Detailed Comparison Between Two Main Phenotypes of CRB1-Related Retinal Dystrophy, Pan-retinopathy and Maculopathy

Abstract

Purpose

To conduct a comprehensive analysis of the distinct manifestations of two primary phenotypes of Crumbs cell polarity complex component 1 (CRB1)-related retinal dystrophy, pan-retinopathy, and maculopathy.

Methods

A cohort of 75 patients with biallelic pathogenic variants in the CRB1 gene was recruited. Clinical evaluations including visual acuity, refractive errors, fundus autofluorescence imaging, optical coherence tomography, visual field testing, and full-field ERG.

Results

Genetic analysis identified 89 disease-causing CRB1 variants. Patients with biallelic loss-of-function variants were significantly more prevalent in the pan-retinopathy group (40.3%) than in the maculopathy group (7.1%) (P = 0.03). The proportion of patients harboring biallelic variants expressing wild-type CRB1-B was significantly higher in the maculopathy group (30.8%) than in the pan-retinopathy group (8.1%). In the pan-retinopathy group, visual function was significantly worse (P < 0.01). The pan-retinopathy group also showed marked hyperopia, shorter axial lengths, severe visual field impairment, and severely attenuated ERG responses (all P < 0.05). Initial visual acuity was better in the maculopathy group, but a critical decline occurred after 18.94 years. Optical coherence tomography and fundus autofluorescence imaging revealed distinct patterns: pan-retinopathy predominantly showed outer retinal atrophy (98.2%), parafoveal thickening (58.9%–60.7%), and diffuse hypofluorescence (87.5%), whereas maculopathy was characterized by macular edema/schisis (66.7%) and bilateral localized macular hypofluorescence (45.5%).

Conclusions

This study establishes CRB1-associated pan-retinopathy and maculopathy as clinically and genetically distinct entities. The maculopathy group initially has relatively stable vision but shows significant deterioration after adulthood.

Crumbs cell polarity complex component 1 (CRB1; OMIM: #604210) is a key constituent of the CRB complex, a critical apical structure involved in maintaining cellular junctions.¹ Within the retina, CRB1 is integral to regulating processes such as cellular adhesion, establishment of polarity, and intercellular signaling, and it is widely recognized as essential for proper retinal development.^2,3 Although multiple CRB1 isoforms have been identified, only three exhibit substantial expression levels in human retinal tissue: CRB1-A [MT470365], CRB1-B [MT470366], and CRB1-C[MT470367].⁴

Pathogenic variants in CRB1 gene have been linked to a spectrum of inherited retinal dystrophies, encompassing conditions such as Leber congenital amaurosis (LCA), early-onset severe retinal dystrophy (EOSRD), autosomal recessive RP, foveal schisis (FRS), and macular dystrophy (MD).^5–8 LCA and EOSRD, along with RP, are classified as pan-retinopathies, characterized by widespread dysfunction across the entire retina, with CRB1 being a frequently implicated gene in these disorders.^9,10 In contrast, FRS and MD represent more recently identified phenotypes associated with CRB1 variants, primarily affecting the macular region. Notably, MD is regarded as a progressive or advanced manifestation of FRS.^5,11

Previous studies on CRB1-related retinopathy highlight significant differences between the two primary phenotypes in terms of retinal morphology, visual function, and overall prognosis.^1,5,7 For instance, CRB1-associated pan-retinopathy is characterized by distinct features such as hypermetropia owing to short axial length, abnormal retinal lamination, thickened parafoveal retina, preserved para-arteriolar RPE (PPRPE), cystoid macular edema, and Coat's-like vasculopathy.⁸ In contrast, CRB1-related maculopathy presents differently, with patients often exhibiting myopia and a well-organized retinal structure, without the aforementioned features.^7,11 Although genotype–phenotype correlations remain incompletely understood, emerging evidence suggests variability in retinal phenotype expression, with potential contributions from genetic modifiers or environmental factors influencing disease manifestation.^1,8,12

In this retrospective observational study, we conducted a comprehensive analysis of a cohort comprising 75 patients with biallelic pathogenic variants in CRB1 gene, aiming to elucidate the distinct manifestations of the two primary phenotypes. Our investigation focused on critical clinical parameters, including refractive error, visual function progression, retinal morphology, and other phenotype-specific features. Notably, this study represents the largest and most direct cohort comparison to date between subgroups of CRB1-related retinopathy, offering novel insights into the diverse underlying pathophysiological mechanisms.

Methods

Recruitment of Subjects

All patients were recruited from Peking Union Medical College Hospital, Beijing, China. The study was approved by the Institutional Review Board of Peking Union Medical College Hospital and adhered to the tenets of the Declaration of Helsinki, as well as the Guidance on Sample Collection for Human Genetic Diseases issued by the Ministry of Public Health of China. The diagnostic criteria used in our study are as follows: LCA is defined as severe congenital blindness with extinguished or markedly diminished full-field ERG responses before the first 6 months of life, as well as the absence of systemic abnormalities.¹³ EOSRD is defined as being milder than LCA and present after the first 6 months of life but before 10 years.¹⁴ RP is defined as sequential degeneration of rods and RPE followed by cones, which presents in adulthood.¹⁴ The FRS features only a subtly altered foveal reflex, which subsequently progresses to a bull's eye pattern of outer retinal atrophy (MD). Optical coherence tomography (OCT) reveals distinct structural changes: FRS is characterized by intraretinal cysts with preserved outer retinal architecture, and MD exhibits outer retinal atrophy, including loss of the ellipsoid zone and RPE thinning in the macular region.⁵ Fundus autofluorescence (FAF) imaging reveals no or only mild changes in the macula in FRS patients, whereas in MD patients, the macula exhibits relatively well-demarcated areas of hypofluorescence. Notably, in the most advanced cases of MD, the retinopathy occasionally extends beyond the macula, affecting retinal areas superior, inferior, and nasal to the optic nerve.⁵ LCA/EOSRD and RP are categorized as CRB1-related pan-retinopathy, whereas FRS and MD are classified as CRB1-related maculopathy.

