Screening Mutations of the Monogenic Syndromic High Myopia by Whole Exome Sequencing From MAGIC Project

Abstract

Purpose: This observational study aimed to identify mutations in monogenic syndromic high myopia (msHM) using data from reported samples (n = 9370) of the Myopia Associated Genetics and Intervention Consortium (MAGIC) project.

Methods: The targeted panel containing 298 msHM-related genes was constructed and screening of clinically actionable variants was performed based on whole exome sequencing. Capillary sequencing was used to verify the identified gene mutations in the probands and perform segregation analysis with their relatives.

Results: A total of 381 candidate variants in 84 genes and 85 eye diseases were found to contribute to msHM in 3.6% (335/9370) of patients with HM. Among them, the 22 genes with the most variations accounted for 62.7% of the diagnostic cases. In the genotype-phenotype association analysis, 60% (201/335) of suspected msHM cases were recalled and 25 patients (12.4%) received a definitive genetic diagnosis. Pathogenic variants were distributed in 18 msHM-related diseases, mainly involving retinal dystrophy genes (e.g. TRPM1, CACNA1F, and FZD4), connective tissue disease genes (e.g. FBN1 and COL2A1), corneal or lens development genes (HSF4, GJA8, and MIP), and other genes (TEK). The msHM gene mutation types were allocated to four categories: nonsense mutations (36%), missense mutations (36%), frameshift mutations (20%), and splice site mutations (8%).

Conclusions: This study highlights the importance of thorough molecular subtyping of msHM to provide appropriate genetic counselling and multispecialty care for children and adolescents with HM.

High myopia (HM) is defined as a refractive error of at least −6.00 diopters (D) or an axial length (AL) ≥26 mm and it is found in 0.5 to 5% of the global population and 15 to 20% of the East Asian population.¹ HM is a leading cause of blindness due to related complications, such as cataracts, open-angle glaucoma, retinal detachment, macular hole, and choroidal degeneration.^2,3 Without proper optical correction during infancy and early childhood, HM can lead to behavioral or developmental issues.^4,5 The clinical management of this disease is a complex and often multidisciplinary process. 

Inheritance of myopia may be complex or monogenic. Monogenic high myopia was further subdivided into an isolated and a syndromic form. Over the past 20 years, 20 genes have been found to be associated with monogenic isolated HM, including SCO2, ZNF644, P4HA2, SLC39A5, BSG, CCDC111, GLRA2, CPSF1, TNFRSF21, NDUFAF7, GRM6, LEPREL1, LRPAP1, DZIP1, XYLT1, CTSH, LOXL3, ARR3, OPN1LW, and FKBP5.^6–8 In contrast, monogenic syndromic HM (msHM) is a very heterogeneous group of diseases with additional ocular or systemic involvement, and requires more complex investigations, optical correction, and myopia control treatments. To date, based on the Human Phenotype Ontology (HPO; https://hpo.jax.org/) database, more than 100 genes have been found to be associated with msHM. 

MsHM can be broadly categorized into three groups based on the clinical characteristics of the disease: (1) ametropic retinal dystrophies (defined as category 1); (2) connective tissue disorders (category 2); and (3) other disorders related to corneal, lens, and optic nerve malformations (category 3).^1,9 For instance, Stickler syndrome and Marfan syndrome are two relatively common heritable connective tissue disorders associated with HM that can have a range of severe ocular and systemic manifestations.¹⁰ Correspondingly, HM can be a first feature of an ametropic retinal dystrophies disease, placing challenge for pediatric ophthalmologists who see such diseases.¹¹ Importantly, the characteristics of these diseases indicate that HM involves multiple signaling pathways and biological processes, making it an excellent model for studying its etiology. For example, at least eight genes have been reported in either complete or incomplete congenital stationary night blindness (CSNB), HM is often seen as an early onset symptom and may have similar pathogenic mechanisms of ON-bipolar cell signal transmission defect.^12,13 

