- Email: [email protected]
Audiological findings of shunt-treated hydrocephalus in children
International Elsevier
PEDOT
Journal of Pediatric
Otorhinolaryngology,
21
18 (1989) 21-30
00601
Audiological findings of shunt- treated hydrocephalus in children Heikki Liippbnen
‘, Martti
Sorri ‘, Willy Serlo 2 and Lennart
von Wendt
3
Depariments of ’ Otola~ngology and 2 Paediatrics, University of Oulu, Oulu (Finland) and ’ Nordic Council for Arctic Medical Research, Oulu (Finland) (Received 9 December 1988) (Revised version vived 19 April 1989) (Accepted 15 June 1989)
Key words: Hydrocephalus; loss
Shunt
surgery;
Audiological
finding;
Sensorineural
high-frequency
hearing
Abstract 47 hydrocephalic children (mean age 10.4 years) were examined on average 7.9 years after initial shunting. The etiology of hydrocephalus was classified into 5 groups as follows: perinatal intraventricular hemorrhage 19, congenital obstructive hydrocephalus 15, intracranial cysts 5, severe intracranial anomalies 4 and central nervous system infections 4 children. Audiological examination included pure-tone audiometry, tympanometry, registration of stapedius reflex thresholds and adaptation. A sensorineural high-frequency hearing loss was found in 18 (38%) of 47 examined shunt-treated hydrocephalic children, and 11 of the losses could be classified as the retrocochlear type. The differences of the mean hearing thresholds between the etiological groups of childhood hydrocephalus were minimal.
Introduction
Disturbed cerebrospinal fluid (CSF) circulation leads to an accumulation of CSF, a condition which is generally called hydrocephalus. After the introduction of a valve-regulated shunt, to ensure unidirectional CSF flow [16] shunting has emerged as the main therapeutic measure in the management of hydrocephalus during infancy and childhood [5]. Although shunting has immensely improved the overall
Correspondence: H. Ltjppanen, SF-90220 Oulu, Finland.
0165-5876/89/$03.50
Department
of Otolaryngology,
0 1989 Elsevier Science Publishers
B.V.
University
of Oulu,
Kajaanintie
52,
22
prognosis for hydrocephalus, it seems that overdrainage may be one major factor contributing to the still not infrequent sub-optimal results [7,19]. Recently, it has been shown that overdrainage of cerebrospinal fluid may cause cranial base hypoplasia [9]. Most oto-neurological studies concerning childhood hydrocephalus have been performed rather soon after the shunting procedure, and there are only a few reports on long-term results. Case reports of adult aqueductal stenosis with hearing loss and vertigo have shown greatly improved oto-neurological test results soon after shunt operations [3,22]. Abnormal auditory brainstem responses (ABR) have been found in almost all hydrocephalus patients, both in children and adults [6,14]. Abnormalities of the ABR were reversed in hydrocephalic children soon after medical or surgical therapy [8,15]. High-frequency hearing losses have been reported to exist in 5% of the adult patients who had hydrocephalus diagnosed in their childhood [lo]. On the other hand, 24% of preterm infants who required shunts for posthemorrhagic hydrocephalus showed hearing impairment at an age varying from 6 to 8 months [4]. In the present study, we have examined hydrocephalic children on average 7.9 years after initial shunting. The aim of this study was to evaluate the audiological findings of these children, and to study the possible origin of hearing losses and, also, elucidate possible differences between etiological groups of childhood hydrocephalus. The vestibular findings of these children will be reported in a subsequent paper.
Subjects and Methods
Subjects Our material consists of children who have been shunt-treated due to hydrocephalus. 47 patients, 28 boys and 19 girls, were chosen from a total series of 148 children diagnosed, treated and followed-up at the Department of Pediatrics in the University Central Hospital of Oulu during the years 1966-1983. These children were invited to follow-up examinations in the years 1984-1985, and the study was aimed at the children who at that time were 4 years old or older (mean age 10.4 years; range 4.6-17.9 years). All the children had been treated by shunt operations (mean/case 3.6 operations, range 1-14 operations) and the examinations were carried out on average 7.9 years (range 1.3-17.1 years) after initial shunting. The etiology of hydrocephalus was classified into 5 groups according to Amacher and Wellington [l]: perinatal intraventricular hemorrhage (IVH) 19, congenital obstructive hydrocephalus 15, intracranial cysts 5, severe intracranial anomalies 4 and central nervous system (CNS) infections 4 children. Children suffering from hydrocephalus caused by spina bifida or tumor were excluded, because factors other than hydrocephalus usually affect their prognosis. The subjects generally cooperated in the examinations, and 85% of them managed well in normal school grades. The intelligence tests (the Wechsler Intelligence Scale for Children and the Wechsler
23
