toxicity endpoints were included in the analysis. No SNP had been studied before in the REQUITE cohort. Deep Sparse AutoEncoders (DSAE) were trained to recognize features (SNPs) identifying patients with no toxicity and tested on a different independent mixed population including patients without and with toxicity. Results: One thousand, four hundred and one patients were included, and toxicity rates were: rectal bleeding 11.7%, urinary frequency 4%, haematuria 5.5%, nocturia 7.8%, decreased urinary stream 17.1%. Twenty-four of the 43 SNPs that were associated with the toxicity endpoints were validated as identifying patients with toxicity. Twenty of the 24 SNPs were associated with the same toxicity endpoint as reported in the literature: 9 SNPs for urinary symptoms and 11 SNPs for overall toxicity. The other 4 SNPs were associated with a different endpoint. Conclusion: Deep learning algorithms can validate SNPs associated with toxicity after radiotherapy for prostate cancer. The method should be studied further to identify polygenic SNP risk signatures for radiotherapy toxicity. The signatures could then be included in integrated normal tissue complication probability models and tested for their ability to personalize radiotherapy treatment planning.
    • Development of a method for generating SNP interaction-aware polygenic risk scores for radiotherapy toxicity

      Franco, N. R.; Massi, M. C.; Ieva, F.; Manzoni, A.; Paganoni, A. M.; Zunino, P.; Veldeman, L.; Ost, P.; Fonteyne, V.; Talbot, C. J.; et al. (2021)
      Aim: To identify the effect of single nucleotide polymorphism (SNP) interactions on the risk of toxicity following radiotherapy (RT) for prostate cancer (PCa) and propose a new method for polygenic risk score incorporating SNP-SNP interactions (PRSi). Materials and methods: Analysis included the REQUITE PCa cohort that received external beam RT and was followed for 2 years. Late toxicity endpoints were: rectal bleeding, urinary frequency, haematuria, nocturia, decreased urinary stream. Among 43 literature-identified SNPs, the 30% most strongly associated with each toxicity were tested. SNP-SNP combinations (named SNP-allele sets) seen in ≥10% of the cohort were condensed into risk (RS) and protection (PS) scores, respectively indicating increased or decreased toxicity risk. Performance of RS and PS was evaluated by logistic regression. RS and PS were then combined into a single PRSi evaluated by area under the receiver operating characteristic curve (AUC). Results: Among 1,387 analysed patients, toxicity rates were 11.7% (rectal bleeding), 4.0% (urinary frequency), 5.5% (haematuria), 7.8% (nocturia) and 17.1% (decreased urinary stream). RS and PS combined 8 to 15 different SNP-allele sets, depending on the toxicity endpoint. Distributions of PRSi differed significantly in patients with/without toxicity with AUCs ranging from 0.61 to 0.78. PRSi was better than the classical summed PRS, particularly for the urinary frequency, haematuria and decreased urinary stream endpoints. Conclusions: Our method incorporates SNP-SNP interactions when calculating PRS for radiotherapy toxicity. Our approach is better than classical summation in discriminating patients with toxicity and should enable incorporating genetic information to improve normal tissue complication probability models.
    • Do polygenic risk scores for cancer susceptibility associate with risk of radiotherapy toxicity

      Kerns, S. L.; Barnett, G.; Dorling, L.; Fachal, L.; Martinez-Calvo, L.; Aguado-Barrera, M.; Dearnaley, D; Coles, C. E.; Burnet, Neil G; Webb, A.; et al. (2020)
      Purpose/Objective(s): Single nucleotide polymorphisms (SNPs) associated with increased susceptibility to cancer frequently lie in genes that might also modulate response to radiation. The purpose of this study is to evaluate whether individuals at increased genetic risk of developing cancer are also more likely to develop late radiotherapy-induced toxicities. We test the hypothesis that polygenic risk scores (PRS) for cancer susceptibility alter toxicity risk among cancer patients who received radiotherapy. Materials/Methods: Analysis included 1,134 breast, 3,521 prostate, and 610 lung cancer patients from radiotherapy cohorts from the Radiogenomics and REQUITE Consortia. All patients received radiotherapy alone or as part of combination treatment and were followed prospectively for development of toxicity in relevant tissues. Germline DNA was genotyped using a genome-wide SNP array with non-typed SNPs imputed using the 1000 Genomes reference data. A PRS was generated for each patient by summing risk alleles from cancer susceptibility SNPs identified in the literature - 352 for breast, 147 for prostate, 24 for lung cancer. A weighted PRS was similarly generated for prostate and lung cancer in which each SNP was first weighted by its odds ratio for cancer susceptibility; such a score was not available for breast cancer as the odds ratios vary by estrogen receptor status. Toxicity was quantified for each patient using STAT score, which is a previously validated measure of overall radiotherapy toxicity that combines multiple individual endpoints. Multivariable logistic regression tested association between PRS and toxicity STAT score (dichotomized at the mean plus one standard deviation) at 2 years after treatment for breast and prostate and up to 1 year for lung cancer patients, controlling for clinical covariates. Individual SNPs comprising the PRS were tested in a secondary analysis. Results: No association was found between PRS and development of late radiotherapy toxicity among breast (PRS OR Z 1.01, 95% CI Z 1.00 to 1.02), prostate (PRS OR Z 1.01, 95% CI Z 1.00 to 1.03; weighted PRS OR Z 1.10, 95%CI Z 0.94 to 1.28) or lung cancer patients (PRS OR Z 0.95, 95% CI Z 0.88 to 1.02; weighted PRS OR Z 0.77, 95% CI Z 0.51 to 1.14). On multivariable analysis of individual SNPs, rs138944387 was associated with breast pain (beta Z 1.12; 95% CI Z 0.62 to 1.61; p Z 1.09x10-5) and rs17513613 was associated with risk of breast edema (beta Z -0.21; 95% CI Z -0.31 to -0.12; p Z 2.01x10-5). Conclusion: Cancer patients with a high polygenic predisposition to breast, prostate or lung cancer show no evidence of an increased risk of late radiotherapy toxicity. Thus, these patients can undergo standard treatment without an anticipated excess toxicity risk. The association between individual SNPs and late toxicity requires validation in independent cohorts and functional studies to elucidate the biologic mechanism underlying this shared risk.
    • Genome wide association study of acute radiation toxicity and quality of life in breast cancer patients - results from the REQUITE cohort study

      Rattay, T.; Veal, C. D.; Azria, D.; Chang-Claude, J.; Davidson, Susan E; Dunning, A.; De Ruysscher, D; Fachal, L.; Gutierrez-Enriquez, S.; Lambin, P.; et al. (2020)
      Background: Around a quarter of breast cancer patients treated by surgery and radiotherapy experienceclinically significant toxicity, which may adversely affect breast cosmesis and qualityof life (QoL). If patients at high risk of toxicity could be identified at diagnosis,this could be taken into account when discussing treatment options. This study wasdesigned to identify common single nucleotide polymorphisms (SNPs) associated withacute radiation toxicity and change in QoL on completion of radiotherapy.