Otoprotection in Children Treated for Cancer: A CODEPEH Review Using PRISMA Methodology on Pharmacological Agents, Biomarkers and Genetic Susceptibility

1. Introduction

Platinum compounds, predominantly cisplatin, are essential drugs in multiple paediatric tumours, including hepatoblastoma, medulloblastoma, neuroblastoma, osteosarcoma and germ-cell tumours. Their antitumour efficacy is accompanied by auditory toxicity that first appears at high frequencies, progresses with cumulative dose and is usually irreversible. In a multicentre North American cohort, 44% of children, adolescents and young adults exposed to cisplatin developed moderate or severe hearing loss; in the global meta-analysis of platinum exposure, the pooled prevalence was 43.2%, close to 49.2% with cisplatin and lower, although not negligible, with carboplatin [1,2].

The biology of cisplatin-related cochlear injury helps to explain why this toxicity is difficult to prevent. Cisplatin accumulates in the cochlea, particularly in the stria vascularis, where it promotes oxidative stress, mitochondrial dysfunction, inflammation and outer hair-cell death. In addition, the drug can be retained in the inner ear for months or years after treatment, providing a biological explanation for late progression of hearing loss after chemotherapy has been completed [3,4].

Sodium thiosulfate (STS) is the only otoprotective intervention with high-level clinical evidence in children. Two randomised paediatric trials, SIOPEL 6 and ACCL0431, demonstrated reduced cisplatin-induced hearing loss, and subsequent analyses clarified both the magnitude of the auditory benefit and the importance of tumour stage when interpreting oncological safety [5,6,7,8]. New evidence from the Japanese Children’s Cancer Group STS-J01 study extends the clinical experience to Japanese children and adolescents with localised solid tumours, and real-world off-label data are beginning to describe implementation outside pivotal trial conditions [9,10].

Regulatory and guideline interpretation remains focused on children with localised, non-metastatic solid tumours. The US Food and Drug Administration, European Medicines Agency and NICE restrict or recommend systemic STS in this disease context, while clinical practice guidance and recent expert implementation proceedings emphasise careful timing, dosing, workflow integration and avoidance of extrapolation to disseminated disease without multidisciplinary review [11,12,13,14,15,16].

Recent clinical studies in children have shown that sodium thiosulfate reduces cisplatin-induced hearing loss [5,6,7,8,9,10].

At present, early detection of damage continues to rely on serial audiological monitoring. Extended high-frequency audiometry and distortion-product otoacoustic emissions can detect changes earlier than conventional audiometry, but there is considerable interest in the possibility of identifying biomarkers of auditory damage at a very early stage [1,2,3].

Genetic susceptibility exists, but does not yet appear to justify routine pharmacogenetic screening. Consequently, current risk stratification should remain predominantly clinical. From a clinical perspective, risk does not depend solely on cumulative dose. In the largest contemporary cohort, risk increased with more intensive daily and per-cycle doses, younger age, specific diagnoses, especially hepatic tumours, neuroblastoma and brain tumours, and concomitant exposures such as vincristine or cranial irradiation. Paediatric ototoxicity should therefore be interpreted as the result of an interaction between treatment, host, target organ and tumour context [1,3]. In young children, the functional cost of even moderate hearing losses is especially high because it coincides with critical stages of speech and learning development [1,2].

The aim of this review is to synthesize the evidence on pharmacological prevention, early detection through audiological monitoring, analysis of potential biomarkers and assessment of genetic risk for ototoxicity in children treated for cancer, with a principal focus on cisplatin and a reasoned extension to other ototoxic antitumour agents such as carboplatin.

2. Materials and Methods

A structured rapid systematic review using the PRISMA 2020 framework was updated to answer the clinical question: which strategies have demonstrated, or suggest, usefulness for preventing, detecting early or stratifying the risk of ototoxicity in children treated for cancer, especially with cisplatin? The Oxford Centre for Evidence-Based Medicine levels of evidence were used to classify the strength of findings [17,18].

The main search was based on PubMed/MEDLINE and expanded reading of full texts available in PMC or indexed abstracts, complemented by targeted searches of clinical guidelines, regulatory documents, conference abstracts and primary sources relevant to paediatric oncology and otology. Search terms included combinations of cisplatin, platinum, ototoxicity, children, paediatric cancer, sodium thiosulfate, Pedmark, Pedmarqsi, delayed sodium thiosulfate, intratympanic, transtympanic, DB-020, hyaluronan gel, biomarkers, prestin, pharmacogenetics, ACYP2, SLC22A2 and PanCareLIFE. The updated search closing date was 30 July 2026.

Paediatric or mixed studies were included when directly relevant to paediatric oncology in four domains: pharmacological prevention, monitoring and early detection, biological biomarkers and genetic risk. Systematic reviews, consensus statements, practical implementation documents, regulatory assessments and high-quality translational studies were also accepted when they provided information critical for clinical decision-making. Adult or preclinical studies were excluded unless they addressed a delivery route, mechanism or biomarker of strong translational relevance and were explicitly interpreted as indirect evidence.

The screening figures should be interpreted as those of a rapid systematic update, not as a Cochrane review with duplicate independent extraction. The original search identified 89 records. After removal of 18 duplicates or overlapping records, 71 titles and abstracts were screened, 40 full texts or complete records were assessed and 42 key studies or documents were included in the updated qualitative synthesis. A meta-analysis was not performed because of heterogeneity in study design, tumour context, delivery route, audiological scales and outcome definitions.

Table 1. Updated PRISMA selection process.

PRISMA phase	Number	Comment
Records identified	89	Original search plus updated STS, intratympanic-delivery, implementation and biomarker-directed searches.
Duplicates or overlapping records removed	18	Bibliographic duplicates, overlapping abstracts and repeated regulatory or review sources.
Records screened by title/abstract	71	Screening by paediatric relevance, oncology context and relationship to ototoxicity prevention, monitoring or risk stratification.
Full texts or complete records assessed	40	Assessment of clinical trials, cohorts, systematic reviews, consensus statements, implementation guidance, regulatory documents and translational studies.
Documents included in qualitative synthesis	42	Key studies or documents included across pharmacological prevention, intratympanic STS, monitoring/biomarkers and pharmacogenomics.

