An integration of physiological, morphological and biochemical assessments in studying guinea pig models of ototoxicity and otoprotection−a systematic review

Highlight

• •Using guinea pig models alongside a test battery can provide valuable insights into the mechanisms underlying ototoxicity and otoprotection.
• •This approach offers a detailed framework for elucidating ototoxic mechanisms and exploring therapeutic strategies for otoprotection.
• •This review encourages researchers to pursue basic research in this field.

Abstract

Purpose

This systematic review aims to explore the development and utilization of a comprehensive test battery comprising physiological, morphological, and biochemical assessments in guinea pig models to study mechanisms of ototoxicity and otoprotection.

Methods

This review was developed from peer-reviewed articles almost exclusively dealing with guinea pigs published in those journals listed on Journal Citation Reports.

Results

Initially, 560 articles were retrieved from 1994 to 2024. Following the guidelines (PRISMA 2020 statement) for reporting reviews, 54 relevant papers were ultimately selected.

Conclusion

Using animal models of ototoxicity and otoprotection, particularly the guinea pig can provide valuable insights into the mechanisms underlying ototoxicity and otoprotection in humans. This approach offers a detailed framework for elucidating ototoxic mechanisms and exploring therapeutic strategies for otoprotection, while also encouraging researchers to pursue basic research in this field.

Introduction

History background

Ototoxicity refers to the ability of certain therapeutic agents and chemical substances to damage the inner ear system, leading to hearing impairment and/or vestibular deficit by selected degeneration of the sensory receptors and neurons. The common culprits of ototoxicity include aminoglycosides identified in the 1940 s (Aran, 1995); loop diuretics recognized in the 1950 s (Ikeda et al., 1997); and chemotherapeutic agents, which became notable in the 1960 s and 1970 s. Among these offenders, aminoglycosides are most well-known for causing toxicity in the inner ear, with varying sensitivities between cochlear and vestibular partitions (Halsey et al., 2005).

Gentamicin, a member of the aminoglycoside family, is particularly known for its ability to selectively destroy vestibular hair cells (Halmagyi et al., 1994). Hence, intratympanic administration of gentamicin has emerged as the preferred approach for treating intractable vertigo in patients with Meniere's disease (Minor, 1999). Various agents have been proposed for otoprotection against ototoxicity, including ginkgo biloba (Krieglstein et al., 1986), d-methionine (Eastman, 1986), glutathione (Hoffman et al., 1988), salicylate (Sha and Schacht, 1999b), geranylgeranylacetone (Mikuriya et al., 2005), N-acetylcysteine (Chen et al., 2021), etc.

Rationale and objective

When evaluating the potential ototoxicity of a drug or its therapeutic efficacy against a known ototxic agent, it is essential to employ a comprehensive test battery integrating physiological, morphological, and biochemical assessments (Cheng et al., 2006, Day et al., 2007, Lo et al., 2017, Tan et al., 2001, Wu et al., 2024, Yang et al., 2011, Young et al., 2001). This comprehensive test battery offers researchers a detailed framework to deepen their understanding of the mechanisms driving ototoxicity and to investigate potential therapeutic strategies for achieving otoprotection. This systematic review aims to explore the development and utilization of a comprehensive test battery in guinea pig models for investigating ototoxicity such as chemotherapeutic agents and aminoglycoside. It highlights the integration of physiological, morphological, and biochemical assessments almost exclusively on the guinea pigs in studying ototoxicity and otoprotection.

Methods based on PRISMA

Study design

This review was conducted following the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020 statement) (Page et al., 2021). According to the systematic review methodology, the PICO question is defined as follows, and studies were selected:

• -P (Participants): Various guinea pig models of ototoxicity or otoprotection were enrolled.
• -I (Intervention): A combination of physiological, morphological, and biochemical assessments was utilized to evaluate the ototoxicity or otoprotection.
• -C (Comparison): Results from physiological, morphological, and biochemical examinations were correlated.
• -O (Outcome): Estimation of otoprotection by comprehensive test battery, so as to evaluate its efficacy.

Search strategy

The literature was initially surveyed by two independent reviewers through the PubMed database of the US National Library of Medicine, Scopus, and Google Scholar, focusing primarily on guinea pig models in English-language publications from 1994 to 2024.

The search terms from databases comprised “cisplatin”, “aminoglycosides”, “amikacin” “gentamicin”, “kanamycin”, and “streptomycin”. Additionally, drugs for otoprotection such as “d-methionine”, “ginkgo biloba”, “salicylate”, “methylprednisolone”, and “geranylgeranylacetone” were also included.