Intelligence Scale for Adults [24,25]) revealed normal intelligence (Intelligence quotient (IQ) exceeding 85) in 47% of patients. The mean IQ was 82 + 28 [19]. Methods
The pure-tone hearing thresholds were examined in a clinical sound isolated audiometry room using a Madsen Electronics OB-70 or OB-77 clinical audiometer calibrated according to the ISO(1975) standard. The air conduction thresholds were measured with TDH-39P earphones with MX-41/AR cushions. Measurements were made employing an ascending-descending technique at frequencies of 0.125, 0.25, 0.5, 1, 2, 3, 4, 6 and 8 kHz using 5 dB steps. The bone conduction thresholds were measured with an Oticon BCX 10 381 oscillator pressed against the mastoid process. Measurements were made at frequencies of 0.25, 0.5, 1, 2 and 4 kHz if the air conduction threshold was 10 dB or worse. The contralateral ear was masked using an insert earphone Phonic Ear PH 78 for the thresholds 1171. There is no international agreement on the audiological limits of normal hearing. In Finland, however, a hearing level of 15 dB has commonly been used as the boundary of hearing impairment [11,18,20]. Therefore, a pure-tone average (PTA; mean of the frequencies 0.5, 1 and 2 kHz) worse than 15 dB was considered in this study to be pathological. A hearing threshold of 20 dB or worse at any of the frequencies 4, 6 and 8 kHz was regarded as high-frequency hearing loss. The impedance measurements included registration of tympanogram, middle ear pressure, contra- and ipsilateral stapedius reflexes and stapedius reflex adaptation. A Madsen model 2073A impedance bridge was used employing a probe tone of 226 Hz. During the tympanometry the ear canal pressure was changed from +200 mmH,O to -200 mmH,O in 25 s. The types of tympanograms were classified into 6 different groups [21]. The middle ear pressure was considered to be pathological when it was lower than - 50 mmH,O. The contralateral stapedius reflex thresholds were measured at stimulus frequencies of 0.5, 1, 2 and 4 kHz and also with white noise. Ipsilaterally the acoustic stimulus was given at frequencies of 0.5, 1 and 2 kHz. A stapedius reflex threshold of 95 dB HL or less was considered normal. Thresholds of 100 dB or more were judged pathological [12,13,21]. The cases with pathological stapedius reflex thresholds were classified as the retrocochlear type, if the difference between reflex threshold and air conduction threshold was more than 60 dB, and middle ear pathology was excluded. The adaptation of the stapedius reflex was measured at frequencies of 0.5 and 1 kHz at 10 dB suprathreshold levels. The decay of the reflex was considered abnormal if it was 50% or more in 10 s [2]. Computations and statistical analyses were performed using the computer service of the University of Oulu. The statistical significance of differences between two means was tested using the Student’s r-test. Results Pure-tone audiometty
The mean air conduction thresholds of the whole group are shown in Fig. 1. The left ears showed worse thresholds, however, there was no statistical difference
24
T
dS (ISO)
I
I
20
o
q
right
x = left
1
i
ear ear
30 0.;25
0.25
0.5
1
2
41
6,
6I
kfiz
Fig. 1. Air conduction thresholds (means and S.D.) of 47 shunt-treated hydrocephalic children.
between the ears. The boys had worse thresholds than the girls but the difference was not statistically significant either (Figs. 2 and 3). Four (8.5%) patients had a mild hearing loss at the middle frequencies according to our criteria. A lCyear-old boy had a sensorineural hearing loss (PTA 17 dB) in both ears. Another 3 patients had mild conductive losses at the middle frequencies. A 6-year-old boy had a conductive loss in both ears (PTA 17 dB), and another 6 year-old-boy had a conductive loss in his right ear (PTA 17 dB). A 7-year-old boy
dS
(ISO)
A = girls o : boys
tn.191 (~26)
30 0.126
0.Z
0.5
1
2
$4
I
I
6
6
Fig. 2. Air conduction thresholds (means and S.D.) of right ears in shunt-treated (n = 19) and boys (n = 28).
kflz
hydrocephalic
girls
25 dB WiO)
T
0
T
girls
A :
q = boys
T
i
I
I 1
I
h:19) (n=28)
30 0.125
0.25
Fig. 3. Air conduction
0.5
thresholds
1
2
3
4
6
(means and SD.) of left ears in shunt-treated (n = 19) and boys (n = 28).
8kHz
hydrocephalic
girls
had a conductive loss of 18 dB in his left ear. All these 3 patients with a conductive hearing loss also had a sloping tympanogram indicating otitis media with effusion. Sensorineural high-frequency hearing losses according to our criteria were found in 18 (38%) of the patients. Nine patients had a bilateral hearing loss, two in the right and 7 in the left ear. Of the patients 14 had a hearing threshold between 20 and 30 dB. Four patients had a threshold of 35 dB or worse at the high frequencies,
dB (ISO) 0
15 p: 0.;25
0.;5
0.5
0.01
0.006
1
1
;
4
;
;
kHz
Fig. 4. Mean air conduction thresholds of right ears in 5 different etiology groups of hydrocephalus. values express the difference between the two major groups (hydrocephalus caused by perinatal hemorrhage and congenital obstructive hydrocephalus).
P i.v.
26 dS
(ISO) 0
,’
0
PERlNATAL
(10)
. CONGENlTAL (15) 0
15
CYST
6)
.
