The protective effect of aspirin-induced temporary threshold shift in an animal model of cisplatin-related ototoxicity

The purpose of this study was to evaluate whether induction of temporary threshold shift (TTS) with aspirin prior to cisplatin exposure can prevent or minimize cisplatin detrimental effects on hearing. We randomly divided BALB mice into three groups: (1) cisplatin only, (2) aspirin only, and (3) combined aspirin/cisplatin. Cisplatin was administered as a single intraperitoneal injection of 14 mg/kg. Aspirin was administered for three weeks via intraperitoneal injection of 200 mg/kg sodium salicylate, twice daily. Air conduction thresholds were recorded using Auditory Brainstem Responses (ABR). Cochleae were harvested and cochlear hair cells were counted using a scanning electron microscope (SEM). Aspirin-induced TTS have reached an average of 30.05±16.9 dB after 2 weeks. At 60 days, cisplatin-only treated mice experienced an average threshold shifts of 50.7 dB at 4 kHz, 35.16 dB at 8 kHz, 70 dB at 16 kHz, 53.1 dB at 32 kHz. All threshold shifts were significantly worse than for cisplatin/aspirin treated mice with TTS of 11.85 dB at 4 kHz, 3.58 dB at 8 kHz, 16.58 dB at 16 kHz, 20.41 dB at 32 kHz (p < 0.01). Cochlear cell count with SEM has shown reduction in the number of both inner and outer hair cells in the mid-turn in cisplatin treated mice. Aspirin induced TTS can protect from cisplatin-induced ototoxicity. This beneficial effect was demonstrated by auditory thresholds as well as SEM. Larger pre-clinical and clinical studies are still needed to confirm these findings.


BID
Twice a day TID Three times a day

Background
Cisplatin is a commonly used chemotherapeutic agent (Rybak et al. 2007(Rybak et al. , 2009). Its main reported side effects are nephrotoxicity, peripheral neurotoxicity, and ototoxicity (Rybak et al. 2007(Rybak et al. , 2009). Cisplatin-induced hearing loss occurs as the result of cochlear damage (Lanvers-Kaminsky et al. 2017) which typically begins at the base of the cochlea and penetrates towards the apex with increasing cumulative exposure (Komune et al. 1981).
Cochlear damage includes injury to the stria vascularis (Lanvers-Kaminsky et al. 2017), spiral ganglion neurons (Boheim and Bichler 1985) inner and outer hair cells  as well as supporting cells (Laurell and Bagger-Sjoback 1991). The degree of cisplatin ototoxicity is dose dependent, but also varies, though to a lesser extent, with different treatment schedules and administration routes Laurell and Bagger-Sjoback 2000).
In clinical use, cisplatin causes a symmetric hearing loss in 69% of patients, measuring between 15 and 65 dB in the 4,000-to 8,000-Hz range, and in 100% of patients if extended high-frequency audiometry (> 8,000 Hz) is used. In general, hearing loss is permanent, with some sporadic and partial recovery (Moroso and Blair 1983).
As the mechanism of cisplatin ototoxicity lies primarily in oxidative damage to the outer hair cells, previous animal studies have evaluated the administration of antioxidants for preventing cisplatin-induced ototoxicity. Systemic administration of either l-or d-methionine before Cisplatin treatment reduced ototoxic and nephrotoxic side effects but were also associated with an unwanted partial tumor-sparing effect (Reser et al. 1999). Similar decreased chemotherapeutic efficacy of cisplatin has been reported for sodium thiosulfate co-treatment (Inoue et al. 1991;Aamdal et al. 1988).
Aspirin, and its active ingredient, salicylate, is a commonly used antipyretic, analgesic and anti-inflammatory drug. In contrast to forementioned antioxidants (methionine and sodium thiosulfate), aspirin did not inhibit the anti-tumor efficacy of cisplatin (Li et al. 2002;Crabb et al. 2017). Several studies evaluated the protective effect of aspirin on the ototoxic effect of cisplatin. Li et al. (Li et al. 2002) and Minami et al. (Minami et al. 2004) demonstrated aspirin protective effect on cisplatin ototoxicity in rats. Both used a short regimen of aspirin treatment and followed the animals for a short period of time and demonstrated significant although partial otoprotection.
Beyond its antioxidative effect, aspirin affects uniquely the inner ear. Continuous aspirin administration at high doses induces for itself hearing loss and tinnitus (Sheppard et al. 2014;McFadden and Plattsmier1983). However, its effect is reversible (Li et al. 2002;Minami et al. 2004;Surnar et al. 2018). In humans receiving high doses of 4 g daily, temporary threshold shift (TTS) of 10-40 dB develops and persists throughout the period of drug administration. Hearing thresholds typically returns to their pre-treatment level after aspirin administration is stopped (McFadden and Plattsmier 1983;Crifò 1975;Mccabe and Dey 1965). TTS induced by high doses of aspirin can theoretically be associated with reduced metabolic activity of cochlear elements and may be associated with improved resilience of the cochlea to damaging insults. Such resilience is similar to tissue preservation achieved in brain and cardiac surgeries using hypothermia to reduce metabolic rate (hibernation) (Engelman et al. 2015;Hong et al. 2014). Hypometabolic effect of hypothermia was found to be otoprotective as well. In a study by Spankovich et al., a reduction in Cisplatininduced hearing loss after cold water irrigation of guinea pigs external auditory canal was demonstrated (Spankovich et al. 2016). Moreover, hypothermia was effective also in preventing noise induced hearing loss as well as cochlear trauma induced by electrode insertion in cochlear implantation in animal models Hildesheimer et al. 1991;Tamames et al. 2016;Balkany et al. 2005). Hypothermia is not feasible as a treatment option for patients receiving cisplatin, however, a pharmacologic agent that can induce hypometabolism of the cochlea has the potential to protect from its ototoxicity.
Clinical studies have failed to demonstrate the hearingprotective effects of aspirin shown in pre-clinical studies. In a phase two, double-blind, randomised controlled trial in patients treated with cisplatin-based chemotherapy for multiple cancer types (Crabb et al. 2017), short pre-treatment with aspirin did not protect from cisplatin-related ototoxicity.
At present, there are no approved treatments to prevent ototoxicity in patients undergoing cisplatin-based chemotherapy.
Our hypothesis was that in addition to its protective antiinflammatory or anti-oxidative effect, high doses of aspirin may reduce cochlear activity, manifested as temporary threshold shift (TTS). Reduction in cochlear activity and metabolism may increase cochlear resilience to the damaging effects of cisplatin. We, therefore, explored in an animal model the protective effects of aspirin associated TTS against cisplatin induced ototoxicity, using auditory brainstem responses and scanning electron microscopy.

