Lebranchu, Pierre MD; Le Meur, Guylène MD, PhD; Magot, Armelle MD; David, Albert MD; Verny, Christophe MD, PhD; Weber, Michel MD, PhD; Milea, Dan MD, PhD

Autosomal dominant cerebellar ataxia (ADCA) is a rare heterogeneous group of diseases characterized by cerebellar symptoms, often associated with other neurological signs. To date, 20 associated genes have been identified, with the phenotypic spectrum ranging from isolated ataxia to multisystemic deficits. Overall, prevalence ranges from 2 to 7 per 100,000 people (¹).

Spinocerebellar ataxia type 1 (SCA1) usually is diagnosed in the 4th decade, with the most common neurologic signs being limb ataxia and dysarthria. Ophthalmological signs may initially include nystagmus and saccadic hypometria, eventually associated with ophthalmoparesis or mild optic neuropathy as the disease progresses. In SCA7, macular dystrophy can be associated with ataxia (²).

All SCA1 (³) and most of SCA7 mutations correspond to a CAG trinucleotide expansion within the gene’s coding region (<2-cm segment on 6p23 chromosome for SCA1 and 3p21.1-p12 for SCA7). Less often, SCA7 patients exhibit a large deletion within the gene. Both SCA1 and SCA7 mutations are usually caused by polyglutamine expansion, manifesting above a threshold of CAG repeats. There is a strong association between the clinical severity of the disease, the age of first symptoms, and the number of CAG repeats (⁴). Anticipation, which explains the younger age of onset of symptoms in successive intrafamilial generations, results from CAG repeat expansion upon transmission.

SCA7 has a specific association with retinal pathology. This degenerative retinopathy initially affects cones, before progressing to cone–rod dystrophy (²). Fundus examination demonstrates macular pigmentary changes, sometimes associated with mild temporal optic disc pallor (⁵). The gene encodes for a protein (ataxin-7), widely expressed in human retina but of unknown function (⁶). This protein may interact with the function of CRX (cone–rod homeobox), a known transcription factor implicated in human cone–rod dystrophy (⁷).

Other maculopathies associated with SCAs are very rare. To our knowledge, only 2 single case reports (^8,9) have been previously described a maculopathy in SCA1. In both cases, the retinal changes were caused by a late onset of clinically detectable pigmentary macular dystrophy, associated with abnormal electroretinography (ERG), revealing rod and cone dysfunction. The aim of this study was to report the first family with genetically identified SCA1 associated with a maculopathy.

METHODS

Four patients from the same family (Fig. 1) were included in the study. All had a complete ophthalmological examination and automated static perimetry, fundus photography, and optical coherence tomography (OCT). Patient 1 underwent additional testing, including microperimetry, visual evoked potentials (VEP), full-field ERG, and multifocal electroretinography (mfERG). Details of previous genetic and neurological examinations were extracted from medical charts with patient’s permission. After informed consent, DNA from family members was analyzed for SCA1 and SCA7 gene mutations (Table 1).

FIG. 1

TABLE 1

RESULTS

Patient 1

A 51-year-old man reported a 6-month history of bilateral, progressive painless visual loss. His medical history included surgical removal of a pituitary adenoma 9 years earlier with full recovery of visual fields. One year before vision loss, the patient was evaluated for mild ataxia and swallowing difficulties. Diagnosis of SCA1 was confirmed via molecular analysis with a 42 CAG repeat expansion in one allele. Visual acuity was 20/32, right eye, and 20/40, left eye. Lanthony desaturated 15-hue color testing was abnormal, with a green–red axis on the right side and without axis on the other side. Visual field testing disclosed bilateral central scotomas (Fig. 2). On funduscopy, there were small drusen around the fovea and mild temporal optic disc pallor (Fig. 3). Retinal autofluorescence revealed central pigment irregularities in both eyes. OCT disclosed disorganization of the macular photoreceptor layer bilaterally (Fig. 4) and thinning of the temporal retinal nerve fiber layer (RNFL) (Table 2). Microperimetry revealed a decrease of retinal sensitivity (threshold ranging from 0 to 8 dB) in the 2-mm central area, with a relative sparing of the perifoveal surrounding area (threshold ranging from 8 dB to 10 dB). Comparison between function and anatomy confirmed the decrease of sensitivity in the area where the photoreceptor layer was laminated (Fig. 5). Full-field ERG was normal but mfERG confirmed macular dysfunction in each eye. VEP disclosed mildly increased P100 latencies and decreased P100 amplitudes.

FIG. 2

FIG. 3

FIG. 4

TABLE 2

FIG. 5

Patient 2

A 73-year-old woman had a long history of ataxia and unexplained visual loss. She had suffered a slow deterioration in her ability to walk since her 50s and dysarthria since her 60s. No cognitive disabilities were reported. Visual acuity was 20/50, right eye, and 20/80, left eye. Visual field results were not reliable. Macular drusen were present bilaterally (Fig. 3), and retinal autofluorescence was suggestive of central atrophy. The optic discs were normal. Fluorescein angiography showed mild granular appearance of the perifoveal retinal pigmentary epithelium (RPE). Macular OCT revealed bilateral macular atrophy with an abnormal foveal lamination pattern between the external layers. Focal thickening of the retinal pigment epithelium complex was present in the right eye (Fig. 4). Measurement of macular and global RNFL thickness was normal (Table 2).

