Pediatric Cardiology

, Volume 32, Issue 8, pp 1147–1157

The Contribution of Chromosomal Abnormalities to Congenital Heart Defects: A Population-Based Study

Authors

  • Robert J. Hartman
    • National Center on Birth Defects and Developmental DisabilitiesCenters for Disease Control and Prevention
    • Oak Ridge Institute for Science and Education
    • National Center on Birth Defects and Developmental DisabilitiesCenters for Disease Control and Prevention
  • Lorenzo D. Botto
    • Department of PediatricsUniversity of Utah School of Medicine
  • Tiffany Riehle-Colarusso
    • National Center on Birth Defects and Developmental DisabilitiesCenters for Disease Control and Prevention
  • Christa L. Martin
    • Department of Human GeneticsEmory University
  • Janet D. Cragan
    • National Center on Birth Defects and Developmental DisabilitiesCenters for Disease Control and Prevention
  • Mikyong Shin
    • National Center on Birth Defects and Developmental DisabilitiesCenters for Disease Control and Prevention
    • RTI International
  • Adolfo Correa
    • National Center on Birth Defects and Developmental DisabilitiesCenters for Disease Control and Prevention
Original Article

DOI: 10.1007/s00246-011-0034-5

Cite this article as:
Hartman, R.J., Rasmussen, S.A., Botto, L.D. et al. Pediatr Cardiol (2011) 32: 1147. doi:10.1007/s00246-011-0034-5
  • 634 Views

Abstract

We aimed to assess the frequency of chromosomal abnormalities among infants with congenital heart defects (CHDs) in an analysis of population-based surveillance data. We reviewed data from the Metropolitan Atlanta Congenital Defects Program, a population-based birth-defects surveillance system, to assess the frequency of chromosomal abnormalities among live-born infants and fetal deaths with CHDs delivered from January 1, 1994, to December 31, 2005. Among 4430 infants with CHDs, 547 (12.3%) had a chromosomal abnormality. CHDs most likely to be associated with a chromosomal abnormality were interrupted aortic arch (type B and not otherwise specified; 69.2%), atrioventricular septal defect (67.2%), and double-outlet right ventricle (33.3%). The most common chromosomal abnormalities observed were trisomy 21 (52.8%), trisomy 18 (12.8%), 22q11.2 deletion (12.2%), and trisomy 13 (5.7%). In conclusion, in our study, approximately 1 in 8 infants with a CHD had a chromosomal abnormality. Clinicians should have a low threshold at which to obtain testing for chromosomal abnormalities in infants with CHDs, especially those with certain types of CHDs. Use of new technologies that have become recently available (e.g., chromosomal microarray) may increase the identified contribution of chromosomal abnormalities even further.

Keywords

Chromosomal abnormalityCongenital heart defectCongenital heart diseasePrevalenceEpidemiology

Introduction

Congenital heart defects (CHDs) are the most common type of birth defect, with an estimated birth prevalence of 1/125 live births [26]. Birth defects are one of the most common causes of infant mortality, resulting in approximately 20% of deaths during the first year of life [20]; among birth defects, CHDs are the most frequent contributor to neonatal mortality [37]. In addition, costs associated with the care of a child with a CHD are significant [5, 36].

The association between CHDs and many types of chromosomal abnormalities is well recognized [23, 28]. However, estimates of the proportion of CHDs associated with a chromosomal abnormality vary widely from 9% to 18% [3, 911, 13, 14, 16, 24, 29, 31] depending on the study location, inclusion criteria for infants with CHDs, and whether the analysis was clinic- or population-based (Table 1). Most estimates of the contribution of chromosomal abnormalities to CHD etiology have only included aneuploidies; information on the contribution of other chromosomal abnormalities, such as 22q11.2 deletion, has not typically been included. Advances in cardiac imaging and in cytogenetic technologies provide an opportunity to determine whether the contribution has changed from previous estimates. We present findings from a population-based study of the contribution of chromosomal abnormalities to CHDs in a recent birth cohort of infants born from 1994 through 2005 to mothers residing in metropolitan Atlanta, GA.
Table 1

Epidemiological studies of the abnormal chromosomal contribution to CHDs during the last 20 years

Reference

Study site (years in study)

Study design

22q11.2 deletion included in analysis (yes/no/not stated)

Source of data on CHD

Excluded infants

Percentage of CHD attributed to chromosomal abnormalities (%)

Stoll et al. 1989 [31]

Departement du Bas-Rhin (France; 1979–1986)

Hospital-based, consecutive-births case-control study

No

Echocardiography, cardiac catheterization, surgery, or autopsy

Infants <26 weeks gestation

11.3

Ferencz et al. 1989 [11]

