Clinical and Genetic Characteristics of Korean Occult Macular Dystrophy Patients

Abstract

Purpose.: We investigated the genetic characteristics of retinitis pigmentosa 1-like 1 (RP1L1) gene in patients with occult macular dystrophy (OMD) and identified genotype-phenotype correlations.

Methods.: We sequenced the entire exons and flanking regions of the RP1L1 gene in 19 Korean OMD patients. Detailed retinal morphologic abnormalities were evaluated using spectral domain optical coherence tomography and infrared reflectance imaging. Clinical features, retinal morphologic abnormalities, and disease progression were compared among the subtypes separated on the basis of genotypes.

Results.: Ten of 19 (52.6%) patients had RP1L1 mutations: 7 (36.8%) had the previously reported mutation, p.Arg45Trp, and the other 3 (15.8%) had novel variants (p.Gln2311Pro, p.Ser676Cys, and p.Asp1425His). p.Gln2311Pro, p.Asp1425His, and p.Ser676Cys were predicted as pathogenic in 3, 2, and none of 5 computational assessment tools, respectively. Photoreceptor disruption across the fovea was symmetric in all eyes with RP1L1 mutations, but asymmetric in 3 eyes without the mutations. Symmetric round hyporeflectance centered on the fovea in infrared reflectance images was seen in all patients with the p.Arg45Trp mutation, but in only 2 of 12 patients without the mutation. The patients with p.Arg45Trp showed greater extent of photoreceptor disruption than those without the mutation. Progression of photoreceptor disruption was remarkable in patients with RP1L1 mutations, but insignificant in those without.

Conclusions.: OMD is clinically and genetically heterogeneous, with different morphologic features and progression in outer retinal pathology according to RP1L1 gene mutations, indicating the genotype-phenotype correlation in patients with OMD.

Introduction

Occult macular dystrophy (OMD) is a hereditary macular disease characterized by progressive visual decline. ^1,2 The diagnosis of OMD is made based on a finding of abnormal macular function on the focal macular electroretinogram (ERG) with normal fundus appearance, as well as normal fluorescein angiogram and full-field ERG. A recent genetic study identified dominant mutations in the retinitis pigmentosa 1-like 1 (RP1L1) gene in Japanese OMD families, and its product, RP1L1, was shown to be localized in rod photoreceptors in the mouse, and in rod and cone photoreceptors in the Cynomolgus monkey. ^3,4 A c.133 C>T with a substitution of tryptophan (TGG) for arginine (CGG) at amino acid position 45 (p.Arg45Trp) of the RP1L1 gene was reported in 4 Japanese OMD families, ³ and another mutation (p.S1199C) in RP1L1 was reported recently in a sporadic OMD patient. ⁵ However, as mutations in RP1L1 have been found only in Japanese families and 1 sporadic patient, genetic studies in a larger number of OMD patients of different ethnicities are required. In a recent report by Scholl et al., there were no RP1L1 mutations in a white family of European descent with OMD. ⁶  

Spectral-domain optical coherence tomography (SD-OCT) can reveal characteristic alterations of the retina to an unprecedented quasi-histologic level in various retinal diseases, including OMD. ^1,2,7–12 Using this imaging modality, several studies have demonstrated morphologic deformities in the photoreceptor layer and suggested that photoreceptor abnormality is the main pathology of OMD. ^7,9–12 Additionally, fundus infrared (IR) reflectance and IR-autofluorescence imaging have been used to delineate structural abnormalities, and provide novel information on retinal pathology in various macular and retinal diseases. ^13–19 From these imaging modalities, detailed information on pathologic changes in the photoreceptor layers in OMD may be obtained in a noninvasive manner. 

The purpose of our study was to examine the genetic characteristics of RP1L1 in patients with OMD and correlate these with the outer retinal pathology revealed using multimodal imaging modalities, including SD-OCT and fundus IR reflectance. More specifically, we investigated the genotype-phenotype correlation in OMD by exploring the relationship of the RP1L1 genotype with pathologic features, disease progression, and visual/electrophysiologic function. 

