Pure GAA repeat expansions in FGF14 are related to late onset cerebellar ataxia (SCA27b).

There are two major issues for counseling: First, the pathogenic thresholds of the length of these repeat expansions are debated in literature and hence uncertain. Second: it was suggested in the literature that maternal transmission leads to expansion, whereas paternal transmission leads to contraction. It is uncertain whether this is always true, whether the size of the intergenerational changes can be predicted and whether large-normal repeat lengths need to be reported with a risk for offspring.

These two uncertainties hamper adequate counseling of patients and their relatives in clinical practice.

Description of studies

A total of nine studies were included in the analysis of the literature. Important study characteristics and results are summarized in table 2.

Kakumoto (2025) performed an observational study to analyze the frequency of FGF14 GAA repeat expansions. In total, 548 patients with multiple system atrophy (MSA), 476 patients with undiagnosed ataxia and 455 healthy participants were included. Patients with undiagnosed ataxia were included if they had progressive ataxia, onset after age of 20 years, had no secondary ataxia due to cerebrovascular diseases, tumor, alcohol abuse, autoimmune diseases, endocrinological diseases, or remote effects of malignancies, and were negative for SCAs 1, 2, 6, 7, 8, 12, 17, 31, and 36, Machado–Joseph disease (SCA3), and dentatorubral-pallidoluysian atrophy. Long-range PCR products were further analyzed using a long-read sequencer.

Laß (2024) performed an observational study to characterize the repeat tract sequence, assess mosaicism, and determine the number of repeat interruptions in individuals with FGF14-related late-onset cerebellar ataxia as well as in unaffected individuals. Participants with progressive cerebellar ataxia of unknown etiology were included. Exclusion criteria were patients with secondary forms of ataxia (lesion, toxic, inflammatory, and paraneoplastic) and known repeat-expansion SCAs (SCA1, 2, 3, 6, and 17). All control participants underwent a structured neurological examination and showed no signs of neurological disease. Long-range PCR, Sanger sequencing, repeat-primed PCR, Nanopore, and PacBio sequencing were used. In total, 27 ataxia patients and 13 unaffected individuals were included with expanded repeat lengths (³250 repeats). 

Livanos (2025) performed a retrospective observational study to examine the prevalence of GAA repeat expansion and the phenotypic profile in a Cypriot cohort with unresolved late-onset cerebellar ataxia. Participants who developed ataxia at or after the age of 20 years and tested negative for the SCA1, SCA2, SCA3, SCA6, SCA7, SCA8, SCA10, SCA12, SCA36, DRPLA, Friedreich’s ataxia (FRDA) and the pathogenic RFC1 repeat expansions were included. Participants with other ataxia-causative recessive variants were excluded. In total, 155 unrelated index patients who presented with late-onset cerebellar ataxia and 229 non-neurological disease controls were included. The repeat locus was analyzed using long-range PCR followed by capillary electrophoresis, agarose gel electrophoresis, and automated electrophoresis. In addition, bidirectional repeat-primed PCR and Sanger sequencing were used to verify the absence of interruptions or non-GAA motifs within the expanded alleles.

Mereaux (2024) performed a cross-sectional study to evaluate the frequency and length of FGF14 expanded alleles in individuals with cerebellar ataxia compared to control participants and to explore the related genetic and clinical variability. In total, 845 patients with cerebellar ataxia and 475 controls were included. Patients were selected with half of the patients having a negative result on a routine diagnostic molecular test for SCAs (consisting of targeted sizing of CAG repeats in ATXN1, -2, -3, -7, and CACNA1A). Whole Exome Sequencing (WES) was conducted in 678 patients to identify single nucleotide variants and repeat expansions with 100% sensitivity. Patients who did not undergo WES were clinically selected for SCA27b testing based on the presence of downbeat nystagmus or episodic symptoms of cerebellar ataxia. The FGF14 GAA locus was amplified by long-range and repeat-primed polymerase chain reactions (PCR).

Miyatake (2024) performed an observational study to elucidate the molecular features of the GAA-FGF14 locus and to explore the relationship between sequence expansion and disease. In total, 460 patients with molecularly undiagnosed adult-onset ataxia and 1114 controls (1022 Japanese and 92 non-Japanese) were included. Of the 1022 Japanese controls, 837 participants were unaffected healthy controls and 185 other disease controls with congenital, pediatric or multiorgan disorders without neurodegeneration. Genetic analyses were performed with genomic DNA. Long-read whole-genome sequencing (LR-WGS) on PromethION was performed or FGF14 repeat expansion was screened using PCR-based assays and then either targeted long-read sequencing (T-LRS) using GridION or fragment analysis of screening-positive samples. 

