gene product, absence of the variant from control population data-
bases (e.g., Exome Sequencing Project, 1000 Genomes Project,
Exome Aggregation Consortium, or gnomAD [22]), and a new
missense substitution at an amino acid residue where a different
substitution has been previously identified as pathogenic
[21]. Other supporting criteria include co-segregation with disease
in multiple affected family members and the presence of multiple
lines of computational evidence (e.g., with regard to residue con-
servation) to support an adverse impact of the variant on the gene
product, although use of in silico prediction algorithms is not
without limitation [21]. In contrast, criteria in favor of classification
as a benign variant include the presence of a high allele frequency
for the variant in population databases (>5% in Exome Sequencing
Project, 1000 Genomes Project, or Exome Aggregation Consor-
tium), observation in healthy (heterozygous) adults where full
penetrance is anticipated at a young age, lack of segregation in
affected family members, or several lines of in silico evidence argu-
ing against a deleterious impact of the variant on the gene product.
By allocating a different weighting to these (and other) pathogenic
and benign criteria, variants relevant to Mendelian disease can be
classified as “pathogenic,” “likely pathogenic,” of “uncertain sig-
nificance,” “likely benign,” or “benign” [21]. While not disease-
specific, such a framework structures the process of reliable rare
variant interpretation, aiding categorization as causal in disease or
likely innocent bystander, a growing and important challenge given
the increasing clinical and research application of high-throughput
DNA sequencing, and use of more extensive gene panels [23].
PRKAG2mutations display autosomal dominant transmission
and largely present in late adolescence or early adult life (mean age
24 years in one case series) [24]. A minority of sporadic cases
present in infancy, likely reflecting de novo mutation, typically
with profound impact on kinase activity and florid clinical sequelae
[25, 26]. While penetrance is close to 100%, there is, in common
with sarcomeric HCM, evidence of significant variability in disease
expressivity, including within families, some of whom carry effec-
tively unique “private” mutations [2]. Thus, while pre-excitation
appears to be highly prevalent in Arg302Gln mutation carriers
[17], families with a high incidence of right bundle branch block
and short PR interval, but no LVH or pre-excitation, and others
with no demonstrable glycogen excess have all been described
[27–29]. Notwithstanding the rarity of the syndrome (with ~200
patients reported in the literature), attempts at genotype-
phenotype correlation have been made, with findings of a higher
frequency of LVH but lower rate of electronic pacemaker implan-
tation, with the Asn488Ile compared to Arg302Gln variant
[17]. At the extreme end of the disease spectrum, mutations at
the Arg531 residue in CBS4 are notable for their association with
early presentation or fulminant phenotype: paroxysmal atrial
PRKAG2 syndrome 587