Inclusion criteria

• -Ototoxicity by cisplatin or aminoglycosides in guinea pigs
• -Otoprotection by d-methionine, methylprednisolone, geranylgeranylacetone, salicylate or extract of ginkgo biloba (EGb 761) in guinea pigs

Exclusion criteria

• -Studies lack of audiovestibular results.
• -Papers published in non-English language or non-SCI journals
• -Duplicate reports

Data analysis

In accordance with the framework of this review, two independent researchers carefully reviewed the titles and abstracts of the selected articles to assess their eligibility. The prevalence of abnormality rate of each test was compared among various ototoxic and otoprotective agents.

Certainty assessment

Quality assessment of the results was performed by two expert researchers. Disputes were resolved by discussion with a third researcher.

Initially, 560 papers were retrieved from 1994 to 2024. After excluding non-English reports (n = 73) and articles from non-SCI journals (n = 35), a total of 452 records were evaluated for eligibility. Subsequently, studies were excluded if they contained unrelated articles (n = 187), did not use animals (n = 18), or covered a similar topic published at a later date (n = 193). As a result, 54 relevant papers focusing mainly on guinea pig studies selected by two reviewers were ultimately included in the final selection.

Methods of assessment

Selection of animals

When selecting experimental animals for hearing measurements, the guinea pig model is often considered superior to the mouse model. Assessing hearing levels in each ear individually is feasible in a guinea pig model but not in a mouse model. This limitation arises because delivering sound stimuli exclusively to one ear in mice is challenging due to the close proximity of their both ears. Additionally, alert animals are often uncooperative during vestibular function testing, and handling rats or mice poses a risk of being bitten. In contrast, guinea pigs can tolerate a comprehensive inner ear test battery, comprising auditory brainstem response (ABR) for testing hearing, caloric testing for evaluating semicircular canal function, and the cervical and ocular vestibular-evoked myogenic potential (cVEMP and oVEMP) tests for assessing the saccular and utricular function, respectively (Young, 2018). Additionally, galvanic VEMP for checking the vestibular afferents. However, the ototoxicity and otoprotection literature contains relatively few studies evaluating vestibular function in guinea pigs, likely because such testing must be performed in awake animals, as general anesthesia abolishes nystagmus. Consequently, only a limited number of reports present caloric nystagmus, cVEMP, and oVEMP data in guinea pigs. In contrast, ABR testing to assess hearing can be conducted under anesthesia.

In contrast, the mouse model offers distinct advantages for studying cochlear ototoxicity because hearing-related genes have been extensively characterized in mice at the molecular level. Accordingly, mouse models are more commonly used in research on genetic disorders and aging, but are less frequently employed in studies of vestibular toxicity. This study was institutional review board approved (no.20050145, 20090265, 20100277), and was conducted in accordance with the guideline for the care and use of laboratory animals by the Animal Research Committee in National Taiwan University, College of Medicine.

Physiological assessment

ABR

Under intraperitoneal injection of pentobarbital sodium (35 mg/kg), one needle electrode was inserted into the ipsilateral retro-auricular area, another electrode was placed on the vertex, and the ground electrode was positioned on the neck. Click stimuli with a frequency range of 1–4 kHz under 0.1 ms were delivered. The repetition rate was 20/s and 400 sweeps were averaged to generate the ABR. The stimulus intensity started at 100 dB sound pressure level (SPL) and was gradually reduced in 10 dB increments until the waveforms became undetectable, enabling the determination of the ABR threshold (Fig. 1). If the ABR was not present at a 100 dB SPL stimulus, the intensity was increased to 135 dB SPL to confirm the absence of ABR (Young, 2018).

Fig. 1.

Auditory brainstem response (ABR) test in a healthy guinea pig shows the threshold of 50 dB SPL on both ears.

Bithermal caloric test

A pair of clip electrodes were attached to the outer canthi on both sides of an alert guinea pig, while a reference electrode was placed on the vertex. Each animal received irrigation of ice water (4°C, 5 mL) in the ear canal for 5 s. Both original wave and slow phase velocity of the caloric nystagmus (Fig. 2) were obtained by an electronystagmographic recorder, respectively (Young et al., 1991).

Fig. 2.

