Hidden Gene Deletions Explain Severe Friedreich Ataxia Cases Misdiagnosed by Standard Testing

Friedreich ataxia has long been considered one of the most genetically predictable of the inherited ataxias, with the overwhelming majority of patients carrying two expanded GAA triplet repeats in the FXN gene. A new study published in Annals of Clinical and Translational Neurology now reveals a subtle genetic trap that has been hiding inside routine diagnostic results: some patients who appear, by standard testing, to carry two identical expanded alleles are in fact compound heterozygotes, with one expanded repeat and a previously invisible deletion of the proximal portion of the FXN gene. These covert deletions, undetectable by conventional polymerase chain reaction methods, are associated with strikingly severe disease courses, and their identification carries immediate consequences for diagnosis, prognosis, and genetic counseling.

The biology underlying Friedreich ataxia begins with the FXN gene, which encodes frataxin, a mitochondrial protein essential for iron-sulfur cluster assembly and cellular energy metabolism. When functional frataxin falls below a critical threshold, patients develop progressive cerebellar ataxia, loss of deep sensory pathways, hypertrophic cardiomyopathy, scoliosis, and an elevated risk of diabetes. Roughly 96 percent of disease-causing alleles involve expansion of a GAA repeat in the first intron of FXN, which silences transcription without altering the protein sequence itself. The remaining 4 percent of pathogenic alleles consist of intragenic variants such as missense changes, nonsense mutations, splice variants, and deletions of varying size. Patients carrying two pure GAA expansions show a reasonably consistent relationship between the length of the shorter repeat, known as GAA1, and clinical severity, which has made genetic testing a powerful prognostic tool.

That predictability breaks down, however, in a subset of patients whose disease course is far more aggressive than their apparent repeat lengths would suggest. A recent investigation into this discrepancy uncovered a surprising mechanism: in roughly half of individuals who seemed by long-range PCR to carry two expanded alleles of the same size, one of the two alleles was not being amplified at all. The test was effectively blind to one allele, producing the false conclusion of homozygosity for identical repeats. Most of these individuals turned out to carry expanded composite repeats, complex rearranged repeat structures that evade standard amplification, but a small number harbored something even more consequential: deletions of the proximal portion of the FXN gene, encompassing the promoter and early exons, which abolish production of frataxin from that allele entirely.

The new report, led by investigators at the Children’s Hospital of Philadelphia using records from the Friedreich Ataxia Clinical Outcome Measures Study, presents the clinical features of patients carrying these covert proximal deletions alongside one expanded GAA allele. The prototypic case is particularly striking. Patient 1 was diagnosed at age six with ataxia superimposed on a pre-existing diagnosis of hypertrophic cardiomyopathy. He lost the ability to walk at age 12, developed diabetes at 13, and suffered severe vision loss at 16. Initial genetic testing had indicated that he inherited two GAA expansion alleles of similar length from his carrier parents. The unraveling of that conclusion came through his younger sister, who developed Friedreich ataxia at age 10 and was found to carry two pure GAA expansions of clearly distinct sizes, 490 and 990 repeats. The genetic mismatch between siblings who shared the same parents prompted deeper analysis.

Long-read whole genome sequencing using Oxford Nanopore Technology, confirmed by a custom PCR assay, resolved the puzzle. Patient 1 had inherited only one GAA-expanded allele; his second allele carried a de novo deletion removing the FXN promoter, exon 1, and part of intron 1. Because the deletion eliminated the sequences targeted by standard diagnostic PCR, the allele appeared simply absent from the analysis, mimicking homozygosity for two similar repeats. The contrast between the siblings is clinically dramatic: Patient 2, carrying two expanded repeats but no deletion, had a much milder course, with no cardiomyopathy and no loss of ambulation at her most recent clinic visit, while her brother with the silent deletion deteriorated rapidly through childhood and adolescence.

The researchers extended the comparison to a broader set of patients with null variants of FXN, alleles that produce no functional frataxin at all. Across sixteen patients with deletions ranging from the proximal promoter region to distal exons, and including frameshift and start codon mutations, a consistent pattern emerged. All affected individuals showed ataxia and at least mild dysarthria early in their course. Those with proximal deletions tended toward early disease onset, with cardiomyopathy and scoliosis appearing in childhood. Poor weight gain and appetite, a recognized endocrine feature of Friedreich ataxia, developed in all patients with deletions, and one patient with a small frameshift variant required placement of a gastric feeding tube for nutrition. Yet the picture was not uniformly grim: a few individuals carrying shorter GAA alleles on their second copy, including one patient with a start codon mutation paired with a repeat of only 68 triplets, remained remarkably mild, with onset as late as age 41 and none of the major complications.

