Ophthalmologic Findings in Children With Sensorineural Hearing Loss

Objectives  To determine the prevalence of refractive and nonrefractive ophthalmologic abnormalities in children with sensorineural hearing loss (SNHL) and to evaluate the overall utility of routine ophthalmological examination in children with SNHL.

Design  An institutional review board–approved retrospective analysis of ophthalmologic findings in children (18 years and younger) with SNHL seen between November 6, 2000, and June 4, 2007.

Setting  Tertiary care university children's hospital.

Patients  Children (18 years and younger) with SNHL.

Main Outcome Measures  Ophthalmologic findings and SNHL.

Results  There were 226 patients with SNHL who underwent ophthalmologic examination. Of these patients, 49 (21.7%) had an ophthalmologic abnormality. Refractive errors were present in 23 patients (10.2%), whereas nonrefractive conditions were present in 29 patients (12.8%). The cause of SNHL was syndromic in 11 patients (4.9%), of whom 5 (2.2%) had syndromes with associated ophthalmologic abnormalities. No statistically significant differences were found in ophthalmologic findings based on severity or laterality of SNHL. The prevalence of ophthalmologic findings was not significantly different between patients whose conditions were diagnosed on the basis of newborn screening and other patients. Ophthalmologic abnormalities were found in 1 of 27 patients (3.7%) with biallelic GJB2 mutations and 22 of 106 patients (20.7%) without any GJB2 mutations (P = .04).

Conclusions  In this study, the overall prevalence of ophthalmologic findings in children with SNHL was 21.7%. Ophthalmologic anomalies were present in 3.7% of children with GJB2 mutations. Routine ophthalmologic examination can be beneficial in the evaluation of children with SNHL.

Sensorineural hearing loss (SNHL) is a condition with profound implications for patients, families, and society.¹ It has been estimated that 1 to 3 children per 1000 live births have at least moderate SNHL and that 4 children per 10 000 live births have profound SNHL.^1-4 Half of all pediatric cases of SNHL result from environmental causes and the other half from genetic causes.^1,5,6 Environmental causes include congenital infections, ototoxic drug exposures, neonatal hyperbilirubinemia, and prematurity.³ Of the children with hereditary hearing impairment, approximately 70% have nonsyndromic SNHL without any other associated abnormalities. Many genes have been shown to be associated with nonsyndromic SNHL; one of them, GJB2 (OMIM 121011), accounts for at least 50% of hereditary nonsyndromic SNHL in whites.^1,5,6 Approximately 30% of heredity SNHL is syndromic and occurs with associated clinical findings in other organ systems.^1,5,6

Initial evaluation of SNHL begins with a comprehensive clinical history with emphasis on family history and the presence of risk factors during the pregnancy and prenatal period. Physical examination, audiologic evaluation, radiographic imaging, laboratory testing, genetic testing, and consultation with other specialists can be helpful in determining the cause of SNHL.^3,6,7 There is controversy about the yield of routine laboratory testing in children with SNHL.⁷ Abnormalities are found on temporal bone computed tomography (CT) in as many as 40% of patients; this technique is considered to be important in routine evaluation.^7,8

Especially early in life, SNHL is associated with delays in language, speech, cognitive, and social development.^9-11 Given the effects of hearing impairment, children with SNHL are particularly dependent on other means of information acquisition. If these children were to have unrecognized ophthalmologic abnormalities that limited visual acuity, there could be further detrimental effects on development.^12-14 For this reason, it is important to diagnose ocular anomalies early so that interventions can be undertaken to maximize visual acuity. Another reason to consider ophthalmologic evaluation is that many of the syndromic and environmental causes of SNHL also result in ophthalmologic findings.^1,15,16 Detecting these abnormalities can be helpful in determining the cause of SNHL. Children suspected or found to have a genetic syndrome can be evaluated by a medical geneticist for anomalies in other organ systems while receiving appropriate genetic counseling. The objectives of this study are to determine the prevalence of refractive and nonrefractive ophthalmologic abnormalities in children with SNHL and to evaluate the overall utility of routine ophthalmologic examination in children with SNHL.

This study is a retrospective analysis of children (18 years and younger) with SNHL seen at Children's Hospital of Pittsburgh (Pittsburgh, Pennsylvania) between November 6, 2000, and June 4, 2007. The study was reviewed by the hospital's institutional review board and approved by expedited review.

