The clinical utility of pediatric cardiomyopathy genetic testing: From diagnosis to a precision medicine-based approach to care

Abstract

Background:

Pediatric-onset cardiomyopathies are rare yet cause significant morbidity and mortality in affected children. Genetic testing has a major role in the clinical evaluation of pediatric-onset cardiomyopathies, and identification of a variant in an associated gene can be used to confirm the clinical diagnosis and exclude syndromic causes that may warrant different treatment strategies. Further, risk-predictive testing of first-degree relatives can assess who is at-risk of disease and requires continued clinical follow-up.

Aim of Review:

In this review, we seek to describe the current role of genetic testing in the clinical diagnosis and management of patients and families with the five major cardiomyopathies. Further, we highlight the ongoing development of precision-based approaches to diagnosis, prognosis, and treatment.

Key Scientific Concepts of Review:

Emerging application of genotype-phenotype correlations opens the door for genetics to guide a precision medicine-based approach to prognosis and potentially for therapies. Despite advances in our understanding of the genetic etiology of cardiomyopathy and increased accessibility of clinical genetic testing, not all pediatric cardiomyopathy patients have a clear genetic explanation for their disease. Expanded genomic studies are needed to understand the cause of disease in these patients, improve variant classification and genotype-driven prognostic predictions, and ultimately develop truly disease preventing treatment.

Introduction

Pediatric onset cardiomyopathy is a heterogenous group of primary myocardial diseases that include hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), restrictive cardiomyopathy (RCM), arrhythmogenic cardiomyopathy (ACM), and non-compaction cardiomyopathy (NCCM).^{1, 2} Overall, the annual incidence is between ~1 per 100,000 children, and diagnosis is more common in the first year of life, with incidence during this time period at ~8 per 100,000 children.^3–6 DCM is the most common pediatric-onset cardiomyopathy, comprising ~50–60% of all childhood cardiomyopathies, followed by HCM at ~25–40%, NCCM at ~5–10%, and RCM at ~2.5–9%.^3–7 ACM is uncommonly diagnosed in childhood as children rarely meet adult diagnostic criteria; however, children can still present with a predominantly arrhythmic phenotype.⁸ Outcomes in pediatric cardiomyopathies have improved over the last decade. For example, death rates of children with DCM decreased from 18% between 1990–1999 to 9% between 2000–2009.⁹ However, cardiomyopathies remain a cause of significant morbidity and mortality, especially when diagnosed at a young age. Roughly 6% of pediatric cardiomyopathy patients experience sudden cardiac death (SCD).¹⁰ Finally, two- to five-year survival varies between 13% to 22% across all cardiomyopathies.^{5, 6, 8, 10–12}

Pediatric cardiomyopathies are caused by diverse mechanisms including genetic variants, infection, neuromuscular disease, inherited metabolic disorders, and mitochondrial disease.^{4, 6, 13} Cardiomyopathies can occur in the context of a syndrome with non-cardiac manifestations of disease, so-called syndromic cardiomyopathies. Non-syndromic cardiomyopathies are believed to be heritable in a predominantly autosomal dominant (AD) fashion. Combinations of pathologic variants have also been identified as disease-causing and are associated with poor prognosis and accelerated disease.¹⁴ While less common, de novo pathogenic variants are also an established cause of cardiomyopathy, affecting up to 18% of pediatric patients.¹⁵ Genetic testing plays a central role in diagnosis confirmation as well as risk-assessment of family members to determine those at-risk of developing disease.

Rather than discrete diagnoses, pediatric cardiomyopathies can present on a spectrum of disease with overlapping phenotypes and causative genes. Within families with multiple individuals positive for the same disease-associated variant, variable expressivity and incomplete penetrance can be observed. For example, families that share the same genetic variant may have genotype positive individuals without disease (incomplete penetrance), while others may have variable disease age of onset, manifestations, and course (variable expressivity). In addition, a single patient may present with features of many cardiomyopathies, such as arrhythmic DCM, or NCCM with HCM or DCM characteristics.¹⁶ This phenotypic overlap is supported by a similarly overlapping genetic basis. For example, variants in sarcomere genes are often shared between HCM, DCM, and NCCM, yet the etiology of incomplete penetrance and variable expressivity is still emerging (Figure 1). In this review, we discuss the clinical utility of genetic testing in pediatric cardiomyopathy and the potential for genetics to inform a precision medicine-based approach to care.

Figure 1: Variable expressivity of cardiomyopathy genes.

Variants in many genes, particularly those encoding myofilament proteins that comprise the sarcomere, are shared between multiple different cardiomyopathies. Syndromic causes of disease are shaded grey.

Overview of Clinical Genetic Testing in Pediatric Cardiomyopathies

Clinical genetic testing begins with the affected proband and predominantly takes the form of targeted gene panel-based sequencing test, although exome sequencing (ES), genome sequencing (GS), and chromosomal microarrays which identify copy number variants (CNVs), have diagnostic roles in specific situations. Initial evaluation of a patient with cardiomyopathy involves describing a detailed 3-generation family history to evaluate inheritance patterns.^17–20 If a P/LP variant is identified in a proband, cascade genetic testing and clinical testing is recommended for first-degree relatives.^17–20 Some guidelines suggest that when a truly pathogenic variant is identified in a proband, genetic testing can precede clinical evaluations in relatives.¹⁹ Importantly, the confidence by which risk-predictive testing can be applied rests solely on the likelihood that the variant is disease-causative. The classification of a variant can evolve as more patients are tested and better methods to validate pathogenicity are developed. Periodic reassessment of disease-associated genes is critical to ensure variants are correctly classified and at-risk relatives are correctly followed. (Figure 2)

Figure 2: Diagram of the multifaceted role of clinical genetic testing in pediatric cardiomyopathy.

Clinical genetic testing plays a central role in both diagnostic testing in a child with concern for cardiomyopathy as well as risk-predictive testing in the family when a disease-associated variant is identified. As our understanding of the genetic causes of cardiomyopathies improves, the door opens for better prognostic predictions and precision medicine approaches to diagnosis and treatment.

