Prevalence and Phenotypic Correlations of Calmodulinopathy-Causative CALM1-3 Variants Detected in a Multi-Center Molecular Autopsy Cohort of Sudden Unexplained Death Victims

Sudden unexplained death (SUD) is a profoundly tragic event for families and their communities. These cases can be categorized into two main groups including sudden infant death syndrome (SIDS, <1year of age), and SUD in the young (SUDY, 1-35 years). In the US alone, SIDS accounts for approximately 3000 sudden deaths each year, while SUDY occurs in up to 5000 individuals annually.¹ Approximately 10% of SIDS and 25% of SUDY may be caused by pathogenic variants in cardiac channelopathy-susceptibility genes.²

Three different calmodulin genes, CALM1 (chr14q21), CALM2 (chr2p21), and CALM3 (chr19q13), each of which has a distinct genomic locus and unique nucleotide sequence, all encode for an identical 149 amino acid calmodulin protein (CaM).³ Since 2012, pathogenic variants in CALM1-3-encoded CaM, approximately 90% of which are de novo, have been implicated as the underlying cause of multiple arrhythmic phenotypes including long QT syndrome (CaM-LQTS), catecholaminergic polymorphic ventricular tachycardia (CaM-CPVT), and idiopathic ventricular fibrillation (CaM-IVF), collectively termed the calmodulinopathies.⁴

Data from the recently published International Calmodulinopathy Registry found that 68% of 74 calmodulinopathic patients have suffered at least one major arrhythmic event at an average onset age of 4 years, and 27% experienced sudden cardiac death (SCD).⁴ However, despite its malignant and potentially lethal phenotype, the prevalence of calmodulinopathic variants in cases of SIDS and SUDY remains unknown.

Here, we determined the spectrum and prevalence of CALM1-3 pathogenic variants in a large multi-center cohort of 599 SIDS and 258 SUDY cases contributed from three international medical centers in the United States, United Kingdom, and Australia. In order to prevent the re-identification of individuals included in this study, individual patient data will not be made available to other researchers. This study complies with the Declaration of Helsinki; locally appointed ethics committees including Mayo Clinic’s Institutional Review Board have approved the research protocol.

Of the SIDS cases, 362 (60%) were male and 237 (40%) were female. The median age at death was 2 months (interquartile range (IQR) 1-4 months), and 349 (58%) of these cases were white. In the SUDY cohort, 176 (68%) cases were male and 82 (32%) were female. The median age at death was 21 years (IQR 16-29 years) with 88% dying between the ages of 1-35 years and 196 (76%) were white.

Postmortem genomic DNA, derived from each decedent, underwent either whole exome or targeted gene panel sequencing followed by a gene-specific analysis of CALM1, CALM2, and CALM3, using Ingenuity Variant Analysis software. Only rare (minor allele frequency ≤ 0.005% in gnomAD) nonsynonymous variants with a call quality score of ≥ 20 and a read depth of ≥ 10 were considered. Identified variants were classified according to the American College of Medical Genetics (ACMG) guidelines.

Overall, we identified a pathogenic CALM1-3 variant in 3 out of 857 SUD cases (0.035%). Interestingly, none of these variants were present in our SIDS cohort (0/599, 0%), but all 3 were identified in cases of SUDY (3/258, 1.2%; p=0.027; Figure 1A). The yield of pathogenic CALM1-3 variants was significantly higher in SUDY cases dying between the ages of 1 and 10 years (3/32, 9.4%) compared to those older than 10 years at age of death (0/226, 0%; p=0.002; Figure 1B). In comparison, approximately 7% of our SIDS cohort and 26% of our SUDY cohort hosted a rare nonsynonymous variant in one of the four major channelopathy genes (KCNQ1, KCNH2, SCN5A, or RYR2).

Figure 1. Yield and Location of CALM Variants Identified in SIDS and SUDY.

A.A table showing the yield of CALM variants in the SIDS and SUDY cohorts and in SUDY cases between the ages of 1-10 years and those > 10 years of age.

B.A schematic of the CaM protein showing the N- and C-domains, each with two EF hands (EF-I through EF-IV) with calcium (red) bound. Blue circles represent WT residues, and white circles represent variants identified in our SUDY cohort. Combined Annotation Dependent Depletion (CADD) score > 20 is considered an in silico threshold for possible pathogenicity.

A CaM-p.Asn54Ile variant (CALM1, c.161A>T) was identified in a 9-year-old female who died suddenly following extreme emotion. This variant is located in the inter-EF-hand I-II linker domain and is known to cause CaM-CPVT (Figure 1C).⁴ The second variant, CaM-р. Phe90Leu (CALM2, c.268T>C), which resides in the inter-EF-hand II-III linker domain was found in a 5-year-old male who experienced sudden death during physical exertion. While this variant has not been identified before in CALM2, CALM1-p.Phe90Leu has been associated previously with CaM-IVF previously (Figure 1C).⁴ The CaM-p.Asn98Ser variant (CALM2, c.293A#x003E;G) was identified in a 2-year-old male who died suddenly while engaging in toddler play. This variant is located in the EF-hand III domain and has been associated with both LQTS and CPVT in CALM1- and CALM2-encoded calmodulin (Figure 1C).⁴ However, patients with the p.Asn98Ser variant typically do not express an overt LQTS phenotype.⁵

Although CaM-LQTS patients exhibit a more malignant phenotype and have a higher rate of SCD than other calmodulinopathy phenotypes, it is not surprising that the CALM variants identified in our SUDY cases have been associated with CaM-CPVT or CaM-IVF and not CaM-LQTS. Typically, CaM-LQTS manifests with severe and readily detectable clinical features (QTc > 550 ms, bradycardia, 2:1 AV block, T-wave alternans) often occurring during infancy and is therefore likely detected prior to the occurrence of sudden death. Thus, the absence of CaM-LQTS variants in SIDS and SUDY may be explained by their high penetrance and marked expressivity. In contrast, CaM-CPVT and CaM−IVF patients do not display the same readily detectable clinical feature and may elude detection until a sentinel event of SCD after the first year of life. Additionally, channelopathies are responsible for a greater percentage of SUDY than SIDS and therefore larger number of cases may be needed to identify CALM-related SIDS cases.

While pathogenic variants in CALM1-3 do not contribute meaningfully to SIDS, about 1% of SUDY overall stems from pathogenic CALM variants. Additionally, CALM variants may account for up to 10% of the SUD cases occurring during childhood. Therefore, the CALM1, CALM2, and CALM3 genes should be included in postmorterm genetic testing (aka, the molecular autopsy), especially in children who have died between the ages of 1-10 years.