Concealed Cardiomyopathy in Autopsy-Inconclusive Cases of Sudden Cardiac Death and Implications for Families
Original Investigation
Central Illustartion
Abstract
Background
Genetic testing following sudden cardiac death (SCD) is currently guided by autopsy findings, despite the inherent challenges of autopsy examination and mounting evidence that malignant arrhythmia may occur before structural changes in inherited cardiomyopathy, so-called “concealed cardiomyopathy” (CCM).
Objectives
The authors sought to identify the spectrum of genes implicated in autopsy-inconclusive SCD and describe the impact of identifying CCM on the ongoing care of SCD families.
Methods
Using a standardized framework for adjudication, autopsy-inconclusive SCD cases were identified as having a structurally normal heart or subdiagnostic findings of uncertain significance on autopsy. Genetic variants were classified for pathogenicity using the American College of Medical Genetics and Genomics guidelines. Family follow-up was performed where possible.
Results
Twenty disease-causing variants were identified among 91 autopsy-inconclusive SCD cases (mean age 25.4 ± 10.7 years) with a similar rate regardless of the presence or absence of subdiagnostic findings (25.5% vs 18.2%; P = 0.398). Cardiomyopathy-associated genes harbored 70% of clinically actionable variants and were overrepresented in cases with subdiagnostic structural changes at autopsy (79% vs 21%; P = 0.038). Nine of the 20 disease-causing variants identified were in genes implicated in arrhythmogenic cardiomyopathy. Nearly two-thirds of genotype-positive relatives had an observable phenotype either at initial assessment or subsequent follow-up, and 27 genotype-negative first-degree relatives were released from ongoing screening.
Conclusions
Phenotype-directed genetic testing following SCD risks under recognition of CCM. Comprehensive evaluation of the decedent should include assessment of genes implicated in cardiomyopathy in addition to primary arrhythmias to improve diagnosis of CCM and optimize care for families.
Introduction
Sudden cardiac death (SCD) is the tragic final outcome of a number of cardiac conditions. While ischemic heart disease is the most common cause of SCD in the general population, inherited cardiomyopathy (CM), such as hypertrophic cardiomyopathy (HCM) or inherited arrhythmia syndromes (IAS), such as long QT syndrome, are important causes in younger SCD victims.1,2 Establishing the cause of SCD is key to providing optimal care for surviving relatives, and where an inherited cardiac condition is identified or suspected, family screening and management of at-risk relatives aim to prevent further tragedy in the family.
Comprehensive autopsy assessment is the gold standard investigation following SCD, and findings guide the care of surviving relatives.3-6 However, surveys show that global practice falls short of this recommendation, with autopsy mandatory in only 30% of jurisdictions and actually performed in an average of 43% of cases following the sudden death of a young person.7,8 When autopsy is performed, it is important for clinicians, who will be using the results to inform care of surviving relatives, to appreciate the challenges and limitations of autopsy examination that may influence the cause of death reported. Differentiating nondiagnostic changes from pathological abnormalities that account for death can be difficult, and there are “degrees of certainty” that need to be considered when attributing causality.9 Cardiac parameters such as ventricular dilation and cardiomegaly are particularly prone to inconsistent assessment on autopsy due to a paucity of normative data to guide cutoffs for abnormality, as well as variable practice between labs in specimen preparation.10 Furthermore, specialist cardiac pathologist review is only undertaken in 30% to 40% of cases despite guideline recommendations and evidence that specialist review improves accuracy of autopsy diagnosis.7,11
Despite these significant limitations of autopsy assessment, guidelines for genetic testing in SCD victims are guided by autopsy findings. Where a diagnosis of a cardiomyopathy such as HCM or arrhythmogenic cardiomyopathy (ACM) is reported on autopsy, phenotype-directed genetic testing is recommended where a sample from the deceased is available.3 Where no cause of death is established at autopsy and no diagnostic clues are apparent from family screening, genetic analysis including IAS-associated genes is recommended, that is, the “molecular autopsy.”3,6,12 This limited approach to genetic assessment in those with a normal autopsy is based on the historical assumption that features of an inherited CM would be identified at autopsy. However, disease-causing variants in CM-associated genes have been reported in both SCD and resuscitated cardiac arrest cohorts with structurally normal hearts, and there is mounting evidence that malignant arrhythmia can occur before overt structural changes in inherited CM, referred to as “concealed cardiomyopathy” (CCM).13-16
In this study, we sought to identify the types of genes (CM- and IAS-associated) implicated in SCD where no definite cause of death is determined at autopsy. We compared the types of genes implicated in those with a structurally normal heart and those with subdiagnostic findings of uncertain significance at autopsy, and highlight the clinical implications of identifying cases of CCM for surviving relatives.
Methods
Study population
Study population
All SCD cases aged between 1 and 64 years where family had been referred to the genetic heart disease team at Royal Prince Alfred Hospital between 1997 and 2020 for family screening or consideration of genetic testing were screened for inclusion. SCD was defined as death within 1 hour of symptom onset, or within 24 hours of last being seen in good health for unwitnessed deaths.17 Cases where histology was not performed or genetic testing not undertaken, and those with a diagnostic abnormality at autopsy were excluded. Cases with a premorbid diagnosis were excluded from analysis. Only the proband was included in families with multiple SCDs. Data were collected with consent from the deceased’s next of kin and participants, according to institutional ethics approval.