Methods

Patients and Clinical Data Collection

The diagnosis of OMD was made based on progressive decline in visual acuity; no abnormal findings on fundus photography, fluorescein angiography (FA), and full-field standard ERG; and reduced foveal multifocal ERG (mfERG) response, defined as local amplitudes significantly lower than those of an age-matched normal population (Fig. 1). This study used genomic DNA samples from 19 patients with OMD who were diagnosed with OMD between January 2008 and June 2012. Approval to conduct this study was obtained from the Institutional Review Board (IRB) of Seoul National University Bundang Hospital, and our study adheres to the tenets of the Declaration of Helsinki. Informed consent was obtained from all patients before genetic analysis. 

Figure 1

 

Results of ocular examinations in the right eye of patient 1. (A) Fundus photograph and (B) fluorescein angiograph show normal appearance. (C) Visual field test obtained by Humphrey field analyzer central 24-2 threshold testing shows central scotoma. (D) Trace arrays of multifocal ERG tested with 61 hexagonal stimuli show decreased responses in the central locus of the retina. (E) Full-field standard ERG demonstrates normal response.

All patients underwent complete ophthalmic examinations, including best-corrected visual acuity (BCVA), fundus examination and photography, FA, full-field ERG, and mfERG (VERIS II; ElectroDiagnostic Imaging 45, Inc., San Francisco, CA). Full-field ERG was performed using procedures proposed by the International Society for Clinical Electrophysiology of Vision (ISCEV), and mfERG was performed using 61 scaled hexagons with procedures conforming to ISCEV guidelines. 

Imaging of Retinal Abnormalities

High-resolution macula imaging was performed using combined confocal scanning laser ophthalmoscopy (cSLO) and SD-OCT (Spectralis OCT; Heidelberg Engineering, Heidelberg, Germany). Fundus IR reflectance (λ = 830 nm; field of view, 30° × 30°; image resolution, 768 × 768 pixels) was obtained from all patients with simultaneous SD-OCT scanning (λ = 870 nm; acquisition speed, 40,000 A-scans per second; scan depth, 1.8 mm; digital depth resolution, approximately 3.5 μm per pixel). ¹⁴ Automated eye tracking and image alignment based on cSLO images allowed for the correlation of cSLO and SD-OCT findings. 

Image Analysis

All images were reviewed independently with respect to structural integrity and the distribution of photoreceptor disruption in a masked fashion by 2 retina specialists (SJW and SJA). If there were any discrepancies in the judgment of qualitative parameters, the disagreement was solved by consensus and arbitrated by a third reviewer. Layer-by-layer morphologic analysis was performed to identify the extent and distribution of photoreceptor disruption. Using the digital caliper tool in the Spectralis OCT system (Heidelberg Engineering), the photoreceptor thickness (between the inner borders of the external limiting membrane [ELM] and the RPE ²⁰ ) was measured at the central fovea. The progression of the disease was evaluated by comparing SD-OCT images taken at the time of diagnosis and at the final visit in patients with follow-up periods ≥ 12 months. 

Genetic Analysis

RP1L1 sequences, including the entire exons and flanking regions, were analyzed. Genomic DNA was extracted from blood using the Puregene DNA isolation kit (Gentra Systems, Inc., Minneapolis, MN). Gene-specific polymerase chain reaction and direct sequencing using a DNA analyzer (3730xl; Applied Biosystems, Foster City, CA) were performed as described previously. ⁵ The identified mutations and coding polymorphisms were assayed in chromosomes from 180 healthy Korean individuals with direct sequencing of the RP1L1 gene. Novel variants were defined as those identified not in normal controls, but in patients with OMD, other than the previously reported mutations c.133 C>T ³ and c.3596 C>G ⁵ . The effect of a missense mutation on the encoded protein was predicted with the PolyPhen-2 (in the public domain at http://genetics.bwh.harvard.edu/pph2/), SIFT (in the public domain at http://sift.jcvi.org/), PMut (in the public domain at http://mmb2.pcb.ub.es:8080/PMut/), AGVGD (in the public domain at http://agvgd.iarc.fr), and MutationTaster (in the public domain at http://www.mutationtaster.org) online tools. 

Data Analysis

Clinical characteristics, such as age, sex, inheritance pattern, laterality, BCVA, refractive status, and symptom duration, were described in subgroups divided based on the RP1L1 genotype. Morphologic features, and quantitative data from the SD-OCT and IR findings also were compared among the genotype subgroups. 