Animals
A total of 20, 8-weeks old male BALB/cJ mice with a weight ranging from 20 to 25 gm were randomly divided into three groups: (1) cisplatin only (n = 7), (2) aspirin only (n = 7), and (3) combined aspirin/cisplatin (n = 6). The BALB/cJ mouse strain was chosen due to a previous study by DeBacker et al., that have demonstrated cisplatin induced ABR threshold shift differences between mouse strains, sexes, and dosing schedules. In comparison with other strains, the BALB/cJ strain had lower mortality rates and marked hearing loss following a single cisplatin injection (DeBacker et al. 2020).
All mice were purchased from ENVIGO and thereafter shipped and housed in a vivarium maintained by Felsenstein Research Center of the Rabin Medical Center and Tel Aviv University. The mice were acclimated to the vivarium for 1 week prior to beginning of ABR analysis, aspirin and/or cisplatin exposures. Following cisplatin administration, all mice were single housed in an isolated area. All animals treated with salicylate or cisplatin were monitored daily for signs of distress, pain and weight loss of above 20% of baseline.
All housing and experimental procedures were reviewed and approved by Rabin Medical Center Institutional Animal Care and Use Committee (020,502).

Auditory brainstem response
Animals were anesthetized with inhalant isoflurane (5% for induction and 1.8% for maintenance of anesthesia with a 1 L/min oxygen flow rate).
Auditory brainstem responses (ABR) were measured using the SmartEP software, version 5.33 and stimuli was provided through the Opti-Amp 8002 and two high frequency transducers (Intelligent Hearing Systems, Miami, USA). A differential active needle electrode was placed subcutaneously below the test ear, a reference electrode below the contralateral ear, and a ground electrode at the vertex. ABR thresholds were recorded for each ear separately in response to 4, 8, 16, and 32 kHz. The stimulus consisted of 15 ms tone bursts, with a rise-fall time of 1 ms. The stimuli were presented to the external auditory meatus in a closed acoustic system through a tube connected to the high frequency transducer. Using a costume software, ABRs were averaged across one thousand and twenty-four tone presentations, delivered at 10 times per second. Hearing threshold was defined as the lowest intensity of stimulation that yielded a repeatable ABR response.

