Genetic polymorphisms in the carbonyl reductase 3 gene CBR3 and the NAD(P)H:quinone oxidoreductase 1 gene NQO1 in patients who developed anthracycline-related congestive heart failure after childhood cancer


  • Presented in part at the 2006 Annual Meeting of The American Society of Clinical Oncology, Atlanta, Georgia, June 206, 2006

  • Childhood Cancer Survivor Study (CCSS) investigators and institutions: Leslie L. Robison, PhD, Melissa Hudson, MD, Greg Armstrong, MD, and Daniel M. Green, MD (St. Jude Children's Research Hospital, Memphis, Tenn); Lillian Meacham, MD and Ann Mertens, PhD (Children's Healthcare of Atlanta/Emory University, Atlanta, Ga); Joanna Perkins, MD and Maura O'Leary, MD (Children's Hospitals and Clinics of Minnesota, St. Paul, Minn); Debra Friedman, MD, MPH and Thomas Pendergrass, MD (Children's Hospital and Medical Center, Seattle, Wash); Brian Greffe, MD and Lorrie Odom, MD (Children's Hospital, Denver, Colo); Kathy Ruccione, RN, MPH (Children's Hospital, Los Angeles, Calif); John Mulvihill, MD (Children's Hospital, Oklahoma City, Okla); Jill Ginsberg, MD and Anna Meadows, MD (Children's Hospital of Philadelphia, Philadelphia, Pa); Jean Tersak, MD, A. Kim Ritchey, MD, and Julie Blatt, MD (Children's Hospital of Pittsburgh, Pittsburgh, Pa); Gregory Reaman, MD and Roger Packer, MD (Children's National Medical Center, Washington, DC); Stella M. Davies, MD, PhD (Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio); Smita Bhatia, MD, MPH (City of Hope, Los Angeles, Calif); Lisa Diller, MD, Holcombe Grier, MD, and Frederick Li, MD (Dana-Farber Cancer Institute/Childrens Hospital, Boston, Mass); Wendy M. Leisenring, ScD and John Potter, MD, PhD (Fred Hutchinson Cancer Research Center, Seattle, Wash); Mark Greenberg, MBChB and Paul C. Nathan, MD (Hospital for Sick Children, Toronto, Ontario); John Boice, ScD (International Epidemiology Institute, Rockville, Md); Vilmarie Rodriguez, MD, W. Anthony Smithson, MD, and Gerald Gilchrist, MD (Mayo Clinic, Rochester, Minn); Charles A. Sklar, MD and Kevin Oeffinger, MD (Memorial Sloan-Kettering Cancer Center, New York, NY); Jerry Finklestein, MD (Miller Children's Hospital, Long Beach, Calif); Roy Wu, PhD and Peter Inskip, ScD (National Cancer Institute, Bethesda, Md); Amanda Termuhlen, MD, Frederick Ruymann, MD, Stephen Qualman, MD, and Sue Hammond, MD (Nationwide Children's Hospital, Columbus, Ohio); Terry A. Vik, MD and Robert Weetman, MD (Riley Hospital for Children, Indianapolis, Ind); Martin Brecher, MD (Roswell Park Cancer Institute, Buffalo, NY); Robert Hayashi, MD and Teresa Vietti, MD (St. Louis Children's Hospital, Mo); Neyssa Marina, MD, Sarah S. Donaldson, MD, and Michael P. Link, MD (Stanford University School of Medicine, Stanford, Calif); Zoann Dreyer, MD (Texas Children's Hospital, Houston, Tex); Kimberly Whelan, MD, MSPH, Jane Sande, MD, and Roger Berkow, MD (University of Alabama, Birmingham, Ala); Yutaka Yasui, PhD (University of Alberta, Edmonton, Alberta); Jacqueline Casillas, MD, MSHS and Lonnie Zeltzer, MD (University of California-Los Angeles, Low Angeles, Calif); Robert Goldsby, MD and Arthur Ablin, MD (University of California-San Francisco, San Francisco, Calif); Raymond Hutchinson, MD (University of Michigan, Ann Arbor, Mich); Joseph Neglia, MD, MPH (University of Minnesota, Minneapolis, Minn); Dennis Deapen, DrPH (University of Southern California, Los Angeles, Calif); Norman Breslow, PhD (University of Washington, Seattle, Wash); Dan Bowers, MD, Gail Tomlinson, MD, and George R. Buchanan, MD (University of Texas Southwestern Medical Center, Dallas, Tex); Louise Strong, MD and Marilyn Stovall, MPH, PhD (University of Texas M. D. Anderson Cancer Center, Houston, Tex).

