Purpose
To report the clinical and genetic findings of 9 Japanese patients with autosomal recessive bestrophinopathy (ARB).
Design
Retrospective, multicenter observational case series.
Methods
Nine ARB patients from 7 unrelated Japanese families that were examined in 3 institutions in Japan were studied. A series of ophthalmic examinations including fundus photography, spectral-domain optical coherence tomography, fundus autofluorescence, electrooculography (EOG), electroretinography, and the results of genetic analysis were reviewed.
Results
Genetic analyses identified 7 pathogenic variants in BEST1 including 2 novel variants, c.478G>C (p.A160P) and c.948+1delG. Homozygous variants were found in 4 families and compound heterozygous variants were found in 3 families. Two patients were diagnosed as ARB only after the whole exome sequencing analyses. The Arden ratio of the EOG was less than 1.5 in all 7 patients tested. Vitelliform lesions typical for Best vitelliform macular dystrophy were not seen in any of the patients. Seven patients shared some of the previously described features of ARB: subretinal deposits, extensive subretinal fluid, and cystoid macular edema (CME). However, the other 2 patients with severe retinal degeneration lacked these features. Focal choroidal excavations were present bilaterally in 2 patients. One case had a marked reduction of the CME and expansion of subretinal deposits over an 8-year of follow-up period.
Conclusions
Japanese ARB patients had some but not all of the previously described features. Genetic analyses are essential to diagnose ARB correctly in consequence of considerable phenotypic variations.
The BEST1 gene was first identified as the gene responsible for Best vitelliform macular dystrophy (BVMD) in 1998 (OMIM *607854; also known as VMD2 ). BEST1 encodes the bestrophin-1 protein, which consists of 585 amino acids and is located in the basolateral plasma membrane of the retinal pigment epithelial (RPE) cells.
Variants of the BEST1 gene cause a wide range of retinal diseases including BVMD, autosomal dominant vitreoretinochoroidopathy (ADVIRC), autosomal recessive bestrophinopathy (ARB), adult-onset vitelliform macular degeneration (AVMD), retinitis pigmentosa, and the microcornea, retinal dystrophy, cataract, and posterior staphyloma (MRCS) syndrome. BVMD is the most well known BEST1 -associated retinal dystrophy, and its clinical features were first described by Best in 1905. BVMD is usually caused by autosomal dominant variants of the BEST1 gene with often incomplete penetrance.
As for ARB, Schatz and associates described a patient with retinopathy and biallelic variants of BEST1 in 2006. In 2008, Burgess and associates reported that ARB was a distinct retinal disease associated with biallelic variants in the BEST1 gene. ARB is caused by either compound heterozygous or homozygous mutations in the BEST1 gene.
ARB is associated with a wide spectrum of fundus abnormalities. Multiple white flecks scattered over the posterior pole of the retina are clearly detected by fundus autofluorescence (FAF). Optical coherence tomography (OCT) images show an accumulation of fluid within and/or beneath the neurosensory retina. The electrooculography (EOG) light peak is absent or markedly reduced. Central visual disturbances and reduced amplitudes of electroretinograms (ERGs) are also associated with ARB. However, the visual acuity and full-field ERGs may remain relatively normal for a long period of time in some cases. The results of several case studies have shown various phenotypes for ARB since the Burgess report. Some cases had marked subretinal deposits while others had mainly RPE atrophy and scar lesions. Because of the rarity of the disease and variable phenotypes, ARB has been reported mostly in a small number of cases and definitive clinical features of ARB have not been determined.
Thus, the purpose of this study was to describe the clinical and genetic findings of 9 Japanese patients with ARB. We show that the phenotypes of these patients varied considerably and examine the alterations of the FAF and spectral-domain OCT (SDOCT) images during the course of the disease in 1 patient followed for 8 years.
Methods
This is a retrospective, multicenter observational case series. The procedures used in this study adhered to the tenets of the Declaration of Helsinki, and an informed consent was obtained from the patients and relatives. An approval of the protocol of this study was obtained from the institutional review board of each institution (approval numbers: Nagoya University 2015-0072, Jikei University 24-231 6997, National Hospital Organization Tokyo -Medical Center R11-087 and R14-050).
The medical records of 9 patients from 7 unrelated Japanese families who were confirmed to carry biallelic pathogenic variants of the BEST1 gene were reviewed. The medical histories and the results of the clinical examinations of the patients and family members including the best-corrected visual acuity (BCVA), intraocular pressure measurements, slit-lamp examinations, and ophthalmoscopy were reviewed. Full-field ERGs were recorded from all patients, and EOGs were recorded from 7 patients, and these tests were performed conforming to the guidelines of the International Society for Clinical Electrophysiology of Vision (ISCEV) standards. Patients also underwent fundus color photography, FAF imaging (Spectralis HRA; Heidelberg Engineering, Heidelberg, Germany), and SDOCT imaging (Spectralis; Heidelberg Engineering, or Cirrus HD-OCT; Carl Zeiss Meditec, Dublin, California, USA). Four patients (NA0050, JU0209, NA0044, and NA0062) underwent fluorescein angiography (FA) and 1 patient (NA0044) also underwent indocyanine green angiography (IA; Spectralis HRA; Heidelberg Engineering).