Mohren (2024) performed an observational study to identify FGF14 repeat expansion and provide molecular, phenotypic and mechanistic insights in how pure AAG repeat expansions lead to adult-onset cerebellar ataxia whereas expansions involving other motifs are non-pathogenic and interrupted expansions may exhibit reduced penetrance. In total, 169 patients and 802 controls were included. FGF14 alleles were analyzed by a combination of long-range PCR, fluorescent gene fragment analysis and targeted nanopore sequencing.

Pellerin (2023) performed an observational study to discover novel disease-causing repeat expansions in individuals with late-onset cerebellar ataxia. Patients with progressive ataxia with onset at or after 30 years of age, with no clinical signs indicative of multiple system atrophy cerebellar subtype (MSA-c) and exclusion of acquired causes and, when appropriate, negative results on an ataxia gene panel as well as on screening for repeat expansions associated with common spinocerebellar ataxias, fragile X–associated tremor/ataxia syndrome, Friedreich’s ataxia, and CANVAS (cerebellar ataxia, neuropathy, and vestibular areflexia syndrome). Six patients with late-onset cerebellar ataxia without underlying genetic cause were included. In addition, a validation cohort of 66 French Canadian, 228 German, 20 Australian and 31 Indian index patients were included. Control cohorts contained 209 French Canadian and 199 German individuals who were neurologically healthy or who were reported not to have ataxia. Long-range PCR and targeted long-read nanopore sequencing were used. 

Rafehi (2023) performed an observational study to identify an intronic (GAA) repeat expansion in fibroblast growth factor 14 (FGF14). PCR and long-read sequence analysis were performed. Participants presenting with clinical features suggestive of a genetic cerebellar ataxia were included. Exclusion criteria included clinical features indicative of an acquired cause, including ataxia onset related to an acute injury or illness such as stroke, encephalitis or sepsis. The Australian cohort consisted of 95 ataxia patients and 215 controls (all healthy). In addition, 96 lymphoblast-derived gDNA samples from UK control individuals were analyzed (all healthy). Besides, the German validation cohort consisted of 104 individuals with ataxia of an unidentified cause and 190 controls. PCR and long-read sequence analysis were used.

Zheng (2024) performed an observational study to establish the prevalence of the GAA repeat expansion in FGF14 among Chinese patients with late-onset cerebellar ataxia (LOCA) and to characterize patterns of tandem repeat inheritance, radiological findings, and sympathetic nervous system involvement. Genetic screening for the GAA repeat expansion was conducted in 664 individuals with undiagnosed LOCA and in 39 family members of the two GAA-FGF14 probands. Long-range PCR, repeated-primed PCR and Sanger sequencing were used. Eighteen GAA-FGF14 patients were included. The control group consisted of 16 unrelated individuals who were healthy at the time of sympathetic skin response assessment and were matched to the 16 GAA-FGF14 patients under follow-up by age and sex.

Table 2. Characteristics of included studies

Results

1. Length of repeat expansion
Kakumoto (2025) reported that 6 undiagnosed ataxia patients (1.3%) had a pure GAA expansion ³250 (ranging from 251 to 361) as compared to 1 healthy control (0.2%). Of these six patients, five carried a (GAA)≥300 expansion, which was classified as falling within the full-penetrance range, while the other patient had a (GAA)250–299 expansion, considered to lie within the incomplete-penetrance range.Besides, 2 undiagnosed ataxia patients (0.4%) had a pure GAA expansion between 200 and 249 (ranging from 212 to 240), as compared to 1 healthy control (0.2%). No pure GAA expansion <200 were found in healthy controls, while this was found in 1 undiagnosed ataxia patient (0.2%).

Laß (2024) reported that the 27 FGF14 ataxia patients had a median repeat number of 342 (IQR 304 to 426) on long-range PCR as compared to 309 (IQR 286 to 326) for the 13 unaffected individuals (healthy controls).

Livanos (2025) reported that eight FGF14 ataxia patients (5.1%) carried a pathogenic (GAA)_³300 repeat expansion, while four (2.6%) had a reduced penetrance (GAA)_250-299 expansion. The median repeat count of patients with (GAA)_³250 repeat expansion was 323 repeat units (range 251 to 489 repeats). Two non-neurological disease controls (0.9%) were identified with a (GAA)₂₈₆ and a (GAA)_275. No control individuals were found to carry the (GAA)_³300 repeat expansion. In addition, the median repeat number was 17 trinucleotide repeat units (IQR: 10 to 37) for FGF14 ataxia patients, while for the non-disease controls, it was 12 repeat units (IQR: 10 to 33).  

