We studied 23 patients (from 11 families) who had Waardenburg's syndrome. Patients were evaluated by conventional audiometric methods and by distortion-product otoacoustic emissions to determine the penetrance and the degree and type of hearing loss. Twelve of the patients had the type I form of the syndrome and 11 had type II. Overall, we found hearing loss in 19 of the 23 patients (83%); hearing loss affected nine type I patients (75%) and 10 type II patients (91%). Five type I patients (42%) and eight type II patients (73%) had a hearing loss of [less than]100 dB. Bilateral symmetrical hearing loss was the most common type of loss, as it was seen in six of the type I patients (50%) and eight of the type II patients (73%). At lower frequencies, distortion-product otoacoustic emission amplitudes were found to be significantly above the noise floor in five of the 11 patients whose hearing thresholds were 60 dB HL or worse by click auditory brainstem response testing. These findings led us to conclude that it is necessary to use otoacoustic emissions in patients with Waardenburg's syndrome in order to provide optimum fitting of hearing aids, especially in children.
Waardenburg's s syndrome was first described in 1951 as a new entity characterized by congenital sensorineural hearing loss and pigmentary disturbances of the skin, hair, and eyes (figure).  Waardenburg's syndrome is classified as one of two types, according to the presence (type I) or absence (type II) of dystopia canthorum.  The penetrance of congenital hearing loss, the most significant clinical finding in patients with this syndrome, has been reported to be 35 to 70% in type I patients and 55 to 85% in type II patients. [3-5] When it occurs, the hearing loss can be either complete or partial and bilateral or unilateral. The incidence of unilateral loss has been reported to be 4 to 1.3%. [3-5]
As is the case with other syndromic hearing losses, a significant number of patients with Waardenburg's s syndrome have both low- and high-frequency hearing losses and impairments, as indicated by U-shaped audiograms. [5,6] The click auditory brainstem response (ABR) is highly sensitive in detecting childhood hearing loss, but its primary drawback is its lack of frequency specificity.
Distortion-product otoacoustic emissions (DPOAEs), first described by Kemp in 1979, are the acoustic energies in the external ear canal that emanate as a result of the simultaneous stimulation of the cochlea by two continuous pure tones, which are referred to as primaries.  The frequency of this acoustic energy is closely related to the frequency of two primary stimuli: f1 and f2. As nonlinear responses that are present in almost all normal ears and that have high-frequency specificity, DPOAEs are useful in the diagnosis of hearing loss, especially in children. [8,9]
Our search of the literature found only one previous study that investigated the use of DPOAEs in Waardenburg's syndrome.  In this article, we describe our study of Waardenburg's syndrome in 23 patients from 11 families. We discuss its penetrance and the degree and type of hearing loss as determined by conventional methods and by DPOAEs in patients with both type I and type II syndrome.
Materials and methods
Twenty-three patients (from 11 families) with Waardenburg's s syndrome were followed at the University of Istanbul's Center for Deaf Children between 1990 and 1997. The 13 females and 10 males ranged in age from 2 to 40 years (mean: 17).
Twelve of these patients met the diagnostic criteria for type I Waardenburg's syndrome as proposed by Farrer et al,  and 11 were diagnosed with type II syndrome, as described by Liu et al.  We also evaluated pigmentary disturbances according to the criteria described by the latter author.
All patients were examined otoscopically, and all were evaluated by tympanometry, DPOAEs, and either pure-tone audiometry or ABR. In 12 of the patients, air-conduction thresholds were determined at 500 Hz and 1,2,4,6, and 8 kHz by pure-tone audiometry. The degree of hearing loss was determined by examining the mean values across the frequency range. The 11 patients who could not be subjected to pure-tone audiometry because of their age had their hearing thresholds determined by click ABR. Hearing threshold results were placed into one of four categories ([less than or equal to]30,31 to 60,61 to 100, and [greater than]100dB HL). In cases of bilateral asymmetrical hearing loss, the classification was based on the better ear.
A Celesta 503 Otoacoustic Emission Analyzer (Madsen Electronics; Taastrup, Denmark) was used to test DPOAEs. DPOAEs were determined at frequencies of 500 and 750 Hz and at 1,1.5,2,3,4,6, and 8 kHz to obtain a "DP-gram." The ratio of the two primaries (f2/f1) was 1.22, and their intensity was set at 65 dB sound pressure level. Distortion products were accepted by the system as valid because the response fell within the range of [+ or -]3 standard deviations of noise and [less than]80 dB of stimulus level. 
All patients were otoscopically normal, and all had type A tympanograms. With the use of conventional methods and DPOAE, we found that 19 of the 23 patients (83%) had a sensorineural hearing loss--nine (75%) type I patients and 10 (91%) type II patients. No progression of hearing loss was observed in any patient during a followup that ranged from 8 months to 7 years. The incidence of hearing loss and pigmentary disturbances are recorded in table 1, along with those reported by Liu et al for comparison purposes. 