Clinical Evaluation

The clinical evaluations included measurements of tests of the best-corrected decimal visual acuity (BCVA) converted to the median logMAR BCVA, refractive errors, FAF imaging (Heidelberg Engineering, Heidelberg, Germany, or Daytona, Optos, Dunfermline, UK), OCT (Topcon, Tokyo, Japan, or Heidelberg Engineering), visual field (VF) testing (Octopus 900 Haag-Streit, Bern, Switzerland), and full-field ERG according to the standards of the International Society for Clinical Electrophysiology of Vision¹⁵ (RETI-port ERG system; Roland Consult, Wiesbaden, Germany). Amplitudes of b-waves in scotopic 3.0 ERG and photopic 3.0 ERG responses are selected for statistical analysis.

Genetic Analysis

Whole exome sequencing was used to identify pathogenic variants associated with CRB1. The target capture was performed using the Nadprep Hybrid Capture Reagents (Nanodigmbio, Nanjing, China) followed by sequencing on the Illumina Novaseq platform. The raw data were analyzed using the Burrow–Wheeler Aligner (BWA) v0.7.12 software, and the reads were compared with the reference sequence of hg19 from the UCSC Genome Browser. A comparison was conducted using the 1000 Genomes Project database, the dbSNP database, and the Exome Aggregation Consortium database to exclude nonpathogenic polymorphisms.

The allele frequency was determined using the gnomAD browser (https://gnomad.broadinstitute.org/). The pathogenicity of the Missense variants and the nucleotide conservation were accessed through VarCards.¹⁶ In silico algorithms, including Mutation Taster,¹⁷ CADD,¹⁸ DANN,¹⁹ and FATHMM_MKL,²⁰ were used to predict the pathogenicity of the Missense variants. Nucleotide conservation was analyzed using GERP.²¹ Sanger sequencing validation and segregation analysis were used to identify putative pathogenic CRB1 variants.

Statistic Analysis

Statistical analysis was performed using R software version 4.4.3. P values of 0.05 or less were considered statistically significant. Because the clinical data from the left and right eyes are not independent, only one eye per patient was randomly selected for statistical analysis. In cross-sectional data analysis, when patients have multiple follow-ups, the data observed for the first time are used for statistical analysis. For longitudinal analysis, a linear mixed model was used to evaluate the progression rate of BCVA. Furthermore, patients should meet the criteria of a minimum of two visits during the follow-up period, and a minimum duration between the initial and final visit of 1 year. Patients’ abilities to count fingers, detect hand movements, and detect light perception were assigned logMAR values of 2.6, 2.7, and 2.8, respectively.²²

Results

Either homozygous or compound heterozygous variants were detected in CRB1 gene in 75 patients from 71 families, including 58 patients with LCA/EOSRD, 4 patients with RP, 8 patients with FRS, and 5 patients with MD. There were 36 males and 39 females in our cohort. Among the patients, 13 patients with LCA/EOSRD had been previously reported by our team.⁹

Genetic Analysis

Overall, 54 patients carried compound heterozygous CRB1 variants, and 21 carried homozygous variants, one patient of which carried two homozygous variants (from consanguineous marriage). Eighty-nine disease-causing variants were identified, among which 33 alleles are novel. More than one-half of the variants (56/89 [62.9%]) were located at exons 6, 7, and 9, which encoded the three laminin G–like domains of the CRB1 protein. The results of the variant analyses are summarized in Table 1 and Figure 1.

Table 1.