Gentamicin-treated guinea pig on the left ear. Normal caloric nystagmus is identified on the right ear, while absent response is noted on the left ear The upper trace is time base (1 s/scale), the middle trace is the eye movement, and the lower trace is the slow phase velocity (^o/s) of caloric nystagmus.

cVEMP test

A vigilant guinea pig was restrained with its head elevated and neck extended while positioned in a prone stance. A pair of needle electrodes were inserted into each neck extensor muscle, with a reference electrode positioned on the vertex. A plastic tube delivering click stimuli (duration 0.1 ms; intensity 120 dB SPL) was used to present stimuli monaurally, and the response was recorded unilaterally. The stimulation rate was 5/s. Each response was analyzed over a 24 ms period, with 100 responses averaged for each run. The biphasic waveforms, with latencies ranging from 6 to 9 ms, were designated as waves I and II, respectively (Fig. 3A). Two to three consecutive runs were conducted to verify the reproducibility of waves I and II, and cVEMP was deemed to be present. The mean latencies of the peak I and peak II for the guinea pigs were 7.24 ± 0.49 and 9.15 ± 0.47 ms, respectively, and the median peak-to-peak amplitudes was 11.9 (ranged, 6.1–16.9) µV (Yang and Young., 2005).

Fig. 3.

Gentamicin-treated guinea pig on the left ear; right ear, control. (A) Present cervical vestibular-evoked myogenic potential (cVEMP) (waveform l-II) on the right ear, and absent cVEMP on the left ear. (B) Present ocular VEMP (oVEMP, waveform nl-pl) on the left eye, and absent oVEMP on the right eye, indicating a cross response of oVEMP.

oVEMP test

Two active needle electrodes were bilaterally inserted into the inferior extraocular muscles of an alert guinea pig, while two reference electrodes were positioned 1 cm below the active electrodes, and the ground electrode was placed on the vertex. A minishaker was used to deliver consistent taps to the forehead of the guinea pig, producing a peak force level of 139 dB. Each response was analyzed over a 24 ms period, with 30 responses averaged for each run. The biphasic waveform at the latencies of 3–5 ms were termed nI and pI, respectively (Fig. 3B). Two to three consecutive runs were performed to verify the nI-pI waveform, and oVEMP was considered present based on the results. The mean latencies of nI and pI for guinea pigs were 3.17 ± 0.37 and 4.72 ± 0.38 ms, respectively (Yang et al., 2010a).

Galvanic VEMP test

A pair of electrodes for delivering galvanic stimuli was placed on the ipsilateral (cathode) and contralateral (anode) retroauricular regions, with a reference electrode positioned on the forehead. Galvanic stimuli with an intensity of 3 mA and a duration of 0.5 ms were applied. Responses to 128 galvanic stimuli were averaged for each run (Young, 2018). Since the original galvanic VEMP waveform contained electrical artifact, the responses obtained without neck muscle contraction were subtracted from those with neck muscle contraction to generate the final galvanic VEMP (Fig. 4).

Fig. 4.

Illustration of how the galvanic cVEMP is obtained in a guinea pig. To remove electrical artifacts, the responses without neck muscle contraction (A) are subtracted from those with neck muscle contraction (B), and a galvanic cVEMP waveform is derived after subtraction (C).

Morphological assessment

Light microscopy

After deep anesthesia with a 50 mg/kg intraperitoneal injection of sodium pentobarbital, the animals were perfused intracardially with saline solution, followed by a fixative containing 2.5% glutaraldehyde in 0.1 M phosphate buffer at pH 7.4. After complete fixation, the temporal bones were harvested and decalcified in a solution of 10% ethylene diamine tetraacetic acid (EDTA) containing 2.5% glutaraldehyde for one week. The tissue blocks were sliced horizontally into 200 μm thick sections. These sections were then dehydrated in an ascending ethanol series, infiltrated with propylene oxide, and finally embedded in an Araldite-Epon mixture. Semi-thin (1 μm) sections of saccular macula were cut and stained with toluidine blue (Day et al., 2007).

Electron microscopy

For electron microscopy, ultrathin sections (60–80 nm) of the specimen were cut, stained with 1% lead citrate, and then observed. A quantitative comparison of the densities of type I and type II hair cells, as well as supporting cells, in the otolithic organs was conducted. Montage images covering a field of 4,000 µm2 were used, with three randomly selected images from each ear. The percentage of damaged hair cells was calculated via dividing the number of damaged hair cells by the total number of respective hair cells in each montage image (Lue et al., 2009).