    The CCSS is a collaborative, multi-institutional project, funded as a resource by the National Cancer Institute, of individuals who survived ≥5 years after a diagnosis of childhood cancer. CCSS is a retrospectively ascertained cohort of 20,346 childhood cancer survivors who were diagnosed before age 21 years between 1970 and 1986 and approximately 4000 siblings of survivors who serve as a control group. The cohort was assembled through the efforts of 26 participating clinical research centers in the United States and Canada. The study currently is funded by a U24 resource grant (National Cancer Institute grant U24 CA55727) awarded to St. Jude Children's Research Hospital. Currently, investigators are in the process of expanding the cohort to include an additional 14,000 childhood cancer survivors who were diagnosed before age 21 years between 1987 and 1999. For information on how to access and use the CCSS resource, visit



Exposure to anthracyclines as part of cancer therapy has been associated with the development of congestive heart failure (CHF). The potential role of genetic risk factors in anthracycline-related CHF remains to be defined. Thus, in this study, the authors examined whether common polymorphisms in candidate genes involved in the pharmacodynamics of anthracyclines (in particular, the nicotinamide adenine dinucleotide phosphate:quinone oxidoreductase 1 gene NQO1 and the carbonyl reductase 3 gene CBR3) had an impact on the risk of anthracycline-related CHF.


A nested case–control study was conducted within a cohort of 1979 patients enrolled in the Childhood Cancer Survivor Study who received treatment with anthracyclines and had available DNA. Thirty patients with CHF (cases) and 115 matched controls were genotyped for polymorphisms in NQO1 (NQO1*2) and CBR3 (the CBR3 valine [V] to methionine [M] substitution at position 244 [V244M]). Enzyme activity assays with recombinant CBR3 isoforms (CBR3 V244 and CBR3 M244) and the anthracycline substrate doxorubicin were used to investigate the functional impact of the CBR3 V244M polymorphism.


Multivariate analyses adjusted for sex and primary disease recurrence were used to test for associations between the candidate genetic polymorphisms (NQO1*2 and CBR3 V244M) and the risk of CHF. Analyses indicated no association between the NQO1*2 polymorphism and the risk of anthracycline-related CHF (odds ratio [OR], 1.04; P = .97). There was a trend toward an association between the CBR3 V244M polymorphism and the risk of CHF (OR, 8.16; P = .056 for G/G vs A/A; OR, 5.44; P = .092 for G/A vs A/A). In line, recombinant CBR3 V244 (G allele) synthesized 2.6-fold more cardiotoxic doxorubicinol per unit of time than CBR3 M244 (A allele; CBR3 V244 [8.26 ± 3.57 nmol/] vs CBR3 M244 [3.22 ± 0.67 nmol/]; P = .01).


The functional CBR3 V244M polymorphism may have an impact on the risk of anthracycline-related CHF among childhood cancer survivors by modulating the intracardiac formation of cardiotoxic anthracycline alcohol metabolites. Larger confirmatory case–control studies are warranted. Cancer 2008. © 2008 American Cancer Society.

Anthracycline-related congestive heart failure (CHF) is an important long-term complication among childhood cancer survivors. Preclinical abnormalities of cardiac structure and function have been reported in 60% of patients who received anthracyclines,1 and the risk of overt, clinically symptomatic CHF has been estimated at from 4% to 5%.2

Potential risk factors for anthracycline-related CHF include total cumulative dose, being a woman, radiation therapy, and younger age at diagnosis.3 The pathogenesis of anthracycline-related chronic cardiotoxicity appears to be mediated by a combination of oxidative damage and perturbations in cardiac iron homeostasis induced by anthracycline C-13 alcohol metabolites.4 The potential role of genetic variability in the pathogenesis of chronic cardiotoxicity remains to be elucidated. Therefore, the objective of the current study was to investigate the impact of 2 functional candidate genetic polymorphisms on the risk of anthracycline-related CHF among childhood cancer survivors.

