Can genetics predict risk for sudden cardiac death? The relentless search for the Holy Grail

This editorial refers to ‘A comprehensive evaluation of the genetic architecture of sudden cardiac arrest†’, by F.N. Ashar et al., on page 3961. Sudden cardiac death (SCD) remains a daunting problem for at least three major reasons: (i) with a prevalence of 20% of total mortality in the industrial world it has a major impact on society; (ii) it has devastating psycho-social effects on the families of the victims who often are still in their prime; and (iii) it represents an unconquered scientific challenge for medicine and cardiology.1 Understandably, major resources are devoted to the attempts at identifying risk factors for SCD and for its precursor, sudden cardiac arrest (SCA), and also to develop a valid risk stratification approach. A consensus document, focusing on the current status and future directions of risk stratification for SCD, was published recently by a group of investigators with long-standing interest in SCD: they identified the autonomic nervous system and genetic screening as the most promising areas for research.1 As to the role of genetics, they concluded that (i) genetic screening has a high potential to contribute to the identification of individuals at risk for SCD, as already proven for genetically based diseases; (ii) significant insights can also be expected for SCD in the general population, but this will require studying homogeneous phenotypes; and (iii) novel technologies, including next-generation sequencing (NGS) and the smart analysis of clusters of common genetic variants as risk modifiers, hold significant promise. One of the limitations to actual progress is represented by the fact that researchers, even after collecting SCA cases for many years, have each collected SCA cohorts that still lack the size and statistical power for the identification of all interacting risk factors which is necessary to design risk scores for SCA occurrence. To overcome these difficulties, a consortium of 16 partners across 10 European countries have joined forces in the ‘European Sudden Cardiac Arrest network: towards Prevention, Education and New Effective Treatments project’ (ESCAPE-NET), recently funded by the European Union’s Horizon 2020 programme. This 5-year project will bring together the largest European cohorts of SCA victims in one joint harmonized database of >94 000 cases of SCA with >15 000 DNA samples.2 Concurrently, other major efforts are ongoing and, in this issue of the European Heart Journal, Ashar and colleagues report their attempt to identify potential genetic loci associated with SCA and risk factors causally associated with SCA.3 They carried out the largest exploration of the genomics of SCA to date. Nine genome-wide association studies (GWASs) of individuals of European descent comprising a total of 3939 cases and 25 989 controls were collected and analysed by a meta-analysis in the discovery stage. This step was followed by a replication stage comprising a total of 4918 additional cases and 21 873 controls obtained from 12 studies with individuals of European, African, and Asian descent. The replication stage consisted of (i) a replication of the top 26 single nucleotide polymorphisms (SNPs) resulting from the discovery stage and (ii) a candidate gene analysis focused on 54 well-known inherited arrhythmia genes. At first glance, this may appear as a negative study because it failed to identify common genetic variants associated with SCA at the genome-wide level either in the main analysis or in secondary analyses focused on specific subgroups (individuals of European descent, men, women, younger participants, or cases with ventricular fibrillation/shockable rhythm). Moreover, there was no evidence that common genetic variants in arrhythmia genes are associated with SCA risk in the general population. This is curious given that electrical instability is an important cause of SCA and that previous studies have already examined inherited arrhythmia genes to identify genetic variants that influence SCA risk.4–6 The authors’ comment on this finding was that rare mutations in ion channel genes and other electrophysiology-related genes increase arrhythmia risk and may also increase SCA risk,7 while common genetic variants might not be relevant for SCA. This conclusion is intriguing and important. Like other GWASs, this study followed the ‘common disease–common variant’ hypothesis according to which a major portion of risk for a common disease in populations is due to a defined number of common genetic variants. Conversely, an alternative and competing theory, the ‘common disease–rare variant’ hypothesis, suggests that multiple rare variants might have a greater impact on SCA risk.8 The results of Ashar et al. support this latter hypothesis. Even though the number of subjects enrolled is objectively large, an adequate power was reached only for analysis of arrhythmia genes, while the study was underpowered to detect effect sizes typically observed in GWASs of complex traits. Notwithstanding this limitation, the analytical strategy adopted by Ashar et al. represents a point of strength and has considerable merit. Indeed, they also applied the Mendelian randomization (MR) approach to investigate clinical and subclinical known risk factors and their causal effects on SCA in depth. In general, understanding the causal role of risk factors in cardiovascular and other diseases is crucial in order to find effective approaches (including pharmacological therapies) for disease treatment or prevention.9 Recently, the use of MR has become popular in medicine, although it is not a new concept, as the origins of this method can be traced back over five decades.10 However, in the post-GWAS era, the enormous increase of knowledge regarding human common genetic variability obtained by GWASs, epigenomics, and other new genome-wide technologies has led to a revival