ANOMALY
A
INFECTION
(4) (4) \ , \
’ ‘\
Fig. 5. Mean air conduction thresholds of left ears in 5 different etiological groups of hydrocephalus. P values express the difference between the two major groups (hydrocephalus caused by perinatal i.v. hemorrhage and congenital obstructive hydrocephalus).
of these 4 patients, 1 boy and 1 girl belonged to the IVH group, 1 boy had congenital obstructive hydrocephalus and 1 boy had an intracranial cyst. The mean air conduction thresholds of the different etiological groups are seen in Figs. 4 and 5. The statistical analysis between the two major etiological groups, hydrocephalus caused by perinatal IVH and congenital obstructive hydrocephalus, showed the IVH group to have worse thresholds. The difference was significant (P < 0.01) at the frequency of 2 kHz in the right ear, and almost significant (P < 0.05) at the frequency of 1 kHz in the right and at the frequencies of 1,2 and 3 kHz in the left ear. The differences between the other etiological groups are not presented because of the small number of subjects belonging to each group. Impedance measurements The types of the tympanograms are shown in Table I. No flat or W-shaped tympanograms existed in this material. Due to poor cooperation or technical problems, no tympanogram could be registered in 5 of the right and in 6 of the left ears. A pathological middle ear pressure was recorded in 6 of the right and in 5 of the left ears (Table II). There was no registration in 7 patients. There was no significant difference between the stapedius reflex thresholds with a contra- or ipsilateral stimulus of 0.5, 1, 2, and 4 kHz. The findings of the contralateral stapedius reflex thresholds with a stimulus of 1 kHz are shown in Table III. All pathological findings were classified as the retrocochlear type, and this kind of pathology was found in 15 patients.
27 TABLE I Tympanogram
rypes. The results are given as the number of patients
Normal V-shaped Deep V-shaped Low V-shaped Sloping Unsuccessful
TABLE
belonging to each categoty (n = 47).
Right ear
Left ear
28 8 3 3 5
24 10 4 3 6
II
Middle ear pressure.
> - 50 mmH,O I -50 mmH,O Not measured
The results are given as the number of patients
belonging to each category (n = 47).
Right ear
Lef ear
34 6 1
35 5 I
TABLE III Contralateral stapedius rejlex (registered ear) with a stimulus ofpatients belonging to each category (n = 47).
Normal Pathological Not measured
of I kHz.
The results are given as the number
Right ear
Lejt ear
29 11 7
28 11 8
A pathological stapedius reflex adaptation was found in 4 of 36 patients studied. One patient had pathological adaptation at a frequency of 0.5 kHz and 3 at a frequency of 1 kHz.
Discussion
The results of this study showed that sensorineural high-frequency hearing losses are found in shunt-treated hydrocephalic children more frequently than in healthy agepeers. According to the impedance measurements, there were hints of retrocochlear pathology in most of the cases with sensorineural hearing losses. This study was performed at the Department of Pediatrics in the University Central Hospital of Oulu which is the only referral unit for children with neurological or neurosurgical disorders in northern Finland, and consequently this series is a comprehensive consecutive series. Although our series is relatively small the results may be considered valid for shunt-treated hydrocephalic children in general.
28
Our material consisted of shunt-operated hydrocephalic children from 5 different etiological groups, but 3 of these subgroups were too small for adequate statistical analysis. The common feature of all the patients was that they had been treated with at least one shunt operation. As our study also included vestibular measurements, the time needed for all the measurements was quite long and therefore ABR measurements had to be abandoned in this pediatric material, and thus, we cannot make direct comparisons to ABR reports in the literature [14]. Our result of only one patient with a mild sensorineural hearing loss at the middle frequencies differs clearly from the results reported by Boynton et al. [4] who found the prevalence of 24% hearing losses in an infant material of hydrocephalus caused by perinatal IVH. They used visual reinforced audiometry and their patients were between 6 and 8 months of age, which may partly explain the different results. However, this is not surprising because it has been clearly shown that IVH patients are especially prone to additional handicaps as a consequence of the vulnerability of the premature brain [7]. In our material the prevalence of sensorineural high-frequency losses was considerably higher than in an earlier report by Jansen et al. [lo], although the mean age in their material was higher than in ours and also most of their patients seemed not to cooperate as well as ours. Kraus et al. [14] found elevated thresholds (> 20 dB HL) in 45% of their shunted hydrocephalic children in ABR studies using acoustic rarefaction clicks. This is in good agreement with our result of 38% high-frequency hearing losses measured with pure-tone audiometry. Contrary to Kraus et al. [14], we found an almost significant difference (P < 0.05) in some air conduction thresholds at the middle frequencies between the two major etiological groups, congenital obstructive hydrocephalus and hydrocephalus caused by perinatal IVH. The IVH group had statistically worse thresholds, although the clinical difference was rather small. It can be concluded that IVH, which is a frequently occurring complication in preterm infants [23], also affects hearing. The pathophysiological mechanisms of ABR abnormalities are not known although, for example, neuronal conduction block and ischemic damage caused by increased intracranial pressure or interstitial edema have been suggested [8,14,15]. Our stapedius reflex measurements indicate pathology of the retrocochlear type in all pathological cases and are thus in good agreement with previous reports. However, all our patients had been treated by shunt operations and there were no signs of increased intracranial pressure clinically or by computerized tomography investigation. Whether these pathological changes had developed during infancy before the shunt operations or later remains unclear. Shunting lasting for a long time may cause overdrainage of the cerebrospinal fluid and probably causes cranial base hypoplasia, as has been shown in a previously published study based on the same patient material as in the present paper [9]. Theoretically cranial base hypoplasia could cause disturbances in the brainstem area and also lead to pathological findings in the audiological and electrophysiological measurements. As a conclusion of this study, we found 18 (38%) of 47 examined shunt-treated hydrocephalus patients to have a sensorineural high-frequency hearing loss, and in 11 cases the findings indicated a retrocochlear pathology. The differences of the
29
mean hearing thresholds between the etiological groups of childhood hydrocephalus were minimal. A new hypothesis for the pathophysiological mechanism, namely anatomical abnormalities of the skull base, causing brainstem disturbances is presented.