Baltimore-Washington Infant Study (1981–1986)

Regional population-based cardiovascular specific birth defects registry

No

Echocardiography, cardiac catheterization, surgical inspection, or autopsy

Pregnancy terminations, fetal deaths, diagnosis occurred after first year of life

12.9

Pradat 1992 [24]

The Registry of Congenital Malformations and The Child Cardiology Registry (Sweden; 1981–1986)

Population-based birth defects registry

No

Echocardiography, cardiac catheterization, surgery, or necropsy

Infants <28 weeks gestation

13

Kidd et al. 1993 [16]

New South Wales and the Australian Capital Territory (Australia; 1981–1984)

Clinic-based

No

Pediatric cardiologists, echocardiography, cardiac catheterization, surgery, or autopsy

Pregnancy terminations, fetal deaths, diagnosis occurred after first year of life

9.5

Ferencz et al. 1997 [10]

Baltimore-Washington Infant Study (1981–1989)

Regional population–based cardiovascular-specific birth-defects registry

No

Echocardiography, cardiac catheterization, surgical inspection, or autopsy

Pregnancy terminations, fetal deaths, diagnosis occurred after first year of life

12.1

Grech & Gatt 1999 [13]

The country of Malta (1990–1994)

Hospital-based

Not stated

Echocardiogram reports, cardiac catheterization, surgery, and postmortem reports

Pregnancy termination, fetal deaths, diagnosis occurred after first year of life

9

Harris et al. 2003 [14]

Combination of the Central-Eastern France Registry (France; 1983–1992), Swedish national population-based system for registration of congenital malformations (Sweden; 1981–1992), and CBDMP (1985–1992)

France and CBDMP: regional population- based registries; Sweden: national population-based birth-defects registry

Not stated

International Society of Cardiology Codes; only CBDMP mentions that diagnoses confirmed by echocardiography, cardiac catheterization, surgery, and/or autopsy

France and Sweden: <28 weeks gestation, CBDMP: <20 weeks gestation; heart conditions excluded: cardiomegaly, cardiomyopathy, fibroelastosis, rate or rhythm anomalies, cardiac valve insufficiency, and patent ductus

Total: 18

Individual registry:

France:12.6

Sweden: 12.8

CBDMP: 20.3

Bosi et al. 2003 [3]

Emilia-Romagana birth-defects registry [1980–2000]

Regional population–based birth-defects registry

Yes

Clinical evaluation, echocardiography, cardiac catheterization, surgery, autopsy

<28 weeks gestation, stillbirths, spontaneous and/or induced abortions, diagnosis occurred after the second year of life

9.10

Schellberg et al. 2004 [29]

Germany (not stated)

Clinic-based

Yes

Clinical evaluation, electrocardiography, chest radiography, echocardiography, cardiac catheterization, MRI, or CT

Pregnancy terminations, fetal deaths

16

Dadvand et al. 2009 [9]

Northern England (1985–2003)

Population-based registry of congenital anomalies

Not stated

Echocardiography, cardiac catheterization, surgery, or autopsy

EUROCAT exclusion list, <20 weeks gestation

11.6

EUROCAT European Registers of Congenital Anomalies and Twins, MRI magnetic resonance imaging, CT computed tomography, CBDMP California Birth Defects Monitoring Program

Methods

Data Source

Information on CHDs and chromosomal abnormalities was obtained from the Metropolitan Atlanta Congenital Defects Program (MACDP), an ongoing population-based birth-defects surveillance system in operation since 1967. For this analysis, we reviewed data on infants delivered from January 1, 1994 through December 31, 2005. Detailed methods of MACDP are described elsewhere [7]. Briefly, MACDP ascertains deliveries affected by major birth defects to mothers whose residence is in the five central counties of the metropolitan Atlanta area at the time of delivery. Trained abstractors actively ascertain information on major structural and genetic birth defects identified among live-born infants before the sixth birthday, among fetal deaths occurring at ≥20 weeks’ gestation, and among pregnancy terminations. Because of concern for the reliability of CHD diagnoses made prenatally, for this analysis, pregnancy terminations were excluded, and live-born infants and fetal deaths with CHDs that were confirmed postnatally were included. MACDP data sources include medical records from birth and pediatric hospitals, cytogenetic laboratories (one academic and one commercial laboratory), and a major referral center for pediatric cardiology. Abstracted information is reviewed by MACDP staff, which includes pediatricians and clinical geneticists. Up to 24 individual defects can be coded per infant using a modified version of the International Classification of Disease, Ninth Revision, Clinical Modification and British Paediatric Association (modified ICD9/BPA) coding classification systems.