The Shapiro-Wilk test was used to assess the normality of distributions of continuous variables. According to the results on Shapiro-Wilk tests, statistical analyses for comparing visual and electrophysiologic function were performed using the ANOVA or Kruskal-Wallis test. P values less than 0.05 were considered statistically significant. Statistical analyses were performed using SPSS for Windows (Ver. 17.0, Statistical Package for the Social Sciences; SPSS, Inc., Chicago, IL). 

Results

Clinical and Genetic Characteristics of Patients With OMD

The data on demographic, clinical, and genetic characteristics of the 19 patients (10 male and 9 female) are shown in Table 1. The mean age was 34.2 ± 17.6 years (range, 8–71 years). The mean follow-up period was 16.6 ± 7.9 months (range, 3–26). Four patients showed unilateral OMD, whereas other 15 patients had bilateral involvement. The BCVA in the affected eye ranged from 20/200 to 20/25 (median, 20/80). Slit-lamp and fundus examination revealed no remarkable findings, except for the presence of cataract in the right eye in patient 19. Eight patients had autosomal dominant inheritance, whereas 9 cases were sporadic, and the data on inheritance were not available in 2 patients. Patients 3, 4, and 19 were from the same family, and patients 6 and 7 were siblings. The other patients were unrelated. 

Table 1

 

Clinical and Genetic Characteristics of 19 Patients With OMD

Patient No./ Sex	Genotype of RP1L1	Age	Inheritance	Laterality	Visual Acuity OD	Visual Acuity OS	SE OD	SE OS	Sx Duration, y	PR Disruption
										Symmetricity Across the Fovea	Symmetricity Between the Eyes	Distribution	PR Thickness, μm	Involved PR Layers
1/M	c.133C>T (p.Arg45Trp)	38	AD	Bilateral	20/200	20/200	−1.38	−1.63	10	Symmetric	Symmetric	More severe toward fovea	51	COST, IS-OS, ELM
3/M		21			20/60	20/60	−4.50	−5.00	9				66	COST, IS-OS, ELM
4/M		8			20/100	20/150	−1.75	−2.50	5.5				50	COST, IS-OS, ELM
6/M		34			20/200	20/200	−6.00	−6.25	6				58	COST, IS-OS, ELM
7/F		36			20/50	20/40	−0.88	−1.00	2				66	COST, IS-OS
14/M		19			20/100	20/100	−3.63	−3.50	3				59	COST, IS-OS, ELM
19/M		51			20/200	20/40	−0.88	−0.25	3				65	COST, IS-OS
8/F	c.6932A>C (p.Gln2311Pro)	47	Unknown	Bilateral	20/40	20/60	−0.38	−0.38	1	Symmetric	Asymmetric	More severe toward fovea	48	COST, IS-OS
15/F	c.2026A>T (p.Ser676Cys)	58	Unknown	Bilateral	20/200	20/30	−0.38	+0.38	3				45	COST, IS-OS
17/F	c.4273G>C (p.Asp1425His)	53	Spo	Unilateral	20/20	20/100	+1.25	+0.50	1	59	COST, IS-OS
2/F	No mutation	33	Spo	Bilateral	20/120	20/40	−3.50	−3.50	2	Symmetric OD, asymmetric OS	Asymmetric	More severe in nasal retina OS	56	COST, IS-OS
5/M		11	Spo	Bilateral	20/60	20/50	−0.50	+0.63	2	Symmetric	Symmetric	More severe toward fovea	33	COST, IS-OS
9/F		50	Spo	Unilateral	20/100	20/25	−1.50	0.00	1	Symmetric	Asymmetric	More severe toward fovea	66	COST, IS-OS
10/F		15	Spo	Unilateral	20/30	20/20	−2.00	−2.13	1	Asymmetric	Asymmetric	More severe in temporal retina	70	COST, IS-OS
11/M		24	Spo	Bilateral	20/80	20/100	−1.50	−2.50	2	Symmetric	Symmetric	More severe toward fovea	55	COST, IS-OS
12/F		40	Spo	Bilateral	20/100	20/100	−3.50	−3.50	5	Symmetric	Asymmetric	More severe toward fovea	55	COST, IS-OS
13/F		26	AD	Bilateral	20/200	20/25	−0.38	−0.25	0.33	Symmetric	Asymmetric	More severe toward fovea	52	COST, IS-OS
16/M		71	Spo	Bilateral	20/40	20/25	−0.63	+0.50	2	Symmetric	Asymmetric	More severe toward fovea	83	COST
18/M		14	Spo	Unilateral	20/15	20/40	+0.75	+1.00	4	Asymmetric	Asymmetric	Irregular	83	COST