Since the launch of precision medicine (PM) and genome medicine (GM), clustering of rare variants based on whole exome sequencing (WES) or targeted panel sequencing has played a crucial role in the diagnosis, prediction, prevention, and treatment of diseases, especially heterogeneous genetic diseases.^14,15 However, the traditional inherited disease diagnostic strategy is based on the “phenotype-first” approach, in which patients with phenotypic disease traits are genotyped to identify gene variants that may be associated with the disease. This strategy is often challenging due to the complexity of the disease phenotypes, especially in pediatric patients and those with incomplete penetrance. Correspondingly, the unbiased “genome-first” approach combined with electronic medical records (EMRs) phenotypic data has the potential to expand our understanding of the relationships between rare variants in specific genes and a range of phenotypes. This approach has greater utility for use in studies of heterogeneous populations or diseases to provide insight into suspected gene ontologies and undiagnosed Mendelian diseases.^16–19 

The present study investigated the utility of WES using an msHM gene panel and screened for related mutations among a population of nearly 10,000 Chinese Han children and adolescents with HM. Additionally, this study verified whether the “genome-first” approach is applicable and reasonable for the accurate diagnosis of these patients. The results of this study provide new ideas for the prevention and control of HM, especially for early-onset, uncorrectable and familial myopia. 

Materials and Methods

Cohort Description and Ethics Statement

For the current study, all Chinese individuals with putative genetic variants of suspected msHM were selected from the MAGIC project, a cohort consists of 9370 individuals with HM (patients from the children and adolescents myopia survey [CAMS] program) who have undergone WES.^8,20,21 The inclusion criteria for this study were as follows: (1) spherical equivalent (SE) ≤–6.00 D in at least one eye with at least 2 measurements under non-cycloplegia; (2) age at study inclusion ≤19 years (Fig. 1A). All participants or their guardians (when participants were <16 years) gave written informed consent for participation. This study was conducted at the Eye Hospital, Wenzhou Medical University (REC reference 2021-015-K-12-01). Clinical evaluations included slit-lamp microscopy, indirect ophthalmoscope examination, uncorrected visual acuity (UCVA), best-corrected visual acuity (BCVA), vision inspection, visual field, intraocular pressure (IOP), fundus photography, and optical coherence tomography (OCT). Additionally, some patients underwent optical coherence tomography angiography (OCTA; Fig. 1C). The corresponding clinical examination data of each proband's parents, family history, and family tree were collected during a clinical recall process. Finally, we established a comprehensive electronic archive with a unique ID for families combined with thier genomic variation and clinical data (Fig. 1D). This study was approved by the Institutional Medical Ethics Committee of the Eye Hospital, Wenzhou Medical University. All study procedures were performed in adherence with the tenets of the Declaration of Helsinki. 

Figure 1.

 

Flowchart of screen mutations of the msHM. (A) Population screening. (B) Sample collection and determination of variants. (C) Panel customization and genotype-phenotype correlation. (D) Clinical intervention/advice.

DNA Extraction and WES

DNA obtained from the oral mucosa of all probands and their parents was extracted using a FlexiGene DNA Kit (Qiagen, Venlo, The Netherlands), according to the manufacturer's protocol. DNA from all probands underwent whole-exome sequencing using the Agilent-V6 targeted capture kit (Agilent) and the Illumina NovaSeq6000 sequencing system (150PE; Berry Genomics Institute, Beijing, China; Fig. 1B). The mean depth and coverage of the target region were approximately 73.0 × and 99.8%, respectively. 

Detection and Analysis of Variants

Sequence reads were aligned to the reference human genome (UCSC hg19) using the Burrows-Wheeler Alignment (BWA) tool²²; the variants, including single nucleotide variants (SNVs) and small insertions or deletions (InDels), were called and annotated with Verita Trekker and Enliven software (Berry Genomics Institute, Beijing, China), respectively. The called variants were then filtered using an in-house standardized pipeline and database. The types of variants were classified as follows: missense, splicing, stop gain, and frameshift. The curation level of the variants was classified as pathogenic (P), likely pathogenic (LP), and variants of uncertain clinical significance (VUS), according to the American Medical Genetics and Genomics guidelines (ACMG) interpretation standards and guidelines²³ (see Fig. 1B). 

