Intratympanic N-acetylcysteine in the prevention of cisplatin-induced ototoxicity: a systematic review and meta-analysis of randomized controlled trials

Abstract

Objective

To evaluate the efficacy of the otoprotective transtympanic application of N-acetylcysteine in preventing chemotherapy-induced ototoxicity in patients subjected to platinum-based chemotherapy.

Data sources

PubMed, Scopus, Web of Science, Cochrane Central, and ClinicalTrials.gov were searched for the following concepts: ((“Acetylcysteine” [Mesh]) AND (“Ototoxicity” [Mesh]) AND (“Cisplatin” [Objective: To evaluate the efficacy of otoprotective transtympanic application of N-acetylcysteine in the prevention of chemotherapy-induced ototoxicity in patients subjected to platinum-based chemotherapy. [Mesh]).

Study selection

Inclusion: randomized controlled trials, Exclusion: (1) case reports or case series; (2) thesis; (3) review articles; (4) conference abstracts; (5) animal studies; (6) non-english studies; (7) studies whose population was other than patients on platinum-based chemotherapy. Data extraction: changes in hearing thresholds measured by pure tone tympanometry, covering high and low frequencies: 250, 500, 1000, 2000, 4000, and 8000 Hz. We used RevMan (Review Manager) version 5.3 to conduct the meta-analysis and GRADE to assess the quality of the evidence. Data synthesis: The literature search yielded 277 unique articles. After reviewing six full-text articles, three RCTs provided data available for meta-analysis. A total of 88 cisplatin-based chemotherapy candidates were included for final analysis. Hearing thresholds showed a significant threshold difference between the ear treated with N-acetylcysteine and the control ear (ear not treated with N-acetylcysteine), especially at high frequencies as high as 8000 Hz (pooled effect size − 10.67, 95% CI [-12.33, -9.02], P < 0.00001). The data favored the Nac group in all frequencies as well, at 4000 Hz (pooled effect size − 2.13, 95% CI [-3.49, -0.77], P = 0.002), at 2000 Hz (pooled effect size − 1.38, 95% CI [-2.69, -0.06], P = 0.04), at 1000 Hz (pooled effect size − 1.58, 95% CI [-2.63, -0.53], P = 0.003), at 500 Hz (pooled effect size − 1.58, 95% CI [-2.62, -0.54], P = 0.003), and at the low frequency of 250 after solving the heterogeneity (pooled effect size − 0.96, 95% CI [-2.88, 0.95], P = 0.32).

Conclusions

Current data justifies the transtympanic administration of N-acetylcysteine for otoprotection in chemotherapy patients, minimizing the enduring consequences of cisplatin-induced ototoxicity and auditory impairment. Given the results’ emphasis on the Sarafraz et al. (2018) study, more randomized controlled trials are necessary with an expanded sample size and standardization of N-acetylcysteine concentration, study population, and assessed outcomes.

Introduction and background

Cancer remains a global health challenge, with millions of individuals worldwide undergoing chemotherapy as a primary treatment modality. While chemotherapy has revolutionized cancer care, several side effects that significantly negatively impact patients’ quality of life can also be associated with it. Among these side effects, cisplatin-induced ototoxicity is a formidable barrier to pursuing curative cancer treatment. Cisplatin, a potent and widely used chemotherapeutic agent, exhibits remarkable efficacy against various malignancies, including testicular, ovarian, and head and neck cancers. However, its clinical utility is marred by the risk of ototoxicity, which can lead to permanent hearing loss and vestibular dysfunction [1, 2].

Ototoxicity is a well-recognized consequence of cisplatin therapy, affecting patients of all ages, from pediatric to geriatric populations. The damage wrought by Cisplatin within the inner ear, specifically the cochlea and the vestibular system, can result in profound and irreversible hearing impairment [3], impacting communication, cognitive development in children, and overall quality of life. Moreover, vestibular disturbances can lead to imbalance and falls, compromising patient well-being. The incidence of cisplatin-induced hearing loss varies markedly among different age demographics. It is predicted to impact approximately one-third of adults and up to two-thirds of paediatric patients, with the affected range spanning from 15 to 65 dB at frequencies of 4000–8000 Hz. Rozencweig et al. documented a prevalence of roughly 30% in their research [2, 4].