 Patients 3, 4, and 19 are from the same family, and patients 6 and 7 are siblings. Other patients are unrelated. AD, autosomal dominant; COST, cone outer segment tips; ELM, external limiting membrane; IS-OS, inner segment-outer segment junction; OD, right eye; OS, left eye; PR, photoreceptor; Spo, sporadic; SE, spherical equivalent; Sx, symptom.

Genetic mutations of RP1L1 also are presented in Table 1. Mutations in the RP1L1 gene were found in 10 of 19 (52.6%) patients. A c.133C>T in exon 2 with a substitution of tryptophan for arginine at amino acid position 45 (p.R45W), previously identified as the dominant mutation in a Japanese OMD family, ³ was the most common mutation (7 of 19, 31.3%) in our cohort. Three novel missense mutations of the gene, not found in normal controls, were found in 3 of 19 (15.8%) patients. The 3 novel variants were c.6932A>C, c.2026A>T, and c.4273G>C in exon 4, resulting in the substitution of proline for glutamine, cysteine for serine, and histidine for aspartate at amino acid positions 2311, 676, and 1425, respectively. In 9 of 19 (47.4%) patients, however, no mutation was detected in the RP1L1 gene. Table 2 shows the results of in silico prediction of the pathogenicity of 3 novel variants. p.Q2311P and p.D1425H were predicted as pathologic in 3 (SIFT, PMut, and AGVGD) and 2 (PolyPhen-2 and PMut) tools, respectively. However, p.S676C was not predicted as pathogenic in any of the 5 prediction tools. 

Table 2. 

 

In Silico Prediction of Novel Variants in the RP1L1 Gene

Patient No.	Variant	PolyPhen-2, Benign/Possibly Damaging/Probably Damaging	SIFT, Tolerated/Deleterious	PMut, Neutral/Pathologic	Mutation Taster, Polymorphism/Mutation	AGVGD, C0 (less likely)–C65 (most likely)
8	c.6932A>C (p.Q2311P)	Benign	Deleterious	Pathologic	Polymorphism	Class C65 (most likely)
15	c.2026A>T (p.S676C)	Benign	Tolerated	Neutral	Polymorphism	Class C0 (less likely)
17	c.4273G>C (p.D1425H)	Probably damaging	Tolerated	Pathologic	Polymorphism	Class C0 (less likely)

 PolyPhen-2 (available in the public domain at http://genetics.bwh.harvard.edu/pph2/); SIFT (available in the public domain at http://sift.jcvi.org/); PMut (available in the public domain at http://mmb2.pcb.ub.es:8080/PMut/); Mutation Taster (available in the public domain at http://www.mutationtaster.org); AGVGD (available in the public domain at http://agvgd.iarc.fr).

All patients with the p.R45W mutation showed autosomal dominant inheritance. All patients had bilateral involvement, and the age at diagnosis was less than 40 in most of these patients (except a 51-year-old patient). The inheritance pattern of the disease was unknown or sporadic in patients with novel RP1L1 variants, and the ages at diagnosis varied from 47 to 58, which generally were older than those with the p.R45W mutation. 

Among 3 patients with novel variants, one showed unilateral involvement. Patients without RP1L1 mutations showed variable clinical characteristics; patient age ranged from 11 to 71, inheritance was autosomal dominant or sporadic, and 3 of 9 patients had unilateral involvement. 

