Genotypic Profile and Clinical Characteristics of CRX-Associated Retinopathy in Koreans

Abstract

This study aimed to investigate the clinical characteristics of Korean patients with retinal dystrophy associated with pathogenic variants of cone rod homeobox-containing gene (CRX). We retrospectively enrolled Korean patients with CRX-associated retinal dystrophy (CRX-RD) who visited two tertiary referral hospitals. Pathogenic variants were identified using targeted panel sequencing or whole-exome sequencing. We analyzed clinical features and phenotypic spectra according to genotype. Eleven patients with CRX-RD were included in this study. Six patients with cone-rod dystrophy (CORD), two with macular dystrophy (MD), two with Leber congenital amaurosis (LCA), and one with retinitis pigmentosa (RP) were included. One patient (9.1%) had autosomal recessive inheritance, and the other ten patients (90.9%) had autosomal dominant inheritance. Six patients (54.5%) were male, and the mean age of symptom onset was 27.0 ± 17.9 years. At the first presentation, the mean age was 39.4 ± 20.6 years, and best-corrected visual acuity (BCVA) (logMAR) was 0.76 ± 0.90 in the better eye. Negative electroretinography (ERG) was observed in seven (63.6%) patients. Nine pathogenic variants were identified, including two novel variants, c.101-1G>A and c.898T>C:p.(*300Glnext*118). Taken together with the variants reported in prior studies, all variants within the homeodomain are missense variants, whereas most variants downstream of the homeodomain are truncating variants (88%). The clinical features of pathogenic variants within the homeodomain are either CORD or MD with bull's eye maculopathy, whereas variants downstream of the homeodomain cause more diverse phenotypes, with CORD and MD in 36%, LCA in 40%, and RP in 24%. This is the first case series in Korea to investigate the CRX-RD genotype-phenotype correlation. Pathogenic variants downstream of the homeodomain of the CRX gene are present as RP, LCA, and CORD, whereas pathogenic variants within the homeodomain are mainly present as CORD or MD with bull's eye maculopathy. This trend was similar to previous genotype-phenotype analyses of CRX-RD. Further molecular biologic research on this correlation is required.

Keywords: CRX; Korean population; Leber congenital amaurosis; cone-rod dystrophy; macular dystrophy; retinitis pigmentosa.

Figure 1

Color fundus photography, optical coherence tomography (OCT), fundus autofluorescence (FAF) and electroretinogram (ERG) of representative cases by each phenotype. In patient 7 (A–D, c.193G>C, cone-rod dystrophy (CORD)), round macular atrophy and peripheral granularity were observed in fundus photography and FAF (A,B); subfoveal atrophy and parafoveal blurring of photoreceptor layers were observed in OCT (B); and cone-dominant reduction in response observed in ERG (D). In patient 5 (E–H, c.128G>A, macular dystrophy (MD)), Perifoveal ring-shaped retinal pigment epithelium atrophy was observed in fundus photography (E) and OCT (G). Outside of the atrophic lesions, ring-shaped hyperautofluorecence was observed in FAF (F). Photoreceptor response in ERG was normal (H). In patient 8 (I–L, c.443del, Leber congenital amaurosis (LCA)) and FAF were not performed, and only color fundus photographs with good quality were attached. Blond fundus and white punctate lesions in peripheral retina (I,J) and diffuse blurring of photoreceptor layers (K) were observed. In patient 9 (M–P, c.684G>C, retinitis pigmentosa (RP)), peripheral retinal degeneration with bone spicule (M,N) and parafoveal sparing of retinal pigment epithelium atrophy (O) were observed. ERG was extinguished in patients 8 and 9 (L,P) and electronegative ERG was observed in patients 7 and 5 (D,H).

Figure 2

Distribution and correlation analysis between age, best-corrected visual acuity, and refractive error of patients. The six plots on the lower left are the distribution of each patient (Blue triangle is cone-rod dystrophy, light blue circle is macular dystrophy, red inverted triangle is rod-cone dystrophy, and purple square is Leber congenital amaurosis). The six plots on the upper right showed the degree of correlation between each variable (The closer to red, the stronger the positive correlation; the closer to purple, the stronger the negative correlation; and bold letters are the correlation coefficients). Age at symptom onset and BCVA (logMAR) showed a negative correlation (p < 0.01). Age at symptom onset and age at first examination showed a positive correlation (p = 0.02). In addition, refraction error and BCVA (logMAR) showed a positive correlation (p = 0.01), but this result includes the extreme values of two LCA patients with hyperopia, so there is a limitation in interpreting it as a linear correlation. ^† Please note that decimal BCVA was used for distribution plots for the convenience of the reader while logMAR visual acuity was used for correlation analysis. BCVA, best-corrected visual acuity; CORD, cone-rod dystrophy; LCA, Leber congenital amaurosis; MD, macular dystrophy; RP, retinitis pigmentosa.

Figure 3

Schematic diagram of CRX structure and phenotype distribution of reported variants from this paper and four previous studies on CRX-associated retinopathy. A total of 35 variants from 83 affected patients were included (cone-rod dystrophy (CORD) or macular dystrophy (MD) in blue letters, retinitis pigmentosa (RP) or rod-cone dystrophy (RCD) in red letters, Leber congenital amaurosis (LCA) in purple letters, and conflicting phenotype in black letters). Hot spots commonly reported in two or more studies are indicated in bold. All mutations within the homeodomain were missense mutations, and most showed phenotypes of CORD or MD. Notably, few patients with other phenotypes were either compound heterozygosity (c.101-1G>A/c.122G>A(p.R41Q), LCA) or homozygosity (p.D65H, RP), unlike all the other patients were simple heterozygosity. On the other hand, the majority of the mutations after the homeodomain were truncating mutations (88%), and various phenotypes appeared. (CORD and MD in 36%, LCA in 40%, and RP in 24%). † This study used a different phenotypic classification system. We performed phenotypic classification based on the test results provided.