Confocal microscopy

Following deep anesthesia, the membranous labyrinth was carefully dissected as a whole-mount preparation from the guinea pigs. The cochlear and vestibular explants were fixated and stained with a conjugated rhodamine-phalloidin probe in phosphate-buffered saline (PBS). When the fluorescent dye rhodamine conjugates with phalloidin, it emits red fluorescence and labels F-actin. The tissues were then washed three times with PBS, mounted on glass slides with Fluoromount, and examined using confocal microscopy (Yang et al., 2011).

Biochemical assessment

Biochemical assay

Cochlear lateral walls of guinea pigs were harvested for analyzing Na⁺, K⁺-ATPase, Ca⁺²-ATPase, lipid peroxidation (LPO), and nitric oxide (NO) levels (Lin et al. 1999). Membrane ATPase activities were assessed by measuring inorganic phosphate release during ATP hydrolysis. Na⁺, K⁺-ATPase and Ca²⁺-ATPase activities were determined by subtracting ouabain-insensitive Mg²⁺-ATPase activity. The LPO was evaluated through malondialdehyde (MDA) detection, with absorbance at 586 nm used to quantify MDA concentrations. The NO levels were measured via a chemiluminescence assay, involving perchloric acid-treated tissue supernatants analyzed for NO oxidation products (Cheng et al., 2006).

Immunocytochemistry

The decalcified tissues block that contained vestibular ganglions were dehydrated and embedded in paraffin mixtures. Paraffin tissue blocks were sectioned at 7 µm thickness using a histomicrotome, and the sections were then processed through a series of procedures. The retrieval sections were treated with rabbit polyclonal anti-Nav1.8 antibodies, mouse anti-growth-associated protein-43 (1:6000, Sigma,MO, USA), rabbit anti-Substance P (1:1000, Peninsula, CA, USA), or neuropeptide Y (1:2000, Peninsula), then incubated and processed. Peroxidase activity was subsequently visualized using a Vector DAB Kit. Finally, the slides were mounted and examined under a light microscope (Cheng et al., 2010, Lin et al., 2010).

Results

1. Cisplatin

Ototoxicity by systemic application of cisplatin

Platinum-based agents are chemotherapeutic drugs containing platinum, such as cisplatin, carboplatin, and oxaliplatin, that treat various cancers by binding to DNA, causing cross-links that induce tumor cell apoptosis. Among them, cisplatin (cis-diamminedichloroplatinum II) has been recognized as an effective anti-neoplastic agent since 1971. Its toxic effects are cumulative and irreversible, comprising nephrotoxicity, peripheral neuropathy, gastrointestinal disturbance, and ototoxicity. Numerous studies have investigated the ototoxic effects of cisplatin in animal models, yet most have focused on single high-dose bolus injections of the drug, which does not align with treatment protocol in clinical patients. To address this, Stanford et al. (2021) proposed a dosing regimen of once-weekly injections of 4 mg/kg cisplatin for three weeks, designed to more accurately replicate clinical treatment protocol.

Morphologically, cisplatin has been shown to cause significant loss of outer hair cells at basal turn of the cochlea in guinea pigs and humans (Cheng et al., 2001; 2005). However, degenerative change was not identified at the vestibular partition, indicating that the cochlear partition is more susceptible to cisplatin-induced ototoxicity than the vestibular partition.

Otoprotection by d-methionine against cisplatin ototoxicity

Since the pathogenesis of cisplatin-induced ototoxicity involves oxidative stress (Ravi et al., 1995), protective strategies aimed at reducing free radical production have been employed to prevent hair cell loss and to minimize damage to the stria vascularis (Campbell et al., 1996). Otoprotective agent i.e. d-methionine has been reported to mitigate cisplatin ototoxicity (Cheng et al., 2005).

Exposure to cisplatin (5 mg/kg/d) for 7 days in guinea pig models results in an increased ABR threshold physiologically, which correlates with hair cell loss at basal turn of the cochlea morphologically. Biochemically, cisplatin significantly reduces the Na⁺, K⁺-ATPase and Ca⁺²-ATPase activities in the cochlear lateral wall, while promoting increased LPO and NO production (Heinrich et al., 2008), and upregulation of heat shock protein (Lo et al., 2017).