Mereaux (2024) reported that the proportion of individuals carrying FGF14 GAA_250-299 alleles was similar for FGF14 ataxia patients and healthy controls (1.7%). Besides, 745 of the 845 FGF14 ataxia patients (88.2%) had <250 GAA repeat for FGF14 as compared to 462 of the 475 healthy controls (97.3%) (p<0.0001). However, 85 of the 845 FGF14 ataxia patients  (10.1%) had ³300 GAA repeat for FGF14 as compared to 5 of the 475 healthy controls (1.1%) (p<0.0001). For the five healthy controls, the GAA-pure alleles ranged from 308 to 380 repeats.

Miyatake (2024) reported that GAA repeat with ³250 repeat units was found in 8 of the 460 FGF14 ataxia patients (1.7%) as compared to 3 of the 1022 healthy controls (0.29%) (OR=6.012, 95%CI 1.802 to 21.03). Besides, the pure GAA repeat with ³200 repeat units was reported in 14 of the 460 FGF14 ataxia patients (3%) as compared to 5 of the 1022 healthy controls (0.49%) (OR=6.385, 95%CI 2.246 to 16.18). The frequencies of the alleles of GAA-GCA repeat with ³200 repeat units and any repeat with ³200 repeat units were not different between FGF14 ataxia patients and healthy controls. These results showed that pure GAA repeat with ≥ 200 repeats may be pathogenic with reduced penetrance, and GAA-GCA repeat expansion is not pathogenic despite the tendency to be longer than pure GAA repeat expansion.

Mohren (2024) reported that AAG repeats between 180 and 249 were found in 14 FGF14 ataxia patients (8.1%) and in 27 healthy controls (3.3%). An AAG repeat between 250 and 299 was reported in 11 ataxia FGF14 patients (7.4%) and in 5 healthy controls (0.6%). In addition, AAG repeats ³ 300 were reported in 31 FGF14 ataxia patients (15.5%) as compared to 2 healthy controls (0.2%). In contrast, large AAGGAG expanded alleles were found at similar frequencies in FGF14 ataxia patients and healthy controls (p=0.74, OR=1.16), confirming that these expansions are non-pathogenic.

Pellerin (2023) reported that in the French Canadian cohort consisting of 66 FGF14 ataxia patients and 209 healthy controls, a (GAA)_³250 was observed in 40 FGF14 ataxia patients (61%) and in 3 healthy controls (1%) (OR=105.60, 95%CI 31.09 to 334.20, p<0.001). In the German cohort consisting of 228 FGF14 ataxia patients and 199 healthy controls, a (GAA)_³250 was observed in 42 FGF14 ataxia patients (18%) and 5 healthy controls (3%) (OR=8.76, 95%CI 3.45 to 20.84, p<0.001). Furthermore, 3 of 20 Australian index patients (15%) and 3 of 31 Indian index patients (10%) carried a repeat expansion in FGF14 . (GAA)_250-300 expansions are likely to be pathogenic, although with reduced penetrance, and (GAA)_>300 expansions are fully penetrant.

Rafehi (2023) reported that LR-PCR analysis of all 95 affected individuals (FGF14 ataxia) and 311 healthy controls resulted in thirteen affected individuals (13.7%) that demonstrated pure (GAA)_>250, compared to only two healthy control subjects (0.6%). Four of the thirteen affected individuals had (GAA)_>335. Pure alleles (GAA)_>300 were exclusively found among affected individuals. Besides, in the German cohort, consisting of 104 affected individuals and 190 healthy controls, an (GAA)_>250 was reported in 15 affected individuals (14.4%) and in 10 healthy controls (5.3%) (p=0.014, OR=3.03 with 95%CI 1.3 to 7.02). Expanded alleles greater than (GAA)_>335 are disease causing and fully penetrant, while (GAA)_>250 are likely to be pathogenic, although with reduced penetrance.

Zheng (2024) reported that of the 18 individuals with FGF14, the GAA-FGF14 repeat number expanded in 10 individuals following maternal transmission (with expansions ranging from 1 to 35 repeats), whereas it contacted during paternal transmission in 3 individuals (with contractions ranging from 93 to 99 repeats).

2. Association with age at onset

Kakumoto (2025) reported that the number of GAA repeat units and the age at onset tended to be inversely correlated, although this was not statistically significant (Spearman’s r = −0.37, p = 0.50). The age at onset was 64.8 ± 5.1 years (range 57–71).

Livanos (2025) reported a weak inverse correlation between repeat size and age of onset (R²=0.016) (after excluding an outlier). However, this was not statistically significant (p=0.70). The mean age at disease onset was 60.0 ± _13.5 years.

