Treatment-specific genetic risk scores for late effects prediction in childhood, adolescent, and young adult cancer survivors

点击翻译按钮获取中文摘要

PROJECT SUMMARY OR ABSTRACT There is growing evidence that polygenic risk scores (PRS), or aggregate measures of genetic risk for disease based on findings from genome-wide association studies (GWAS), may be useful in predicting risks for a wide array of chronic health conditions (CHCs) in the general population. But are PRS useful for predicting late effects risks in survivors of childhood, adolescent, and young adult (CAYA) cancer? Recently, we have shown that PRS based on GWAS conducted in the general population are relatively poor predictors of risks for various CHCs in survivors, in part because they do not consider the modifying effects of specific curative therapies for cancer on genetic risks for CHCs, i.e., treatment-specific genetic effects. To date, no existing analytic method bridges this research gap: this project is innovative because it aims to develop a new method for PRS that can also incorporate treatment-specific genetic effects. The overall objective of the proposed project is to create novel genetic risk predictors that combine established genetic risk signals identified in general population GWAS with treatment-specific genetic effects to equitably predict CHC risks in CAYA cancer survivors. We will investigate this approach using a preliminary set of ten CHCs and complex traits, including subsequent cancers and cardiovascular and endocrine diseases. The long-term goal of this project is to apply these newly developed treatment-specific genetic risk predictors to inform personalized follow-up care programs for survivors, which responds to RFA-CA-20-028 application domains 5 (development of risk predictors of late effects) and 6 (development of targeted interventions to reduce the burden of cancer). Utilizing existing genetic/phenotype data from ~13,000 survivors, the proposed study will pursue the following specific aims: (1) compile a comprehensive catalog of treatment-specific genetic effects for each of the selected phenotypes; (2) develop and evaluate corresponding “survivor-enriched” PRS that incorporate these treatment-specific genetic effects; and (3) assess whether survivor-enriched PRS can decrease disparities in late effects risk prediction in survivors from racial/ethnic minority populations. To accomplish the first aim, we will perform strategic genome-wide association analyses to identify genetic variants whose risk associations with CHCs are modified by therapeutic exposures. We then intend to adapt existing methods shown to improve the cross-population generalizability of PRS to combine discoveries from general population GWAS with results from genetic analyses in survivors. The relative predictive ability of these survivor-enriched PRS will be evaluated in an independent survivor sample. Lastly, given that treatment-specific genetic effects may be trans-ethnic, we will assess survivor-enriched PRS in racial/ethnic minority survivor subgroups that have been traditionally excluded from Eurocentric GWAS. The proposed research is significant in its potential to help identify survivors at high risk for adverse health outcomes with treatment-specific genetic risk scores, which would represent an important advance over current guidelines for survivorship care.