Autopsy review
Structured, physician-led autopsy review was based on current guidelines using a standardized framework previously described by Raju et al18 to identify cases as having a normal autopsy, or subdiagnostic (uncertain) or diagnostic (pathological) autopsy findings.9 Borderline cases were independently adjudicated by a second physician. Adjusted predicted heart weight was calculated and actual heart weight stratified using an online calculator.19 Autopsies were considered inconclusive if there were subdiagnostic findings alone or a structurally normal heart.
Genetic analysis
The technology used reflected the best available at the time of family assessment and ranged from segregation testing of a variant identified in a family member, to commercial gene panel, and exome and genome sequencing (Supplemental Figure 1). Gene lists analyzed by patient are detailed in Supplemental Table 1. Detailed genetic methods can be found in the data supplement (Supplemental Methods). Evidence for variant pathogenicity was reviewed by a multidisciplinary team comprising genetic counselors, cardiologists, and clinical geneticists and scientists, and variant classification was determined using the American College of Medical Genetics and Genomics (ACMG) criteria with adaptations to key criteria as previously described.20-23 Variants meeting likely pathogenic or pathogenic ACMG criteria were considered disease-causing and thus clinically actionable. Rare variants of uncertain significance (VUS) were defined as variants with a minor allele frequency of ≤0.004% in the Genome Aggregation Database (gnomAD version 2.1.1) and were only reported in patients who had research-based testing with raw data available for analysis.21 Nontruncating variants in TTN and intronic variants (with the exception of canonical splice site variants) were excluded from analysis.
Statistical analysis
SAS Studio (SAS Institute) was used for statistical analysis. Descriptive statistics were used to determine the characteristics of the cohort with continuous variables reported as means and standard deviations and categorical variables as counts and percentages of available data. Categorical variables were compared using chi-square tests or Fisher exact test as appropriate, and continuous variables were compared using Student’s t-test. Statistical significance was defined as P < 0.05. The data underlying this article will be shared on reasonable request to the corresponding author.
Results
Cohort characteristics
Cohort characteristics
Ninety-one cases of autopsy-inconclusive SCD were identified, after excluding 35 cases where a definitive diagnosis was identified at autopsy and 78 cases with no sample available for genetic testing. Forty-seven cases (51.6%) had subdiagnostic abnormalities on autopsy, whereas 44 (48.4%) had a structurally normal heart. The characteristics of the cohort are detailed in Table 1. The mean age at time of death was 25.4 ± 10.7 years (Supplemental Figure 2), and over two-thirds of decedents were male (n = 65, 71.4%). In the majority of cases with available data, SCD events occurred in the home (n = 62, 68.1%), or during rest or sleep (n = 55, 68.8%). Only 8 cases (9.3%) had a family history of SCD, and 10 (11.9%) had a known history of syncope before death. There were no demographic differences between victims with a normal heart and those with subdiagnostic findings at autopsy. Twenty-eight cases were part of a cohort described previously.2
| Total | Normal Autopsy | Subdiagnostic Autopsy Findings | P Value | |
|---|---|---|---|---|
| Patients | 91 (100) | 44 (48.4) | 47 (51.6) | |
| Male | 65 (71.4) | 32 (72.7) | 33 (70.2) | 0.791 |
| Age <35 y | 70 (76.9) | 36 (72.3) | 34 (81.8) | 0.283 |
| Family history of SCDa | 8 (9.3) | 2 (4.9) | 6 (13.3) | 0.250 |
| History of syncopea | 10 (11.9) | 6 (14.6) | 4 (9.3) | 0.451 |
| Periexertionala | 25 (31.3) | 13 (34.2) | 12 (28.6) | 0.578 |
Of patients with subdiagnostic findings at autopsy, 38.3% (n = 18) had a single subdiagnostic structural change, 29.8% (n = 14) had 2 subdiagnostic findings, whereas 31.9% (n = 15) had 3 or 4 subdiagnostic abnormalities identified. Left ventricular fibrosis was the most common subdiagnostic finding, seen in 22.0% (n = 20) of cases. Cardiomegaly was present in 16.5% of cases (n = 15), whereas left ventricular hypertrophy (n = 11) and right ventricular fat (n = 11) were each seen in 12.1% of cases (Supplemental Figure 3).
Genetic testing
Twenty disease-causing variants were identified among the 91 SCD victims (Table 2). The majority of disease-causing variants resided in CM-associated genes (70%, n = 14) and only one-quarter in IAS-associated genes (25%, n = 5). One pathogenic variant was identified in COL3A1, a definitive vascular-type Ehlers-Danlos syndrome gene, but no features of macrovascular dissection or aneurysm were found on autopsy. Disease-causing variants were identified more commonly in females (38.5% vs 15.4%; P = 0.016) (Supplemental Table 2), but there was no significant difference in rate of disease-causing variants based on age at death (<35 years = 24.3% compared with 14.3% ≥35 years; P = 0.332), activity at time of death (28.0% periexertional vs 18.2% rest/sleep; P = 0.320), or known history of syncope before death (10.3% vs 18.8%; P = 0.347). Although not statistically significant, disease-causing variants were identified at nearly twice the rate in those with left ventricular fibrosis compared with those without (35.0% vs 18.3%; P = 0.069).