Given the pivotal role of Cisplatin in cancer treatment, strategies to mitigate its ototoxic effects are of paramount importance [5]. Many interventions have been explored, from altering Cisplatin dosing regimens to developing otoprotective agents. One such candidate is N-acetylcysteine (NAC), a well-known mucolytic agent and antioxidant. NAC’s multifaceted pharmacological properties have made it an attractive candidate for mitigating cisplatin-induced ototoxicity.

NAC’s mechanism of action in the context of ototoxicity revolves around its ability to neutralize harmful reactive oxygen species (ROS) generated by Cisplatin. Oxidative stress within the cochlea is a crucial driver of cisplatin-induced hearing loss, and NAC, through its antioxidant capabilities, has the potential to counteract this oxidative insult. Furthermore, NAC’s mucolytic properties may enhance drug clearance from the inner ear, minimizing Cisplatin’s local toxicity [6].

Using NAC as an otoprotective agent, delivered directly to the inner ear via transtympanic application, represents an innovative approach to addressing cisplatin-induced ototoxicity. By bypassing systemic administration, transtympanic NAC aims to maximize local drug concentrations in the cochlea while minimizing systemic exposure, potentially reducing the risk of side effects.

However, the efficacy of transtympanic NAC in preventing cisplatin-induced ototoxicity remains a subject of debate within the scientific community. Previous studies that investigated this intervention had mixed results, so a thorough review of the evidence was needed to determine the actual effect of transtympanic NAC on keeping hearing while on platinum-based chemotherapy.

This systematic review and meta-analysis seek to provide a comprehensive, evidence-based evaluation of the otoprotective efficacy of transtympanic NAC in patients undergoing platinum-based chemotherapy. We intend to address essential inquiries regarding the effectiveness, safety, and therapeutic significance of this intervention by combining data from randomized controlled trials (RCTs) focused on it. Through a thorough examination of the current literature, we seek to offer doctors, researchers, and patients a deeper insight into the viability of transtympanic NAC as a method to protect patients’ hearing and enhance their overall cancer treatment experience.

Methods

We followed the preferred reporting of systematic reviews and meta-analysis (PRISMA statement) guidelines when reporting this manuscript [7]. This work was conducted by the authors in adherence to the Cochrane Handbook of Systematic Reviews of Interventions [8]. This study was prospectively registered on PROSPERO (CRD42023446111).

Criteria for considering studies in this review

Studies satisfying the following inclusion criteria were included in the systematic review:

• 1. Population: studies on patients liable to ototoxicity and hearing loss due to platinum-based chemotherapy, for example, Cisplatin.

• 2. Intervention: All doses were eligible in studies where the experimental group received N-acetylcysteine.

• 3. Comparator: studies where the control group received a placebo or standard treatment.

• 4. Outcome: studies reporting at least the changes in hearing thresholds.

• 5. Study design: studies described as a randomized controlled trial, where patients were assigned to the treatment groups in a random allocation method.

We excluded articles that were (1) case reports/case series, (2) thesis, (3) review articles, (4) conference abstracts, (5) animal studies, (6) non-English studies, and (7) studies whose population was other than patients on platinum-based chemotherapy.

Literature search and keywords

We searched PubMed, Cochrane Central, ClinicalTrials.gov, Scopus, and the Web of Science for relevant studies in July 2023. For a sensitive search strategy, we used MESH keywords. The keywords were N-acetylcysteine, dexamethasone, ototoxicity, hearing loss, transtympanic, intratympanic, Cisplatin, cisplatin-induced ototoxicity, cisplatin-induced hearing loss, and platinum-based chemotherapy. The search strategy for all databases was: ((“Acetylcysteine” [Mesh]) OR (N-Acetyl-L-cysteine) OR (N Acetyl L cysteine) OR (N-Acetylcysteine) OR (N Acetylcysteine) OR (Mercapturic Acid) OR (Acetylcysteine AL) OR (NAC AL)) AND ((“Ototoxicity” [Mesh]) OR (“Hearing Loss” [Mesh]) OR (Loss, Hearing) OR (Hypoacusis) OR (Hypoacuses) OR (Hearing Impairment) OR (Deafness)) AND ((“Cisplatin” [Mesh]) OR (cis Diamminedichloroplatinum) OR (Platinum Diamminodichloride) OR (cis Platinum) OR (Dichlorodiammineplatinum) OR (NSC-119875) OR (“Carboplatin” [Mesh]) OR (CBDCA) OR (J.M. 8) OR (NSC 241240) OR (“Oxaliplatin” [Mesh]) OR (ACT 078))

