The primary objective of the study is to investigate the prevalence of hearing loss after cancer treatment in a prospective cohort of pediatric CNS and solid tumor patients. Secondary objectives focus on examining the prevalence of tinnitus and vertigo after cancer treatment, identifying clinical determinants for ototoxicity, and investigating the feasibility of testing patients at risk for ototoxicity with standardized audiological examinations.

We will perform a national prospective cohort study in the Netherlands. This study is embedded in the Princess Máxima Center, a national health care center that has been set up to treat all children diagnosed with cancer. Collaboration with the audiological department of Wilhelmina Children’s Hospital in the Netherlands has been established to perform audiological testing of children of all ages. The selection, invitation, and inclusion of patients takes place in the Princess Máxima Center. The audiological assessment is age-adjusted, according to recently published guidelines [24]. It will be performed before the start of treatment, during cancer treatment according to treatment protocols, and within 3 months after the end of treatment (Table 1). Information about the screening outcome of the neonatal hearing screening is retrieved for all patients [25].

Children aged 0 to 19 years who will be diagnosed with a CNS or solid tumor between January 2021 and January 2023 are eligible for participation in the study. They will be divided in two strata (Table 1). Stratum 1 consists of patients treated with platinum agents (cisplatin, carboplatin, or oxaliplatin), CNS/ear-nose-throat (ENT) irradiation, or CNS/ENT surgery, and audiological examinations are part of standard care. Stratum 2 consists of patients not treated with therapies described in stratum 1 or any patients in which audiological examinations are not part of standard care. The independent contribution of any comedication on ototoxicity development will be studied in stratum 2.

Pediatric oncologists from the Princess Máxima Center will select patients for invitation. For patients in whom audiological examinations are not standard of care, written informed consent will be obtained from parents/guardians (patients) according to good clinical practice regulations. Withdrawal is possible at any time during the study without providing a reason.

All patients will be examined by standard audiological examinations (Table 2), which always includes otoscopic inspection of the ears and tympanometry. Depending on the age of the child, (a combination of) brainstem-evoked response audiometry (BERA), distortion product otoacoustic emissions (DPOAEs), visual reinforcement audiometry (VRA), conditioned play audiometry, and extended high-frequency pure tone audiometry (PTA) will be applied. Anamnestic screening for tinnitus and vertigo will be performed during audiological examinations as a standard procedure for patients 8 years and older and 10 years and older, respectively [24,26,27]. We will schedule at least two visits during the study period, including a baseline visit before the start of cancer treatment to exclude pre-existent hearing loss and a follow-up visit within 3 months after the end of cancer treatment.

Otoscopy is used for inspection of the external auditory canal and tympanic membrane, which focuses on accumulation of cerumen, infections, and perforations of the tympanic membrane [28,29]. Tympanometry is routinely applied as an indicator of conductive hearing loss due to middle ear pathology. Static air pressure is varied systematically in the ear canal while an acoustic stimulus of 226 or 1000 Hz is delivered. This procedure measures the ability of the middle ear system to transfer low-frequency acoustic energy to the cochlea as a function of static air canal pressure [30]. Abnormal functioning of the middle ear due to a thickened tympanic membrane; presence of middle ear fluid, for example, caused by acute otitis media; or dysfunction of the Eustachian tube can be detected [31].

BERA will be performed for objectively measuring hearing thresholds to quantify the type and severity of hearing loss [32]. Electric activity of the VIII nerve and its central connections are evoked by a broadband click stimulus or frequency-specific stimulus through a headphone, insert phone, or bone conductor. The responses can be measured by electrodes placed on the scalp and is represented as a registration of measured electrical potentials as a function of time after stimulus presentation, characterized by reproducible peaks at specific time delays (latencies). No active cooperation is required, but a prerequisite is that the patient is asleep or lies in a relaxed position to avoid artifacts due to muscle activity.

Evoked otoacoustic emissions (OAEs) are sounds generated within the inner ear after presentation of a stimulus. OAEs are related to the nonlinear behavior of the cochlea, often referred to as the cochlear amplifier. A clear relationship exists between the presence of normal OAEs and functional outer hair cells, which implies that OAEs can only be evoked in ears with normal hearing thresholds or very mild hearing loss (<25 dB HL) [33,34]. Different types of OAEs exist, but often DPOAEs and transiently evoked OAEs (TEOAEs) are measured. TEOAEs are evoked by using a click stimulus and frequency range up to 4 kHz, in contrast to the DPOAEs that are evoked by using pairs of primary tones with particular intensity. The frequency range that can be monitored is much larger than with TEOAEs and ranges up to 10 kHz [31,33,34]. As ototoxicity will start in the high frequencies, measuring DPOAEs is preferred in addition to TEOAEs in our study.

VRA and conditioned play audiometry are both subjective tests based on a conditioning procedure. For VRA, the child is conditioned to make a head turn and to look at a “reward,” for example, a short movie or picture, every time a frequency-specific stimulus is presented [35,36]. For conditioned play audiometry, children are conditioned to perform a motivation enhancing task when an auditory stimulus is presented, for example, putting a block into a box [31,35]. Frequencies up to 8 kHz can be measured, but for ototoxicity monitoring, the range is preferably extended to the high frequencies up to 12 kHz, as in an early stage, ototoxicity may affect the high frequencies [37].