Filtering of the Dataset and Verification Analysis

“High myopia” was entered as a keyword query in the human phenotype ontology database (hpo.jax.org/app/browse/term/HP:0011003),²⁴ and 18 categories, 298 genes, and 315 disease subtypes involving msHM were obtained (HM as one of the necessary phenotypes, not the only one). Then, a reference panel was designed, defined as the Target_msHM_298_V1 panel (Table 1, Supplementary Table S1, Fig. 1C). A further 6 gene datasets were selected as control sets: 778 genes from Target_Eye_792_V2 chip, the largest custom-made capture panel of known inherited eye diseases (defined as IEDs-778), by Beijing Genomics Institute (Shenzhen, China)²⁵ (Supplementary Table S2); 278 genes in RetNet, a comprehensive analysis gene dataset of known inherited retinal dystrophies (RN-278; Supplementary Table S3); 77 genes associated with retinitis pigmentosa (RP) from a Chinese population (RPC-77)²⁶; 56 genes associated with RP from a Spanish population (RPS-56)²⁷; and 2 datasets of Chinese patients with early-onset high myopia (eoHM), containing 33 genes and 34 genes (eoHM-33 and eoHM-34, respectively).^28,29 Finally, the RPC-77 and RPS-56 datasets were defined as the RP-genes dataset and the eoHM-33 and eoHM-34 datasets were defined as the eoHM-genes dataset (Supplementary Table S4). 

Table 1.

 

Content of Target_msHM_298_V1 Panel

 

Functional Enrichment Analysis

WebGestalt (http://www.webgestalt.org/)³⁰ and g:Profiler (http://biit.cs.ut.ee/gprofiler/gost)³¹ were used to study the characteristics and specificity of gene functions and signaling pathways. A total of 298 gene symbols in the Target_msHM_298_V1 panel served as the input to generate Gene Ontology (GO) terms (GO_BP, GO_MF, and GO_CC), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and WIKI pathways. R packages (ggplot2, ggpubr) were used to draw images. 

Confirmation of Suspected Variants

Genomic DNA samples were collected from the oral mucosa of the recalled patients and their parents (see Fig. 1B). The candidate variants were reviewed by clinical geneticists and ophthalmologists, and then capillary sequencing was performed. Patients with positive variants in more than one gene were excluded (n = 2). Co-segregation analysis was performed by capillary sequencing when a mutation was detected and members of the proband's family were available. 

Statistical Analysis

The diagnostic yield was calculated based on the number of variants classified as pathogenic, likely pathogenic, or VUS, or based on the number of probands who received a definite clinical diagnosis. Novel variations were those that were not registered in the Human Gene Mutation Database, ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/),³² InterVar (http://wintervar.wglab.org/),³³ or the in-house variant database (including information on 400,000 Chinese whole-exome variants). Statistical analyses were performed using Stata software version 15.1 (StataCorp LP, College Station, TX, USA). 

Results

Characteristics of the Cohort and Gene Panel

The MAGIC cohort consisted of 9370 unrelated Chinese patients with HM (4467 female patients and 4903 male patients; 47.7%/52.3%). Data from this cohort are available in our publicly published dataset for the new pathogenic and susceptible genes of HM.⁸ The mean age of the cohort was 14.8 ± 2.1 years (range = 6–19 years, median = 15 years) and the mean SE was as follows: OS = −6.805 ± 1.641 D, OD = −7.031 ± 1.496 D (range = −1.25 to −20.875 D; Supplementary Table S5). 

