Discussion

Nowadays, audiological monitoring is strongly recommended for ototoxicity surveillance [32] in pediatric oncology, especially if treated with platinum-derived compounds. In fact, it has been recently demonstrated, that the overall incidence of hearing loss can be significantly decreased by auditory monitoring during treatment because of early identification of ototoxicity and therapy shift to carboplatin [15]. Furthermore, late onset or progression of hearing loss after the end of therapy has been described [7, 8, 29] even though few studies have been focused on long-term audiological monitoring. However, specific guidelines on long-term interventions in reducing late ototoxicity are still under discussion for health-care professionals in pediatric oncology.
Our aim was to provide evidence about the effectiveness of long-term surveillance in preventing platinum-induced hearing effects in children and adolescents surviving to cancer. In our experience, hearing status is currently measured after each cycle of cisplatin and adjustment of chemotherapy can be helped with suitable audiological counseling. When a child is diagnosed with grade > 2 ototoxicity, cisplatin administration is stopped and changed to carboplatin. This protocol is able to achieve hearing loss diagnosis at an early stage and to provide, jointly with pediatric oncologists, therapeutic strategies to reduce further progression of ototoxic side effect shifting by cisplatin to carboplatin [15].
In our series we confirmed that hearing loss may develop and progress in children during platinum-based therapy [29], however deafness may increase over the years after the end of therapy. Furthermore, we interestingly demonstrated the high risk to develop a late onset/progressive hearing loss at 5-years follow-up (SIOP criteria > 0) that occurred in 15/38 (39.5%) patients. These results are particularly important to identify cases requiring early hearing rehabilitative strategies to prevent harmful effect of hearing deprivation. In fact, we observed that 25% of monitored children needed hearing aid intervention.
How the ototoxicity appears or progresses over time is still unclear, nevertheless a prolonged maintenance of platinum in the body has been detected up to 20 years after treatment [33]. Furthermore, in experimental models, cisplatin has been demonstrated to target numerous cell types including outer hair cells and synapses between inner hair cells and terminal fibers of spiral ganglion neurons with a following neuronal degeneration that can appear at months after synaptic damage [10, 34, 35]. Thus, as in the noise induced hearing loss and aging [36], the cisplatin-induced synaptopathy and neural degeneration might explain the compromised speech understanding in noisy environments especially observed in children that may lead to difficulties in school performances and psychosocial functioning [37, 38].
Platinum-induced hearing loss is irreversible [2, 15, 29], in fact in our experience no improvement of hearing was found over time, as expected we only observed a worsening of hearing after platinum therapy that was even more common than a late onset of hearing loss (i.e 26.3% vs 13.1% respectively). Among children having a progressive hearing loss, a SIOP grade 4 was found in 8/10 (80%) indicating that the presence of hearing loss at the end of treatment was a negative prognostic factor for long-term ototoxicity outcomes (Odds Ratio: 7.2 – CI: 1.67–31.1; p<0.01). Furthermore, median time of the late onset and progression of ototoxicity was 24 and 18 months respectively by the end of treatment. The early onset of ototoxicity as detected by the auditory monitoring is suggestive for a worse audiological outcome over time likely depending by individual susceptibility that is still understudied. A genetic predisposition has been proposed based upon observations of substantial inter-individual variability in the prevalence and severity of ototoxicity including polymorphisms for genes involved in the oxidative stress, which is the major target of cisplatin ototoxicity [15, 39]. Unfortunately, a routinely analysis of genetic polymorphism is still lacking [40], therefore the long-term audiological monitoring remains a major target for prevention of ototoxicity.
Traditionally, cumulative doses, age at treatment, cranial irradiation and type of malignances have been considered as risk factors for platinum-induced ototoxicity in pediatric population [40, 41]. In contrast to previous findings [43], we found that the cumulative dose of cisplatin, affecting the early onset of hearing loss [15], did not correlate with long-term development of ototoxicity.
