Non‐GAA Repeat Expansions in FGF14 Are Likely Not Pathogenic—Reply to: “Shaking Up Ataxia: FGF14 and RFC1 Repeat Expansions in Affected and Unaffected Members of a Chilean Family”

Saffie Awad and colleagues described 12 affected individuals from a Chilean family carrying an FGF14 repeat expansion. All affected family members exhibited a complex phenotype of cerebellar ataxia (without oculomotor signs), moderate-to-severe sensorimotor polyneuropathy, rapid eye movement sleep behavior disorder, autonomic dysfunction, and neuropsychiatric features. A wide range of age at onset was observed, varying from 17 to 50 years. Notably, the results of long-range polymerase chain reaction (LR-PCR) and repeat-primed PCR (RP-PCR) identified a non-GAA repeat expansion in all four tested FGF14 expansion carriers from this family. The repeat configuration of the expansion was, however, not established. The expansion did not segregate with disease in this family in which all affected individuals shared the same phenotype, raising questions about its pathogenicity. Intronic GAA repeat expansions in FGF14 have recently been shown to cause spinocerebellar ataxia 27B (SCA27B), which is characterized by a late-onset, slowly progressive pancerebellar syndrome with frequent cerebellar oculomotor signs and rarely associated with mild axonal polyneuropathy. The disease develops on average at about age 60, although it can rarely manifest before age 30 in patients carrying biallelic GAA expansions. The phenotype of the Chilean family therefore differs from that previously reported in the published series. The FGF14-SCA27B repeat locus is highly polymorphic, in terms of both repeat length and configuration. Expansions of non-GAA repeats at the FGF14-SCA27B locus have previously been identified in controls and patients with ataxia. To this date, only GAA repeat expansions have been shown to cause SCA27B. We have previously reported that (GAAGGA)n expansions did not segregate with disease in two Australian and three German families with adult-onset ataxia. Furthermore, we have recently identified a Turkish family with unsolved adult-onset spastic ataxia in whom a complex [(GAA)n(GCA)m]z expansion did not segregate with disease (Fig. 1A–C), thus further suggesting that nonGAA expansions in FGF14 are not pathogenic. These results are in keeping with previous reports showing that non-GAA expansions in the frataxin (FXN) gene are nonpathogenic. We have also observed that in addition to producing a flat profile with a short stretch of GAA repeats at either end of the locus on RP-PCR, non-GAA expansions produce a cluster of non-bell-shaped peaks on LR-PCR, compared to the characteristic bell-shaped appearance of GAA expansions (Fig. 1B–F). The FGF14 expansion identified in the Chilean family also exhibited little-to-no intergenerational instability upon maternal transmission across three generations, therefore unlikely accounting for the observed genetic anticipation. Pathogenic GAA-pure repeat expansions are associated with significant instability upon intergenerational transmission: GAA repeats are more likely to expand with maternal transmission and contract with paternal transmission. In comparison, nonGAA-pure expansions are much more stable upon transmission, similar to what has been observed in the Chilean family. In conclusion, we agree with Saffie Awad and colleagues that it is critical to determine the repeat configuration of FGF14 expansions, in addition to their size, as growing evidence suggests that non-GAA expansions are not pathogenic. This last point highlights the importance of adopting standardized protocols to diagnose SCA27B. Moreover, segregation studies in affected families, tests of association in large case–control series, and, ultimately, functional studies will be needed in the future to establish the pathogenicity of alternative motifs at the FGF14-SCA27B repeat locus.