We designed a custom gene panel named “Target_msHM_298_V1” which comprised 298 genes. From this, screening of msHM-related variants was performed based on the exome sequencing data. In the GO analysis of the targeted gene panel, the GO_BP terms were mainly involved in visual perception, sensory perception of light stimulus, detection of visible light, and phototransduction (Fig. 2A); the GO_MF terms were mainly involved in the structural constituents of the eye lens and intracellular cyclic guanosine monophosphate (cGMP)-activated cation channel activity (Supplementary Fig. S1A); the GO_CC terms were mainly involved in photoreceptor disc membrane, photoreceptor inner/outer segment, and photoreceptor cell cilium (Supplementary Fig. S1B). The KEGG and WIKI pathways were mainly involved in phototransduction and retinol metabolism (Fig. 2B). Furthermore, to determine the similarities and differences between the inherited retinal dystrophies datasets versus the other inherited eye disease panels, we compared 2 comprehensive datasets: IEDs-778 and RN-278. A total of 177 genes were present in all 3 datasets and 28 genes were unique to our dataset (Fig. 2C). 

Figure 2.

 

Characteristics of gene panel. (A) Gene ontology analysis (GO_BP) of the Target_msHM_298_V1 gene panel. (B) Pathway analysis of the Target_msHM_298_V1 gene panel. (C) Overlap analysis of the reference genes of msHM-298, RN-278, and IEDs-778. (D) Overlap analysis of the reference genes of msHM-84, RPC-77, RPS-56, eoHM-33, and eoHM-34. (E) Overlap analysis of the reference genes of msHM-TOP, RPC-TOP, and eoHM-TOP.

Potential Candidate Patients With msHM and Variants

Based on the Target_msHM_298_V1 panel and rigorous in-house bioinformatics pipeline, 381 variants (324 SNVs and 57 InDels) were identified from 335 patients, rendering an overall yield of 3.6% (335/9370) in the study population. Of all the candidate variants (267 unique), 48 were novel and not present in HGMD, ClinVar, or InterVar (Supplementary Table S6). The mean age of the 335 candidate patients with msHM was 14.5 ± 2.4 years (range = 6–19 years, median = 15 years) and the mean SE was as follows: OS = −6.829 ± 1.683 D, OD = −6.981 ± 1.694 D (range = 0.25 to −19.50 D; Table 2). 

Table 2.

 

Baseline Characteristics of the Candidate Patients With MsHM

 

Collectively, variations spanned 84 genes in our cohort (defined as msHM-84). According to the ACMG/AMP interpretation standards and guidelines, the number of missense, splicing, stop gain, and frameshift variations was 250, 29, 46, and 56, respectively, and the number of pathogenic, likely pathogenic, and VUS variants was 36, 229, and 116, respectively (Figs. 3A, 3B). Of note, 14 genes involved in autosomal recessive inherited eye diseases explained 13.7% (46/335) of the solved cases. Among these variants, 43 were compound heterozygous and 3 were homozygous. Mutations in six genes were detected in at least three patients. The USH2A gene accounted for the greatest number of cases (26.1%, 12/46). Eight genes involved in X-linked inherited eye disease explained 8.4% (28/335) of the solved cases. The CACNA1F gene explained the greatest number of solved cases (42.9%, 12/28). The majority of patients had autosomal dominant inherited eye disease (77.9%, 261/335), and most common mutation occurs in the OPA1 gene (11.9%, 31/261; Fig. 3C). Approximately 62.7% (210/335) of the variants in probands were found in the top 22 genes (OPA1, SNRNP200, BEST1, CACNA1F, USH2A, RHO, LPR5, EYS, COL2A1, MIP, CFI, NR2E3, GUCY2D, MYOC, FBN1, FZD4, CRB1, RP1, CA4, ABCA4, PDE6B, and CRX) in our cohort (defined as msHM-TOP). The top 8 of these genes with no fewer than 10 variants accounted for 35.3% (119/335) of all identified mutations (Fig. 3D). In addition to the top 22 genes, the remaining 62 genes accounted for 37.3% (125/335) of variants; 33 genes had at least 2 variants, and 29 genes with 1 mutation were found in only 1 patient. 