Table 1. Updated PRISMA selection process.

PRISMA phase	Number	Comment
Records identified	89	Original search plus updated STS, intratympanic-delivery, implementation and biomarker-directed searches.
Duplicates or overlapping records removed	18	Bibliographic duplicates, overlapping abstracts and repeated regulatory or review sources.
Records screened by title/abstract	71	Screening by paediatric relevance, oncology context and relationship to ototoxicity prevention, monitoring or risk stratification.
Full texts or complete records assessed	40	Assessment of clinical trials, cohorts, systematic reviews, consensus statements, implementation guidance, regulatory documents and translational studies.
Documents included in qualitative synthesis	42	Key studies or documents included across pharmacological prevention, intratympanic STS, monitoring/biomarkers and pharmacogenomics.

3. Results

The updated synthesis identified four clinically relevant evidence blocks. First, systemic delayed STS remains the only otoprotective intervention with high-level paediatric evidence. Second, new publications address implementation of systemic STS in daily practice and extend the population evidence base. Third, intratympanic and transtympanic STS represent a promising but still investigational strategy, with adult early-phase and preclinical evidence but no paediatric clinical evidence. Fourth, early detection and genetic stratification remain important adjuncts but are not yet sufficiently validated to replace audiological monitoring or clinical risk assessment.

Table 2. Studies and documents included on pharmacological prevention.

Ref.	Design/source	Population/ context	Intervention	Main outcome or contribution	OCEBM level
5	Randomised trial	109 children with standard-risk hepatoblastoma	Delayed IV STS 6 h after cisplatin	Brock grade >=1 hearing loss was 33% with STS versus 63% with cisplatin alone; no apparent detriment to 3-year event-free or overall survival.	2
6	Randomised, open-label phase III trial	125 paediatric patients with diverse solid tumours	Delayed IV STS versus observation	Reduced cisplatin-induced hearing loss overall; oncological interpretation is most reassuring in localised disease.	2
7	Updated survival analysis	ACCL0431 cohort	Delayed IV STS	Maintained concern for lower survival in disseminated disease, supporting restriction to localised, non-metastatic tumours.	2
8	Secondary reanalysis of an RCT	121/125 ACCL0431 participants evaluable	Central rereading using SIOP scale	Confirmed otoprotection using a consensus paediatric ototoxicity scale.	2
9	Open-label single-arm phase II with historical control	31 Japanese children/adolescents; 25 evaluable primary cohort	Anhydrous STS 12.8 g/m2 IV, 6 h after cisplatin	76% free of ASHA-defined ototoxicity and 84% without Brock hearing loss; RR 0.42 versus historical controls; response rate 95.8%; no STS-related serious adverse events.	3
10	Multicentre retrospective real-world study	Children, adolescents and young adults receiving off-label pentahydrate STS	Pentahydrate STS in clinical practice	Supported feasibility and tolerability outside pivotal trials, but without the certainty of randomised evidence.	3
15	Narrative review and implementation-focused synthesis	Patients with cancer treated with platinum compounds	Systemic STS	Summarised efficacy, pharmacokinetics and safety; highlighted importance of timing and disease extent.	5
16	Expert implementation proceedings	Paediatric daily clinical practice	Implementation of systemic STS	Identified practical challenges including timing, integration into protocols, drug interactions, eligibility and costs.	5
19	Cochrane systematic review	Children with cancer receiving platinum compounds	Medical interventions	Confirmed limited evidence for most otoprotectants before modern STS approvals.	1
20	Systematic review and meta-analysis	4 STS clinical trials; 278 patients	STS during platinum chemotherapy	STS was associated with lower risk of ototoxic effects; interpretation limited by route, timing and oncological heterogeneity.	1
21	Non-randomised comparative cohort	379 children with standard- or high-risk medulloblastoma	Amifostine	Protective signal in standard-risk disease, not confirmed in high-risk disease; not considered standard.	3
22	Non-randomised controlled phase I study	52 paediatric patients with non-metastatic tumours	IV N-acetylcysteine 4 h after cisplatin	Feasible and showed a favourable signal, but requires randomised confirmation.	3
23	Phase II randomised controlled trial in adults	94 adults with diverse tumours	Aspirin	Negative; did not reduce cisplatin-related hearing loss.	2
24	Phase II randomised controlled trial in adults	50 adults; 27 analysable	Oral D-methionine	Lower high-frequency threshold shift; indirect evidence for paediatrics.	2

Table 2. Studies and documents included on pharmacological prevention.

Ref.	Design/source	Population/ context	Intervention	Main outcome or contribution	OCEBM level
5	Randomised trial	109 children with standard-risk hepatoblastoma	Delayed IV STS 6 h after cisplatin	Brock grade >=1 hearing loss was 33% with STS versus 63% with cisplatin alone; no apparent detriment to 3-year event-free or overall survival.	2
6	Randomised, open-label phase III trial	125 paediatric patients with diverse solid tumours	Delayed IV STS versus observation	Reduced cisplatin-induced hearing loss overall; oncological interpretation is most reassuring in localised disease.	2
7	Updated survival analysis	ACCL0431 cohort	Delayed IV STS	Maintained concern for lower survival in disseminated disease, supporting restriction to localised, non-metastatic tumours.	2
8	Secondary reanalysis of an RCT	121/125 ACCL0431 participants evaluable	Central rereading using SIOP scale	Confirmed otoprotection using a consensus paediatric ototoxicity scale.	2
9	Open-label single-arm phase II with historical control	31 Japanese children/adolescents; 25 evaluable primary cohort	Anhydrous STS 12.8 g/m2 IV, 6 h after cisplatin	76% free of ASHA-defined ototoxicity and 84% without Brock hearing loss; RR 0.42 versus historical controls; response rate 95.8%; no STS-related serious adverse events.	3
10	Multicentre retrospective real-world study	Children, adolescents and young adults receiving off-label pentahydrate STS	Pentahydrate STS in clinical practice	Supported feasibility and tolerability outside pivotal trials, but without the certainty of randomised evidence.	3
15	Narrative review and implementation-focused synthesis	Patients with cancer treated with platinum compounds	Systemic STS	Summarised efficacy, pharmacokinetics and safety; highlighted importance of timing and disease extent.	5
16	Expert implementation proceedings	Paediatric daily clinical practice	Implementation of systemic STS	Identified practical challenges including timing, integration into protocols, drug interactions, eligibility and costs.	5
19	Cochrane systematic review	Children with cancer receiving platinum compounds	Medical interventions	Confirmed limited evidence for most otoprotectants before modern STS approvals.	1
20	Systematic review and meta-analysis	4 STS clinical trials; 278 patients	STS during platinum chemotherapy	STS was associated with lower risk of ototoxic effects; interpretation limited by route, timing and oncological heterogeneity.	1
21	Non-randomised comparative cohort	379 children with standard- or high-risk medulloblastoma	Amifostine	Protective signal in standard-risk disease, not confirmed in high-risk disease; not considered standard.	3
22	Non-randomised controlled phase I study	52 paediatric patients with non-metastatic tumours	IV N-acetylcysteine 4 h after cisplatin	Feasible and showed a favourable signal, but requires randomised confirmation.	3
23	Phase II randomised controlled trial in adults	94 adults with diverse tumours	Aspirin	Negative; did not reduce cisplatin-related hearing loss.	2
24	Phase II randomised controlled trial in adults	50 adults; 27 analysable	Oral D-methionine	Lower high-frequency threshold shift; indirect evidence for paediatrics.	2