Hypertrophic Cardiomyopathy

Clinical Diagnosis

Pediatric HCM affects ~3 per 100,000 children with most children diagnosed either in infancy or adolescence.^{3, 5, 21} HCM is characterized by increased left ventricular or interventricular septal thickness in the absence of an identifiable trigger for hypertrophy. Classically, HCM has preserved to hyperdynamic ejection fraction and can present with varying degrees of left ventricular outflow tract obstruction with or without systolic anterior movement of the mitral valve.^{1, 22, 23} Diagnosis of pediatric HCM is typically made based on the 2020 AHA/ACC diagnostic criteria.²⁰

Genetics of HCM

The classical cause of HCM is autosomal dominant variants in genes which encode components of the cardiac sarcomere, so-called sarcomeric HCM. Causes of non-syndromic, non-sarcomeric HCM are less common and heterogenous including variants in Z-disc and calcium-handling proteins (Table 1). In contrast to “cardiac only” disease, HCM can be phenotypically mimicked by genetic syndromes including Pompe disease (GAA), Noonan syndrome (PTPN11, RAF1, SOS1, etc.), and Freidrich ataxia (FXN), all of which can present primarily with left ventricular hypertrophy and have associated extra-cardiac manifestations.^{22, 24, 25} Familial HCM is typically AD with incomplete penetrance and variable expressivity. X-linked causes have also been described, especially when the individual manifests evidence of syndromic HCM such as in Barth syndrome or Danon disease.²⁴

Table 1 –

Genetic etiology of hypertrophic cardiomyopathy

Gene	Protein	Chr	Mode of inheritance	Pediatric frequency	OMIM151	Reference
Giant Filament
TTN	Titin	2q31	AD, AR	< 1%	188840	60
Thick Filament
MYH7	β-Myosin heavy chain 7	14q11-q12	AD	~18–49%	160760	26, 27
MYL2	Regulatory myosin light chain	12q23-q24	AD	Rare	160781	26, 27
MYL3	Essential myosin light chain	3p21-p21	AD, AR	Rare – 5%	160790	26, 27
Intermediate Filament
MYBPC3	Cardiac myosin-binding protein C	11p11	AD, AR	~21–36%	600958	26, 27
Thin Filament
ACTC1	α-Cardiac actin	15q14	AD	Rare – 5%	102540	26, 27
TNNC1	Cardiac troponin C	3p21	AD	Rare	191040	47, 48
TNNI3	Cardiac troponin I	19p13	AD	~2–3%	191044	26, 27
TNNT2	Cardiac troponin T	1q32	AD	~4–6%	191045	26, 27
TPM1	α-Tropomyosin	15q22	AD	Rare – 6%	191010	26, 27
Non-sarcomere genes
JPH2	Junctophilin-2	20q13	AD, AR	Rare	605267	152–154
Metabolic Cardiac Hypertrophy (HCM Mimickers)
FXN	Frataxin	9q13	AR	Rare	606829	155, 156
GAA	Lysosomal α-glucosidase	17q25	AR	Rare	606800	49,156
GLA	α-Galactosidase A	Xq22	XL	Rare	300644	157
LAMP2	Lysosome-associated membrane protein 2	Xq24	XD	Rare	309060	157
PRKAG2	AMP-activated protein kinase	7q35-q36	AD	Rare	602743	157
PTPN11	Tyrosine-protein phosphatase nonreceptor type 11	12q24	AD	Rare	176876	35
RAF1	RAF serine/threonine kinase	3p25	AD	Rare	164760	157

AD, autosomal dominant; AR, autosomal recessive; XD, X-linked dominant; XL, X-linked; XR, X-linked recessive.

While classically seen as a monogenic, autosomal dominant disease of the sarcomere, only 50–65% of children with HCM have conclusive genetic tests.²⁶ Variants in MYH7 and MYBPC3 are the most common genetic cause of HCM and account for at least 75% of causative variants detected in pediatric and adult HCM.^{22, 24, 26, 27} Other sarcomeric genes commonly tested in HCM panels include TNNT2, MYL3, and ACTC1.²⁸ Further, variants in Z-disc or calcium-handling proteins are also associated with HCM, including JPH2 and PLN.²⁸ The yield of P/LP variants in sarcomere genes ranges from ~70% of adults with a positive family history, to ~30% of adults with a negative family history.^{29, 30} Similarly, sequencing eight sarcomere genes (MYH7, MYBPC3, TNNT2, TNNI3, TPM1, MYL3, MYL2, and ACTC1) in 84 pediatric patients with idiopathic HCM identified variants in ~50% of patients without family history and ~65% of patients with positive family history.²⁶ Adult HCM is phenotypically heterogenous and far more common than pediatric HCM with estimated prevalence between 1:200 to 1:500.^{31, 32} Although adult and pediatric HCM share causative variants, the mechanisms that drive the early onset phenotypes observed in pediatric HCM are poorly understood.²⁶ Finally, young patients with syndromic HCM are common in pediatric cohorts. In the Pediatric Cardiomyopathy registry, ~8% of HCM patients have neuromuscular disorders and ~9% of HCM patients have inherited metabolic disorders.⁶ Given the numerous causes of HCM, which can present with or without syndromic features, defining disease etiology based on clinical evaluation alone can be challenging.

Diagnostic and Risk-Predictive Genetic Testing

HCM guidelines universally recommend genetic testing in affected individuals to confirm a clinical diagnosis.^17–20 Given that the majority of HCM involves sarcomeric genes, initial testing is often limited to sarcomeric genes with the strongest association with HCM and include syndrome-associated genes if there is an atypical presentation or extracardiac symptoms.^17–20 Expanded panels, ES, and GS are options for follow-up testing should the gene panel be negative. The utility of expansive gene testing is emerging. One study of 2,912 non-syndromic HCM patients found that genetic yield increased only modestly when the HCM CardioChip panel expanded from 11 genes to over 50.³³ GS in 46 families with inconclusive panel sequencing identified P/LP variants in 20%, including in genes excluded from panel tests, deep intronic and splice site variants within genes on the gene panel which were missed by the panel test, and mitochondrial genes.³⁴ Finally, in a comparison study between panel-based and GS in 41 patients, GS identified 19/20 variants identified by panel testing, and identified one novel variant in PTPN11, a gene classically associated with Noonan syndrome.³⁵ The ability to detect atypical manifestations of disease, such as syndromic HCM with very mild extracardiac manifestations, is a strength of expansive genetic testing.