CRB1 Variants Analysis

No.	Diagnosis	Zygosity	Variant 1	Protein 1	Impact on CRB1-B	Impact on CRB1-C	Variant 2	Protein 2	Impact on CRB1-B	Impact on CRB1-C
1	EOSRD	Heterozygous	c.2363T>C	p.L788P	M	WT	c.2413delA	p.I805Lfs*10	M	WT
2	LCA	Heterozygous	c.70+5G>A		WT	M	c.2512A>T	p.K838*	M	WT
3-1	EOSRD	Homozygous	c.238C>T†	p.Q80*	WT	M
3-2	EOSRD	Homozygous	c.238C>T†	p.Q80*	WT	M
4	RP	Heterozygous	c.2590_2591dup†	p.F865Sfs*18	M	WT	c.4207G>C	p.E1403Q	WT	WT
5	RP	Homozygous	c.3862G>A	p.G1288S	M	WT
6	EOSRD	Heterozygous	c.1478delT	p.F493Sfs*9	M	M	c.3658_3659delAG	p.H1221Pfs*10	M	WT
7	LCA	Heterozygous	c.2187delC†	p.D730Mfs*24	M	WT	c.3632G>A†	p.C1211Y	M	WT
8	EOSRD	Heterozygous	c.294T>A†	p.C98*	WT	M	c.3164T>A	p.V1055E	M	WT
9	LCA	Heterozygous	c.1571T>C	p.L524P	M	M	c.3442T>C	p.C1148R	M	WT
10	LCA	Heterozygous	c.1576C>T	p.R526*	M	M	Exon 1 del		WT	M
11	LCA	Homozygous	c.4089_4096dup	p.S1366Lfs*34	WT	WT
12	LCA	Two homozygous variants	c.3054_3055insGC†	p.Y1019Afs*4	M	WT	c.3057_3058insTGACA†	p.M1020*fs*1	M	WT
13	LCA	Heterozygous	c.1369C>T†	p.R457*	M	M	c.1965T>A†	p.Y655*	M	M
14	EOSRD	Heterozygous	c.2512A>T	p.K838*	M	WT	c.2416G>C	p.E806Q	M	WT
15	LCA	Homozygous	c.1369C>T†	p.R457*	M	M
16	LCA	Homozygous	c.2866G>T	p.G956*	M	WT
17	LCA	Heterozygous	c.1369C>T†	p.R457*	M	M	c.3741dupA†	p.F1248Ifs*15	M	WT
18	EOSRD	Heterozygous	c.2234C>T	p.T745M	M	WT	c.3695A>G	p.H1232R	M	WT
19	EOSRD	Homozygous	c.1831T>C	p.S611P	M	M
20	EOSRD	Heterozygous	c.70+2T>A		WT	M	c.3355delA†	p.K1119Nfs*22	M	WT
21	EOSRD	Heterozygous	c.2371G>C	p.G791R	M	WT	c.2681_2682insC	p.C896Lfs*13	M	WT
22	EOSRD	Heterozygous	c.1576C>T	p.R526*	M	M	c.2714G>A	p.R905Q	M	WT
23	EOSRD	Homozygous	c.3878G>A	p.W1293*	M	WT
24	LCA	Heterozygous	c.70+2T>A		WT	M	c.1576C>T	p.R526*	M	M
25	EOSRD	Heterozygous	c.1147T>C	p.C383R	WT	M	c.1571T>C	p.L524P	M	M
26	LCA	Heterozygous	c.456T>A†	p.C152*	M	WT	c.2396G>A†	p.C799Y	M	WT
27-1	EOSRD	Homozygous	c.1505G>T†	p.G502V	M	M
27-2	EOSRD	Homozygous	c.1505G>T†	p.G502V	M	M
28	LCA	Homozygous	c.1624T>C†	p.S542P	M	M
29	LCA	Homozygous	c.1369C>T†	p.R457*	M	M
30	EOSRD	Heterozygous	c.1399G>A†	p.E467K	M	M	c.3341T>A†	p.L1114*	M	WT
31-1	LCA	Heterozygous	c.2512A>T	p.K838*	M	WT	c.3676G>T	p.G1226*	M	WT
31-2	LCA	Heterozygous	c.2512A>T	p.K838*	M	WT	c.3676G>T	p.G1226*	M	WT
32-1	EOSRD	Heterozygous	197325773_197326261dup†		WT	M	c.1633delinsAA	p.S545Nfs*6	M	M
32-2	LCA	Heterozygous	197325773_197326261dup†		WT	M	c.1633delinsAA	p.S545Nfs*6	M	M
33	EOSRD	Heterozygous	c.1201T>A	p.C401S	M	M	c.1576C>T	p.R526*	M	M
34	EOSRD	Heterozygous	c.2056C>T	p.R686C	M	M	c.3676G>T	p.G1226*	M	WT
35	LCA	Heterozygous	c.523_532dup†	p.Y178Cfs*15	WT	M	c.2128+2T>G		M	M
36	EOSRD	Homozygous	c.1841G>T	p.G614V	M	M
37	LCA	Homozygous	c.3676G>T	p.G1226*	M	WT
38	LCA	Heterozygous	c.1831T>C	p.S611P	M	M	c.2540_2541delTC	p.F848Qfs*60	M	WT
39	LCA	Heterozygous	c.2290C>T	p.R764C	M	WT	c.3676G>T	p.G1226*	M	WT
40	LCA	Heterozygous	c.1576C>T	p.R526*	M	M	c.3023T>C	p.L1008S	M	WT
41	LCA	Homozygous	c.1576C>T	p.R526*	M	M
42	LCA	Heterozygous	c.1202G>A	p.C401Y	M	M	c.2462C>G	p.T821R	M	WT
43	LCA	Homozygous	c.1576C>T	p.R526*	M	M
44	LCA	Heterozygous	c.652+2G>T†		WT	M	c.1831T>C	p.S611P	M	M
45	EOSRD	Heterozygous	c.2815T>G	p.C939G	M	WT	c.3676G>T	p.G1226*	M	WT
46	EOSRD	Heterozygous	c.2842+5G>A		M	WT	c.3631T>G†	p.C1211G	M	WT
47	EOSRD	Homozygous	c.3442T>C	p.C1148R	M	WT
48	LCA	Homozygous	c.1831T>C	p.S611P	M	M
49	EOSRD	Heterozygous	c.4005+2T>G		M	WT	Exon 1 Del		WT	M
50	RP	Heterozygous	c.3373T>C†	p.F1125L	M	WT	c.2981_2982delAA	p.K994Rfs*3	M	WT
51	EOSRD	Heterozygous	c.2172T>A	p.Y724*	M	WT	c.2234C>T	p.T745M	M	WT
52	RP	Heterozygous	c.1913C>T	p.S638L	M	M	c.2729C>G†	p.S910C	M	WT
53	LCA	Heterozygous	c.3307G>A	p.G1103R	M	WT	c.1543C>T	p.Q515*	M	M
54	EOSRD	Heterozygous	c.1576 C>T	p.R526*	M	M	c.2021A>T†	p.D674V	M	M
55	EOSRD	Heterozygous	Exon 1 Del		WT	M	c.1A>G†		WT	M
56	EOSRD	Heterozygous	c.4207G>C	p.E1403Q	WT	WT	c.601T>C	p.C201R	WT	M
57	EOSRD	Heterozygous	c.2234C>T	p.T745M	M	WT	Exon9-Exon 12 del†		M	WT
58	EOSRD	Homozygous	c.2498G>A	p.G833D	M	WT
59	FRS	Heterozygous	c.1568T>A†	p.L523Q	M	M	c.3991C>T	p.R1331C	M	WT
60	FRS	Heterozygous	c.2512A>T	p.K838*	M	WT	c.3554C>T†	p.P1185L	M	WT
61	MD	Heterozygous	c.1366T>C	p.C456R	M	M	c.1381C>T†	p.Q461*	M	M
62	MD	Homozygous	c.4006-10A>G		WT	WT
63	FRS	Heterozygous	c.1405T>G	p.C469G	M	M	c.2942G>T†	p.R981I	M	WT
64	FRS	Heterozygous	c.2172T>A	p.Y724*	M	WT	c.3862G>A	p.G1288S	M	WT
65	FRS	Heterozygous	c.2230C>T	p.R744*	M	WT	c.4207G>C	p.E1403Q	WT	WT
66	FRS	Heterozygous	c.3307G>A	p.G1103R	M	WT	c.1928T>C†	p.L643P	M	M
67	MD	Heterozygous	c.1055_1063delGGGAATGTG	p.G352_C354delGEC	WT	M	c.4207G>C	p.E1403Q	WT	WT
68	FRS	Heterozygous	c.664G>A	p.E222K	WT	M	c.2963_2971delTAATATTGCinsAA†	p.I988Kfs*32	M	WT
69	MD	Heterozygous	c.80G>T	p.C27F	WT	M	c.419A>G†	p.Y140C	WT	M
70	FRS	Heterozygous	c.455G>A	p.C152Y	WT	M	c.683G>A	p.C228Y	WT	M
71	MD	Heterozygous	c.1543C>T	p.Q515*	M	M	c.3862G>A	p.G1288S	M	WT