NAD(P)H:quinone oxidoreductase 1 (NQO1) participates in protection against intracellular oxidative stress, and many pro-oxidant drugs, including anthracyclines, induce basal NQO1 activity.5 NQO1 catalyzes the 2-electron reduction of anthracyclines to yield hydroquinone derivatives, thereby preventing the generation of unstable semiquinone radicals that may promote oxidative damage mediated by reactive oxygen species.6 NQO1 also contributes to maintain intracellular levels of bioactive vitamin E in situations of oxidative stress, and vitamin E administration has been evaluated as a cardioprotective strategy during therapy with anthracyclines.1 Therefore, we hypothesized that relatively low NQO1 activity dictated by the NQO1*2 allele (q = 0.16–0.20) may have an impact on the risk of anthracycline-related CHF. In the human myocardium, cytosolic carbonyl reductases (CBRs), and aldo/keto reductases (AKRs) catalyze the 2-electron reduction of the anthracycline side chain C-13 carbonyl group to form cardiotoxic alcohol metabolites (eg, doxorubicinol, daunorubicinol). A polymorphism in the carbonyl reductase 3 gene CBR3 (the CBR3 valine [V] to methionine [M] substitution at position 244 [V244M]) is relatively common in different ethnic groups (whites in the U.S.: q = 0.31; blacks in the U.S.: q = 0.51) and has an impact on CBR3 catalytic activity.7 We hypothesized that functional genetic polymorphisms in CBRs may have an impact on the risk of anthracycline-related cardiotoxicity.

Thus, childhood cancer survivors with anthracycline-related CHF (cases) and matched controls were genotyped for the NQO1*2 (oxidative damage) and CBR3 V244M polymorphisms (perturbations in cardiac iron homeostasis induced by cardiotoxic anthracycline C-13 alcohol metabolites), and multivariate analyses were used to evaluate the risks of anthracycline-related CHF associated with each genotype. In addition, kinetic assays with recombinant CBR3 isoforms (CBR3 V244 and CBR3 M244) and the substrate doxorubicin were used to determine the functional impact of CBR3 V244M.


The Childhood Cancer Survivor Study Cohort

The Childhood Cancer Survivor Study (CCSS) is a multi-institutional, retrospective cohort of 5-year survivors of childhood cancer that was designed to study the late effects of cancer therapy.8 Eligibility criteria for participation in the CCSS cohort were: 1) diagnosis of leukemia, brain tumor, Hodgkin lymphoma, non-Hodgkin lymphoma, Wilms tumor, neuroblastoma, soft tissue sarcoma, or bone tumor between 1970 and 1986 at 1 of the 26 participating institutions; 2) age <21 years at diagnosis; and 3) and survival ≥5 years after diagnosis. From this CCSS cohort of 14,352 individuals, we limited subsequent eligibility for the current study to those participants who were exposed to anthracyclines and provided a buccal cell sample (n = 1979 participants). Data were collected on a wide range of exposures and outcomes, including the development of CHF, as well as demographic characteristics, smoking and alcohol consumption, and family history. Each participating center's institutional review board reviewed and approved the CCSS protocol.

Cancer Therapy

Detailed therapy-related information was obtained from medical records, including cumulative doses of anthracyclines. Radiation exposure to the heart was summarized as the total cumulative dose in centigrays to the heart.

Ascertainment of CHF

Congestive heart failure was ascertained through self-report by the study participants. Positive cases were validated by the subsequent administration of a telephone interview, during which the signs and symptoms of CHF as well as the use of medications for management of the CHF were reviewed. The details of the script used to validate this outcome are presented in Figure 1. The responses from the script were reviewed independently by 2 physicians, and the information was used to validate CHF.

Figure 1.

