Genomic Prediction of Doxorubicin-Induced Cardiotoxicity

点击翻译按钮获取中文摘要

Project Summary The anthracycline doxorubicin used in approximately 60% of pediatric cancer patients with metastatic solid tumors (sarcomas), blastomas, leukemia, and lymphoma. Treatments using doxorubicin are complicated by its well-established cardiotoxic side effect, which affects approximately 16% of pediatric patients, can lead to heart failure requiring heart transplant, and limits doxorubicin’s clinical utilization. Despite more than 50 years of research in this field, there is still, at present, little potential for either predicting or preventing cardiotoxicity. There is an obvious need for novel and innovative approaches to overcome this hurdle. Candidate gene association studies and genome–wide association studies (GWAS) have identified many single nucleotide polymorphisms (SNPs) that are statistically correlated with doxorubicin–induced cardiotoxicity (DIC), yet experimental validation of these SNPs has not been feasible due to the difficulty in isolating and culturing human cardiomyocytes in vitro. In our recent work, we showed that patient–specific human induced pluripotent stem cell–derived cardiomyocytes (hiPSC–CM) are efficient predictors of a patient’s likelihood of developing DIC, confirming for the first time that there is a genomic basis to DIC. Although GWAS has proven to be a powerful methodology for informing such genomic bases, it detects correlation rather than causation, and identified SNPs commonly fail to be replicated in subsequent studies. Here, we hypothesize that hiPSC-CMs can be utilized in three different modalities to study genetic variants associated with DIC: firstly, to discover novel predictive SNPs; secondly, to validate SNPs; and thirdly, to examine the modulated pathways and determine genotype-specific cardioprotective methodologies. In Aim 1, we will recruit 100 pediatric cancer patients who were exposed to doxorubicin and assess the response of patient-derived hiPSC-CM to doxorubicin in vitro to validate our previous findings in a large pediatric cohort with diverse biological covariates to verify the power of this tool. In Aim 2, we will use these 100 patient-specific lines to identify drug response differential expression quantitative trait loci (deQTL), assessing biological covariates such as dose, age, sex, SF, and cancer diagnosis both individually and combined. We will then validate these variants with genome editing, and mechanistically examine pathways causative to DIC susceptibility concentrating on genes with known roles in cardiomyopathy, cardioprotection, and doxorubicin metabolism. In Aim 3, we will interrogate the rigor and reproducibility of >40 existing DIC SNP studies, using CRISPR/Cas9 to edit the gene of interest in control isogenic hiPSC lines then assess the response of hiPSC-CM to doxorubicin. We will then use the discoveries above to discover/repurpose genome-informed cardioprotective drugs to prevent DIC in a genotype-specific manner. In summary, this work will deliver us the genetic rationale for why patients experience DIC and provide 1, fully human validated SNP data for clinical application, and 2, novel cardioprotective drugs to attenuate DIC.