US hereditary ataxia mutation frequencies compared to EFNS molecular testing guidelines

Objective: To describe the mutation frequencies in ataxia genes referred to a clinical laboratory for genetic testing and compare the data with the 2014 European Federation of Neurological Societies (EFNS) testing guidelines. Methods: Specimens (n=3,394) submitted for hereditary ataxia testing included a 20-gene comprehensive profile (n=2,165), a 14-gene autosomal dominant (AD) profile (n=812), or a 6-gene autosomal recessive (AR) profile (n=417). The genes associated with AD inheritance included ATN1 (DRPLA), ATXN1 (SCA1), ATXN2 (SCA2), ATXN3 (SCA3), SPTBN2 (SCA5), CACNA1A (SCA6), ATXN7 (SCA7), ATXN8OS (SCA8), ATXN10 (SCA10), PPP2R2B (SCA12), KCNC3 (SCA13), PRKCG (SCA14), TBP (SCA17), and AFG3L2 (SCA28). The genes associated with AR inheritance included APTX, FXN [Friedreich ataxia (FRDA)], POLG, SETX, SIL1, and TTPA. Repeat expansions were analyzed by PCR, repeat-primed PCR, and Southern blot assay. Sequence variants were analyzed by Sanger sequencing. Results: The comprehensive profile yielded positive results for one or more genes in 254 (11.7%) specimens including 204 (80.3%) AD, and 50 (19.7%) AR. Most positive findings (92.9%) were repeat expansions with 7.1% pathogenic variants. Expansions in SCA8 (18.1%) were the most common positive results followed by SCA3 and SCA6 (16.1% each), FRDA (15.7%), SCA2 (13.8%), SCA1 (6.7%), SCA10 (2.4%), SCA17 (2.0%), SCA7 (1.6%), and SCA12 (0.4%). SCA1, 2, 3, 6, 7, and 17 accounted for 81.6% of positive results in the AD profile while FRDA accounted for 78.9% in the AR profile. Conclusion: The mutation frequencies reported, except for SCA8 and SCA10, are consistent with the EFNS genetic testing recommendations for ataxia. Correspondence to: Joseph J. Higgins, MD, FAAN, Quest Diagnostics, 200 Forest Street, 2nd Floor, Marlborough, MA 01752 USA, Tel: 774.843.3246, Fax: 508.659.0820, E-mail: Joseph.J.Higgins@QuestDiagnostics.com


Introduction
The hereditary ataxias are characterized by progressive incoordination that mainly affects ambulation but often involves the upper extremities, speech, and eye movements [1]. Prevalence varies across populations with estimates ranging from 1 to 9 per 100,000 people [2][3][4][5]. Inheritance patterns can be autosomal dominant (AD), autosomal recessive (AR), X-linked, or maternal if part of a mitochondrial genetic syndrome [1]. The most frequent inheritance pattern is AD, which is generally associated with an adult onset. The most common spinocerebellar ataxia (SCA) loci 1, 2, 3, 6, and 7 are characterized by CAG trinucleotide repeat expansions in their respective genes [1]. Among the AR ataxias, which usually have a childhood onset, Friedreich ataxia (FRDA) is the most common with a disease frequency of 1:20,000 to 1:50,000 [1].
Diagnosing specific ataxia subtypes has important treatment and prognostic implications. The diagnosis typically requires ataxia, a detailed family history, the identification of additional clinical features, and neuroimaging. Molecular testing plays a prominent diagnostic role because of the difficulty differentiating subtypes based on clinical grounds alone [1]. In 2014, the European Federation of Neurological Societies (EFNS) recommended initial genetic testing for AD ataxia, including SCA1, 2, 3, 6, 7, and 17 and one gene, FXN (FRDA); it recommended testing for AR ataxia when cerebellar atrophy is absent on brain MRI [6,7]. These seven loci (SCA1, 2, 3, 6, 7, 17, and FRDA) represent the majority of hereditary ataxias in Europe and more than 60% of AD cases worldwide [8]. The EFNS did not include SCA10 or dentatorubral-pallidoluysian atrophy (DRPLA) in their recommendations because pathogenic variants in these genes are restricted to patients with ancestral ties to Latin American or Asian countries [6,7,9]. Although SCA8 accounts for 2% to 5% of AD ataxia, the EFNS did not recommend testing for this subtype in initial screening because the causative expanded allele has a low penetrance [6,7,10].
The EFNS recommendations for molecular diagnosis of ataxia are designed to assist European neurologists in using genetic testing to diagnose specific ataxia subtypes [6,7]. To evaluate the EFNS recommendations for testing specific ataxia genes, we retrospectively assessed the frequencies and types of mutations in specimens referred to a clinical laboratory in the USA.