Telephone script for validating cases of congestive heart failure.

Genomic DNA Collection, Extraction, and Storage

Participants were instructed to rinse their mouth with mouthwash and to return the used mouthwash to the laboratory in a sterile container. Samples were centrifuged, and cells were pelleted and washed. DNA samples were extracted by using the Puregene kit (Gentra Systems, Minneapolis, Minn). A median of 64 μg of DNA per sample has been obtained for the cohort overall.

Nested Case–Control Study

The cohort of 1979 patients who had exposure to anthracyclines and availability of buccal cell specimens provided us with a sampling frame from which to select controls. For each case of CHF, multiple controls (median, 4 controls; range, 1–4 controls) were selected randomly from within the cohort by using the following matching criteria: age at diagnosis (95%: ±5 years), race/ethnicity (white, black, or other), duration of follow-up (controls were followed at least until the case developed the event), anthracycline dose (<100 mg/m2, 100–349 mg/m2, 350–499 mg/m2, or >500 mg/m2), and radiation to the heart (matched for 81% of controls; this factor was included in all models to adjust for incomplete matching). Thirty cases with CHF and 115 matched controls were included in this analysis.


Laboratory personnel were blinded to case–control status. The NQO1*2 polymorphism (reference single-nucleotide polymorphism [SNP] identification number [rs1800566]) was examined with a validated polymerase chain reaction (PCR)-restriction fragment length polymorphism technique as described.9 The CBR3 V244M polymorphism (rs1056892) was analyzed with a validated assay for allelic discrimination with specific fluorescent probes (Applied Biosystems) also as described.7

Kinetic Analysis

Recombinant CBR3 V244 and CBR3 M244 were obtained as described previously.7 The amounts of doxorubicinol synthesized by CBR3 V244 and CBR3 M244 were quantified with a validated high-performance liquid chromatography (HPLC) coupled to tandem mass spectrometry (MS) assay (HPLC-MS/MS) as described.10 Incubation mixtures (1.0 mL) contained 100 mM potassium phosphate buffer, pH 7.4; 200 μM nicotinamide adenine dinucleotide phosphate (NADPH); 500 μM doxorubicin; and enzyme (CBR3 V244 or CBR3 M244). Mixtures were incubated for 5 hours at 37 °C, and reactions were stopped by placing samples on ice. Samples were stored at −80 °C until HPLC-MS/MS analysis. Data shown are the averages ± standard deviation of 2 independent kinetic experiments for each CBR3 isoform.

Statistical Analysis

Analyses of the case–control study measured the degree of association of genetic polymorphisms with adverse outcome (CHF) by estimation of odds ratios (ORs) using conditional logistic regression to investigate the simultaneous effects of several variables. Candidate confounding variables for inclusion in the multivariate model were sex (with men as the reference group), smoking history (current, former, never, with never as the reference group), first recurrence of original disease, heart in radiation beam (yes/no), and family history of heart disease (yes/no). Selection of covariates for the final multivariate model included any factors that changed the effect of the genotype on CHF by >10% or that were significant themselves. Two-sided P values <.05 were considered significant. CBR3 and NQO1 genotype distributions were consistent with those predicted under conditions of Hardy-Weinberg equilibrium, with P values >.05. The overall distributions of CBR3 and NQO1 alleles were similar (P > .05) to the distributions reported previously.7, 11


The distribution of clinical characteristics (including those used for matching) of the 30 cases with anthracycline-related CHF and the 115 matched controls, as well as their corresponding CBR3 and NQO1 genotypes, are summarized in Table 1. Univariate and multivariate analyses of CHF risk indicated no apparent association between NQO1 genotype status and the risk of CHF (univariate: OR, 1.26 [P = .84] for C/C vs T/T; OR, 0.65 [P = .72] for C/T vs T/T) (Table 2). There was no marked change to this result in multivariate models that were adjusted for sex and primary disease recurrence (multivariate: OR, 1.04 [P = .97] for C/C vs T/T; OR, 0.52 [P = .60] for C/T vs T/T; not shown in Table). However, multivariate analysis adjusting for the same risk factors revealed a trend toward an association between CBR3 genotype status and the risk of CHF after treatment with anthracyclines (OR, 8.16 [P = .056] for G/G vs A/A; OR, 5.44 [P = .092] for G/A vs A/A) (Table 2). Kinetic assays with recombinant CBR3 protein isoforms and the substrate doxorubicin revealed a 2.6-fold higher rate of doxorubicinol synthesis for CBR3 V244 (G allele) compared with CBR3 M244 (A allele: CBR3 V244 [8.26 ± 3.57 nmol/] vs CBR3 M244 [3.22 ± 0.67 nmol/]; P = .01).