Acknowledgement
This study was financially supported by the Bracke Gstergard Research Foundation, Gothenburg, Sweden.
References 1 Amacher, A.L. and Wellington, J., Infantile hydrocephalus: Long-term results of surgical therapy, Child’s Brain, 11 (1984) 217-229. 2 Anderson, H., Barr, B. and Wedenberg, E.. The early detection of acoustic tumors by stapedius reflex test. In G. Wolstenholme and J. Knight (Eds.), Ciba Foundation Symposium on Sensorineural Hearing Loss, 1970, pp. 275-294. 3 Barlas, O., Gokay, H., Turantan, I. and Baserer, N., Adult aqueductal stenosis with fluctuating hearing loss and vertigo, J. Neurosurg., 59 (1983) 703-705. 4 Boynton, B.R., Boynton, C.A., Merritt, T.A., Vaucher, Y.E., James, H.E. and Bejar, R.F., Ventriculoperitoneal shunts in low birth weight infants with intracranical hemorrhage: neurodevelopmental outcome, Neurosurgery, 18 (1986) 141-145. 5 Choux, M., Introduction, Monogr. Neurol. Sci., 8 (1982) 1-6. 6 Edwards, C.G., Durieux-Smith, A. and Picton, T.W., Auditory brainstem response audiometry in neonatal hydrocephalus, J. Otolaryngol. (Suppl.), 14 (1985) 40-46. 7 Femell, E., Infantile hydrocephalus. Epidemiology and neuropediatric aspects of the outcome in Swedish children born 1967-82. Thesis, Gothenburg, Sweden, 1987, 51 pp. 8 Hall, J.W., Brown, D.P. and Mackey-Hargadine, J.R., Pediatric applications of serial auditory brainstem and middle-latency evoked response recordings. Int. J. Pediatr. Otorhinolaryngol., 9 (1985) 201-218. 9 Huggare, J.A., Kantomaa, T.J., Riinning, O.V. and Serlo, W.S., Craniofacial morphology in shunttreated hydrocephalic children, Cleft Palate J., 23 (1986) 261-269. 10 Jansen, J., Gloerfelt-Tarp, B., Pedersen, H. and Zilsdorff, K., Prognosis of infantile hydrocephalus. Follow-up in adult patients, born 194661955, Acta Neurol. Stand., 65 (1982) 81-93. 11 Jauhiainen, T., Kuulonhuolto. In T. Jauhiainen (Ed.), Kuulo, 1st edn., Kuulonhuoltoliitto ry, Mikkeli, 1969, pp. 91-100. 12 Jepsen, O., The tympanic muscles in normal material examined by means of the impedance, Acta Otolaiyngol., 39 (1955) 406-408. 13 Jerger, J., Mauldin, L. and Lewis, N., Temporal summation of the acoustic reflex, Audiology, 16 (1977) 177-200. 14 Kraus, N., &damar, b., Heydemann, P.T., Stein, L. and Reed, N.L., Auditory brainstem responses in hydrocephalic patients, Electroencephalogr. Clin. Neurophysiol., 59 (1984) 310-317. 15 McPherson, D.L., Arnlie, R. and Foltz, E., Auditory brainstem response in infant hydrocephalus, Child. Nerv. Syst., 1 (1985) 70-76. 16 Nulsen, F.E. and Spitz, E.B., Treatment of hydrocephalus by direct heart shunt from ventricle to jugular vein, Surg. Forum, 2 (1952) 399-403. 17 Palva, T. and Palva, A., Masking in audiometry. III. Reflections upon the present position, Acta Otolaryngol., 54 (1962) 521-531.
30 18 Salmivalli, A., Kuulon tutkiminen. In T. Palva (Pd.), Korva-, nenl- ja kurkkutaudit, 2nd edn., Vammalan kirjapaino, V-ala, 1983, pp. 22-29. 19 Serlo, W., Shunt treatment of hydrocephalus in children. Thesis, Acta Univ. Oul. D. 130. Med. Intema et Paediatr., (1985) 26. 20 Sonninen, A., Pyorti, T. and Klemetti, A., Screening for neonatal hearing disorders in the province of central Finland, Ann. Chir. Gynaecol., 64 (1975) 180-188. 21 Sorri, M., Impedance audiometry as a postoperative study. Thesis, Acta Univ. Oul. D. 45. Ophtalmol. Oto-rhino-laryngol., (1979) 4. 22 Suzuki, M., Harada, Y., Ishida, M., Wada, H., Ohta, M., Sakoda, K. and Uozumi, T., Aqueductal stenosis-results of vestibular function tests, J. Laryngol. Otol., 99 (1985) 151-161. 23 Volpe, J.J., Neonatal intraventricular hemorrhage, N. Engl. J. Med., 304 (1981) 886-891. 24 Wechsler, D., Wechsler Intelligence Scale for Children. Manual. The Psychological Corporation. New York, 1949. 25 Wechsler, D., Wechsler Adult Intelligence Scale. Manual, The Psychological Corporation, New York, 1955.
