- Email: [email protected]
Auditory brainstem responses in patients with neonatal hyperbilirubinemia and bilirubin encephalopathy
Auditory Brainstem Responses in Patients with Neonatal Hyperbilirubinemia and Bilirubin Encephalopathy Kun-Long Hung, MD
To understand the neurotoxicity of neonatal hyperbilirubinemia, auditory brain stem response (ABR) studies were performed in 75 jaundiced patients, who were divided into 4 groups in either a retrospective or prospective way. Retrospective ABR study in 10 known kemicteric patients (group I) showed elevation of the hearing threshold and delay of wave I, as well as prolongation of central brainstem transmission, in various degrees, in most cases. Six jaundiced infants (group II) with ABR testing before and after blood exchange transfusion (BET) showed shortening of wave latencies and increase in amplitude after BET. There were additional 20 infants with BET (group III) and 39 with phototherapy (group IV) receiving ABR testing after the therapeutic procedures. Prospective follow-up in groups II, III and IV showed normalization of the ABRs in all except one patient in the following months. These findings demonstrate the nature of bilirubin neurotoxicity and the prognostic value of ABRs in the monitoring of bilirubin toxicity. ABR testing is sensitive in reflecting the effect of hyperbilirubinemia, and provides a valuable guide for the early recognition and close follow-up of bilirubin neurotoxicity. Key words: Auditory brainstem response (ABR), hyperbilirubinemia, kernicterus, bilirubin neurotoxicity. Hung K-L. Auditory brainstem responses in patients with neonatal hyperbilirubinemia and bilirubin encephalopathy. Brain Dev 1989;11:297-301
Neonatal hyperbilirubinemia is one of the major problems in neonatal intensive care. Its sequelae include many neurological deficits [1-3]. The occurrence of kernicterus, which is one of these sequelae, has been reduced because of recent advances in mUltiple exchange transfusion, phototherapy and frequent serum bilirubin concentration measurement. Nevertheless, bilirubin encephalopathy still continues to occur because of inconsistency in clinical presentation of early bilirubin toxicity [4]. The application of the auditory brainstem response (ABR) in the study of neonatal hyperbilirubinemia seems reasonable because of its potential value in assessing hearing and brainstem functions [5-10]. The relationship between neonatal hyperbilirubinemia and ABR has been mentioned in a few studies [11-15]. Chisin et al [12] found that the occurrence of hearing loss following neonatal hyper-
From the Department of Pediatrics, Cathay General Hospital Taipei, Taiwan. Received for publication: March 4, 1989. Accepted for pUblication: July 22, 1989. Correspondence address: Dr. Kun-Long Hung, Department of Pediatrics, Cathay General Hospital, 280, Jen-Ai Rd, Sec 4, Taipei, Taiwan, Republic of China.
bilirubinemia showed a good correlation with the absence of wave I on ABR testing. Perlman et al [13] reported that jaundice was associated with significant transient aberrations of ABRs, suggestive of a transient brainstem encephalopathy. Studies on ABR changes in overt bilirubin encephalopathy have been rather scarce [16, 17]. To understand the neurotoxicity of hyperbilirubinemia, as reflected by the ABRs, a retrospective and prospective study was conducted. The present report describes the late changes on ABRs in kernicteric patients and acute changes of ABRs in neonates suffering from hyperbilirubinemia. SUBJECTS AND METHODS During the period from January 1984 to December 1986, ABR studies were performed in patients suffering from neonatal hyperbilirubinemia in our nursery and neonatal intensive care unit (NICU), as well as those with overt bilirubin encephalopathy from outpatient clinics. All the patients were full-term at birth with birth weights of greater than 2,500 gm. The causes of jaundice were mostly ABO incompatibility, glucose-6-phosphate dehydrogenase (G-6-P D) deficiency and unknown factors. Patients with prematurity, sepsis, meningitis, asphyxia and intracranial hemorrhage were excluded from this study. A total
of 75 patients were enrolled and divided into four groups according to their clinical status. Group I subjects comprised 10 older patients who were known to have kernicterus. ABRs were examined retrospectively to determine the posticteric changes. Group II consisted of 6 jaundiced patients with ABR testing before and after blood exchange transfusion (BET). They were tested within 2 hours before BET and on the same day after BET. Group III consisted of 20 jaundiced patients who only had ABR testing after BET. Group IV consisted of 39 infants who needed only phototherapy (PT), with ABR testing after the procedure. The criteria for procedure selection followed the traditional concept [18]. For the group II to IV patients, follow-up testing was done at 1 month old and then at 1 year old if the initial ABRs showed abnormalities. Besides the ABR testings, neurological and audiological evaluations were performed for the patients with kernicterus (group I) and those with ABR abnormalities in the following-up of group II to IV patients. Skin disc electrodes were applied to the vertex (Cz), earlobe ipsilateral to the signal (AI or A2) and middle forehead (ground). Bioelectric activities were recorded for the potential difference between Cz and Al or A2. Positivity at the vertex was recorded as an upward deflection. A total of 2,000 responses was amplified (l05 times), filtered (50 to 1 kHz) and then analyzed with a signal averager (NIHON-KOHDEN Neuropack II) for the initial 10 msec after stimulus. Acoustic stimuli (clicks) were generated by passing alternating-polarity square waves at a frequency of 2 kHz from a stimulator into a set of attached earphones. The stimuli were presented monaurally at a rate of 10 clicks per second. Intensity was adjusted to normal hearing level (HL) referring to the average adult threshold in the recording environment. All patients were tested as to the hearing threshold [19] (i.e. minimal intensity of stim-
Table I Clinical and ABR findings in kern icteric patients
(group /)
Peak Type of ABR changes Case Sex Present bilirubin cerebral Thresh- Latency Ty e* no age (mg/dl) palsy old I V I-V P
1 2 3 4 5 6 7 8
9 10
M M F F M F M M M M
7m 9m 2y 2y 7m 6y 2y 10 y 10 y 12 y
34 36 36 51 30 30 33 19 29 ?