All children with SNHL for whom results of ophthalmologic evaluation were available were included in the study. The patients' medical records were reviewed for demographic information, prenatal history, patient medical history, family history of hearing loss, results of audiometric evaluation, genetic test results, findings from CT and magnetic resonance imaging (MRI), causes of SNHL (for cases in which a cause was found), and results from ophthalmologic evaluation. In addition, the reason for evaluation by an otolaryngologist for hearing loss (eg, failing newborn screen) was also recorded. Severity of hearing loss was defined as follows: mild, 26 to 40 dB; moderate, 41 to 55 dB; moderately severe, 56 to 70 dB; severe, 71 to 90 dB; and profound, greater than 90 dB.¹⁷

Comparisons of categorical variables were conducted using the χ² test and, when appropriate, the Fisher exact test. Continuous variables were analyzed using the Mann-Whitney test. All P values were calculated using 2-sided hypotheses, and the threshold for statistical significance was set at P < .05. All statistical analyses were conducted using STATA statistical software, version 9.0 (StataCorp, College Station, Texas).

There were 226 children with SNHL for whom results of ophthalmologic evaluation were available. Of these, 112 (49.5%) were female and 114 (50.5%) were male. The mean age at the time of diagnosis was 37 months (median, 33 months). Of the 226 children with SNHL in this study, 66 (29.2%) were initially detected by newborn screening and then seen in our clinic. Sensorineural hearing loss was bilateral in 174 patients (76.9%) and unilateral in 52 (23.0%). On the basis of audiometric evaluation, hearing loss was mild in 37 patients (16.4%), moderate in 58 (25.7%), moderately severe in 20 (8.8%), severe in 36 (15.9%), and profound in 75 (33.2%).

After ophthalmologic examination, 49 patients (21.7%) were diagnosed as having an ophthalmologic abnormality (Table 1). Refractive errors, such as myopia, hyperopia, and astigmatism, were present in 23 patients (10.2%). Nonrefractive conditions, such as strabismus and retinitis pigmentosa, were diagnosed in 29 patients (12.8%).

Of all the patients, a syndromic cause of SNHL was found in 11 patients (4.8%). Of these, 5 patients (2.2%) had syndromes with associated ophthalmologic abnormalities. Two patients had Waardenburg syndrome (1 had pale irises and 1 had white eyelashes). One patient with myopia was diagnosed as having Stickler syndrome. Another patient was found to have peripheral vision impairment and was diagnosed as having Usher syndrome. In these 4 patients, the ophthalmologic examination was helpful because it suggested or confirmed the diagnosis of the respective syndromes. The fifth patient had bilateral persistent hyperplastic primary vitreous that resulted in blindness; multiple syndromes were in the differential diagnosis. He was diagnosed as having SNHL many years after his ophthalmologic anomalies were detected. Two additional patients may have had Usher syndrome. On ophthalmologic examination, 1 had retinitis pigmentosa and 1 had retinal degeneration. These 2 patients were siblings and had other organ abnormalities; multiple syndromic causes are being considered.

The severity of hearing loss was categorized as mild, moderate, moderately severe, severe, or profound. The prevalence of all ophthalmologic findings, refractive errors, and nonrefractive conditions was determined within each of these groups (Table 2). No statistically significant differences were found in the prevalence of ocular anomalies based on severity of SNHL. The prevalence of ophthalmologic findings was compared between children with unilateral and bilateral SNHL, and no statistically significant differences were found between these groups (Table 3). Hearing loss was initially detected by newborn screening in 66 patients. In this subgroup of children, 14 of 66 (21.2%) had an ophthalmologic abnormality, 6 of 66 (9.0%) had refractive errors, and 9 of 66 (13.6%) had nonrefractive conditions. These prevalences were not significantly different between the children who failed newborn screening and the other children with SNHL.

Patients in the study were offered genetic testing for mutations in GJB2. The results of genetic testing revealed that 27 of 144 patients (18.8%) had biallelic mutations in GJB2 and 11 of 144 (7.6%) had a single allele mutation in GJB2. Ophthalmologic abnormalities were found in 1 of 27 patients (3.7%) with biallelic GJB2 mutations, none of the patients with single allele GJB2 mutations, and 22 of 106 patients (20.8%) with no mutations in GJB2. The difference in ophthalmologic findings between patients with biallelic mutations and those without GJB2 mutations was statistically significant (P = .04). The only ocular finding in patients with biallelic GJB2 mutations was bilateral ptosis. Of the 106 patients with no GJB2 mutations, 101 had nonsyndromic, non-GJB2 SNHL. Among the patients with nonsyndromic, non-GJB2 SNHL, 19 of 101 (18.8%) had ophthalmologic abnormalities, a percentage which neared statistical significance when compared with that of the patients with biallelic GJB2 mutations (18.8% vs 3.7%; P = .07).