CHDs: Classification, Inclusion, and Exclusion Criteria

Clinical information on infants with CHDs in MACDP (ICD9/BPA codes 745.000 to 747.499, 747.640, 759.040, and 759.300 to 759.399) was reviewed by physicians with expertise in pediatric cardiology to classify the structural heart defects using a modified nomenclature from the Society of Thoracic Surgeons [27]. Individual CHDs and aggregates based on presumed morphogenetically similar developmental mechanisms have been described elsewhere [27]. Patent ductus arteriosus (PDA) is abstracted only if another major birth defect is present and was not classified as a CHD if the infant was <36 weeks’ gestational age; if a term infant had a PDA identified and closed before 6 weeks of life; or if the PDA was necessary to sustain life due to the presence of other CHDs. Bicuspid aortic valve was only counted if it occurred as an isolated CHD. This classification system excludes patent foramen ovale and other isolated minor conditions, such as valve insufficiency. An infant could be classified as having ≤4 CHDs; however, infants with a single ventricle (e.g., tricuspid atresia or double-inlet ventricle) or heterotaxy were classified as having only one major CHD, regardless of the presence of other CHDs. For this analysis, cases of isolated cardiomyopathy and nonstructural heart defects (e.g., isolated neonatal arrhythmias or persistent fetal circulation) were excluded.

Ascertainment of Cytogenetic Abnormalities

After identifying infants with CHDs, we selected those who had been identified as having a chromosomal abnormality by using the modified ICD9/BPA codes for chromosomal abnormalities (758.000 to 758.999) (including any cytogenetic results other than 46,XX, 46,XY, or variants believed to be benign). During the time period included in this analysis, fluorescence in situ hybridization (FISH) testing for 22q11.2 deletion was widely available. Of note, the decision to perform cytogenetic testing was at the discretion of the health care providers caring for the infants; thus, not all infants with CHDs were tested. Infants with chromosomal abnormality codes that are used only for specific cytogenetically confirmed diagnoses were included as having a chromosomal abnormality without further review. These included infants with a code for trisomy 21 (758.000, 758.010, 758.020, or 758.030), mosaic trisomy 21 (758.040), trisomy 13 (758.100, 758.110, 758.120, or 758.130), trisomy 18 (758.200, 758.210, 758.220, or 758.230), 22q11.2 deletion (758.370), 45,X (758.600), or Klinefelter syndrome (758.700). In addition, abstracted results of chromosome analyses or FISH studies on infants who had nonspecific codes for any other chromosomal abnormality were reviewed to ensure that the child had a documented chromosomal abnormality and to allow classification into an appropriate category. Altogether, infants were classified into 22 mutually exclusive categories of chromosomal abnormalities (see Table 3). Results were reviewed by a cytogeneticist (C.L.M.) to ensure appropriate classification. If an apparently balanced translocation was present, the case was considered to have a chromosomal abnormality only if the translocation was de novo or if parental results were not available. When a chromosomal abnormality was suspected but not documented by cytogenetic testing, the infant was assumed to not have a chromosomal abnormality.

Data Analysis

To calculate the birth prevalence of CHDs, we used as the denominator the number of live infants born to residents of the MACDP surveillance area for the years 1994 to 2005 as obtained from vital records of the state of Georgia. We calculated the proportion of infants with CHDs who had chromosomal abnormalities for all CHDs and by CHD type. The infant was only counted once for all types of CHDs if there were multiple types of CHDs present. Risk ratios (RRs) and 95% confidence intervals (CI) were used to compare the proportion of infants with CHDs who have chromosomal abnormalities by infant sex, gestational age (for live-born singletons only: preterm [<37 completed weeks], term [≥37 completed weeks]), maternal race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other), maternal age at birth (<35 years, ≥35 years), one or more than one CHD type, and birth status (live-born infants, fetal deaths). RRs and 95% CI were generated using SABER [6]. Results were considered statistically significant if P was ≤0.05 and 95% CI did not include 1.0.