The Morphology and Distribution of Photoreceptor Disruption and the Association With Genotype

The morphology and distribution of photoreceptor disruption differed according to the RP1L1 genetic mutation present (Fig. 2). Patients with the p.R45W mutation showed almost identical patterns of photoreceptor disruption. In these patients, the disruption of photoreceptors was symmetric across the fovea and between the 2 eyes. The disruption involved the cone outer segment tips (COST) and inner segment-outer segment (IS-OS) junction layers in all patients, and the ELM layer in 5 of 7 (71.4%) patients. The disruption typically was most severe in the fovea and became less severe toward the peripheral retina. Fundus IR findings showed symmetric and round hyporeflectance centered on the fovea in both eyes. Although 3 patients with novel RP1L1 variants had 3 different mutations in the RP1L1 gene, SD-OCT images commonly showed photoreceptor disruption that was symmetric across the fovea, but asymmetric between the 2 eyes. In particular, a 58-year-old female patient with the c.2026A>T genotype showed foveal thinning in OCT images of both eyes, and the photoreceptor disruption and retinal thinning appeared to be more severe in the left eye. The COST and IS-OS junction layers were involved in all patients, whereas the ELM was not involved in any patient. Fundus IR imaging showed asymmetric hyporeflectance between the 2 eyes in all these patients. In patients without RP1L1 mutations, asymmetric photoreceptor disruption between the 2 eyes was noted in 7 of 9 (77.8%) patients, and 3 eyes showed asymmetric patterns of photoreceptor abnormality across the fovea. Three of 9 patients had unilateral involvement. The distribution of photoreceptor abnormality was not centered on the fovea in 3 eyes. The COST layer was involved in all patients and the IS-OS junction layer was involved in 7 of 9 (77.8%) patients, whereas the ELM was not involved in any patient. An asymmetric pattern of photoreceptor abnormality between the 2 eyes was associated with asymmetric visual acuity between the eyes (0.62 ± 0.29 logarithm of the Minimum Angle of Resolution [logMAR] in the more disrupted eye versus 0.17 ± 0.23 logMAR in the less disrupted eye, P = 0.001 by paired t-test). 

Figure 2

 

IR reflectance and SD-OCT images in patients with OMD and their genotypes. In patients with p.R45W mutation (left column), photoreceptor disruption is symmetric across the fovea and between the 2 eyes. The degree of photoreceptor disruption is most severe in the fovea, and IR images show symmetric and round hyporeflectance centered on the fovea in both eyes. In patients with p.Q2311P, p.S676C, and p.D1425H, (middle column) SD-OCT images show symmetric distribution of photoreceptor disruption across the fovea, but photoreceptor abnormalities are not symmetric between the 2 eyes. Some patients without RP1L1 mutation (right column) show asymmetric patterns of photoreceptor disruption across the fovea.

Visual/Electrophysiological Function and Its Relevance With Genotype

Figure 3 demonstrates the comparison of BCVA (logMAR scale), mean threshold value at the central 4 points of Humphrey perimetry, and foveal P1 amplitude on mfERG among the genotypes of the RP1L1. Median BCVAs of the eyes in patients with the p.R45W mutation, novel variants, and no mutation were 20/100 (range, 20/200–20/40), 20/60 (range, 20/200–20/30), and 20/60 (range, 20/200–20/25), respectively. The median of the mean threshold value at the central 4 points of Humphrey perimetry was −5.88 (range, −9.25 to −3.75), −4.25 (range, −8–0), and −7.25 (range, −14.75–0) in patients with p.R45W, novel variants, and no mutation, respectively. The median foveal P1 amplitude on mfERG was 96.5 (range, 77.4–131.6), 82.1 (range, 75.7–99.7), and 85.4 (range, 74.4–148.8) nV/deg ² in patients with p.R45W, novel variants, and no mutation, respectively. The differences in visual and electrophysiologic function were not statistically significant among the 3 groups (P = 0.24, 0.35, and 0.67 for BCVA, perimetric result, and foveal P1 amplitude, respectively, using the Kruskal-Wallis test). 

Figure 3

 

Box plots indicating visual and electrophysiologic function among genotypes of the RP1L1. Visual function, BCVA in logMAR scale (left column), and mean threshold value at central 4 points of Humphrey perimetry (middle column) are not significantly different among the genotypes of the RP1L1 (P = 0.24 and 0.35 for BCVA and perimetric results, respectively, Kruskal-Wallis test). Foveal P1 amplitude on multifocal ERG, which represents electrophysiologic function, is not significantly different among the genotypes (P = 0.67).

