Ocular manifestations of CHARGE syndrome in a pediatric cohort with genotype/phenotype analysis

CHARGE syndrome is a rare multi‐system condition associated with CHD7 variants. However, ocular manifestations and particularly ophthalmic genotype–phenotype associations, are not well‐studied. This study evaluated ocular manifestations and genotype–phenotype associations in pediatric patients with CHARGE syndrome. A retrospective chart review included pediatric patients under 20 years‐old with clinical diagnosis of CHARGE syndrome and documented ophthalmic examination. Demographics, genetic testing, and ocular findings were collected. Comprehensive literature review enhanced the genotype–phenotype analysis. Forty‐two patients (20 male) underwent eye examination at an average age of 9.45 ± 6.52 years‐old. Thirty‐nine (93%) had ophthalmic manifestations in at least one eye. Optic nerve/chorioretinal colobomas were most common (38 patients), followed by microphthalmia (13), cataract (6), and iris colobomas (4). Extraocular findings included strabismus (32 patients), nasolacrimal duct obstructions (11, 5 with punctal agenesis), and cranial nerve VII palsy (10). Genotype–phenotype analyses (27 patients) showed variability in ocular phenotypes without association to location or variant types. Splicing (10 patients) and frameshift (10) variants were most prevalent. Patients with CHARGE syndrome may present with a myriad of ophthalmic manifestations. There is limited data regarding genotype–phenotype correlations and additional studies are needed.

occur in the iris (Zentner et al., 2010).Colobomas are the predominant ocular feature associated with CHARGE syndrome, but other ophthalmic manifestations including microphthalmos, microcornea, refractive error (usually myopia), and strabismus have been reported in affected individuals as well (Zentner et al., 2010).CHARGE syndrome is caused by heterozygous pathogenic variants in CHD7, a gene encoding chromodomain-helicase-DNA-binding protein 7 (CHD7), a transcriptional regulator crucial in embryonic development.Previous studies are limited in descriptions of the various ophthalmic findings and have not identified specific genotypephenotype correlations (Dana et al., 2023;Janssen et al., 2012;Jongmans et al., 2006;Lalani et al., 2006;Legendre et al., 2017;Legendre et al., 2018;Sohn et al., 2016;Sun et al., 2022;Vissers et al., 2004;Zentner et al., 2010).Since other studies rarely comment on the breadth of ophthalmic manifestations in CHARGE syndrome and ophthalmic genotype-phenotype associations, we aimed to address these gaps with providing both new data and a review of the literature.First, we aimed to evaluate and characterize ophthalmic manifestations of CHARGE syndrome in a large cohort of pediatric patients and examine potential genotype-phenotype correlations.Second, we reviewed the literature to report on ophthalmic genotype-phenotype associations for the variants identified in this cohort.Third, we completed a literature review to summarize the prevalence of ophthalmic findings observed in other groups with CHARGE syndrome.The variants that were classified as of uncertain significance on initial laboratory reports were re-interpreted by genetic counselors with variant curation expertise.Variants were assessed and classified in accordance with ACMG-AMP 2015 sequence (Richards et al., 2015) or 2020 ACMG-ClinGen CNV guidelines (Riggs et al., 2020).These analyses included a review of publicly available databases (e.g., ClinVar (Landrum et al., 2018), Leiden Open Variation Database (Fokkema et al., 2021)) and the latest biomedical literature for data relevant to each variant.Variant minor allele frequency and gene constraint data were obtained from the Genome Aggregation Database (Karczewski et al., 2020) (gnomAD) v2.1.1 and v3.1.2.In silico missense pathogenicity predictions were performed using REVEL (Ioannidis et al., 2016) (benign: ≤0.290; pathogenic: ≥0.644 (Pejaver et al., 2022)) and the splicing prediction algorithms SpliceSiteFinder-like (Shapiro & Senapathy, 1987), MaxEntScan (Yeo & Burge, 2004), NNSplice (Reese et al., 1997), and GeneSplicer (Pertea et al., 2001) (predicted impact: Δ ≥10% between wild-type and variant), and SpliceAI (Jaganathan et al., 2019) (predicted impact: ≥0.2; no impact: <0.2).Updated variant classifications were then compared to the original classifications.