FRS, foveal retinoschisis; M mutant; WT, wild type.

^†Novel variant.

Figure 1.

The 12 exons of the canonical transcript encoding isoform A and the equivalent exons comprising the smaller transcript encoding isoform B and isoform C (A). Schematic representation of the CRB1 gene (B) and protein (C), highlighting functional domains. Missense variations and inframe deletions are indicated below the gene structure, loss of function variants are noted above. In green: specific variants responsible for EOSRD/LCA/RP. In orange: specific variants responsible for FRS/MD. In black: specific variants responsible for both EOSRD/LCA/RP and FRS/MD.

Loss-of-function (LoF) variants are genetic variants that disrupt gene function, typically leading to reduced or absent protein activity. These include nonsense mutations (premature stop codons), frameshift indels (disrupting the reading frame), splice-site mutations (affecting canonical GT–AG sites), large deletions/copy number variations (removing exons), and start–loss mutations (disrupting translation initiation). The percentage of patients with two LoF variants in the pan-retinopathy and maculopathy groups was 40.3% and 7.1%, respectively. The ratio difference was statistically significant by Fisher's exact test (P = 0.03).

All identified variants affect the CRB1-A isoform. The schematic representation of CRB1 isoforms in retina is shown in Figure 1, and predicated impact of the variants on the isoforms CRB1-B and CRB1-C is described in Table 1. The proportion of patients harboring biallelic wild-type CRB1-B variants was significantly greater in the maculopathy group (30.8%) than in the pan-retinopathy group (8.1%) (P = 0.04, Fisher's exact test). When considering at least one wild-type CRB1-B variant, the difference between groups (27.4% in maculopathy vs. 46.1% in pan-retinopathy) was not statistically significant (P = 0.20). For the CRB1-C isoform, no significant differences were observed between the pan-retinopathy and maculopathy groups in patients with biallelic wild-type variants (37.1% vs. 30.8%; P = 0.76) or at least one wild-type variant (62.9% vs. 76.9%; P = 0.52).

Clinical Manifestation Comparison Between the Pan-retinopathy and Maculopathy Groups

In the pan-retinopathy group (n = 62), nystagmus was observed in 32 patients (51.6%), and 19 patients (30.7%) reported nyctalopia. Additionally, 7 patients (11.3%) exhibited a shallow anterior chamber, 13 patients (21.0%) presented with cataracts, 6 patients (9.7%) demonstrated the oculodigital sign, and 9 patients (14.5%) had strabismus. In contrast, in the maculopathy group (n = 13), only one patient exhibited cataracts, and no other symptoms or physical signs were observed (Table 2).

Table 2.

