Albertina De Lauretis,

Bruno De Capua, ABR evaluation of ototoxicity

Maria Teresa Barbieri,
Luisa Bellussi,
in cancer patients receiving
Desiderio Passàli cisplatin or carboplatin

E.N.T. Department, University of
Siena, Italy
KEY WORDS:
Auditory brainstem, carboplatin,
cisplatin, ototoxicity
RECEIVED/ACCEPTED:
February 1998/July 18, 1998
ADDRESS FOR CORRESPONDENCE:
Desiderio Passàli, Clinica ORL,
Policlinico Le Scotte, V. le Bracci,
I-53100 Siena, Italy. Tel: ‡39 577
40035, fax: ‡39 577 47940, e-mail:
PASSALI@unisi.it
The development of ototoxicity was evaluated using auditory
brainstem response (ABR) in cancer patients randomized to receive
a cisplatin-based chemotherapy (cisplatin dose: 70 mg/m2) or a
carboplatin-based chemotherapy (carboplatin dose: 250 mg/m2). The
ABR measurements were performed in a sound-treated room using
2000 clicks of alternating polarity at an intensity of 100 dB PESPL
presented to the patients at a rate of 21 clicks per second. Of 59
patients, 21 (9 in the cisplatin group and 12 in the carboplatin group)
met our pre-established criteria and were included in the ototoxicity
study. Two patients of the cisplatin group developed evidence of
clinically occult ototoxicity after two cycles of chemotherapy; the
latency of wave V of the ABR increased significantly from 5.874 to
6.336 msec and from 5.826 to 6.458 msec in both patients; these
patients had a hearing loss detected by conventional audiometry (125
to 8,000 Hz) after five and six cycles of chemotherapy, respectively.
None of the 12 examined carboplatin patients developed ABR-
measured ototoxicity or abnormal audiograms during treatment. Our
results suggest that ABR might prove to be useful in detecting early
hearing deterioration from cisplatin.

Scand Audiol 1999;28:139–43

Introduction cisplatin ranges from 20% to 90% following

Cisplatin is an antineoplastic drug which is
widely used in the treatment of many tumours.
Its side effects include nausea and vomiting,
neurotoxicity, nephrotoxicity and myelotoxicity;
cisplatin ototoxicity is also well documented,
although its exact mechanism is still unknown
(Helson et al., 1978). The suggested mechanisms
of cisplatin-mediated ototoxicity are probably
multifactorial, and it is postulated that this
toxicity may be related to cisplatin-induced
inhibition of Na‡- K‡- ATPase in the outer
hair cells of the cochlea (Wright & Schaefer,
1982; Schweitzer, 1993). Carboplatin is a
second-generation platinum compound which
is less nephrotoxic and neurotoxic than cisplatin.
It was originally thought that carboplatin did not
have the ototoxic potential of cisplatin, while
recent prospective trials have reported a 0% to
19% incidence of hearing loss from carboplatin
(Castello et al., 1990; Kennedy et al., 1990).
In most studies with audiologic monitoring,
the clinical evidence of ototoxicity from
 1999 Scandinavian University Press
standard doses of this drug up to 120 mg/m2
per course (Aguilar-Markulis et al., 1981;
Gratton 1990). This great variability in the
incidence of ototoxicity can be influenced by
mode of drug administration, cumulative dose,
total dose, age and prior otologic disease
(Reddel et al., 1982). Most studies on cisplatin
ototoxicity are based on serial audiometry for
frequencies up to 8,000 Hz, while other authors
have tried to compare auditory brainstem
response (ABR) versus conventional or high
frequency audiometry (up to 20,000 Hz) for
detecting hearing loss from antineoplastic
agents (Rybak, 1981; van der Hulst et al.,
1988; Fausti et al., 1992). However, because of
the various definitions used for ototoxicity and
the lack of uniformity with regard to dose
schedules, it is not fully defined whether ABR
is useful in early detecting and monitoring the
hearing loss from ototoxic drugs such as
cisplatin.
