Machine Learning Identifies Clinical and Genetic Factors Associated With Anthracycline Cardiotoxicity in Pediatric Cancer Survivors

Marie-A Chaix, Neha Parmar, Caroline Kinnear, Myriam Lafreniere-Roula, Oyediran Akinrinade, Roderick Yao, Anastasia Miron, Emily Lam, Guoliang Meng, Anne Christie, Ashok Kumar Manickaraj, Stacey Marjerrison, Rejane Dillenburg, Mylène Bassal, Jane Lougheed, Shayna Zelcer, Herschel Rosenberg, David Hodgson, Leonard Sender, Paul Kantor, Cedric Manlhiot, James Ellis, Luc Mertens, Paul C Nathan, Seema Mital, Marie-A Chaix, Neha Parmar, Caroline Kinnear, Myriam Lafreniere-Roula, Oyediran Akinrinade, Roderick Yao, Anastasia Miron, Emily Lam, Guoliang Meng, Anne Christie, Ashok Kumar Manickaraj, Stacey Marjerrison, Rejane Dillenburg, Mylène Bassal, Jane Lougheed, Shayna Zelcer, Herschel Rosenberg, David Hodgson, Leonard Sender, Paul Kantor, Cedric Manlhiot, James Ellis, Luc Mertens, Paul C Nathan, Seema Mital

Abstract

Background: Despite known clinical risk factors, predicting anthracycline cardiotoxicity remains challenging.

Objectives: This study sought to develop a clinical and genetic risk prediction model for anthracycline cardiotoxicity in childhood cancer survivors.

Methods: We performed exome sequencing in 289 childhood cancer survivors at least 3 years from anthracycline exposure. In a nested case-control design, 183 case patients with reduced left ventricular ejection fraction despite low-dose doxorubicin (≤250 mg/m2), and 106 control patients with preserved left ventricular ejection fraction despite doxorubicin >250 mg/m2 were selected as extreme phenotypes. Rare/low-frequency variants were collapsed to identify genes differentially enriched for variants between case patients and control patients. The expression levels of 5 top-ranked genes were evaluated in human induced pluripotent stem cell-derived cardiomyocytes, and variant enrichment was confirmed in a replication cohort. Using random forest, a risk prediction model that included genetic and clinical predictors was developed.

Results: Thirty-one genes were differentially enriched for variants between case patients and control patients (p < 0.001). Only 42.6% case patients harbored a variant in these genes compared to 89.6% control patients (odds ratio: 0.09; 95% confidence interval: 0.04 to 0.17; p = 3.98 × 10-15). A risk prediction model for cardiotoxicity that included clinical and genetic factors had a higher prediction accuracy and lower misclassification rate compared to the clinical-only model. In vitro inhibition of gene-associated pathways (PI3KR2, ZNF827) provided protection from cardiotoxicity in cardiomyocytes.

Conclusions: Our study identified variants in cardiac injury pathway genes that protect against cardiotoxicity and informed the development of a prediction model for delayed anthracycline cardiotoxicity, and it also provided new targets in autophagy genes for the development of cardio-protective drugs. (Preventing Cardiac Sequelae in Pediatric Cancer Survivors [PCS2]; NCT01805778).

Keywords: AUC, area under the curve; CI, confidence interval; DMSO, dimethyl sulfoxide; DOX, doxorubicin; GSEA, gene set enrichment analysis; H2AX, H2A family member X; IC50, half-maximal inhibitory concentration; LV, left ventricular; LVEF, left ventricular ejection fraction; MAF, minor allele frequency; OR, odds ratio; PGP, Personal Genome Project; RF, random forest; SKAT, sequence kernel association test; SNV, single-nucleotide variant; anthracycline; cancer survivorship; cardiomyopathy; echocardiography; genomics; hiPSC-CM, human induced pluripotent stem cell–derived cardiomyocyte; mRNA, messenger RNA; machine learning; risk prediction.

