Demographics

A total of 41 subjects with two alleles of USH2A variants were enrolled in our study; 19 (46%) were male, and 22 (54%) were female. The mean age at the first visit was 47.66 (SD ± 13.54) years. The mean age at symptom onset, which was referred specially as the age that the patients reported nyctalopia and/or visual field constriction, was 31.21 (SD ± 15.24) years. Hearing impairment was reported in 15 (37%) patients. The mean follow-up period, defined by the length of time between the first and last visits, was 46.89 (SD ± 28.31) months. Detailed clinical presentations and genotype results are presented in Table 1.

Table 1 Clinical presentations and genotypes of the patients with USH2A-related retinal dystrophiesGenetic spectrum of USH2A-related retinal dystrophies in Taiwanese

Capture-based NGS was performed for all recruited patients to identify 46 variants distributed throughout the entire USH2A gene sequence (Fig. 1A); four subjects had structural variants (Table 1). USH2A variants were confirmed according to American College of Medical Genetics and Genomics guidelines. Regarding the types of the variants, 23 (50%) were missense, 10 (22%) were splicing variants, 8 (17%) were nonsense, and 5 (11%) were frameshift mutations (Fig. 1B). The “splicing variant” was defined in our cohort as the variant located in the intronic regions which resulted in RNA splicing defect and was considered as a truncated mutation. The two most predominant variants in this study were c.2802T>G (p.Cys934Trp) and c.4576G>A (p.Gly1526Arg) (Table 2), with relative allele frequencies of 15.85% and 7.32%, respectively. The most common position of the variants was exon 13, with a frequency of 21% (Fig. 1C). The most common variant domain documented in this study was the fibronectin type-III domain, with 63% (n = 29). Seven variants were in the laminin G domain, four were in the laminin-type epidermal growth factor-like domain, and the remaining were in the laminin G-like jelly roll fold, laminin N-terminal, and transmembrane domains.

Fig. 1

Detailed genetic information in our cohort. A Schematic of the USH2A gene structure with the locations of the variants showed in our cohort. The two most predominant variants in our cohort were highlighted in red. B Types of mutations in our cohort. C Percentage of subjects with exon 13 (EX.13 +) and others (EX.13-). D Percentage of subjects with/without the existence of truncated alleles. Group T-: without truncated alleles; Group T + : with at least one truncated allele

Table 2 Results of the genetic analysis of the USH2A variants in our cohortClinical presentations of USH2A-related retinal dystrophies based on the existence of truncated alleles

To investigate the correlation between the existence of truncated alleles and clinical presentation, such as age at symptom onset, age at the first visit, visual acuity at the first and recent visits, EZ line progression, and fundus autofluorescence (FAF) lesion progression, a detailed comparison was performed among groups T- (those with no truncated variants) and T+ (those with at least one truncated variant), which comprised 17 (41%) and 24 (59%) patients, respectively (Fig. 1D).

The ages at symptom onset and first visit were much lower in group T+ (mean = 26.55 and 43.67 years, respectively) than in group T- (mean = 37.24 and 53.29 years, respectively). The differences between the age at symptom onset and at the first visit in these two groups were statistically significant (p = 0.02 and 0.03, respectively).

Visual acuity (average of both eyes, calculated as logarithm of the minimum angle of resolution, logMAR) at the first visit was 0.77 in group T+; whereas, that in group T- was 0.56. At a recent visit, the visual acuities (average of both eyes) were 0.83 for group T+ and 0.73 for group T-. Figure 2A showed the baseline average visual acuities of both eyes. The visual acuity progression seemed to be faster in group T+ than in group T- along with increasing age.

Fig. 2

Comparison of the clinical progression of patients with/without truncated alleles in our cohort. Every point represents an individual patient. The blue and orange lines are the linear regression lines for group T- (without truncated alleles) and T + (with at least one truncated allele), respectively. OD, right eye; OS, left eye. A LogMAR of the average of both eyes at the first visit, presented as a function of age. B The average intact area of macula of both eyes on fundus autofluorescence (FAF, %) at the first visit, presented as a function of age. C EZ line length of the average of both eyes at the first visit, presented as a function of age. D EZ line loss progression rates (mm/year) of the average of both eyes, presented as a function of age

Figure 2B showed the average preserved areas of the macula at the first visit of both eyes. The mean baseline intact areas of the macula were 42.45% in group T+ and 34.17% in group T-. The decrease in the intact macular area was faster in group T+ with increasing age. As for the hypofluorescent area progression rate at the macula, the average rates were 3.88 %/year in group T+ and 3.76 %/year in group T-, respectively.