The purpose of this study was to evaluate the
effectiveness of ABR in identifying the early
140
hearing damages in cancer patients receiving
cisplatin or carboplatin.
Material and methods
The ototoxicity study was performed on bladder
cancer patients randomized to receive either a
cisplatin-based chemotherapy (M-VEC: metho-
trexate, vinblastine, epirubicin and cisplatin) or a
carboplatin-based chemotherapy (M-VECa:
methotrexate, vinblastine, epirubicin and carbo-
platin).
The eligibility criteria included a histologi-
cally proven diagnosis of recurrent or metastatic
bladder cancer, an Eastern Cooperative Oncol-
ogy Group (ECOG) performance status of 2 or
less (Oken et al., 1982); an age of 72 years or
younger; an absolute neutrophil count (ANC) of
1500/mm3 or more; a normal platelet count
(>100.00/mm3); a serum creatinine level of
1.6 mg/dl or less and a creatinine clearance of
>65 mL/min; a serum bilirubin level of 1.5 mg/
dl or less; a haemoglobin level of 11.0 g/dl or
more; a normal left ventricular ejection fraction
(>50%); no previous systemic therapy for
recurrent or metastatic disease; no prior radio-
therapy to the head and neck. Bladder cancer
patients who had previously received adjuvant
chemotherapy that had been discontinued at least
1 year before entering the study and patients
receiving adjuvant chemotherapy after cystect-
omy were also eligible. Adjuvant chemotherapy
was administered in order to reduce the risk of
recurrent disease after surgery.
The combination chemotherapy consisted of
cisplatin (70 mg/m2 intravenous [i.v.] by 1 h
infusion on day 2) in the M-VEC arm and
carboplatin (250 mg/m2 i.v. by 1-h infusion on
day 1) in the M-VECa arm, plus methotrexate
i.v., vinblastine i.v. and epirubicin i.v. at the
same doses in the same days in both treatment
arms. The chemotherapy cycles were scheduled
at 28-day intervals. The patients received at least
1 litre of 0.9% normal saline solution and
mannitol diuresis during cisplatin administration
in order to protect against cisplatin-induced
nephrotoxicity. Chemotherapy was not adminis-
tered as scheduled if any haematologic toxicity
occurred. Cisplatin, carboplatin, epirubicin and
vinblastine were reduced by 50% if the ANC
was less than 1500/mm3, if the leukocyte count
was less than 2500/mm3, or if the platelet count
was less than 75000/mm3. Cisplatin and carbo-
⋅ ⋅ ⋅
Scandinavian Audiology Vol 28 No 3 1999
A De Lauretis et al.
platin were also reduced by 50% if the glomer-
ular filtration rate was less than 60 mL/min.
Aminoglycosides or loop diuretics were not
administered. Withdrawal from the study was
based predominantly on failure to control dis-
ease progression or life-threatening toxicity.
The development of ototoxicity was evaluated
by measuring ABR (Jewett & Williston, 1971).
All patients with pre-existing diseases of the
acoustic system identified by means of conven-
tional pure-tone audiometry and impedance (125
to 8,000 Hz) were excluded from the ABR
studies in order to evaluate the risk of develop-
ment of ototoxicity in a homogeneous group of
normal hearing subjects. The eligibility criteria
were normal hearing thresholds of 25 dB or less
at each frequency between 125 and 8,000 Hz, a
normal tympanogram, normal stapedius reflex
threshold (both ipsilateral and contralateral) and
normal ABR. Normal tympanograms and nor-
mal stapedius reflex thresholds were defined in
agreement with Jerger’s classification and with
Metz’s test, respectively (Jerger, 1975; Metz,
1946). The ABR examinations were recorded in
a sound-treated room using 2000 clicks of
alternating polarity at constant intensity of 100
dB PESPL presented to the patients through 49-
TDH earphone at a rate of 21 clicks per second.