Comparison of Ocular Findings Between Pan-retinopathy and Maculopathy Associated With CRB1 Variants

Characteristics	Pan-Retinopathy	Maculopathy
No. of cases	62 (31 males, 31 females)	13 (5 males, 8 females)
Nystagmus present	32 (51.6)	None
Nyctolopia	19 (30.7)	None
Shallow anterior chamber	7 (11.3)	None
Cataract	13 (21.0)	1 (7.7)
Oculo-digital sign	6 (9.7)	None
Strabismus	9 (14.5)	None

Values are number (%).

LogMAR BCVA

The mean logMAR BCVA was significantly worse in the pan-retinopathy group (n = 53) compared with the maculopathy group (n = 13). In the pan-retinopathy group, the mean logMAR BCVA was 1.64 ± 0.90 in the right eye and 1.67 ± 0.88 in the left eye. In contrast, the maculopathy group exhibited better visual acuity, with mean logMAR BCVA values of 0.72 ± 0.75 in the right eye and 0.84 ± 0.82 in the left eye. The Kruskal–Wallis rank-sum test confirmed that the logMAR BCVA was statistically significantly worse in the pan-retinopathy group (P < 0.01) (Fig. 3A).

Figure 3.

Boxplots comparing ocular parameters between pan-retinopathy and maculopathy groups in CRB1-related retinal dystrophy. (A) LogMAR BCVA showed worse visual acuity in pan-retinopathy. (B) SE indicated higher hyperopia in pan-retinopathy. (C) Axial length was shorter in pan-retinopathy. (D) MD in VF was greater in pan-retinopathy. (E) Scotopic 3.0 and (F) photopic 3.0 b-wave amplitude were reduced in pan-retinopathy. All differences were statistically significant (P < 0.05).

Progression Rates of LogMAR BCVA From Longitudinal Analysis

Linear mixed models fit by restricted maximum likelihood were used to determine the overall progression rate for logMAR BCVA. The mean ± SD annual progression of logMAR BCVA were 0.03 ± 0.01 logMAR units (P < 0.01) in the pan-retinopathy group (follow-up visits = 62), and 0.05 ± 0.01 logMAR units (P < 0.01) in maculopathy group (follow-up visits = 16), respectively. There was no statistical difference between the progression rates of the two phenotype groups by t test (P = 0.24).

For patients with pan-retinopathy, segmented regression analysis revealed a critical age threshold at 5.52 years. Before this breakpoint, age was negatively correlated with logMAR BCVA (P = 0.02), suggesting early visual improvement. However, beyond 5.52 years, age exhibited a significant positive association with logMAR BCVA (P < 0.01) (Fig. 2A). In contrast, patients with maculopathy displayed a later breakpoint at 18.94 years. Before this age, no significant relationship was observed between age and logMAR BCVA (P = 0.44). However, after 18.94 years, age demonstrated a strong positive correlation with worsening logMAR BCVA (P < 0.01) (Fig. 2B).

Figure 2.

Patients with pan-retinopathy (A) exhibited a breakpoint at 5.52 years. Before this age, logMAR BCVA improved with age (coefficient = −0.072; P = 0.02), whereas afterward it progressively worsened (slope = 0.040; P < 0.01). Patients with maculopathy (B) showed a later breakpoint at 18.94 years. The initial negative trend (coefficient = −0.022; P = 0.44) was nonsignificant, but after 18.94 years, logMAR BCVA worsened significantly (slope = 0.063; P < 0.01).

Spherical Equivalent (SE)

The mean SE was significantly higher in the pan-retinopathy group (n = 43) compared with the maculopathy group (n = 12). In the pan-retinopathy group, the mean SE values were 6.13 ± 3.82 diopters (D) in the right eye and 5.89 ± 4.26 D in the left eye. In contrast, the maculopathy group exhibited much lower mean SE values of 0.80 ± 1.76 D in the right eye and 0.86 ± 1.74 D in the left eye.

Additionally, the proportion of eyes with an SE of +6.00 D or greater was significantly higher in the pan-retinopathy group: 57.5% (23/40) in the right eye and 47.5% (19/40) in the left eye. In the maculopathy group, no eyes (0/12) met this criterion in either eye. The Kruskal–Wallis rank-sum test confirmed that the SE was statistically significantly higher (indicating hyperopia) in the pan-retinopathy group, with a significant difference between the two phenotypic groups (P < 0.01) (Fig. 3B).

Axial Length

The mean axial length was significantly shorter in the pan-retinopathy group compared with the maculopathy group. In the pan-retinopathy group (n = 17), the mean axial lengths were 20.07 ± 1.21 mm in the right eye and 20.05 ± 1.31 mm in the left eye. In the maculopathy group (n = 4), the mean axial lengths were notably longer, with values of 21.70 ± 1.15 mm in the right eye and 21.59 ± 1.13 mm in the left eye. The Kruskal–Wallis rank sum test demonstrated that the axial length was statistically significantly shorter in the pan-retinopathy group (P = 0.04) (Fig. 3C).

Mean Deviation (MD) in VF

The MD in the VF was significantly higher in the pan-retinopathy group (n = 21) compared with the maculopathy group (n = 6). Specifically, the MD values were 27.93 ± 5.84 in the right eye and 27.84 ± 6.41 in the left eye for the pan-retinopathy group, whereas the maculopathy group exhibited MD values of 10.45 ± 3.87 in the right eye and 9.47 ± 4.20 in the left eye. The Kruskal–Wallis rank-sum test confirmed that the MD in the pan-retinopathy group was statistically significantly greater (P < 0.01) (Fig. 3D).