Overall, the greatest paediatric certainty remains concentrated in delayed systemic STS. The 2026 STS-J01 study [9] strengthens external validity across populations but does not replace the pivotal randomised evidence, because it is single-arm and relies on historical comparison. Other systemic agents remain investigational or unsupported for routine paediatric use [19,20,21,22,23,24].

Table 3. Evidence on intratympanic or transtympanic sodium thiosulfate and local delivery systems.

Ref.	Design/source	Population/ model	Intervention/ route	Main outcome or contribution	OCEBM level
[25]	Randomised phase Ib clinical trial	Adults receiving high-dose cisplatin	Intratympanic DB-020, 12% or 25%, one ear; placebo in the contralateral ear	DB-020 25% reduced ototoxicity compared with placebo ears, with low systemic thiosulfate exposure and no apparent effect on free cisplatin plasma concentrations.	3
[26]	Phase I safety, tolerability and pharmacokinetic study	Adult cancer patients receiving cisplatin	Novel intratympanic thiosulfate	Supported local tolerability and very low systemic exposure, providing rationale for further development.	4
[27]	Randomised clinical trial	Adults with head and neck cancer	Transtympanic STS gel	Showed a non-significant trend toward less hearing loss; response was variable.	3
[28]	Systematic review and meta-analysis of intratympanic therapy	Adults receiving cisplatin; multiple intratympanic agents	Intratympanic dexamethasone, NAC, STS and other agents	No agent demonstrated consistent pooled benefit; STS data were limited and not sufficiently mature for paediatric extrapolation.	1
[29]	Phase III trial protocol	100 adults with head and neck cancer planned for cisplatin >=200 mg/m2	Transtympanic STS before each cisplatin infusion; within-patient ear-level control	Ongoing SOUND trial designed to test clinically relevant threshold shift benefit.	5
[30]	Review of advanced delivery systems	Translational and formulation literature	Nanoparticles, hydrogels, permeation enhancers and local delivery systems	Maps strategies to overcome the 6-h systemic window and improve cochlear exposure.	5
[31]	Preclinical animal study	Guinea-pig cisplatin ototoxicity model	Intratympanic STS-hyaluronan gel at pH 6.5 or 8.0	Reduced outer hair-cell loss at both pH levels; NaCl control was ineffective.	5
[32]	Preclinical formulation study	Inner-ear delivery models	STS-loaded solid lipid nanoparticles	Demonstrated rationally designed nanoparticle delivery and sustained-release concepts.	5

Table 3. Evidence on intratympanic or transtympanic sodium thiosulfate and local delivery systems.

Ref.	Design/source	Population/ model	Intervention/ route	Main outcome or contribution	OCEBM level
[25]	Randomised phase Ib clinical trial	Adults receiving high-dose cisplatin	Intratympanic DB-020, 12% or 25%, one ear; placebo in the contralateral ear	DB-020 25% reduced ototoxicity compared with placebo ears, with low systemic thiosulfate exposure and no apparent effect on free cisplatin plasma concentrations.	3
[26]	Phase I safety, tolerability and pharmacokinetic study	Adult cancer patients receiving cisplatin	Novel intratympanic thiosulfate	Supported local tolerability and very low systemic exposure, providing rationale for further development.	4
[27]	Randomised clinical trial	Adults with head and neck cancer	Transtympanic STS gel	Showed a non-significant trend toward less hearing loss; response was variable.	3
[28]	Systematic review and meta-analysis of intratympanic therapy	Adults receiving cisplatin; multiple intratympanic agents	Intratympanic dexamethasone, NAC, STS and other agents	No agent demonstrated consistent pooled benefit; STS data were limited and not sufficiently mature for paediatric extrapolation.	1
[29]	Phase III trial protocol	100 adults with head and neck cancer planned for cisplatin >=200 mg/m2	Transtympanic STS before each cisplatin infusion; within-patient ear-level control	Ongoing SOUND trial designed to test clinically relevant threshold shift benefit.	5
[30]	Review of advanced delivery systems	Translational and formulation literature	Nanoparticles, hydrogels, permeation enhancers and local delivery systems	Maps strategies to overcome the 6-h systemic window and improve cochlear exposure.	5
[31]	Preclinical animal study	Guinea-pig cisplatin ototoxicity model	Intratympanic STS-hyaluronan gel at pH 6.5 or 8.0	Reduced outer hair-cell loss at both pH levels; NaCl control was ineffective.	5
[32]	Preclinical formulation study	Inner-ear delivery models	STS-loaded solid lipid nanoparticles	Demonstrated rationally designed nanoparticle delivery and sustained-release concepts.	5

Table 4. Practical comparison of systemic delayed STS and intratympanic/transtympanic STS.