To quantify the impact of null alleles, the team compared the deletion and null variant group with a large cohort of 478 patients homozygous for two expanded GAA alleles. The mean shorter repeat length was statistically indistinguishable between groups, 667 plus or minus 235 repeats in the homozygous cohort versus 734 plus or minus 228 in the deletion group, yet the clinical outcomes diverged sharply. Patients with null variants had a lower average age of onset, 8.3 years versus 12 years, and dramatically higher rates of cardiomyopathy, 93 percent versus 61 percent, a difference that reached statistical significance. Age at death also differed significantly, with 32 percent of the deletion group deceased compared with 9 percent of the homozygous repeat group. Trends toward more frequent diabetes, scoliosis surgery, and vision loss from optic neuropathy were also observed. The conclusion is inescapable: a null allele, which contributes no frataxin whatsoever, produces a more severe phenotype than a second expanded repeat, which still permits a trickle of residual gene expression.

These findings carry a direct message for clinical practice. Patients who appear by commercial testing to carry two identical expanded alleles but who show an unexpectedly severe, even if otherwise typical, Friedreich ataxia phenotype should be evaluated further for an undetected deletion. Conversely, patients with one expanded allele, one apparently normal allele, and a severe phenotype should prompt a search for a null non-GAA variant, while a single expanded allele with a normal second allele and an atypical phenotype more often points toward a missense variant, though exceptions exist. The diagnostic gap exists because standard PCR-based assays target sequences within the proximal gene that are precisely the regions missing in these covert deletions. Long-read whole genome sequencing, which can traverse the entire locus and reveal structural rearrangements, proved essential for detection, and the authors demonstrated that a targeted custom PCR assay can confirm the findings once the deletion is known.

Beyond diagnosis, the study refines the statistical architecture of Friedreich ataxia prognosis. Conventional severity models treat GAA1, the shorter of the two expanded repeats, as the primary determinant of disease course, and for good reason: GAA1 length alone accounts for an R-squared of roughly 0.4 to 0.5 in predicting various clinical parameters, because it best indexes residual FXN transcription. GAA2, the longer repeat, is usually omitted from models because its values are correlated with GAA1 and its independent contribution is difficult to isolate. The new data suggest that in extreme cases, particularly when one allele is a complete null, GAA2 length can further refine severity predictions, explaining some of the variance that GAA1 alone leaves unaccounted for. Patient 11, with a 68-repeat allele paired with a null variant and a remarkably mild course, illustrates how much protective information a short second allele can carry even when its partner produces no protein at all.

The broader significance of this work extends to the growing recognition that Friedreich ataxia is more genetically heterogeneous than standard testing implies. As long-read sequencing becomes more accessible, previously invisible structural variants such as composite repeats and proximal deletions are likely to be uncovered in increasing numbers, reshaping prevalence estimates and improving the accuracy of genetic counseling for families in which recurrence risk depends on knowing exactly what each parent carries. For patients, an accurate molecular diagnosis is the gateway to emerging therapies, including frataxin-enhancing compounds and gene-based approaches currently in development, many of which assume intact target sequences within the FXN gene. Knowing whether a patient’s second allele is a long repeat or a deletion may ultimately influence which therapeutic strategy is appropriate. What began as a puzzling discrepancy between two siblings has exposed a systematic blind spot in Friedreich ataxia genetics, one that the field can now begin to correct.

Subject of Research: Covert proximal FXN gene deletions in compound heterozygote Friedreich ataxia patients

Article Title: Compound Heterozygote Friedreich Ataxia Patients With Covert Proximal FXN Gene Deletions

Article References: Lazaropoulos, M. P., Devore, M. C., Lam, C., Park, C., Bidichandani, S., & Lynch, D. R. (2026). Compound Heterozygote Friedreich Ataxia Patients With Covert Proximal FXN Gene Deletions. Annals of Clinical and Translational Neurology, 13(10), 2134-2137. https://doi.org/10.1002/acn3.70408

Image Credits: AI Generated

DOI: 10.1002/acn3.70408

Keywords: Friedreich ataxia, FXN gene, frataxin, GAA repeat expansion, genetic testing, long-read sequencing, compound heterozygote, gene deletion, cardiomyopathy, ataxia, null variants, genetic diagnosis