Methods
This study used de-identified laboratory results from consecutive specimens submitted for ataxia genetic testing between 2011 and 2014. Less than 1% of specimens were received from outside the USA. The results from sequencing and repeat expansion assays were extracted from an internal database by gene name without identifying information. In all cases, healthcare providers ordered genetic testing on a standardized requisition form and provided the clinical indications and ICD codes. Detailed clinical information and the relatedness between patients were unavailable.
The study included 3,394 patient specimens, 2,165 of which were tested using a comprehensive profile of 20 genes: 14 associated with AD and 6 associated with AR ataxias ( Table 1). The remaining specimens were tested with a 14-gene AD profile (n=812) or a 6-gene AR profile (n=417) ( Table 1). Non-repeat expansions were detected by DNA sequencing, which was performed by PCR amplification of purified genomic DNA, followed by Sanger DNA sequencing of the coding regions of all the genes except POLG and KCNC3. POLG gene exons 7, 13, and 21 and KCNC3 gene exon 2 were sequenced and analyzed. Abnormal sequence variants were confirmed using bi-directional sequencing.
Repeat expansions were detected by PCR amplification of the repeat region followed by high-resolution electrophoresis to determine the number of tandem repeats in each allele. Repeat primed assay or Southern blot analysis was used to confirm the homozygosity of normal alleles. A two-tailed Fisher's exact test was used to compare the differences in the positive rates between the EFNS [7] and the clinical laboratory datasets.

Discussion
The positive rates in this study may be overestimates due to the lack of detailed clinical information and the possibility that more than one family member was tested. Overall, the data agrees with the 2014 EFNS guidelines for genetic testing for ataxia with two exceptions [6,7]. The first exception is that SCA10 is not included in the EFNS recommendation because it has been only been reported in individuals from Latin American or of Native American descent [9]. However, in our study population, SCA10 accounted for 2.4% of our positive specimens; we noted a regional clustering in California and Texas likely due to founder effects. The second exception is that repeat expansions in SCA8 were the most common positive result (18.1%) in the comprehensive profile. The high rate of co-occurrence with pathogenic variants in other AD ataxia genes in this study (10/46 specimens: 4 SCA8 expansions on both alleles, 4 with other repeat expansions, and 2 with a sequencing variant in POLG) suggest that repeat expansions in SCA8 may not cause ataxia directly in some cases. In addition, not all SCA8 expansions are associated with progressive ataxia suggesting the co-occurrence is due to its low penetrance [8,10]. The role of SCA8 as a genetic modifier and the mechanisms for incomplete penetrance require further study.
The comprehensive profile detected AD mutations in nearly 10% of ataxia patients and AR mutations in only 2.3% in our heterogeneous sample. When an AD or AR profile was ordered, the frequency of positive results roughly doubled. This suggests that the AD-only or AR-only profiles should be performed first in patients with a known inheritance pattern. Table 1. Ataxia genes (loci) included in multigene profiles tested in a US clinical laboratory   [7]. The frequencies of positive results in the comprehensive and AR profiles were similar to the EFNS data. The frequency of positive results for the AD profile was significantly (P<0.001) higher compared to the EFNS data.