PEDOT
Journal of Pediatric
Otorhinolaryngology,
21
18 (1989) 21-30
00601
Audiological findings of shunt- treated hydrocephalus in children Heikki Liippbnen
‘, Martti
Sorri ‘, Willy Serlo 2 and Lennart
von Wendt
3
Depariments of ’ Otola~ngology and 2 Paediatrics, University of Oulu, Oulu (Finland) and ’ Nordic Council for Arctic Medical Research, Oulu (Finland) (Received 9 December 1988) (Revised version vived 19 April 1989) (Accepted 15 June 1989)
Key words: Hydrocephalus; loss
Shunt
surgery;
Audiological
finding;
Sensorineural
high-frequency
hearing
Abstract 47 hydrocephalic children (mean age 10.4 years) were examined on average 7.9 years after initial shunting. The etiology of hydrocephalus was classified into 5 groups as follows: perinatal intraventricular hemorrhage 19, congenital obstructive hydrocephalus 15, intracranial cysts 5, severe intracranial anomalies 4 and central nervous system infections 4 children. Audiological examination included pure-tone audiometry, tympanometry, registration of stapedius reflex thresholds and adaptation. A sensorineural high-frequency hearing loss was found in 18 (38%) of 47 examined shunt-treated hydrocephalic children, and 11 of the losses could be classified as the retrocochlear type. The differences of the mean hearing thresholds between the etiological groups of childhood hydrocephalus were minimal.
Introduction
Disturbed cerebrospinal fluid (CSF) circulation leads to an accumulation of CSF, a condition which is generally called hydrocephalus. After the introduction of a valve-regulated shunt, to ensure unidirectional CSF flow [16] shunting has emerged as the main therapeutic measure in the management of hydrocephalus during infancy and childhood [5]. Although shunting has immensely improved the overall
Correspondence: H. Ltjppanen, SF-90220 Oulu, Finland.
0165-5876/89/$03.50
Department
of Otolaryngology,
0 1989 Elsevier Science Publishers
B.V.
University
of Oulu,
Kajaanintie
52,
22
prognosis for hydrocephalus, it seems that overdrainage may be one major factor contributing to the still not infrequent sub-optimal results [7,19]. Recently, it has been shown that overdrainage of cerebrospinal fluid may cause cranial base hypoplasia [9]. Most oto-neurological studies concerning childhood hydrocephalus have been performed rather soon after the shunting procedure, and there are only a few reports on long-term results. Case reports of adult aqueductal stenosis with hearing loss and vertigo have shown greatly improved oto-neurological test results soon after shunt operations [3,22]. Abnormal auditory brainstem responses (ABR) have been found in almost all hydrocephalus patients, both in children and adults [6,14]. Abnormalities of the ABR were reversed in hydrocephalic children soon after medical or surgical therapy [8,15]. High-frequency hearing losses have been reported to exist in 5% of the adult patients who had hydrocephalus diagnosed in their childhood [lo]. On the other hand, 24% of preterm infants who required shunts for posthemorrhagic hydrocephalus showed hearing impairment at an age varying from 6 to 8 months [4]. In the present study, we have examined hydrocephalic children on average 7.9 years after initial shunting. The aim of this study was to evaluate the audiological findings of these children, and to study the possible origin of hearing losses and, also, elucidate possible differences between etiological groups of childhood hydrocephalus. The vestibular findings of these children will be reported in a subsequent paper.
Subjects and Methods
Subjects Our material consists of children who have been shunt-treated due to hydrocephalus. 47 patients, 28 boys and 19 girls, were chosen from a total series of 148 children diagnosed, treated and followed-up at the Department of Pediatrics in the University Central Hospital of Oulu during the years 1966-1983. These children were invited to follow-up examinations in the years 1984-1985, and the study was aimed at the children who at that time were 4 years old or older (mean age 10.4 years; range 4.6-17.9 years). All the children had been treated by shunt operations (mean/case 3.6 operations, range 1-14 operations) and the examinations were carried out on average 7.9 years (range 1.3-17.1 years) after initial shunting. The etiology of hydrocephalus was classified into 5 groups according to Amacher and Wellington [l]: perinatal intraventricular hemorrhage (IVH) 19, congenital obstructive hydrocephalus 15, intracranial cysts 5, severe intracranial anomalies 4 and central nervous system (CNS) infections 4 children. Children suffering from hydrocephalus caused by spina bifida or tumor were excluded, because factors other than hydrocephalus usually affect their prognosis. The subjects generally cooperated in the examinations, and 85% of them managed well in normal school grades. The intelligence tests (the Wechsler Intelligence Scale for Children and the Wechsler
23
Intelligence Scale for Adults [24,25]) revealed normal intelligence (Intelligence quotient (IQ) exceeding 85) in 47% of patients. The mean IQ was 82 + 28 [19]. Methods