Patient 3

A 55-year-old woman, with severe ataxia that began 13 years earlier complained of bilateral, progressive painless visual loss. Serial neurological examinations revealed slowly progressive kinetic and static cerebellar syndrome, associated with tetrapyramidal signs and speech and swallowing difficulties. Marked ponto-cerebellar atrophy was detected on magnetic resonance imaging (MRI). Visual acuity was 20/200, right eye, and 20/100, left eye. Visual fields showed bilateral central scotomas. Funduscopy revealed only mild granular appearance of the foveal RPE (Fig. 3). There was RNFL thinning on OCT (Table 2) and abnormal foveal cavitation was visible, between the outer segment layer and the layer representing the junction of the inner and outer segments. The layer corresponding to the retinal epithelium complex exhibited local foveal thickening (Fig. 4).

Patient 4

A 52-year-old man, with no visual complaints, had a history of slowly progressive cerebellar–pyramidal syndrome. MRI showed marked cerebellar atrophy, and genetic analysis confirmed SCA1 diagnosis. Other causes of genetic spinocerebellar ataxia were excluded (SCA 2, 3, 6, 7, 12, and 17). Visual acuity was 20/20 bilaterally. Macular visual field testing revealed asymptomatic, asymmetric central scotomas. Funduscopy was normal (Fig. 3), but OCT revealed central alteration of the photoreceptor outer segment layer, with focal thickening (Fig. 4). RNFL thickness was normal (Table 2).

DISCUSSION

Optic neuropathy has been reported previously as a cause of visual impairment in SCA1. Abe et al (¹⁰) reported 6 SCA1 patients from 3 families with decreased visual acuity and optic atrophy. Using OCT, Stricker et al (¹¹) demonstrated significant thinning of temporal retinal nerve fibers, suggesting preferential involvement of the papillomacular bundle. Abnormalities of both latency and amplitude of VEPs also have been found in SCA1 patients (^12–14). In most of these studies, the retinal function was tested using full-field ERG; mfERG was not performed.

Among ADCA, SCA7 has been associated with retinal disease (¹⁵), ranging from minor retinal findings (slightly attenuated retinal arteries, mild temporal disc pallor, normal fundus autofluorescence (¹⁶)) to severe visual loss (¹⁷). Macular OCT in SCA7 patients has confirmed both quantitative macular thinning and qualitative altered foveal lamination of the photoreceptor layer at early stages of disease (^18,19). There is an increasing evidence of macular abnormalities occurring in ADCA patients, being 2.7 times more common in ataxic patients (²⁰). Stricker et al (¹¹) found decreased total macular volume with OCT in 6 SCA1 patients, but this was not statistically significant. Using OCT, Pula et al (²¹) found a significant thinning of macular volume at 3 mm in 7 SCA1 patients, and Vaclavik et al (²²) described a SCA1 patient with maculopathy and retinal dysfunction. Thurtell et al (⁹) reported a case of a genetically proven SCA1 patient with visual loss also secondary to rod–cone dystrophy, whose clinical findings were very similar to those of our patients: bilateral, progressive painless visual loss with macular drusen and mild pigmentary alterations on fundus examination. Other members of this reported family (⁹) exhibiting the SCA1 mutation also complained of visual loss, but no details of their fundus examination were reported.

We can only speculate why several related patients exhibiting the same SCA1 mutation were affected by a maculopathy, with heterogeneous intrafamilial expression. A common finding in SCA1 patients is phenotypic variability because of the number of CAG repeats among affected patients (²³). To explain the occurrence of both neurological and ophthalmological diseases in this family, the hypothesis of 2 independently segregated traits cannot be excluded, but the evidence is weak. The probability of an SCA1 family exhibiting an hereditary macular disease is equal to that of the general population, less than 1/40,000 (²⁴). The probability of 4 relatives harboring 2 separate autosomal dominant diseases is even lower. It is more likely that both diseases are genetically linked, either with 2 independent mutations nearby on the same chromosome or because of a “local effect” of one mutation on another gene nearby.

Pathogenesis of SCA1 is due partly to direct mutation of the gene (6p22,3), translated into an abnormal protein (ataxin-1) that has an abnormally long stretch of glutamine. This leads to polyQ protein aggregation in the cell and aberrant protein interactions, with specific transcriptional complexes in the nucleus (²⁵). Trinucleotide repeat disorder could exhibit other pathogenic manifestations, secondary to direct accumulation of the mutant messenger RNA in the nucleus, indirect gain of function (²⁶), or inhibition of adjacent gene (²⁷). In our patients, OCT demonstrated alterations of the external layers of the fovea, suggesting a loss of the structural integrity of the photoreceptors. Clinically, this foveolar lamination is compatible with an adult-onset vitelliform maculopathy, but the thinning of the whole macular area (including the surrounding perifoveolar retina) could be related to a cone dystrophy. Cone dystrophy is a clinically and genetically heterogeneous group, and no standard screening test has been developed. More than 10 different genes and loci have been identified in autosomal dominant cone dystrophy (https://sph.uth.edu/Retnet/sum-dis.htm#A-genes). Interestingly, a cluster of genes (GUCA1A, PRPH2) implicated in a spectrum of macular disease is located in 6p21.1, nearby SCA1 mutation. The gene GUCA1A (guanylate cyclase activator 1A) is implicated in cone dystrophy (COD3), which has phenotypic variability, even in the same family (^28,29). The gene PRPH2 (Peripherin 2) has been also identified in adult-onset vitelliform macular dystrophy (³⁰) or adult-onset foveomacular dystrophy (³¹). On inhibiting one of its promoters, CAG trinucleotide repeat could interfere with the regulation of one of these genes.

Establishing a definite association of a maculopathy with SCA1 is limited, in part, by our small patient sample size. More extensive study of relatives, especially those who do not have the SCA1 mutation, would be very informative. Other family members declined evaluation because most were free of symptoms and did not want genetic testing. However, none of them complained of neurological or visual symptoms. Whether there is truly a macular disorder associated with SCA1 awaits further study.

REFERENCES