When comparing children with no GJB2 mutations, a single GJB2 mutation, and biallelic GJB2 mutations, the mean ages were 38, 42, and 20 months, respectively. A significant difference in age was found between the children with no mutations and those with biallelic mutations (P = .04).

The results of imaging studies were available in 216 patients (95.6%); 168 patients underwent evaluation by CT only, 37 by CT and MRI, and 11 by MRI only. Abnormal imaging findings (on CT and/or MRI) were present in 56 of 216 patients (25.9%). The most common anomaly detected on imaging was large vestibular aqueduct (LVA), which was present in 23 of 216 patients (10.6%). The prevalence of ophthalmologic anomalies was not significantly different between children who had an imaging finding and children with normal imaging study results, or between children with LVAs and children without LVAs (Table 4).

The overall prevalence of ophthalmologic findings from this study was compared with the prevalence in a group of children from the general population. As part of the Baltimore Vision Screening Project,¹⁸ 285 elementary schoolchildren from inner-city Baltimore, Maryland, were evaluated for ophthalmologic abnormalities. The mean age of these children was 5.1 years, and 40 of 285 (14.0%) were found to have an ophthalmologic abnormality. The prevalence of ophthalmologic findings in this sample was significantly different from the prevalence in the current study of children with SNHL (14.0% vs 21.7%; P = .02).

The aims of this study were to determine the prevalence of ophthalmologic findings in children with SNHL and to examine the role of ophthalmologic consultation in their evaluation. Children with SNHL are known to be at increased risk for delayed language, speech, cognitive, and social development.^9-11 Like hearing, vision plays a key role in gathering information from the environment. For that reason, it is important to ensure that visual function is optimized in children who have SNHL, especially in the first few years of life, during which there are many key developmental milestones. The importance of hearing and vision is one of the reasons to advocate ophthalmologic evaluation in children with SNHL. Previous studies^13,19 have indicated that the prevalence of refractive errors and nonrefractive conditions is higher in children with SNHL than in other children, a finding which also supports the need for routine ophthalmologic examination in this population.

Among children with SNHL, we found that the prevalence of ophthalmologic findings was 21.7%. There are limited studies that evaluate the overall prevalence of ophthalmologic abnormalities in the general pediatric population. A large study from the Centers for Disease Control and Prevention reported that 2.5% of American children have visual impairment. However, this finding was based on parental surveys and did not include children who already benefited from corrective lenses.²⁰ For these reasons, the reported prevalence is likely an underestimation of all ophthalmologic anomalies. A study¹⁸ by the Baltimore Vision Screening Project evaluated inner-city elementary schoolchildren for ophthalmologic abnormalities. This study found that 14% of children had an ocular abnormality, which was significantly lower than the prevalence in children with SNHL from the present study.

Many of the children in our study had ocular conditions for which interventions (eg, refractive lenses or surgical procedures) and continued ophthalmologic care can restore normal visual function or minimize further vision loss. In addition to suggesting ways to optimizing vision, ophthalmologic consultation can be helpful in evaluating the causes of SNHL. Some causes of SNHL (eg, Waardenburg syndrome or congenital rubella) have associated ophthalmologic abnormalities that can lead physicians who evaluate hearing loss to the correct cause. Knowledge of the cause can be comforting to patients and family members while providing relevant prognostic information. In this study, ophthalmologic evaluation was helpful in diagnosis of a syndrome in 4 patients. In cases in which syndromic causes are found, patients can be evaluated for other anomalies and offered genetic counseling.

Previous studies^{12-14,19,21-26} have reported ocular findings in 8% to 61% of children with SNHL. We believe that the variation in the reported prevalence of ophthalmologic abnormalities in the literature is likely a reflection of factors such as the age of the children and the specific population studied (clinic^12,19 vs institutional^13,24; only cochlear implant candidates).²² Causes of SNHL varied notably among these studies; for example, some of these studies had many more patients with congenital rubella (25% of patients in one study²⁴) than others. This factor is important because congenital rubella is associated with ocular abnormalities, such as cataracts and retinopathy, and a sample with many patients with congenital rubella would likely also have a high prevalence of ophthalmologic anomalies. The studies also had variations in the inclusion and exclusion criteria (all pediatric SNHL^13,19,24 vs unexplained SNHL¹²), which are another source of selection bias. Furthermore, there are differences in institutional referral patterns, the scope of the ophthalmologic examination, and the definition of ophthalmologic abnormalities, all of which likely affected the studies' results.

