This retrospective observational analysis of the Dutch national Fabry disease database (Amsterdam UMC) examined how the adoption of broader sequencing techniques and the addition of the GLA-gene to NGS/WES panels altered the types of patients diagnosed with Fabry disease. The cohort included 319 individuals with pathogenic or likely pathogenic GLA variants (data export December 2, 2025). The authors report a significant increase in detection of non-classical/attenuated variants and a shift toward presentations dominated by cardiomyopathy after broader sequencing was introduced (~2011). This shift has implications for genetic counseling, clinical follow-up, and interpretation of historical outcome data.
Fabry disease (FD) is an X-linked lysosomal storage disorder caused by deleterious variants in the GLA-gene, leading to deficiency of alpha-galactosidase A. Disease expression ranges from classical, severe childhood-onset multisystem disease in males (acroparesthesia, cornea verticillata, angiokeratoma, and later renal, cardiac, and cerebrovascular complications) to milder non-classical or cardiac-predominant phenotypes associated with residual enzyme activity and lower substrate levels (Gb3, lyso-Gb3).
The GLA-gene is genetically heterogeneous, with many distinct pathogenic variants described worldwide. Diagnostic practice expanded from targeted Sanger sequencing to include Next Generation Sequencing (NGS) and Whole Exome Sequencing (WES)/WGS in Amsterdam UMC from around 2011. These broader tests allow inclusion of GLA on panels for conditions such as hypertrophic cardiomyopathy (HCM), renal disorders, and polyneuropathies.
Data were extracted from the national FD database maintained by the Inherited Metabolic Disease Expert Center at Amsterdam UMC. The registry contains prospective clinical and genetic data collected during outpatient visits since 1999. The analysis included individuals with GLA variants classified as pathogenic or likely pathogenic at the time of diagnostics. Variants currently deemed benign (P60L, A143T, D313Y, T385A, W277C) were excluded; R112H is included although its pathogenicity has been subject to debate.
Male patients were classified as classical if they had at least one classical symptom (cornea verticillata, angiokeratoma, acroparesthesia) and untreated plasma lyso-Gb3 > 40 nmol/L. Males not meeting these criteria were classified as non-classical. Female patients were grouped by whether they carried variants typically causing classical disease in males (deleterious) or less deleterious variants. The study included only variants labeled pathogenic/likely pathogenic at testing; classification largely remained stable over time for most variants.
Prior to ~2011, diagnostics were typically single-gene Sanger testing of GLA. From ~2011 onward, broader NGS and WES-based techniques were offered; analysis pipelines ranged from targeted panels to mendeliome or open exome analysis. The GLA-gene was added to several diagnostic panels, most notably those for hypertrophic cardiomyopathy. Dutch diagnostic laboratories operate under ISO 15189 accreditation; Amsterdam reported high coverage and sensitivity for GLA sequencing with additional Sanger sequencing for low coverage regions when needed.
The authors compared the proportion of classical versus non-classical index cases identified by single-gene testing versus broad sequencing techniques, and compared diagnoses before versus after 2011. Both NGS and WES/WGS were grouped as broad sequencing. A Chi-squared test was used for comparisons; reported p-values included p < 0.001 for differences in detection route between classical and non-classical variants and p = 0.004 for differences in variant detection before versus after 2011.
After exclusions (benign variants, indeterminate phenotypes, one enzyme-positive but genotype-negative male), 319 individuals with pathogenic GLA variants remained for analysis, representing 104 index patients and 64 unique GLA variants. Most variants were inherited (n = 275); 8 were reported as de novo and 36 could not be established. Among index patients, broad sequencing identified 68% of non-classical variant carriers versus 22% of classical variant carriers (p < 0.001). Across the cohort, 24% of non-classical cases were diagnosed via broad sequencing compared to 7% of classical cases (p < 0.001). More non-classical variants were identified after 2011 compared to before (p = 0.004), showing a temporal shift.
Six variants were notably recurrent in the cohort (I319T, P389A, R342Q, F18S, R220*, R112H). The most frequent, I319T (n = 43), associates with a non-classical phenotype and was independently established in 19 patients. Genealogical reconstruction suggested that many I319T carriers share a likely distant common ancestor from the 18th–19th century; degrees of kinship ranged from fourth-degree to much more distant (up to 15th degree). The authors modelled descendant numbers and estimated that a single affected female ancestor from 1850 could give rise to roughly 80 affected descendants alive today; their pedigree with four ancestors (1840–1860) identified 43 affected descendants, suggesting underascertainment of carriers in the population.
For classical/severe variant carriers (n = 169 with known reason for testing), cascade family screening was the most frequent indication (113, 67%), followed by classical symptoms (29, 17%), cardiomyopathy (14, 8%), and renal insufficiency (6, 4%). For attenuated/non-classical variant carriers (n = 115), cascade screening remained most common (74, 64%), but cardiomyopathy accounted for a larger share (28, 24%). The inclusion of GLA in HCM panels made cardiomyopathy a prominent indication for index case testing in more recent years.
The diagnostic shift toward identifying individuals with less deleterious variants and milder, often cardiac-predominant phenotypes has practical consequences. Comparing current cohorts to historical cohorts may be misleading without accounting for changes in testing strategy and case mix. Clinical follow-up and counseling should be tailored to avoid over- or undertreatment: individuals identified through broad panels may require different surveillance intensity and treatment thresholds than those with classical FD.
This analysis is retrospective and based on a single-country national registry. Full raw data were not made publicly available for privacy reasons, though data are reportedly accessible under data-sharing agreements. Variant pathogenicity classifications and some phenotype assignments can evolve over time (noted for R112H), and linkage of some index cases to family histories was limited by available records. Non-paternity testing was not systematically performed in de novo assignments.
Adding the GLA-gene to broader NGS/WES panels has increased detection of non-classical, cardiomyopathy-predominant Fabry disease and shifted the phenotypic composition of the diagnosed cohort. This change necessitates awareness when interpreting clinical and treatment outcome data across time periods, and highlights the need for individualized genetic counseling and follow-up strategies to balance risks of over- and undertreatment.