Results

We identified 4430 infants with CHDs, for a birth prevalence of 7.9/1000 live births. Among infants with CHDs, 547 (12.3%) had a chromosomal abnormality, and most of these infants (77.1%) had only 1 type of CHD. CHDs most likely to be associated with a chromosomal abnormality were interrupted aortic arch (IAA), type B and not otherwise specified (69.2%); atrioventricular septal defect (AVSD) (67.2%); double-outlet right ventricle (DORV) (33.3%); partial anomalous pulmonary venous return (33.3%); and truncus arteriosus (32.3%). CHDs that were least likely to have had a chromosomal abnormality diagnosed were heterotaxy (2.2%), Ebstein anomaly (2.6%), and pulmonary valve stenosis (3.3%) (Table 2). The most commonly observed chromosomal abnormalities among infants with CHDs were trisomy 21 (52.8%), trisomy 18 (12.8%), 22q11.2 deletion (12.2%), and trisomy 13 (5.7%) (Table 3). Deletion 22q11.2 was the largest single type of chromosomal abnormality among many infants with conotruncal CHDs, but a variety of other chromosomal abnormalities were observed in these infants as well. Trisomy 21 was observed most often in infants with an AVSD or a secundum atrial septal defect (ASD). Most of the infants with ventricular septal defect, not otherwise specified (VSD NOS), with a chromosomal abnormality had trisomy 18, 21, or 13 (Table 4).
Table 2

Proportion of infants (live births and fetal deaths) with CHDs who have chromosomal abnormalities by CHD classification (MACDP 1994–2005)

CHD classification

All infants with CHD

Infants with single CHD

Any CHDa

Chromosomal abnormality and CHDa

Any CHD

Chromosomal abnormality and CHD

n

n

%

n

n

%

All types of CHD

4430

547

12.3

3903

422

10.8

 Conotruncal

  Interrupted aortic arch, type B or NOS

26

18

69.2

17

12

70.6

  DORV

27

9

33.3

21

7

33.3

  Truncus arteriosus

31

10

32.3

21

6

28.6

  Tetralogy of Fallot

263

52

19.8

242

48

19.8

  Vascular rings

70

9

12.9

51

1

2.0

  VSD, Conotruncal

22

2

9.1

15

0

N/A

  d-TGA

132

6

4.5

113

5

4.4

 AVSD

220

148

67.2

153

108

70.6

 Abnormal cell growth

  Partial anomalous pulmonary venous return

6

2

33.3

1

0

N/A

  Total anomalous pulmonary venous return

41

3

7.3

34

2

5.9

  ASD, sinus venosus

19

1

5.3

15

0

N/A

 VSD

  VSD, NOS

181

37

20.4

159

29

18.2

  VSD, perimembranous

593

92

15.5

407

51

12.5

  VSD, muscular

1,356

59

4.4

1,212

40

3.3

 ASD

  ASD secundum

517

102

19.7

274

38

13.9

  ASD OS/NOS

127

21

16.5

89

11

12.4

 Left-sided obstructive defects

  HLHS

123

12

9.8

116

12

10.3

  Mitral valve stenosis

11

2

18.2

10

2

20.0

  Coarctation of the aorta

249

30

12.0

120

10

8.3

  Aortic stenosis

59

5

8.5

46

2

4.3

  Bicuspid aortic valve

45

3

6.7

45

3

6.7

 PDA

170

28

16.5

121

11

9.1

 Right-sided obstructive defects

  Tricuspid valve stenosis

13

1

7.7

12

1

8.3

  Pulmonary atresia

31

3

9.7

31

3

9.7

  Pulmonary stenosis, other

23

1

4.3

23

1

4.3

  Pulmonary stenosis, valvar

304

10

3.3

215

3

1.4

 Single ventricle/complex CHDb

56

5

8.9

56

5

8.9

 Ebstein anomaly

38

1

2.6

37

1

2.7

 Heterotaxy

91

2

2.2

91

2

2.2

 Other CHDs

  Coronary artery anomaly

19

2

10.5

18

2

11.1

  Other vascular anomalies

58

5

8.6

57

5

8.8

OS otherwise specified, N/A not applicable, TGA transposition of the great arteries, HLHS hypoplastic left heart syndrome

aCHD type is not mutually exclusive, but infants were only counted once for all types of CHD

bIncludes the CHD types double-inlet single ventricle, single ventricle OS/NOS, unbalanced AVSD, and mitral or tricuspid atresia

Table 3

Frequencies of chromosomal abnormalities among infants with CHDs (MACDP 1994–2005)

Chromosomal abnormality category

n = 547

%

Numerical

428

78.2

 Autosomal

409

74.8

  Trisomy 21

289

52.8

  Trisomy 21, other

9

1.6

   Mosaic trisomy 21

4

0.7

   46,XX,der(21;21)(q10;q10)

1

0.2

   45,XX,−21[11]/46,XX,+21,der(21;21)(q10;q10)[8]

1

0.2

   47,XY,t(4;13)(p15.2;q21.2),+21

1

0.2

   47,XY,+21[36]/47,XY,+i(8)(p10)[14]