| Code | Sex | Premorbid Clinical Data | Autopsy | Subdiagnostic Abnormalities | Genetic Testing Type | Gene (Associated Condition) | Genomic Position (GRCh38) | Transcript: cDNA | Protein | Consequence | gnomAD Allele Frequency | Variant Classification (Criteria Used) | Genotype-Positive First-Degree Relatives | Case Published Previously |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MG1 | Male | Syncope while walking 3 weeks before SCD. No premorbid investigations performed (awaiting assessment at time of death). | Subdiagnostic | Cardiomegaly, Hypertrophy | Exome | RIT1 (HCM, Noonan syndrome) | chr1:155904739C>T | NM_006912.5: c.229G>A | p.Ala77Thr | Missense | Absent | Pathogenic (PS2, PS3_supporting, PS4, PM1, PM2, PP3) | De novo variant | No |
| QW1 | Male | Presyncope on exertion 2 days before death. ECG: left anterior hemiblock and incomplete right bundle branch block. | Subdiagnostic | Cardiomegaly, LV dilation | Exome | PKP2 (ACM) | chr12:32796207G>T | NM_004572.4: c.2391C>A | p.Tyr797Ter | Nonsense | Absent | Pathogenic (PVS1, PM2, PM6) | De novo variant | Yes39 |
| QF6 | Male | No premorbid cardiac symptoms or investigations. | Subdiagnostic | LV fibrosis | Segregation | PKP2 (ACM) | chr12:32802500_32802505delinsC | NM_004572.4: c.2197_2202CACACC>G | p.His733Alafs∗8 | Frameshift | Absent | Pathogenic (PVS1, PS4, PP1_strong, PM2) | 2: G+P+ and G+P? (Figure 2C) | No |
| RJ1 | Female | Possible palpitations, time course unclear and no investigations performed prior to death. | Subdiagnostic | LV/RV Fibrosis, RV fat | Exome | DSP (ACM) | chr6:7579323C>T | NM_004415.4: c.3133C>T | p.Arg1045Ter | Nonsense | Absent | Pathogenic (PVS1, PM2, PS4_moderate) | 1: G+P− | No |
| RM1 | Female | No premorbid cardiac symptoms or investigations. | Normal | NA | Gene panel | KCNH2 (LQTS) | chr7:150948995G>A | NM_000238.3: c.2453C>T | p.Ser818Leu | Missense | 1/251432 | Pathogenic (PP1_strong, PS3_moderate, PM2, PS4_moderate, PM6, PP3) | 4: 2 × G+P+, 1 × G+P− and 1 × G+P? | Yes40 |
| RO1 | Male | Presyncope for 1 week before death Known mild mitral valve prolapse with mild mitral valve regurgitation. | Subdiagnostic | Mitral valve prolapse, LV fibrosis | Exome | TTN (DCM) | chr2:178547522_178547526del | NM_001256850.1: c.89180_89184CTTAAA>C | p.Ile29727Serfs∗34 | Frameshift | 2/248122 | Likely pathogenic (PVS1_strong, PM2, PS4_moderate) | 2: G+P+, G+P− (Figure 2A) | No |
| ALB2 | Female | No premorbid cardiac symptoms or investigations. | Normal | NA | Segregation | ACTN2 (HCM, DCM) | chr1:236719007G>A | NM_001103.3: c.355G>A | p.Ala119Thr | Missense | Absent | Likely pathogenic (PP1_strong, PM2, PS4_supporting, PP3) | 1: G+P+ (Figure 2C) | Yes41 |
| APG1 | Female | No premorbid cardiac symptoms or investigations. | Subdiagnostic | RV fibrosis, RV fat | Exome | KCNH2 (LQTS) | chr7:149981783-150777008del | a | a | Whole gene deletion | a | Pathogenic 1A, 2A, 3Ab | Family declined genetic testing | Yes42 |
| ARH1 | Male | No premorbid cardiac symptoms or investigations. | Subdiagnostic | LV disarray, fibrosis | Exome | MYH7 (HCM) | chr14:23426833C>T | NM_000257.2: c.1988G>A | p.Arg663His | Missense | 4/282842 | Pathogenic (PS3, PS4, PP1_strong, PM2) | Lost to follow-up | Yes2 |
| AWK1 | Male | No premorbid cardiac symptoms or investigations. | Subdiagnostic | Inflammation | Gene panel | TTN (DCM) | chr2:178616508G>A | NM_001256850.1: c.43360C>T | p.Arg14454Ter | Nonsense | 1/247622 | Likely pathogenic (PVS1_strong, PM2, PP1_moderate, PS4_supporting) | Family assessment in progress | No |
| AYD1 | Male | No premorbid cardiac symptoms or investigations. | Normal | NA | Gene panel | PKP2 (ACM) | chr12:32879001_32879004del | NM_004572.4: c.253_256TGAGT>T | p.Glu85Metfs∗26 | Frameshift | 1/251410 | Likely pathogenic (PVS1, PM2, PS4_supporting) | Family assessment in progress | Yes2,39 |
| WH1 | Male | Sore throat and flu-like illness for 3 days before SCD. No premorbid cardiac symptoms or investigations. | Normal | NA | Gene panel | SCN5A (BrS, DCM) | chr3:38551447C>T | NM_198056.2: c.4925G>A | p.Gly1642Glu | Missense | Absent | Likely pathogenic (PM2, PP1, PP3, PS4_moderate) | 2: G+P+, G+P− | No |
| AAN2 | Female | Nonspecifically unwell for the month before death. No premorbid cardiac history or investigations. | Subdiagnostic | Inflammation, LV fibrosis | Segregation | DSP (ACM) | chr6:7559281C>T | NM_004415.3: c.478C>T | p.Arg160Ter | Nonsense | Absent | Pathogenic (PVS1, PM2, PS4_moderate) | 2: G+P+, G+P− | No |
| BFC11 | Female | Mother and grandmother had diagnosis of CPVT. Clinical screening in progress. Normal ECG and echocardiogram 2 weeks before death. Awaiting stress test at time of death. | Normal | NA | Segregation | CASQ2 (CPVT) | chr1:115732968T>C | NM_001232.3: c.539A>G | p.Lys180Arg | Missense | Absent | Pathogenic (PS3, PP1_strong, PM2, PP3) | 1: G+P+ | Yes43 |