The skin was cleaned with acetone before
application of the electrodes in order to keep
the impedance of their connections below 5
Kohm. The high pass filter was 200 Hz and the
low pass filter 2 kHz. Electrodes were placed on
the vertex and the ipsilateral and contralateral
mastoid processes (the latter acting as the ground
electrode). Vertex-positive ABR waves were
numbered from I to V; the latency waves I, III,
V, and the interpeak latency interval (IPLI) I–III,
III–V, and I–V were considered for the analysis.
The tests were performed using an Amplaid
MK15 with a preamplifier; two tests were
performed on each session in order to ascertain
the reproducibility of the ABR. The ABR
evaluations were repeated after two cycles of
chemotherapy (1 week after the last cisplatin or
carboplatin administration), and then every two
cycles or when the patient was withdrawn from
the study. Normal mean latency values and 2
standard deviations (SDs) were Wave I,
1.8  0.8 msec; Wave III, 3.9  0.7 msec; Wave
V, 5.7  0.8 msec. The normal ABR latency
values were determined on the basis of a group
of 84 neurologically intact persons who were
ABR evaluation of ototoxicity
Table I. Characteristics of 21 bladder cancer patients treated with a cisplatin-based regimen or a carboplatin-
based regimen
Patient characteristics
Evaluated patients
Sex
Male
Female
Age (median), (years)
Range (years)
Performance status (ECOG)
1
2
Adjuvant chemotherapy
Chemotherapy for advanced disease
monitored by ABR in our Institute in a period
spanning 1 year. A latency shift of 0.4 msec plus
2 SDs from the mean in an individual patient was
considered significant. ABR recording was
performed by one audiometrician who was not
aware of the detailed status of the patients
regarding chemotherapy. Interpretation of the
ABR wave latency and amplitude was done on-
line.
Pure-tone thresholds from 125 Hz to 8,000 Hz
were determined using an Amplifon A309
audiometer, and impedance was measured with
an Amplaid 720 impedance audiometer. Audio-
grams were evaluated before treatment and then
every two cycles of chemotherapy. A threshold
shift was defined as a decrease of greater than 10
dB at one or more frequencies (125, 250, 500,
1000, 1500, 2000, 4000, 6000, 8000 Hz).
Statistical methods
The mean latencies of two replications of the
ABR waves during each investigation were used
for statistical analysis. Statistical significance
between post-treatment and baseline ABR re-
cordings was performed using the t-test for
paired data.
Results
A total of 59 patients entered the study: 31 were
randomly allocated to the cisplatin-based regi-
men and 28 to the carboplatin-based regimen
arm. Most of these patients took part in a
randomized study for treatment of advanced
bladder cancer, and a few in a randomized study
of adjuvant treatment of resected bladder cancer
(Table I).
141
Cisplatin Carboplatin
9 12
7 10
2 2
64 62
47–70 51–71
6 7
3 5
2
7 12
Of the 59 patients in the study, 21 (9 in the
cisplatin group and 12 in the carboplatin group)
were included in the ototoxicity study (36 had a
pre-existing hearing loss, and 4 of the 23 patients
initially with normal hearing were excluded
because they received only one cycle of che-
motherapy).
Of these 21 patients, 1 in the cisplatin group
and 3 in the carboplatin group had not received
full-dose cisplatin or carboplatin during the
second cycle because of myelotoxicity.
Table II gives the ABR changes in mean wave
latency and IPLI: at baseline there was no
significant statistical difference in mean wave
V latency between cisplatin and carboplatin
patients (p = 0.46). A significant shift in the
latencies of ABR waves was observed in 2 of the
9 cisplatin patients examined after 2 cycles of
chemotherapy at a cumulative cisplatin dose of
240 mg and 220 mg, respectively; the latency of
wave V of the ABR increased from 5.874 to
6.336 msec and from 5.826 to 6.458 msec in the
two patients, and the differences in I–V IPLI
compared to the baseline value were 0.502 msec
and 0.512 msec. After another two cycles of
chemotherapy, latency of wave V of the ABR
remained prolonged in these two patients, both
of whom were treated with chemotherapy for
advanced disease. No significant ABR impair-
ment was observed in the other seven cisplatin
patients during chemotherapy.