Feature	Systemic delayed STS	Intratympanic/transtympanic STS	Main refs
Evidence level	Randomised paediatric phase III evidence; FDA, EMA and NICE authorisation/recommendation in localised, non-metastatic solid tumours.	Adult phase Ib, adult randomised or protocol data, and preclinical studies; no paediatric clinical data.	[5,6,7,8,9,10,11,12,13,14,15,16,25,26,27,28,29,30,31]
Timing	Delayed administration, usually 6 h after cisplatin, to reduce risk of tumour protection.	Can be given before cisplatin, commonly within 3 h before infusion in adult protocols.	[5,6,25,29]
Oncological interaction	Theoretical and observed concern if systemic chelation overlaps antitumour exposure, especially in disseminated disease.	Theoretical advantage of low systemic exposure and local cochlear chelation only; still requires oncological safety confirmation.	[7,15,25,26]
Clinical applicability	Current paediatric standard only for selected localised, non-metastatic solid tumours.	Investigational; potentially attractive for adults receiving high-dose cisplatin and future paediatric scenarios where systemic STS is problematic.	[11,12,13,14,15,16,25,26,27,28,29]
Main limitations	Requires strict workflow timing; electrolyte abnormalities, nausea/vomiting and implementation costs; uncertain safety in metastatic disease.	Variable round-window permeability, middle-ear conditions, need for repeated injections, procedural burden and absence of paediatric data.	[15,16,28,29]

Table 4. Practical comparison of systemic delayed STS and intratympanic/transtympanic STS.

Feature	Systemic delayed STS	Intratympanic/transtympanic STS	Main refs
Evidence level	Randomised paediatric phase III evidence; FDA, EMA and NICE authorisation/recommendation in localised, non-metastatic solid tumours.	Adult phase Ib, adult randomised or protocol data, and preclinical studies; no paediatric clinical data.	[5,6,7,8,9,10,11,12,13,14,15,16,25,26,27,28,29,30,31]
Timing	Delayed administration, usually 6 h after cisplatin, to reduce risk of tumour protection.	Can be given before cisplatin, commonly within 3 h before infusion in adult protocols.	[5,6,25,29]
Oncological interaction	Theoretical and observed concern if systemic chelation overlaps antitumour exposure, especially in disseminated disease.	Theoretical advantage of low systemic exposure and local cochlear chelation only; still requires oncological safety confirmation.	[7,15,25,26]
Clinical applicability	Current paediatric standard only for selected localised, non-metastatic solid tumours.	Investigational; potentially attractive for adults receiving high-dose cisplatin and future paediatric scenarios where systemic STS is problematic.	[11,12,13,14,15,16,25,26,27,28,29]
Main limitations	Requires strict workflow timing; electrolyte abnormalities, nausea/vomiting and implementation costs; uncertain safety in metastatic disease.	Variable round-window permeability, middle-ear conditions, need for repeated injections, procedural burden and absence of paediatric data.	[15,16,28,29]

Table 5. Studies included on monitoring, early detection and biomarkers.

Ref.	Design/source	Population/ model	Marker type	Main finding	OCEBM level
[44]	Harmonised survivor guideline	Childhood, adolescent and young adult cancer survivors	Audiological surveillance	Recommends long-term surveillance after exposure to platinum compounds and/or cranial radiotherapy.	1
[45]	Multidisciplinary SIOP consensus	During childhood cancer treatment	Audiological monitoring	Recommends baseline, during-treatment and end-of-treatment assessment; ideally before cisplatin cycles according to risk and resources.	5
[46]	Prospective paediatric study	17 children/adolescents	Extended high-frequency audiometry and DPOAE	Detected early changes before conventional audiometry.	3
[47]	Translational preclinical study	Rodents	Serum prestin	Prestin increased transiently after cisplatin and suggested value as a marker of cochlear injury.	5
[48]	Prospective human study	42 patients	Prestin in blood, CSF and perilymph	Prestin was detectable but variable, with no robust clinical blood correlation.	3
[49]	Experimental validation study	Mouse model	Serum prestin by ELISA	Concluded that prestin was not a sensitive or reliable serum biomarker with the ELISAs evaluated.	5
[50,51,52]	Scoping review and reliability studies	Human hearing and sensorineural hearing-loss literature	Inner-ear and prestin biomarkers	Support biological plausibility but show that analytical and clinical validation remain incomplete.	4-5
[53,54,55,56,57,58]	Preclinical or exploratory biomarker studies	Inner-ear disorders, animal ototoxicity models and cochlear explants	OTOLIN-1, miRNAs, extracellular vesicles and inflammatory cytokines	Promising exploratory signals, but not ready for routine paediatric oncology use.	5

Table 5. Studies included on monitoring, early detection and biomarkers.

Ref.	Design/source	Population/ model	Marker type	Main finding	OCEBM level
[44]	Harmonised survivor guideline	Childhood, adolescent and young adult cancer survivors	Audiological surveillance	Recommends long-term surveillance after exposure to platinum compounds and/or cranial radiotherapy.	1
[45]	Multidisciplinary SIOP consensus	During childhood cancer treatment	Audiological monitoring	Recommends baseline, during-treatment and end-of-treatment assessment; ideally before cisplatin cycles according to risk and resources.	5
[46]	Prospective paediatric study	17 children/adolescents	Extended high-frequency audiometry and DPOAE	Detected early changes before conventional audiometry.	3
[47]	Translational preclinical study	Rodents	Serum prestin	Prestin increased transiently after cisplatin and suggested value as a marker of cochlear injury.	5
[48]	Prospective human study	42 patients	Prestin in blood, CSF and perilymph	Prestin was detectable but variable, with no robust clinical blood correlation.	3
[49]	Experimental validation study	Mouse model	Serum prestin by ELISA	Concluded that prestin was not a sensitive or reliable serum biomarker with the ELISAs evaluated.	5
[50,51,52]	Scoping review and reliability studies	Human hearing and sensorineural hearing-loss literature	Inner-ear and prestin biomarkers	Support biological plausibility but show that analytical and clinical validation remain incomplete.	4-5
[53,54,55,56,57,58]	Preclinical or exploratory biomarker studies	Inner-ear disorders, animal ototoxicity models and cochlear explants	OTOLIN-1, miRNAs, extracellular vesicles and inflammatory cytokines	Promising exploratory signals, but not ready for routine paediatric oncology use.	5

The monitoring and biomarker evidence shows a clear hierarchy of clinical readiness. Serial audiological monitoring is implementable now, extended high-frequency audiometry and DPOAE are useful functional early-warning tools, and serum or molecular biomarkers remain exploratory. Prestin is biologically attractive because it reflects outer hair-cell biology, but current studies are inconsistent and analytically fragile.