As few genetic variants are completely penetrant, a family member who is genotype-positive but phenotype-negative may or may not develop HCM in the future. For example, a study of 285 individuals from 156 families with HCM history found 46% of sarcomere variant carriers developed HCM within 15 years of familial genetic screening.³⁶ Thus, potentially lifelong clinical follow-up could be considered to facilitate early diagnosis and intervention.^{17, 18} Finally, risk-predictive genetic testing of a known pathogenic variant in a family can free gene-negative individuals from further clinical evaluation. Among a cohort of 90 HCM patients and first-degree relatives who underwent clinical and genetic evaluation for HCM, 41% of relatives were both genotype and phenotype-negative and could be excluded from further cardiac follow-up.³⁷ Some guidelines suggest clinical or genetic screening of children with a family history of HCM should begin after age 10.¹⁹ Despite this, variable expressivity of shared variants can cause early-onset disease, thus limiting screening to early adolescence may not account for early phenotypic presentation and shrink a window for early disease intervention.³⁸ One study of 1,198 children under 18-years-old who were referred for family screening for HCM found 5% met diagnostic criteria for HCM on clinical evaluation.³⁹ Of the 5% of children who were diagnosed with HCM, 69% carried pathogenic variants, suggesting that risk-predictive genetic testing in children can be useful for targeting early clinical management to those at-risk.³⁹

Prognosis and Therapeutics

Beyond confirming clinical diagnosis, genetic testing in HCM is useful for predicting clinical course and prognosis throughout a child’s life. Multiple studies demonstrate that patients with sarcomere variants tend to be more severely affected than patients with non-sarcomeric variants.^{30, 40} Within the sarcomere genes, some argue that specific sarcomere variants are inherently more pathogenic than others.^41–43 However, comparing unrelated families with the same variant highlights the marked heterogeneity in disease severity, suggesting other genetic and environmental factors may influence phenotypic expression of disease.⁴¹ Roughly 5% of HCM patients have two or more rare, traditionally AD variants in cardiomyopathy genes, and higher genetic variant burden is associated with significantly increased risk of transplant, sudden cardiac arrest (SCA), or death.^{14, 44–46} There is emerging evidence for gene-specific phenotypes, such as variants in pathogenic hotspots of TNNC1 predisposing to a proarrhythmic phenotype with pediatric sudden death.^{47, 48} These prognostic genotype-phenotype correlations hold the promise of informing a child’s clinical course and prognosis when fully elucidated.

In pediatric HCM patients with a clear genetic etiology, there is opportunity for true disease-modifying treatments based on genetic findings, particularly in metabolic forms of HCM. For example, in GAA-positive Pompe disease, alpha-glucosidase enzyme replacement therapy is associated with HCM reversal in affected children.⁴⁹ Similarly, in two patients with HCM associated with Noonan syndrome, MEK-inhibition via trametinib significantly reduced cardiac hypertrophy and NT-proBNP measurements.⁵⁰ In sarcomeric pediatric HCM, there are no widely used changes to therapy based on genetic testing; however, this may change. Recent efforts to target sarcomere positive HCM with myofilament small molecule inhibitors have demonstrated early success. The small-molecule mavacamten, which inhibits sarcomere contractility, is associated with hypertrophy prevention in Mhy7 variant HCM mouse models.⁵¹ The EXPLORER-HCM randomized, double blind, placebo-controlled trial in 429 adult HCM patients found mavacamten improved exercise capacity, outflow tract obstruction, and ventricular function, though only 31% of the mavacamten group had P/LP variants in sarcomere genes.⁵² No data on mavacamten usage in pediatric HCM patients with or without sarcomere variants has been reported to date. Finally, the ongoing VANISH trial is investigating the efficacy of valsartan in preventing disease progression in patients with early stage, sarcomere positive HCM; however, final results are not yet available.⁵³

Precision Medicine

While there is emerging evidence for the prognostic and therapeutic utility of genetic testing, challenges must be overcome to achieve a true precision medicine-based approach to HCM management. Although expanded sequencing can improve diagnostic testing yield in patients with HCM, ES and GS may exacerbate the challenge of VUS identification in rare cardiomyopathy genes. Moreover, broad use of diagnostic ES and GS in infants and children without a specific clinical concern for cardiomyopathy raises the possibility of incidentally findings in cardiomyopathy genes.³⁴ For example, population-based “background” genetic variation in HCM genes in the general population is relatively high at roughly 5% which creates a hurdle for disease risk assessment in a child who has no evidence of cardiomyopathy on clinical evaluation.⁵⁴

Dilated Cardiomyopathy

Clinical Diagnosis

Non-ischemic DCM is the most common pediatric cardiomyopathy and affects between 0.6–1.0 per 100,000 children.^{3, 5, 21} Non-ischemic DCM is characterized by reduced systolic dysfunction and LV end-diastolic diameter, and diagnosis is based on echocardiography and/or cardiac MRI, and rarely endocardial biopsy if amyloidosis or an infectious cause is suspected.⁵⁵.⁵⁶ Like in the case of pediatric HCM, the causes of DCM in children are diverse, ranging from isolated familial and sporadic DCM, to mitochondrial disease, neuromuscular disease, infectious myocarditis, and inherited metabolic disorders.⁵⁷

Genetics of DCM

DCM is typically associated with autosomal dominant inheritance; however, autosomal recessive inheritance has also been reported, for example, in JPH2.⁵⁸ Over 111 genes are offered on some DCM testing panels as being associated with DCM developing, including sarcomere genes (MYH7, TPM1, and, TTN), calcium-sensitive and calcium-signaling genes (JPH2, and PLN), z-disk genes (ACTN2, BAG3, and CRYAB), and nuclear envelope genes (LMNA, EMD), nearly all which are rare causes of pediatric DCM (Table 2).⁵⁹ Truncating titin variants (TTNtvs) are by far the most common genetic cause of DCM and are implicated in up to 25% of familial cases.⁶⁰ CNVs, such as 1p36 deletion syndrome, are also associated with syndromic DCM with a NCCM phenotype.^{61, 62} Indeed, there is overlap between genes associated with DCM, HCM, and the other cardiomyopathies which suggests to some extent all cardiomyopathies exist on the same phenotypic spectrum. For example, in a cohort of 639 patients with sporadic or familial DCM, 31% of identified variants were also associated with ACM and 16% with HCM.⁶³ In clinical practice, the yield of genetic testing in familial DCM ranges from 30% to 40%.^{64, 65}