Cisplatin and salicylate exposure and ABR protocol
Cisplatin was administered as a single intraperitoneal injection of 14 mg/kg as previously performed by Hill et al. .
Aspirin was administered for three weeks via intraperitoneal injection of 200 mg/kg sodium salicylate (Sigma-St. Louis, MO), twice daily. In the combined aspirin/cisplatin group, aspirin was administered for 14 days prior and 7 days after cisplatin administration. Our protocol of long-term aspirin administration and dosage was selected to attain a sustained threshold shift (Hu et al. 2014;Myers and Bernstein 1965;Yu et al. 2008).
ABR recordings were performed at the following intervals: Aspirin-Cisplatin group: (a) baseline, prior to administration of both drugs, (b) immediately prior to cisplatin administration, after 14 days of aspirin treatment, (c) 7 days after cisplatin administration at the completion of aspirin treatment, (d) 30 days and (e) 60 days after cisplatin exposure. Cisplatin-only group: (a) baseline, prior to cisplatin exposure, (c) 7 days after cisplatin administration, (d) 30 days and (e) 60 days after cisplatin exposure. Aspirin-only group: (a) baseline, prior to administration of aspirin, (b) after 14 days of aspirin treatment, (c) at the completion 21 days of aspirin treatment, (d) 30 days and (e) 60 days after measurement b. Figure 1 presents the study protocol for aspirin and cisplatin administration and ABR measurements.

Scanning electron microscopy
Mice were sacrificed according to ethical guidelines and inner ears were dissected in cold PBS buffer and fixed in glutaraldehyde (2.5% v/v in PBS) overnight at 4 °C. The organ of Corti (n = 3 from each treatment group) was exposed, Fig. 1 Study protocol for aspirin and cisplatin administration and ABR measurements and the sample was treated with alternating incubations in osmium tetroxide and thiocarbohydrazide (OTOTO method;Hunter-Duvar, 1978). Treated samples were critical point dried and gold-coated at the Faculty of Life Sciences Electron Microscopy Unit at Tel Aviv University and imaged with a JSM 540A scanning electron microscope (SEM) (Jeol). Using the images compiled in Adobe Photoshop, hair cells were counted using the SEM images. Intact stereocilia bundles were counted directly and considered as one cell. In case of impaired or highly disrupted bundle cell was considered damaged and were not counted. Apex and Mid turn were each counted using three biological repeats with two fields of view acquired each at 340× magnification and averaged, while Basal turn was counted using two biological repeats with two fields of view per treatment in similar magnification.

Statistical analysis
Statistical analysis was performed with SPSS software, version 21.0 (Armonk, NY: IBM Corp.). All values are presented as mean ± SD. Comparison between means of non-parametric variables were compared with the Mann-Whitney test. Normal distributed variables were compared with student t test. A p value of < 0.05 was considered significant.

Results
Average auditory thresholds of the three study groups, in response to 4,8,16 and 32 kHz TB stimuli are presented in Table 1 and Figs. 2a-d respectively.
Comparison between 'cisplatin-only' and 'combined cisplatin/aspirin' groups demonstrated a significant otoprotective effect. Sixty days after cisplatin administration, cisplatin-only treated mice had an average threshold shifts of 50.7 ± 14.7 dB at 4 kHz, 35.1 ± 22.6 dB at 8 kHz, 70 ± 19.9 dB at 16 kHz and 53.1 ± 22.1 dB at 32 kHz, which were significantly worse than for cisplatin/aspirin treated mice that experienced an average threshold shift of 11.85 ± 10.89 dB at 4 kHz, 3.58 ± 8.1 dB at 8 kHz, 16.58 ± 16.57 dB at 16 kHz, 20.41 ± 16.84 dB at 32 kHz (p < 0.01). Comparison between aspirin-only and cisplatin/ aspirin groups demonstrated no significant difference in thresholds between the groups after 60 days. Figure 3 presents the thresholds changes for the three treatment group.
Cochlear scanning electron microscopy results are presented in Table 2 and Fig. 4. In comparison with mice from the aspirin-only group, mice treated with cisplatin had reduction in the number of both inner hair cells (IHC) and outer hair cells (OHC). The damaging effect of cisplatin was more complete at the basal turn, where the protective effect of aspirin was minimal. However, at the middle turn cisplatin caused partial destruction of hair cells, and aspirin was found to be protective with similar counts of both inner hair cells and outer hair cells in aspirin-only and aspirin-cisplatin groups (OHC 154,151 IHC 49,47,respectively) and better than in the cisplatin-only group (OHC 70; IHC 33).