Screening and study selection process

We used Rayyan [9] for semi-automated screening of the literature search results. Studies were screened in two phases. The first phase was title/abstract screening for potential clinical studies. In the second phase, we retrieved the full-text articles of the selected abstracts for further eligibility screening. Two review authors (Rewan M. Ibrahim and Taif Al-Saraireh) independently conducted the literature search.

Data extraction

Data were extracted from a uniform online data extraction sheet for all included studies. Extracted data were divided into four domains: (1) study characteristics, (2) characteristics of the included studies’ population, (3) risk of bias domains, and (4) study outcomes.

Risk of bias assessment

We assessed the risk of bias in the included studies using the Cochrane risk of bias (ROB) tool. The Cochrane ROB tool examines the potential bias in seven study domains, including:

• 1. Random sequence generation.

• 2. Allocation concealment.

• 3. Blinding of the investigators and patients.

• 4. Blinding of the outcome assessors.

• 5. Incomplete outcome data.

• 6. Selective outcome reporting.

• 7. Other sources of bias.

In each domain, each study was tagged as “low risk,” “high risk,” or “unclear” after careful revision of the data presented in the published articles. We used the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) method to assess the certainty of the evidence for each outcome. Rewan M. Ibrahim conducted grade assessments.

Measures of treatment effect

Studies assess the impact of N-Acetylcysteine in patients liable for platinum-based chemotherapy ototoxicity and hearing loss. The primary outcome measure for this systematic review is the change in hearing thresholds measured mainly by pure tone audiometry covering high and low frequencies of 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz, and 8000 Hz. A reduction in pure-tone audiometry indicates symptom worsening. Other secondary outcomes may include adverse effects.

Evidence synthesis

Yoo et al. (2014) published the data for each patient. Using Jamovi version 2.3, the data was not normally distributed according to the Shapiro-Wilk test, so median and interquartile ranges of 25, 50, and 75 were calculated, together with each frequency’s maximum and minimum values.

In the case of the absent mean and S.D. in the published article, we calculated them from the median and interquartile ranges according to the methods of Wan et al. [10]. The standard deviation of change in audiometry between pre- and post-treatment was calculated assuming a weak positive correlation of 0.2.

We combined pure tone audiometry data to find the mean difference between the two study groups from the start of the study to the end, along with the 95% confidence intervals. This was done in a random effect meta-analysis model. We used RevMan (Review Manager) version 5.3 for Windows.

Publication bias

According to Egger and colleagues [11, 12], publication bias assessment is not reliable for less than ten studies. Therefore, in the present study, we could not assess the existence of publication bias by Egger’s test for funnel plot asymmetry.

Results

Results of the literature search

Two hundred seventy-seven records were obtained from the literature search. Rayyan identified 120 of them as duplicates. After excluding irrelevant abstracts, six articles were eligible for full-text screening. Three articles describing 3 RCTs were included in this systematic review and meta-analysis. The PRISMA flow diagram of the study selection process is shown in Fig. 1.

Fig. 1

Shows the PRISMA flow diagram of the study selection process

Characteristics of the included studies

All the studies included are randomized controlled trials in which candidates for cisplatin-based chemotherapy received transtympanic infusions of N-acetylcysteine in one ear. The N-acetylcysteine was diluted from 20 to 10% by adding Ringer lactate, except for (Yoo et al. 2014) study, where it was diluted to 2%. The other ear acted as a control in the (Riga et al. 2013) and (Yoo et al. 2014) studies, while (Sarafraz et al. 2018) used dexamethasone as a control. (Riga et al. 2013) evaluated the patients one month after the end of the cisplatin chemotherapy session; (Yoo et al. 2014) assessed them 1–2 months after chemotherapy, and (Sarafraz et al. 2018) evaluated them six months later. All the studies measured the hearing acuity using pure tone tympanometry. Riga et al. (2013) and Sarafraz et al. (2018) identified N-acetylcysteine as an efficacious otoprotective agent against cisplatin-induced ototoxicity, particularly at elevated frequencies (up to 8000 Hz), whereas Yoo et al. (2014) reported the overall effect to be insignificant, with the exception of two patients who exhibited favorable outcomes. A summary of the included study characteristics is shown in Table 1. The characteristics of the included studies’ populations are shown in Table 2.