Table 2 –

Genetic etiology of dilated cardiomyopathy

Gene	Protein	Chr	Mode of inheritance	Pediatric frequency	OMIM	Reference
Sarcomeric Genes
MYBPC3	Cardiac myosin-binding protein C	11p11.2	AD, AR	Rare	600958	59, 63, 67, 158–160
MYH6	β-Myosin heavy chain 6	14q11	AD	Rare	160710	59, 63, 67, 158–160
MYH7	β-Myosin heavy chain 7	14q11-q12	AD	Rare	160760	59, 63, 67, 158–160
TNNI3	Cardiac troponin I	19p13	AD	Rare	191044	59, 63, 67, 158–160
TNNT2	Cardiac troponin T	1q32	AD	Rare	191045	59, 63, 67, 158–160
TPM1	α-Tropomyosin	15q22	AD	Rare	191010	59, 63, 67, 158–160
TTN	Titin	2q31	AD, AR	~15–25%	188840	60, 72, 90, 161, 162
Non-sarcomeric genes
ACTN2	Actinin, α 2	1q43	AD	Rare	103573	59
BAG3	BAG family molecular chaperone regulator 3	10q26	AD	Rare	603883	156,163
CRYAB	Crystallin, α B	11q23	AD	Rare	123590	59
FLNC	Filamin C	7q32	AD	Rare	102565	79, 80
JPH2	Junctophilin 2	20q13	AD, AR	Rare	605267	58, 164
LMNA	Lamin A/C	1q22	AD	Rare	150330	24
PLN	Phospholamban	6q22	AD	Rare	172405	59
RBM20	RNA-binding motif protein 20	10q25	AD	Rare	613171	81
Syndromic DCM
CPT2	Carnitine palmitoyltransferase 2	1p32	AD, AR	Rare	600650	24
DMD	Dystrophin	Zp21	XR	Rare	300377	24
EMD	Emerin	Xq28	XR	Rare	300384	24
TAZ	Tafazzin	Xq28	XR	Rare	300394	24,133

AD, autosomal dominant; AR, autosomal recessive; XD, X-linked dominant; XL, X-linked; XR, X-linked recessive

Diagnostic and Risk-Predictive Genetic Testing

Genetic testing can help distinguish between cardiac-only DCM and syndromic/neuromuscular DCM such as Duchenne muscular dystrophy (DMD), Emery-Dreifuss muscular dystrophy (EMD, SYNE1, SYNE2), or carnitine palmitoyltransferase II deficiency (CPT2). DCM gene panels are commonly used for first line testing; however, ES and GS have also been explored given the genetic heterogeneity of DCM. ES in 83 DCM patients identified P/LP variants in 12% of all patients, and 18% in the subset of 11 pediatric patients.⁶⁶ In a 21 patient pediatric DCM cohort, genetic yield of ES approached 50%, though only variants in RRAGC and TAF1A in two patients would have been missed by standard targeted-sequencing panels.⁶⁷ Finally, GS can improve genetic yield by capturing CNVs along with more traditional means such as chromosomal microarrays.^{61, 62}

Whether it is by expansive gene panels, ES, or GS, increasing detection of pathologic variants in rare DCM genes may reduce the specificity of the findings. Yet, confidently identifying pathogenic variants can be challenging, especially when background variation for many rare genes is higher than the prevalence of variants among those with disease. Recent efforts, led by ClinGen, have re-evaluated variant pathogenicity in the 100+ genes associated with DCM. One study sequenced 56 DCM genes in 2538 patients and found strong associations with rare variants in only 12 genes.⁶⁸ The yield of these genes remains low, explaining only 1 in 4 children with familial DCM.⁶⁸ Other genes previously considered monogenic causes of DCM may still contribute to disease progression but in a more complex and combinatorial way than previously thought.

Roughly 20–35% of DCM is familial, and risk predictive genetic testing can help capture affected or at-risk individuals who share a P/LP variant with an affected child.^{69, 70} Like in the case of HCM, cascade genetic and cardiac screening of family members is should be offered when a P/LP variant is identified in an affected proband.

Prognosis and Therapeutics

Established genotype-phenotype correlations in pediatric patients with DCM can provide prognostic guide to treatment plans. For example, some TTNtvs, which are not associated with DCM development in isolation increase susceptibility to stress-induced cardiomyopathies, such as in the setting of pregnancy or doxorubicin chemotherapy.^71–73 It follows that TTN variant carriers, especially those with affected family members who share a variant, be monitored carefully in these contexts. Further, LMNA variants, which are often shared with ACM, are associated with an arrhythmic DCM phenotype, with or without the fibrofatty replacement characteristic of ACM.^74–77 Other genes associated with arrhythmic DCM phenotypes include FLNC, RBM20, and PLN.^78–82 Finally, TNNC1 variants are associated with early onset DCM, and increased rates of heart transplant and SCD compared other genes.⁴⁷ Identification of genes with arrhythmic potential can guide careful electrophysiological evaluation and potential implantable cardiac defibrillator (ICD) placement in at-risk individuals.

As in HCM, novel sarcomere-targeted small molecules have potential as disease modifying treatments.⁸³ The myosin activator omecamtiv mecarbil was evaluated in the COSMIC-HF and GALATIC-HF trials in patients with systolic heart failure, and was found to improve cardiac function and long-term outcomes.^{84, 85} Though there have not been clinical trials testing efficacy of omecamtiv mecarbil in DCM patients specifically, in vitro studies with patient-derived induced pluripotent stem cell cardiomyocytes (iPSC-CMs) demonstrated improved cardiomyocyte function in certain TNNT, TPM1, and MYH7 variants.^86–88 Finally, development of gene therapy, such as in Duchenne muscular dystrophy, holds promise to target and rescue functional expression of pathologic gene variants. For instance, exon 51 skipping via eteplirsen in the PROMOVI trial holds promise to slow progression of cardiomyopathy alongside muscular dystrophy.⁸⁹” Application of similar gene therapy strategies to DCM may represent a new therapeutic avenue to target DCM at the level of individual variants.