Figure 2C showed the patients’ average EZ line lengths of both eyes plotted at the first visit. The mean baseline EZ line length was 1.218 mm in group T+, while in group T-, it measured 1.267 mm. The decrease in the EZ line length was faster in group T+ with increasing age. As for the analysis of the EZ line loss progression rate at the fovea, the average rates were 0.088 mm/year in group T+ and 0.071 mm/year in group T-, respectively. In the linear model showing the progression rate of EZ line loss as a function of age (Fig. 2D), group T- demonstrated faster progression rate along with increasing age; whereas, group T+ showed a decreasing progression rate along with age.

More patients in group T+ reported hearing impairment. Twelve (50%) in group T+ suffered from hearing loss, and twelve (50%) did not; whereas, three (18%) patients in group T- had hearing loss, and fourteen (82%) patients did not. Differences were analyzed using Pearson's chi-square test and showed statistical significance (p = 0.034).

Clinical presentations of USH2A-related retinal dystrophies based on the existence of C.2802T>G Variant

c.2802T>G (p.Cys934Trp), being the most predominant variant in our study, accounted for 15.85% (13/82); i.e., 32% (n = 13) of subjects had c.2802T>G variant, and 68% (n = 28) did not. The ages at symptom onset and the first visit were higher in patients with the c.2802T>G variant (mean = 46.93 and 50.77 years, respectively) than in patients without (mean = 29.54 and 34.54 years, respectively). However, the differences between the age at symptom onset and the age at first visit in these two groups were not statistically significant.

Visual acuity (average of both eyes, logMAR) at the first visit was 0.56 in patients with the c.2802T>G variant; whereas, that in those without the c.2802T>G variant was 0.72. Visual acuities (average of both eyes) at a recent visit were 0.72 for patients with the c.2802T>G variant and 0.94 for those without the c.2802T>G variant. Figures 3A showed the average baseline visual acuities of the right and left eyes and both groups showed similar trend.

Fig. 3

Comparison of the clinical progression of patients with/without c.2802 T > G in our cohort. Every point represents an individual patient. The blue and orange lines are the linear regression lines for those with and without c.2802 T > G, respectively. A, B LogMAR of the average of both eyes at the first visit, presented as a function of age. B The average intact area of macula of both eyes on fundus autofluorescence (FAF, %) at the first visit, presented as a function of age. C EZ line length of the average of both eyes at the first visit, presented as a function of age. D EZ line loss progression rates (mm/year) of the average of both eyes, presented as a function of age

Figure 3B showed the average preserved areas of the macula at the first visit of both eyes. The average intact areas of the macula at the first visit for those with c.2802T>G variant were 33.24%; while for those without the c.2802T>G variant, the average intact areas were 42.84%. Regarding the progression rate of hypofluorescent areas at the macula, the average rate was 3.95%/year in those with the c.2802T>G variant and 3.69 %/year in those without the c.2802T>G variant.

Figure 3C showed the patients’ average EZ line lengths of both eyes plotted at the first visit. The mean baseline EZ line length was 1.447 mm in patients with the c.2802T>G variant, whereas it measured 1.182 mm in those without the variant. In the analysis of the EZ line loss progression rate at the fovea, the average rates were 0.081 mm/year in patients with the c.2802T>G variant and 0.079 mm/year in those without the c.2802T>G variant. In the linear model showing the progression rate of EZ line loss as a function of age (Fig. 3D), patients with the c.2802T>G variant showed a faster EZ line progression rate along with increasing age.

Because c.2802T>G is a missense mutation, we further analyzed 13 subjects with the c.2802T>G variant to determine the effect of the other allele. Only two (15.4%) had the other allele with a truncated variant, and eleven (84.6%) did not. However, the two subgroups had no significant differences in visual acuity, EZ line length, and remaining FAF areas.