Conflict of interest statement

The study was supported by the Canadian Cancer Society, Toronto, Canada; Canadian Institutes of Health Research, Toronto, Canada; the Pediatric Oncology Group of Ontario, Toronto, Canada; the Ontario Institute for Cancer Research, Toronto, Canada; Children’s Cancer and Blood disorders, Toronto, Canada; Ted Rogers Centre for Heart Research, Toronto, Canada; and the Labatt Family Heart Center at the Hospital for Sick Children, Toronto, Canada. Dr. Sender is Senior Vice President, Medical Affairs for Pediatric, Adolescent and Young Adult Oncology at NantKwest. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

© 2020 The Authors.

Figures

Graphical abstract
Graphical abstract
Central Illustration
Central Illustration
An Integrated Precision Approach to Predict Anthracycline Cardiotoxicity Whole-exome sequencing data from 289 pediatric cancer survivors with extreme phenotypes identified a higher burden of rare variants in control patients compared to case patients in 31 biologically relevant genes. The top-ranked genes were functionally evaluated in human induced pluripotent stem cell–derived cardiomyocytes (hiPSC-CMs) and variant enrichment was confirmed in a replication cohort. Targeted pathway inhibitors were more effective at reducing anthracycline-induced injury than dexrazoxane. Using random forest, clinical and genetic predictors were integrated into a prediction model for anthracycline cardiotoxicity.
Figure 1
Figure 1
Study Cohort and CONSORT Diagram (A) Study cohort: Scatterplot showing the distribution of left ventricular ejection fraction (LVEF) against anthracycline cumulative dose, with control patients in blue (n = 121), case patients in red (n = 238), and the rest in gray (n = 357). (B) Consolidated Standards of Reporting Trials (CONSORT) diagram: selection of case and control patients for exome sequencing in the discovery cohort. ∗Participants with complete data and good-quality DNA. PCS2 = Preventing Cardiac Sequelae in Pediatric Cancer Survivors.
Figure 2
Figure 2
Genes Associated With Anthracycline Cardiotoxicity (A) A total of 28 genes showed differential enrichment between case and control patients by at least 2 methods (p < 0.001), and 3 additional biologically relevant genes (orange) were significant by at least 1 method. (B) Burden of rare and low-frequency single-nucleotide variants in the 31 prioritized genes was higher in control patients compared to case patients (p < 0.001). (C) Forest plot showing the estimated odds ratios (95% confidence intervals) for the 31 top genes in case versus control patients using Fisher exact test. Genes in bold were prioritized for functional studies.
Figure 2
Figure 2
Genes Associated With Anthracycline Cardiotoxicity (A) A total of 28 genes showed differential enrichment between case and control patients by at least 2 methods (p < 0.001), and 3 additional biologically relevant genes (orange) were significant by at least 1 method. (B) Burden of rare and low-frequency single-nucleotide variants in the 31 prioritized genes was higher in control patients compared to case patients (p < 0.001). (C) Forest plot showing the estimated odds ratios (95% confidence intervals) for the 31 top genes in case versus control patients using Fisher exact test. Genes in bold were prioritized for functional studies.
Figure 3
Figure 3
Pathways Associated With Anthracycline Cardiotoxicity (A) GeneMania analysis identified 46 interacting genes, including 26 of the 31 top genes. Large circles represent significantly associated genes; small circles represent other interacting genes. Physical interaction (pink lines), coexpression (purple lines), colocalization (blue lines), shared protein domains (gray-yellow lines), genetic interaction (green lines), and predicted (orange lines). (B) Gene set enrichment analysis identified the top-ranked pathways to which the genes mapped (p < 0.001). The solid bar shows number of significant genes in each pathway (p < 0.001); the dashed bar represents the total genes.