| BNL1 | Female | No premorbid cardiac history or investigations. Autopsy-indeterminate SCD in mother aged 42 y. | Subdiagnostic | Cardiomegaly | Exome | FLNC (HCM, DCM, ACM) | chr7:128846336_128846351del | NM_001458.5: c.4000_4015TCGCTGTGCCCAAGAGC>T | p.Ala1334Profs∗6 | Frameshift | Absent | Likely pathogenic (PVS1_strong, PM2) | 1: G+P− | No |
| CDC1 | Male | History of recurrent hydropneumothorax. No history of syncope. Normal ECG, echocardiogram, Holter monitor, stress test. | Normal | NA | Exome | COL3A1 (Ehlers-Danlos) | chr2:188988099G>A | NM_000090.3: c.547G>A | p.Gly183Ser | Missense | Absent | Pathogenic (PS2, PS3, PS4, PM2, PP2, PP3) | De novo variant | No |
| CHC1 | Female | Recent onset dizziness, awaiting assessment at time of death. No premorbid cardiac investigations. | Subdiagnostic | LV dilation, fibrosis, RV fat | Exome | FLNC (HCM, DCM, ACM) | chr7:128848981_128848982insACGTCACA | NM_001458.5: c.4926_4927T>TCGTCACAA | p.Val1643Thrfs∗26 | Frameshift | 1/243620 | Likely pathogenic (PVS1_strong, PM2, PS4_supporting) | 1: G+P+ (Figure 2C) | No |
| CHT1 | Male | No premorbid cardiac symptoms or investigations. Father had unexplained cardiac arrest 9 years prior to the death of CHT1. | Subdiagnostic | LV fibrosis | Segregation | FLNC (HCM, DCM, ACM) | chr7:128835418G>T | NM_001458.5: c.445G>T | p.Glu149Ter | Nonsense | Absent | Likely pathogenic (PVS1_strong, PM2) | 1: G+P+ | No |
| CBT4 | Female | No premorbid cardiac history or investigations. | Normal | NA | Segregation | NKX2.5 (ACM) | chr5:173232865_173232868del | NM_004387.4: c.677_680delGCCAT>G | p.Asp226Alafs∗5 | Frameshift | Absent | Pathogenic (PVS1, PM1, PM2) | 1: G+P+ | No |
| CLH2 | Female | Syncopal episodes 4 days and 4 years before death. Premorbid ECG not available. | Normal | NA | Segregation | KCNH2 (LQTS) | chr7:150958402del | NM_000238.3: c.577GG>G | p.Ala193Profs∗8 | Frameshift | Absent | Likely pathogenic (PVS1, PM2) | 1: G+P+ | No |
The yield of genetic testing was similar between SCD victims with subdiagnostic structural changes at autopsy and those with a structurally normal heart at autopsy (25.5% vs 18.2%; P = 0.398). However, the type of genes harboring disease-causing variants were different between the groups. Disease-causing variants in CM-associated genes were seen at nearly 4 times the rate in those with subdiagnostic structural abnormalities than those with a normal autopsy (79% vs 21%, n = 11 vs 3; P = 0.038) (Figure 1A). Indeed 11 of 12 clinically actionable variants in SCD cases with subdiagnostic findings at autopsy were in CM-associated genes. Notably, the majority of CM-associated genes with disease-causing variants (n = 9, 64.3%), are implicated in ACM (Figure 1B).
Concealed CM in SCD With Subdiagnostic Abnormalities at Autopsy
(A) Disease-causing variants in cardiomyopathy (CM)-associated genes (blue) were more common in sudden cardiac death (SCD) victims with subdiagnostic findings at autopsy than in those with structurally normal hearts (79% vs 21%; P = 0.038). (B) The majority of disease-causing variants in the cohort were identified in CM-associated genes (blue) with a predominance of genes implicated in arrhythmogenic cardiomyopathy, denoted by an asterisk. †COL3A1 is associated with vascular Ehlers-Danlos syndrome and not associated with primary cardiomyopathy or arrhythmia. IAS = inherited arrhythmia syndrome.
Cascade testing in first-degree relatives
There was a total of 84 first-degree relatives (78 living) among the 20 SCD probands with a disease-causing variant. All families were offered cascade testing in conjunction with clinical assessment, and the results of testing were available in 47 individuals. Twenty (42.6%, n = 20/47) of first-degree relatives who underwent genetic testing were positive for the familial variant, and 13 (65%) of these genotype-positive patients had an observable phenotype. The remaining 27 genotype-negative first-degree relatives were released from follow-up (Central Illustration). Of the 14 families where the decedent had CCM, clinical information was available in 12 genotype-positive first-degree relatives from 9 families (2 cases were de novo, and no family follow-up information was available in 3 families) and is outlined in Table 2. Pedigrees from 3 of these families are shown in Figure 2. When extended relatives of decedents with CCM were considered, 20 genotype-positive relatives were identified, 10 of whom demonstrated a concordant phenotype, and 47 individuals who did not carry the family variant were able to be reassured and released from follow-up.