In this study, the prevalence of ophthalmologic abnormalities was not significantly affected by the severity (mild, moderate, moderately severe, severe, or profound) or laterality (unilateral vs bilateral) of SNHL. Furthermore, we found that ophthalmologic abnormalities were not significantly affected by the presence of LVAs or other imaging anomalies, in confirmation of similar findings by Mafong et al.¹² Although audiometric and imaging studies can be helpful in the overall evaluation of SNHL, they do not seem to be informative with respect to ocular findings.

Recent molecular findings have contributed to our understanding of hearing and hearing loss. Many genes have been implicated in syndromic and nonsyndromic hearing loss. Mutations in GJB2, which codes for connexin 26, are present in at least 50% of white patients with hereditary nonsyndromic SNHL.^1,5,6 Although many mutations have been described, the 35delG mutation is most common in the majority of populations and results in a frameshift with premature termination of protein translation. Connexin 26 is a gap junction protein and has been localized to the inner ear. It plays a role in intercellular communication, possibly by affecting the electrochemical milieu and potassium recirculation. The exact mechanism by which connexin mutations result in SNHL is an area of active investigation.^1,5,6 We found that children with biallelic GJB2 mutations had a significantly lower prevalence (3.7%) of ophthalmologic anomalies than children with normal GJB2 (20.8%). This is consistent with the impression that GJB2 mutations result in SNHL but not in additional anomalies or syndromes.^5,6 However, the children with biallelic GJB2 mutations were significantly younger than those without any mutations, which could explain the decreased prevalence of ophthalmologic abnormalities.

The present study has limitations. Even though ophthalmologic evaluation was recommended for all children with SNHL seen in our clinic, not all families followed this recommendation. Families with specific concerns regarding visual impairment were probably more likely to see an ophthalmologist, which may have biased our sample of patients. Furthermore, we did not specify certain tests in the ophthalmologic examination, so it is possible that some patients had more thorough evaluations than others.

A multidisciplinary approach is important in the evaluation and treatment of children with SNHL to ensure that their medical, educational, and social needs are met. A battery of various laboratory tests has been used as part of routine evaluation in the past; however, it may be more appropriate to use information gathered from patient history or physical examination to guide laboratory testing and studies.⁷ On the other hand, radiographic anomalies are common in children with SNHL, and the use of imaging studies (usually CT) has been advocated for all patients.^7,8 Similarly, GJB2 testing should be performed with children for whom a clear cause is not evident. Further genetic tests (such as for syndrome-specific and mitochondrial mutations) can be used when indicated.⁶ In addition to these tests and studies, we recommend that all children with SNHL should undergo routine ophthalmologic examination. The Joint Committee on Infant Hearing has also recommended that ophthalmologic evaluation be performed on children with SNHL.²⁷ Ophthalmologic evaluation can be beneficial for patients by allowing ophthalmologists to diagnose (and possibly treat) coexisting disorders that affect vision and by helping otolaryngologists to determine the cause of SNHL.

In this study, the overall prevalence of ophthalmologic findings in children with SNHL was 21.7%. Notably, 10.2% and 12.8% of children with SNHL had refractive and nonrefractive abnormalities, respectively. The only subgroup of patients to have a significantly different prevalence of ophthalmologic abnormalities was children with biallelic GJB2 mutations (3.7%). Thus, routine ophthalmologic examination can be beneficial in all children with SNHL.

Correspondence: David H. Chi, MD, Children's Hospital of Pittsburgh, 3705 Fifth Ave, Ste 1555-ENT, Pittsburgh, PA 15213 (david.chi@chp.edu).

Submitted for Publication: December 5, 2007; final revision received April 18, 2008; accepted May 7, 2008.

Author Contributions: Dr Chi had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Chi. Acquisition of data: Ruscetta. Analysis and interpretation of data: Sharma, Ruscetta, and Chi. Drafting of the manuscript: Sharma. Critical revision of the manuscript for important intellectual content: Ruscetta and Chi. Statistical analysis: Sharma. Administrative, technical, and material support: Ruscetta. Study supervision: Chi.

Financial Disclosure: None reported.