Table 6. Studies included on genetic risk of ototoxicity.

Ref.	Design/ source	Cohort	Markers/ objective	Main finding	OCEBM level
[59]	PanCareLIFE candidate-marker cohort	900 paediatric survivors	14 SNPs in 11 genes	SLC22A2 rs316019 retained a signal, but overall predictive performance was poor compared with clinical factors.	3
[60]	Systematic review and meta-analysis	32 articles; 4406 participants	Genetic variation in platinum ototoxicity	ACYP2 rs1872328 showed an association; other effects depended on therapeutic context.	1
[61]	Systematic review	30 studies	Genetic vulnerability to cisplatin	ACYP2 showed a marked risk signal, but heterogeneity was substantial.	1
[62]	Multicentre GWAS meta-analysis	GO-CAT and UK MAGIC cohorts	Genome-wide discovery	No definitive genome-wide findings; TSPAN5 and other loci remained suggestive.	3
[63]	Paediatric cohort and systematic review/meta-analysis	United Kingdom	TPMT, COMT and ACYP2	Contradictory results reinforced field heterogeneity.	3
[64,65,66]	Candidate and GWAS studies	Paediatric platinum-treated cohorts	Clinical and genetic risk markers including TCERG1L	Support genetic susceptibility but do not yet justify isolated SNP-based clinical decision-making.	3

Table 6. Studies included on genetic risk of ototoxicity.

Ref.	Design/ source	Cohort	Markers/ objective	Main finding	OCEBM level
[59]	PanCareLIFE candidate-marker cohort	900 paediatric survivors	14 SNPs in 11 genes	SLC22A2 rs316019 retained a signal, but overall predictive performance was poor compared with clinical factors.	3
[60]	Systematic review and meta-analysis	32 articles; 4406 participants	Genetic variation in platinum ototoxicity	ACYP2 rs1872328 showed an association; other effects depended on therapeutic context.	1
[61]	Systematic review	30 studies	Genetic vulnerability to cisplatin	ACYP2 showed a marked risk signal, but heterogeneity was substantial.	1
[62]	Multicentre GWAS meta-analysis	GO-CAT and UK MAGIC cohorts	Genome-wide discovery	No definitive genome-wide findings; TSPAN5 and other loci remained suggestive.	3
[63]	Paediatric cohort and systematic review/meta-analysis	United Kingdom	TPMT, COMT and ACYP2	Contradictory results reinforced field heterogeneity.	3
[64,65,66]	Candidate and GWAS studies	Paediatric platinum-treated cohorts	Clinical and genetic risk markers including TCERG1L	Support genetic susceptibility but do not yet justify isolated SNP-based clinical decision-making.	3

4. Discussion

4.1. Pathophysiological Basis of Platinum-Induced Cochlear Damage

Understanding the pharmacodynamics of platinum agents is essential for anticipating cochlear damage. Cisplatin crosses the blood-labyrinth barrier through CTR1 copper transporters, which are found in hair cells, the stria vascularis and the spiral ganglion and accumulates in vulnerable cochlear structures for months or even years after treatment has ended, explaining the late progression of damage. It generates reactive oxygen species, injures mitochondria, activates inflammatory and apoptotic pathways and damages outer hair cells first in the basal turn of the cochlea. This explains the characteristic initial high-frequency pattern and subsequent progression toward speech frequencies [3,4].

The pathophysiology also explains why simple dose reduction is not an adequate solution in paediatric oncology. Reducing cisplatin exposure may jeopardise tumour control, while cochlear retention means that injury may progress even after completion of treatment. Effective prevention therefore requires an intervention that protects the cochlea without reducing antitumour efficacy.

4.2. Systemic Sodium Thiosulfate: Mechanisms, Clinical Evidence and Tumour Context

STS acts primarily through chemical neutralisation and sequestration of reactive platinum species, while also influencing antioxidant pathways. The decisive clinical principle is delayed administration. Giving STS approximately 6 h after cisplatin aims to reduce systemic neutralisation of the antitumour drug while still limiting cochlear injury [5,6,14,15].

The strongest evidence exists in standard-risk hepatoblastoma, where SIOPEL 6 showed a major reduction in hearing loss without an apparent survival penalty. ACCL0431 broadened evidence to diverse tumours but also introduced the key caution: in disseminated disease, updated survival analyses raised concern that systemic STS could compromise outcomes. This distinction explains why regulatory indications and recommendations focus on localised, non-metastatic solid tumours [5,6,7,8,9,10,11,12,13].

The 2026 STS-J01 study is important because it confirms reproducibility of delayed anhydrous STS in a Japanese paediatric and adolescent population. It used 12.8 g/m2 IV 6 h after cisplatin and reported that most evaluable patients remained free of ASHA-defined ototoxicity, with preserved tumour response and no serious adverse events attributed to STS. However, because it was single-arm and compared with historical controls, it should be viewed as supportive external-validity evidence rather than as a new pivotal trial [9].

Implementation is now one of the central practical issues. Expert proceedings from the SIOP Supportive Care Network emphasise eligibility selection, exact timing of cisplatin and STS infusion, integration with existing protocols, coordination between oncology, pharmacy and audiology, management of electrolyte abnormalities and economic considerations. These are not minor operational details: incorrect timing could reduce otoprotection or increase the risk of tumour interaction [15,16].

Interpretation by tumour type remains essential. In hepatoblastoma the evidence is most direct. In other localised solid tumours treated with cisplatin, the evidence is less tumour-specific but supported by ACCL0431, regulatory reasoning and expert guidelines. In metastatic or disseminated disease, routine systemic STS remains difficult to justify outside trials or carefully individualised multidisciplinary decisions [7,13,14,15,16].

