High incidence of cisplatin-induced ototoxicity in paediatric patients in the Western Cape , South Africa

Cancer is the second most commonly diagnosed non-communicable disease and one of the leading causes of death globally. It accounted for an estimated 8 million deaths globally in 2015.1 Approximately 14 million new cancer cases are reported annually and up to 10% affects children below the age of 15 years.1 In South Africa, over 100 000 individuals are diagnosed with cancer every year.2 Specific to paediatrics, between 33.4 and 47.2 per million children were diagnosed with cancer between 2003 and 2007, while an estimated 62.6 and 87.8 per million children were diagnosed with cancer in the Western Cape and Free State provinces, respectively.2 With regard to the survival rate, South African children were reported to have a 51% survival rate for childhood cancer in 2014.2


Introduction
Cancer is the second most commonly diagnosed non-communicable disease and one of the leading causes of death globally.It accounted for an estimated 8 million deaths globally in 2015. 1 Approximately 14 million new cancer cases are reported annually and up to 10% affects children below the age of 15 years. 1 In South Africa, over 100 000 individuals are diagnosed with cancer every year. 2 Specific to paediatrics, between 33.4 and 47.2 per million children were diagnosed with cancer between 2003 and 2007, while an estimated 62.6 and 87.8 per million children were diagnosed with cancer in the Western Cape and Free State provinces, respectively. 2With regard to the survival rate, South African children were reported to have a 51% survival rate for childhood cancer in 2014. 2 Cisplatin-based chemotherapy is the common anti-cancer treatment for both children and adults because of its high success rate, lower cost and availability. 3Cisplatin, a platinum-based chemotherapeutic, has been used effectively in the treatment of various soft tissue cancers for over 30 years. 3 Unfortunately, it is also associated with high incidences of ototoxicity.Ototoxicity refers to drug or chemical-induced damage to the structures of the inner ear. 4,5The incidence of cisplatin-induced ototoxicity in paediatric patients varies considerably, ranging from 13% to 96%. 4,5This variability could be attributed to the use of different diagnostic criteria for ototoxicity and/or use of different ototoxicity grading scales and use of different audiometric tests to diagnose ototoxicity-induced ototoxicity.
Cisplatin-induced ototoxicity is characterised by symmetrical sensorineural hearing loss accompanied by tinnitus, otalgia, poor speech discrimination and aural fullness. 4,6Furthermore, the hearing loss is permanent and usually starts in the higher frequencies on the audiogram before progressing to the lower frequency range. 5,6,7Hearing loss during childhood can leave debilitating effects on speech-language acquisition, socioemotional development, and reading and writing abilities.Furthermore, childhood hearing loss can lead to poor scholastic performance, academic achievement and overall quality of life, for both the child and his or her family. 8It is, therefore, imperative that cisplatin-induced hearing loss should be prevented, especially in this patient group.
One of the ways to prevent or minimise the occurrence of treatment-induced hearing loss in children during cisplatinbased chemotherapy is to closely monitor their hearing status during treatment (i.e.ototoxicity monitoring).Ototoxicity monitoring involves prospective collection of serial audiometric data at regular intervals to ensure early detection of changes in hearing thresholds presumably attributed to treatment regimen. 9If a deterioration in patient's hearing thresholds is detected in time, the oncologist has an option to explore alteration to the treatment to avoid more serious hearing loss, for instance discontinuing the use of cisplatin and replacing it with a less ototoxic alternative (e.g.carboplatin). 9splatin-induced ototoxic damage typically starts at the basal end of the cochlear, the region of the inner ear that is responsible for high-frequency encoding. 10The continued use of cisplatin may lead to the damage extending to the apical end of the cochlea, where the low-frequency information is encoded. 10Therefore, ototoxicity monitoring protocols must include audiological tests that are sensitive to changes in high-frequency thresholds (i.e.test frequencies > 8 kHz) or otoacoustic emissions to ensure early detection of changes in patients' hearing thresholds. 9Studies that investigated the incidence of cisplatin-induced ototoxicity in paediatric oncology patients are scarce in the African continent.This study therefore aims to address this information gap and will determine the incidence of cisplatininduced ototoxicity, as well as document factors associated with the risk for developing cisplatin-induced ototoxicity in a paediatric population in South Africa.