ERGs

The scotopic 3.0 b-wave amplitude was significantly lower in the pan-retinopathy group compared with the maculopathy group. In the pan-retinopathy group (n = 32 for the right eye; n = 34 for the left eye), the mean amplitudes were 12.25 ± 29.37 µV in the right eye and 12.21 ± 31.02 µV in the left eye. In contrast, the maculopathy group (n = 11) demonstrated significantly higher amplitudes, with mean values of 246.35 ± 159.40 µV in the right eye and 251.55 ± 159.40 µV in the left eye. The Kruskal–Wallis rank-sum test indicated a statistically significant difference between the two groups (P < 0.01) (Fig. 3E).

Similarly, the photopic 3.0 b-wave amplitude was significantly reduced in the pan-retinopathy group compared with the maculopathy group. The pan-retinopathy group (n = 32 for the right eye; n = 34 for the left eye) exhibited mean amplitudes of 5.31 ± 11.7 µV in the right eye and 5.67 ± 13.45 µV in the left eye. In the maculopathy group (n = 11), the mean amplitudes were notably higher, with values of 82.03 ± 50.04 µV in the right eye and 89.36 ± 51.54 µV in the left eye. The Kruskal–Wallis rank-sum test confirmed a statistically significant difference between the groups (P < 0.01) (Fig. 3F). The visual function results of patients with pan-retinopathy and maculopathy are listed in Table 3.

Table 3.

Comparison of Measured Parameters Between Pan-retinopathy and Maculopathy Associated With CRB1 Variants

	Diagnosis
	Pan-Retinopathy		Maculopathy
Parameters	OD	OS	OD	OS	P Value
LogMAR BCVA (available, n)	53	53	13	13	<0.01*
Mean ± SD	1.64 ± 0.90	1.67 ± 0.88	0.72 ± 0.75	0.84 ± 0.82
Range	0.3 to 2.8	0.2 to 2.8	0 to 2.6	0.1 to 2.6
SE (available, n)	40	40	12	12	<0.01*
Mean ± SD, D	6.13 ± 3.82	5.89 ± 4.26	0.80 ± 1.76	0.86 ± 1.74
Range, D	−2.88 to +13.88	−2.00 to +14.75	−2.62 to +3.50	−1.50 to +4.38
+2 D ≤ SE < +6 D (n, %)	8, 20.0%	10, 25.0%	3, 25.0%	3, 25.0%	NA
SE ≥ +6 D (n, %)	23, 57.5%	19, 47.5%	0, 0.0%	0, 0.0%
Axial length (available, n)	16	17	4	4	0.04†
Mean ± SD, mm	20.07 ± 1.21	20.05 ± 1.31	21.70 ± 1.15	21.59 ± 1.13
Range, mm	18.00 to 22.12	17.77 to 22.48	20.22 to 22.99	20.25 to 22.96
MD in VF (available, n)	21	21	6	6	<0.01*
Mean ± SD	27.93 ± 5.84	27.84 ± 6.41	10.45 ± 3.87	9.47 ± 4.20
Range	14.0 to 34.0	11.0 to 34.0	4.5 to 16.0	3.1 to 16.0
Scotopic 3.0 ERGs (available, n)	32	34	11	11	<0.01*
Mean ± SD, µV	27.93 ± 5.84	27.84 ± 6.41	246.35 ± 159.40	251.55 ± 159.40
Range, µV	0.0 to 115.2	0.0 to 120.8	122.0 to 626.0	116.0 to 649.0
Photopic 3.0 ERGs (available, n)	32	34	11	11	<0.01*
Mean ± SD, µV	5.31 ± 11.7	5.67 ± 13.45	82.03 ± 50.04	89.36 ± 51.54
Range, µV	0.0 to 37.0	0.0 to 48.0	20.0 to 197.0	30.0 to 209.0

^* P ≤ 0.01.

^† P ≤ 0.05.

OCT and FAF

In pan-retinopathy, OCT was available for 56 patients. Outer retinal atrophy was observed in 55 patients (98.2%) in both eyes. Thinning of the fovea was noted in 23 patients (41.1%) bilaterally. Parafoveal thickening was present in 33 eyes (58.9%) in the right eye and 34 eyes (60.7%) in the left eye. A scalloped macular appearance was seen in 5 patients (8.9%) bilaterally, and a coarse retinal structure was observed in 10 eyes (17.9%) in the right eye and 9 eyes (16.1%) in the left eye. Macular edema or schisis was identified in six patients (10.7%) bilaterally.

In maculopathy, OCT was available for 12 patients. Outer retinal atrophy was noted in 5 patients (41.7%) bilaterally. Thinning of the fovea was observed in one patient (8.3%) bilaterally. However, macular edema or schisis was present in eight patients (66.7%) bilaterally. Parafoveal thickening, scalloped macular appearance, and coarse retinal structure were not reported in any cases. The typical OCT features are illustrated in Figure 4.

Figure 4.

(Top) OCT findings in CRB1-related pan-retinopathy (A-C). (Bottom) OCT features of CRB1-related maculopathy (D, E). (A) Loss of ellipsoid zone with intraretinal cysts. (B) Parafoveal retinal thickening and outer retinal atrophy. (C) Scalloped macular appearance with a coarse retinal structure (loss of laminar structure). (D) Macular ellipsoid zone disruption. (E) Prominent macular edema or schisis.