Identification of Concealed Cardiomyopathy Following Autopsy-Inconclusive SCD Guides Management of Families
Disease-causing variants in cardiomyopathy (CM)-associated genes are identified in autopsy-inconclusive sudden cardiac death (SCD) victims, more commonly in those with subdiagnostic autopsy findings. These variants are useful in guiding the management of surviving relatives. IAS = inherited arrhythmia syndrome.
Pedigrees of 3 Concealed Cardiomyopathy Families
Pedigrees of 3 families where decedent died of concealed cardiomyopathy. (A) A family (of proband RO1) where the decedent had a degree of myxoid degeneration of the mitral valve and nonspecific “patchy minor interstitial and perivascular fibrosis” at autopsy with a pathogenic TTN truncating variant on autopsy genetic testing. This variant segregated to the decedent’s mother, who, despite normal initial clinical screening, was found to have a dilated cardiomyopathy with left ventricular ejection fraction of 40% on follow-up. (B) The ALB and CHC families are joined by nonconsanguineous marriage. The genotype-positive siblings of both decedents have subtle structural changes with left ventricular noncompaction but normal left ventricular function in II:3 (ACTN2) and a small area of late gadolinium enhancement but normal left ventricular function on cardiac magnetic resonance (CMR) imaging in II:2 (FLNC). (C) Pedigree of sudden cardiac death (SCD) victim QF6. The disease-causing PKP2 variant was only identified after the SCD proband’s sister (II:1) was diagnosed with arrhythmogenic right ventricular cardiomyopathy 4 years following her brother’s death after presenting with conscious ventricular tachycardia while playing sport. ARVC = arrhythmogenic right ventricular cardiomyopathy; DCM = dilated cardiomyopathy; LVNC = left ventricular noncompaction.
Seven first-degree relatives across 5 families did not have cascade testing after the inheritance pattern was established (eg, de novo disease or variant identified in one parent so other not tested, and so on), testing was ongoing in 4 families (17 first-degree relatives), testing was not performed or not known to have been performed in 11 first-degree relatives (due to lack of sample in deceased relative or loss to follow-up). Testing was deferred in 2 phenotype-negative children.
Variants of uncertain significance
Forty-seven of the 63 decedents (74.6%) who underwent research-based testing had at least 1 VUS (98 individual variants) (Supplemental Table 3), but this decreased to 26 decedents (41.3%) when only variants in genes known to be associated with CM or channelopathy (Supplemental Table 4) were considered. The mean number of VUS in literature-based genes per patient was 0.54 ± 0.74 (range 0 to 3).
Discussion
The majority of clinically actionable genetic variants in this cohort of autopsy-inconclusive SCD victims were identified in CM-associated genes. Historically, genetic testing has been limited to IAS-associated genes in cases of unexplained SCD. We show that disease-causing variants in CM-associated genes are more commonly identified in individuals with subdiagnostic findings on autopsy than those with a structurally normal heart. Identification of CCM impacts follow-up assessment of SCD families, aiding diagnosis and targeting clinical management. These results suggest that CCM should be considered in all SCD victims where no definitive cause of death is identified at autopsy and supports the use of multiphenotype genetic testing, analyzing genes with known disease association in CM or IAS, as part of the comprehensive assessment of SCD probands in conjunction with, rather than dependent on autopsy examination.
CCM in autopsy-indeterminate SCD
Fourteen of 91 autopsy-indeterminate SCD cases (15.4%) in the present cohort are considered to have died from CCM due to the presence of a disease-causing variant in a CM-associated gene. Several groups have identified disease-causing variants in CM-associated genes in autopsy-inconclusive SCD victims. In a prospective SCD cohort previously reported by our group, 20 disease-causing variants were identified among 113 unexplained SCD victims. Interestingly, 7 of the disease-causing variants in the current cohort were in genes not analyzed in the previous study (FLNC, TTN, RIT1, NKX2.5).2 More recently, Neves et al13 published a cohort including unexplained SCD victims who remained genotype-negative following guideline-directed, IAS-focused genetic testing, and 8 (12%) cases had a clinically actionable variant in a CM-associated gene (PKP2, DSP, TTN, MYBPC3, MYH7, BAG3) despite no apparent phenotype on autopsy. Lahrouchi et al14 reported 6 variants in CM-associated genes (TTN, PLN, PKP2, MYH7) of 40 clinically actionable variants in a large cohort of SCD victims with structurally normal hearts. Interestingly, only 1 case of CCM (likely pathogenic truncating variant in TTN in a case with isolated left ventricular fibrosis at autopsy) was reported in a cohort of 29 SCD victims with autopsy findings of uncertain significance by the same group.24 The significantly lower rate of CCM reported by the latter work may be explained by the meticulous and rigorous autopsy review. All cases were reviewed by 2 expert pathologists, and findings were then confirmed by an expert cardiac pathologist. It should be noted that CCM has also been described in clinically idiopathic sudden cardiac arrest cohorts.13,16,25
Although CM-associated variants are seen across the spectrum of autopsy-inconclusive SCD, we show for the first time in a combined cohort that CCM appears to be more common in SCD victims with subdiagnostic structural changes at autopsy than in those with a normal heart. Given the extent of structural abnormalities (ejection fraction, wall thickness, and so on) are considered to indicate risk of SCD events in inherited CM, such as HCM and ACM,26,27 we hypothesis that subdiagnostic findings at autopsy may represent an intermediate phase between a structurally normal heart with potentially lower risk, and an overt phenotype sufficient for diagnosis with increased risk of malignant arrhythmia.