Table 1 Shows a summary of included studies

Table 2 Shows baseline characteristics of the included studies

Results of syntheses

Primary outcome (change in pure tone tympanometry)

The overall Mean difference of change in the pure tone tympanometry at 8000 Hz favored the study group (pooled effect size − 10.67, 95% CI [-12.33, -9.02], P < 0.00001).

Pooled studies were homogenous (Chi² = 1.18, (P = 0.55); I² = 0%).

The overall Mean difference of change in the pure tone tympanometry at 4000 Hz favored the study group (pooled effect size − 2.13, 95% CI [-3.49, -0.77], P = 0.002).

Pooled studies were homogenous (Chi² = 0.28, (P = 0.87); I² = 0%).

The overall Mean difference of change in the pure tone tympanometry at 2000 Hz favored the study group (pooled effect size − 1.38, 95% CI [-2.69, -0.06], P = 0.04).

Pooled studies were homogenous (Chi² = 0.58, (P = 0.75); I² = 0%).

The overall Mean difference of change in the pure tone tympanometry at 1000 Hz favored the study group (pooled effect size − 1.58, 95% CI [-2.63, -0.53], P = 0.003).

Pooled studies were homogenous (Chi² = 0.22, (P = 0.89); I² = 0%).

The overall Mean difference of change in the pure tone tympanometry at 500 Hz favored the study group (pooled effect size − 1.58, 95% CI [-2.62, -0.54], P = 0.003).

Pooled studies were homogenous (Chi² = 1.97, (P = 0.37); I² = 0%).

The overall Mean difference of change in the pure tone tympanometry at 250 Hz did not favor either of the two groups (pooled effect size − 0.96, 95% CI [-2.88, 0.95], P = 0.32).

Pooled studies were not homogenous (Chi² = 2.42, (P = 0.30); I² = 17%).

To resolve the heterogenicity, we conducted a sensitivity analysis in multiple scenarios, excluding one study in each scenario. Heterogenicity was best resolved by excluding the study of Yoo 2014 (P = 0.69; I² = 0%). After removing Yoo et al. from the meta-analysis model, the overall Mean Difference favored the study group (MD − 1.45, 95% CI [-2.52, -0.38],P = 0.008).

To statistically test the robustness of evidence, we conducted a sensitivity analysis, removing one study in each scenario, to ensure that the effect size was not dependent on a single study. Excluding Sarafraz et al. 2018 study changes the overall effect estimate of variables, which may indicate weak evidence (need further confirmation using the GRADE assessment tool).

Quality assessment

Results of the risk of bias assessment showed that the quality of the included studies mostly ranged from moderate to low quality. A summary of the risk of bias assessment is shown in Fig. 2. No significant risk of bias was observed regarding selection bias (random sequence generation and allocation concealment), reporting bias (selective reporting), and other sources of bias. All the studies used “the internal control method,” where the included patients acted as an experimental and control group simultaneously, where one ear acted as experimental and the other acted as a control, to ensure fully matching comparisons in limited populations. (Yoo et al.2014) is a nonblind, open-label clinical trial. In (Riga et al. 2013), only 20 patients completed the study out of 24; in (Yoo et al. (2014), only 11 out of 13; and in (Sarafraz et al. (2018), only 57 out of 60. Missing outcome data were not handled to treat analysis.

Fig. 2

Shows a summary of the risk of bias assessment in the included studies

Grade assessment

Table 3 Grade assessments and evidence profile of different frequencies measured by pure tone audiometry in platinum-based chemotherapy patients

Discussion

Summary of the key findings

The study’s primary outcome is hearing acuity, measured with pure tone tympanometry covering low and high frequencies: 250, 500, 1000, 2000, 4000, and 8000 Hz. It showed that N-acetylcysteine protected the hearing acuity of patients undergoing chemotherapy. The treated ear represented fewer hearing thresholds than the control ear, which is especially evident at high frequencies (8000 Hz). The data of the pure tone tympanometry at all frequencies favored the N-acetylcysteine group.