Laß (2024) reported that there was an association between GAA repeat length and age at onset (p=0.016, R²=0.256) in patients with a pure GAA tract >250 repeat units.

Mereaux (2024) reported that compared to the GAA_<250 group, the GAA_³300 group had the latest median age at onset (55 versus 40 years in the GAA_<250 group; p < 0.0001). However, there was no correlation between the age at onset and the expanded GAA allele size above 300 GAA.

Miyatake (2024) reported a negative correlation between age of onset and repeat count (R²=0.2590, p=0.0631). Similar clinical characteristics were seen in patients with ≥ 250 repeat units or with 200–249 units, further supporting the assertion that a repeat count of 200-249 may also cause SCA27B.

Mohren (2024) reported an inverse correlation between the number of AAG repeats and the age at onset (R²=0.27, p=0.00043). Patients with FGF14 expansions between 250 and 299 repeats are significantly later affected on average (63.4 years; n = 36) than patients with ≥ 300 repeats (58.5 years; n = 115; p=0.046).

Pellerin (2023) reported a weak inverse correlation between the age at onset and the size of the repeat expansion (R²= 0.10).

Rafehi (2023) reported an inverse correlation between FGF14 RE length and ataxia onset (R²=0.44). The regression suggests that for every (GAA)₁₀ increase above 250 repeats, the age of onset is reduced by about 1.16 years.

A systematic review of the literature was performed to answer the following question(s):

What is the pathogenic threshold for the length of repeat expansions in FGF-14/SCA27b related to ataxia?

Table 1. PICO

Relevant outcome measures

The guideline panel considered length of repeat expansion as a critical outcome measure for decision making; and anticipation as an important outcome measure for decision making.

Search and select (Methods)

A systematic literature search was performed by a medical information specialist using the following bibliographic databases: Embase.com and Ovid/Medline all. Both databases were searched from 2005 to April 30, 2025, for systematic reviews and observational studies. Systematic searches were completed using a combination of controlled vocabulary and natural language keywords. The overall search strategy was derived from the following primary search concepts: (1) FGF14/ SCA27b; (2) repeat expansion. Duplicates were removed using EndNote software. After deduplication a total of 72 records were imported for title/abstract screening. Initially, 20 studies were selected based on title and abstract screening. After reading the full text, eleven studies were excluded (see the exclusion table under the tab ‘Evidence tabellen’), and nine studies were included.

Given the type of literature that was found, the working group concluded that a systematic literature review according to the GRADE methodology (Guyatta, 2011) would not provide any added value.