The otoprotective effect of d-methionine (300 mg/kg/d) against cisplatin (5 mg/kg/d) ototoxicity for 7 days in animal models could be attributed to its antioxidant properties, its ability to restore ATPase activities, and its influence on reducing cisplatin levels when co-administered (Ekborn et al., 2002, Cheng et al., 2006, Lo et al., 2015). Although systemic administration of d-methionine has been investigated for otoprotection against ototoxicity, its limited penetration into the inner ear raises major concerns regarding its effectiveness in preserving hearing.

Otoprotection: systemic vs. transtympanic administration

There are two routes for administering otoprotective agents: systemic and transtympanic. As systemic administration may attenuate the antitumor efficacy of cisplatin, transtympanic administration represents an alternative approach that minimizes systemic side effects and avoids potential interactions with cisplatin.

A critical step in transtympanic administration is the diffusion of otoprotective agents across the round window membrane. This process depends on both the physicochemical properties of the agents and the characteristics of the round window membrane (van den Berg et al., 2006). Therefore, it is essential to establish an effective local administration strategy that provides otoprotection without compromising the antitumor efficacy of cisplatin (Waissbluth, 2020, Tang et al., 2021).

Recently, nanoparticles have been introduced in various fields of medicine because of their ability to deliver drugs to specific target organs or cells. Most studies reported that nanoparticles are safe, without significant reduction in cell viability or evidence of ototoxicity (Barbara et al., 2022). Various nanotechnological approaches have been employed to improve drug pharmacokinetics and therapeutic efficiency, including encapsulation, polymerization, surface functionalization, and controlled drug release. Thus, these strategies enhance drug transport across the round window membrane, which is in evolution.

Otoprotection by locally applied dexamethasone against cisplatin ototoxicity

Antioxidants including thiol-containing compounds such as d-methionine, lipoic acid, N-acetylcysteine, and glutathione have been tested for their ability to scavenge free radicals and protect against cisplatin-induced ototoxicity (Rybak and Whitworth, 2005, Chen et al., 2021, Plane et al., 2025).

Cheng et al. (2005) reported that the elevated ABR threshold physiologically correlates with a decrease in both Na⁺, K⁺-ATPase and Ca⁺²-ATPase activities in the cochlear lateral wall of cisplatin-treated guinea pigs biochemically. Murphy and Daniel (2011) utilized intratympanic dexamethasone administration to offer otoprotection against cisplatin ototoxicity in guinea pig models. Guinea pigs that received intraperitoneal cisplatin (14 mg/kg) injection aligned with intratympanic dexamethasone (24 mg/ml) application, showed reduced ABR threshold at 8 kHz. Hence, intratympanic administration of dexamethasone appears to be an effective strategy for minimizing cisplatin-induced ototoxicity.

Additionally, dexamethasone has been shown to inhibit tumor necrosis factor-alpha induced cytokine secretion and inflammatory response in murine cultured fibroblasts (Maeda et al., 2005). Hence, combining physiological, biochemical, and morphological assessments in animal models may help evaluate the efficacy of otoprotective agents against ototoxicity.

Mechanisms of otoprotection against cisplatin ototoxicity

Integrating physiological (ABR threshold comparison and vestibular response), biochemical (enzyme activity measurement), and morphological (hair cell degeneration analysis) assessments in animal models may help evaluate the effect of otoprotectants against ototoxicity. Accordingly, G-protein–coupled receptors, such as adenosine A receptors and cannabinoid 2 receptors, have shown efficacy in animal models by promoting ROS scavenging, inhibiting ROS generation, and attenuating inflammatory responses. Inflammation may be triggered by activation of transient receptor potential vanilloid 1 (TRPV1) channels in the cochlea. Additionally, cisplatin-induced DNA damage and subsequent activating apoptotic pathways reveal potential therapeutic targets for preventing cisplatin-induced ototoxicity (Sheth et al., 2017). Furthermore, among mitochondria-targeted antioxidants, the protective effect of MitoQ against cisplatin-induced ototoxicity has been evaluated in animal models and/or auditory cell lines of mice (Fujimoto and Yamasoba, 2019).

2. Aminoglycosides

Ototoxicity by systemic application of aminoglycosides

Among members of aminoglycoside, some aminoglycosides show a predilection for cochlear toxicity, while others primarily affect the vestibular partition. Studies from animal models revealed that the ototoxic potential for gentamicin, tobramycin was similar and about 4 times higher than that of amikacin (Parravicini et al., 1982). The variations in ototoxicity among these aminoglycosides are influenced by several factors, namely, 1) dosage and duration of treatment, 2) half-life of drug, 3) renal function, and 4) the route of administration (Aran, 1995).