Thresholds at pure-tone audiometry were
reported as the average measurements for right
and left ears. No hearing loss greater than 10 dB
at one or more frequencies was detected in any
cisplatin patient after two or four cycles of
chemotherapy. One of the two patients who
showed an early impairment of ABR received a
⋅ ⋅ ⋅
Scandinavian Audiology Vol 28 No 3 1999
142
Table II. Mean  standard deviation of latencies in milliseconds of ABR waves and interpeak latency interval
in 21 bladder cancer patients treated with a cisplatin based regimen or a carboplatin based regimen
Cisplatin (9 pts)
Baseline 2 cycles
Wave I 1.79  0.08 1.85  0.09
Wave III 3.89  0.22 3.98  0.25
Wave V 5.89  0.16 6.04  0.29
I–III (IPLI) 2.09  0.22 1.99  0.20
III–V (IPLI) 2.00  0.19 2.20  0.25
I–V (IPLI) 4.12  0.13 4.18  0.24
total of six cycles of chemotherapy, and at that
time the pure-tone audiometry revealed changes
in hearing threshold levels of 20 dB at 8,000 Hz,
15 dB at 6,000 Hz, and 10 dB at 4,000 Hz, which
remained constant after another 3 months. The
other patient with an early impairment of ABR
received five cycles of chemotherapy, following
which the treatment was stopped for progression
of the disease; at that time, pure-tone audiometry
demonstrated changes in hearing threshold
levels of 25 dB at 8,000 Hz, 20 dB at 6,000 Hz
and 15 dB at 4,000 Hz. No progressive hearing
losses were observed after 1 month. This patient
died from the cancer 2 months after the disease
progression. No significant hearing loss at any
frequency was observed in the other seven
cisplatin patients.
None of the 12 examined carboplatin patients
developed ABR-measured ototoxicity or abnor-
mal audiograms during treatment, even in the
patients receiving nine cycles of chemotherapy.
Discussion
Cisplatin ototoxicity is well known, and the
clinical manifestations of the ototoxic damage
vary from a high frequency hearing loss to a
severe hearing loss affecting communication.
Ototoxicity is not a life-threatening complica-
tion, but it may cause a notable deterioration in
cancer patient life quality. The early detection of
ototoxic effects leads to the identification of
patients at risk for a further deterioration in
hearing function and, in this respect, it is
possible to change a cisplatin treatment into a
different dose schedule or into a less ototoxic
agent with an equal antitumour activity.
Some authors reported that ABR can detect
subclinical prolonged wave V and waves I–V
IPLI in patients receiving chemotherapy and
radiotherapy for head and neck cancer, in cancer
⋅ ⋅ ⋅
Scandinavian Audiology Vol 28 No 3 1999
A De Lauretis et al.
Carboplatin (12 pts)
P value Baseline 2 cycles P value
n.s. 1.75  0.10 1.76  0.11 n.s.
n.s. 3.86  0.17 3.88  0.11 n.s.
n.s. 5.78  0.13 5.80  0.15 n.s.
n.s. 2.11  0.18 2.12  0.19 n.s.
n.s. 1.91  0.16 1.91  0.13 n.s.
n.s. 4.03  0.15 4.04  0.17 n.s.