S S A A H S H A+S S A+S
t t t t N N t t t t
P P N P P N
P P P P P (-)
P P P N N (-)
(-) (-) (-) (-) (-) (-) (-) (-) (-)
P P
N
M: male, F: female, S: spastic, A: athetoid, H: hypotonia, t: elevated, P: prolonged, N: normal, (-): absence of peak latencies (wave I or V) or I-V interpeak latencies, * see text.
298 Brain & Development, Vol 11, No 5,1989
IV IV III II II III I I II
ulation eliciting recognizable wave V) and their brainstem responses to supra threshold stimuli. Waves I, III and V were identified, and their latencies as well as interpeak latencies (I-V) were measured. Abnormalities of ABRs were analyzed, comparing with a set of age-matched controls (30 newborns, 55 infants and 59 older children) in the same laboratory which had been described before [20,21]. RESULTS The main clinical and ABR findings in the group I cases are summarized in Table 1. There were 7 males and 3 females. Their present ages ranged from 7 months to 12 years. As shown in the table, all of them had a history of overt hyperbilirubinemia and received BET one to four times during the neonatal period, except case 10. The peak bilirubin level ranged from 19 to 51 mg/dl. Only case 10 had not received any treatment and his peak bilirubin level was not checked. The current neurological status of all these patients included cerebral palsy and mental subnormality. All except case 5 had hearing and/ or speech impairment. The motor deficits consisted of spasticity, athetosis, hypotonia and mixed type. The ABR abnormalities can be classified on the basis of the audiological and neurolOgical status as below: Type I: No ABR appeared up to 100 dB clicks. This
ABR Type
Latency ( .. sec)
dB
I _
I
-= 100
-
v
I-V
6.5
3.8
V
II~90
2.7
I
III~70
2.0
I V IV~85
J
3_0
8.4
5.4
O.4Iv
1.sec Fig 1 Four types of ABR abnormalities encountered in this study. Type I: no response obtained up 100 dBHL, the patient has profound hearing loss with or without brainstem involvement. Type II: prolongation of wave I latency with a normal interpeak interval suggesting peripheral involvement. Type III: normal wave I latency with a prolonged 1- V interval or absence of later components, which is considered to indicate brainstem dysfunction. Type IV: both the 1-V interval and wave I latencies prolonged, which means that audiological and neurological disorders coexist. In this figure, the examples of ABR types I, II, III and IV were taken from the recordings of cases 7, 4, 6 and 1 in group I patients, respectively.
means that the patient is deaf to signals up to maximal intensity. But since no ABR is available for analysis, the patient's neurological status is obscure. Type II: The I-V interval fell within normal range for age, but wave I latency was prolonged to 60 dBHL clicks and the click threshold was elevated. These patients are considered to suffer from hearing loss without additional neurological disorders. Type III: The I-V interval was prolonged, but wave I was normal to 60 dBHL clicks. Patients in this group are considered to suffer predominantly from neurological disorders. Type IV: Both the I-V interval and wave I latency were prolonged. It seems that both audiological and neurological disorders exist in these patients. Selective case demonstrations of these four types of abnormalities of ABRs in group I patients are shown in Fig 1. For group II patients, the bilirubin levels and the
Table 2 Clinical data and ABR changes in patients who received BET (group II) No Sex
F
2 3
Age (day)
Bilirubin (mg%) Before BET (After BET)
ABR changes Before BET (After BET)
5
22.5 (14.4)
(N)
F F
4
4
M
3
5
M
5
6
M
5
14.7 ( 9.4) 20 (10 ) 23.7 (13 ) 19.8 (12.8) 21.6 (11.5)
changes of ABRs before and after BET are shown in Table 2. Five patients had abnormal ABRs before BET. Three of them became normal and the other 2 showed improvement after BET. Table 3 shows the average parameter changes of ABR before and after BET. The latency of each component was shortened and the amplitude increased after exchange transfusion. Statistically, the improvement of wave I was the most remarkable after BET. Both group II and III patients who received BET were screened with ABR testing. As shown in Table 4, 9 of the 26 patients had abnormal ABRs initially. The findings were no response (type I abnormality) in one patient, prolongation of wave I latency (type II) in 6 and prolonged wave III or wave V latency (type III) in the other 2 patients. Follow-up ABR testings one month later showed normalization in all except one patient. That patient had bilirubin level up to 32.7 mg/dl at the 6th day of life. The cause of jaundice was considered to be G-6-P D deficiency. He received BET twice and, after the 2nd BET, his ABRs were not obtainable even at one year old. The audiological test proved his deafness. No ABR data could be traced before BET. For group IV patients, as shown in Table 4, the ABR screening showed 6 with minor abnormal findings initial-
Table 4 ABR screening for patients with blood exchange transfusion (BET) and phototherapy (PT)*
N
Prolonged wave 1*, Lt
BET 26 cases
(N)
Elevated threshold, Bil (N)
Prolonged wave I & III-V, Rt (Prolonged wave I, Rt) Low amplitude, Bil (N)
Prolonged wave I, Bil (Prolonged wave I, improved, Bil)
Lt: left, Rt: right, Bil: bilateral, N: normal, BET: blood exchange transfusion, * prolonged latency of wave I.
Table 3 ABR changes before and after blood exchange transfusion (BET) BeforeBET*
AfterBET*
P value**
Latencies (ms) I III V Amplitudes (j.lV)
2.48 ± 0.69 5.22 ± 0.58 7.44 ± 0.63
2.18 ± 0.36 5.10 ± 0.47 7.39 ± 0.63
< 0.005 <0.05 < 0.05
I III V
0.71 ± 0.30 0.34 ± 0.09 0.59 ± 0.22
0.86 ± 0.31 0.39 ± 0.09 0.64 ± 0.23
< 0.005 < 0.005 < 0.05
No of tests = 12 (based on measurements from both ears of 6 patients), * values and mean ± SD, **paired t tests.
PT
39 cases
/
/".
(17) N
ABR within 1st week ABR at 1 month
(9) P
/\
(8) (1) N P
(33) N
".