Identification of BEST1 Variants
Mutations detection was performed by direct sequencing and/or whole exome sequencing. Primal direct sequencing was used in 7 patients from 6 families. Primal whole exome sequencing was performed on 2 patients (NA0044 and NA1044) from a family, followed by direct sequencing for confirmation. Co-segregation analysis was performed with direct sequencing in 6 families (family numbers I–IV; Supplemental Figure , available at AJO.com ). Blood samples were obtained from all probands and their family members when available. Genomic DNA was extracted from peripheral blood leukocytes by a Gentra Puregene Blood Kit (Qiagen, Hilden, Germany). Direct sequencing of the BEST1 gene was performed according to an established method (mRNA reference sequence: NM_004183.3). The protein-coding exons, exons 2-11, of the BEST1 gene were sequenced with an automated sequencer (3730xl DNA Analyzer; Applied Biosystems, Foster City, California, USA) with a BigDye Terminator Kit (V3.1, Applied Biosystems). Whole exome sequencing and targeted sequence analysis for the BEST1 gene were done according to the published protocol of the National Institute of Sensory Organs, a customized analysis protocol for the Japanese population. In silico bioinformatic analyses were performed to predict the pathogenicity of all of the identified BEST1 variants that have been reported. The details of the in silico molecular genetic analyses are described as supplemental information (Supplemental Material available at AJO.com ).
Results
The demographic and clinical characteristics of the 9 patients at the time of the diagnosis of ARB are shown in the Table . The pedigrees of the 9 patients are shown in the Supplemental Figure . All the patients were Japanese, and 2 families were consanguineous (Families I and VI). The mean age was 31 ± 10 years with a range of 14–45 years at the most recent examination. The age at the time of diagnosis varied from 12 to 35 years. The initial symptom of most of the patients was blurred vision, but 4 patients had good visual acuity of 20/20 or better at that time. The patients also complained of other symptoms, such as metamorphopsia, photophobia, night blindness, and floaters. One patient (KA160) had no symptoms and was found unexpectedly by fundus examination. The visual acuity at the time of diagnosis by gene analyses ranged from 20/20 to 20/200. Three patients (NA1050, JU773, and KA160) had visual acuity of 20/20 in at least 1 eye. The refractive errors ranged from hyperopia to mild myopia, with a range from −2.5 to +2.0 diopter. None of the patients had a shallow anterior chamber, and none had angle-closure glaucoma. The Arden ratio, light peak/dark trough, was severely reduced to less than 1.2 in all 7 patients who underwent EOG examinations. The amplitudes of the full-field scotopic and photopic ERGs were within normal range in 4 patients (NA1050, JU0773, NA0062, and KA160) and were reduced by different degrees in the other 5 patients. The 3 patients with poor visual acuity (NA1044, JU0209, and JU0645) had severely reduced ERGs. Fundus examinations of the parents were performed in 6 families (Families I–VI). The father of the JU0645 had vitelliruptive macular degeneration but the other parents showed no specific fundus abnormalities that would be considered to be related to BVMD or ARB.
| Family Number, Case ID, Age (y), Sex | Age at Onset (y) | Symptoms at Onset | BCVA (OD,OS) | Spherical Equivalent (D) (OD,OS) | Arden Ratio of EOG (OD,OS) | Amplitudes of Full-Field ERG | Nucleotide Changes of BEST1 Gene (Allele 1, Allele 2) | Amino Acid Changes (Allele 1, Allele 2) |
|---|---|---|---|---|---|---|---|---|
| I, NA1050, 27, F | 22 | Floaters | 20/20, 20/25 | +0.5, 0 | 1.1, 1.1 | Normal | c.763C>T, c.763C>T | p.R255W, p.R255W |
| I, NA0050, 23, M | 19 | Night blindness, blurred vision | 20/40, 20/32 | +2.0, +1.5 | 1.0, 1.0 | Moderate decrease | c.763C>T, c.763C>T | p.R255W, p.R255W |
| II, JU0209, 35, M | 12 | Blurred vision | 20/100, 20/200 | −0.5, −2.5 | 1.0, 1.0 | Severe decrease | c.763C>T, c.763C>T | p.R255W, p.R255W |
| III, NA1044, 45, M | 35 | Blurred vision | 20/80, 20/50 | +1.0, +1.0 | N/A | Severe decrease | c.73C>T, c.584C>T | p.R25W, p.A195V |
| III, NA0044, 42, F | 41 | Metamorphopsia, floaters | 20/32, 20/25 | 0, −0.25 | N/A | Mild decrease | c.73C>T, c.584C>T | p.R25W, p.A195V |
| IV, NA0062, 28, F | 20 | Photophobia, blurred vision | 20/25, 20/25 | 0, −0.75 | 1.1, 1.1 | Normal | c.102C>T, c.584C>T | p.G34G, p.A195V |
| V, JU0773, 14, F | 14 | Metamorphopsia | 20/20, 20/20 | +1.0, +0.5 | 1.2, 1.2 | Normal | c.478G>C, c.584C>T | p.A160P, p.A195V |