In both type I and II patients, the most common degree of hearing loss category was [greater than]100 dB HL. Profound hearing loss was detected in five type I patients (42%) and eight type II patients (73%); according to [[chi].sup.2] analysis, the difference between the two groups was statistically significant (p[less than]0.01). There were no statistically significant differences with respect to the penetrance of hearing loss in the other ranges of hearing thresholds (table 2).
The most common type of hearing loss was bilateral hearing loss, which was seen in 18 of the 23 patients (78%) (table 3). Bilateral symmetrical hearing loss was more common than unilateral and bilateral asymmetrical hearing loss in both types of the syndrome. Only one patient (type II) had unilateral hearing loss. Moreover, bilateral symmetrical hearing loss was more common in type II patients than in type I patients.
DPOAEs were normal in the four patients (three type I and one type II) who had normal hearing as determined by conventional audiometry. In these patients, emission amplitudes were above the noise floor at all frequencies between 500 Hz and 8 kHz, and these responses were accepted by the system as valid. No patient had a notch in the DP-gram.
In 11 of the 16 patients whose hearing thresholds were 60 dB HL or worse, thresholds were determined by click ABR. In five of these 11 patients, results in seven ears were incompatible with DP-gram findings. In five of these seven ears, DPOAE amplitudes were significantly above the noise floor at 500 Hz, 750 Hz, and 1 kHz and below the noise floor at the higher frequencies. In the other two ears, emission amplitudes were above the noise floor only at 500 and 750 Hz.
Audiologic tests of the 23 patients showed that the penetrance of sensorineural hearing loss was 83%. Penetrance was 75% in the type I group and 91% in the type II group. In a study by Hageman and Delleman, the penetrance of hearing loss was 36 and 57% in type I and type II patients, respectively;  these rates are remarkably lower than those in our study. However, in two more recent studies, Newton  and Liu et al  reported penetrance rates of 69 and 58%, respectively, in type I patients and 87 and 78% in type II patients; these findings are more in line with our own. One common finding among all four studies was that the penetrance of hearing loss in type II patients was remarkably higher than in type I patients.
In our study, the most common category of degree of hearing loss in both types of patients was [greater than]100 dB HL--42% in type I patients and 73% in type II patients; the difference between the two groups was statistically significant (p[less than]0.01). Newton reported hearing losses of [greater than]100 dB HL in 34% of type I patients and 36% of type II patients.  Liu et al observed this degree of hearing loss in 58% of type I patients and 47% of type II patients (table 2).  It is interesting, however, that while the penetrance and degree of hearing loss is greater in type II patients, the expression of pigmentary disturbances on the skin and abnormalities of the facial skeleton are more common in type I patients (table 1). Type II patients are more likely to be identified by their families and primary care physicians by severe hearing loss than by skeletal abnormalities or pigmentary disturbances. Many type II patients who do not have hearing loss, especially those in developing countries, are considered by their families to be normal, and therefore they do not seek medical attention. On the other hand, because of obvious phenotypic characteristics such as dystopia canthorum, type I patients are easily recognized when no hearing loss is present.
In our study, bilateral symmetrical sensorineural hearing loss was the most common form, present in 50% of type I cases and 73% of type II cases (table 3). In a study published in 1977, Hageman reported bilateral symmetrical hearing loss in about 25% of type I patients and in almost 50% of type II patients.  These percentages are lower than those reported by Newton  and those of our study. However, our findings are in agreement with those of Hageman with respect to the fact that the penetrance of bilateral symmetrical hearing loss was greater in type II patients than in type I patients. By contrast, Newton  and Liu et al  both reported an equal distribution of bilateral symmetrical hearing loss in the two types of patients. As mentioned earlier, the higher penetrance in our study can probably be attributed to the fact that the diagnosis was made by identification of hearing loss rather than by phenotype. Newton's data on 79 patients, who belonged to 10 families with Waardenburg's syndrome, were not affected by the reasons that these patients had visited their physician.
DPOAEs are ideal for predicting hearing loss and puretone audiogram shape because they are (1) present in all ears with normal hearing, (2) absent in ears with a hearing threshold [less than]50 dB HL, and (3) frequency-specific. [8,9,15]
In a small study published jointly, Liu and Newton reported the detection of a notch between 1,000 and 3,000 Hz in the DP-grams of seven of eight (88%) normal-hearing patients with Waardenburg's syndrome.  The notch was present in all five of their type II patients and in two of their three type I patients. They speculated that the notch was the result of either a subclinical pathologic change or some other physical phenomenon. We found no such notch in our study, and there are no other reports in the literature of DPOAE findings in patients with Waardenburg's syndrome. If we combine the four patients in our study who had normal hearing with Liu and Newton's eight patients, we can extrapolate that somewhat more than half (7 of 12 in these two studies) of patients with Waardenburg's syndrome who have normal hearing can be expected to have subclinical pathologic changes.
