2017 Sec 1 Green Book

Statistical Analyses Data distributions were reported as medians with ranges or interquartile ranges for continuous variables and frequencies with proportions for categorical variables. Comparisons were made among subjects who had ipsilateral hearing loss, contra- lateral hearing loss, and bilateral hearing loss relative to the ear with the identified EVA. Due to the non-normal distribution of the majority of the data, nonparametric statistics were used in the analyses. Continuous data were compared using the Kruskal–Wallis test or the Wilcoxon rank sums test, as appro- priate. Categorical data were compared among groups using v 2 analysis or the Fisher exact test, as appropriate. Correlations between vestibular measurements and continuous variables such as PTA values were conducted using the Spearman rank correlation coefficient. For assessing which factors may be inde- pendently related (independent predictors) to having a progressive hearing loss in an ear, multivariate analysis was conducted using generalized estimating equations (GEE). GEE allows for the clustering of ears in the analysis (two ears belonging to the same individual). GEE was used for both mod- eling of odds or risk progression as a dichotomous variable and modeling the rate of progression (as a decibel loss per year). For all analyses, a P value of .05 or less was considered statistically significant. All analyses were performed using SAS for Windows, version 9.2 (SAS Institute, Cary, NC). RESULTS Our database search identified 144 patients (67 males, 77 females) who met our inclusion criteria. Uni- lateral EVA was identified in 74 (51.4%) of these patients (Fig. 1); 42 (56.8%) had left-sided EVA, and 32 had right-sided EVA. The median age at which the hear- ing loss was identified was 59.5 months (range, 0–324.5 months). The median follow-up time was 37.8 months (range, 0–812.5 months), and 116 (80.5%) patients were followed for at least 3 months. As shown in Table I, patients with unilateral EVA were identified at a later age than patients with bilateral EVA (70 vs. 54.5 months; P ¼ .01). They also had a shorter period of audiometric follow-up (32.8 vs. 56.3 months; P ¼ .02) and were more likely to have unilateral hearing loss (46% vs. 15.7%; P < .0001). Unilateral Versus Bilateral Hearing Loss Forty-five patients (31.3%) with EVA had unilateral hearing loss. In 23 (51%) of these patients, the hearing loss was left-sided. Among the 74 patients with unilat- eral EVA, 34 (46%) had unilateral hearing loss; in 1 patient, the unilateral loss was in the contralateral ear. The remaining 39 patients had bilateral hearing loss. More than 55% of patients with unilateral EVA pre- sented with hearing loss in the contralateral ear (Fig. 1). Among the 70 patients with bilateral EVA, only 11 (15.7%) had unilateral hearing loss. The median age at which hearing loss was identified in all EVA patients (n ¼ 144) was significantly higher among those with uni- lateral hearing loss as compared to those with bilateral loss (78.5 vs. 45.5 months; range, 16.5–158 vs. 0–324.5 months; P < .0001). This finding was consistent when patients were stratified into a unilateral EVA group (80.5 vs. 42.8 months; range, 16.5–158 vs. 0–324.5

Sequencing products were analyzed using the ABI Prism 3730 Capillary Sequence Detection System, and raw data were obtained and compared to the published consensus sequences using Sequencher 4.8 (Gene Codes Corporation, Ann Arbor, MI) sequencing analysis software. The 342-kb deletion in the genetic region of D13S1830 ( GJB6 -D13S1830) locus on chromo- some 13q12 was analyzed by PCR and gel electrophoresis. Detailed methodologies have previously been published. 19,25 Mutation nomenclature is based on the recommendations of the American College of Medical Genetics and the Human Genome Variation Society. Several in silico analysis methods for the prediction of functional consequences of sequence variants were utilized to provide initial classification. All sequence variants were verified by the Human Gene Mutation Database, the National Center for Biotechnology Information Single Nucleotide Polymorphism database, and locus-specific mutation databases, such as Connexin Deafness Homepage (http://davinci.crg.es/deafness/ index.php?seccion ¼ mut_db&db ¼ nonsynd&nonsynd ¼ cx26mut). For novel mutations, we used SIFT (http://blocks. fhcrc.org/sift/SIFT.html), which utilizes evolutionary infor- mation from homologous proteins, 16 and PolyPhen (http:// www.bork.embl-heidelberg.de/PolyPhen/), which incorporates structural information into classification rules. 17 The Grantham Scale 18 was also used to evaluate the significance of amino acid substitutions, and a determination of the likelihood of the sequence change being pathogenic was made by the Molecular Genetics laboratory. Fig. 2. Schema for acquiring vestibular aqueduct measurements. (A) Operculum measurement. (B) Midpoint measurements: half of the dis- tance between the opercular and crus plane (A to B). OP ¼ opercu- lum; PSSC ¼ posterior semicircular canal. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Laryngoscope 123: June 2013

Greinwald et al.: Unilateral Enlarged Vestibular Aqueduct

135