of MR and of its applications. MR allows estimation of a causal effect from observational data in the presence of confounding factors. This analysis uses genetic polymorphisms with well-known effects as instrumental variables to determine whether an observational association between a risk factor and an outcome is consistent with a causal effect. The MR approach involves finding genetic variants which are associated with an exposure, and then testing the association between these variants and the outcome. The causal relationship between exposure and outcome can be estimated only if some essential conditions are respected: the genetic variants must be truly associated with the exposure; they should not be associated with confounders of the exposure–outcome relationship; and they should only be related to the outcome of interest through the exposure under study.11 Ashar et al. used a multi-SNP genetic risk score association (GRSA) model to compare genetic associations of 18 known SCA risk factors with those of SCA. Briefly, for each of the 18 putative SCA risk factors, a GRSA was obtained using SNPs which were significantly associated with the trait in previous GWASs. The associations of those SNPs with SCA risk factors and the associations emerged with SCA were then used to calculate an inverse-variance weighted multi-SNP GRSA. To simplify the concept, the GRSA could be described as the result of a linear regression between the effect of those SNPs on SCA risk factors and the effect that they have on SCA outcome. The MR approach allowed the authors to unmask significant causal associations between SCA and several clinical and subclinical risk factors. Coronary artery disease (CAD) and CAD risk factors such as blood pressure, lipids, and diabetes were found to be causally associated with SCA, as were anthropometric traits such as height and body mass index (BMI). Last, but not least, the authors reported significant causal associations between SCA and electrical instability traits such as the QT interval and atrial fibrillation, suggesting aetiological roles for these traits in SCA risk. This last finding supports the current understanding of the mechanisms underlying SCA, pointing to the fact that factors associated with electrical instability play a major role in increasing risk for SCA and thereby for SCD. Interestingly, this elegant and complex analysis fully confirms the first evidence, published exactly 40 years ago,12 that a prolongation of the QT interval in patients who had suffered an acute myocardial infarction significantly increases the risk of SCD. The 1978 study had a much simpler design, as 55 patients with an acute myocardial infarction were followed for 7 years with an ECG taken every 2 months. The QTc of the 28 patients who died suddenly was much longer than that of the survivors, and a constantly prolonged QTc carried a 2.16 greater risk of SCD. In science, it is reassuring when totally different studies reach the same conclusions 40 years apart. A significant limitation of the study, correctly recognized by Ashar et al., is the lack of common variants exhibiting large effect sizes associated with SCA which at the moment limits their potential clinical utility for risk prediction. However, during the last 10 years, MR studies have generated much progress in cardiovascular disease causality, pathogenesis, prevention, and treatment,9 and the present study confirms the power of this approach by exploring for the first time causal relationships behind SCA and relevant risk factors, enhancing our understanding of SCA pathophysiology and offering new insights for potential interventions. For example, those factors which are causally associated with increased SCA risk could represent a novel rational target for prevention. The observation that QT interval prolongation increases risk also in the general population may shift the current paradigm by which the use of IKr blockers must be avoided in patients with the congenital long QT syndrome (LQTS)13 and is recognized as a major cause of drug-induced LQTS.14 Indeed, it may be reasonable to consider that, whenever possible and medically appropriate, drugs without IKr-blocking activity should be preferred. In conclusion, the advances in NGS technology have opened up new opportunities to explore whole genomes and whole exomes, expanding our knowledge of genetic variability and enabling the ability to search for potential ‘modifier genes’. Indeed, there is growing interest in these genetic variants capable of influencing in either direction the arrhythmic risk associated with a disease-causing mutation or with an arrhythmogenic substrate (e.g. an acute myocardial infarction).15 Further studies based on NGS approaches might help in integrating exome sequence variation into a Genomic Risk Score, which could be combined with risk scores from common variants, from epigenomics and from the ‘exposome’, defined as the environmental influence acting on the genetic background. This combined approach, schematically represented in Take home figure, might represent the right cocktail to come closer to understanding the predictive power of genetics for sudden cardiac death. The scheme proposed represents an approach that aims at integrating exome or genome sequence variation into a ‘Genomic Risk Score’. This score combines the effects of rare and common genetic variants, and also considers epigenetic variability and the ‘exposome’. The exposome is defined as the environmental influence acting on the genetic background. This integrated ‘omic’ approach might represent a possible novel strategy to obtain risk scores applicable for direct clinical use. The authors are grateful to Pinuccia De Tomasi for expert editorial support. This work has received funding from the European Union’s Horizon 2020 research and innovation programme under acronym ESCAPE-NET, registered under grant agreement no. 733381. Conflict of interest: none declared.