Arrhythmogenic cardiomyopathy (ACM) associated genes in autopsy-inconclusive SCD
Nearly two-thirds of the CCM cases in this autopsy-inconclusive SCD cohort involved genes implicated in ACM. Although the historical term arrhythmogenic right ventricular cardiomyopathy reminds us that the pathognomonic changes were first identified in the right ventricle, we now know that the left ventricle is frequently involved in ACM. In a large autopsy-based ACM series, 87% of cases had left ventricular involvement, and although the majority had biventricular involvement, more cases demonstrated isolated left ventricular disease than isolated right ventricular disease (17% vs 13%).28 Left ventricular changes have also been shown to precede right ventricular abnormalities in vivo, on electrocardiogram and cardiac magnetic resonance imaging, and clinical guidelines have recently been proposed to improve diagnosis of left ventricular–dominant ACM.29,30 Current autopsy guidelines underappreciate left ventricular involvement in ACM, and indeed, left ventricular fat infiltration is not considered when attributing cause of death.9 This anachronism will lead to ACM cases being considered autopsy-indeterminate with subdiagnostic findings.
Ventricular fibrosis and genetically defined CM
Ventricular fibrosis is the common substrate for malignant arrhythmia in CM, and disease-causing variants have been identified in up to 10% of SCD victims with “isolated” ventricular fibrosis at autopsy.1 Neves et al13 have recently demonstrated that the presence of left ventricular fibrosis significantly increases the yield of genetic testing in autopsy-inconclusive SCD victims to 36%. We saw a similar rate in the current study (35%). The difference in reported yields can likely be accounted for by the autopsy findings of the cohorts; Junttila et al1 only included cases with isolated fibrosis and excluded patients with a heart weight >420 g (regardless of sex and body weight), whereas the other studies included patients with other uncertain findings. Ventricular fibrosis is a nonspecific finding of a number of cardiac conditions, but there is growing recognition that some patterns may point to genetically defined cardiomyopathies, for example, DSP/FLNC CM.31 These conditions can be challenging to diagnose phenotypically and provide further justification for considering comprehensive autopsy assessment to include genetic testing.
IAS-associated genes in SCD victims with subdiagnostic structural findings at autopsy
Our findings parallel previous studies showing that disease-causing variants in IAS-associated genes are very rare in those with subdiagnostic structural changes at autopsy.1,15,24 Only 1 case with subdiagnostic findings at autopsy had a disease-causing variant in an IAS-associated gene (APG1). It is important to highlight that although the yield of disease-causing IAS-associated variants in those with subtle structural changes at autopsy is low, IAS remain an important cause of death in this group. Papadakis et al11 demonstrated that IAS are identified in up to 50% of families where the SCD victim had subdiagnostic findings at autopsy. Interestingly, family screening revealed a diagnosis of Brugada syndrome (BrS) in two-thirds of families where a diagnosis was made (14/21, 67%).11 Although BrS was historically considered a channelopathy, there is growing evidence of structural changes in the condition, suggesting it may in fact be a focal CM.32,33 Furthermore, the genetic architecture of BrS is complex, and a monogenic cause of disease is identified in only around 20% of cases. Together, these findings should break the assumption that individuals with primary arrhythmic syndromes would have structurally normal hearts but reminds us that although the rate of IAS variants in SCD victims with subtle structural abnormalities is low, clinical family screening is vital.
Potential sex differences
Although low numbers precluded multivariable analysis, it was interesting to note the apparent increased yield of genetic testing in females, despite no significant difference between the sexes in age, autopsy findings, activity at time of death, presence of ventricular fibrosis, or type of gene implicated. We hypothesize that this is partly due to the higher event rate in long QT syndrome in females34 (notably all 3 SCD victims with disease-causing KCNH2 variants in this cohort were female) but also to the potentially more subtle structural changes in females. In this study, 60% of clinically actionable variants in women were identified in CM-associated genes, most of whom had subdiagnostic findings at autopsy. Males have been shown to have more significant structural abnormalities in some inherited CMs than women. In HCM, men have increased left ventricular wall thickness and cavity size compared with women, and in TTN-related dilated CM, males are more likely to have reduced left ventricular ejection fraction and increased cavity dimensions at diagnosis compared with females.35,36 Given these differences, it is possible that the autopsy findings of inherited CM may be more subtle in females and not reach diagnostic thresholds, explaining why women are more likely to have no cause of death identified at autopsy.2
Importance of identifying CCM for family screening following SCD
Two of the families highlighted in Figure 2 had first-degree relatives who were phenotype negative at baseline assessment and developed clinical evidence of disease years after the SCD proband’s death (2a, I:2 and 2c, II:1). Clinical screening aims to prevent further SCD events in families by identifying at-risk first-degree relatives where a diagnosis of a potentially heritable condition is made in the proband, or the heart is structurally normal on autopsy. However, genetic conditions demonstrate variable expression, and because an initial negative clinical screen does not exclude disease, periodic follow-up is recommended.3,4 Age-related penetrance of inherited CM is well described, and emergence of clinical abnormalities over time has been seen in patients with an initial diagnosis of idiopathic ventricular fibrillation, where 21% of patients developed an observable phenotype during follow-up.37 Incomplete penetrance and variable expression are also described in family screening following SCD.15
Identifying a causative variant in a SCD family alters the care of surviving relatives, both genotype-positive and genotype-negative. Lifestyle modifications may be important in helping reduce risk and delaying or preventing the emergence of a clinical phenotype. Given the predominance of ACM-related variants in this cohort, exercise restriction would be an important intervention for these genotype-positive relatives (Figure 2C). Directed investigations beyond the recommended electrocardiogram and echocardiogram may uncover subtle phenotypes (Figure 2B). Identifying the diagnosis can aid care in complex families such as that shown in Figure 2B where CCM had caused SCD events in both maternal and paternal lineages due to 2 unrelated variants (FLNC and ACTN2). Individuals II:2 and II:3 have mild, nonspecific phenotypes that would be of limited use in guiding screening for their children, but the knowledge of the familial variants has allowed for focused care and close follow-up (III:2) or release from screening (III:1). Implantable cardioverter-defibrillators are generally not recommended for genotype-positive, phenotype-negative individuals, but further work is needed to understand the risk to this cohort, particularly in highly arrhythmogenic genetically defined CM such as LMNA-CM, RBM-20-CM, and DSP-CM. Identifying a disease-causing genetic variant can provide reassurance to genotype-negative relatives who can be released from ongoing clinical screening. The care of SCD families is best managed in specialized multidisciplinary clinics with detailed phenotyping, rigorous assessment of variant pathogenicity, careful management of VUS, and periodic reassessment of variants.38 Investigating genes with the potential to cause SCD is important in the research setting to identify new genetic culprits and potential risk alleles, but only genes with an established association with an inherited cardiovascular pathology should be analyzed in the clinical setting following SCD. This approach is paramount if broader testing of cardiovascular genes following autopsy-inconclusive SCD is undertaken to ensure optimal care of surviving relatives.