1

0.2

   “Mosaic Down syndrome/trisomy X”a

1

0.2

  Trisomy 18

70

12.8

  Trisomy 18, other

3

0.5

   48,XXY,+18

1

0.2

   48,XXX,+18

1

0.2

   47,XY,+18[24]/46,XY[6]

1

0.2

  Trisomy 13

31

5.7

  Trisomy 22

3

0.5

  Trisomy 22, other

3

0.5

   47,XX,+22[2]/46,XX[48]

1

0.2

   47,XY+22[3]/46,XY[13]

1

0.2

   47,XX,+22[5]/46,XX[15]

1

0.2

  Triploidy

1

0.2

   69,XXX

1

0.2

  Other

1

0.2

   45,XY,−21[4]/46,XY[46]

1

0.2

 Sex chromosome

19

3.5

  Turner syndrome

6

1.1

  Turner syndrome, other

7

1.3

   45,X,t(8;13)(q21.1;q22)[16]/46,XX,t(8;13)(q21.1;q22)[14]

1

0.2

   46,X,i(X)(q10)

1

0.2

   46,X,idic(X)(q22.1)

1

0.2

   45,X[26]/46,XY[4]

1

0.2

   45,X[19]/46,X,idic(X)(qter → p11.2)[11]

1

0.2

   45,X[35]/46,X,idic(Y)(p11.32)[15]

1

0.2

   45,X[23]/47,X,+?dic r(Y)[10]

1

0.2

  Klinefelter syndrome

2

0.4

  Klinefelter syndrome, other

2

0.4

   48, XXYY

1

0.2

   47,XXY[16]/46,XY[14]

1

0.2

  47, XXX

1

0.2

  47, XYY

1

0.2

Structural

115

21.0

 Unbalanced

110

20.1

  Deletion

84

15.4

   22q11.2

67

12.2

   46,XX,del(11)(q24.2)

2

0.4

   46,XX,der(1)(pter → 36.3305::p36.3205 → p36.2300::p36.3100 → qter)

1

0.2

   46,XY,del(1)(q42.13q42.3)

1

0.2

   46,XX,del(2)(p25.1)

1

0.2

   46,XX,del(4)(p15.32)

1

0.2

   46,XY,del(4)(q28.2q31.3)

1

0.2

   46, XY,del(5)(q11.2q13.1)

1

0.2

   “Deletion of the short arm of the sixth chromosome”a

1

0.2

   46,XX,del(6)(p25)

1

0.2

   46,XX,add(7)(p21) or del(7)(p21p22)

1

0.2

   46,XX,del(9)(p22.1)

1

0.2

   46,XX,del(9)(p23)

1

0.2

   46,XY,del(10)(p15.1)

1

0.2

   46,XY,del(17)(p11.2p11.2)

1

0.2

   46,XX,del(18)(q21.31)

1

0.2

   46,XY,del(18)(q23)

1

0.2

  Duplication

9

1.6

   46,XY,dup(3)(q26.2q27.1)mat 

1

0.2

   46,XY,add(4)(p14)

1

0.2

   46,XY,add(5)(p15.1)

1

0.2

   46,XX,dup(6)(pter → q25.3::q25.1 → qter)

1

0.2

   46,XX,add(9)(p24.1)

1

0.2

   47,XX,+der(11)(q11).ish der(11)(wcp11+,D11Z1+)

1

0.2

   46,XX,dup(17)(q23.3q24.2)

1

0.2

   46,XX,dup(20)(q11.2q13.1) OR 46,XX,ins(20;?)(q13.1;?)

1

0.2

   46,XY,add(21)(q22.3)

1

0.2

  Unbalanced translocation

9

1.6

   46,XX,der(4)t(4;?)(q31.1;?)

1

0.2

   47,XY,der(6)t(6;17)(p25.1;q25.2)mat,+mar.ish der(7) 

1

0.2

   46,XY,add(12)(p13.3).ish der(12)t(4;12)(wcp4+)

1

0.2

   47,XY,+der(15)t(15;16)(q14;q13)mat

1

0.2

   46,XY,der(15)t(15;17)(q26.3;p13.1).ish der(15)t(15;17) (wcp15+,D15Z+,FES+,15qsubtel+,D1752199+,LIS1+)

1

0.2

   46,XY,der(21)t(3;21)(q21;q22.3)

1

0.2

   46,XX,der(22)t(1;22)(q32.1;q13.3)

1

0.2

   46,X,der(X)t(X;12)(p11.23;q22)mat

1

0.2

   46,X,der(Y)t(X;Y)(p11.4;p11.2)

1

0.2

  Recombinant chromosome from inversion

1

0.2

   46,XX,rec(3)dup(3q)inv(3)(p25q25.1)mat

1

0.2

  Other

9

1.6

   46,XX,i(8)(q10)