The pure-tone hearing thresholds were examined in a clinical sound isolated audiometry room using a Madsen Electronics OB-70 or OB-77 clinical audiometer calibrated according to the ISO(1975) standard. The air conduction thresholds were measured with TDH-39P earphones with MX-41/AR cushions. Measurements were made employing an ascending-descending technique at frequencies of 0.125, 0.25, 0.5, 1, 2, 3, 4, 6 and 8 kHz using 5 dB steps. The bone conduction thresholds were measured with an Oticon BCX 10 381 oscillator pressed against the mastoid process. Measurements were made at frequencies of 0.25, 0.5, 1, 2 and 4 kHz if the air conduction threshold was 10 dB or worse. The contralateral ear was masked using an insert earphone Phonic Ear PH 78 for the thresholds 1171. There is no international agreement on the audiological limits of normal hearing. In Finland, however, a hearing level of 15 dB has commonly been used as the boundary of hearing impairment [11,18,20]. Therefore, a pure-tone average (PTA; mean of the frequencies 0.5, 1 and 2 kHz) worse than 15 dB was considered in this study to be pathological. A hearing threshold of 20 dB or worse at any of the frequencies 4, 6 and 8 kHz was regarded as high-frequency hearing loss. The impedance measurements included registration of tympanogram, middle ear pressure, contra- and ipsilateral stapedius reflexes and stapedius reflex adaptation. A Madsen model 2073A impedance bridge was used employing a probe tone of 226 Hz. During the tympanometry the ear canal pressure was changed from +200 mmH,O to -200 mmH,O in 25 s. The types of tympanograms were classified into 6 different groups [21]. The middle ear pressure was considered to be pathological when it was lower than - 50 mmH,O. The contralateral stapedius reflex thresholds were measured at stimulus frequencies of 0.5, 1, 2 and 4 kHz and also with white noise. Ipsilaterally the acoustic stimulus was given at frequencies of 0.5, 1 and 2 kHz. A stapedius reflex threshold of 95 dB HL or less was considered normal. Thresholds of 100 dB or more were judged pathological [12,13,21]. The cases with pathological stapedius reflex thresholds were classified as the retrocochlear type, if the difference between reflex threshold and air conduction threshold was more than 60 dB, and middle ear pathology was excluded. The adaptation of the stapedius reflex was measured at frequencies of 0.5 and 1 kHz at 10 dB suprathreshold levels. The decay of the reflex was considered abnormal if it was 50% or more in 10 s [2]. Computations and statistical analyses were performed using the computer service of the University of Oulu. The statistical significance of differences between two means was tested using the Student’s r-test. Results Pure-tone audiometty
The mean air conduction thresholds of the whole group are shown in Fig. 1. The left ears showed worse thresholds, however, there was no statistical difference
24
T
dS (ISO)
I
I
20
o
q
right
x = left
1
i
ear ear
30 0.;25
0.25
0.5
1
2
41
6,
6I
kfiz
Fig. 1. Air conduction thresholds (means and S.D.) of 47 shunt-treated hydrocephalic children.
between the ears. The boys had worse thresholds than the girls but the difference was not statistically significant either (Figs. 2 and 3). Four (8.5%) patients had a mild hearing loss at the middle frequencies according to our criteria. A lCyear-old boy had a sensorineural hearing loss (PTA 17 dB) in both ears. Another 3 patients had mild conductive losses at the middle frequencies. A 6-year-old boy had a conductive loss in both ears (PTA 17 dB), and another 6 year-old-boy had a conductive loss in his right ear (PTA 17 dB). A 7-year-old boy
dS
(ISO)
A = girls o : boys
tn.191 (~26)
30 0.126
0.Z
0.5
1
2
$4
I
I
6
6
Fig. 2. Air conduction thresholds (means and S.D.) of right ears in shunt-treated (n = 19) and boys (n = 28).
kflz
hydrocephalic
girls
25 dB WiO)
T
0
T
girls
A :
q = boys
T
i
I
I 1
I
h:19) (n=28)
30 0.125
0.25
Fig. 3. Air conduction
0.5
thresholds
1
2
3
4
6
(means and SD.) of left ears in shunt-treated (n = 19) and boys (n = 28).
8kHz
hydrocephalic
girls
had a conductive loss of 18 dB in his left ear. All these 3 patients with a conductive hearing loss also had a sloping tympanogram indicating otitis media with effusion. Sensorineural high-frequency hearing losses according to our criteria were found in 18 (38%) of the patients. Nine patients had a bilateral hearing loss, two in the right and 7 in the left ear. Of the patients 14 had a hearing threshold between 20 and 30 dB. Four patients had a threshold of 35 dB or worse at the high frequencies,
dB (ISO) 0
15 p: 0.;25
0.;5
0.5
0.01
0.006
1
1
;
4
;
;
kHz
Fig. 4. Mean air conduction thresholds of right ears in 5 different etiology groups of hydrocephalus. values express the difference between the two major groups (hydrocephalus caused by perinatal hemorrhage and congenital obstructive hydrocephalus).
P i.v.
26 dS
(ISO) 0
,’
0
PERlNATAL
(10)
. CONGENlTAL (15) 0
15
CYST
6)
.