Ototoxicity involves two distinct phases. First, the binding of the positively charged aminoglycoside to the negatively charged cell membrane. Second, dynamic changes that occur within the cell after the aminoglycoside enters (Nakashima et al. 2000). Imamura and Adams (2003) compared gentamicin uptake and retention in guinea pig models and found no significant difference in cochlear degeneration between local and systemic administration, likely because gentamicin selectively damages vestibular hair cells while preserving cochlear function. Nevertheless, bilateral profound sensorineural hearing loss has been reported in a patient receiving gentamicin therapy, although the incidence is low (2–3%) (Moffat and Ramsden, 1977).

Otoprotection by salicylate against gentamicin ototoxicity

Aminoglycosides can chelate iron, forming a redox-active iron complex that promotes the production of free radicals. Song and Schacht (1996) proposed that sodium salicylate could protect hair cells from gentamicin-induced ototoxicity through two primary mechanisms. First, sodium salicylate acts as both an iron chelator and a free radical scavenger. Second, under conditions of oxidative stress, salicylate is oxidized to 2,3-dihydroxybenzoate, which serves as a highly effective iron chelator. However, salicylate neither affected serum gentamicin levels nor diminished its antibacterial efficacy (Sha and Schacht, 1999b).

Physiologically, co-therapy with salicylate significantly reduced gentamicin-induced ABR threshold shifts in guinea pig models, lowering a severe hearing threshold shift from over 60 dBSPL to less than 20 dBSPL. Morphological assessment confirmed that salicylate provided protection to cochlea hair cells of guinea pigs. In other words, salicylate, known for its iron-chelating and antioxidant properties, effectively prevents gentamicin-induced hearing loss in guinea pig models.

In contrast, sodium salicylate provided partial otoprotection against cisplatin-induced ototoxicity, specifically protecting the outer hair cells of the cochlea. Yet, this protection was limited, as it did not preserve the functional integrity of these cells (Hyppolito et al., 2006).

Local delivery of therapeutic agents

The local delivery of agents through the round window membrane has emerged as an effective approach for treating inner ear disorders in humans. Various therapeutic agents i.e., aminoglycosides, steroids, antioxidants, growth factors, and others, are locally applied to the round window membrane of guinea pigs to facilitate the entry of these agents into the inner ear (Gianoli and Li, 2001, Keithley et al., 1998, Korver et al., 2002). However, aminoglycosides (i.e., gentamicin, streptomycin and kanamycin) exhibit varying degrees of toxicity in the cochlear and vestibular sensory cells (Halsey et al. 2005). Therefore, prior to the clinical usage of a new agent, it is essential to conduct a comprehensive inner ear test battery in animal models to assess the potential for ototoxicity and to determine the optimal dosage for clinical usage (Yang et al., 2010b, Yang and Young, 2019).

Furthermore, mitochondria are the primary sources of ROS, and ROS-induced oxidative stress plays a critical role in cochlear damage. Among mitochondria-targeted antioxidants, the protective effects of MitoQ and SkQR1 against ototoxicity have been evaluated in animal models and/or auditory cell lines of mice. Either MitoQ or SkQR1 confers otoprotection against gentamicin-induced ototoxicity. However, it has not yet been evaluated in large-scale clinical trials (Fujimoto and Yamasoba, 2019).

Ototoxicity by local application of aminoglycoside eardrops

Aminoglycoside eardrops have been used to treat Gram-negative bacterial infections of the ears; however, their ototoxicity raises significant concerns (López-González et al., 1998, Sha and Schacht, 1999a). Guinea pig models were treated with three kinds of aminoglycosides—gentamicin, tobramycin, and amikacin—via local application onto the round window membrane to compare their relative ototoxicity. Physiologically, abnormality rates in the ABR, caloric test, and cVEMP test were 30%, 100%, and 100%, respectively, in gentamicin-treated guinea pigs (40 mg/ml). These findings correlated with more pronounced pathological changes in the vestibular partition than in the cochlear partition (Dobie et al., 2006).

In tobramycin-treated guinea pigs (40 mg/ml), ABR, caloric test and cVEMP test revealed 100%, 40%, and 30% abnormality rates, respectively, indicating that tobramycin has selective cochleotoxicity, with less effect on the vestibular partition. In amikacin-treated guinea pigs (12.5 mg/ml), abnormality rates of ABR, caloric test and cVEMP test were 30%, 40%, and 40% respectively, showing less toxic than gentamicin (Day et al., 2007). Thus, among aminoglycoside group, gentamicin is the most potent in inducing vestibular deficit and eliminating vestibular function (Dubie et al., 2006), while cochlear damage from gentamicin depends on the kinetics of local middle ear administration and the total drug dosage (Wagner et al., 2005).

