Genetic and clinical analysis of nonsyndromic hearing impairment in pediatric and adult cases

This prospective study included 263 NSHI patients who first visited the Ear, Nose and Throat (ENT) Department at the Central Hospital of Zhumadian, Zhumadian City, Henan Province, People’s Republic of China (PRC) between March 2008 and March 2013. The 168 males (63.9%) and 95 females (36.1%) had a mean age of 13.1 + 10.8 years (range: 2 months to 60 years). The mean age of NSHI onset was 5.72 years. The study excluded SHL patients and patients with complications resulting from causes such as tympanitis, meningocephalitis, late-stage Meniere’s disease, trauma, acoustic neuroma and ototoxic drugs. The subjects were divided into age groups according to their ages at the first hospital visit and at onset of NSHI: adult group (>18 years old) and pediatric group (<18 years); the pediatric group was subdivided into infants (<3 years), preschool (3-6 years) and school-age (7-18 years) groups. The age of onset was defined as the age at which a patient or patient’s family discovered the patient’s deafness, or the age at which objective audiometry identified the condition. This study was approved by the Hospital Ethics Committee, and written informed consent was obtained from the patients or their families.

All the patients were assessed by audiometry, including pure tone audiometry (PTA) or auditory brainstem response (ABR) tests and multiple-frequency auditory steady-state response (ASSR). The criteria for analysis and classification of deafness were in accordance with published guidelines [11] as follows: 1) hearing loss was classified by frequency measurements as full-frequency hearing loss (0.25-8.0 kHz), high-frequency hearing loss (2.0-8.0 kHz), mid-frequency hearing loss (0.5-2.0 kHz), or low-frequency hearing loss (0.25-0.5 kHz); 2) according to the stages of speech development, deafness was classified into post-linguistic deafness (>3 years) or pre-linguistic deafness (<3 years); and 3) the severity of hearing impairment was judged by the better-hearing ear as being mild [20-40 decibels (dB), hearing level (HL)], moderate (41-70 dB HL), severe (71-95 dB HL), or profound (>95 dB HL).

Peripheral venous blood (4 mL) was collected from each subject. The Universal DNA Isolation Kit (BioTeke Corporation, Beijing, PRC) was used to isolate genomic DNA according to the manufacturer’s instructions. Polymerase chain reaction (PCR) was used to amplify the GJB2 gene for mutation analysis. The PCR primers, synthesized by Sangon Biotech (Shanghai, PRC) were as follows: downstream primer (5’-GGG CAA TGC TTA AAC TGG C-3’); upstream primer (5’-TAT GAC ACT CCC CAG CAC AG-3’) [12]. The PCR product was recovered for sequence determination. The sequencing results were compared with the published GJB2 gene sequence (GenBank accession number M86849) to determine the presence of deafness-related sequence variants.

Primers for PCR amplification covered mtDNA 1988-2007 and mtDNA 618-635, and PCR amplification involved all fragments of the mitochondrial 12S rRNA gene. Twenty-four sets of primers, covering the whole mitochondrial genome with partially overlapping fragments, were used to perform PCR amplification for mtDNA from patients with the A1555G/C1494T mutations [13]. After the PCR-amplified DNA was recovered from gel with the Agarose Gel DNA Purification Kit Ver. 2.0 (Code No. DV805; TaKaRa Biotech Co. Ltd., Dalin, PRC), the BigDye® Terminator Cycle Sequencing Kit (Microread Genetics, Beijing, PRC) was used to perform sequencing on the ABI PRISM™ 3700 DNA automated sequencer (Biocoen Biotechnology, Beijing, PRC). Sequencing results were compared with the Cambridge Reference Sequence (GenBank accession number NC_001807).

EpiData version 3.1 was used to create a data bank using double data entry, and logic checks were performed. SAS 9.2 (SAS Institute, Cary, NC, USA) was used to analyze data by the χ² test. A value of p <0.05 was considered to indicate a statistically significant difference.

Of the 263 NSHI patients, 49 (18.6%) exhibited a GJB2 gene mutation: 29 cases were homozygous for a mutation, eight had compound heterozygous mutations and 12 had a single heterozygous mutation (Table 2). The types of mutation included 235delC, 176del6bp, 512insAACG, and 299delAT; 235delC was the most common type of GJB2 gene mutation, occurring in a total of 42 cases, 30 of whom were homozygous and the other 12 were heterozygous. Thirty patients (11.4%) carried mtDNA mutations; 28 cases had the C1494T mutation and two cases had the A1555G mutation.

Patients were categorized by age at the ENT Department visit (Figure 1, Table 3), resulting in 67 adults and 196 pediatric cases (72 infants, 54 preschool cases, and 70 school-age cases). The GJB2 gene mutations were detected in 5.97% adults and 22.96% pediatric cases; this difference was statistically significant (χ² = 9.506, p = 0.002). In contrast, mtDNA A1555G/C1494T mutations were detected in 31.34% adults and 4.59% pediatric cases; this difference was also statistically significant (χ² = 35.359, p <0.001). Within the pediatric group, the distributions of both GJB2 gene and mtDNA mutations among the subcategories were statistically similar (each p value >0.05).

Patients were divided by age of onset of NSHI, resulting in 26 adult cases and 237 pediatric cases (186 infants, 21 preschool cases, and 30 schoolage cases) (Figure 2, Table 4). The GJB2 gene mutations were not detected in any of the adult-onset cases, but were present in 20.68% of pediatric cases; this difference was statistically significant (p = 0.006). In contrast, mtDNA A1555G/C1494T mutations were detected in 15.38% of adult-onset patients and 8.86% of pediatric cases; this difference was not statis-tically significant (p = 0.288). Differences in mutation distributions in the pediatric subcategories were not statistically significant for either the GJB2 gene or mtDNA A1555G/C1494T mutations (each p value >0.05).

For the pediatric cases, hearing loss was further classified as either pre-lingual or post-lingual (Table 5). The difference in GJB2 gene mutation between these two groups was statistically significant (χ² = 4.683, p = 0.031); in contrast, the difference in mtDNA A1555G/C1494T mutations between these groups was not statistically significant (χ² = 0.695, p = 0.404).

Patients who were positive for any mutation were subdivided into three age categories by mutation status and age of onset (Table 6). Patients with the GJB2 gene mutation primarily experienced onset within the first year of life (65.31%). Those with mtDNA A1555G/C1494T mutations were nearly evenly divided between the three onset age groups. However, the difference in onset age distributions between those with GJB2 gene mutations and those with mtDNA mutations was statistically significant (p <0.05).

Finally, patients were categorized according to their hearing loss phenotypes and mutation status. Patients with mtDNA A1555G/C1494T mutations generally exhibited mild-to-moderate hearing loss. In contrast, most patients with GJB2 gene mutations exhibited profound hearing loss. This difference in phenotypic distributions was statistically significant (p <0.05) (Table 7). Of the 30 patients with mtDNA A1555G/C1494T mutations, 22 underwent complete PTA. Severity of hearing loss was again judged according to the multiple grading standards: published recommendations [11], ISO-1964 criteria, and ISO-1997 criteria, as well as the mean hearing thresholds at 0.25-8.0 kHz (full frequency band), 1.0-4.0 kHz and 4.0-8.0 kHz (Table 8). The difference between hearing loss judged according to the ISO-1964 criteria and the one judged according to the mean hearing threshold at 4.0-8.0 kHz, was statistically significant (p <0.001); however, the differences between hearing loss judged according to the mean hearing threshold at 4.0-8.0 kHz and the one according to the other criteria were not statistically significant (each p value >0.05).