1

0.2

   47,XY,+del(8)p21/46,XY

1

0.2

   46,XX,der(17)(pter→p13.1::p1205→p11.2::p1205→qter)

1

0.2

   46,XX,der(18)(pter→q21.32::q21.32→q11.2::qter)

1

0.2

   46,XY,idic(20)(p11.1)[14]/46,XY[16]

1

0.2

   46,XX,r(21)(p13q22.2)

1

0.2

   46,X,dir dup(X)(q13q24)

1

0.2

   46,X,del(Y)(q11.2)

1

0.2

   47,XX, +mar de novo.ish der(14/22) (D14Z1+/D22Z1+,wcp14−, wcp22−,rDNA++)

1

0.2

 Apparently balanced

6

1.1

  Translocation

3

0.5

   46,XY,t(1;6)(q44;q23.3)

1

0.2

   45,XX,der(13;22)(q10;q10)

1

0.2

   45,XX,der(13;14)(q10;q10)[7]/46,XX[13]

1

0.2

  Inversion

3

0.5

   46,XY,2qs pat,inv(5)(q13.3q23.2)mat

1

0.2

   46,XY,inv(7)(p22q11.23)

1

0.2

   46,XX,inv(11)(q21q23.3)mat

1

0.2

aChromosomal abnormalities listed in quotes were shown when the cytogenetic analysis report was not available in the abstracted records

Table 4

Chromosomal abnormalities observed with specific CHDs listed by CHD type (MACDP 1994–2005)

 

n

%

Conotruncal

 IAA (type B or NOS)

18

 

  22q11.2 deletion

17

94.4

  Unbalanced deletiona

1

5.6

 DORV

9

 

  Trisomy 18

6

66.7

  Trisomy 13

2

22.2

  22q11.2 deletion

1

11.1

 Truncus arteriosus

10

 

  22q11.2 deletion

9

90.0

  Trisomy 13

1

10.0

 Tetralogy of Fallot

52

 

  22q11.2 deletion

23

44.2

  Trisomy 21

9

17.3

  Trisomy 18

7

13.5

  Trisomy 13

5

9.6

  Unbalanced deletiona

2

3.8

  Trisomy 21, other

1

1.9

  Trisomy 22, other

1

1.9

  Unbalanced, other

1

1.9

  Unbalanced Translocation

1

1.9

  47, XXX

1

1.9

  Balanced translocation

1

1.9

AVSD

148

 

 Trisomy 21

131

88.5

 Trisomy 18

11

7.4

 Trisomy 13

2

1.4

 Trisomy 21, other

2

1.4

 Unbalanced, other

1

0.7

 Triploidy

1

0.7

VSD, NOS

37

 

 Trisomy 18

17

45.9

 Trisomy 21

8

21.6

 Trisomy 13

4

10.8

 Trisomy 22

2

5.4

 Turner syndrome

2

5.4

 Trisomy 18, other

1

2.7

 22q11.2 deletion

1

2.7

 Unbalanced duplication

1

2.7

 Unbalanced deletiona

1

2.7

ASD, Secundum

102

 

 Trisomy 21

80

78.4

 Trisomy 13

7

6.9

 22q11.2 deletion

7

6.9

 Trisomy 18

3

2.9

 Unbalanced duplication

2

2.0

 Unbalanced translocation

2

2.0

 Trisomy 18, other

1

1.0

aInfants with unbalanced deletions other than 22q11.2 deletion

The proportion of infants with CHDs who had a chromosomal abnormality did not differ by infant sex or gestational age (Table 5). Infants with more than one CHD diagnosis were more likely to have a chromosomal abnormality (RR = 2.2, 95% CI 1.8–2.6) than those with only one CHD diagnosis. Infants with CHDs born to non-Hispanic black mothers were slightly more likely to have a chromosomal abnormality diagnosed compared with those born to non-Hispanic white mothers (RR = 1.3, 95% CI 1.1–1.5). Fetal deaths with CHDs were nearly three times as likely to have a chromosomal abnormality diagnosed (33.9%) as live-born infants (12.1%) (RR = 2.8, 95% CI 2.0–4.1). Infants with CHDs born to mothers who were ≥35 years old were more likely to have a chromosomal abnormality than infants with CHDs born to mothers <35 years of age (RR = 2.1, 95% CI 1.8–2.5) (Table 5). However, when infants with CHDs and trisomy 21, 18, or 13 were excluded from the maternal age analysis, there was no longer a statistically significant association (data not shown).
Table 5

Clinical and demographic characteristics of infants with CHDs and chromosomal abnormalities (MACDP 1994–2005)