Progression of OMD and Its Relevance With Genotype

Figure 4 shows the progression of photoreceptor disruption within 2 years in subgroups divided based on RP1L1 mutations. Patients with RP1L1 mutations showed morphologic progression of photoreceptor disruption within 2 years, as indicated by the area between the 2 dotted lines in Figure 4. In contrast, patients without RP1L1 mutations showed no apparent progression. Median photoreceptor thickness decreased from 58 to 54 μm, 59 to 54 μm, and 61 to 60 μm in patients with the p.R45W mutation, those with novel RP1L1 variants, and those without RP1L1 mutations, respectively. The observed decrease in photoreceptor thickness might be within the variability of manual measurement of photoreceptor thickness. Furthermore, the differences in photoreceptor thickness change among the 3 groups was not statistically significant (P = 0.156 by the Kruskal-Wallis test). 

Figure 4

 

SD-OCT images demonstrating progression of photoreceptor disruption within 2 years in patient groups separated on the basis of genotype. Areas between the 2 dotted lines indicate the progression of photoreceptor disruption in the patients. Patients with the RP1L1 mutation showed prominent morphologic progression of photoreceptor disruption within 2 years, compared to those without the mutation.

Table 3 shows a summary of clinical characteristics and pathologic features in the 3 subgroups divided based on the RP1L1 genotype. 

Table 3. 

 

Summary of Clinical Characteristics in Subtypes of OMD Classified by Genotypes.

	RP1L1 Mutation
	c.133C>T, p.R45W, n = 7	Novel Variants,* n = 3	No Mutation, n = 9
Age of onset, median (range)	34 (8–51)	53 (47–58)	33 (11–71)
Inheritance	AD	Sporadic in 33.3%, unknown in 66.7%	AD in 11.1%, sporadic in 88.9%
Laterality	Bilateral	Bilateral in 66.7%	Bilateral in 66.7%
Visual function
Visual acuity, median (range)	20/100 Snellen equivalent (20/200–20/40)	20/60 Snellen equivalent (20/200–20/30)	20/60 Snellen equivalent (20/200–20/25)
Perimetry results, median (range)†	−5.88 (−9.25 to −3.75) dB	−4.25 (−8–0) dB	−7.25 (−14.75–0) dB
Photoreceptor disruption
Involved layers Distribution Symmetricity	COST (100%), IS-OS junction (100%), ELM (71.4%) More severe toward the fovea Symmetric across the fovea, symmetric between 2 eyes	COST (100%), IS-OS junction (100%), no ELM More severe toward the fovea Symmetric across the fovea, asymmetric between 2 eyes	COST (100%), IS-OS junction (77.8%), no ELM More severe toward the fovea in 12 (80%) eyes Symmetric across the fovea in 12 (80%) eyes asymmetric between 2 eyes in 7 patients (77.8%)
IR finding	Symmetric and round hyporeflectance centered on the fovea	Asymmetric hyporeflectance	Symmetric and round hyporeflectance centered on the fovea in 6 (40%) eyes asymmetric hyporeflectance in 4 (26.7%) eyes no hyporeflectance in 5 (33.3%) eyes
Progression within 2 years	Prominent	Prominent	Less prominent

*  c.6932A>C (p.Q2311P), c.2026A>T (p.S676C), c.4273G>C (p.D1425H). All novel variants are missense mutations in exon 4.

†  Mean threshold at central 4 points in Humphrey perimetry.

Discussion

Our study showed genotype-phenotype correlation of OMD from clinicogenetic analyses of the RP1L1 gene using multimodal imaging modalities. The morphology and distribution of photoreceptor disruption were associated with the RP1L1 genotype, and visual function varied according to the genotypes. In our patients, those with different genotypes showed different rates of progression of photoreceptor disruption on quasi-histologic sectional images of the retina obtained by SD-OCT. 

Our study revealed that the previously identified mutation in the RP1L1 gene from Japanese families, R45W ³ , also was found in Korean OMD patients. The mutation was the most common mutation of the RP1L1 gene among our patients and it involved 36.8% of our Korean OMD patients. However, another mutation in RP1L1 identified from a sporadic case from Japanese OMD patient, p.S1199C ⁵ , was not identified among our cases. In addition to these missense mutations of the RP1L1 gene, we identified 3 novel RP1L1 variants, p.Gln2311Pro, p.Asp1425His, and p.Ser676Cys, among our patients. However, p.Ser676Cys does not appear to be pathogenic, as 5 computational tools all evaluated this mutation as not pathogenic. 