Study limitations
The variable genetic testing methods used in this retrospective study reflect the evolution of our approach to genetic testing over time. Given that only cases who had genetic testing were included, this is a selected unexplained SCD cohort, at risk of ascertainment bias, and results should not be considered a “yield” of genetic testing autopsy-inconclusive SCD. Although the inclusion of cases where only limited genetic panels were tested may underestimate the results of genetic testing by missing variants in genes not analyzed, inclusion of cases with segregation testing likely enriched the cohort for inherited diseases. This approach was taken to maximize our understanding of types of genes implicated in autopsy-inconclusive SCD. The size of the cohort precluded multivariable analysis. Autopsy findings were not reviewed by a specialist cardiac pathologist, but physician-led autopsy review reflects real-world management of SCD families.
Conclusions
CCM is an important but under-recognized cause of death in autopsy-inconclusive SCD, particularly where subtle, subdiagnostic changes are seen at autopsy. Disease-causing variants in ACM-associated genes are a common cause of CCM in SCD and identification of these variants can alter management of surviving relatives. Comprehensive proband assessment, including genetic testing of evidence-based CM and IAS-associated genes, may increase diagnosis and improve care of living relatives in SCD families.
Perspectives
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Victims of sudden cardiac death in whom autopsy fails to reveal a definitive cause may have disease-causing genetic variants associated with cardiomyopathy or arrhythmic syndromes. Identification of these variants coupled with genetic counselling can improve the care of living relatives.
TRANSLATIONAL OUTLOOK: Longer-term follow-up of family members of victims of sudden cardiac death may identify risk factors for disease development in genotype-positive, phenotype-negative individuals, informing prognosis and improving diagnosis and management strategies.
Funding Support and Author Disclosures
This work was supported by the National Health and Medical Research Council grants #102568 to L. Yeates, #1162929 to Dr Ingles, #2003997 and #1154992 to Dr Semsarian, National Heart Foundation of Australia grant #191351 to Ms Yeates, and NSW Health grants to Drs Semsarian and Bagnall. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Abbreviations and Acronyms
| ACM | arrhythmogenic cardiomyopathy |
| BrS | Brugada syndrome |
| CCM | concealed cardiomyopathy |
| CM | cardiomyopathy |
| HCM | hypertrophic cardiomyopathy |
| IAS | inherited arrhythmia syndrome |
| SCD | sudden cardiac death |
| VUS | variant of uncertain significance |
References
1. "Primary myocardial fibrosis as an alternative phenotype pathway of inherited cardiac structural disorders". Circulation 2018;137:2716-2726.
2. "A prospective study of sudden cardiac death among children and young adults". N Engl J Med 2016;374:2441-2452.
3. "2020 APHRS/HRS expert consensus statement on the investigation of decedents with sudden unexplained death and patients with sudden cardiac arrest, and of their families". Heart Rhythm 2021;18:1: e1-e50. https://doi.org/10.1016/j.hrthm.2020.10.010.
4. "2015 ESC guidelines for the management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: the Task Force for the Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death of the ESC". Eur Heart J 2015;36:2793-2867.
5. "European recommendations integrating genetic testing into multidisciplinary management of sudden cardiac death". Eur J Hum Genet 2019;27:1763-1773.
6. "European Heart Rhythm Association (EHRA)/Heart Rhythm Society (HRS)/Asia Pacific Heart Rhythm Society (APHRS)/Latin American Heart Rhythm Society (LAHRS) expert consensus statement on the state of genetic testing for cardiac diseases". Europace 2022;24:8: 1307-1367. https://doi.org/10.1093/europace/euac030.
7. "Investigation on Sudden Unexpected Death in the Young (SUDY) in Europe: results of the European Heart Rhythm Association Survey". Europace 2022;24:2: 331-339. https://doi.org/10.1093/europace/euab176.
8. "Global approaches to cardiogenetic evaluation after sudden cardiac death in the young: A survey among health care professionals". Heart Rhythm 2021;18:1637-1644.
9. "Guidelines for autopsy investigation of sudden cardiac death: 2017 update from the Association for European Cardiovascular Pathology". Virchows Arch 2017;471:691-705.
10. "Prevalence of abnormal heart weight after sudden death in people younger than 40 years of age". J Am Heart Assoc 2020;9:e015699.