(6) P
/\
(5) (1) N P
1
1
P (NR)
N
ABR at 1 year
N: normal ABR recording, P: pathologic ABR recording, NR: no response, ( ): number of cases, * see text.
Table 5 Peak bilirubin levels and percentages of ABR abnormalities in kern icteric and non-kernicteric patients
Kernicterus (n =9) Non-kernicterus (n = 24) P value**
Day of BET*
Peak b iliru bin level after day 3 (mg%)
ABR abnormality
5.3 d
33.1 (n = 9)
100%
4.1 d
23.2 (n
= 17)
< 0.005
(%)
16% < 0.005
* mean age of blood exchange transfusion, ** student t tests.
Hung: ABR in jaundiced patients 299
ly. Among them, 4 had prolongation of wave I latency (type II) and 2 had borderline prolongation of I-V interpeak latency (type III). By the 2nd month of life, most of these abnormalities had disappeared. One remaining abnormal patient showed recovery at one year of age. All group II, III and IV patients with abnormal initial ABRs were followed-up carefully. The neurological and audiological outcomes were excellent except the deaf case. It is easy to see in Table 5 that the patients who developed kernicterus after BET had significantly higher bilirubin levels than those without kernicterus. The rate of ABR abnormalities was more remarkable in the kernicteric group. DISCUSSION Kernicterus or bilirubin encephalopathy is a neurologic syndrome caused by serious neonatal jaundice arising from various etiologies [22]. The clinical findings of kernicterus follow a complicated course. Usually lethargy and feeding problem are the earliest clinical features during the acute stage although sometimes these are obscure. Then, hypertonia and opisthotonus develop in the subacute stage. The late picture of kernicterus usually develops a few months or years later. This includes hearing loss, dyskinesia, cerebral palsy, upward gaze palsy, mental retardation and minimal brain dysfunction [22]. Athetosis is the major type of cerebral palsy, but other features such as hypotonia, spasticity or mixed type can also be found depended on clinical stage and the predominant pathological lesion [23]. The kernicteric patients in this study also showed variety in presentation: 4 athetosis (including 2 mixed), 4 spasticity and 2 hypotonia. There is great controversy as to the relationship of the bilirubin level and the occurrence of bilirubin toxicity, although the concept of "20 mg/dl" for an exchange transfusion in an uncomplicated full-term infant has been traditionally accepted for many years [4, 24]. Most of the group I patients in this study were transferred patients and had peak bilirubin levels far beyond this safety level. Although they had received BET from one to four times within the first week of life, kernicterus still occurred because of the delay in treatment. Therefore, early recognition of bilirubin toxicity remains a challenge to both parents and physicians. From Table 5, it is easy to see that those who developed kernicterus (mostly group I patients) had delay in treatment and higher bilirubin level. To prevent this, some monitoring systems are necessary besides clinical observation. The application of ABR sheds some light on this problem. Posticteric sequelae such as hearing loss, as well as CNS deficits, can be demonstrated by recording of the auditory brainstem responses. The absence or latency prolongation of wave I indicates peripheral hearing loss [12]. The total absence of ABR is thought to show severe peri-
300 Brain & Development, Vol 11, No 5,1989
pheral involvement and total deafness may occur, but accompanying brainstem damage cannot be ruled out [11, 19]. The prolongation of I-V interpeak latency and deformity or absence of later components (wave II-V) suggest brainstem derangement, either functional or anatomical [10]. With the help of ABR study, it has been found that hyperbilirubinemia can cause transient brainstem encephalopathy with levels as low as 17.7 mg/dl [11] and 15 mg/ dl [14]. Most cases are reversible if the jaundice is resolved in a short time. It has been hypothesized that kernicterus will occur if the integrity of the blood-brain barrier breaks down at a certain bilirubin level [25]. What the critical point for irreversible brain damage is remains a major subject of debate. The traditional concept can only be used as a rough guide. Kuriyama et al [17] found an irreversible change of ABR in a neonate with a bilirubin level of 18 mg/dl. An individual's susceptibility to bilirubin toxicity, in fact, depends on multiple risk factors such as asphyxia, acidosis, infection, hypothermia, prematurity etc [22] . In a study of 25 kernicteric patients, Kaga et al [16] found that 88% showed ABR threshold elevation, most of their ABRs were peripheral involvement with prolonged latency of wave I or absence of ABRs, but no brainstem lesion pattern. However, other reports showed the typical brainstem involvement in patients with neonatal jaundice at acute stage [11, 13, 14]. In the present study, both peripheral and brains tern involvement were demonstrated in group I patients to be the late sequelae of bilirubin neurotoxicity. Since these patients had delayed or no treatment for their jaundice, it is possible that the results reflect more of the natural fate of posticteric changes. But the selective fate of ABR changes in an individual remains obscure. The beneficial effect of therapy could also be demonstrated by ABR testing. Nwaesei et al [15] showed acute brainstem toxicity in 9 jaundiced infants before BET and its reversibility after the procedure. They found that the interpeak latencies of I-III and III-V were significantly reduced after BET. In this study, the acute effect of BET was demonstrated in both peripheral and central transmissions, in varied degrees. For group II to IV patients, the ABR screening showed that peripheral involvement seemed to be more common than the central effect. However, most of the group I patients were shown to have more central lesions than those in the other groups. It is possible that bilirubin neurotoxicity might affect the peripheral system at the early stage. The central lesions will appear later if more bilirubin damage exists. This assumption partially corresponds with the idea of Lenhardt et al [14] that, when bilirubin concentration reaches a critical value, hearing loss can occur and wave I may not be well resolved; then the later waves might serve as early signals of bilirubin inmtration of neural system.