Figure 3.

 

Potential findings of candidate variants and diseases. (A) Type, number, and proportion of variation according to ExonicFunc. (B) Type, number, and proportion of variation according to curation level. (C) Type, number, and proportion of variation according to inheritance mode. (D) The TOP 22 genes and the number of variants. (E) The TOP 20 diseases and the number of patients.

Further, to confirm the similarities and differences in the gene mutation spectrums of different populations and ocular disease types, four potential gene datasets were referenced (RP-genes: RPC-77 and RPS-56; eoHM-genes: eoHM-33 and eoHM-34). Five genes, including RHO, CRX, RPGR, PRPH2, and PRPF8 were present in all five datasets (Fig. 2D). We then focused on the top (with the highest number of mutations) RP-genes (RP-TOP; total: 24 genes) and eoHM-genes (eoHM-TOP; total: 25 genes; Supplementary Table S7) and found that the SNRNP200, USH2A, and RHO genes appeared in three datasets (Fig. 2E). A total of 85 clinical msHM cases were identified, mainly involving tissues of the optic nerve, retina, lens, vitreous, and cornea. The 10 most common diseases accounted for 47.5% (159/335) cases, including optic atrophy 1 (OPA1; OMIM #166500), retinitis pigmentosa 33 (RP33; OMIM #610359), macular dystrophy, vitelliform, 4 (VMD4; OMIM #616151), night blindness, congenital stationary (incomplete), 2A, X-linked (CSNB2A; OMIM #300071), retinitis pigmentosa 39 (RP39; OMIM #613809), congenital stationary, autosomal dominant 1 (CSNBAD1; OMIM #610445), familial exudative vitreoretinopathy 4 (FEVR4; OMIM #601813), retinitis pigmentosa 25 (RP25; OMIM #602772), cataract 15, multiple types (CTRCT15; OMIM #615274), and retinitis pigmentosa 37 (RP37; OMIM #611131; Fig. 3E). 

Ocular or Systemic Features Known to be Associated With Causal Genes

A total of 201 probands (85 female patients and 116 male patientss; 201/335, 60%) among the candidate patients with msHM were recalled for clinical evaluations. The mean age of the patients was 14.7 ± 2.6 years (range = 6–19 years, median = 15 years); 17% were under the age of 12 years. The mean SE was as follows: OS = −6.808 ± 1.641 D and OD = −7.037 ± 1.501 D. The mean AL was as follows: OS = 26.3 ± 1.2 mm and OD = 26.4 ± 1.1 mm. Among them, 60 probands with both parents, 102 probands with only 1 parent (33 cases with a father, and 69 cases with a mother), and 27 cases with other family members were included in the segregation analysis. Among these patients, tessellated fundus was the most common clinical characterization according to fundus photography (78.6%, 158/201; Fig. 4A, Table 3). 

Figure 4.

 

Ocular or systemic features known to be associated with the causal gene. (A) Leopard-shaped fundus. (B) Lens opacity through the slit lamp. (C) Type, number, and ratio of congenital cataract in our study. (D) Anterior segment examination of CDA and granular turbidity. (E) Papilledema. (F) Posterior staphyloma. (G) Abnormal vascular morphology of Marfan syndrome.

Table 3.

 

Clinical Characteristics of Probands and Their Relatives

 

A total of 215 candidate variants spanning 68 genes were identified (185 SNVs and 30 InDels) in 201 candidate patients with msHM. Among the candidate variants (165 unique), there were five novel variations. In combination with the clinical examination data, 25 patients (9 female patients and 16 male patients, 12.4%, 25/201) received determined genotype-phenotype diagnoses, including congenital cataracts (n = 12), stickler syndrome (n = 1), corneal dystrophy (n = 1), congenital stationary night blindness/retinitis pigmentosa (n = 5), cone-rod dystrophy (n = 1), familial exudative vitreoretinopathy (n = 1), Marfan syndrome (n = 2), and primary open-angle glaucoma (n = 2). Among them, 60% (15/25), 28% (7/25), and 12% (3/25) of solved cases belong to category 3, category 1, and category 2, respectively. Nine patients (4.5%, 9/201) received secondary determined genotype-phenotype diagnoses, with some specific phenotypes lacking critical diagnostic evidence, including retinitis pigmentosa (n = 4), congenital cataracts (n = 1), stickler syndrome (n = 1), corneal dystrophy (n = 1), familial exudative vitreoretinopathy (n = 1), and Marfan syndrome (n = 1; Table 4). 