4.3. Intratympanic and Transtympanic Sodium Thiosulfate

The intratympanic route offers a compelling theoretical solution to the pharmacokinetic paradox of systemic STS. Systemic STS must be delayed to avoid neutralising cisplatin during its antitumour exposure, leaving the cochlea unprotected during the early peak of platinum exposure. Local intratympanic or transtympanic administration could be given before cisplatin, producing high inner-ear exposure with minimal systemic thiosulfate and therefore lower theoretical risk of interfering with antitumour efficacy [25,26,27].

The most mature clinical signal comes from the adult DB-020 programme. In the randomised phase Ib trial, DB-020 was administered intratympanically to one ear and placebo to the contralateral ear before high-dose cisplatin. The 25% formulation produced a clinically meaningful reduction in ototoxicity, and systemic thiosulfate levels were far below those expected to affect cisplatin efficacy. These findings are encouraging but remain adult data, not paediatric evidence [25,26].

Other transtympanic studies have been less definitive. The Duinkerken trial in adults with head and neck cancer showed only a non-significant trend toward benefit, while the 2026 systematic review of intratympanic therapy concluded that no intratympanic agent had shown consistent pooled benefit. Importantly, the STS evidence was too limited for confident pooling, and paediatric clinical data are absent [27,28].

The ongoing SOUND trial is therefore highly relevant. It will evaluate transtympanic STS in 100 adults with head and neck cancer receiving high-dose cisplatin, using a within-patient design in which one ear is treated and the contralateral ear serves as control. If positive, it will strengthen the rationale for broader development but still will not resolve paediatric feasibility, acceptability or safety [29].

Formulation science is developing rapidly. Advanced delivery strategies include hyaluronan hydrogels designed to prolong middle-ear residence, solid lipid or polymeric nanoparticles, and non-invasive transtympanic approaches with permeation enhancers. A 2026 guinea-pig study showed that STS-hyaluronan gel reduced outer hair-cell loss at pH 6.5 and pH 8.0, while NaCl control was ineffective, supporting the biological plausibility of sustained local STS delivery [30,31,32].

A broader 2026 review of sodium thiosulfate also underlines that otoprotection is part of a wider clinical-development landscape for this compound, but paediatric cisplatin otoprotection remains the most mature oncology indication [33].

At present, however, intratympanic STS should be described in the review as an investigational route. Its potential advantages are substantial, especially for settings where systemic STS is contraindicated or controversial, but important limitations remain: variability in round-window permeability, middle-ear effusion or radiation-related middle-ear disease, need for repeated procedures, procedural discomfort, local adverse events and the complete absence of paediatric oncology trials [25,26,27,28,29,30,31,32].

4.4. Other Pharmacological Strategies

Amifostine showed a protective signal only in standard-risk medulloblastoma, but the study was not randomized and failed to demonstrate high-risk benefit. Therefore, amifostine cannot be considered standard; It is more prudent to understand it as a historical proof of concept of antioxidant protection [21].

N-acetylcysteine is probably the most promising candidate after thiosulfate. Its biological rationale is solid: it promotes glutathione synthesis and buffers oxidative stress, one of the central axes of cisplatin-related cochlear injury. The paediatric phase I study demonstrated feasibility, a recommended phase II dose and a clinically interesting efficacy signal; however, it remains a developing strategy because it lacks a confirmatory randomised paediatric phase III trial [22].

Aspirin and D-methionine illustrate the distance between pathophysiological plausibility and clinical applicability. Aspirin failed in a randomised trial despite the anti-inflammatory and anti-ROS hypothesis; D-methionine showed protection at high frequencies in adults, but the evidence remains indirect for paediatrics and has not changed paediatric oncology practice [23,24].

Amifostine, N-acetylcysteine, aspirin and D-methionine illustrate the gap between mechanistic plausibility and paediatric clinical certainty.

Other plausible treatments not included in this review include the debated otoprotective effect of steroids, with the counterpoint that they may reduce the antitumour effectiveness of platinum-based medicines when administered systemically; alternative routes are therefore being sought, in addition to other safety problems associated with prolonged or chronic doses, especially in children [34]. More otoprotective candidates include vitamins, nutritional antioxidants, ebselen, ORC-13661, curcumin, ferulic acid, statins, anti-inflammatory approaches and hydrogen. Many have preclinical or adult evidence, but none has the paediatric evidence base required to replace or match systemic STS in its approved indication [34,35,36,37,38,39,40].

Some possible treatments are based on the view that inflammation is crucial in the pathogenesis of acquired sensorineural hearing loss, although the precise mechanism involved remains difficult to establish. Acetylsalicylic acid could be useful in aminoglycoside ototoxicity but not in cisplatin ototoxicity [23], as discussed above; statins could also be otoprotective agents because of their cholesterol-lowering, antioxidant and anti-inflammatory properties [41].

Tumour necrosis factor alpha (TNF-α) plays a key role in cisplatin ototoxicity, and the combination of IFN-γ and TNF-α appears to increase cisplatin cytotoxicity in cochlear sensory cells ex vivo, opening a pathway for the use of anti-TNF and anti-IFN drugs to prevent cochlear damage. Studies on TNF-α and otoprotection exist. Preclinical evidence indicates that blocking TNF-α or modulating its inflammatory/oxidative pathways may protect the inner ear against noise, drugs, cochlear implantation and autoimmune processes. However, TNF-α also fulfils physiological functions, and clinical data with anti-TNF agents in humans remain limited and sometimes contradictory [42].

Aqueous and inhaled gaseous hydrogen (H2) has also shown potential for preventing ischaemia in several animal models of ototoxicity and nephrotoxicity, presumably by reducing oxidative stress. In vivo cisplatin-induced hearing loss has been shown to decrease when hydrogen gas inhalation is used, although its effects on the tumour are unknown [43].

Table 7. Pharmacological prevention: mechanism and applicability by tumour context.