Figure 3
Figure 3
Pathways Associated With Anthracycline Cardiotoxicity (A) GeneMania analysis identified 46 interacting genes, including 26 of the 31 top genes. Large circles represent significantly associated genes; small circles represent other interacting genes. Physical interaction (pink lines), coexpression (purple lines), colocalization (blue lines), shared protein domains (gray-yellow lines), genetic interaction (green lines), and predicted (orange lines). (B) Gene set enrichment analysis identified the top-ranked pathways to which the genes mapped (p < 0.001). The solid bar shows number of significant genes in each pathway (p < 0.001); the dashed bar represents the total genes.
Figure 4
Figure 4
Effect of Anthracycline in hiPSC-CMs (A) The 24-h DOX treatment caused a dose-dependent decrease in cell function and viability in hiPSC-CMs measured using the cell index. (B) Presto blue cell viability assay demonstrated a decrease in metabolic activity and proliferation with increasing DOX doses. The values represent the average relative fluorescence from 3 independent experiments. (C) Representative immunofluorescence images showing increased γ-H2AX staining (green) (white arrow), a DNA damage marker, in the nuclei (blue DAPI staining) of DOX-treated cells. (D) DOX treatment increased average γ-H2AX foci per nucleus compared to untreated cells. Error bars represent mean ± SD for 3 independent biological replicates. ∗p < 0.05; ∗∗∗p < 0.001. CMC = combined multivariate and collapsing; DAPI = 4′,6-diamidino-2-phenylindole; DMSO = dimethyl sulfoxide; DOX = doxorubicin; H2AX = H2A family member X; hiPSC-CM = human induced pluripotent stem cell–derived cardiomyocyte; μM = μmol/l.
Figure 5
Figure 5
Effect of Targeted Gene Inhibition on DOX-Induced Cardiotoxicity in hiPSC-CMs (A) RT-qPCR analysis of PGP17 hiPSC-CMs (3 biological replicates, each containing 3 technical replicates) treated with 0.1 μmol/L DOX demonstrated a significant increase in gene expression levels of ZNF827, ELAC2, and PI3KR2.(B) PGP17 and (C) PGP14 hiPSC-CMs were pre-treated with DMSO or an inhibitor for 24 h and then exposed to DOX for 24 h. Dose-response curves for cell index, that is, CM viability, and IC50 values for DOX with and without inhibitors are shown. Data are mean ± SD for 3 independent replicates. ∗p < 0.05; ∗∗p < 0.01. PGP = Personal Genome Project; RT-qPCR = quantitative reverse transcription polymerase chain reaction; other abbreviations as in Figure 4.
Figure 6
Figure 6
Accuracy Measures of Prediction Models Using Random Forest Boxplots representing the accuracy measures for the overall dataset, the testing set, and the training set for the (A) clinical model, (B) genetic model, (C) combined model, and (D) receiver operator characteristic AUC for model-derived prediction of anthracycline cardiotoxicity. Clinical model AUC: 0.59 (black), genetic model AUC: 0.71 (blue), combined model AUC: 0.72 (red). AUC = area under the curve; FNR = false negative rate; FPR = false positive rate; MC = misclassification; NPV = negative predictive value; PPV = positive predictive value; Sens = sensitivity; Sn = sensitivity; Sp = specificity; Spec = specificity.
Figure 6
Figure 6
Accuracy Measures of Prediction Models Using Random Forest Boxplots representing the accuracy measures for the overall dataset, the testing set, and the training set for the (A) clinical model, (B) genetic model, (C) combined model, and (D) receiver operator characteristic AUC for model-derived prediction of anthracycline cardiotoxicity. Clinical model AUC: 0.59 (black), genetic model AUC: 0.71 (blue), combined model AUC: 0.72 (red). AUC = area under the curve; FNR = false negative rate; FPR = false positive rate; MC = misclassification; NPV = negative predictive value; PPV = positive predictive value; Sens = sensitivity; Sn = sensitivity; Sp = specificity; Spec = specificity.

