Children with Sensorineural Hearing Loss Should Be Tested with GJB2 Screen

Newswise — It is estimated that 40,000 children are born in the United States each year with significant hearing impairment, with an incidence four of 1,000 of moderate to profound sensorineural hearing loss (SNHL) at birth. Traditionally, the cause of congenital SNHL has been broadly classified as either hereditary or acquired. It is estimated that up to 50 percent of all congenital moderate-to-profound SNHL cases are genetic in origin. Of these, the majority (approximately 80 percent) is transmitted in an autosomal recessive manner. To date, there is no consensus diagnostic protocol for children presenting with SNHL. The wide range of available tests and specialties, coupled to the costs involved with an exhaustive work-up, pose an evaluation dilemma for the practitioner.

The cause of childhood SNHL can often be traced in the history to acquired, environmental causes such as: intrauterine infections, ototoxic medications, maternal or neonatal metabolic disorders, maternal illicit drug use, prematurity, low Apgar scores, and exposure to teratogens in up to 35 percent of cases. Physical examination can further demonstrate obvious associated syndromes. Unfortunately, non-syndromic hereditary SNHL is difficult to diagnose on history and physical alone. Subsequent exhaustive testing remains controversial.

A past research effort reported on a sequential diagnostic algorithm where molecular GJB2 genetic testing is obtained first. In these reports, it was suggested that temporal bone imaging should then be obtained secondarily only if the GJB2 screen is negative in cases of bilateral SNHL, and primarily in cases of unilateral SNHL. Furthermore, the authors recommended that laboratory investigation should not be routinely obtained, but based on the clinical history, genetic, and imaging results.

A new study to identify a useful diagnostic protocol was performed in a prospective fashion in as all children with idiopathic SNHL were evaluated identically in order to determine if a stepwise paradigm is more diagnostically efficient and cost-effective than the more commonly used simultaneous testing approach in pediatric SNHL due to unknown reasons. The authors of "Improved Diagnostic Effectiveness with a Sequential Diagnostic Paradigm in Idiopathic Pediatric Sensorineural Hearing Loss" are Diego A. Preciado MD, Colm Madden MD, Daniel I. Choo MD, John H. Greinwald MD, David Myer, Chris Ngo, and John K. Bradshaw, and Louise Lawson, PhD, all from the Cincinnati Children's Hospital Medical Center, Cincinnati, OH. Their findings will be presented at The Twentieth Annual Meeting of the American Society of Pediatric Otolaryngology (ASPO) http://www.aspo.us/ being held May 27-30, 2005, at the J.W. Marriott Las Vegas Resort in Las Vegas, NV.

Methodology: A prospective, longitudinal case series study at a tertiary referral children's hospital, which included 150 children who had presented with idiopathic SNHL in the last two years. All children were evaluated with full diagnostic evaluations including GJB2 screens, temporal bone CT scans, and laboratory investigations. Specific laboratory data which was obtained included: CBC, thyroid function tests, syphilitic blood tests, cholesterol and triglyceride levels, blood chemistries, ECG, and ophthalmologic testing.

Age appropriate audiometric data with air and bone thresholds was available for all patients. Auditory brainstem response (ABR) potentials were performed on children who were either too young or unable to cooperate with behavioral audiometry. The children were categorized as having either mild-to-moderate (21-54 dB HL), moderate-to-severe (55-69 dB HL), severe-to-profound (>70 dB HL), SNHL based on the means of the available bone pure tone thresholds at 500, 1000, 2000, 4000, and 8000 Hz on their initial audiometric evaluation. Children with unilateral SNHL had normal (<20 dB HL) pure tone averages (PTA) in one ear. Children with asymmetric SNHL (>15 dB HL PTA difference) were grouped according to the mean PTAs of both ears.

The main outcome measures of the study were to compare the diagnostic yields of GJB2 screens, imaging, and laboratory results per SNHL category, and determine the cumulative test-specific price estimates comparing a sequential (where some testing is avoided) vs. a simultaneous testing approach.

Results: The mean age of the 150 subjects 4.8 years, with a range of 1 week to 18 years of age. The categorization of hearing loss was as follows: 36 (24.0 percent) children had bilateral severe-to-profound SNHL, 29 (19.3 percent) had moderately severe SNHL, 64 (42.7 percent) had mild-to-moderate SNHL, and 21 (14 percent) had unilateral SNHL. Key findings were:

GJB2 screens: Eighteen patients of the150 patients tested (12 percent) were found to have bi allelic mutations in the GJB2 gene that accounted for the SNHL. The following mutations were identified: homozygous 35delG/35delG nonsense mutations leading to premature transcription stops (six patients), other bi allelic nonsense/nonsense mutations (313del14, 235delC, 312del14) leading to premature transcription stops (four patients), and at least one missense (missense/any mutation) (M34T, V37I, K15T, and L90V) leading to amino-acid substitutions in the protein (eight patients).

Radiologic imaging: Radiologic abnormalities were identified in 45 of the 150 children (30.0 percent). The diagnostic yield of CT scanning was 10/36 (27.8 percent) for patients with bilateral severe-to-profound loss, 11/29 (37.9 percent) for those with bilateral moderately severe loss, 15/64 (23.4 percent) for those with bilateral mild-to-moderate loss, and 9/21 (42.8 percent) for children with unilateral SNHL. The most frequent finding on imaging was an enlarged vestibular aqueduct (24/45 [53.3 percent]), followed by cochlear dysplasia (6/45, [13.3 percent]), and cochlear hypoplasia (2/45, [4.4 percent]). Thirteen patients (28.9 percent) had multiple abnormalities identified.

Laboratory testing: The following six lab test categories were ordered for each patient: CBC, thyroid function tests, syphilitic blood tests, cholesterol and triglyceride levels, blood chemistries, and ECG. Patients without laboratory data were excluded from this study. Thirty-five patients had ophthalmologic evaluations as well. Of the 630 laboratory tests, 45 were found to be abnormal. None contributed towards a clinical diagnosis.

Conclusions: Based upon test-specific medical savings analyses, it appears that children with severe to profound SNHL should undergo GJB2 screens first, as opposed to children with milder types of SNHL, whom should undergo imaging first (see below). For example, at the research facility, the cost of a full simultaneous SNHL evaluation is $1990 ($500 + $1100 + $360). Therefore, a GJB2 screen diagnostic yield of 22.2 percent in children with severe to profound SNHL (as per the results) would entail savings of $24,420 in imaging costs per 100 children tested. On the other hand, in this same population, a radiologic imaging yield of 27.8 percent (as per our results) would entail savings of only $13,900 in GJB2 screening costs per 100 children tested. Therefore, there are higher test specific savings when a GJB2 screen is obtained first in children with severe-to-profound SNHL. Given the diagnostic yields seen in our results, the reverse appears to be true in children with milder forms of hearing loss, where imaging obtained as a first step entails higher savings.

Accordingly, the researchers conclude that a stepwise diagnostic paradigm in which children with bilateral SNHL is more diagnostically efficient and cost effective than the more commonly used full, simultaneous testing approach. Initial results suggest children with severe to profound SNHL should first be tested with a GJB2 screen, as opposed to those with milder SNHL, who should undergo imaging as the initial testing step. Laboratory investigation should not be routine, but based on clinical history. Children with unilateral SNHL should only be investigated with imaging.