Table 1. Clinical Characteristics of the Study Population
VariableNo. of patients (%)
Cases: CHF, n = 30*Controls, n = 115
  • CHF indicates congestive heart failure; SD. standard deviation from the mean; cGy, centigrays; CRB3, carbonyl reductase 3 gene; V244M, valine to methionine substitution at position 244; G, guanosine; A, adenosine; NQO1*2, polymorphism of the nicotinamide adenine dinucleotide phosphate:quinone oxidoreductase 1 (NQO1) gene; C, cytidine; T, thymidine.

  • *

    Demographic and clinical data were collected for 30 CHF cases and 115 matched controls. CBR3 and NQO1 genotypes were analyzed in 29 CHF cases for both genotypes and in 111 controls and 110 controls, respectively.

  • Matching variables. Because there is some variation in the number of controls per case, the percentage of controls and cases in each category of a specific matching variable may not be identical.

Mean±SD age at primary diagnosis, y10.3 ± 6.59.1 ± 5.8
 White28 (93)108 (94)
 Black1 (3)2 (2)
 Other1 (3)5 (4)
Cumulative anthracycline exposure, mg/m2  
 <1001 (3)2 (2)
 100–35013 (43)46 (40)
 350–5007 (23)31 (27)
 >5009 (30)36 (31)
Radiation to the heart, cGy  
 No radiation12 (40)52 (45)
 <250012 (40)44 (38)
 2500–45006 (20)19 (17)
 >45000 (0)0 (0)
Primary diagnosis  
 Hodgkin lymphoma8 (27)17 (15)
 Leukemia7 (23)43 (37)
 Bone cancer6 (20)23 (20)
 Soft tissue sarcoma3 (10)15 (13)
 Non-Hodgkin lymphoma5 (17)7 (6)
 Other1 (3)10 (9)
History of smoking; n = 111 controls  
 Never smoked26 (87)82 (71)
 Former smoker2 (7)11 (10)
 Current smoker2 (7)18 (16)
 Women20 (67)58 (50)
 Men10 (33)57 (50)
Family history of heart disease3 (10)11 (10)
CBR3 V244M genotype; n = 29 cases, n = 111 controls  
 A/A2 (7)14 (12)
 G/A15 (50)54 (47)
 G/G12 (40)43 (37)
NQO1*2 genotype; n = 29 cases, n = 110 controls  
 T/T1 (3)4 (4)
 C/T6 (20)37 (32)
 C/C22 (73)69 (60)
Table 2. Univariate and Multivariate Evaluation of Risk Factors for Anthracycline-related Congestive Heart Failure
VariableUnivariate modelMultivariate model*
  • OR indicates odds ratio; CI, confidence interval; CRB3, carbonyl reductase 3 gene; V244M, valine to methionine substitution at position 244; G, guanosine; A, adenosine; NQO1*2, polymorphism of the nicotinamide adenine dinucleotide phosphate:quinone oxidoreductase 1 (NQO1) gene; C, cytidine; T, thymidine.

  • *

    Factors that are included in the multivariate model are those with results shown under “Multivariate.” All models, including the univariate model, also are adjusted for heart in radiation beam to account for partial case–control matching on this factor.