Methods
This was a retrospective records review of consecutive paediatric patients (< 15 years old) who underwent cisplatinbased chemotherapy and ototoxicity monitoring at Red Cross War Memorial Children's Hospital from 01 January 2016 to 31 December 2017.Patients' records were selected and included in this study if they had a baseline audiogram showing normal hearing thresholds, at least one monitoring audiogram, and the patient was aged between 5 and 18 years old.Records of patients who received prior radiation therapy in the head and neck region and those who were previously treated with ototoxic medications were excluded from the study.Primary endpoints were hearing loss, distortion product otoacoustic emission (DPOAE) findings expressed as a 'pass' or 'refer' (where a pass indicates normal cochlear functioning and, a refer indicates cochlear pathology), presence of tinnitus, vestibular dysfunction and reports of otalgia.Secondary outcomes were to identify treatment and patient factors associated with ototoxicity following cisplatin chemotherapy.
Data abstracted from patients' records were captured using an online password-protected form designed with Google Forms and subsequently transferred onto an Excel spreadsheet.Limiting access and using proxy patient identifiers maintained patient confidentiality.The following variables were captured: age, sex, ethnicity, type of cancer, cumulative cisplatin dose received, treatment duration, pure tone audiometric and extended high-frequency audiometry results, DPOAE results, otoscopic examination results and tinnitus reports.The American Speech-Language-Hearing Association's (ASHA) 11 significant threshold shift (STS) criteria were used to determine the presence or absence of ototoxicity-induced hearing loss.The grading of the ototoxicity was done according to the International Society of Paediatric Oncology Boston (SIOP) Scale. 12atistical analysis was performed using the IBM SPSS Statistics version 24 for Macintosh software.Categorical variables were expressed as frequencies and percentages.Logistic regression was used to determine associations between patient and treatment variables and the likelihood of developing ototoxicity during chemotherapy treatment.

Ethical considerations
Ethical approval for the study was sought and obtained from the University of Cape Town's Faculty of Health Sciences Human Research Ethics Committee (HREC REF: 465/2017).

Results
A total of 122 records of paediatric patients who underwent cisplatin-based chemotherapy at Red Cross War Memorial Children's Hospital during the study period were accessed for review.Fifty-eight records met the inclusion criteria and thus were selected for inclusion in this study.However, nine of the records had missing information and therefore were also excluded.In the end, 49 records were included in the final review (see Table 1 for a description of patients' characteristics).Bilateral sensorineural hearing loss was observed in 80% (n = 39) of the patients when utilising extended highfrequency audiometry (9 kHz -12 kHz) in comparison to 49% (n = 24) when conventional audiometry (0.25 kHz -8 kHz) was used.Consistent with pure tone audiometry, DPOAEs were absent in 32 (67%) (n = 32) participants.There was also a strong positive correlation (Pearson's r = 0.899) between DPOAE and extended high-frequency pure tone audiometry.The majority of the participants (56%) presented with the SIOP grades 2-4 ototoxic hearing loss (see Figure 1).Six (10%) participants reported high-pitched tinnitus.Other observed auditory pathologies in the sample included cerumen impaction, which was observed in 29 (60%) participants at the end of chemotherapy.
Further analysis of the results using logistic regression analysis showed that a higher cumulative dose (> 200 mg/m 2 ) was associated with a higher incidence of hearing loss during cisplatin-based chemotherapy (odds ratio [OR]: 1.81; 95% confidence interval [CI]: 0.67-17.34;p = 0.044).With respect to age, in comparison to patients older than 10 years, those below the age of 10 years were four times more likely to develop hearing loss; however, this was not statistically significant (OR: 4.00, 95% CI: 0.82-19.46;p = 0.085).Sex and treatment duration also showed higher odds of developing hearing loss, but these associations were not statistically significant (see Table 2).

Discussion
This was one of the first studies to document the incidence of cisplatin-induced ototoxicity in paediatric oncology in South Africa.The findings of this study indicated that a high proportion (80%) of paediatric patients who underwent cisplatin chemotherapy at Red Cross War Memorial Children's Hospital during January 2016 -December 2017 developed hearing loss following cisplatin-based chemotherapy.The incidence of hearing loss following cisplatin chemotherapy in the paediatric population reported in previous studies is known to vary considerably (42% 13 to 94% 14,15 ).However, despite this variability, these studies consistently reported a high incidence of cisplatin-induced hearing loss, especially if more sensitive tests of detecting changes in patient's hearing thresholds are used as was the case in the current study.About 55% of the patients in this study developed moderateto-severe hearing loss, which is classified as disabling 16 because of its known negative impact on function and quality of life.Disabling hearing loss 16 can have a debilitating impact, particularly in this study's age group, in terms of difficulty in understanding speech, psychosocial development, reading and writing abilities, scholastic performance and academic achievement and quality of life, for both the patients and their families. 16,17,18splatin-induced ototoxicity is believed to be related to the formation of reactive-oxygen-species (ROS) and depletion of anti-oxidant scavenger molecules, subsequently inducing calcium influx and cell apoptosis. 19Yancey et al. 13 further illustrated that one target of ROS-induced cochlear damage is the outer hair cells (OHCs), which may lead to sensory hearing loss.Therefore, an abnormality of DPAOE, a test that is known to be sensitive to changes in OHCs function, seems to confirm this as a possible site of lesion.
A strong positive correlation was found between DPOAEs and extended high-frequency audiometry results with respect to early detection in patients' auditory status.These findings therefore suggest that in cases where it is not possible to perform behavioural audiometric assessments, such as extended high-frequency audiometry, DPOAEs can be used as an alternative test. 20,21This is especially important in paediatric oncology where obtaining reliable extended    high-frequency audiometry results can be challenging. 15ehavioural audiometric tests, such as extended highfrequency audiometry, require a patient to be fully awake and concentrating for the full duration of the test.This may be difficult in children who are undergoing cancer treatment because they can get tired quickly and lose their concentration, thus resulting in unreliable results. 21th respect to diagnostic protocols, 80% of the patients were diagnosed with ototoxicity when utilising extended highfrequency audiometry (9 kHz -16 kHz), when compared to only 49% when using conventional audiometry (0.25 kHz -80 kHz).This shows that extended high-frequency audiometry was more sensitive in terms of early identification of changes in hearing threshold than conventional audiometry.Several studies have also reported on the effectiveness of extended high-frequency audiometry and its superiority to conventional audiometry when it comes to early detection of ototoxicityinduced hearing loss. 10,14Owing to the demographic characteristics of this study's sample, particularly age, early identification of hearing loss would, therefore, be of paramount importance to overcome the negative impact of hearing loss.