Otoprotection by EGb761 against locally applied gentamicin ototoxicity

Gentamicin-induced ototoxicity occurs through the generation of ROS. Elevated ROS levels trigger signaling pathways that promote cellular apoptosis, ultimately damaging the hair cells of the inner ear. Gentamicin ototoxicity can potentially be prevented through the simultaneous administration of an anti-oxidant i.e. EGb761.

EGb 761 is supposed to protect the inner ear through multiple mechanisms, including scavenging free radicals, inhibiting NO synthesis, reducing LPO (Kobuchi et al., 1997), and prevent apoptosis. EGb 761 thereby neutralizes various free radicals, shielding tissues from oxidative stress (Oyama et al., 1996). Physiologically, in guinea pigs treated with gentamicin (40 mg/kg/d) applied to the round window membrane, the mean ABR threshold (68 ± 6 dBSPL) was significantly higher than the controls (38 ± 4 dBSPL). When EGb 761 (100 mg/kg/d) was administered orally to guinea pigs, hearing thresholds were significantly preserved (41 ± 3 dBSPL) compared to gentamicin-treated animals (Yang et al., 2011). Morphologically, EGb 761 exhibits a protective effect against ototoxicity in cochlear cultures derived from gentamicin-treated guinea pig models. EGb 761 accomplishes this by reducing the production of ROS and NO, while inhibiting apoptosis in cochlear hair cells. As a result, EGb 761 may help preserve more cochlear hair cells and improve hearing in gentamicin-treated guinea pig models.

Neurotoxicity by local application of gentamicin

Gentamicin has the potential to induce ototoxicity in vestibular receptors and their afferents. Physiological assessment demonstrated that absence of cVEMP in gentamicin-treated guinea pigs correlated with morphological findings, specifically damage to type I hair cells at the saccular macula observed by electron microscopy (Lue et al., 2009). Additionally, absence of galvanic VEMP was correlated with damage to the terminals of vestibular ganglion neurons. Biochemically, the proportion of voltage-gated sodium channel Nav1.8–like immunoreactive (Nav1.8-LI) neurons was significantly reduced in gentamicin-treated guinea pigs (Cheng et al., 2010). This finding is consistent with downregulation of Nav1.8 in peripheral nerve injury (Dib-Hajj et al., 1996).

Furthermore, biomarkers such as growth-associated protein-43 (GAP-43), a membrane- and cytoskeletal-associated phosphoprotein, and Substance P (SP) as well as neuropeptide Y (NPY) are expressed in damaged neurons (Schreyer and Skene, 1993, Hokfelt et al., 1994, Tsai et al., 2007). Double-labeling results revealed that 82% of SP-LI neurons and 16% of NPY-LI neurons co-expressed GAP-43, suggesting that SP accumulation coincided with a decrease in NPY expression in regenerating vestibular ganglion neurons following gentamicin treatment in guinea pig models (Lin et al., 2010).

Discussion

Synergistic ototoxicity vs. Otoprotection

Cisplatin-induced ototoxicity arises from multiple mechanisms, i.e., oxidative stress, inflammation, DNA damage, and apoptosis. Exploring optimal dosing strategies and potential interactions with other protective agents may improve its efficacy while reducing cisplatin-associated adverse effects, thereby providing a more comprehensive approach alleviating cisplatin-induced toxicity (Elmorsy et al., 2024).

Combination of chemotherapeutic agents such as cisplatin, carboplatin, and oxaliplatin can create increased synergistic ototoxic effects. In contrast, combination regimens pairing cisplatin with natural compounds i.e., garlic, curcumin, and ginger, have shown some promising outcomes (Agarwal, 1996, Wahab et al., 2020, Zadorozhna and Mangieri, 2021). Medicinal plants harbor numerous bioactive constituents with strong potential for the treatment of human diseases. Experimental data suggest that natural products may achieve therapeutic benefit by activating apoptotic processes through modulation of mitogen activated protein kinase and p53 signaling, while simultaneously enhancing chemosensitivity to cisplatin (Tomar et al., 2017). Moreover, by influencing key transcription factors, natural products can reduce cisplatin-induced organ toxicity while triggering cell death through apoptotic and/or necrotic pathways.