References

    1. Octavia Y., Tocchetti C.G., Gabrielson K.L., Janssens S., Crijns H.J., Moens A.L. Doxorubicin-induced cardiomyopathy: from molecular mechanisms to therapeutic strategies. J Mol Cell Cardiol. 2012;52:1213–1225.
    1. Nathan P.C., Amir E., Abdel-Qadir H. Cardiac outcomes in survivors of pediatric and adult cancers. Can J Cardiol. 2016;32:871–880.
    1. Lipshultz S.E., Lipsitz S.R., Mone S.M. Female sex and higher drug dose as risk factors for late cardiotoxic effects of doxorubicin therapy for childhood cancer. N Engl J Med. 1995;332:1738–1743.
    1. Hershman D.L., McBride R.B., Eisenberger A., Tsai W.Y., Grann V.R., Jacobson J.S. Doxorubicin, cardiac risk factors, and cardiac toxicity in elderly patients with diffuse B-cell non-Hodgkin’s lymphoma. J Clin Oncol. 2008;26:3159–3165.
    1. Pein F., Sakiroglu O., Dahan M. Cardiac abnormalities 15 years and more after adriamycin therapy in 229 childhood survivors of a solid tumour at the Institut Gustave Roussy. Br J Cancer. 2004;91:37–44.
    1. Aminkeng F., Bhavsar A.P., Visscher H. A coding variant in RARG confers susceptibility to anthracycline-induced cardiotoxicity in childhood cancer. Nat Genet. 2015;47:1079–1084.
    1. Schneider B.P., Shen F., Gardner L. Genome-wide association study for anthracycline-induced congestive heart failure. Clin Cancer Res. 2017;23:43–51.
    1. Wang X., Sun C.L., Quinones-Lombrana A. CELF4 variant and anthracycline-related cardiomyopathy: a Children’s Oncology Group genome-wide association study. J Clin Oncol. 2016;34:863–870.
    1. Wells Q.S., Veatch O.J., Fessel J.P. Genome-wide association and pathway analysis of left ventricular function after anthracycline exposure in adults. Pharmacogenet Genomics. 2017;27:247–254.
    1. Linschoten M., Teske A.J., Cramer M.J., van der Wall E., Asselbergs F.W. Chemotherapy-related cardiac dysfunction: a systematic review of genetic variants modulating individual risk. Circ Genom Precis Med. 2018;11
    1. Visscher H., Ross C.J., Rassekh S.R. Pharmacogenomic prediction of anthracycline-induced cardiotoxicity in children. J Clin Oncol. 2012;30:1422–1428.
    1. Skitch A., Mital S., Mertens L. Novel approaches to the prediction, diagnosis and treatment of cardiac late effects in survivors of childhood cancer: a multi-centre observational study. BMC Cancer. 2017;17:519–528.
    1. Children’s Oncology Group . October 2013. Long-term follow up guidelines for survivors of childhood, adolescent, and young adult cancer. Available at: Accessed September 23, 2020.
    1. Lang R.M., Bierig M., Devereux R.B. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr. 2005;18:1440–1463.
    1. Cardinale D., Colombo A., Bacchiani G. Early detection of anthracycline cardiotoxicity and improvement with heart failure therapy. Circulation. 2015;131:1981–1988.
    1. DePristo M.A., Banks E., Poplin R. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat Genet. 2011;43:491–498.
    1. International HapMap C., Frazer K.A., Ballinger D.G. A second generation human haplotype map of over 3.1 million SNPs. Nature. 2007;449:851–861.
    1. Lek M., Karczewski K.J., Minikel E.V. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536:285–291.
    1. Tate J.G., Bamford S., Jubb H.C. COSMIC: the catalogue of somatic mutations in cancer. Nucleic Acids Res. 2018;47:D941–D947.
    1. Li B., Leal S.M. Methods for detecting associations with rare variants for common diseases: application to analysis of sequence data. Am J Hum Genet. 2008;83:311–321.
    1. Wu M.C., Lee S., Cai T., Li Y., Boehnke M., Lin X. Rare-variant association testing for sequencing data with the sequence kernel association test. Am J Hum Genet. 2011;89:82–93.
