The study included 24 cochleas from 12 guinea pigs. The subjects underwent examination for distortion product otoacoustic emissions (DPOAE) and brainstem auditory evoked potentials (BAEP) under anesthesia with intramuscular ketamine hydrochloride (40 mg/kg) and xylazine (10 mg/kg). Included subjects had to have DPOAEs and BAEPs. Guinea pigs with otitis, ear wax, and narrow ear canals that prevented the placement of an auditory catheter were excluded. The subjects were divided into four groups:

• •

  Group 1: Three guinea pigs (six cochleas) were given subcutaneous saline solution for eight days. Forty-eight hours after the last injection they were submitted to three sessions of HBOT (two ATA for 60 minutes) on days 11, 13, and 15 of the experiment.

• •

  Group 2: Three guinea pigs (six cochleas) were administered subcutaneous amikacin (600 mg/kg/day) for eight consecutive days.

• •

  Group 3: Three guinea pigs (six cochleas) were administered subcutaneous amikacin (600 mg/kg/day) for eight consecutive days. They rested for another seven days and were given subcutaneous saline solution.

• •

  Group 4: Three guinea pigs (six cochleas) were administered subcutaneous amikacin (600 mg/kg/day) for eight consecutive days. Forty-eight hours after the last injection they were submitted to three sessions of HBOT (two ATA for 60 minutes) on days 11, 13, and 15 of the experiment.

Auditory function assessment was based on DPOAEs and BAEPs. The devices used were the SMART DPOEA® and the SMART EP® from Intelligent Hearing Systems - Miami/Florida (USA).

Auditory function tests were carried out in the beginning of the experiment (in all groups), before HBOT sessions (groups one and four), before resting (group three), and immediately before the subjects were slaughtered (in all groups).

DPOAE tests followed the list of frequencies 2F(1) -F(2) at a ratio F(1)/F(2) = 1.22. Otoacoustic emissions starting at 1.5 kHz were considered. The size of the ear canals of the guinea pigs prevented emissions under this frequency from being assessed, as responses could be mistaken for noise10. As described by other authors, the intensity used in this study was 70 dB NPS for F(1) and F(2)2,16,17.

The hair in the vertex area and in the region between the orbits was shaved to aid in BAEP capture. The posterior portion of the pinna was cleaned with a steel sponge to remove oil and improve the contact between the electrode and the skin18. Surface electrodes were placed as follows: one positive electrode on the cranial vertex, two negative electrodes on the posterior portion of the pinna, and the ground wire on the forehead of the subjects, between their orbits. A layer of electrolytic paste was placed between the tip of the electrodes and the skin of the subjects to improve the conduction of the electric signals18.

Insertion earbuds were used to convey the sound stimuli. Alternated clicks of 0.1 ms were played at a rate of 27.7 stimuli per second. Fundamental frequency was 1,000 Hz at an intensity of 90 dB nHL. The signal captured by the electrodes was submitted to high pass and low pass filters of 150 Hz and 3,000 Hz respectively. Filtered amplified signals had their mean values calculated based on 1024 scans of BAEP records with a 12 ms window. The result was provided in the form of electric potential waves18. Only wave I was considered in this study, as it recorded the potential of the auditory nerve and is part of the peripheral auditory system18. Electrophysiological thresholds were calculated as sound stimuli were played in decreasing steps of 10 dB down to the approximated threshold.

HBOT sessions were carried in the Department of Experimental Surgery in a small-size multiple-subject experimental hyperbaric chamber. Three sessions were held in alternating days15. Each session lasted 100 minutes at 2 ATA, sixty minutes of which at absolute pressure15.

The guinea pigs were anesthetized as per the protocol and slaughtered at the end of the observation period by decapitation19.

The bulla tympani was opened and 2.5% glutaraldehyde slowly injected through an opening made in the apex of the cochlea and the round window; the specimens were immersed in this solution for four hours at 4°C9,15,17-20. After fixation in glutaraldehyde, the specimens were washed five times18 in 0.1 M phosphate buffer solution and dissected to expose the cochlear turns9,15,17-20. After washing, the cochleas were fixated once again in 1% osmium tetroxide and 0.1 M phosphate buffer solution for two hours at 4°C, and washed in 0.1 M phosphate buffer solution at a pH of 7.39,15,17-20. The specimens were dipped in ethanol in growing concentrations of 50%, 70%, 90%, and 95% for 10 minutes each to be dehydrated. Then they were dipped in 100% ethanol in three cycles of 20 minutes each, and in one last cycle of 12 hours in ambient temperature9,15,17-20. The water still present in the specimens after dehydration was removed through the carbon dioxide critical point drying method. A BAL-TEC CPD 030® (Critical Point Dryer) was used in the drying process. The ethanol in the specimens was removed by dipping them in carbon dioxide at 4°C; then the specimens were heated to 40°C to remove the carbon dioxide by evaporation, which happened at 31°C - the critical point9,15,17-20.