Precision Medicine

Advances in precision medicine and application to DCM are needed to aid in the interpretation of VUS and increase understanding of how rare and common variants can synergistically increase likelihood of disease. For example, TTN variants found in 1–3% of the general population without evidence of disease expression.^{55, 90, 91} Not every TTN variant carrier will develop DCM or have increased susceptibility to pregnancy or chemotherapy-induced cardiomyopathy.^{71, 72} Recent evidence has suggested that TTNtvs falling in an exon that can be spliced-out in the population may be less likely to cause disease.⁹² Conversely, TTNtvs which fall into exons frequently spliced-in are associated with cardiomyopathic disease.⁹² In addition, amino acid signal-to-noise analysis has demonstrated that pathogenic TTNtvs fall in conserved hot-spots which can help stratify benign and pathogenic TTNtvs.⁹³

Arrhythmic Cardiomyopathy

Clinical Diagnosis

Arrhythmic cardiomyopathy (ACM) is an umbrella term that includes arrhythmogenic right ventricular cardiomyopathy (ARVC), as well as arrhythmic left ventricular cardiomyopathy and biventricular phenotypes.^{94, 95} ACM is rarely diagnosed in children, and the estimated overall population prevalence is between 1:2000 to 1:5000.^{3, 5, 96} Importantly, ACM is associated with a high degree of mortality and is estimated to cause at least 16% of sudden cardiac deaths in children and young adults between 1 and 19 years old.⁹⁷ ACM is characterized by fibro-fatty infiltration of the right ventricle, systolic dysfunction, and ventricular arrhythmias that predispose sudden cardiac death.^{94, 98, 99} Diagnosis of right ventricular ACM is typically based on the 2010 Task Force Criteria, while the more recent Padua criteria include guidance for left and bi-ventricular disease.^{96, 99–101}

Genetics of ACM

The inheritance of ACM, like other cardiomyopathies, is typically autosomal dominant with incomplete penetrance and variable expressivity. The most common genes associated ACM include desmosomal genes (PKP2, JUP, DSC2, DSG2, DSP), as well as LMNA, RBM20, and TMEM43, among others (Table 3).⁹⁴ Autosomal recessive variants, such as in JUP or DSP, have also been described and are associated with Naxos disease which is characterized by syndrome manifesting with palmoplantar keratoderma and woolly hair.^102–104 Other case reports associate DSP variants with ACM, epidermolysis bullosa and aplasia cutis congenita.¹⁰⁵ The genetic yield of desmosome gene panel testing in ACM is high relative to other cardiomyopathies. In a study of 93 ACM probands, 58% of affected individuals had variants in a panel of PKP2, DSP, DSG2, DSC2, and TMEM43 versus 16% of healthy controls.¹⁰⁶ Another study of 158 patients with borderline or definite ACM tested for variants in a panel of desmosomal genes and LMNA.⁷⁴ The genetic yield was similar to the previous study with 57% genotype-positive individuals in desmosomal genes and 4% had variants in LMNA.⁷⁴

Table 3 –

Genetic etiology of arrhythmogenic cardiomyopathy

Gene	Protein	Chr	Mode of inheritance	Pediatric frequency	OMIM	Reference
DES	Desmin	2q35	AD	~3%	125660	115
DSC2	Desmocollin 2	18q 12	AD, AR	~3%	125645	74, 94, 106, 115
DSG2	Desmoglein 2	18q12-q12	AD	~10%	125671	74, 94, 106, 115
DSP	Desmoplakin	6p24	AD, AR	~6–13%	125647	74, 106, 115
JUP	Plakoglobin	17q21	AD, AR	Rare	173325	74, 94, 115
LMNA	Lamin A/C	1q22	AD	~10%	150330	95, 107, 115
PKP2	Plakophilin 2	12p 11	AD	30–45%	602861	94, 106, 113, 115
PLN	Phospholamban	6q22	AD	Rare	172405	115
TMEM43	Transmembrane protein 43	3p25	AD	Rare	612048	106, 115

AD, autosomal dominant; AR, autosomal recessive; XD, X-linked dominant; XL, X-linked; XR, X-linked recessive

Diagnostic and Rick-Predictive Genetic Testing

Genetic testing is a critical component of ACM diagnosis in children, as many children presenting with signs of ACM do not meet the adult diagnostic criteria. Importantly, pediatric patients may present with a more arrhythmic phenotype than adults, and LMNA variant carriers may lack significant fibrofatty replacement characteristic of typical ACM.^{95, 107} Guidelines for genetic testing in ACM typically focuses on evaluating a panel of desmosome genes and a number of other non-desmosomal genes, which capture ~60% of patients.^{18, 108} Use of broad panels containing non-desmosomal genes such as LMNA, can help distinguish between ACM and arrhythmic DCM, or potentially arrhythmia syndromes such as catecholaminergic polymorphic ventricular tachycardia characterized by variants in RYR2. ES or GS can be performed to capture other associated genes and CNVs that may be missed by traditional panels.^{109, 110} For example, deletions in PKP2 were identified in two patients with ACM, while another study found CNVs in ~5% of ACM patients.^{111, 112} Further, ES and/or Multiplex-Ligation-Dependent-Probe-Amplification in 70 ACM patients without desmosome variants identified PKP2 deletions affecting at least one exon in 5.7% of patients.¹¹³ Compound heterozygous or digenic variants in desmosome genes are common in ACM. One study found roughly 24% of patients had compound heterozygous variants in PKP2, and up to 42% of patient had additional variants in desmosome genes, in addition to PKP2.¹¹⁴ Multiple variants are associated with more severe disease, and variability in variant burden may explain, in part, the incomplete penetrance and variable expressivity seen within families.¹¹⁴