Strategy	Proposed mechanism	Current best-fit scenario	Practical status
Systemic delayed STS	Chelation/sequestration of reactive platinum species and reduction of oxidative stress; delayed administration to minimise tumour protection.	Standard-risk hepatoblastoma and other localised, non-metastatic solid tumours treated with cisplatin.	Selective paediatric standard in approved scenarios.
Intratympanic/ transtympanic STS	Local cochlear platinum neutralisation with minimal systemic exposure; avoids systemic 6-h delay in theory.	Investigational adult high-dose cisplatin settings; possible future role where systemic STS is problematic.	Promising but not paediatric standard.
Amifostine	Thiol prodrug with free-radical scavenging and tissue-protective effects.	Historical signal in standard-risk medulloblastoma.	Not standard.
N-acetylcysteine	Glutathione repletion and redox modulation.	Localised tumours treated with cisplatin; still without phase III evidence.	Promising/investigational.
Aspirin	Modulation of inflammation and reactive oxygen species.	No convincing paediatric subgroup.	Not recommended as cisplatin otoprotectant.
D-methionine	Sulphur-containing antioxidant and support for thiol/glutathione systems.	Indirect adult signal; no paediatric validation.	Experimental/translational.

Table 7. Pharmacological prevention: mechanism and applicability by tumour context.

Strategy	Proposed mechanism	Current best-fit scenario	Practical status
Systemic delayed STS	Chelation/sequestration of reactive platinum species and reduction of oxidative stress; delayed administration to minimise tumour protection.	Standard-risk hepatoblastoma and other localised, non-metastatic solid tumours treated with cisplatin.	Selective paediatric standard in approved scenarios.
Intratympanic/ transtympanic STS	Local cochlear platinum neutralisation with minimal systemic exposure; avoids systemic 6-h delay in theory.	Investigational adult high-dose cisplatin settings; possible future role where systemic STS is problematic.	Promising but not paediatric standard.
Amifostine	Thiol prodrug with free-radical scavenging and tissue-protective effects.	Historical signal in standard-risk medulloblastoma.	Not standard.
N-acetylcysteine	Glutathione repletion and redox modulation.	Localised tumours treated with cisplatin; still without phase III evidence.	Promising/investigational.
Aspirin	Modulation of inflammation and reactive oxygen species.	No convincing paediatric subgroup.	Not recommended as cisplatin otoprotectant.
D-methionine	Sulphur-containing antioxidant and support for thiol/glutathione systems.	Indirect adult signal; no paediatric validation.	Experimental/translational.

4.5. Early Diagnosis with Audiological and Molecular Biomarkers

In current practice, the most useful early-detection biomarker is still functional and audiological rather than serum-based. International recommendations support baseline assessment, monitoring during therapy, end-of-treatment assessment and long-term surveillance after platinum compounds or cranial radiotherapy. High-risk patients should be monitored more intensively, ideally around cisplatin cycles when feasible [44,45].

Extended high-frequency audiometry and DPOAE can detect cochlear changes earlier than conventional audiometry, which is clinically important because early damage begins outside the speech-frequency range. These methods should be interpreted as functional early-warning tools rather than replacements for conventional audiometry or age-adapted objective testing such as auditory brainstem responses in very young children [44,45,46].

Serum biomarkers remain promising but immature. Prestin is the most attractive biomarker because of its biological specificity. It is the membrane motor protein of the outer hair cells, precisely one of the structures most vulnerable to cisplatin. This plausibility has led to its exploration as a serum marker of subclinical cochlear damage. Animal studies show a transient post-exposure increase and human studies have confirmed that it can be detected in blood, cerebrospinal fluid and perilymph. A scoping review on blood prestin levels concluded that the findings are promising but still heterogeneous, and a reliability study in humans showed that serum levels may be reproducible in healthy subjects, which is analytically important, although insufficient to validate its clinical use in ototoxicity [47,48,49,50,51,52]. The problem is that kinetics are volatile, depend on sampling time and analytical method, and the most recent validation studies warn of specificity and sensitivity problems in the available ELISAs. Overall, prestin is currently the most biologically plausible serum candidate, but it cannot yet be recommended for routine clinical use. Its future usefulness, if confirmed, will probably not be to replace audiology, but to complement early risk stratification and activate more intensive audiological surveillance. Current evidence does not support using serum prestin to make routine cisplatin or STS decisions [53,54].

OTOLIN-1 is another inner-ear protein detectable in blood. It has been proposed as a serological marker of labyrinthine pathology, and serum elevation has been observed in Ménière disease and sudden deafness. Its main limitation in paediatric oncology is lower specificity for cisplatin-related cochlear damage, so its current value is more exploratory than clinical [55].

Among molecular biomarkers, circulating microRNAs are a promising line of research. MicroRNAs are small non-coding RNAs that regulate gene expression and have exceptional stability in plasma and serum, facilitating their use as diagnostic biomarkers.

The miR-183 family is a cluster comprising miR-96, miR-182 and miR-183, which is abundantly expressed in spiral ganglion neurones and sensory cells of the inner ear. Its increase in blood reflects processes of cell destruction and inflammation in the cochlea.

In animal models of ototoxicity, miR-205 has also shown an early and relatively sustained serum increase, suggesting that it may reflect inner-ear injury before clinical consolidation of the damage. This miRNA has been identified as migrating from the cochlea to the blood after ototoxic damage, and the extent of its elevation is related to the degree of auditory injury. However, evidence in humans and, above all, in paediatric patients treated with cisplatin remains insufficient [56].

miR-34a is used as a marker of oxidative stress and cellular ageing. In experimental models, its overexpression exacerbates cisplatin ototoxicity by inducing mitochondrial dysfunction.

In addition to their diagnostic value, these biomarkers provide routes for future gene therapies. For example, targeted suppression of certain miRNAs, such as miR-34a or miR-29b, has been observed to reduce apoptosis and oxidative stress in the auditory epithelium.

More recently, experimental studies of small extracellular vesicles derived from cochlear explants exposed to cisplatin have identified multi-omic signatures with potential value as biomarkers and therapeutic targets. This, however, remains a clearly preclinical line of work [57].

