Congenital heart disease (CHD) is the most common human birth defect worldwide1, striking a tremendous toll on affected families, caregivers, and healthcare systems. Approximately 40,000 children are born each year in the United States with a heart malformation, and at least another 40,000 are born annually with sub-clinical malformations that result in heart disease later in adulthood. Significant advances in cardiac care and surgery have lowered mortality, and there are now over 1 million survivors of CHD in the U.S. As a result, the economic effects of CHD are substantial, particularly when considering lifetime costs of management.
As an increasing proportion of the CHD population reaches reproductive age, questions of the genetic contribution to disease and risk of transmission have become paramount. Such individuals also often suffer age-dependent complications in heart function that may be related to the initial developmental and/or genetic insult that resulted in the CHD. While their causes remain generally unknown, most CHDs are thought to have a multifactorial etiology with an interplay of genetic and environmental effects2. However, the relative contributions of genes and the environment have been difficult to discern. In this edition of Circulation, Oyen et al. utilize a uniquely sized and annotated population to estimate recurrence risk for specific CHDs in families and, thereby, indirectly assess the role of genetic inheritance in CHD.
A number of studies have attempted to quantify the risks conferred by a family history of CHD, demographic qualities, or environmental exposures3, 4. Gestational insults, such as rubella infection and gestational diabetes, can predispose to CHD, as can exposure to ethanol and other teratogens, such as retinoic acid3, 5. While the incidence of CHD is higher in the setting of these exposures, most fetuses remain unaffected, suggesting that only a subpopulation may be at risk. In contrast, several syndromic and familial cases of CHD are caused by rare single-gene mutations that have major effects, sometimes with 100% penetrance6–8. Thus, the more common forms of CHD that appear to be sporadic may, in fact, be caused by inherited genetic variants that modestly affect protein expression or function and only manifest as disease when combined with additional genetic, epigenetic, environmental or hemodynamic insults. However, experimental evidence for this theory remains elusive.
Oyen et al. address the epidemiologic aspect of this theory by examining familial aggregation of CHD based on an unusually large and well-defined Danish population that has been annotated in multiple registries. The authors used a population-based design that uniquely captured all residents of Denmark—over 1.7 million—over a 28-year period (Reference for Oyen this issue). They identified approximately 18,000 individuals with CHD and capitalized on the Danish Family Relations Database to link affected individuals with first-, second-, and third-degree relatives. Disease information for relatives was also available in the database and allowed phenotype-based development of pedigrees. Using this population, the authors estimated the contribution of a family history of CHD to an individual's risk of CHD, and they estimated the population risk conferred by such family histories.
They found the relative risk of recurrence for all types of CHD to be approximately 3 when a first-degree relative had CHD. This relative risk diminished when the family history of CHD was in only second- and third-degree relatives. These findings are consistent with the commonly used empiric risks provided to families faced with a potential recurrence of CHD. Even after the authors accounted for cases with known chromosome abnormalities or other congenital anomalies, the increased relative risks of recurrence persisted. Therefore, their findings might be applicable to the most common scenario in which CHD is an isolated finding. Furthermore, the authors analyzed the relative risk of recurrence of disease in unlike-sex twin pairings (presumably dizyogtic twins), and this estimate was similar to the relative risk found in those individuals with affected first-degree relatives. Interestingly, same-sex twins, which likely include some monozygotic twins, demonstrated a ~threefold higher relative risk of recurrence than unlike-sex twins. These data strongly suggest a genetic component to “sporadic” congenital heart disease (Fig. 1).
Interestingly, when the authors analyzed recurrence of the same type of CHD within families, the relative risk of recurrence was significantly higher for certain malformations. For example, heterotaxy, atrioventricular septal defect, and left and right ventricular outflow tract obstructive lesions had particularly higher relative risks of recurrence, with heterotaxy having a relative risk of ~80. Although the numbers available for analysis were inevitably smaller than when CHD was analyzed as one collective group, the large initial size of this population still provides compelling evidence for a strong genetic effect on many forms of CHD. These findings are similar to the high incidence of aortic valve disease found in first-degree relatives of individuals with aortic valve atresia associated with hypoplastic left-heart syndrome9, 10. It is intriguing to consider that the high relative risk of recurrence for lesions, such as heterotaxy or aortic valve disease, may reflect the presence of inherited variants of genes already implicated in autosomal dominant disease that have more modest functional effects resulting in incomplete penetrance7, 11.
Despite these interesting findings, the study by Oyen et al. found that the contribution of CHD family history to the overall prevalence of CHD was only ~2–4%, suggesting that most of the cases of CHD in the population did not have a family history of disease. This observation suggests that predisposition to disease involves multiple factors, including different genetic loci, epigenetic factors (e.g., DNA methylation or histone modifications that affect gene expression (Fig. 2)), environmental influences, subtle hemodynamic factors during cardiac development, or a combination of these factors. In this scenario, a potential disease-susceptibility allele could lead to disease penetrance or non-penetrance (Fig. 1), depending on the size of the effect of the susceptibility allele and the presence of “second hits” that modify the phenotype.
Given these findings from Oyen et al. and related studies10, 12, where should we direct our efforts to better understand and combat cardiac birth defects? This study emphasizes the growing need to understand cardiac phenotypes that may have multifactorial etiologies so that we can direct efforts toward prevention and, eventually, novel therapies. Family-based genetic mapping studies and population-based association studies will play important roles in elucidating the rare and common genetic variants that predispose to CHD. Such studies are becoming increasingly feasible with rapidly evolving genome-wide technologies that can survey the genome for potential genetic changes. For example, high-density SNP detection using microarrays and, more recently, next-generation deep sequencing for whole exome analyses are tremendously powerful tools to detect human genetic variation. However, the ability to associate genetic changes with disease involves complex bioinformatics analyses that will need to be developed as the compendium of human genetic variation is discovered. In addition, a rate-limiting step will likely be access to sufficient numbers of patients with similar heart defects for association studies. This effort will require nation-wide biobanks with high-quality phenotypic information.
Despite the current and future advances in genetic discovery in cardiac disease, there will undoubtedly be further complexities that underlie sporadic and familial disease. For example, non-coding regions of the genome have been traditionally understudied or overlooked altogether, and these and other regions of the genome need to be effectively interrogated to identify small non-coding RNAs (e.g., microRNAs, see Fig. 2), introns, and novel sequences potentially associated with disease13, 14. Furthermore, as the genetic bases of CHDs are elucidated, our understanding of environmental contributions and fetal hemodynamics to disease predisposition needs to grow substantially.
What does this mean practically for a family dealing with a child suffering from CHD? Currently, there is often no clear explanation of causality for the family. In fact, CHD, as with many other conditions, appears to fall into the conundrum of probabilistic causality15. That is, it cannot be assumed that the purported “cause” (be it genetic or environmental) is always related to the expected “effect.” Given this uncertainty, when dealing with potential future pregnancies for a family already affected by disease, the current status of the field promotes the use of the empiric recurrence risks to provide further information. The challenge that lies ahead is to provide better insight into the likelihood of disease. Identification of the predisposing genetic variants may lead to approaches involving modification of the environmental factors that might be able to lower penetrance even in the presence of susceptibility allele. We hope that studies such as those by Oyen et al. can translate into further research from the field toward more precise testing for disease, sensitive and universal prenatal screening and improved genetic counseling. For the sake of families faced with recurrent disease and for those that will encounter their first birth defect, it is incumbent upon the field to engage in a focused effort to determine the underlying cause for the high recurrence risk reported here for sub-types of CHD and to ultimately identify effective preventive measures.