ANOMALY
A
INFECTION
(4) (4) \ , \
’ ‘\
Fig. 5. Mean air conduction thresholds of left ears in 5 different etiological groups of hydrocephalus. P values express the difference between the two major groups (hydrocephalus caused by perinatal i.v. hemorrhage and congenital obstructive hydrocephalus).
of these 4 patients, 1 boy and 1 girl belonged to the IVH group, 1 boy had congenital obstructive hydrocephalus and 1 boy had an intracranial cyst. The mean air conduction thresholds of the different etiological groups are seen in Figs. 4 and 5. The statistical analysis between the two major etiological groups, hydrocephalus caused by perinatal IVH and congenital obstructive hydrocephalus, showed the IVH group to have worse thresholds. The difference was significant (P < 0.01) at the frequency of 2 kHz in the right ear, and almost significant (P < 0.05) at the frequency of 1 kHz in the right and at the frequencies of 1,2 and 3 kHz in the left ear. The differences between the other etiological groups are not presented because of the small number of subjects belonging to each group. Impedance measurements The types of the tympanograms are shown in Table I. No flat or W-shaped tympanograms existed in this material. Due to poor cooperation or technical problems, no tympanogram could be registered in 5 of the right and in 6 of the left ears. A pathological middle ear pressure was recorded in 6 of the right and in 5 of the left ears (Table II). There was no registration in 7 patients. There was no significant difference between the stapedius reflex thresholds with a contra- or ipsilateral stimulus of 0.5, 1, 2, and 4 kHz. The findings of the contralateral stapedius reflex thresholds with a stimulus of 1 kHz are shown in Table III. All pathological findings were classified as the retrocochlear type, and this kind of pathology was found in 15 patients.
27 TABLE I Tympanogram
rypes. The results are given as the number of patients
Normal V-shaped Deep V-shaped Low V-shaped Sloping Unsuccessful
TABLE
belonging to each categoty (n = 47).
Right ear
Left ear
28 8 3 3 5
24 10 4 3 6
II
Middle ear pressure.
> - 50 mmH,O I -50 mmH,O Not measured
The results are given as the number of patients
belonging to each category (n = 47).
Right ear
Lef ear
34 6 1
35 5 I
TABLE III Contralateral stapedius rejlex (registered ear) with a stimulus ofpatients belonging to each category (n = 47).
Normal Pathological Not measured
of I kHz.
The results are given as the number
Right ear
Lejt ear
29 11 7
28 11 8
A pathological stapedius reflex adaptation was found in 4 of 36 patients studied. One patient had pathological adaptation at a frequency of 0.5 kHz and 3 at a frequency of 1 kHz.
Discussion
The results of this study showed that sensorineural high-frequency hearing losses are found in shunt-treated hydrocephalic children more frequently than in healthy agepeers. According to the impedance measurements, there were hints of retrocochlear pathology in most of the cases with sensorineural hearing losses. This study was performed at the Department of Pediatrics in the University Central Hospital of Oulu which is the only referral unit for children with neurological or neurosurgical disorders in northern Finland, and consequently this series is a comprehensive consecutive series. Although our series is relatively small the results may be considered valid for shunt-treated hydrocephalic children in general.
28
Our material consisted of shunt-operated hydrocephalic children from 5 different etiological groups, but 3 of these subgroups were too small for adequate statistical analysis. The common feature of all the patients was that they had been treated with at least one shunt operation. As our study also included vestibular measurements, the time needed for all the measurements was quite long and therefore ABR measurements had to be abandoned in this pediatric material, and thus, we cannot make direct comparisons to ABR reports in the literature [14]. Our result of only one patient with a mild sensorineural hearing loss at the middle frequencies differs clearly from the results reported by Boynton et al. [4] who found the prevalence of 24% hearing losses in an infant material of hydrocephalus caused by perinatal IVH. They used visual reinforced audiometry and their patients were between 6 and 8 months of age, which may partly explain the different results. However, this is not surprising because it has been clearly shown that IVH patients are especially prone to additional handicaps as a consequence of the vulnerability of the premature brain [7]. In our material the prevalence of sensorineural high-frequency losses was considerably higher than in an earlier report by Jansen et al. [lo], although the mean age in their material was higher than in ours and also most of their patients seemed not to cooperate as well as ours. Kraus et al. [14] found elevated thresholds (> 20 dB HL) in 45% of their shunted hydrocephalic children in ABR studies using acoustic rarefaction clicks. This is in good agreement with our result of 38% high-frequency hearing losses measured with pure-tone audiometry. Contrary to Kraus et al. [14], we found an almost significant difference (P < 0.05) in some air conduction thresholds at the middle frequencies between the two major etiological groups, congenital obstructive hydrocephalus and hydrocephalus caused by perinatal IVH. The IVH group had statistically worse thresholds, although the clinical difference was rather small. It can be concluded that IVH, which is a frequently occurring complication in preterm infants [23], also affects hearing. The pathophysiological mechanisms of ABR abnormalities are not known although, for example, neuronal conduction block and ischemic damage caused by increased intracranial pressure or interstitial edema have been suggested [8,14,15]. Our stapedius reflex measurements indicate pathology of the retrocochlear type in all pathological cases and are thus in good agreement with previous reports. However, all our patients had been treated by shunt operations and there were no signs of increased intracranial pressure clinically or by computerized tomography investigation. Whether these pathological changes had developed during infancy before the shunt operations or later remains unclear. Shunting lasting for a long time may cause overdrainage of the cerebrospinal fluid and probably causes cranial base hypoplasia, as has been shown in a previously published study based on the same patient material as in the present paper [9]. Theoretically cranial base hypoplasia could cause disturbances in the brainstem area and also lead to pathological findings in the audiological and electrophysiological measurements. As a conclusion of this study, we found 18 (38%) of 47 examined shunt-treated hydrocephalus patients to have a sensorineural high-frequency hearing loss, and in 11 cases the findings indicated a retrocochlear pathology. The differences of the
29
mean hearing thresholds between the etiological groups of childhood hydrocephalus were minimal. A new hypothesis for the pathophysiological mechanism, namely anatomical abnormalities of the skull base, causing brainstem disturbances is presented.