CBR3 V244M      
 G/G vs A/A5.630.80–39.57.0838.160.95–70.10.056
 G/A vs A/A3.660.64––39.14.092
NQO1*2 genotype      
 C/C vs T/T1.260.14–11.39.84   
 C/T vs T/T0.650.06–6.77.72   
Smoking history      
 Current smoker0.320.06–1.65.17   
 Former smoker0.530.10–2.79.45   
First recurrence4.741.02–22.05.0486.351.13–35.75.036
Family history of heart disease0.920.25–3.45.90   


The objective of this pilot study was to evaluate the impact of candidate SNPs on the risk of anthracycline-related CHF among childhood cancer survivors. Our selection criteria for candidate genetic polymorphisms was based on a “unifying hypothesis” postulating that anthracycline-related chronic cardiotoxicity results from a combination of oxidative stress and metabolic perturbations induced by the intracardiac formation of C-13 anthracycline alcohol metabolites.4

The common NQO1*2 polymorphism (whites, q = 0.16–0.20; blacks, q = 0.16) encodes for an NQO1 protein variant with decreased intracellular stability; and, in consequence, low NQO1 activity has been detected in individuals with heterozygous (NQO1*1/*2) or homozygous (NQO1*2/*2) genotypes.9, 12, 13 Our data failed to demonstrate an impact of NQO1 genotype status on the risk of anthracycline-related CHF and are similar to those reported by Wojnowski et al.14 It is possible that other proteins involved in protection against oxidative stress compensate for the potential deficit resulting from low NQO1 activity dictated by the NQO1*2 allele.

In humans, CBR activity is the main source of quinone detoxification, and biochemical studies have demonstrated that AKRs have 7- to 18-fold lower catalytic efficiencies (VMax/Km) for the reduction of anthracycline substrates than CBRs.9, 15, 16 There are 2 monomeric CBRs (CBR1 and CBR3) encoded for genes located in chromosome 21 (CBR1 21q22.13 and CBR3 21q22.2), and the resulting proteins have 72% of identity at the amino-acid level. The key role of CBR activity during the pathogenesis of anthracycline-related cardiotoxicity has been pinpointed in biochemical and murine studies. For example, mice with a null allele of Cbr1 (Cbr1+/−) had low plasma levels of doxorubicinol and had a significantly lower incidence of anthracycline-related cardiotoxicity compared with animals that had 2 active Cbr1 alleles (Cbr1+/+).17 Conversely, mice that overexpressed human CBR1 in the heart had high intracardiac levels of doxorubicinol and increased signs of doxorubicin-related cardiotoxicity.18 Recently, we demonstrated that the CBR3 V244M polymorphism results in CBR3 protein isoforms (CBR3 V244 and CBR3 M244) with distinctive catalytic properties toward the prototypical quinone substrate menadione.7 Comparative 3-dimensional analysis suggested that the V244M amino-acid substitution is positioned in a region that is critical for interactions with the NADPH cofactor. In the current study, we extended our observations by measuring the maximal activities of CBR3 V244 and CBR3 M244 with the anthracycline substrate doxorubicin. CBR3 V244 had a 2.6-fold higher rate of doxorubicinol synthesis than CBR3 M244. It is noteworthy that we detected a 8-fold increased risk of anthracycline-related CHF among individuals who were homozygous for the G allele (CBR3 V244 isoform) compared with individuals who were homozygous for the A allele (CBR3 M244 isoform). On the basis of our genetic findings and the current kinetic evidence, we hypothesize that individuals with the CBR3 V244M homozygous G genotype may present with an increased rate of myocardial synthesis of cardiotoxic C-13 alcohol metabolites compared with individuals who are homozygous for the A allele. CBR3 messenger RNA is detected readily in the human heart by reverse transcriptase-PCR (J.G.B., unpublished observations), but the precise contribution of polymorphic CBR3 to the intracardiac formation of cardiotoxic anthracycline C-13 alcohol metabolites remains to be determined. Our findings support the notion that polymorphic CBRs may have an impact on the pharmacodynamics of anthracyclines and, concomitantly, on the risk of cardiotoxicity.

To our knowledge, this pilot study is the first of its kind that attempts to identify genetic risk factors for anthracycline-related CHF among pediatric cancer survivors. These preliminary results are limited by a relatively small sample. Larger confirmatory studies are warranted and are underway to address these limitations.