| METHODS
A comprehensive literature search on Human Gene Mutation Database (HGMD; Stenson et al., 2014) was performed to identify publications reporting the same variants identified in this study.Ophthalmic findings were extracted from the literature and then compared with our patient cohort.The literature was also reviewed to identify the prevalence of ophthalmic manifestations of CHARGE syndrome.

| Ophthalmic manifestations
Thirty-nine patients (93%) had ophthalmic manifestations in at least one eye.Optic nerve and chorioretinal colobomas were identified in 38 patients (90%) of which 23 patients had bilateral colobomas and 15 had a unilateral coloboma (Table 2).Seventeen patients (40%) had at least one eye without posterior coloboma.Only five eyes of four patients had iris coloboma, all of which also had posterior segment colobomas.Thirteen patients (31%) had microphthalmia of one eye (10 patients) or both eyes (2 patients).Microphthalmia was diagnosed either by measurement of axial length, radiological imaging of the orbits, or clinical observation.All patients with microphthalmia had other ophthalmic findings: posterior segment colobomas (14 eyes of 12 patients; 1 additional patient had no view to the posterior pole due to corneal opacification), iris colobomas (3 eyes of 3 patients), and cataracts (4 eyes of 3 patients).Six patients (14%) had a cataract (2 bilateral) that was associated with optic nerve/chorioretinal in 5 patients and iris colobomas in 2 patients.Three patients with cataracts had retinal detachments.Of the six patients (eight eyes) with cataracts, two patients had unilateral cataract extraction.Five eyes in four patients (9%) had retinal detachments, all of which were associated with a confirmed or suspected chorioretinal coloboma (one eye presented with a detachment), and four eyes of four patients underwent pars plana vitrectomy (four eyes) with or without scleral buckle placement (two eyes).Two patients (5%) were diagnosed with glaucoma.One patient had isolated chorioretinal coloboma while the other had a history of chorioretinal coloboma complicated by retinal detachment repaired with silicone oil and scleral buckle.
Strabismus was documented in 32 patients (76%) and was the most common extraocular finding.Eleven patients (26%) had nasolacrimal duct obstructions in one eye (1 patient) or both eyes (10 patients).Five of those patients were noted to have punctal agenesis in at least one puncta.Ten patients (24%) had unilateral (eight patients) or bilateral (two patients) cranial nerve VII palsies.
Optotype LogMAR best corrected visual acuity (BCVA) was available for 22 patients.Average LogMAR BCVA of the better seeing eye of each patient was 0.34 ± 0.31 (median 0.30, range 0-1.1).The remainder of the patients were either unable to do optotype visual acuity testing due to young age, developmental delays, or low vision.Systemic manifestations of our patients with genetic testing are described in Table S1.
Literature review found reports describing ophthalmic findings associated with 4 of the 27 variants identified in this study.Table S2 provides the ophthalmic findings in patients with these four variants from this study and the literature.

| Genetic analysis
Genetic testing results were available for 27 patients with CHD7 variants (Table 3).The 27 identified variants are mapped on a pictoral representation of CHD7 in Figure 1.Twenty-five (93%) patients had positive diagnostic yield after initial testing; 26 (96%) had positive diagnostic yield following re-interpretation of the identified variants.
All disease-causing (likely pathogenic and pathogenic) variants were presumably loss-of-function in nature, likely resulting in severe protein truncation or, more likely, loss of allelic expression due to nonsense-mediated decay.This includes the missense variant in patient 36, in which the variant alters the last nucleotide of an exon and is strongly predicted to cause loss of the donor splice site, and should therefore be expected to confer disease due to its effect on splicing compared to the typical effect of missense variants.In all cases in which parental testing was completed (n = 2) the variants were found to be de novo.Except for patients 21 and 22 who are monozygotic twins, no studied individuals had family histories significant for CHARGE syndrome.Splicing (n = 10 patients) and frameshift (n = 10) variants were most commonly identified, followed by nonsense variants (n = 6) and an entire gene deletion (n = 1).Variants  Yellow circles correspond to VUS, Orange circles with LP-P and Red circles P variants.The solid/dashed/dotted edges are frameshift/splicing/nonsense variants, respectively.to be underreported.In the current study, we present a cohort of patients with CHARGE syndrome and in addition to comprehensive eye analysis, we also assessed genotype-phenotype associations.