Explanation of the study results

Cisplatin-induced ototoxicity is a well-documented and significant concern in oncology [13]. Damage to the inner ear structures, particularly the cochlea and vestibular system, distinguishes it. Cisplatin-induced ototoxicity has a multifactorial mechanism that includes oxidative stress, inflammation, and DNA damage within the inner ear cells [14, 15]. N-acetylcysteine (NAC) has precursors that synthesize the endogenous antioxidant glutathione [16, 17], and it has shown promise in protecting against a wide range of toxicity, including that caused by chemotherapy agents such as cisplatin [18].

An innovative method of reducing the negative effects of cisplatin-induced ototoxicity is to administer N-acetylcysteine (NAC), an otoprotective drug, directly to the inner ear via transtympanic administration. Transtympanic N-acetylcysteine (NAC) is a method designed to enhance medication concentrations in the cochlea while reducing systemic exposure, hence potentially alleviating side effects. This approach avoids systemic administration [19,20,21]. Transtympanic NAC delivers a high concentration of this antioxidant directly to the inner ear [22],18), which can counteract the oxidative stress caused by cisplatin, potentially reducing the severity of ototoxicity.

The scientific basis for the observation that high frequencies are frequently affected first in hearing loss is grounded in the anatomical and physiological characteristics of the human auditory system. This phenomenon is commonly referred to as “audiometric notch” or “noise-induced hearing loss,” where the high-frequency region of the audiogram is the first to exhibit hearing impairment. The primary reasons for this phenomenon include the basilar membrane mechanics: In the cochlea, its stiffness and width vary along its length [23]. High-frequency sounds are encoded near the basal end of the cochlea, where the membrane is narrow and stiff. This region is more susceptible to damage from loud noises or prolonged exposure to high-frequency sounds [24]. Cisplatin-induced ototoxicity is also based on cisplatin’s known pharmacokinetics and toxicodynamics [25]. Cisplatin, a potent chemotherapy agent, primarily targets sensory hair cells in the inner ear, which detect and transmit auditory signals. Because of their high metabolic activity and direct exposure to the drug via the bloodstream, these hair cells, particularly those in the cochlea responsible for high-frequency hearing, are particularly vulnerable to cisplatin-induced damage (14, 15). Because of this, when cisplatin enters the bloodstream during chemotherapy, it preferentially accumulates in the cochlea, causing an initial impairment of high-frequency hearing thresholds. This new scientific knowledge shows how important it is to use treatments like N-acetylcysteine to keep these sensitive sensory cells safe and patients’ hearing clear, especially at high frequencies, during and after cisplatin-based chemotherapy (18).

Clinical relevance of PTA threshold changes

The pooled effect size (mean change in PTA threshold) in our systematic review varies from 1.5 to 11 dB across different frequencies and is statistically significant; however, the clinical significance of these changes may be constrained. Clinically evident hearing impairment typically manifests as a change in pure tone audiometry thresholds over 25 dB, as indicated by toxicity grading criteria like CTCAE version 5.0 and the SIOP Boston Ototoxicity scale. Consequently, while our results suggest a protective effect of transtympanic NAC, the extent of the change may not attain the threshold generally linked to clinically significant hearing impairment. This constraint must be acknowledged when analysing the data, and further research should strive to evaluate both statistical and clinical significance to offer a more thorough assessment of the intervention’s effects.

Previous research on this topic has yielded mixed results, with some suggesting that NAC may protect against cisplatin-induced hearing loss, while others have found no significant benefits. Variations in study design, patient populations, and NAC dosing regimens could explain these discrepancies.

Some of the included studies have a limited sample size, such as (Riga et al. 2013), which had only 20 patients complete the trial, and (Yoo et al. 2014), which even had 11 patients. Additionally, the studies have slightly contradictory results.

This contributed to the significant heterogeneity in the evidence, which appeared at 250 Hz, and the dependency of the overall effect size on the Sarafraz 2018 study, which has a larger sample size (57) than Riga 2013 and Yoo 2014.