Table 4.

 

Detailed Information of Determined Genotype-Phenotype Diagnosed Patients

 

Congenital cataract³⁴ was the most common type of msHM in our cohort. Twelve probands and 18 family members were identified, and the determined diagnosis rate was 40% (12/30). This included congenital cataract types 1, 3, 5, 6, 9, 15, and 17 (Figs. 4B, 4C). One patient with the variant in FZD4 (c.313A>G, p.Met105Val), which is one of the genes that cause exudative vitreoretinopathy (EVR1; OMIM #133780),³⁵ presented with irregular gray exudative foci. Moreover, one patient with the variant TGFBI (c.371G>A, p.Arg124His) was diagnosed with corneal dystrophy, Avellino type (CDA; OMIM #607541),³⁶ a type of granular corneal dystrophy (GCD; Fig. 4D). One patient had a variant in TEK (c.226C>T, p.Gln76*), which caused the glaucoma 3, primary congenital E (OMIM #617272), and presented with high IOP, visual field defects, and papilledema (Fig. 4E). Two patients were diagnosed with Marfan syndrome; OMIM #154700) involving two FBN1 variants (c.1957_1958del, p.Val653*fs*1; c.679C>T, p.Gln227*), fundus photographs of one patient showed posterior staphyloma and vascular attenuation³⁷ (Figs. 4F, 4G). 

One family (F1) was diagnosed with stickler syndrome, type I (STL1; OMIM #108300).³⁸ This family comprised four generations and six members. STL1 is characterized by early-onset HM and short height (<−1 SD proportionate short stature). One member (II:1) presented with macular cysts and posterior vitreous detachment in addition to the above manifestations (Figs. 5A–D). One proband (F2, II:1) with night blindness, congenital stationary (complete), 1C (CSNB1C; OMIM #613216)³⁹ exhibited HM, night blindness, tessellated fundus, and temporal arc-shaped spots, her parents (F2, I:1 and F2, I:2) were both variant carriers of TRPM1 (c.270del, p.Asp91Ilefs*10; Figs. 5E–G). For 2 cases (patients 15 and 23) with night blindness, congenital stationary (incomplete), 2A (CSNB2A; OMIM #300071). Patient F3, I:2 exhibited macular degeneration exudates, macular schisis, and thinning of the retinal nerve fiber layer (RNFL; Figs. 5H–J). The remaining 166 patients had no determined genotype-phenotype associations, including 102 variants that have been reported as pathogenic or likely pathogenic in the HGMD, ClinVar, InterVar, and in-house variant databases. 

Figure 5.

 

Family pedigrees, Capillary sequencing, and EMR results of patients. (A–D) Stickler syndrome, type I. (E–G) Night blindness, congenital stationary (complete), 1C, and autosomal recessive (CSNB1C). (H–J) Night blindness, congenital stationary (incomplete), 2A, X-linked (CSNB2A). Filled symbols indicate affected individuals, unfilled symbols indicate unaffected individuals, and gray-filled symbols indicate a heterozygous carrier.