Clinical presentations of USH2A-related retinal dystrophies based on the rate of EZ line progression

Among the 41 patients in our cohort, 19 underwent regular follow-ups with high-quality imaging more than four times. The follow-up period was from 20.43 to 110.10 months, with a mean of 48.34 months. Thirteen of them lost the follow-up of optical coherence tomography (OCT)images, and nine initially showed a destroyed EZ line.

In our cohort, the overall average rate of EZ line progression was 0.082 mm/year in both eyes. To assess the difference based on the rate of EZ line progression, clinical presentations and genotype were compared among the fast progression group (rate > 0.1 mm/year), moderate progression group (rate between 0.05–0.1 mm/year), and slow progression group (rate < 0.05 mm/year), which comprised 6 (32%), 7 (36%), and 6 (32%) patients, respectively.

The age of symptom onset showed the mean value of 32.17, 31.71, and 38.40 years for the fast, moderate, and slow progression subgroups, respectively. Although the age of onset was lower in the fast and moderate progression subgroups than in the slow progression subgroup, the analysis of variance did not show statistical significance.

To further distinguish the difference, we focused on the fast and slow subgroups to analyze the visual acuity and FAF hypofluorescent area progression at the macula. Visual acuities (average of both eyes, logMAR) at the first visit were 0.32 in the slow progression subgroup and 0.34 in the fast progression subgroup, respectively. Visual acuity (average of both eyes) at a recent visit was 0.34 in the slow progression subgroup; whereas, it measured 0.36 in the fast progression subgroup. The average baseline intact areas of the macula of both eyes were 56.94% in the slow progression subgroup and 44.67% in the fast progression subgroup, respectively. However, the difference was not statistically significant.

Case presentation

Two patients were selected from groups T- and T+ for a more detailed comparison.

Case 1

The patient (proband 17; 30 years old, with a follow-up period of 37.2 months), who had one truncated variant from group T+, had variants c.5836C>T and c.10859T>C. At the first visit, the patient had an early onset age of 21 years and a visual acuity of 0.22 logMAR for both eyes. Color fundus and FAF images of the right eye at first and recent visits were showed from Fig. 4A-D. FAF images (Fig. 4C and D) showed typical-type RP and a macular hypofluorescent lesion progression rate of 0.813 %/year. OCT imaging (Fig. 4E and F) showed that the EZ line loss rate in the fovea was faster at 0.210 mm/year in the right eye.

Fig. 4

Case presentation from a patient of group T + . Right eye sequential images of color fundus, FAF and optical coherence tomography (OCT) of a 30-year-old woman with one truncated variant and a follow-up period of 37.2 months. A, B Color fundus at the first and latest visit. C, D FAF showed typical-type RP and a macula hypofluorescent lesion progression rate of 0.813%/year. E, F In OCT imaging, the EZ line was intact between the two solid lines at the first visit, whereas the two dotted lines revealed the boundary of the EZ line at a recent visit. The EZ line loss rate at the fovea performed faster, with 0.210 mm/year

Case 2

The patient (proband 20; 50 years of age) with no truncated variants from group T- had variants c.2802T>G and c.13007G>A. The patient had a relatively late onset age of 46 years. The mean follow-up period was 37.77 months. Visual acuity at the first visit showed mild impairment with 0.05 logMAR in the right eye and 0.15 logMAR in the left eye. Color fundus and FAF images of the right eye at first and recent visits showed from Fig. 5A-D. FAF images (Fig. 5C and D) showed pericentral-type RP and a macular hypofluorescent lesion progression rate of 2.832 %/year in the right eye. OCT imaging (Fig. 5E and F) showed that the EZ line loss rate at the fovea was slower at 0.020 (mm/year) in the right eye.

Fig. 5

Case presentation from a patient of group T-. Right eye sequential images of color fundus, FAF and optical coherence tomography (OCT) of a 50-year-old woman without truncated variants and with a follow-up period of 37.77 months. A, B Color fundus at the first and latest visit. C, D FAF showed pericentral-type RP and a macula hypofluorescent lesion progression rate of 2.832%/year. E, F In OCT imaging, the EZ line was intact between the two solid lines at the first visit, whereas the two dotted lines revealed the boundary of the EZ line at a recent visit. EZ line loss rate at the fovea performed slower, with 0.020 mm/year