Longitudinal study of ABRs in the control group [21] showed that ABR is a reliable test to reflect the maturational process. Some screening or follow-up studies [26, 27] about hyperbilirubinemia had mentioned that ABR testing is a sensitive procedure for the early detection of hearing impairment. The predictive value of a positive test was low since some improvement and reversibility could be found later. Nevertheless, the predictive value of a negative test was very high. Those who initially passed screening were later found to have normal hearing if they did not suffer from other damages [27]. Thus in this study only those with abnormal initial ABRs were followed-up serially. Majority of the group II, III and IV patients got normalization of their ABRs finally. The results of ABR testing were more fruitful when done at one month old or later, which is compatible with the previous report [26]. Therefore follow-up ABR testing is necessary. It is concluded from this ABR study that the conventional criteria for the management of hyperbilirubinemia seems effective. But some modifications are needed, especially in patients with high risk factors. The ABR study can provide as a reference in the detection of the occurrence of bilirubin toxicity and monitoring of the effect of therapy. Minor abnormality of ABRs may be non-significant in clinial implication. However, during the initial testing, if the ABRs showed evidence of severe brainstem dysfunction (marked I-V interval prolongation or absence of later components) or profound peripheral involvement (no response or severe depression), intensive management should be considered to avoid permanent neural damage, and what's more, close follow-up to determine any possible sequela.
ACKNOWLEDGMENTS The author expresses his thanks to Miss Fune-Jiun Wang for technical assistance and Miss Cyndia Lo for preparing the manuscript.
REFERENCES 1. Bengtsson B, Verneholt J. A follow-up study of hyperbilirubinemia in healthy, full-term infants without isoimmunization. Acta Pediatr Scand 1974;63:70-80. 2. Johnson WH, Angara V, Baumel R, et at Erythroblastosis fetalis and hyperbilirubinemia: a five-year follow-up with neurological, psychological and audiological evaluation. Pediatrics 1967;39:88-92. 3. Abramovich SJ, Gregory S, Slemick K, Stewart A. Hearing loss in very low birthweight infants treated with neonatal intensive care. Arch Dis Child 1979;54:421-6. 4. Lucey JF. Bilirubin and brain damage - A real mess. Pediatrics 1982;69:381-2. 5. Starr A, Achor L. Auditory brainstem responses in neurologic disease. Arch Neuro/1975;32:761-8. 6. Stockard n, Stockard JE, Sharbrough FW. Detection and localization of occult lesions with brainstem auditory responses. Mayo Clin Proc 1977;52:761-9.
7. Hecox KE, Cone B, Blaw ME. Brainstem auditory evoked response in the diagnosis of pediatric neurologic disease. Neurology 1981;31:832-40. 8. Jewett DL, Williston JS. Auditory-evoked far-field potentials averaged from the scalp of humans. Brain 1971;94:681-96. 9. Starr A, Hamilton A. Correlation between confirmed sites of neurological lesions and abnormalities of far-field auditory brainstem responses. Electroencephalogr Clin Neurophysiol 1976;41:595-608. 10. Stockard n, Rossiter VS. Qinical and pathologic correlates of brainstem auditory response abnormalities. Neurology 1977;27: 316-25. 11. Wennberg RP, Ahlfors CE, Bickers R, McMurtry CA, Shetter JL. Abnormal auditory brainstem responses in a newborn infant with hyperbilirubinemia: improvement with exchange transfusion. J Pedia.tr 1982;100:624-6. 12. Chisin R, Perlman M, Sohmer H. Cochlear and brainstem responses in hearing loss following neonatal hyperbilirubinemia. Ann Otol Rhinol Laryngol 1979;88: 352-7. 13. Perlman M, Fainmesser P, Sohmer H, Tamari H, Wax Y, Persmer B. Auditory nerve-brainstem evoked responses in hyperbilirubinemic neonates. Pediatrics 1983;72:658-64. 14. Lenhardt ML, McArtor R, Bryant B. Effects of neonatal hyperbilirubinemia on the brain stem electric response. J Pediatr 1984;104:281-4. 15. Nwaesei CG, Van Aerde J, Boyden M, Perlman M. Changes in auditory brainstem responses in hyperbilirubinemic infants before and after exchange transfusion. Pediatrics 1984;74: 800-3. 16. Kaga K, Kitazumi E, Kodama K. Auditory brainstem responses of kernicteric infants. Int J Pediatr Otorhinolaryngol 1979;1:255-64. 17. Kuriyama M, Konishi Y, Mikawa H. The effect of neonatal hyperbilirubinemia on the auditory brainstem response. Brain Dev (Tokyo) 1986;8:240-5. 18. Gartner LM, Lee K-Y. Jaundice and liver disease. In: Behrman RE, ed. Neonatal-perinatal medicine. St Louis: CV Mosby, 1977:394-429. 19. Despland PA, Galambos R. The auditory brainstem response (ABR) is a useful diagnostic tool in the intensive care nursery. Pediatr Res 1980;14:154-8. 20. Hung KL. Auditory brainstem response in normal Chinese newborns. J Formosan Med Asso 1984;83:1287-93. 21. Hung KL. Development of auditory brainstem evoked responses in normal Chinese children. Acta Paed Sin 1989;30: 23-9. 22. Behrman RE, Kliegman RM. Noninfectious disorders of the newborn. In: Behrman RE, Vaughan VC III, eds. Nelson textbook of pediatrics. 12th ed. Philadelphia: WB Saunders, 1983:364-97. 23. Sarnat M. Neonatal bilirubin encephalopathy. In: Sarnat HB, ed. Topics in neonatal neurology. Orlando: Grune & Stratton, 1984: 109-36. 24. American Academy of Pediatrics. Standards and recommendations for hospital care of newborn infants, ed 6. Evanston (lL): American Academy of Pediatrics, 1977: 95. 25. Levine RL, Fredericks WR, Rapoport SI. Entry of bilirubin into the brain due to opening of the blood-brain barrier. Pe· diatrics 1982;69:255-9. 26. Stockard JE, Stockard JJ, Kleinberg F, Westmoreland BF. Prognostic value of brainstem auditory evoked potentials in neonates. Arch Neuro/1983;40:360-5. 27. Shannon DA, Felix JK, Krumholz A, Goldstein PJ, Harris KC. Hearing screening of high-risk newborns with brainstem auditory evoked potentials: a follow-up study. Pediatrics 1984;73:22-6.