Discussion

MsHM is a highly heterogeneous disease both clinically and genetically, and some symptoms are insidious and delayed. Meanwhile, msHM can present with ametropic retinal dystrophies, connective tissue disorders, or other disorders related to corneal or lens malformations, it greatly increases the likelihood of blindness and disability. Therefore, early accurate and cost-effective genetic diagnosis is crucial for later intervention and treatment with the decreasing cost of sequencing and the popularization of clinical applications. The current results revealed a total of 381 msHM-related candidate variants in 335 unique individuals based on a targeted sequencing panel. In the genotype–phenotype correlations, nearly 17% of subjects (25 determined and 9 secondary determined from 201 patients) had clinically actionable variants associated with 1 of 18 msHM. The variants were in genes involved in retinal or corneal dystrophies, connective tissue diseases, glaucoma, and disorders of lens development. This study demonstrates that among patients who do not appear to have a Mendelian inheritance pattern of HM, subclinical systemic symptoms might be overlooked without molecular diagnosis. For example, defects in lens or night blindness, as the majority (68%) of pathogenic genes in determined patients. 

This is the largest Chinese cohort that has been used to evaluate the diagnostic yield of msHM based on WES data. Furthermore, this study verified the availability of the disease-specific gene panel and analyzed the differences relative to known reference panels for eye diseases. For this comparative analysis, the most widely used reference gene panel associated with inherited eye diseases was obtained from the RetNet database.^28,29,40 A total of 185 genes were present in this dataset and our own dataset, with 113 genes unique to our dataset. Moreover, 62.7% of variants were found in the top 22 genes, among which, the number of variations was greater than or equal to 5. This suggests that screening these top 22 genes as the first step would provide a more rapid and cost-effective means of offering genetic testing for patients with msHM. 

Our results showed some differences in the gene spectrum compared to other disease types and populations. For example, the genes with highest numbers of variants associated with eoHM in the Chinese population and RP in the Spanish population were COL2A1 and USH2A, respectively, compared to OPA1 and SNRNP200 in our cohort. Interestingly, compared to a recent study of HM, Haarman et al. evaluated a panel containing 500 eye disease genes based on 113 enrolled patients with HM and found that 11 patients displayed additional ocular or systemic involvement (11/113, 9.7%), with a lower genetic diagnostic rate in the case of the same categories (ametropic retinal dystrophies genes, connective tissue disorders genes, other disorders related to corneal, lens, and optic nerve malformations genes) as in our study.⁹ These results highlight the importance of obtaining a reference gene panel for specific populations and heterogeneous diseases as a cost-effective diagnostic tool. However, the role of the original WES data in expanding the mutation spectrum and discovering more new genes for msHM cannot be ignored. 

The current findings are more reliable and have clinical reference value due to the large number of patients analyzed and the fact that we assessed the most common disease types prone to genetic variants in the Chinese population. However, some diagnoses in our study were made more complicated than expected. For example, one patient with the pathogenic variant FZD4 (c.313A>G, p.Met105Val) showed no obvious abnormalities, indicating heterogeneity of the gene or mutation site.⁴¹ The parents of the patient with CSNB1C were carriers of the TRPM1 variant; both of them exhibited HM before the age of 18 years, suggesting that carriers of a pathogenic mutation of this gene may have abnormal phenotypes. 

HM, especially in children, is a rare condition with a different pattern of etiology to that seen in older children. The clinical management of such children, in terms of investigation, optical correction, and use of myopia control treatments, is a complex and often multidisciplinary process. There are currently no clinical guidelines for msHM. Therefore, we established a comprehensive clinical management protocol and applied it to the patients. First, patients with family histories accounted for 56% (14/25) of the yield of determined genotype-phenotype diagnosis patients, and the number of patients verified by additional family members was higher than that from probands alone. Therefore, an extensive medical history should be taken from each patient that presents with HM. Second, we established an electronic archive for the 335 patients with msHM to monitor clinical manifestations and provide more accurate genotypic-phenotypic associations. Finally, we offered genetic counseling and clinical intervention advice to 10 families, including 1 patient diagnosed with CDA who was advised against undergoing LASIK surgery so as to avoid postoperative protein deposition and corneal opacity.⁴² 