Clinical implication

Since the inner ear histopathology of guinea pigs closely resemble that of humans, they are frequently used to develop animal models for studying human inner ear disorders (Young, 2018, Yang and Young, 2019). Owing to the dedicated efforts of researchers over the past three decades, the guinea pig models for studying inner ear disorder have substantially advanced. Physiological assessment using an inner ear test battery comprising ABR and cVEMP, oVEMP, caloric and galvanic VEMP tests helps evaluate the function of cochlea, saccule, utricle, semicircular canals and vestibular afferents, respectively. Morphological assessment using light, electron, and confocal microscopy robustly identifies lesion sites and correlates well with physiological results. Biochemical assessment, including the measurement of biomarkers such as Na⁺, K⁺-ATPase, Ca²⁺-ATPase, LPO, NO, GAP-43, SP, NPY, and more, can help evaluate the extent of the lesion and provide insights into potential regeneration processes. By identifying the mechanisms of ototoxicity and otoprotection, it may be possible to develop strategies that mitigate toxicity and restore deficits. In summary, the use of guinea pig models of ototoxicity and otoprotection, together with a comprehensive test battery integrating physiological, morphological, and biochemical assessments, can yield valuable insights into the mechanisms underlying ototoxicity and otoprotection.

Potential value of small model organisms

The lateral-line organ of fish and amphibian senses water movement and vibration and plays a crucial role in schooling behavior as well as in detecting predators and prey. Aside from guinea pig models, the lateral-line system in zebrafish is increasingly adopted in ototoxicity research, which provides an ideal platform for examining stimulus encoding by sensory hair cells (Oliveira Cunha et al., 2025).

Intact zebrafish larvae were immobilized and mounted to allow optimal stimulation of lateral-line hair cells. The hair-cell microphonics —extracellular voltage changes generated by currents through mechanotransduction channels—are recorded (Trapani and Nicolson, 2010). In parallel, action potentials from single afferent neurons in response to water-jet stimulation of their associated hair cells are also measured. Collectively, these methods permit precise analysis of lateral-line sensory function in zebrafish using controlled, repeatable, and physiologically relevant stimulation. Zebrafish therefore represent a useful model for evaluating xenobiotic ototoxicity (Oliveira Cunha et al., 2025), probably owing to the morphological and physiological similarities between their superficial lateral-line hair cells and human inner ear hair cells. Consequently, zebrafish serve as a platform for screening otoprotective interventions against aminoglycosides and chemotherapeutic agents (Lim et al., 2024).

New approach methodologies in the evaluation ototoxicity pathways

New approach methodologies (NAMs) are innovative, animal-free tools that enhance the assessment of chemical and drug safety and efficacy. These approaches encompass human-relevant in vitro, in chemico, and in silico models and are designed to complement, rather than substitute, for conventional animal experiments. It focuses on three-dimensional (3D) in vitro human models, including tissue chips, organoids, and explants, collectively referred to as microphysiological systems. By more faithfully simulating human biology than animal models, these platforms provide greater predictive accuracy, reduced costs, and accelerated outcomes (Cogliati and Fox, 2024).

Progress in NAMs, together with advances on in vitro testing technologies, has stimulated the establishment of dedicated in vitro and ex vivo approaches for nonclinical safety evaluation. It focuses on case studies from recent projects in which animal models alone proved limited or inadequate for safety assessment. This collective effort aims to promote the use of NAMs for more human-relevant safety assessments, ultimately contributing to a reduction in animal testing in drug development, which is in evolution (Beilmann et al., 2025).

Conclusion

Using animal models of ototoxicity and otoprotection, particularly the guinea pig, can provide valuable insights into the mechanisms underlying ototoxicity and otoprotection in humans. This approach offers a detailed framework for elucidating ototoxic mechanisms and exploring therapeutic strategies for otoprotection, while also encouraging researchers to pursue basic research in this field.

Author contributions

Yi-Ho Young: correspondent, drafting, data analysis, final approval, accountability for all aspects of the work.

Po-Hsuan Wu: drafting, data analysis, accountability for all aspects of the work;

Hsing-Hua Liu: supervisor, drafting, data analysis, accountability for all aspects of the work.

Financial disclosure

None.

Availability of data and materials

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Funding

This study was supported by National Science Council, Taipei, Taiwan (Grant no. NSTC 113–2314-B418-004).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.