11. "Sudden cardiac death with autopsy findings of uncertain significance: potential for erroneous interpretation". Circ Arrhythm Electrophysiol 2013;6:588-596.
12. "Sudden cardiac death in the young: the molecular autopsy and a practical approach to surviving relatives". Eur Heart J 2015;36:1290-1296.
13. "Exome sequencing highlights a potential role for concealed cardiomyopathies in youthful sudden cardiac death". Circ Genom Precis Med 2022;15:1: e003497 https://doi.org/10.1161/CIRCGEN.121.003497.
14. "Utility of post-mortem genetic testing in cases of sudden arrhythmic death syndrome". J Am Coll Cardiol 2017;69:2134-2145.
15. "Genomic autopsy of sudden deaths in young individuals". JAMA Cardiol 2021;6:11: 1247-1256. https://doi.org/10.1001/jamacardio.2021.2789.
16. ""Concealed cardiomyopathy" as a cause of previously unexplained sudden cardiac arrest". Int J Cardiol 2021;324:96-101.
17. "Sudden cardiac death". Circulation 1998;98:2334-2351.
18. "Insights into sudden cardiac death: exploring the potential relevance of non-diagnostic autopsy findings". Eur Heart J 2019;40:831-838.
19. "Postmortem heart weight modelled using piecewise linear regression in 27,645 medicolegal autopsy cases". Forensic Sci Int 2015;252:157-162.
20. "Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the ACMG and the Association for Molecular Pathology". Genet Med 2015;17:405-424.
21. "Adaptation and validation of the ACMG/AMP variant classification framework for MYH7-associated inherited cardiomyopathies: recommendations by ClinGen's Inherited Cardiomyopathy Expert Panel". Genet Med 2018;20:351-359.
22. "Technical standards for the interpretation and reporting of constitutional copy-number variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen)". Genet Med 2020;22:245-257.
23. "Variant interpretation for dilated cardiomyopathy: refinement of the ACMG/ClinGen guidelines for the DCM Precision Medicine Study". Circ Genom Precis Med 2020;13:e002480.
24. "The yield of postmortem genetic testing in sudden death cases with structural findings at autopsy". Eur J Hum Genet 2020;28:17-22.
25. "Importance of genetic testing in unexplained cardiac arrest". Eur Heart J 2022;43:32: 3071-3081. https://doi.org/10.1093/eurheartj/ehac145.
26. "2020 AHA/ACC guideline for the diagnosis and treatment of patients with hypertrophic cardiomyopathy: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines". J Am Coll Cardiol 2020;76:25: e159-e240. https://doi.org/10.1016/j.jacc.2020.08.04.
27. "Risk stratification in arrhythmogenic right ventricular cardiomyopathy". Circulation 2017;136:2068-2082.
28. "Sudden death and left ventricular involvement in arrhythmogenic cardiomyopathy". Circulation 2019;139:1786-1797.
29. "Clinical and genetic characterization of families with arrhythmogenic right ventricular dysplasia/cardiomyopathy provides novel insights into patterns of disease expression". Circulation 2007;115:1710-1720.
30. "Diagnosis of arrhythmogenic cardiomyopathy: the Padua criteria". Int J Cardiol 2020;319:106-114.
31. "Dilated cardiomyopathy and arrhythmogenic left ventricular cardiomyopathy: a comprehensive genotype-imaging phenotype study". Eur Heart J Cardiovasc Imaging 2020;21:326-336.
32. "Cardiac histological substrate in patients with clinical phenotype of Brugada syndrome". Circulation 2005;112:3680-3687.
33. "Biventricular myocardial fibrosis and sudden death in patients with Brugada syndrome". J Am Coll Cardiol 2021;78:1511-1521.
34. "Modulating effects of age and gender on the clinical course of long QT syndrome by genotype". J Am Coll Cardiol 2003;42:103-109.
35. "Clinical phenotypes and prognosis of dilated cardiomyopathy caused by truncating variants in the TTN gene". Circ Heart Fail 2020;13:e006832.
36. "Associations between female sex, sarcomere variants, and clinical outcomes in hypertrophic cardiomyopathy". Circ Genom Precis Med 2021;14:e003062.
37. "Long-term outcome of patients initially diagnosed with idiopathic ventricular fibrillation: a descriptive study". Circ Arrhythm Electrophysiol 2016;9:e004258.
38. "Genetic testing for inherited cardiovascular diseases: a scientific statement from the AHA". Circ Genom Precis Med 2020;13:e000067.
39. "Concealed arrhythmogenic right ventricular cardiomyopathy in sudden unexplained cardiac death events". Circ Genom Precis Med 2018;11:e002355.
40. "Clinical and genetic features of Australian families with long QT syndrome: a registry-based study". J Arrhythm 2016;32:456-461.
41. "Exome sequencing identifies a mutation in the ACTN2 gene in a family with idiopathic ventricular fibrillation, left ventricular noncompaction, and sudden death". BMC Med Genet 2014;15:99.
42. "Characterization of clinically relevant copy-number variants from exomes of patients with inherited heart disease and unexplained sudden cardiac death". Genet Med 2021;23:86-93.
43. "A novel heterozygous mutation in cardiac calsequestrin causes autosomal dominant catecholaminergic polymorphic ventricular tachycardia". Heart Rhythm 2016;13:1652-1660.
Appendix
Supplemental DataFootnotes
Listen to this manuscript's audio summary by Editor-in-Chief Dr Valentin Fuster on www.jacc.org/journal/jacc.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