Demographic characteristics
Treatment factors such as duration of cisplatin chemotherapy treatment and dose have been reported to be associated with higher likelihood of developing cisplatin-induced hearing loss during cisplatin chemotherapy. 6,13,14In this study, a higher cumulative dose (> 200 mg/m 2 ) was found to be associated with a high likelihood of developing hearing loss during cisplatin-based chemotherapy, which was consistent with the findings of previous studies.However, with regard to cumulative cisplatin dose, the present study found a relatively lower cumulative cisplatin dose (200 mg/m 2 ) to be associated with hearing loss, when compared to 400 mg/m 2 , which is reported by other scholars in the field. 13,22,23tient factors such as age and sex have also been identified as risk factors for developing hearing loss. 12Association between younger age and a higher likelihood of developing hearing loss was reported in previous studies. 15,23In the present study, children below the age of 10 years were more likely to develop hearing loss when compared to those above 10 years.However, this was not found to be statistically significant, possibly because of the small sample size of the current study.
Existing literature on sex as a risk factor for developing cisplatin-induced ototoxicity is variable; Yancey et al. 13 reported that males were four times more likely to develop ototoxicity, while Li et al. 24 indicated that female sex is associated with higher incidences of ototoxicity.The current study found that males were 34% (OR: 1.34) more likely to develop hearing loss following cisplatin chemotherapy when compared to females.However, this was not statistically significant (p = 0.47), possibly because of the small sample size of this study.
An unexpected finding in this study was the fact that a high proportion of patients developed impacted cerumen following cisplatin chemotherapy.All the participants had clear external auditory meatus with insignificant amounts of cerumen pre-cisplatin chemotherapy, whereas at the end of treatment almost half of the participants (49%) had cerumen impaction.This finding may seem to suggest that cisplatin may alter the cerumen production mechanism within the external auditory meatus.While literature linking cisplatin chemotherapy to increased cerumen production is limited, this finding of a high incidence of cerumen impaction in the study cohort is worth noting because cerumen impaction can contribute to an increase in the prevalence of hearing loss. 25his finding, therefore, highlights the need for routine otoscopic examinations in paediatric patients undergoing cisplatin chemotherapy to identify patients who might have cerumen impaction and need this to be removed.

Study limitations
The findings of this study must be interpreted with caution owing to its methodological limitations -a retrospective record review with a relatively small sample size.However, despite these limitations, this was one of the first studies to determine the incidence of cisplatin-induced ototoxicity in a paediatric oncology population in South Africa.We believe that these findings will prompt action from clinicians managing these patients to try to put appropriate interventions in place to prevent hearing loss in children who undergo cisplatin chemotherapy.

Conclusion
This study showed that a high proportion (80%) of patients who underwent cisplatin-based chemotherapy at Red Cross War Memorial Children's Hospital during the study duration ended up developing ototoxic hearing loss.A higher cumulative dose (> 200 mg/m 2 ) and younger age (< 10 years old) were associated with an increased likelihood of developing hearing loss following cisplatinbased chemotherapy.Given the high proportion of patients in this study who developed hearing loss following cisplatin-based chemotherapy, it is important that such patients are closely monitored (ototoxicity monitoring) to enable early detection of cisplatin-induced ototoxicity.Extended high-frequency audiometry and DPOAE compared to conventional audiometry were found to be more successful in detecting changes in patients' hearing thresholds following treatment.It is also important that routine otoscopic examinations should be part of the ototoxicity monitoring protocol.

FIGURE 1 :
FIGURE 1: Hearing loss grading using the International Society of Paediatric Oncology (SIOP) Boston ototoxicity scale.
of the records did not have the required hearing loss information and were therefore excluded from the final review.CNS, central nervous system.