Recently, as with other cardiomyopathies, efforts have been made to re-evaluate both genes and variants associated with ACM. Re-evaluation of ACM genes found only eight (PKP2, DSP, DSG2, DSC2, JUP, TMEM43, PLN, DES) are definitively associated with ACM, and only these genes should be considered major criteria in the diagnostic paradigm detailed in the 2011 Task Force Criteria.¹¹⁵ Although RYR2 was previously associated with both ACM and CPVT, this most recent study re-classified RYR2 as only associated with CPVT, which speaks to the phenotypic overlap between ACM and channelopathies.¹¹⁵ Another study re-evaluated 39 ACM-associated variants after 5 years, and 30% of these variants were reclassified.¹¹⁶ This included both up- and down-grading of variants: 16% of variants originally classified as VUS became benign, and 12.5% of VUS were reclassified as LP.¹¹⁶

Studies estimate ~40–50% of relatives that carry desmosome variants will meet at least some diagnostic criteria for ACM between clinical evaluation and median 5-year follow-up.^{108, 117} Further, there is emerging evidence that exercise restriction in genotype-positive but phenotype-negative family members may protect against disease progression, emphasizing the importance and actionable implications of risk-predictive screening for family members.¹¹⁸

Prognosis and Therapeutics

Genetic testing can aid in establishing a plan for ACM management and overall prognosis. As in HCM, patients who carry multiple variants, either in desmosomal genes or other cardiomyopathy-associated genes have earlier onset and more severe disease than single variant carriers.^{108, 119, 120} PKP2 variants are associated with a more rapid progression to heart failure.¹²¹

Patients with LMNA variants share phenotypic characteristics with DCM and tend to lack fatty infiltration.¹⁰⁷ Further, in children and adolescents, LMNA variants are associated with left ventricular-dominant disease.⁹⁵ Finally, histological analysis of 60 explanted ACM hearts identified four distinct groups based on the degree of fibrofatty replacement and associated desmosome variants with the most severe RV involvement, while DSP variants and genotype-negative individuals demonstrated more LV or biventricular involvement.¹²²

Genetic testing can guide clinical management by confirmation of an ACM diagnosis versus an arrhythmic syndrome secondary to a cardiac channelopathy. For example, variants in PKP2 and TMEM43 are associated with exercise-induced disease progression and risk of SCA, therefore knowledge of genotype may help to guide decisions on sports participation in specific patients.^{123, 124}

Precision Medicine

Of key importance is understanding the background variation of ACM-associated genes to aid in determining variant pathogenicity, especially when variants are identified incidentally. One study found that the frequency of rare missense variants was similar in control and ACM patients (16% and 21%, respectively). Conversely, radical variants, such as indels, splice site variants, and nonsense variants, were far more common in ACM patients than healthy controls (0.5% and 43% respectively).¹⁰⁶ A screen of 61,019 patients who underwent ES identified 140 patients with predicted loss-of-function variants in ACM associated genes, however penetrance of the ACM phenotype was estimated to be only 6.0%.¹²⁵ Screening ES from 7,066 pediatric patients identified ACM-associated gene variants in 14% of patients, all of whom lacked clinical characteristics of ACM.¹²⁶ Notably, proven pathogenic variants localized to critical functional domains of ACM genes, while incidental variants did not.¹²⁶ There is overlap of genes associated with ACM and DCM, especially with LMNA variants which can cause an arrhythmic form of DCM.^{74, 76} Using precision medicine to understand the genomic architecture of ACM will allow us to better predict disease course in a given patient, including age of onset, predisposition to arrhythmias, degree of fibrofatty replacement, and ultimately develop disease-modifying treatment strategies.

Non-Compaction Cardiomyopathy

Clinical Diagnosis

Non-compaction cardiomyopathy (NCCM), which includes left ventricular non-compaction cardiomyopathy, is rare in children and the prevalence of NCCM in the adult population is estimated to be 0.05%.¹²⁷ NCCM is characterized by prominent trabecular meshwork and can be associated with both left ventricular dilation or hypertrophy as well as potentially loss of systolic function.^{2, 128, 129} Diagnosis of NCCM is typically made based on the Jenni and Peterson criteria.^{129, 130}

Genetics of NCCM

NCCM typically exhibits AD inheritance and shares associated sarcomere genes with the other cardiomyopathies, including MYH7, MYBPC3, TTN, TNNT2, and TNNC1, among others (Table 4).^{128, 131, 132} NCCM is also associated with the mitochondrial disease Barth syndrome, which is characterized by variants in TAZ and presents with skeletal and cardiac myopathy, neutropenia, and 3-methylglutaconic aciduria.¹³³ A study of 128 pediatric NCCM patients found 9% had underlying metabolic or syndromic etiologies.¹³¹ Given the genetic overlap with other cardiomyopathies and syndromic etiology in some cases, the combination of thorough clinical and genetic evaluations are particularly useful in NCCM.

Table 4 –

Genetic etiology of non-compaction cardiomyopathy

Gene	Protein	Chr	Mode of inheritance	Pediatric frequency	OMIM	Reference
ACTC1	α-Cardiac actin	15q14	AD	Rare	102540	128, 131, 132
ACTN2	Actinin, α 2	1q43	AD	Rare	102573	128, 131, 132
LDB3	LIM-binding domain 3	10q23	AD	Rare	605906	128, 131, 132
MYBPC3	Cardiac myosin-binding protein C	11p11	AD, AR	~8%	600958	128, 131, 132, 137
MYH7	β-Myosin heavy chain	14q11-q12	AD	~19%	160760	128, 131, 132, 137
TNNT2	Cardiac troponin T	1q32	AD	Rare	191045	128, 131, 132
TPM1	α-Tropomyosin	15q22	AD	Rare	191010	128, 131, 132
TTN	Titin	2q31	AD, AR	Rare	188840	128, 131, 132, 137
TAZ	Tafazzin	Zq28	XR	Rare	300394	133

AD, autosomal dominant; AR, autosomal recessive; XD, X-linked dominant; XL, X-linked; XR, X-linked recessive