 

All CHDs

Infants with CHDs and chromosomal abnormalities

RR (95% CI)

 

n

n

%

 

Infant sex

 Male

2170

241

11.1

Ref

 Female

2255

305

13.5

1.2 (1.0–1.4)

Gestational agea (week)

 ≥37

3,137

371

11.8

Ref

 <37

895

128

14.3

1.2 (1.0–1.5)

Birth outcome

 Live born

4371

527

12.1

Ref

 Fetal death

59

20

33.9

2.8 (2.0–4.1)

Frequency of CHD types

 Only 1 CHD

3903

422

10.8

Ref

 >1 CHD

527

125

23.7

2.2 (1.8–2.6)

Maternal race/ethnicity

 Non-Hispanic white

1960

215

11.0

Ref

 Non-Hispanic black

1,564

222

14.2

1.3 (1.1–1.5)

 Hispanic

664

81

12.2

1.1 (0.9–1.4)

 Other

188

22

11.7

1.1 (0.7–1.6)

Maternal age (y)

 <35

3482

347

10.0

Ref

 ≥35

939

198

21.1

2.1 (1.8–2.5)

aLive-born singletons only

Discussion

This study provides an updated estimate of the contribution of chromosomal abnormalities to the etiology of CHDs overall and by type of CHD, and suggests that a chromosomal abnormality detected by G-banding analysis or targeted FISH is responsible for CHDs in approximately one in eight live born infants and fetal deaths. Despite the addition of targeted FISH analyses, the contribution is similar to previous estimates. The lack of change could be related to the increasing prevalence in recent years of mild heart defects, such as muscular VSDs, which are less likely to be associated with chromosomal abnormalities. Trisomies 21, 18, and 13 and 22q11.2 deletion comprise the majority of chromosomal abnormalities seen in infants with CHDs in our study. It is important to note that laboratory testing for chromosomal abnormalities continues to improve. Array comparative genomic hybridization (chromosomal microarray) testing has recently become clinically available and can be used to identify chromosomal imbalances not detectable by previous technologies [1, 30]. Our study uses data from a time period when chromosomal microarray testing was not routinely available; thus, these results underestimate the contribution of chromosomal abnormalities to the etiology of CHDs [18]. However, this study will be useful as a comparison to future studies that use microarray testing to better understand the proportion of chromosomal imbalances among infants with CHDs [21].

Our estimated frequency of chromosomal abnormalities among infants with CHDs was 12.3%, similar to that reported in most previous studies estimating the contribution between 11.3% and 13% [911, 24, 31]. Several studies showed a lower frequency of chromosomal abnormalities among infants with CHDs (9% to 9.5%), but these studies included live-born infants only [3, 13, 16]. A greater frequency of chromosomal abnormalities was found in a study by Harris et al. that combined data from three population-based birth-defects registries. These investigators attributed their greater proportion (18%) to different inclusion criteria in California’s birth-defects monitoring system (20.3% of infants with CHDs from California had chromosomal abnormalities compared with 12.6% in infants from France and 12.8% in infants from Sweden) [14]. Schellberg et al. estimated that 16% of infants with CHDs had chromosomal abnormalities; however, their study was clinic-based and used a stepwise approach to obtaining different cytogenetic tests in living infants that presented to their clinic [29].

Similar to our findings, the association between CHDs and trisomies 21, 18, and 13 has long been recognized [3, 911, 13, 14, 16, 24, 29, 31]. We found that many infants with conotruncal CHDs are also frequently observed to have 22q11.2 deletions, a phenomenon that has previously been reported [4, 15, 17, 29]. Our study extends our previous study of 22q11.2 deletions using data from MACDP [4] and provides data on the frequency of 22q11.2 deletion in relation to other common chromosomal abnormalities among infants with CHDs. Among all of the infants with CHDs in our study, 22q11.2 deletions were detected in a similar proportion of CHD infants as trisomy 18 and were observed more frequently than trisomy 13, which has not previously been reported. Our analysis showed more than two thirds of infants with AVSDs had a chromosomal abnormality, and most had trisomy 21. This supports the well-known association between AVSDs and chromosomal abnormalities, in particular, trisomy 21 [10, 12, 19, 34]. chromosomal abnormalities were also frequently seen in infants with DORV. In our study, one-third of infants with DORV had a chromosomal abnormality. Obler et al. showed a greater estimate of 40%; however, their study included infants that did not have confirmed chromosomal abnormalities [22]. The large number of infants with a VSD, NOS or ASD who had a chromosomal abnormality is likely due to the finding of additional CHD types in infants with these defects (e.g., AVSD and ASD).