There are some limitations of this study that should be acknowledged. First, oral mucosa cells as a DNA source for WES may potentially introduce a higher mutation rate than most human cells due to their high replication rate. Second, a total of 167 patients with EMRs (83.1%, 167/201) did not receive determined or secondary determined genotype-phenotype association results, possibly due to genomic polygenicity, phenotypic heterogeneity, limited panel or exome sequencing performance, and lack of relevant clinical symptoms.⁴³ Importantly, diagnosis of inherited retinal dystrophies can be improved by increasing the use of electroretinogram (ERG), visual evoked potential (VEP), and ultra-widefield fundus photography examinations to identify structural abnormalities or diminished function of the peripheral retina. Additionally, whole genome sequencing (WGS) has the potential to unravel some of the yet unsolved cases, especially those that involve intergenic regions of DNA.⁴⁴ Our results suggest that although the “genome-first” approach has certain advantages in heterogeneous populations and for complex phenotypic diseases, unbiased clinical symptom examinations, family member information, and long-term follow-up remain essential. Meanwhile, the number of genes, cost-performance, and clinical utility of the panel need to be weighed against WES and WGS. 

Conclusions

This study demonstrated the utility of the “genome-first” approach and WES with a monogenic syndromic HM disorder gene panel. This approach was used to identify patients with msHM from an HM cohort derived from the MAGIC project and CAMS program. These results not only help to distinguish subtypes, extend the phenotype, and deepen our understanding of these rare diseases, but also serve as an efficient reference for the design of panel-based genetic diagnostic testing, genetic counseling, and future therapy for patients with msHM and their family members. Taken together, these findings contribute to the broader implementation of genomic medicine for other ocular diseases or nonselective patients with specific abnormal phenotypes. 

Acknowledgments

The authors thank all the members of our laboratory. Additionally, we appreciate the support from the School of Ophthalmology & Optometry, Eye Hospital, and School of Biomedical Engineering at Wenzhou Medical University (Wenzhou 325027, P. R. China). 

Supported by the Key Research and Development Program of Zhejiang Province (grant numbers 2023C03031, 2021C03102, and 2020C03036), the National Natural Science Foundation of China (grant numbers 32370679 and 81830027), the National Key Research and Development Program for Active Health and Aging Response (grant numbers 2023YFC3604000 and 2020YFC2008200), the Medicine and Health of Science and Technology Program of Zhejiang (grant number 2023KY151), the Major Scientific and Technological Innovation Project of WenZhou (grant number ZY2020013), the Science and Technology Project of Wenzhou (grant numbers Y20220774 and Y20220154) and the Internal Fund Project of Eye Hospital of Wenzhou Medical University (grant number KYQD20210702). 

Author Contributions: H.W., L.D.X., J.Q., and J.Z.S. contributed to the study design. C.C., G.A., X.G.Y., P.L., and G.S.Z. contributed to sample and clinical data collection. J.Y., S.Y.W., Y.B., Y.Y.Z., and X.Y.L. contributed to data interpretation. C.C., L.D.X., J.Q., H.W., and J.Z.S. wrote the main manuscript text. All authors reviewed the manuscript. The authors read and approved the final manuscript. 

Web Resources: This article includes these web resources: wANNOVAR: https://wannovar.wglab.org/, ClinVar database: https://www.ncbi.nlm.nih.gov/clinvar/, Genome aggregation database (gnomAD): https://gnomad.broadinstitute.org/, Combined Annotation Dependent Depletion (CADD score): https://add.gs.washington.edu/, The Human Phenotype Ontology (HPO): https://hpo.jax.org/app/. 

Data Availability Statements: The data that support the findings of this study are available from the MAGIC and CAMS program, but restrictions apply to the availability of these data, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the MAGIC and CAMS institutional review board. 

Disclosure: C. Chen, None; G. An, None; X. Yu, None; S. Wang, None; P. Lin, None; J. Yuan, None; Y. Zhuang, None; X. Lu, None; Y. Bai, None; G. Zhang, None; J. Su, None; J. Qu, None; L. Xu, None; H. Wang, None