Acknowledgement
This study was financially supported by the Bracke Gstergard Research Foundation, Gothenburg, Sweden.
References 1 Amacher, A.L. and Wellington, J., Infantile hydrocephalus: Long-term results of surgical therapy, Child’s Brain, 11 (1984) 217-229. 2 Anderson, H., Barr, B. and Wedenberg, E.. The early detection of acoustic tumors by stapedius reflex test. In G. Wolstenholme and J. Knight (Eds.), Ciba Foundation Symposium on Sensorineural Hearing Loss, 1970, pp. 275-294. 3 Barlas, O., Gokay, H., Turantan, I. and Baserer, N., Adult aqueductal stenosis with fluctuating hearing loss and vertigo, J. Neurosurg., 59 (1983) 703-705. 4 Boynton, B.R., Boynton, C.A., Merritt, T.A., Vaucher, Y.E., James, H.E. and Bejar, R.F., Ventriculoperitoneal shunts in low birth weight infants with intracranical hemorrhage: neurodevelopmental outcome, Neurosurgery, 18 (1986) 141-145. 5 Choux, M., Introduction, Monogr. Neurol. Sci., 8 (1982) 1-6. 6 Edwards, C.G., Durieux-Smith, A. and Picton, T.W., Auditory brainstem response audiometry in neonatal hydrocephalus, J. Otolaryngol. (Suppl.), 14 (1985) 40-46. 7 Femell, E., Infantile hydrocephalus. Epidemiology and neuropediatric aspects of the outcome in Swedish children born 1967-82. Thesis, Gothenburg, Sweden, 1987, 51 pp. 8 Hall, J.W., Brown, D.P. and Mackey-Hargadine, J.R., Pediatric applications of serial auditory brainstem and middle-latency evoked response recordings. Int. J. Pediatr. Otorhinolaryngol., 9 (1985) 201-218. 9 Huggare, J.A., Kantomaa, T.J., Riinning, O.V. and Serlo, W.S., Craniofacial morphology in shunttreated hydrocephalic children, Cleft Palate J., 23 (1986) 261-269. 10 Jansen, J., Gloerfelt-Tarp, B., Pedersen, H. and Zilsdorff, K., Prognosis of infantile hydrocephalus. Follow-up in adult patients, born 194661955, Acta Neurol. Stand., 65 (1982) 81-93. 11 Jauhiainen, T., Kuulonhuolto. In T. Jauhiainen (Ed.), Kuulo, 1st edn., Kuulonhuoltoliitto ry, Mikkeli, 1969, pp. 91-100. 12 Jepsen, O., The tympanic muscles in normal material examined by means of the impedance, Acta Otolaiyngol., 39 (1955) 406-408. 13 Jerger, J., Mauldin, L. and Lewis, N., Temporal summation of the acoustic reflex, Audiology, 16 (1977) 177-200. 14 Kraus, N., &damar, b., Heydemann, P.T., Stein, L. and Reed, N.L., Auditory brainstem responses in hydrocephalic patients, Electroencephalogr. Clin. Neurophysiol., 59 (1984) 310-317. 15 McPherson, D.L., Arnlie, R. and Foltz, E., Auditory brainstem response in infant hydrocephalus, Child. Nerv. Syst., 1 (1985) 70-76. 16 Nulsen, F.E. and Spitz, E.B., Treatment of hydrocephalus by direct heart shunt from ventricle to jugular vein, Surg. Forum, 2 (1952) 399-403. 17 Palva, T. and Palva, A., Masking in audiometry. III. Reflections upon the present position, Acta Otolaryngol., 54 (1962) 521-531.
30 18 Salmivalli, A., Kuulon tutkiminen. In T. Palva (Pd.), Korva-, nenl- ja kurkkutaudit, 2nd edn., Vammalan kirjapaino, V-ala, 1983, pp. 22-29. 19 Serlo, W., Shunt treatment of hydrocephalus in children. Thesis, Acta Univ. Oul. D. 130. Med. Intema et Paediatr., (1985) 26. 20 Sonninen, A., Pyorti, T. and Klemetti, A., Screening for neonatal hearing disorders in the province of central Finland, Ann. Chir. Gynaecol., 64 (1975) 180-188. 21 Sorri, M., Impedance audiometry as a postoperative study. Thesis, Acta Univ. Oul. D. 45. Ophtalmol. Oto-rhino-laryngol., (1979) 4. 22 Suzuki, M., Harada, Y., Ishida, M., Wada, H., Ohta, M., Sakoda, K. and Uozumi, T., Aqueductal stenosis-results of vestibular function tests, J. Laryngol. Otol., 99 (1985) 151-161. 23 Volpe, J.J., Neonatal intraventricular hemorrhage, N. Engl. J. Med., 304 (1981) 886-891. 24 Wechsler, D., Wechsler Intelligence Scale for Children. Manual. The Psychological Corporation. New York, 1949. 25 Wechsler, D., Wechsler Adult Intelligence Scale. Manual, The Psychological Corporation, New York, 1955.