patients receiving intra-arterial cisplatin, and in
patients receiving treatment with ototoxic
agents, indicating some type of disturbance in
brain-stem auditory pathway (Kingston et al.,
1986; Lau et al., 1992; Maiese et al., 1992). In
our study the ABR method was used to evaluate
its effectiveness in detecting early ototoxicity
from cisplatin and its analogue carboplatin:
ototoxicity measured by means of ABR changes
was observed in two of the nine examined
cisplatin patients after two cycles of chemother-
apy. These two patients had not received any
other ototoxic agents, and the other administered
cytotoxic drugs, i.e. vinblastine, methotrexate
and epirubicin, are not known to provoke a
hearing deterioration. Moreover, computed to-
mography excluded the presence of brain me-
tastases in these two patients. Therefore the
significant prolongation of wave V latency and
I–V interpeak latency interval in these two cases
indicated early and clinically occult ototoxicity
due to cisplatin administration. These results
suggest that ABR might prove to be useful in
detecting early ototoxicity due to cisplatin. In
this study the ABR data showed a change before
the conventional audiometry (125 to 8,000 Hz)
in two patients; indeed these patients had a
hearing loss detected by audiometry after five
and six cycles of chemotherapy, respectively.
However, the sample size was too small to make
any sweeping conclusions about a major sensi-
tivity of ABR over conventional audiometry.
Other authors have reported that high frequency
audiometry might prove to be superior to
conventional audiometry or ABR in detecting
ototoxicity, while high-frequency, tone-burst-
evoked ABRs seemed to be significantly more
effective than click-evoked ABRs in the early
detection of ototoxicity from ototoxic agents
(Coupland et al., 1991; Fausti et al., 1992). ABR
testing was also used in young children, but it
ABR evaluation of ototoxicity
seemed to be less sensitive than pure-tone
threshold audiometry to the early effects of
cisplatin, above all because of the relatively
high incidence of recurrent middle ear disease in
this age group (Weatherly et al., 1991). Never-
theless, a low level of ototoxicity was encoun-
tered in our patients and it was difficult to
compare the ABR sensitivity over audiometry in
detecting and monitoring ototoxic deterioration.
Indeed the moderate cisplatin dose (70 mg/m2 by
bolus administration) repeated at least for two
cycles caused hearing loss in only two out of
nine patients examined (22%), and no further
hearing deterioration was observed in patients
receiving up to six cycles of chemotherapy for a
cumulative cisplatin dose level of 420 mg/m2.
These findings are in line with other studies
which report only mild ototoxic changes in
patients receiving cisplatin 50 mg/m2 for a
cumulative dose up to 400 mg, and a high
incidence of ototoxicity with cisplatin doses of
100–120 mg/m2 (Vermorken et al., 1983; Laur-
ell & Jungnelius, 1990).
With regard to carboplatin, the absence of any
significant ABR changes in our 12 examined
patients may confirm the apparent lack of
ototoxicity associated with moderate doses of
this agent (250 mg/m2) (Delb et al., 1993). On
our findings, it seems that there is no significant
difference in the development of ototoxicity
between patients receiving a low-to-moderate
cisplatin dose and patients receiving a conven-
tional carboplatin dose.
Conclusion
ABR is an easily available and non-invasive
procedure which seems to be useful in detecting
early hearing deterioration in cancer patients
receiving cisplatin. However, because of the
small size of the study and the low level of
otoxicity encountered in our patients, further
investigations are needed before the use of ABR
can be advocated as a method of audiologic
monitoring in patients receiving ototoxic drugs.
References
Aguilar-Markulis NV, Beckley S, Priore R, Mettlin C.
Auditory toxicity effects of long-term cis-dichlorodiam-
mine platinum II therapy in genito urinary cancer
patients. J Surg Oncol 1981;16:111–23.
Castello MA, Clerico A, Deb G, Dominici C, Fidani P,
Donfrancesco A. High dose carboplatin in combination
143
with etoposide (JET regimen) for childhood brain
tumors. Am J Pediatr Hematol Oncol 1990;12:297–300.
Coupland SG, Ponton CW, Eggermont JJ, Bowen TJ, Grant
RM. Assessment of cisplatin-induced ototoxicity using
derived-band ABRs. Int J Pediatr Otorhinolaryngol
1991;22:237–48.