    1. Lee S., Emond M.J., Bamshad M.J. Optimal unified approach for rare-variant association testing with application to small-sample case-control whole-exome sequencing studies. Am J Hum Genet. 2012;91:224–237.
    1. Warde-Farley D., Donaldson S.L., Comes O. The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Res. 2010;38:W214–W220.
    1. Subramanian A., Tamayo P., Mootha V.K. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–15550.
    1. Hildebrandt M.R., Reuter M.S., Wei W. Precision health resource of control iPSC lines for versatile multilineage differentiation. Stem Cell Reports. 2019;13:1126–1141.
    1. Reuter M.S., Walker S., Thiruvahindrapuram B. The Personal Genome Project Canada: findings from whole genome sequences of the inaugural 56 participants. CMAJ. 2018;190:E126–E136.
    1. Zhang X., Guo L., Zeng H. Multi-parametric assessment of cardiomyocyte excitation-contraction coupling using impedance and field potential recording: a tool for cardiac safety assessment. J Pharmacol Toxicol Methods. 2016;81:201–216.
    1. Rao X., Huang X., Zhou Z., Lin X. An improvement of the 2ˆ(-delta CT) method for quantitative real-time polymerase chain reaction data analysis. Biostat Bioinforma Biomath. 2013;3:71–85.
    1. Wang S., Konorev E.A., Kotamraju S., Joseph J., Kalivendi S., Kalyanaraman B. Doxorubicin induces apoptosis in normal and tumor cells via distinctly different mechanisms. intermediacy of H2O2- and p53-dependent pathways. J Biol Chem. 2004;279:25535–25543.
    1. Foglesong P.D., Reckord C., Swink S. Doxorubicin inhibits human DNA topoisomerase I. Cancer Chemother Pharmacol. 1992;30:123–125.
    1. Conomos D., Reddel R.R., Pickett H.A. NuRD-ZNF827 recruitment to telomeres creates a molecular scaffold for homologous recombination. Nat Struct Mol Biol. 2014;21:760–770.
    1. Zilinyi R., Czompa A., Czegledi A. The cardioprotective effect of metformin in doxorubicin-induced cardiotoxicity: the role of autophagy. Molecules. 2018;23:1184–1196.
    1. Kitani T., Ong S.G., Lam C.K. Human-induced pluripotent stem cell model of trastuzumab-induced cardiac dysfunction in patients with breast cancer. Circulation. 2019;139:2451–2465.
    1. Division of Biomedical Statistics and Informatics, Mayo Clinic Research. Biomedical statistics and informatics software packages. 2013. Available at. Accessed November 2020.
    1. Drafts B.C., Twomley K.M., D’Agostino R., Jr. Low to moderate dose anthracycline-based chemotherapy is associated with early noninvasive imaging evidence of subclinical cardiovascular disease. J Am Coll Cardiol Img. 2013;6:877–885.
    1. Li D., Lewinger J.P., Gauderman W.J., Murcray C.E., Conti D. Using extreme phenotype sampling to identify the rare causal variants of quantitative traits in association studies. Genet Epidemiol. 2011;35:790–799.
    1. Gomes A.V., Venkatraman G., Davis J.P. Cardiac troponin T isoforms affect the Ca(2+) sensitivity of force development in the presence of slow skeletal troponin I: insights into the role of troponin T isoforms in the fetal heart. J Biol Chem. 2004;279:49579–49587.
    1. Hassan B., Akcakanat A., Holder A.M., Meric-Bernstam F. Targeting the PI3-kinase/Akt/mTOR signaling pathway. Surg Oncol Clin N Am. 2013;22:641–664.
    1. Liu E.Y., Xu N., O’Prey J. Loss of autophagy causes a synthetic lethal deficiency in DNA repair. Proc Natl Acad Sci U S A. 2015;112:773–778.
    1. Deng S., Yan T., Jendrny C. Dexrazoxane may prevent doxorubicin-induced DNA damage via depleting both topoisomerase II isoforms. BMC Cancer. 2014;14:842–854.
    1. Chen S., Fisher R.C., Signs S. Inhibition of PI3K/Akt/mTOR signaling in PI3KR2-overexpressing colon cancer stem cells reduces tumor growth due to apoptosis. Oncotarget. 2017;8:50476–50488.

Source: PubMed

3
Subskrybuj