1. Laß J, Thomsen M, Borsche M, Lüth T, Prietzsche JC, Schaake S, Milovanović A, Macpherson H, Gustavsson EK, Saffie Awad P, Dragašević-Mišković N, Laabs BH, König IR, Westenberger A, Pearson CE, Brüggemann N, Klein C, Trinh J. FGF14 repeat length and mosaic interruptions: modifiers of spinocerebellar ataxia 27B? Brain. 2025 Nov 4;148(11):4072-4083. doi: 10.1093/brain/awaf183. PMID: 40379261; PMCID: PMC12588676.
2. Livanos I, Votsi C, Michailidou K, Pellerin D, Brais B, Zuchner S, Pantzaris M, Kleopa KA, Zamba Papanicolaou E, Christodoulou K. The FGF14 GAA repeat expansion is a major cause of ataxia in the Cypriot population. Brain Commun. 2025 Jan 3;7(1):fcae479. doi: 10.1093/braincomms/fcae479. PMID: 39801711; PMCID: PMC11724429.
3. Méreaux JL, Davoine CS, Pellerin D, Coarelli G, Coutelier M, Ewenczyk C, Monin ML, Anheim M, Le Ber I, Thobois S, Gobert F, Guillot-Noël L, Forlani S, Jornea L, Heinzmann A, Sangare A, Gaymard B, Guyant-Maréchal L, Charles P, Marelli C, Honnorat J, Degos B, Tison F, Sangla S, Simonetta-Moreau M, Salachas F, Tchikviladzé M, Castelnovo G, Mochel F, Klebe S, Castrioto A, Fenu S, Méneret A, Bourdain F, Wandzel M, Roth V, Bonnet C, Riant F, Stevanin G, Noël S, Fauret-Amsellem AL, Bahlo M, Lockhart PJ, Brais B, Renaud M, Brice A, Durr A. Clinical and genetic keys to cerebellar ataxia due to FGF14 GAA expansions. EBioMedicine. 2024 Jan;99:104931. doi: 10.1016/j.ebiom.2023.104931. Epub 2023 Dec 27. PMID: 38150853; PMCID: PMC10784672.
4. Miyatake S, Doi H, Yaguchi H, Koshimizu E, Kihara N, Matsubara T, Mori Y, Kunieda K, Shimizu Y, Toyota T, Shirai S, Matsushima M, Okubo M, Wada T, Kunii M, Johkura K, Miyamoto R, Osaki Y, Miyama T, Satoh M, Fujita A, Uchiyama Y, Tsuchida N, Misawa K, Hamanaka K, Hamanoue H, Mizuguchi T, Morino H, Izumi Y, Shimohata T, Yoshida K, Adachi H, Tanaka F, Yabe I, Matsumoto N. Complete nanopore repeat sequencing of SCA27B (GAA-FGF14 ataxia) in Japanese. J Neurol Neurosurg Psychiatry. 2024 Nov 18;95(12):1187-1195. doi: 10.1136/jnnp-2024-333541. PMID: 38816190.
5. Mohren L, Erdlenbruch F, Leitão E, Kilpert F, Hönes GS, Kaya S, Schröder C, Thieme A, Sturm M, Park J, Schlüter A, Ruiz M, Morales de la Prida M, Casasnovas C, Becker K, Roggenbuck U, Pechlivanis S, Kaiser FJ, Synofzik M, Wirth T, Anheim M, Haack TB, Lockhart PJ, Jöckel KH, Pujol A, Klebe S, Timmann D, Depienne C. Identification and characterisation of pathogenic and non-pathogenic FGF14 repeat expansions. Nat Commun. 2024 Sep 3;15(1):7665. doi: 10.1038/s41467-024-52148-1. PMID: 39227614; PMCID: PMC11372089.
6. Pellerin D, Danzi MC, Wilke C, Renaud M, Fazal S, Dicaire MJ, Scriba CK, Ashton C, Yanick C, Beijer D, Rebelo A, Rocca C, Jaunmuktane Z, Sonnen JA, Larivière R, Genís D, Molina Porcel L, Choquet K, Sakalla R, Provost S, Robertson R, Allard-Chamard X, Tétreault M, Reiling SJ, Nagy S, Nishadham V, Purushottam M, Vengalil S, Bardhan M, Nalini A, Chen Z, Mathieu J, Massie R, Chalk CH, Lafontaine AL, Evoy F, Rioux MF, Ragoussis J, Boycott KM, Dubé MP, Duquette A, Houlden H, Ravenscroft G, Laing NG, Lamont PJ, Saporta MA, Schüle R, Schöls L, La Piana R, Synofzik M, Zuchner S, Brais B. Deep Intronic FGF14 GAA Repeat Expansion in Late-Onset Cerebellar Ataxia. N Engl J Med. 2023 Jan 12;388(2):128-141. doi: 10.1056/NEJMoa2207406. Epub 2022 Dec 14. PMID: 36516086; PMCID: PMC10042577.
7. Rafehi H, Read J, Szmulewicz DJ, Davies KC, Snell P, Fearnley LG, Scott L, Thomsen M, Gillies G, Pope K, Bennett MF, Munro JE, Ngo KJ, Chen L, Wallis MJ, Butler EG, Kumar KR, Wu KH, Tomlinson SE, Tisch S, Malhotra A, Lee-Archer M, Dolzhenko E, Eberle MA, Roberts LJ, Fogel BL, Brüggemann N, Lohmann K, Delatycki MB, Bahlo M, Lockhart PJ. An intronic GAA repeat expansion in FGF14 causes the autosomal-dominant adult-onset ataxia SCA50/ATX-FGF14. Am J Hum Genet. 2023 Jan 5;110(1):105-119. doi: 10.1016/j.ajhg.2022.11.015. Epub 2022 Dec 8. Erratum in: Am J Hum Genet. 2023 Jun 1;110(6):1018. doi: 10.1016/j.ajhg.2023.05.005. PMID: 36493768; PMCID: PMC9892775.
8. Zheng ZH, Cao CY, Cheng B, Yuan RY, Zeng YH, Guo ZB, Qiu YS, Lv WQ, Liang H, Li JL, Zhang WX, Fang MK, Sun YH, Lin W, Hong JM, Gan SR, Wang N, Chen WJ, Du GQ, Fang L. Characteristics of tandem repeat inheritance and sympathetic nerve involvement in GAA-FGF14 ataxia. J Hum Genet. 2024 Sep;69(9):433-440. doi: 10.1038/s10038-024-01262-5. Epub 2024 Jun 12. PMID: 38866925.