Colobomas associated with CHARGE syndrome are most often localized to the posterior segment (optic nerve/chorioretinal) with rates ranging from 57% to 83% (Table S3; Aramaki et al., 2006;Dana et al., 2023;Delahaye et al., 2007;Issekutz et al., 2005;Jongmans et al., 2006;McMain et al., 2008;Russell-Eggitt et al., 1990;Sohn et al., 2016;Tellier et al., 1998).While this study found a higher rate of posterior colobomas (90% of patients with unilateral or bilateral coloboma), it is possible our number is higher since we only included patients who had ophthalmologic exams at our institution.Our cohort may have been specifically referred to us due to eye findings, and there may have been other patients with CHARGE syndrome without colobomas followed by outside ophthalmologists or optometrists.Iris colobomas are less common in CHARGE syndrome, with a reported rate range of 0%-26% (Aramaki et al., 2006;Jongmans et al., 2006;McMain et al., 2008;Russell-Eggitt et al., 1990;Sohn et al., 2016;Tellier et al., 1998).Similarly, our study found only 10% of patients with iris colobomas, which may be due to a greater role of CHD7 in retinal patterning versus more anterior structures of the optic cup.
In practice, microphthalmia and coloboma are commonly associated with cataract, retinal detachment and glaucoma, however, there are few reports of these findings in CHARGE syndrome.This may be due to incomplete ocular phenotype information, especially in uncooperative children with limited visual function.Fourteen percentage of patients in our cohort had cataract, which was higher than a report of 50 patients with CHARGE syndrome that showed 4 patients with cataract and 1 with lens dislocation (Russell-Eggitt et al., 1990).This included two patients with cataract secondary to persistent fetal vasculature, which was not found in our cohort.Further, Russell-Eggitt et al. reported one patient with cataract and retinal detachment, although it is not clear whether this was also in the context of coloboma.Posterior colobomas are often complicated by retinal detachment due to weakness of the retina at the edges of the coloboma (Lingam et al., 2021), nevertheless there are only a few case reports of retinal detachment in CHARGE syndrome (El Hamichi et al., 2020;Russell-Eggitt et al., 1990;Yu-Chuan Kang et al., 2022) In contrast, we found that almost 10% of CHARGE patients had retinal detachment and all were associated with coloboma and underwent surgical repair.Likewise, there are only a few case reports of glaucoma in CHARGE syndrome (Benson et al., 2017;Katai et al., 1997).Two patients in our cohort had glaucoma with varying mechanisms underlying the elevated intraocular pressure.
There are also extraocular manifestations in CHARGE syndrome that can affect eye function and vision (Table S4).CN VII palsies can be found in CHARGE syndrome with a wide range of reported prevalence between 9% and 89% (Bergman et al., 2008;Dana et al., 2023;Jongmans et al., 2006;Jyonouchi et al., 2009;Legendre et al., 2017;McMain et al., 2008;Russell-Eggitt et al., 1990;Sohn et al., 2016;Strömland et al., 2005;Tellier et al., 1998;Vissers et al., 2004).In our cohort, the rate of CN VII palsies was 24%, which is on the low end of the reports from the literature.This may be due to under-diagnosis as sometimes these findings are subtle or not always considered by ophthalmologists.Strabismus is also reported in 34%-78% of CHARGE patients, although the type of eye misalignment varies and includes both sensory as well as extraocular motility abnormalities such as upgaze deficiencies.We found similarly that 76% of our patients had strabismus, often but not always associated with low vision.Interestingly, we found a high rate of nasolacrimal duct obstruction (26%), which had only been commented on in one prior study which showed a rate of 4% (Russell-Eggitt et al., 1990).We also found that almost half of the patients with nasolacrimal duct obstruction had punctal agenesis; to our knowledge, this has not previously been reported in individuals with CHARGE syndrome and may represent a slight expansion of the ophthalmological phenotype.Both nasolacrimal duct obstruction and punctal agenesis are associated with facial anomalies, and is therefore not a surprising finding (Dohlman et al., 2019), though has likely been under-reported in the past due to focus on some of the more severe disease manifestations.
