Phenotypic implications of pathogenic variant types in Pompe disease

Newborn screening and therapies for Pompe disease (glycogen storage disease type II, acid maltase deficiency) will continue to expand in the future. It is thus important to determine whether enzyme activity or type of pathogenic genetic variant in GAA can best predict phenotypic severity, particularly the presence of infantile-onset Pompe disease (IOPD) versus late-onset Pompe disease (LOPD). We performed a retrospective analysis of 23 participants with genetically-confirmed cases of Pompe disease. The following data were collected: clinical details including presence or absence of cardiomyopathy, enzyme activity levels, and features of GAA variants including exon versus intron location and splice site versus non-splice site. Several combinations of GAA variant types for individual participants had significant associations with disease subtype, cardiomyopathy, age at diagnosis, gross motor function scale (GMFS), and stability of body weight. The presence of at least one splice site variant (c.546 G > C/p.T182 = , c.1076–22 T > G, c.2646 + 2 T > A, and the classic c.−32–13T > G variant) was associated with LOPD, while the presence of non-splice site variants on both alleles was associated with IOPD. Enzyme activity levels in isolation were not sufficient to predict disease subtype or other major clinical features. To extend the findings of prior studies, we found that multiple types of splice site variants beyond the classic c.−32–13T > G variant are often associated with a milder phenotype. Enzyme activity levels continue to have utility for supporting the diagnosis when the genetic variants are ambiguous. It is important for newly diagnosed patients with Pompe disease to have complete genetic, cardiac, and neurological evaluations.