Hung: ABR in jaundiced patients 301
To understand the neurotoxicity of neonatal hyperbilirubinemia, auditory brain stem response (ABR) studies were performed in 75 jaundiced patients, who were divided into 4 groups in either a retrospective or prospective way. Retrospective ABR study in 10 known kemicteric patients (group I) showed elevation of the hearing threshold and delay of wave I, as well as prolongation of central brainstem transmission, in various degrees, in most cases. Six jaundiced infants (group II) with ABR testing before and after blood exchange transfusion (BET) showed shortening of wave latencies and increase in amplitude after BET. There were additional 20 infants with BET (group III) and 39 with phototherapy (group IV) receiving ABR testing after the therapeutic procedures. Prospective follow-up in groups II, III and IV showed normalization of the ABRs in all except one patient in the following months. These findings demonstrate the nature of bilirubin neurotoxicity and the prognostic value of ABRs in the monitoring of bilirubin toxicity. ABR testing is sensitive in reflecting the effect of hyperbilirubinemia, and provides a valuable guide for the early recognition and close follow-up of bilirubin neurotoxicity. Key words: Auditory brainstem response (ABR), hyperbilirubinemia, kernicterus, bilirubin neurotoxicity. Hung K-L. Auditory brainstem responses in patients with neonatal hyperbilirubinemia and bilirubin encephalopathy. Brain Dev 1989;11:297-301
Neonatal hyperbilirubinemia is one of the major problems in neonatal intensive care. Its sequelae include many neurological deficits [1-3]. The occurrence of kernicterus, which is one of these sequelae, has been reduced because of recent advances in mUltiple exchange transfusion, phototherapy and frequent serum bilirubin concentration measurement. Nevertheless, bilirubin encephalopathy still continues to occur because of inconsistency in clinical presentation of early bilirubin toxicity [4]. The application of the auditory brainstem response (ABR) in the study of neonatal hyperbilirubinemia seems reasonable because of its potential value in assessing hearing and brainstem functions [5-10]. The relationship between neonatal hyperbilirubinemia and ABR has been mentioned in a few studies [11-15]. Chisin et al [12] found that the occurrence of hearing loss following neonatal hyper-
From the Department of Pediatrics, Cathay General Hospital Taipei, Taiwan. Received for publication: March 4, 1989. Accepted for pUblication: July 22, 1989. Correspondence address: Dr. Kun-Long Hung, Department of Pediatrics, Cathay General Hospital, 280, Jen-Ai Rd, Sec 4, Taipei, Taiwan, Republic of China.
bilirubinemia showed a good correlation with the absence of wave I on ABR testing. Perlman et al [13] reported that jaundice was associated with significant transient aberrations of ABRs, suggestive of a transient brainstem encephalopathy. Studies on ABR changes in overt bilirubin encephalopathy have been rather scarce [16, 17]. To understand the neurotoxicity of hyperbilirubinemia, as reflected by the ABRs, a retrospective and prospective study was conducted. The present report describes the late changes on ABRs in kernicteric patients and acute changes of ABRs in neonates suffering from hyperbilirubinemia. SUBJECTS AND METHODS During the period from January 1984 to December 1986, ABR studies were performed in patients suffering from neonatal hyperbilirubinemia in our nursery and neonatal intensive care unit (NICU), as well as those with overt bilirubin encephalopathy from outpatient clinics. All the patients were full-term at birth with birth weights of greater than 2,500 gm. The causes of jaundice were mostly ABO incompatibility, glucose-6-phosphate dehydrogenase (G-6-P D) deficiency and unknown factors. Patients with prematurity, sepsis, meningitis, asphyxia and intracranial hemorrhage were excluded from this study. A total
of 75 patients were enrolled and divided into four groups according to their clinical status. Group I subjects comprised 10 older patients who were known to have kernicterus. ABRs were examined retrospectively to determine the posticteric changes. Group II consisted of 6 jaundiced patients with ABR testing before and after blood exchange transfusion (BET). They were tested within 2 hours before BET and on the same day after BET. Group III consisted of 20 jaundiced patients who only had ABR testing after BET. Group IV consisted of 39 infants who needed only phototherapy (PT), with ABR testing after the procedure. The criteria for procedure selection followed the traditional concept [18]. For the group II to IV patients, follow-up testing was done at 1 month old and then at 1 year old if the initial ABRs showed abnormalities. Besides the ABR testings, neurological and audiological evaluations were performed for the patients with kernicterus (group I) and those with ABR abnormalities in the following-up of group II to IV patients. Skin disc electrodes were applied to the vertex (Cz), earlobe ipsilateral to the signal (AI or A2) and middle forehead (ground). Bioelectric activities were recorded for the potential difference between Cz and Al or A2. Positivity at the vertex was recorded as an upward deflection. A total of 2,000 responses was amplified (l05 times), filtered (50 to 1 kHz) and then analyzed with a signal averager (NIHON-KOHDEN Neuropack II) for the initial 10 msec after stimulus. Acoustic stimuli (clicks) were generated by passing alternating-polarity square waves at a frequency of 2 kHz from a stimulator into a set of attached earphones. The stimuli were presented monaurally at a rate of 10 clicks per second. Intensity was adjusted to normal hearing level (HL) referring to the average adult threshold in the recording environment. All patients were tested as to the hearing threshold [19] (i.e. minimal intensity of stim-
Table I Clinical and ABR findings in kern icteric patients
(group /)
Peak Type of ABR changes Case Sex Present bilirubin cerebral Thresh- Latency Ty e* no age (mg/dl) palsy old I V I-V P
1 2 3 4 5 6 7 8
9 10
M M F F M F M M M M
7m 9m 2y 2y 7m 6y 2y 10 y 10 y 12 y
34 36 36 51 30 30 33 19 29 ?