In pan-retinopathy, FAF was available for 40 patients. Diffuse hypofluorescence was observed in 35 patients (87.5%) bilaterally, while PPRPE was present in 16 patients (40.0%) bilaterally. Peripapillary sparing was noted in 7 patients (17.5%) bilaterally, and peripheral retina sparing was observed in 18 patients (45.0%) bilaterally. In maculopathy, FAF was available for 11 patients. In contrast, macular hypofluorescence was observed in five patients (45.5%) bilaterally, and macular hyperfluorescence was noted in two patients (18.2%) bilaterally. Sporadic hypofluorescence was present in three patients (27.3%) bilaterally, and normal FAF was observed in one patient (9.1%) bilaterally. No instances of diffuse hypofluorescence, PPRPE, peripapillary sparing, or peripheral retina sparing were observed in maculopathy. The typical fundus manifestations of the two phenotypes are showed in Figure 5 and Table 4.

Figure 5.

Representative fundus manifestations of CRB1-related retinopathies: LCA showed generalized retinal discoloration with pigmentary changes and preserved perivascular retinal coloration (A1), exhibiting diffuse hypofluorescence with PPRPE, peripapillary sparing, and peripheral sparing on autofluorescence (B1), along with outer retinal atrophy, foveal thinning, and parafoveal thickening on OCT (C1). EOSRD demonstrated posterior pole pigmentary changes with preserved peripheral retina (A2), presenting posterior diffuse hypofluorescence with PPRPE (B2) and outer retinal atrophy with foveal thinning (C2). RP features generalized posterior pole pigmentary changes, most prominent in vascular arcades and macula (A3), showing macular and perivascular hypofluorescence with PPRPE (B3) and outer retinal atrophy with foveal thinning (C3). MFS displayed macular center discoloration (A4) with near-normal autofluorescence (B4), but revealed schisis-like changes with ellipsoid zone discontinuity on OCT (C4). MD manifested macular atrophy (A5) with posterior pole hypofluorescence mixed with focal hyperfluorescence (B5) and outer retinal atrophy with foveal thinning (C5).

Table 4.

OCT and FAF Comparison Between Pan-retinopathy and Maculopathy Related to CRB1 Variants

Characteristics	Pan-Retinopathy		Maculopathy
OCT (available, n)	56	56	12	12
Outer retinal atrophy, n (%)	55 (98.2%)*	55 (98.2%)*	5 (41.7%)†	5 (41.7%)†
Thinning of the fovea, n (%)	23 (41.1%)†	23 (41.1%)†	1 (8.3%)	1 (8.3%)
Parafovea thickening, n (%)	33 (58.9%)†	34 (60.7%)†	None	None
Scalloped macular, n (%)	5 (8.9%)	5 (8.9%)	None	None
Coarse retinal structure, n (%)	10 (17.9%)	9 (16.1%)	None	None
Macular edema/schisis, n (%)	6 (10.7%)	6 (10.7%)	8 (66.7%)†	8 (66.7%)†
FAF (available, n)	40	40	11	11
Diffuse hypofluorescence	35 (87.5%)*	35 (87.5%)*	None	None
PPARPE	16 (40.0%)†	16 (40.0%)†	None	None
Peripapillary sparing	7 (17.5%)	7 (17.5%)	None	None
Peripheral retina sparing	18 (45.0%)†	18 (45.0%)†	None	None
Macular hypofluorescence	3 (7.5%)	3 (7.5%)	5 (45.5%)†	5 (45.5%)†
Macular hyperfluorescence	1 (2.5%)	1 (2.5%)	2 (18.2%)	2 (18.2%)
Sporadic hypofluorescence	2 (5.0%)	2 (5.0%)	3 (27.3%)	3 (27.3%)
Normal	None	None	1 (9.1%)	1 (9.1%)

^*Incidence ≥80%.

^†Incidence ≥ 40%.

Discussion

This study represents the first direct comprehensive comparison of CRB1-associated pan-retinopathy and maculopathy, providing novel insights into their distinct clinical and genetic characteristics. Our cohort includes the largest reported series of CRB1-related maculopathy patients (n = 13) to date, enabling robust characterization of this less common phenotype. The findings demonstrate significant differences in genetic architecture, clinical presentation, and structural–functional correlations between these two forms of CRB1-related retinopathy.

Genetic analysis identified a significant number of CRB1 variants, with a high proportion in exons 6 to 9, consistent with previous reports in Chinese population,²³ underscoring the functional importance of these domains in CRB1-related pathogenesis. Although Western cohorts with CRB1-related maculopathy frequently harbor the p.Ile167_Gly169del variant,^1,5,11 this variant was absent in our Chinese patients. Moreover, unlike Western populations, no definitive hotspot variants emerged in our cohort, highlighting differences in genetic backgrounds across ethnic groups. The study underscores the need for population-specific genetic screening approaches and tailored therapeutic development, as these differences may impact the efficacy of mutation-specific treatments like gene therapy or CRISPR-based interventions across diverse ethnic groups.