The dry specimens were mounted and glued to cylindrical metal stubs using carbon conduction paste9,17. In order to allow proper analysis and observation of structures on SEM, the cochleas were covered by a thin layer of gold using a BAL-TEC SCD 050 device. After preparation, the cochleas were electrically conductive for SEM9,16,17.

SEM images were analyzed for OHC counts in the basal, T2 and T3 turns of the cochlea and for presence, absence, or partial presence of inner hair cells in all turns of the cochlea and of outer hair cells in the apical turn of the cochlea.

The Friedman and the Wilcoxon tests were used to assess the differences between DPOAEs. Both tests were used because the study entailed random variables from the same subjects that changed with time. The tests were performed considering a significance level of 95%. The Friedman test was used to assess BAEPs with a significance level of 95%. Each cochlea represented a different, independent experiment, which reflected the temporal evolution of each cochlea. Morphological data was treated using the Kruskal-Wallis test with a significance level of 95%. This analysis considered each different turn of the cochlea, namely the basal, T2, T3, and apical turns, and analyzed them separately. The effects of amikacin and the offered procedures (HBOT and rest) were considered in each case.

Statistical analysis was performed using Microsoft Excel® and software package EXStat® for non-parametric tests21.

HBOT has been effectively used in other contexts to treat sudden deafness13 and offer protection against noise-induced auditory injuries15 and ototoxicity by cisplatin16.

The auditory protection provided by HBOT has not been clearly described. However, one of the mechanisms involved in amikacin ototoxicity may be related to hair cell ischemia and excessive production of oxygen reactive species accompanied by the typically associated injuries.

HBOT reportedly improves oxidative stress by increasing the production of antioxidants and antioxidant enzymes such as glutathione and superoxide dismutase (SOD), in addition to reducing the levels of markers of oxidative stress such as malondialdehyde and consequently promoting auditory protection by eliminating the oxygen reactive species produced by amikacin.

This study looked into the effects of HBOT upon ototoxic injuries introduced by amikacin to further the understanding of the pathophysiological mechanisms related to inner ear protection.

Ototoxicity was analyzed based on two parameters: morphology and function.

Studies on ototoxicity have routinely resorted to SEM to assess hair cell morphology9,10,15,17,19,23.

The data gathered in this study suggest that the subcutaneous administration of amikacin introduced morphologic alterations in the cochleas of the included guinea pigs. HBOT did not offer any protection against cochlear damage.

DPOAEs have been effectively used to assess OHC function15,16,23.

The functional findings related to amikacin ototoxicity reported in this study were in agreement with the findings published in other studies9,10,17.

Authors studying ultrastructural and functional cochlear alterations in animal models found extensive OHC lesions after systemic administration of amikacin for 12 consecutive days. These findings were supported by this study and corroborate the ototoxic potential of amikacin10.

The guinea pigs treated with amikacin and submitted to HBOT did not have DPOAEs after eight days of drug treatment and had reduced SNR at the end of the experiment, revealing a possible deterioration of OHC function. These findings were also seen in the group submitted to rest after amikacin administration.

Authors have suggested that the degree of DPOAE impairment seen in patients on aminoglycosides is directly correlated to the length of treatment4,24,25.

Based on the literature, one may assume that such functional deterioration could be the outcome of the cumulative effect of amikacin, as these findings were seen in the rest and HBOT groups.

Wave I thresholds were increased (> 90 dB nHL) in all groups given amikacin and remained altered after HBOT. Similar findings were seen in the rest group. Neither of the procedures produced statistically significant improvements, as supported by the morphologic findings seen in the subjects.

DPOAEs may be present in subjects with hearing loss of up to 50 dB NA26. The data gathered in this study, in which DPOAEs were absent, match the BAEP findings and show the devastating effects of the administered dosage of amikacin as reported in the literature on aminoglycoside-induced ototoxicity4,9,10,17,27-29.

This study failed to show the auditory protective effects of HBOT reported in the literature16; in the cited paper, the authors used intraperitoneal injections of cisplatin (8 mg/kg/day) administered for three consecutive days to induce ototoxicity, submitted the included guinea pigs to HBOT, and reported morphologic and functional improvements in the subjects’ outer hair cells. The number of animals used in this study is consistent with the literature9,10,15,16 and was based on the recommendations of the Research Ethics Committee; the differences found in our study cannot be attributed to this variable. However, it is possible that differences in metabolic pathways and modes of injury may have had a determining role.

The hyperoxia caused by HBOT increases tissue tolerance to ischemia and enhances the biologic response against oxygen reactive species. Future studies with similar methodology should consider different times of exposure to the same dosages of amikacin and the same HBOT protocol described in this study. Oxidative stress improvements secondary to HBOT by increases in the production of antioxidants and antioxidant enzymes such as glutathione and superoxide dismutase could be more thoroughly assessed, as well as reductions in oxidative stress markers such as malondialdehyde.