S S A A H S H A+S S A+S
t t t t N N t t t t
P P N P P N
P P P P P (-)
P P P N N (-)
(-) (-) (-) (-) (-) (-) (-) (-) (-)
P P
N
M: male, F: female, S: spastic, A: athetoid, H: hypotonia, t: elevated, P: prolonged, N: normal, (-): absence of peak latencies (wave I or V) or I-V interpeak latencies, * see text.
298 Brain & Development, Vol 11, No 5,1989
IV IV III II II III I I II
ulation eliciting recognizable wave V) and their brainstem responses to supra threshold stimuli. Waves I, III and V were identified, and their latencies as well as interpeak latencies (I-V) were measured. Abnormalities of ABRs were analyzed, comparing with a set of age-matched controls (30 newborns, 55 infants and 59 older children) in the same laboratory which had been described before [20,21]. RESULTS The main clinical and ABR findings in the group I cases are summarized in Table 1. There were 7 males and 3 females. Their present ages ranged from 7 months to 12 years. As shown in the table, all of them had a history of overt hyperbilirubinemia and received BET one to four times during the neonatal period, except case 10. The peak bilirubin level ranged from 19 to 51 mg/dl. Only case 10 had not received any treatment and his peak bilirubin level was not checked. The current neurological status of all these patients included cerebral palsy and mental subnormality. All except case 5 had hearing and/ or speech impairment. The motor deficits consisted of spasticity, athetosis, hypotonia and mixed type. The ABR abnormalities can be classified on the basis of the audiological and neurolOgical status as below: Type I: No ABR appeared up to 100 dB clicks. This
ABR Type
Latency ( .. sec)
dB
I _
I
-= 100
-
v
I-V
6.5
3.8
V
II~90
2.7
I
III~70
2.0
I V IV~85
J
3_0
8.4
5.4
O.4Iv
1.sec Fig 1 Four types of ABR abnormalities encountered in this study. Type I: no response obtained up 100 dBHL, the patient has profound hearing loss with or without brainstem involvement. Type II: prolongation of wave I latency with a normal interpeak interval suggesting peripheral involvement. Type III: normal wave I latency with a prolonged 1- V interval or absence of later components, which is considered to indicate brainstem dysfunction. Type IV: both the 1-V interval and wave I latencies prolonged, which means that audiological and neurological disorders coexist. In this figure, the examples of ABR types I, II, III and IV were taken from the recordings of cases 7, 4, 6 and 1 in group I patients, respectively.
means that the patient is deaf to signals up to maximal intensity. But since no ABR is available for analysis, the patient's neurological status is obscure. Type II: The I-V interval fell within normal range for age, but wave I latency was prolonged to 60 dBHL clicks and the click threshold was elevated. These patients are considered to suffer from hearing loss without additional neurological disorders. Type III: The I-V interval was prolonged, but wave I was normal to 60 dBHL clicks. Patients in this group are considered to suffer predominantly from neurological disorders. Type IV: Both the I-V interval and wave I latency were prolonged. It seems that both audiological and neurological disorders exist in these patients. Selective case demonstrations of these four types of abnormalities of ABRs in group I patients are shown in Fig 1. For group II patients, the bilirubin levels and the
Table 2 Clinical data and ABR changes in patients who received BET (group II) No Sex
F
2 3
Age (day)
Bilirubin (mg%) Before BET (After BET)
ABR changes Before BET (After BET)
5
22.5 (14.4)
(N)
F F
4
4
M
3
5
M
5
6
M
5
14.7 ( 9.4) 20 (10 ) 23.7 (13 ) 19.8 (12.8) 21.6 (11.5)
changes of ABRs before and after BET are shown in Table 2. Five patients had abnormal ABRs before BET. Three of them became normal and the other 2 showed improvement after BET. Table 3 shows the average parameter changes of ABR before and after BET. The latency of each component was shortened and the amplitude increased after exchange transfusion. Statistically, the improvement of wave I was the most remarkable after BET. Both group II and III patients who received BET were screened with ABR testing. As shown in Table 4, 9 of the 26 patients had abnormal ABRs initially. The findings were no response (type I abnormality) in one patient, prolongation of wave I latency (type II) in 6 and prolonged wave III or wave V latency (type III) in the other 2 patients. Follow-up ABR testings one month later showed normalization in all except one patient. That patient had bilirubin level up to 32.7 mg/dl at the 6th day of life. The cause of jaundice was considered to be G-6-P D deficiency. He received BET twice and, after the 2nd BET, his ABRs were not obtainable even at one year old. The audiological test proved his deafness. No ABR data could be traced before BET. For group IV patients, as shown in Table 4, the ABR screening showed 6 with minor abnormal findings initial-
Table 4 ABR screening for patients with blood exchange transfusion (BET) and phototherapy (PT)*
N
Prolonged wave 1*, Lt
BET 26 cases
(N)
Elevated threshold, Bil (N)
Prolonged wave I & III-V, Rt (Prolonged wave I, Rt) Low amplitude, Bil (N)
Prolonged wave I, Bil (Prolonged wave I, improved, Bil)
Lt: left, Rt: right, Bil: bilateral, N: normal, BET: blood exchange transfusion, * prolonged latency of wave I.
Table 3 ABR changes before and after blood exchange transfusion (BET) BeforeBET*
AfterBET*
P value**
Latencies (ms) I III V Amplitudes (j.lV)
2.48 ± 0.69 5.22 ± 0.58 7.44 ± 0.63
2.18 ± 0.36 5.10 ± 0.47 7.39 ± 0.63
< 0.005 <0.05 < 0.05
I III V
0.71 ± 0.30 0.34 ± 0.09 0.59 ± 0.22
0.86 ± 0.31 0.39 ± 0.09 0.64 ± 0.23
< 0.005 < 0.005 < 0.05
No of tests = 12 (based on measurements from both ears of 6 patients), * values and mean ± SD, **paired t tests.
PT
39 cases
/
/".
(17) N
ABR within 1st week ABR at 1 month
(9) P
/\
(8) (1) N P
(33) N
".