Genotype–phenotype analysis suggested that biallelic LoF variants may predispose to severe, pan-retinal involvement, likely owing to complete CRB1 dysfunction, which aligns with prior reports.¹ In contrast, maculopathy often associated with missense or in-frame variants, might reflect partial protein function retention, permitting milder, localized retinal damage.⁵ In the retina, Müller cells express exclusively CRB1-A, the canonical transcript, encoding a protein with 19 EGF-like domains and three laminin G domains. Photoreceptors predominantly express the CRB1-B transcript. Transcript A is the most abundant during development and transcript B is predominant in adult retina.⁴ Additionally, a CRB1-C transcript has been identified. This isoform lacks both transmembrane and cytoplasmic domains, and it is predicted to encode a secreted protein.⁴ The findings of this study reveal a significant association between biallelic wild-type CRB1-B variants and maculopathy, contrasting with previous reports that the isoform B has no impact on the retinal phenotype.¹ This finding suggests that retaining two functional copies of CRB1-B may predispose individuals to a more localized retinal degeneration phenotype, with preservation of outer retinal layers at an early age. However, the presence of one mutant allele of CRB1-B may be insufficient to drive a distinct retinopathy pattern. Neither pan-retinopathy nor maculopathy was influenced by isoform C expression. These findings underscore the need for an isoform-independent approach, such as CRISPR-based genome editing, which may offer a viable alternative to gene augmentation for treating CRB1-associated inherited retinal diseases.²⁴ However, the CRB1-B and CRB1-C–specific exonic sequences (i.e., exon 5a and exon 11alt for isoform B, and exon 6alt for isoform C) are notably absent from the capture regions of most clinically used exome sequencing panels. Consequently, potential pathogenic variants within these isoform-specific regions would remain undetected through routine diagnostic screening. This inherent limitation leads to a systematic under-representation of variants that selectively impair CRB1-B or CRB1-C isoforms while sparing the canonical CRB1-A isoform in clinical datasets. Furthermore, the statistical analysis is confounded by structural overlap among isoforms. These inherent correlations in mutation patterns complicate efforts to attribute biological effects specifically to individual isoforms.

Clinically, pan-retinopathy patients showed more severe ocular involvement and worse visual function than maculopathy patients. Previous studies demonstrated slow progressive reduction of visual acuity in patients with CRB1 biallelic variants.^23,25,26 In our cohort, both groups exhibited statistically significant progression in logMAR BCVA over time, with comparable annual rates. However, the age-dependent progression pattern in maculopathy revealed a critical breakpoint at 18.94 years, after which logMAR BCVA worsened significantly with age. This finding suggests that maculopathy may initially remain stable in younger patients but accelerates in adulthood, possibly owing to cumulative macular damage or age-related metabolic changes. In contrast, pan-retinopathy patients exhibited severe vision impairment in early childhood, followed by a modest visual improvement phase (likely reflecting normal ocular development), before entering the progression of visual decline. The profoundly distinct clinical progression patterns of CRB1-associated pan-retinopathy and maculopathy dictate fundamentally different therapeutic strategies regarding intervention timing, whereas pan-retinopathy necessitates immediate intervention during infancy or even presymptomatic stages to preserve retinal architecture, capitalize on developmental plasticity, and circumvent the diminishing returns of late-stage treatment, maculopathy offers a more adaptable therapeutic window owing to its gradual progression and localized nature, permitting either cautious early intervention in childhood to potentially prevent macular schisis/edema (although requiring a thorough risk–benefit analysis in asymptomatic cases) or, alternatively, targeted treatment during the critical young adult period to achieve optimal therapeutic efficacy while maintaining safety.

The pan-retinopathy group exhibited marked hyperopia and shorter axial lengths, suggesting a developmental arrest in global retinal maturation. In contrast, the maculopathy group displayed near-emmetropic refraction and normal axial lengths, indicating a more localized pathology. Furthermore, the pan-retinopathy group demonstrated severe VF impairment, consistent with diffuse retinal dysfunction, whereas the maculopathy group showed relatively preserved peripheral function. Electrophysiological findings further corroborate this distinction. The pan-retinopathy group exhibited severely attenuated or extinguished ERG responses. In contrast, the maculopathy group maintained significantly greater amplitudes. These findings suggest that CRB1 variants leading to pan-retinopathy cause widespread photoreceptor and bipolar cell dysfunction, while maculopathy variants may preferentially disrupt macular retinal circuitry while preserving extramacular function.

Multimodal imaging further highlights the distinct structural differences between CRB1-associated pan-retinopathy and maculopathy. Pan-retinopathy demonstrated near-universal outer retinal atrophy with frequent parafoveal thickening, suggesting widespread degenerative changes. In contrast, maculopathy showed less frequent outer atrophy but a greater prevalence of macular edema/schisis, indicating a more localized pathology. FAF findings reinforced this dichotomy, with pan-retinopathy exhibiting diffuse hypofluorescence and maculopathy displaying focal macular abnormalities. These imaging findings provide objective biomarkers for phenotypic differentiation and may serve as valuable outcome measures in future clinical trials.

In conclusion, this comprehensive comparison establishes CRB1-associated pan-retinopathy and maculopathy as clinically and genetically distinct entities. The preserved peripheral retinal function in maculopathy patients may offer a better prognosis for mobility and orientation, whereas the global dysfunction in pan-retinopathy suggests more severe visual disability. Our analysis of the largest maculopathy cohort to date provides a foundation for phenotype-specific management strategies and highlights the need for tailored therapeutic approaches. However, the rarity of this maculopathy phenotype represents a key limitation of our study. Given the relatively small sample size of the maculopathy subgroup, findings in our study should be interpreted with caution and require further validation in larger, prospective studies.