PTA is a subjective behavioral measurement of hearing thresholds. Conventional PTA measures the faintest tone a person can hear at selected frequencies. Hearing thresholds are obtained via air conduction (0.25-8 kHz) by using headphones and via bone conduction (0.5-4 kHz) by using a vibrating transducer. Additionally, extended high-frequency hearing thresholds may be determined up to 12 to 14 kHz, depending on the age of the child [31,34,38].

Anamnestic screening for tinnitus and vertigo will be performed during audiological assessments as a standardized procedure. Regarding tinnitus, children 8 years and older will be asked whether they hear noise in their ears. If the answer is yes, questions follow on the type of noise (ringing, buzzing, humming, whistling, sea noise, banging, blowing sound, like a machine, clicking, beep, like the wind, or like cars), the moment of onset (dates, after which chemotherapy course), laterality (left ear, right ear, or both), pitch (high vs low), perceived loudness (graded according to a Likert scale 0-5) at the peak moment, duration (intermittent, most, some of the time, or continuous), annoyance degree (always annoyed, seldom annoyed, very annoyed, little annoyed), and severity with regard to causing worry (not at all, slightly, sometimes, more severely) [26]. Regarding vertigo, children 10 years and older will be asked whether they ever get dizzy. If the answer is yes, questions follow on the moment of onset (dates, after which chemotherapy course), presence of light-headedness (yes/no), tendency to fall (yes/no), experiencing spinning or turning objects (yes/no), sensation of turning or spinning (yes/no), loss of balance when walking or running (yes/no), duration (intermittent, most/some of the time, or continuous), and severity (graded according to a Likert scale 0-5) [27].

In case of hearing loss at the end of cancer treatment, speech audiometry is recommended to test the ability to hear and understand speech. Several lists of words are offered to the child through headphones or a free field loudspeaker. The child is asked to repeat the words. The number of correctly repeated phonemes or words are recorded. The results are registered in a speech audiogram. The test reveals the amount of speech discrimination at various stimulus intensities and provides an indication on the influence of hearing loss on communication. Different test procedures are available, including determination of the ear’s ability to discriminate speech in silence or in different types of noise. Speech audiometry in children is often used as a diagnostic test. In addition, the test can be useful to evaluate the outcome of an intervention with hearing aids [31,34,39].

A power analysis based on two-sided CIs for one proportion has been performed. The Clopper-Pearson exact formula for computing binomial CIs was used. The method is based on the cumulative probabilities of the binomial distribution rather than an approximation [40,41]. A sample size of 367 patients produces a two-sided 95% CI with a width (distance between lower and the upper limit for the CI) equal to 1% when the sample size proportion is 35%. PASS software (NCSS, LLC) has been used for computing the sample size.

Continuous variables will be reported with medians and ranges while categorical variables with percentages. To assess differences between patients with and without hearing loss, tinnitus, or vertigo, chi-square and Student t tests will be used for categorical and continuous variables, respectively. In case of violation of normality, Mann-Whitney U tests will be used.

A logistic regression model will be estimated to investigate the effect of risk factors such as platinum use and cumulative dose, type of CNS/ENT surgery, type of CNS/ENT irradiation and dose, type of comedication and its dose and levels, age at diagnosis, and gender on ototoxicity at the end of treatment. To quantify the effect of prognostic factors on the risk of ototoxicity at the end of treatment, odds ratios along with 95% CIs will be estimated. Statistical analysis will be performed with SPSS, version 26.0.0.1. (IBM Corp) [44].

In the presence of missing data, imputing techniques are used [45].

The study protocol has been approved by the Clinical Research Committee of the Princess Máxima Center and by the Medical Research Ethics Committee Utrecht (Identifier 20-417/M), and has been registered in the Netherlands Trial Register (NL8881).

Ototoxicity is a serious adverse event of childhood cancer treatment. However, the prevalence of hearing loss, tinnitus, and vertigo during and shortly after treatment has never been studied in large prospective pediatric oncology patient cohorts. To date, the association of clinical characteristics and treatment components on the development of ototoxicity is not fully understood. Therefore, we will investigate the feasibility of testing all patients with a potential risk on ototoxicity with standardized audiological examinations. The prevalence and determinants of hearing loss, tinnitus, and vertigo will be investigated in a prospective cohort of childhood cancer patients. Accurate audiological testing and timely detection of ototoxicity will identify novel compounds at risk for ototoxicity and create awareness, which may lead to improved patient care and improved quality of life. Furthermore, we aim to collect high-quality data on the relation between disease status, type of treatment, and audiological functioning in children with solid and CNS tumors before and shortly after treatment. In the future, these data could be used to identify children who are at risk and might benefit from otoprotective agents that are currently under development.

First, it is important to realize that pediatric oncology patients can be critically ill at the time of presentation at the hospital. In that case, it may not be feasible to perform baseline audiological examination. For these selected children, we decided to use audiological anamnesis to exclude any type of pre-existent hearing loss. Second, it is possible that patients will receive medication (eg, antibiotics and diuretics) in shared care centers, which are not reported and may lead to underreporting comedication.