Variants in CHD7 are commonly associated with CHARGE syndrome.While previous studies from 2004 to 2009 have shown positive diagnostic yields of 67% in patients clinically diagnosed with CHARGE syndrome (Zentner et al., 2010), a report in 2014 described a positive yield of 90%-95% of clinically diagnosed patients (Hsu et al., 2014) and we found a 93% positive yield on initial testing which increased to 96% after re-classification.This discrepancy with older reports may be due to improved genetic testing in recent years such as including more intronic variants (Tan et al., 2020).Due to our very high positive diagnostic yield, it was not valuable to compare phenotypes of genotype positive versus negative patients as has been done in other studies (Lalani et al., 2006;Legendre et al., 2017;Russell-Eggitt et al., 1990).However, it is important to note, that genetic results were available from only 27 of the 42 patients in our cohort and that the diagnostic yield might have been lower if all patients had undergone testing.
Our combined rate of 59% of frameshift (37%) and nonsense (22%) variants was lower than the previously published studies as well as the gene databases.In contrast, we found that the rate of intronic splicing variants (37%) was higher than prior studies (13%-28%) and the database (11%).This may be related to our overall higher diagnostic yield rate and the greater inclusion of intronic regions on test.
Numerous studies have attempted to correlate variant location and type with phenotypes, but rarely considered ophthalmic findings.For example, loss-of-function variants near the 3 0 end of CHD7 that are not expected to trigger nonsense-mediated decay have been reported in individuals with more mild systemic phenotypes (Dana et al., 2023).
However, there does not seem to be a correlation with eye phenotypes as both Dana et al. and our data showed that patients with lossof-function variants in exons 36 and 37 had colobomas.Intron 25 has been proposed to be a hotspot (Legendre et al., 2018), with two recurrent pathogenic variants (c.5405-7G>A, c.5405-17G>A) in this region identified in our study population (Table S2).Nevertheless, the phenotypes associated with variants in this region vary as seen in our two patients, one of whom had unilateral posterior coloboma and CN VII palsy, the other without abnormal ophthalmic findings.Interestingly, the five patients in our study who had nasolacrimal duct obstruction with punctal agenesis had either nonsense or frameshift variants clustered in exons 30 and 31.Since this is the first report of this phenotype and we have a relatively small number of patients, more studies are needed to determine whether this is a true genotype-phenotype association; as loss-of-function variants in these exons are expected to result in loss of protein expression, it is likely that this is an artifact of our sample size.Further, splice site and missense variants, which are less common may also yield a more mild overall phenotype that in some patients may lead to less severe undiagnosed disease (Janssen et al., 2012).While two of our patients without ophthalmic manifestations had splicing variants, the third (patient 27) had a frameshift variant in exon 38.This lack of ocular phenotype in this patient is likely due to its position at the 3 0 end of the gene, causing a relatively small truncation of the CHD7's C-terminus, as was noted previously by Dana et al. (2023).In contrast, nonsense and frameshift variants expected to result in loss of protein expression are thought to cause more severe systemic manifestations, although there does not seem to be a difference in rates of colobomas (Legendre et al., 2017).Overall, while these studies suggest there may be some genotype-phenotype associations, specifically with variants affecting splice sites, the overwhelming analysis suggests significant genotype-phenotype variability.This is consistent with the significant variability in both intra-and inter-familial phenotypes associated with identical CHD7 variants in this study and the literature (Delahaye et al., 2007;Janssen et al., 2012;Jongmans et al., 2006;Lalani et al., 2006;Tellier et al., 1998).