Agreement and disagreement with previous studies

Although Riga et al. 2013 reported a significant change in auditory thresholds at the 8000 Hz frequency between the N-acetylcysteine and the controlled ear, and Yoo et al. 2014 even reported no overall change except for two patients, the subsequent Sarafraz et al. 2018 study with a much larger sample size population revealed that the interventional group had better post-chemotherapeutic auditory thresholds than the control group.

Further studies investigating the correlation between the routine and duration of cisplatin administration could explain why, in (Yoo et al. 2014), some patients did not respond to NAC, and two excelled.

Strengths and limitations

Several drugs have been investigated for the prevention of chemotherapy-induced ototoxicity, such as Sodium Thiosulfate (X), Amifostine (X), D-methionine (X), Atorvastatin (X), etc. A significant amount of literature has been checked for efficacy and validity. To date, no systematic review or meta-analysis article has been published studying the otoprotective role of N-acetylcysteine in preventing cisplatin-induced ototoxicity.

All the included studies used “the internal control method,” in which the experimental group is the control group, to overcome baseline imbalances that may arise due to the type of malignancy treated, the cycles and doses of chemotherapy used, and the resulting toxicities. It guarantees fully matched comparisons in limited populations. Additionally, it was considered more ethical to leave one ear untreated to avoid the harmful results of unexpected side effects (Riga 2013).

The study focuses on the intratympanic application of N-acetylcysteine to ensure delivery’s control of the proper concentration of NAC and, subsequently, maximum effect to avoid interference with chemotherapy and possible systemic side effects.

Potential expectation bias

In the studies conducted by Riga et al. and Yoo et al., the design involved employing the same patients for both the intervention and comparison groups, with no placebo or active control. This design decision creates the potential for expectation bias, which may affect the study results in many manners.

Expectation bias in subjects

When patients recognise that they are undergoing an intervention, their anticipations of the treatment’s effectiveness may affect their subjective assessments of symptoms and outcomes. This is especially pertinent in research evaluating hearing thresholds because patient-reported outcomes may be affected by their perceptions of the treatment administered.

Expectation bias in outcome assessors

Likewise, outcome assessors cognisant of the treatment allocation may intentionally or unintentionally interpret the data to favour the intervention. This may happen even without deliberate prejudice, as nuanced signals and anticipations can affect the evaluation process.

Mitigation strategies

To alleviate this bias, subsequent studies should integrate a placebo or active control group and guarantee that both patients and outcome evaluators are blinded to the treatment assignment. Blinding mitigates the impact of expectations on study outcomes, resulting in more reliable and objective findings.

Current Study limitations

Although the present trials offer vital insights into the possible efficacy of transtympanic N-acetylcysteine, the absence of blinding and control groups constitutes a serious drawback. This constraint must be acknowledged when analysing the findings, and further research should endeavour to rectify these methodological issues to enhance the evidence foundation.

Some limitations are present in the review. The quality of the studies was questioned in the steps that could carry attrition bias, as they lacked the intention to treat the missing patients in each study.

Yoo et al. (2014) used only 2% of the NAC that animal studies recommended rather than 10% as in (Riga et al. 2013) and (Sarafraz et al. 2018), forgetting that humans have a much thicker oval window than guinea pigs (X).

Data on possible side effects was insufficiently reported. Riga reported only temporary pain (5 min) after intratympanic infusion of 10% NAC; Sarafraz denied tinnitus in the NAC group; and Yoo (2014) reported no incidence of local toxicity or pain associated with 2% L-NAC.

The significance of the work (implications)

Researchers have been engaged in a long-lasting debate seeking the most effective and safe otoprotective drug for high-risk patients treated with platinum-based chemotherapy. This study expands the literature by summarizing and synthesizing evidence from published RCTs on the efficacy of transtympanic application of N-acetylcysteine.

Authors’ conclusion

The transtympanic administration of N-acetylcysteine for otoprotection in chemotherapy patients is supported by current research, which helps to avoid the long-term consequences of cisplatin-induced ototoxicity and hearing loss.

Given the results’ reliance on the Sarafraz et al. (2018) study, more randomized controlled trials are necessary with an expanded sample size and standardization of N-acetylcysteine concentration, study population, and assessed outcomes.