| VI, JU0645, 38, M | 12 | Blurred vision | 20/125, 20/100 | −0.75, 0 | 1.0, 1.1 | Severe decrease | c.908A>G, c.908A>G | p.D303G, p.D303G |
| VII, KA160, 23, F | 17 | No complaint | 20/20, 20/20 | 0, +0.25 | 1.0, 1.0 | Normal | c.948+1delG, c.948+1delG | N/D |
Analyses of BEST1 Gene Mutations
The pathogenic variants of BEST1 found in our cohorts are shown in the Table . Co-segregation analyses revealed heterozygous/homozygous status; 4 families had homozygous variants and 3 families had compound heterozygous variants. Seven different variants were found in our cohorts. Variants p.A195V and p.G34G (c.102C>T) have been reported to cause ARB with another variant in the compound heterozygous state. Four variants, p.R25W, p.A195V, p.R255W, and p.D303G, have been reported to be causative mutations for autosomal dominant BVMD. p.A160P (c.478G>C) and c.948+1delG were found in this study as new candidates of pathogenic variants. p.A160P was not listed in 5 different single nucleotide polymorphism databases, viz, dbSNP, EVS, ExAC, HGVD, and HTD. The in silico bioinformatic program predicted the pathogenicity of the new variant, p.A160P (PolyPhen-2) as possibly damaging (0.777), SIFT as damaging (0.001), and PROVEAN as deleterious (−4.78). Variant c.948+1delG was predicted to result in a loss of a genuine splice donor site by the in silico prediction performed by the splice site prediction tools (HSF, NNSPLICE, and NetGene2).
Findings From Fundus Imaging
The fundus color photographs, FAF images, and OCT images of the left eye of the 9 patients are shown in Figure 1 . The patients had similar abnormalities in both eyes, although the abnormalities varied considerably among the patients. Six of the patients had different degrees of central yellowish subretinal deposits in the fundus color images (NA1050, NA0050, NA0044, NA0062, JU0773, and KA160). Vitelliform lesions, which are usually seen in eyes with BVMD, were not present in any of the patients, although the fundus of 2 cases resembled that of the vitelliruptive stage of BVMD, with massive subretinal deposits (NA0062 and JU0773). Three patients (JU209, NA1044, and JU0645) had prominent RPE atrophy, which appeared as diffuse grayish green scars over the posterior pole.
The FAF images showed different types of abnormalities with a mixture of hyper- and hypoautofluorescence throughout the posterior pole in all of the patients. Subretinal deposits were detected as hyperautofluorescent spots in the FAF images and were more recognizable than in the color photographs of the fundus. The areas of atrophic RPE were detected as hypoautofluorescent areas.
The SDOCT images showed different kinds of abnormalities. Cystoid edema was seen in 4 patients (NA1050, NA0050, JU0209, and NA0062) in the SDOCT images. The edema was diffusely spread over the macular area and mainly within the inner nuclear layer (INL).
Four patients (NA0044, NA0062, JU0773, and KA160) had serous retinal detachments with elongated photoreceptor outer segments. One patient (JU0209) also had serous retinal detachment, but the ellipsoid zone (EZ) was absent, with a severe thinning of the outer nuclear layer. The EZ was also not observed in 2 patients, NA1044 and JU0645, who had severe RPE atrophy. A disruption of the EZ was seen adjacent to an area of serous detachment in 2 patients (NA0044 and KA160) and under the cystoid edema in 2 patients (NA1050, NA0050). The SDOCT images also showed focal choroidal excavations bilaterally in 2 patients (KA160 in Figure 1 and NA0044 in Figure 2 ). One patient (NA0044) had multiple choroidal excavations and pigment epithelial detachments that corresponded with the areas of increased choroidal vascular permeability in the IA images ( Figure 2 ). Fluorescein angiography showed patchy hyperfluorescence over the posterior pole in the early phase in 4 cases, which indicated the presence of window defects owing to the RPE atrophy. A pooling of fluorescein dye in the cystoid spaces was seen in the late phase in 2 patients (NA0044 and NA0062; images of NA0044 are shown in Figure 2 ).
Case With 8-Year Follow-up
Case NA0062 had marked changes in the fundus images over an 8-year period in spite of only minor changes of her BCVA, which decreased from 20/20 to 20/25 ( Figure 3 ). The yellowish subretinal deposits shown in the fundus color images in 2007 were decreased in 2015. On the other hand, the areas of hyperautofluorescence were expanded in 2015. Cystoid macular edema was present in both INL and outer nuclear layer (ONL) in 2007, but the edema in the ONL disappeared almost completely in 2015.