| CONCLUSIONS
The spectrum of ophthalmic manifestations in patients with CHARGE syndrome varies widely, with most patients having at least one ophthalmic abnormality.There is sparse information in the literature on these ocular findings and even less information on genotype-phenotype associations with variants in CHD7.This report provides ophthalmic genotype-phenotype information for a large cohort of patients with CHARGE syndrome as well as a review of the literature for the prevalence of previously reported ophthalmic manifestation.
To the best of our knowledge, this is the first study to describe punctal agenesis in association with CHARGE syndrome.This report also suggests high positive diagnostic yield with genetic testing in patients with clinical diagnoses of CHARGE.Ophthalmic genotype-phenotype heterogeneity was found in terms of location and type of variants amongst study patients and literature review, however more detailed reporting is needed to better understand any potential associations.

A
retrospective case series identified pediatric patients (<20 years old) clinically diagnosed with CHARGE Syndrome who were evaluated at the Ann & Robert H. Lurie Children's Hospital of Chicago between 2004 and 2023.Patients also had at least 1 documented eye examination.This study was approved as exempt and granted a full waiver of HIPAA authorization due to its retrospective nature by the Institutional Review Board of Lurie Children's Hospital, abided by the tenets of the Declaration of Helsinki, and conducted in accordance with the Health Insurance Portability and Accountability Act.Data collected included age at presentation and most recent visit, self-reported race and ethnicity, family history of CHARGE syndrome, intraocular and extraocular phenotypes, systemic findings, and genetic testing results.From the most recent ophthalmic examination, best corrected visual acuity (BCVA), cycloplegic refraction, and external, slit lamp and fundus findings were documented.Visual acuity was assessed with age-appropriate optotypes and then converted to Log-MAR visual acuity.CHD7 variant nomenclature, zygosity, inheritance, and classification were obtained when available.
T A B L E 2 Ophthalmologic findings.Ophthalmologic finding (number of patients, percentage), laterality (number of patients), total N = 42 patients Optic nerve/chorioretinal coloboma (38, 90%) spanned almost the entirety of the gene from exon 2 to exon 38.During re-classification of variants of uncertain significance (VUS), patient #5's splicing variant c.2096+3del in intron 3 was still considered a VUS.Patient #12 had 2 VUSs: one was upgraded in this study to pathogenic (c.5210+3A>G in intron 23), the other downgraded to benign (c.3697G>A [p.Gly1233Ser] in exon 15).Ophthalmic phenotype information for each patient with genetic testing is summarized in Table 4.All variants (n = 4) identified in patients with nasolacrimal duct obstructions (NLDO) associated with agenesis of at least one punctum were clustered in exons 30 and 31; these comprised nonsense and frameshift variants, but no splicing variants.Strabismus, cranial nerve (CN) VII palsy, microphthalmia and T A B L E 3 Variants in CHD7 (NM_017780.4) for each patient with CHARGE.

F
I G U R E 1 Variant mapping on CHD7.Yellow circles correspond to VUS, orange circles with LP-P, and red circles P variants.The solid/ dashed/dotted edges are frameshift/splicing/nonsense variants, respectively.optic nerve/chorioretinal colobomas were present in individuals with various types of variants along the length of the gene.The one patient with iris coloboma and cataract was patient #21 with a frameshift variant in exon 31.This patient's monozygotic twin (#22) had a different ocular phenotype with a posterior coloboma in both eyes but no iris coloboma or cataract.Patient #19 with a retinal detachment and likely secondary cataract and glaucoma had a frameshift variant in exon 31.Patient #23, with a splicing variant in intron 32, had glaucoma in the right eye, microphthalmia in the left eye, and bilateral optic nerve/chorioretinal colobomas.The three patients without any abnormal eye findings were #12, #15, and #27 with a splicing variant in intron 23, a splicing variant in intron 25 and a frameshift variant in exon 38, respectively.4 | DISCUSSION Eye function and vision can be affected by both intraocular and extraocular manifestations in CHARGE syndrome.The most commonly reported of these ocular findings are colobomas and cranial nerve seven (CN VII) palsies since they are part of the diagnostic criteria.Nevertheless, there are additional ophthalmic findings that tend T A B L E 4 Ocular findings for each patient with genotype information.