The RP1L1 gene encodes a large, highly polymorphic, retina-specific protein, RP1L1. ²¹ Immunohistochemistry of RP1L1 in the retina section of the cynomolgus monkey showed expression in rod and cone photoreceptors. ³ This suggests a pathogenic role for defective RP1L1 in photoreceptor disruption of OMD. Our study supports the role of RP1L1 in OMD, because more than half the patients (10 of 19) had mutations in the RP1L1 gene, and these patients showed structural disruption confined to the photoreceptor layers. However, the other 9 patients had no RP1L1 mutation, indicating that OMD also may be caused by other genes or nongenetic causes. From our results, we believe that OMD may be a genetically heterogeneous macular disease. 

We showed that multimodal imaging modalities that enable quasi-histologic imaging of the retina are useful to investigate genotype-phenotype correlations in OMD patients. In particular, patients with the p.R45W mutation showed homogeneous and unique morphologic characteristics in SD-OCT and IR imaging. SD-OCT images showed photoreceptor disruption to be most severe in the fovea, and symmetric across the fovea and between the 2 eyes. Our OCT findings in patients with the p.R45W mutation are compatible with those by Tsunoda et al. from a large family with the mutation. ¹⁰ Additionally, IR images demonstrated round hyporeflectance around the fovea in both eyes. Furthermore, the genotype was associated with increased severity of photoreceptor disruption, as represented by more layers showing disrupted photoreceptors and functionally worse visual acuity. When comparing photoreceptor disruption in previously reported cases ¹⁰ and ours with the p.R45W mutation to that in other RP1L1 mutations ⁵ or no RP1L1 mutation, ¹⁰ the p.R45W mutation showed more disrupted photoreceptors, indicating that the genotype results in an OMD phenotype of greater severity. From the genotype-phenotype correlation of OMD, which shows that the clinical heterogeneity of OMD originates from genetic heterogeneity in the RP1L1 mutation, we suggested that OMD can be divided into subtypes on the basis of genotypes, which also has clinical importance (Table 3). 

The genotype of RP1L1 also may be important for the prediction of disease progression, as patients with and without RP1L1 mutations showed differences in disease progression in our study. Specifically, patients with RP1L1 mutations showed morphologic progression and more disrupted photoreceptors within the 2-year follow-up period, whereas those without the mutation showed no apparent progression. Faster disease progression may be associated with earlier disease onset and more disrupted photoreceptors, as seen in patients with the p.R45W mutation. Our data on disease progression may be useful for genetic counseling and support the necessity of genetic testing for the RP1L1 mutation in patients with OMD. 

However, our data were insufficient to draw a definite conclusion regarding the effect of the mutation on disease progression due to the small number of patients and short follow-up periods. Furthermore, our data were obtained from only one ethnic group (Korean), and may not be applicable to other ethnic groups. To generalize our results, data on genetic characteristics from a larger number of patients from other ethnic groups should be obtained and analyzed with clinical characteristics. Furthermore, the etiology of OMD cases without RP1L1 mutations was not investigated in our study. OMD cases without RP1L1 mutations comprised approximately half of the total cases, suggesting that other causative genes may exist. Further genetic analyses on other candidate genes from a larger number of cases are necessary. Additionally, functional implications of the novel RP1L1 variants should be investigated further, as in silico prediction tools are inadequate to confirm or refute pathogenicity without careful validation through functional studies. 

In conclusion, OMD is a clinically and genetically heterogeneous disease. We demonstrated the genotype-phenotype correlation of OMD based on RP1L1 sequence data and in vivo quasi-histologic images obtained by SD-OCT and IR imaging. The genetic data on the RP1L1 mutation may provide useful information on the anatomic features of photoreceptor disruption, visual function, and disease progression in OMD. 

Acknowledgments

Supported by a grant from the Korea Health technology R&D Project, Ministry of Health & Welfare, Republic of Korea (A111161) and National Research Foundation of Korea Grant 2012R1A2A2A02012821 (SJW). The authors alone are responsible for the content and writing of the paper. 

Disclosure: S.J. Ahn, None; S.I. Cho, None; J. Ahn, None; S.S. Park, None; K.H. Park, None; S.J. Woo, None