Delb W, Feilen S, Koch A, Federspil P. Experimental
investigations on the ototoxicity of cisplatin and carbo-
platin. Laryngo-Rhino-Otol 1993;72/1:24–7.
Fausti SA, Frey HR, Henry JA, Olson DJ, Schaffer HI.
Early detection of ototoxicity using high-frequency,
tone-burst-evoked auditory brainstem responses. J Am
Acad Audiol 1992;3:397–404.
Gratton MA. The interaction of cisplatin and noise on the
peripheral auditory system. Hear Res 1990;50:211–224.
Helson L, Okonkwo E, Anton L, Cvitkovic E. Cis-platinum
ototoxicity. Clin Toxicol 1978;13:469–79.
Jerger J. Diagnostic use of impedance measures. In:
Handbook of clinical impedance audiometry. Dobbs
Ferry, N.Y.: Morgan Press, 1975;149–74.
Jewett DL, Williston JS. Auditory evoked far fields
averaged from the scalp of humans. Brain 1971;94:
681–96.
Kennedy IC, Fitzharris BM, Colls BM, Atkinson CH.
Carboplatin is ototoxic. Cancer Chemother Pharmacol
1990;26:232–4.
Kingston JE, Abramovich S, Billings RJ, Malpas JS, Fuller
AP. Assessment of the effect of chemotherapy and
radiotherapy on the auditory function of children with
cancer. Clin Otolaryngol 1986;11:403–9.
Lau SK, Wei WI, Sham JS, Choy DTK, Hui Y. Early
changes of auditory brain stem evoked response after
radiotherapy for nasopharyngeal carcinoma. A prospec-
tive study. J Laryngol Otol 1992;106:887–92.
Laurell G, Jungnelius U. High-dose cisplatin treatment:
hearing loss and plasma concentrations. Laryngoscope
1990;100:724–34.
Maiese K, Walker RW, Gargan R, Victor JD. Intra-arterial
cisplatin-associated optic and otic toxicity. Arch Neurol
1992;49:83–6.
Metz O. The acoustic impedance measured on normal and
pathological ears. Acta Otolaryngol (Stockh) 1946;
(Suppl 63):397–405.
Oken MM, Creech RH, Tormey DC. Toxicity and response
criteria of the Eastern Cooperative Oncology Group. Am
J Clin Oncol 1982;5:649–55.
Reddel RR, Kefford RF, Grant JM, Coates AS, Fox RM,
Tattersal HN. Ototoxicity in patients receiving cisplatin:
importance of dose and method of drug administration.
Cancer Treat Rep 1982;66:19–23.
Rybak LP. Cisplatinum associated hearing loss. J Laryngol
Otol 1981;95:745–7.
Schweitzer VG. Cisplatin-induced ototoxicity: effect of
pigmentation and inhibitory agents. Laryngoscope 1993;
103 (Suppl 59):1–52.
van der Hulst RJAM, Dreschler WA, Urbanus NAM. High
frequency audiometry in prospective clinical research of
ototoxicity due to platinum derivatives. Ann Otol Rhinol
Laryngol 1988;97:133–7.
Vermorken JB, Kapteijn TS, Hart AAM, Pinedo HM.
Ototoxicity of cis-diammine dichloroplatinum (II).
Influence of dose, schedule and mode of administration.
Eur J Cancer Clin Oncol 1983;19:53–8.
Weatherly RA, Owen JJ, Catlin FI, Mahoney DH.
Cisplatinum ototoxicity in children. Laryngoscope 1991;
101:917–24.
Wright CG, Schaefer SD. Inner ear histopathology in
patients treated with cis-platinum. Laryngoscope 1982;
92:1408–13.
⋅ ⋅ ⋅
Scandinavian Audiology Vol 28 No 3 1999
Copyright of Scandinavian Audiology is the property of Scandinavian University Press
International Division and its content may not be copied or emailed to multiple sites or posted
to a listserv without the copyright holder's express written permission. However, users may
print, download, or email articles for individual use.