Remarkably, we found that the risk of hearing loss was high associated with cisplatin-based multimodal therapy that included carboplatin and/or radiotherapy rather than a cumulative dose of cisplatin alone that is currently suggested as indicator for hearing screening. Brain irradiation is then an independent factor for the occurrence of ototoxicity during platinum chemotherapy [42, 43]. In fact, most of children treated with carboplatin were those who developed ototoxicity during cisplatin treatment and then shifted to less ototoxic carboplatin; probably susceptibility to cisplatin damage remains major trigger for development of hearing loss. Furthermore, children who underwent carboplatin therapy alone developed a SIOP grade 4 when subjected to prior or concurrent cranial irradiation for brain cancers. Therefore, irradiation remains even in our experience a prognostic factor for long-term ototoxicity outcomes (Odds Ratio of 5.25 – CI: 1.26–21.86; p < 0.01).
Hence, in the pre-treatment counseling, it is important for pediatric oncologist to inform families on the possibility to change chemotherapy avoiding ototoxicity during treatment and indication for a long-term audiologic follow-up, especially in children affected by brain cancers treated with radiotherapy. In these cases, whether shifting by cisplatin to carboplatin the overall survival can be affected is a controversial issue that needs to be further discussed with family [1].
We found a worse hearing in older children, conversely in literature a higher ototoxicity is reported in younger age [4, 32, 41, 43]. In our study, although the limitation in size sample, older children were affected by medulloblastoma, which is even a rare cancer, and experienced a worse hearing loss. This could be explained as the most of them was treated with both platinum compound and irradiation therapy. While, younger children affected by neuroblastoma and retinoblastoma were mainly treated with carboplatin.
A great variability of ototoxic outcomes has been reported [41, 43] that might depend on differences in assessment strategies, such as inclusion criteria, patient population, methodological techniques, auditory measurements and duration of follow-up. Worthy of note, among different scales we use the Boston SIOP ototoxicity scale, which is sensitive to high frequencies of hearing, the most affected by platinum-derived ototoxicity, even if in younger children they are more difficult to detect [32,41,]. Probably there is not a comprehensive scale for all conditions considering the variability of audiological assessment for age at treatment, status of illness of children during testing, concomitant cognitive and sensorial impairment. However, this grading system seems to be able to include a larger number of affected children.
In our series, five out of children who progressed hearing at the long-term follow-up required hearing aid rehabilitation. Deafness, especially during childhood and even when only borderline to mild, can have important implications. Therefore, a parental counselling on auditory diagnosis and benefits of hearing aids is recommended. The early and appropriate hearing rehabilitation that better suits children needed is crucial for brain plasticity maintenance during sensitive period and language development, social integration [44] and quality of life considering that most of these are multi-handicap patients, mainly with sensory deprivation (i.e. visual) that further affects outcomes. While in teenager the high-frequency hearing loss can reduce speech recognition in noisy environment affecting learning and communicative abilities, especially during school performances. Thus, hearing aids and the use of assistive listening devices such as frequency modulation systems are suitable. Additionally, audiological surveillance is useful in children affected by central nervous system tumours since hearing loss is associated with worse physical well-being, poorer relationship with peers and greater need for social support compared to normal hearing [45]. Furthermore, most of survival children live far from tertiary oncologic institutions, therefore many years after the end of therapies oncological check-ups become less frequent; many families continue their sporadic radiologic and clinical controls at their hometown hospital, making it difficult to obtain a long-term reliable hearing assessment.
Parents may be advocated by the supporting services so that their child is not lost to follow-up once they enter survivorship. In addition to monitoring hearing provides beneficial effects on children and young adults in understanding their hearing status, its impact on communication and learning, hearing conservation, and their options for ongoing management.
In conclusion, platinum compounds can lead to bilateral, late onset or progressive and permanent sensorineural hearing loss affecting language development. Therefore, long-term monitoring is recommended in all children considering that deafness is dependent by treatments and individual susceptibility that likely trigger the onset and progression of ototoxicity. To date, the long-term monitoring at least of 5 years, especially in children who underwent multimodal therapy including irradiation, may offer the highest level of care to our patients and their families improving quality of life after successful oncological treatment.