(6) P
/\
(5) (1) N P
1
1
P (NR)
N
ABR at 1 year
N: normal ABR recording, P: pathologic ABR recording, NR: no response, ( ): number of cases, * see text.
Table 5 Peak bilirubin levels and percentages of ABR abnormalities in kern icteric and non-kernicteric patients
Kernicterus (n =9) Non-kernicterus (n = 24) P value**
Day of BET*
Peak b iliru bin level after day 3 (mg%)
ABR abnormality
5.3 d
33.1 (n = 9)
100%
4.1 d
23.2 (n
= 17)
< 0.005
(%)
16% < 0.005
* mean age of blood exchange transfusion, ** student t tests.
Hung: ABR in jaundiced patients 299
ly. Among them, 4 had prolongation of wave I latency (type II) and 2 had borderline prolongation of I-V interpeak latency (type III). By the 2nd month of life, most of these abnormalities had disappeared. One remaining abnormal patient showed recovery at one year of age. All group II, III and IV patients with abnormal initial ABRs were followed-up carefully. The neurological and audiological outcomes were excellent except the deaf case. It is easy to see in Table 5 that the patients who developed kernicterus after BET had significantly higher bilirubin levels than those without kernicterus. The rate of ABR abnormalities was more remarkable in the kernicteric group. DISCUSSION Kernicterus or bilirubin encephalopathy is a neurologic syndrome caused by serious neonatal jaundice arising from various etiologies [22]. The clinical findings of kernicterus follow a complicated course. Usually lethargy and feeding problem are the earliest clinical features during the acute stage although sometimes these are obscure. Then, hypertonia and opisthotonus develop in the subacute stage. The late picture of kernicterus usually develops a few months or years later. This includes hearing loss, dyskinesia, cerebral palsy, upward gaze palsy, mental retardation and minimal brain dysfunction [22]. Athetosis is the major type of cerebral palsy, but other features such as hypotonia, spasticity or mixed type can also be found depended on clinical stage and the predominant pathological lesion [23]. The kernicteric patients in this study also showed variety in presentation: 4 athetosis (including 2 mixed), 4 spasticity and 2 hypotonia. There is great controversy as to the relationship of the bilirubin level and the occurrence of bilirubin toxicity, although the concept of "20 mg/dl" for an exchange transfusion in an uncomplicated full-term infant has been traditionally accepted for many years [4, 24]. Most of the group I patients in this study were transferred patients and had peak bilirubin levels far beyond this safety level. Although they had received BET from one to four times within the first week of life, kernicterus still occurred because of the delay in treatment. Therefore, early recognition of bilirubin toxicity remains a challenge to both parents and physicians. From Table 5, it is easy to see that those who developed kernicterus (mostly group I patients) had delay in treatment and higher bilirubin level. To prevent this, some monitoring systems are necessary besides clinical observation. The application of ABR sheds some light on this problem. Posticteric sequelae such as hearing loss, as well as CNS deficits, can be demonstrated by recording of the auditory brainstem responses. The absence or latency prolongation of wave I indicates peripheral hearing loss [12]. The total absence of ABR is thought to show severe peri-
300 Brain & Development, Vol 11, No 5,1989
pheral involvement and total deafness may occur, but accompanying brainstem damage cannot be ruled out [11, 19]. The prolongation of I-V interpeak latency and deformity or absence of later components (wave II-V) suggest brainstem derangement, either functional or anatomical [10]. With the help of ABR study, it has been found that hyperbilirubinemia can cause transient brainstem encephalopathy with levels as low as 17.7 mg/dl [11] and 15 mg/ dl [14]. Most cases are reversible if the jaundice is resolved in a short time. It has been hypothesized that kernicterus will occur if the integrity of the blood-brain barrier breaks down at a certain bilirubin level [25]. What the critical point for irreversible brain damage is remains a major subject of debate. The traditional concept can only be used as a rough guide. Kuriyama et al [17] found an irreversible change of ABR in a neonate with a bilirubin level of 18 mg/dl. An individual's susceptibility to bilirubin toxicity, in fact, depends on multiple risk factors such as asphyxia, acidosis, infection, hypothermia, prematurity etc [22] . In a study of 25 kernicteric patients, Kaga et al [16] found that 88% showed ABR threshold elevation, most of their ABRs were peripheral involvement with prolonged latency of wave I or absence of ABRs, but no brainstem lesion pattern. However, other reports showed the typical brainstem involvement in patients with neonatal jaundice at acute stage [11, 13, 14]. In the present study, both peripheral and brains tern involvement were demonstrated in group I patients to be the late sequelae of bilirubin neurotoxicity. Since these patients had delayed or no treatment for their jaundice, it is possible that the results reflect more of the natural fate of posticteric changes. But the selective fate of ABR changes in an individual remains obscure. The beneficial effect of therapy could also be demonstrated by ABR testing. Nwaesei et al [15] showed acute brainstem toxicity in 9 jaundiced infants before BET and its reversibility after the procedure. They found that the interpeak latencies of I-III and III-V were significantly reduced after BET. In this study, the acute effect of BET was demonstrated in both peripheral and central transmissions, in varied degrees. For group II to IV patients, the ABR screening showed that peripheral involvement seemed to be more common than the central effect. However, most of the group I patients were shown to have more central lesions than those in the other groups. It is possible that bilirubin neurotoxicity might affect the peripheral system at the early stage. The central lesions will appear later if more bilirubin damage exists. This assumption partially corresponds with the idea of Lenhardt et al [14] that, when bilirubin concentration reaches a critical value, hearing loss can occur and wave I may not be well resolved; then the later waves might serve as early signals of bilirubin inmtration of neural system.