Diagnostic and Risk-Predictive Genetic Testing

There is considerable overlap in genes associated with isolated NCCM and those associated with other cardiomyopathies, and as such, broad HCM and DCM panels are often used for diagnostic testing in pediatric patients presenting with NCCM.^{18, 24} One study screened 327 non-compaction cardiomyopathy patients for 45 cardiomyopathy genes and found causative variants in 32% of patients.¹³² Approximately 71% of detected variants were in MYH7, MYBPC3, and TTN.¹³² Notably, the genetic yield was significantly greater in children than in adults, at 44% and 30% respectively.¹³² In one study of pediatric patients specifically, the genetic yield of cardiomyopathy gene panel testing in one study was only 9%, and patients with concomitant cardiovascular malformation or HCM were most likely to have an LP/P variant, suggesting the etiology of isolated NCCM remains to be elucidated.¹³¹ Genetic testing is also useful to distinguish isolated NCCM from metabolic or mitochondrial causes such as Barth syndrome.¹³³ ES or GS are also options in evaluating NCCM; however, the increased VUS burden and genetic heterogeneity observed in NCCM with expansive testing makes interpretation challenging.¹³⁴

An estimated 32%−40% of pediatric NCCM patients have a positive family history.^{131, 132} Further, a study of 143 NCCM patients and 113 first- or second-degree relatives revealed 51% of probands had family members affected by NCCM, DCM, or HCM.¹²⁸ This study also revealed non-penetrance as 37% of genotype-positive kindred did not have evidence of cardiomyopathy on cardiac screening.¹²⁸ Importantly, a study of 2501 athletes identified prominent LV hypertrabeculation in ~1% of athletes, while only 0.1% of this cohort had overt NCCM, suggesting a spectrum of disease that overlaps with a normal spectrum of myocardial morphology.¹³⁵ Genetic screening paired with clinical evaluation of relatives can help distinguish potentially benign hypertrabeculation from NCCM disease in genotype-positive families.¹³⁵

Prognosis and Therapeutics

Emerging evidence suggests that genetic testing can play a role in prediction of prognosis and disease progression in NCCM. One study found 10% of pediatric NCCM patients had multiple variants and that variant burden was associated with worse LV function, earlier age of presentation, and increased risk of major adverse cardiac events.¹³² Further, patients with sarcomere gene variants were more likely to have a mixed HCM or DCM phenotype rather than isolated NCCM highlighting the role of continued echo monitoring of individuals who host sarcomeric gene variants in a family with NCCM.¹³⁶ Finally, a handful of studies describe differences in outcomes attributable to genotype. ES in 95 NCCM patients identified RBM20 variant carriers had poor prognosis, including need for heart transplant, presumably due to alterations in RBM20 regulated TTN-splicing.¹⁴¹ Similarly, a systematic review of 561 patients found that patients with MYBPC3 and TTN variants had greater risk of major adverse cardiac events (including stroke, heart failure, SCA/SCD, LVAD placement, or transplant) than patients with MYH7 variants.^{136, 137} To date, there are no genotype-guided therapies; however, if NCCM occurs with evidence of another cardiomyopathy phenotype, such as DCM or HCM, treating concomitant cardiomyopathy manifestations is warranted.¹³⁸

Precision Medicine

Precision medicine offers the ability to clarify the phenotypic and genetic heterogeneity inherent in NCCM, although actionable studies are still emerging. Understanding the basis of pathogenic versus benign hypertrabeculation is critical to understanding who may be at-risk for life-threatening disease. Further, given that genes associated with NCCM are also associated with other, more common cardiomyopathies, understanding what drives isolated versus complicated disease will aid in risk stratification. Given the genetic heterogeneity, ES and GS may be helpful in identifying new disease-associated genes and CNVs that may otherwise be missed by cardiomyopathy panel testing. A full understanding of the genetic architecture of NCCM holds the promise of defining the origin of the phenotypic variability seen in individuals and distinguishing between benign hypertrabeculation vs malignant NCCM disease.

Restrictive Cardiomyopathy

RCM is exceptionally rare in children and affects roughly 0.03 per 10,000 children.³

RCM is characterized by decreased myocardial compliance and diastolic dysfunction resulting in heart failure, atrial dilation, and arrhythmias.²⁴ Diagnosis of RCM in adults is primarily based on evidence of myocardial noncompliance and diastolic dysfunction on echocardiography and cardiac catheterization; however, there are no clear diagnostic guidelines for children.²⁴

Genetics of RCM

The genetic causes of pediatric-onset RCM are distinctly different from the causes of RCM that affect older adults, which include cardiac amyloidosis, sarcoidosis, and hemochromatosis.¹³⁹ Infantile RCM is often idiopathic, but is associated with sarcomere variants, including in ACTC1, MYH7, TNNI3, and TTN (Table 5).⁴ Sarcomere gene and DES sequencing in 12 pediatric patients with RCM identified pathogenic variants in 33%, namely in TNNT2, TNNI3, and ACTC1.¹⁴⁰ Importantly, pediatric RCM is distinctly different from transthyretin-cardiac amyloidosis observed in older adults, which is caused by TTR variants.^{141, 142} Although TTR variants are unlikely to cause pediatric-onset restrictive cardiomyopathy, identification of TTR variants in children can inform clinical follow-up in adulthood. Syndromic causes of pediatric RCM include early phenotypes of Anderson-Fabry disease, hemochromatosis, or hematologic disorders that require frequent transfusion.²⁴

Table 5 –

Genetic etiology of restrictive cardiomyopathy

Gene	Protein	Chr	Mode of inheritance	Pediatric frequency	OMIM	Reference
FLNC	Filamin C	7q32	AD	Rare	102565	146
MYBPC3	Cardiac myosin-binding protein C	11p11	AD, AR	Rare	600958	165
MYH7	β-Myosin heavy chain	14q11-q12	AD	Rare	160760	166
ACTC1	α-Cardiac actin	15q14	AD	Rare	102540	140
TNNC1	Cardiac troponin C	3p21	AD	Rare	191040	145
TNNI3	Cardiac troponin I	19p13	AD	Rare	191044	47,140
TNNT2	Cardiac troponin T	1q32	AD	Rare	191045	140, 143
TTN	Titin	2q31	AD, AR	Rare	188840	167

AD, autosomal dominant; AR, autosomal recessive; XD, X-linked dominant; XL, X-linked; XR, X-linked recessive