Discussion
Cisplatin is an indispensable antitumor agent, and currently there is no effective treatment for preventing its toxicity to the inner ear. In the present study we have demonstrated that aspirin, in high doses of 200 mg/kg, twice daily, at a long-term administration of 21 days, produces a temporary threshold shift that is mostly reversible. Aspirin administration at these levels can significantly reduce cisplatin-related sensorineural hearing loss in BALB mice, and can protect from the associate damage to inner and outer hair cells.
In our animal model, we have effectively induced the ototoxic effect of cisplatin. Cisplatin detrimental effect on hearing was maximal at 30 days after exposure in all frequencies (average difference from baseline of 58.16 ± 9.48 dB across all frequencies). At the last measurement 60 days after cisplatin exposure, mice demonstrated threshold shift from baseline of 52.26 ± 14.78 dB (averaged for the four frequencies tested).
Aspirin effect has reached a maximal average threshold shift of 30.05 ± 16.9 dB (across four frequencies) after 21 days of aspirin treatment. This effect was almost completely reversible with a threshold shift from baseline of only 13.14 ± 14.31 dB after 60 days. Interestingly, threshold shift did not reverse after 30 days from end of aspirin administration (30.51 ± 6.94 dB) and continued to improve between 30 and 60 days. This gradual improvement in threshold after exposure to aspirin was different than the recovery period demonstrated by earlier studies in humans. (Mccabe and Dey 1965) were among the first studies that measured the effect of short-term systemic aspirin administration in humans. Their regimen consisted of 4 days period of 5 g per day of aspirin. Hearing loss recovery was noted 72 h after aspirin cessation in all subjects. A similar recovery period was noted in a study by (Myers and Bernstein 1965), that examined hearing loss in patients that were Fig. 2 Threshold shifts in dB SPL in 4, 8, 16 and 32 kHz for the three study groups along the study period from baseline to 60 days after cisplatin exposure. A Average ABR measure before aspirin administration, B Average ABR measure before cisplatin administration, C Average ABR measure 7 days after cisplatin administration, D Average ABR measure 30 days after cisplatin administration, E Average ABR measure 60 days after cisplatin administration administrated high aspirin dose of 6-8 g per day for a period that was not mentioned. A recent animal study by Huang et al. (Huang et al. 2005) used salicylate in high dose regimens and had shown a longer recovery period. They measured distortion product otoacoustic emissions (DPOAEs) in guinea pigs after two weeks administration of salicylate (200 mg/kg sodium salicylate, twice daily) and demonstrated a similar slow reduction of the distortion after salicylate cessation that returned to normal levels after four weeks. In the present study we have demonstrated threshold improvement between the 30-and 60-days measurements in all groups. These differences in the timing of threshold improvement may relate to a longer treatment regimen in our study that was intended to achieve maximal TTS. Previous studies examined the protective effect of shortterm aspirin treatment prior to cisplatin administration. Li et al. examined aspirin protective effect on cisplatin ototoxicity in rats (Li et al. 2002). In their study, they administered 100 mg/kg salicylate BID from 2 days before to 3 days after cisplatin treatment. Salicylate significantly attenuated the cisplatin-induced threshold shift from approximately 20 dB to 5 dB at 10 days post-treatment. Similar results were presented by Minami et al. (Minami et al. 2004) and Surnar et al. (Surnar et al. 2018). However, in a phase two, doubleblind, randomized controlled trial aspirin failed to protect from cisplatin ototoxicity (Crabb et al. 2017). As aspirin protocol regimen relied on the pre-clinical results by Li et al. (Li et al. 2002), cisplatin treated patients were randomized for a co-administration of 975 mg TID aspirin, one day prior and four days after cisplatin treatment. It is possible that the short-term administration of aspirin did not produce a temporary threshold shift.
Both aspirin induced TTS and ototoxic effect of cisplatin were observed in high and low frequencies suggesting that their effect involved different regions of the cochlea. A scanning electron microscopy images have shown severe hair cells damage at the basal turn compared with partial destruction at the middle turn. Basal turn hair cell loss combined with high frequency hearing loss were observed in previous studies. A study by Kamimura et al. has shown an extensive damage to outer hair cells, especially in the basal turn of the cochlea by Cisplatin (Kamimura et al. 1999). Similar basal turn damage with high frequency hearing loss was noted in a study by  and Hyppolito (Hyppolito et al. 2006).
At the last ABR recording, hearing threshold in the aspirin-only group has improved significantly. However, threshold at the end of the study in that group was increased by 13.14 ± 14.31 dB compared to baseline measurement. At that time point, mice were aged 5-6 months. The BALB/c mice strain is characterized by early hearing loss and therefore also serves as a model for age-related hearing loss. In a study by Szepesy et al. (Szepesy et al. 2021) a similar progression of hearing thresholds was noted in all frequencies. Willott et al. (Willott et al. 1998), demonstrated in BALB/c mice model that aging was associated with progressive loss of hair cells and spiral ganglion cells (SGCs), most prominently between 4 and 10 months of age. In addition, they have noted that although loss of hair cells was most severe in the cochlear base, aging BALB/c mice had relatively more SGC loss in the apex. In our study, we did not enroll a control group that received no treatment and therefore cannot determine whether threshold shift at 60 days after aspirin treatment is the result of aspirin exposure or aging.

Conclusions
This study demonstrated that aspirin at doses inducing auditory TTS can protect from cisplatin ototoxicity. The protective effect was manifested both in ABR recordings and in scanning electron microscopy findings. These promising results suggest that aspirin can be used as a protective agent in patients treated with cisplatin. Larger pre-clinical and clinical studies are still needed to confirm these findings.