In Waardenburg's syndrome, as in other hearing losses of genetic origin, patients can experience low- and high-frequency hearing losses and exhibit U-shaped audiographic results. [4,6] Newton found either unilateral or bilateral residual low-frequency (Fisch type I) losses in audiographic shapes in 29 of 60 patients with Waardenburg's syndrome. [4,16] Liu reported that the penetrance of low-frequency hearing loss in Waardenburg's syndrome was 14%.  It is not usually possible to determine these types of hearing losses by click ABR, which is not frequency-specific. In five patients (seven ears) in our study whose hearing thresholds were 60 dB HL or worse by click ABR, DP-grams showed that emission amplitudes were significantly above the noise floor at 500 Hz, 750 Hz, and 1 kHz in five ears. When one considers that emissions are absent in ears that have a hearing threshold of 50 dB HL or worse, one can speculate that hearing might be better than expected, perhaps even normal, at these low frequencies. 
We conclude that patients whose hearing thresholds are determined by click ABR testing should undergo DPOAE testing to estimate low-frequency hearing. The routine use of DPOAEs in children with Waardenburg's syndrome can detect high-frequency hearing loss, and thus it can prevent overamplification in the lower frequencies during the fitting of a hearing aid, which is the primary therapy for patients with this syndrome.
From the Department of Otolaryngology--Head and Neck Surgery, Taksim State Hospital, Istanbul (Dr. Oysu), and the Department of Otolaryngology--Head and Neck Surgery, Istanbul School of Medicine, University of Istanbul (Dr. Baserer and Dr. Tinaz).
(1.) Waardenburg PJ. A new syndrome combining developmental anomalies of eyelids, eyebrows and nose root with pigmentary defects of iris and head hair and congenital deafness. Am J Hum Genet 1951;3:195-253.
(2.) Arias S. Genetic heterogeneity in Waardenburg syndrome. Birth Defects Orig Artic Ser 1971;7:87-l0l.
(3.) Hageman MJ. Audiometric findings in 34 patients with Waardenburg's syndrome. J Laryngol Otol 1977;91:575-84.
(4.) Newton V. Hearing loss and Waardenburg's syndrome: Implications for genetic counselling. J Laryngol Otol 1990;104:97-103.
(5.) Liu XZ, Newton VE, Read AP. Waardenburg syndrome type II: Phenotypic findings and diagnostic criteria. Am J Med Genet 1995;55:95-100.
(6.) Gorlin RJ, Toriello HV, Cohen MM Jr. Hereditary Hearing Loss and Its Syndromes. Oxford Monographs on Medical Genetics, No. 28. New York: Oxford University Press, 1995:370.
(7.) Kemp DT. Evidence of mechanical nonlinearity and frequency selective wave amplification in the cochlea. Arch Otorhinolaryngol 1979;224:37-45.
(8.) Brown AM, Kemp DT. Suppressibility of the 2f1-f2 stimulated acoustic emissions in gerbil and man. Hear Res 1984;13:29-37.
(9.) Lonsbury-Martin BL, Martin GK. The clinical utility of distortion-product otoacoustic emissions. Ear Hear 1990;11:144-54.
(10.) Liu XZ, Newton VE. Distortion product emissions in normal-hearing and low-frequency hearing loss carriers of genes for Waardenburg's syndrome. Ann Otol Rhinol Laryngol 1997;106:220-5.
(11.) Farrer LA, Grundfast KM, Amos J, et al. Waardenburg syndrome (WS) type I is caused by defects at multiple loci, one of which is near ALPP on chromosome 2: First report of the WS consortium. Am J Hum Genet 1992;50:902-13.
(12.) Liu X, Newton V, Read A. Hearing loss and pigmentary disturbances in Waardenburg syndrome with reference to WS type II. J Laryngol Otol 1995;109:96-l00.
(13.) User manual for the Celesta 503 Otoacoustic Emission Analyzer, version 3.00. Taastrup, Denmark: Madsen Electronics, p. 36.
(14.) Hageman MJ, Delleman JW. Heterogeneity in Waardenburg syndrome. Am J Hum Genet 1977;29:468-85.
(15.) Harris FP. Distortion-product otoacoustic emissions in humans with high frequency sensorineural hearing loss. J Speech Hear Res 1990;33:594-600.
(16.) Fisch L. Deafness as part of an hereditary syndrome. J Laryng 1959;73:355-82.
(17.) Liu XZ. Clinical and Molecular Studies of Waardenburg Syndrome Type II [thesis]. Manchester, England: University of Manchester Press, 1995.
COPYRIGHT 2000 Medquest Communications, Inc.
COPYRIGHT 2000 Gale Group