Diagnostic and Risk-Predictive Genetic Testing

Like in other cardiomyopathies, genetic testing should be performed to confirm the clinical diagnosis and evaluate whether or not causative variants are present. Current ACMG recommendations suggest using genes found in the HCM and DCM gene panels in initial genetic evaluation of RCM. As with DCM, ES and GS are also options for further evaluation.¹⁸ The estimated yield of causative variants in RCM varies widely from 10–60% for all ages, however a pediatric-specific estimate of genetic yield in RCM is not known.¹⁸ In a 12-patient pediatric RCM cohort, 33% of probands had family history of RCM or another cardiomyopathy.¹⁴⁰ As with the other cardiomyopathies, identification of a pathogenic variant or VUS in family members warrants continued clinical follow-up

Prognosis and Therapeutics

In general, pediatric-onset RCM has a poor prognosis, and prognostic genotype-phenotype correlations in RCM are poorly understood. Limited reports and mechanistic studies have implicated TNNT2 variants as a driver of RCM with arrhythmia and sudden death.¹⁴³ Further, variants localizing to other troponin macromolecular complex component genes, such as TNNI3 and TNNC1, have been associated with arrhythmic RCM phenotypes through AD and AR modes of inheritance.^{144, 145} Finally, FLNC missense variants are an example of a non-sarcomeric cause of familial, pediatric onset RCM.¹⁴⁶ Nevertheless, the relatively small patient population of pediatric RCM has slowed identification of genotype-phenotype correlations. Given the poor outcomes of pediatric-onset RCM, disease-modifying therapies targeted at the molecular cause of cardiac dysfunction and remodeling are sorely needed to improve prognosis in this vulnerable patient population. To date, there is no universal genotype-associated guidance for therapy.

Precision Medicine

ES and GS are well-suited to defining the underlying causes of RCM in young patients and offer the ability to look outside the canonical cardiomyopathy genes in DCM and HCM panels. One study used transcriptomics and pathway analysis to identify dysregulated alternative splicing patterns unique to RCM mediated by RBM20 which is typically associated with DCM development.¹⁴⁷ Finally, the use of patient-derived iPSC-CMs is a powerful tool in the study of any cardiomyopathy but will be particularly useful in delineating the pathogenic mechanisms of RCM given the limited patient population.

Future Directions in Pediatric Cardiomyopathy Genetic Testing

Although advances in genetic testing platforms have made gene testing widely accessible for pediatric cardiomyopathy patients and families, there are still significant gaps in our understanding of the genetic etiology of cardiomyopathy that limit the diagnostic and prognostic power of genetic testing. It is likely that extrinsic factors, exposures, or environmental triggers may exacerbate disease expression associated with genetic variants not typically sufficient to cause penetrant disease in isolation. Further, the potential role of non-coding regulatory regions, which either directly cause cardiomyopathy or influence disease penetrance or expressivity, is an exciting area with tremendous potential to inform precision medicine. Finally, combinatorial effects of rare and common variants in cardiomyopathy genes remain poorly understood but may explain familial and idiopathic disease in the absence of a putatively pathogenic variant. For example, the presence of multiple rare variants, regardless of pathogenicity classification, are associated with a more severe disease course.¹⁴ Thus, understanding the complete genomic architecture of cardiomyopathic disease, and defining the modulatory influence of both comorbidities and environment, represents the great challenge of applying genomics to the prediction of disease manifestation and progression.

VUS interpretation remains a challenge with significant ramifications for patients and families. When a VUS is identified in family members of an affected patient, they should be followed clinically as if the variant were disease-causing. Methods to improve variant classification and are developing. For example, a recent genome-saturation editing approach has been taken with BRCA variants, where screening 96.5% of possible single nucleotide variants in critical domains for functional effects accurately predicts pathogenicity in known pathogenic variants and generates comprehensive pathogenicity predictions for nearly all possible variants.¹⁴⁸ Further, leveraging CRISPR genome edited mouse and iPSC models to explore the pathologic influence of cardiomyopathy-associated variants promises to provide clarity. When resolved, the potential for genomics to influence precision medicine will be great.

ES and GS are particularly powerful diagnostic tools in the pediatric population. The Undiagnosed Disease Network frequently employs ES and GS to detect a disease-causing locus in 40% of patients, while trio-based GS of sick neonates generates diagnostic information in 45% of patients.^{149, 150} Though the utility of ES and GS relative to panel sequencing is debated for first-line testing of probands, these platforms should be further explored for use in patients with initial negative genetic tests. Most importantly, as we develop our understanding of the genetic causes of pediatric cardiomyopathy, the door opens for the development of better disease-modifying treatments. The development of small molecule myosin inhibitors or activators for HCM and DCM represent a significant advance in cardiomyopathy treatment, and application to other sarcomere and non-sarcomere causes is a needed step towards truly personalized medicine.

Conclusions

Genetic testing is a critical component of pediatric cardiomyopathy diagnosis and management. Beyond confirming clinical diagnosis and assessing risk of first-degree relatives, the implementation of genetic testing in clinical practice allows for better prognostic predictions and the possibility of disease-modifying treatments targeted at the underlying genetic cause of cardiomyopathy in children.

Highlights.

• Genetic testing is a key component of the diagnostic evaluation of pediatric cardiomyopathy

• Risk-predictive genetic testing of a variant found in a proband can identify at-risk family members

• Genotype-phenotype correlations illustrate the potential for genetic testing to inform prognosis

• Understanding genomic drivers opens the door to a precision medicine-based approach to care

Abbreviations:

ACM

arrhythmogenic cardiomyopathy

AD

autosomal dominant

AR

autosomal recessive

CNVs

copy number variants

DCM

dilated cardiomyopathy

ES

exome sequencing

GS

genome sequencing

HCM

hypertrophic cardiomyopathy

ICD

implantable cardiac defibrillator

iPSC-CMs

induced pluripotent stem cell cardiomyocytes

LV

left ventricle

NCCM

non-compaction cardiomyopathy

P/LP

pathogenic/likely pathogenic

RCM

restrictive cardiomyopathy

SCA

sudden cardiac arrest

SCD

sudden cardiac death

TTNtvs

truncating titin variants

VUS

variant of uncertain significance