Our results support the hypothesis that genetic loci on many different chromosomes are involved in the causation of CHDs [23]. Van Karnebeek and Hennekam performed a search in the Human Cytogenetics DataBase looking for unbalanced structural chromosomal abnormalities that had been listed in at least three individuals with CHDs [35]: they identified unbalanced structural abnormalities involving all chromosomes except 12, 14, 19, 21, and X. In our study, there were three infants with CHDs and trisomy 22; most infants with trisomy 22 are known to have CHDs [33]. The association between Turner syndrome and CHDs has been well established [2], and in our analysis, there were eight infants with Turner syndrome and CHD. We also observed infants with CHDs in our study with other sex chromosomal abnormalities, such as Klinefelter syndrome (XXY), XXX, and XYY; these chromosomal abnormalities were also observed in infants with conotruncal CHDs in a recent study [17].

Our study has many strengths. Infants were classified using a clinically standard nomenclature and morphogenetically-based aggregation system for CHDs. This system differentiated subtle but important types of CHDs and minimized misclassification [32]. Data on chromosomal abnormalities were reviewed by a cytogeneticist to ensure that infants were classified into appropriate categories. Such systematic classifications may allow for comparisons with other population-based studies using similar systematic classifications. We used a population-based study design to ensure that all infants born with CHDs, not just those seen at a particular referral center, were included in the analysis. In addition, because our study used data from a birth-defects surveillance system, we were able to include fetal deaths with CHDs in our analysis to give a more complete picture of the contribution of chromosomal abnormalities to CHDs. The population-based design and inclusion of fetal deaths help minimize potential selection bias and allow for more reliable estimation of the contribution of chromosomal abnormalities to CHDs. For example, approximately 5% of infants with tetralogy of Fallot in our analysis had trisomy 18 or trisomy 13, associations that might not be observed in clinic- or hospital-based analyses of tetralogy of Fallot [25].

However, our study also has several limitations. Although most infants with CHDs exhibit symptoms and have them detected early in life, for some defects with limited clinical manifestations in infancy and childhood (e.g., isolated coronary artery anomalies, bicuspid aortic valve, mild coarctation), ascertainment is likely to be incomplete; thus, our estimate of the contribution of chromosomal abnormalities applies to CHDs diagnosed in infancy or early childhood. Decisions about cytogenetic testing were made by clinicians caring for the infant. Based on our data, chromosomal analyses were performed in 38.9% of the infants with CHD. Therefore, our results are likely to underestimate the true contribution of chromosomal abnormalities to CHDs. In addition, only infants on whom cytogenetic testing had been performed before the child’s sixth birthday and documented as abnormal were considered to have a chromosomal abnormality. If cytogenetic studies were never performed or were performed at an age >6 years, then the chromosomal abnormality would have been missed. MACDP does not ascertain all cytogenetic testing results from all sources, so there is a possibility that some infants with CHDs could have an undocumented chromosomal abnormality. Because of concern for the reliability of CHD diagnoses made prenatally, our study did not include pregnancy terminations, many of which might have had chromosomal abnormalities because severe birth defects are found more often among fetal deaths and terminations [8]. Finally, we assumed that chromosomal abnormalities were responsible for the CHDs with which they were observed; however, it is possible that some chromosomal abnormalities (e.g., apparently balanced rearrangements) were not causative but rather were chance observations.

Based on our analysis, approximately one in eight live births and fetal deaths with CHDs have a chromosomal abnormality, and a much greater frequency was observed with certain types of CHDs. The 22q11.2 deletion was observed nearly as often among infants with CHDs as those with trisomy 18 and more frequently than those with trisomy 13. Clinicians should maintain a low threshold at which to obtain testing for chromosomal abnormalities in infants with CHDs, especially in infants with specific types of CHDs or multiple CHDs. A consensus group has recently recommended chromosomal microarray as the first-tier diagnostic test for infants with developmental disabilities, autism spectrum disorders, or multiple congenital anomalies [21]. Inclusion of chromosomal microarray analysis in future studies of CHDs will allow development of evidence-based guidelines about its utility in the evaluation of isolated heart defects.

Acknowledgments

We thank Cheryl Broussard, Suzanne Gilboa, Assia Miller, and Sarah Tinker for their assistance with the statistical analyses. The authors acknowledge the dedication and contributions of the abstractors, staff, and scientists who contribute to the MACDP. This research was supported in part by an appointment to the Research Participation Program at the CDC administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the United States Department of Energy and the CDC.

Disclaimer

The findings and conclusions in this article are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention (CDC).

Copyright information

© Springer Science+Business Media, LLC (outside the USA)  2011