The analysis of inflammatory cytokine biomarkers such as CXCL1 is also at a preclinical stage. Aameri demonstrated that the chemokine CXCL1 increased in serum and cochlea 24 hours after cisplatin administration [58]. CXCL1 is classified as an inflammatory cytokine with biological sources in both serum and cochlear tissue. The molecular mechanism involves the CXCR2 receptor pathway, which regulates the production of inflammatory mediators.

Changes in CXCL1 mRNA levels were observed as early as 6 hours after cisplatin administration, before any audiometric change. At 24 hours, significant increases in CXCL1 levels occurred (p<0.0003) despite the absence of outer hair-cell loss or shifts in the ABR threshold. This early elevation pattern indicates the potential of CXCL1 as an early warning biomarker.

Transtympanic administration of SB225002, a CXCR2 inhibitor, reduced immune-cell migration, protected against hearing loss and preserved hair-cell integrity. The authors concluded that the CXCL1 chemokine acts as an early player in cisplatin ototoxicity by initiating the immune cascade, with CXCR2 as a relevant therapeutic target [58].

All these other molecular candidates, including circulating microRNAs, extracellular-vesicle signatures and inflammatory cytokines such as CXCL1, may reflect different stages of ototoxic injury. Their future value may lie in multibiomarker panels combined with audiological and clinical risk, but they remain exploratory, with very limited direct validation in children receiving cisplatin for cancer [53,54,55,56,57,58].

4.6. Assessment of Genetic Risk of Ototoxicity

Genetics plays a key role in susceptibility to ototoxicity, explaining between 38% and 47% of individual variability in hearing loss induced by drugs such as cisplatin. Whereas some children tolerate high doses without damage, others develop severe hearing loss after the first cycle because of their genetic profile. The genetic literature demonstrates that inherited susceptibility to platinum-related ototoxicity exists, but it does not yet show that it can be used robustly in clinic to decide who should or should not receive an otoprotectant.

The most reasonable current framework is one of polygenic risk, involving genes related to cisplatin transport, oxidative stress, DNA repair and cochlear biology [59]. The most replicated candidate signals involve SLC22A2 and ACYP2, while candidate-gene and GWAS studies have also explored TPMT, COMT, TCERG1L and other loci. Across studies, heterogeneity in phenotype definitions, treatment exposures, age, cranial irradiation and statistical power limits clinical translation [59,60,61,62,63,64,65,66].

The most important practical message is that clinical predictors still outperform isolated genetic markers. In the large PanCareLIFE cohort, age at treatment, cranial radiotherapy and platinum exposure patterns explained risk more robustly than candidate SNPs. Therefore, pharmacogenomics should be considered a complementary research layer, not a substitute for clinical risk assessment or audiological monitoring [59,60,61,62,63,64,65,66,67].

The major interest in pretreatment genetic screening is that it would allow personalized care, enabling oncologists to practice precision medicine by adjusting doses, intensifying audiological monitoring or selecting less toxic alternative drugs, such as carboplatin, in patients identified as high risk

The final realistic short-term application is the development of integrated prediction models combining age, tumour type, platinum dose and schedule, cranial irradiation, baseline hearing status, audiological early-warning measures, serum biomarkers and polygenic information. Such models will require prospective validation before they can guide preventive treatment decisions.

4.7. Limitations and Open Questions

This review has several limitations. First, it is a rapid systematic review update using a PRISMA format, not a full Cochrane review with duplicate independent extraction and formal meta-analysis. Second, the evidence combines different audiological scales, including Brock, SIOP, ASHA and Chang, which limits quantitative comparison. Third, the intratympanic STS literature is predominantly adult or preclinical, so paediatric extrapolation must remain cautious. Fourth, biomarker and pharmacogenomic evidence is promising but not yet validated for routine clinical decisions.

The main open questions are whether systemic STS can be safely extended beyond localised disease; whether strict implementation can reproduce trial-level benefit in real-world settings; whether intratympanic STS can provide reliable cochlear protection without systemic tumour interaction; whether paediatric intratympanic protocols are feasible and acceptable; and whether biomarker-genomic models will outperform clinical risk factors alone.

From the perspective of other antitumour agents, the main drug to consider is carboplatin. Its ototoxic potential is lower than that of cisplatin, but it is not zero, especially in high-dose contexts, combinations with radiotherapy or very young patients. No otoprotectant has so far shown evidence comparable with that of thiosulfate in cisplatin; prevention in carboplatin therefore continues to rely mainly on patient selection, close monitoring and avoidance of ototoxic co-exposures. [2,3,19]

4.8. Practical Implications

With current evidence, clinical management should rest on five pillars:

• Baseline age-appropriate audiological assessment before starting cisplatin.
• Serial monitoring during chemotherapy, after treatment and during survivorship, intensified in high-risk patients.
• Use of delayed systemic STS in children with localised, non-metastatic solid tumours when the therapeutic regimen and workflow permit.
• Avoidance of routine systemic STS in disseminated disease outside trials or carefully documented multidisciplinary decisions.
• Prospective collection of audiological, biomarker and genetic data to build validated risk-prediction models and to prepare future paediatric studies of local STS delivery.

5. Conclusions

Delayed systemic sodium thiosulfate remains the reference pharmacological intervention for preventing cisplatin ototoxicity in children with localised, non-metastatic solid tumours, with the strongest evidence in standard-risk hepatoblastoma and supportive evidence in other localised solid tumours. The 2026 STS-J01 data strengthen cross-population reproducibility, while implementation guidance highlights the need for precise timing and multidisciplinary workflow.

Intratympanic or transtympanic STS is one of the most important emerging developments. It offers a plausible way to protect the cochlea from the start of cisplatin exposure while minimising systemic interaction, but current evidence is not paediatric and remains insufficient for routine clinical use. The field should therefore present it as a promising investigational route, not as an alternative standard.

Early detection should continue to rely on serial audiological monitoring, ideally including extended high-frequency audiometry and DPOAE when feasible. Serum biomarkers, including prestin, OTOLIN-1, microRNAs, extracellular vesicles and inflammatory signatures, remain investigational. Genetic susceptibility is real but not yet actionable as a stand-alone clinical screen. The future is likely to depend on integrated models that combine clinical, audiological, biomarker and genomic data.