Longitudinal study of ABRs in the control group [21] showed that ABR is a reliable test to reflect the maturational process. Some screening or follow-up studies [26, 27] about hyperbilirubinemia had mentioned that ABR testing is a sensitive procedure for the early detection of hearing impairment. The predictive value of a positive test was low since some improvement and reversibility could be found later. Nevertheless, the predictive value of a negative test was very high. Those who initially passed screening were later found to have normal hearing if they did not suffer from other damages [27]. Thus in this study only those with abnormal initial ABRs were followed-up serially. Majority of the group II, III and IV patients got normalization of their ABRs finally. The results of ABR testing were more fruitful when done at one month old or later, which is compatible with the previous report [26]. Therefore follow-up ABR testing is necessary. It is concluded from this ABR study that the conventional criteria for the management of hyperbilirubinemia seems effective. But some modifications are needed, especially in patients with high risk factors. The ABR study can provide as a reference in the detection of the occurrence of bilirubin toxicity and monitoring of the effect of therapy. Minor abnormality of ABRs may be non-significant in clinial implication. However, during the initial testing, if the ABRs showed evidence of severe brainstem dysfunction (marked I-V interval prolongation or absence of later components) or profound peripheral involvement (no response or severe depression), intensive management should be considered to avoid permanent neural damage, and what's more, close follow-up to determine any possible sequela.
ACKNOWLEDGMENTS The author expresses his thanks to Miss Fune-Jiun Wang for technical assistance and Miss Cyndia Lo for preparing the manuscript.
REFERENCES 1. Bengtsson B, Verneholt J. A follow-up study of hyperbilirubinemia in healthy, full-term infants without isoimmunization. Acta Pediatr Scand 1974;63:70-80. 2. Johnson WH, Angara V, Baumel R, et at Erythroblastosis fetalis and hyperbilirubinemia: a five-year follow-up with neurological, psychological and audiological evaluation. Pediatrics 1967;39:88-92. 3. Abramovich SJ, Gregory S, Slemick K, Stewart A. Hearing loss in very low birthweight infants treated with neonatal intensive care. Arch Dis Child 1979;54:421-6. 4. Lucey JF. Bilirubin and brain damage - A real mess. Pediatrics 1982;69:381-2. 5. Starr A, Achor L. Auditory brainstem responses in neurologic disease. Arch Neuro/1975;32:761-8. 6. Stockard n, Stockard JE, Sharbrough FW. Detection and localization of occult lesions with brainstem auditory responses. Mayo Clin Proc 1977;52:761-9.
7. Hecox KE, Cone B, Blaw ME. Brainstem auditory evoked response in the diagnosis of pediatric neurologic disease. Neurology 1981;31:832-40. 8. Jewett DL, Williston JS. Auditory-evoked far-field potentials averaged from the scalp of humans. Brain 1971;94:681-96. 9. Starr A, Hamilton A. Correlation between confirmed sites of neurological lesions and abnormalities of far-field auditory brainstem responses. Electroencephalogr Clin Neurophysiol 1976;41:595-608. 10. Stockard n, Rossiter VS. Qinical and pathologic correlates of brainstem auditory response abnormalities. Neurology 1977;27: 316-25. 11. Wennberg RP, Ahlfors CE, Bickers R, McMurtry CA, Shetter JL. Abnormal auditory brainstem responses in a newborn infant with hyperbilirubinemia: improvement with exchange transfusion. J Pedia.tr 1982;100:624-6. 12. Chisin R, Perlman M, Sohmer H. Cochlear and brainstem responses in hearing loss following neonatal hyperbilirubinemia. Ann Otol Rhinol Laryngol 1979;88: 352-7. 13. Perlman M, Fainmesser P, Sohmer H, Tamari H, Wax Y, Persmer B. Auditory nerve-brainstem evoked responses in hyperbilirubinemic neonates. Pediatrics 1983;72:658-64. 14. Lenhardt ML, McArtor R, Bryant B. Effects of neonatal hyperbilirubinemia on the brain stem electric response. J Pediatr 1984;104:281-4. 15. Nwaesei CG, Van Aerde J, Boyden M, Perlman M. Changes in auditory brainstem responses in hyperbilirubinemic infants before and after exchange transfusion. Pediatrics 1984;74: 800-3. 16. Kaga K, Kitazumi E, Kodama K. Auditory brainstem responses of kernicteric infants. Int J Pediatr Otorhinolaryngol 1979;1:255-64. 17. Kuriyama M, Konishi Y, Mikawa H. The effect of neonatal hyperbilirubinemia on the auditory brainstem response. Brain Dev (Tokyo) 1986;8:240-5. 18. Gartner LM, Lee K-Y. Jaundice and liver disease. In: Behrman RE, ed. Neonatal-perinatal medicine. St Louis: CV Mosby, 1977:394-429. 19. Despland PA, Galambos R. The auditory brainstem response (ABR) is a useful diagnostic tool in the intensive care nursery. Pediatr Res 1980;14:154-8. 20. Hung KL. Auditory brainstem response in normal Chinese newborns. J Formosan Med Asso 1984;83:1287-93. 21. Hung KL. Development of auditory brainstem evoked responses in normal Chinese children. Acta Paed Sin 1989;30: 23-9. 22. Behrman RE, Kliegman RM. Noninfectious disorders of the newborn. In: Behrman RE, Vaughan VC III, eds. Nelson textbook of pediatrics. 12th ed. Philadelphia: WB Saunders, 1983:364-97. 23. Sarnat M. Neonatal bilirubin encephalopathy. In: Sarnat HB, ed. Topics in neonatal neurology. Orlando: Grune & Stratton, 1984: 109-36. 24. American Academy of Pediatrics. Standards and recommendations for hospital care of newborn infants, ed 6. Evanston (lL): American Academy of Pediatrics, 1977: 95. 25. Levine RL, Fredericks WR, Rapoport SI. Entry of bilirubin into the brain due to opening of the blood-brain barrier. Pe· diatrics 1982;69:255-9. 26. Stockard JE, Stockard JJ, Kleinberg F, Westmoreland BF. Prognostic value of brainstem auditory evoked potentials in neonates. Arch Neuro/1983;40:360-5. 27. Shannon DA, Felix JK, Krumholz A, Goldstein PJ, Harris KC. Hearing screening of high-risk newborns with brainstem auditory evoked potentials: a follow-up study. Pediatrics 1984;73:22-6.
Hung: ABR in jaundiced patients 301