Povidone-Iodine Fails to Eradicate Chronic Suppurative Otitis Media and Demonstrates Ototoxic Risk in Mice

Abstract

Hypothesis:

Commercially available povidone-iodine solution can eliminate biofilms and persister cells rapidly in in vivo achievable concentrations without inducing ototoxicity.

Background:

Chronic suppurative otitis media (CSOM) is a substantial global problem. Current treatment options often induce a temporary remission without leading to a permanent cessation of symptoms secondary to the treatments’ inability to eliminate persister cells. Povidone-iodine has been shown to be able to clear biofilm and planktonic cells in in vitro assays, but there are reports of ototoxic effects limiting its clinical utility.

Methods:

Bacterial and biofilm growth with quantification by spectrophotomer, murine auditory brainstem response (ABR), and distortion product otoacoustic emissions, immunohistochemistry, in vivo povidone-iodine treatment of murine CSOM, persister cell assay.

Results:

Commercially available 10% povidone-iodine solution is able to completely eradicate multiple clinical strains of Pseudomonas aeruginosa and Staphylococcus aureus in vitro with 10 minutes of exposure. Mice that have received a transtympanic injection of 1% povidone-iodine solution did not have significantly different auditory brainstem response or distortion product otoacoustic emission results compared with the control. Mice that received a povidone-iodine scrub or 10% povidone-iodine solution had significantly worsened hearing (25- and 13-dB increase in threshold, respectively; p < 0.05). In vivo CSOM infection recurred in all mice after the completion of treatment with 10% povidone-iodine solution, and there was no improvement in the bacterial load after treatment, indicating in vivo failure of therapy.

Conclusion:

Povidone-iodine solution is effective at eliminating biofilm and persister cells in vitro at in vivo achievable concentrations but fails in vivo most likely because of kinetics of distribution in vivo. Even if drug distribution could be improved, the therapeutic window is likely to be too small given that the diluted solution does not have ototoxic potential, whereas while the scrub variant, which contains detergents, and the undiluted solution are ototoxic after a single treatment.

INTRODUCTION

Chronic suppurative otitis media (CSOM), as defined by the World Health Organization, is a recalcitrant infection that leads to a chronically draining ear through a tympanic membrane perforation for at least 2 weeks, but oftentimes much longer (1). CSOM causes a significant worldwide health burden affecting up to 330 million individuals, with it accounting for almost 400 of 1,000 years of lost life (2). The disease is unevenly distributed, with it primarily affecting lower socioeconomic classes (3). In fact, it is the leading cause of sensorineural hearing loss in children in the developing world (4).

There has been little change in treatment modalities for CSOM over the last few decades partly related to its neglected tropical disease status (5). Current treatment regimens are focused on ototopicals antibiotics and antiseptics (6). Numerous studies have shown that these are “effective” treatments, but this is misrepresentative because the follow-up of these studies is too short to detect recurrence (7,8). When treated with ototopical antibiotics alone, a majority of patients will have a recurrence within 4 months of the initiation of treatment (1). These antibiotics are likely converting the disease from an active phase to an inactive phase but leaving residual disease behind to restart the process when treatment resolves (9).

This failure of treatment is multifactorial, but it is primarily driven by the continued presence of persister cells (10). These are metabolically inactive bacterial cells that are able to evade antibiotic-induced cell death, which oftentimes require cell division to exert their effect (11). Furthermore, repeated courses of antibiotics cause sublethal DNA damage within persister cells, which has been linked to the development of both in-class and out-of-class antibiotic resistance (12). This inability to eliminate the persister cells paves the way to the establishment of a chronic and resistant infection.

Previously, we showed that currently used ototopicals, acetic acid, boric acid, and ofloxacin fail to eradicate in vitro biofilms. All three agents failed at eliminating clinical isolates of Pseudomonas aeruginosa (PA) and/or Staphylococcus aureus (SA) strains at commercially available concentrations (13). Povidone-iodine was the only option that showed efficacy against planktonic and biofilm versions in both clinical and laboratory strains of PA and SA. Povidone-iodine is readily available, has a low cost, and is easily stored for long time periods (14,15). Povidone-iodine is already in use for treatment of CSOM in low-resource areas (16). However, there are questions regarding the safety of povidone-iodine near the inner ear because of ototoxicity (17,18). Therefore, we set out to test the efficacy of povidone-iodine against a number of clinically relevant strains of PA and SA in vitro and then looked to test in a validated model of CSOM, while also determining the ototoxic potential of povidone-iodine.

METHODS

Animals and Ethics Approval

All animal procedures were approved by the Institutional Animal Ethics Committee at Stanford University under APLAC 32855. Mice (6- to 8-wk-old CBA/CaJ) were purchased from or bred from stock acquired from Jackson Laboratory and housed in the Stanford University animal care facility with ad libitum access to food and water. Experiments were performed independently at least twice with three to five animals per group. Euthanasia was performed with carbon dioxide followed by secondary method of cervical dislocation.

Human Ethics Approval

Human bacterial samples from CSOM patients were collected under approval by the Institutional Review Board at Stanford University under IRB 43344. Bacterial samples of CSOM patients from India were imported under CDC PHS permit no. 20200213-0580A.

Bacterial Preparation, Biofilm Growth, and Quantification of Biofilm Matrix

The well-studied, fully sequenced and quorum sensing proficient PAO1, which is a laboratory reference strain originally isolated from an infected burn/wound of a patient in Melbourne, Australia (American Type Culture Collection ATCC 15692) and SA ATCC 25923, MSSA, which is a clinical isolate with the designation Seattle 1945, were used as the control strain in the experiments. The clinical isolates were collected from patients with CSOM attending the Ear Clinic in the Stanford Ear Institute, Palo Alto, CA; Dartmouth University, Hanover, NH; King George’s Medical University, Uttar Pradesh, India. The clinical CSOM isolates were kept stored in −80°C frozen glycerol stock culture at our CSOM culture bank. All the overnight cultures were prepared from the −80°C frozen culture stocks in Lysogeny broth (LB) and cultivated at 37°C with shaking at 200 rpm. Biofilms were prepared as follows: stationary overnight cultures of PA01, PA CSOM-1 to CSOM-14, SA 25923, and SA CSOM-1 to CSOM-28 were inoculated into wells of 96-well microtiter plates containing 150 μl LB medium. Experiments were performed in biological replicates of eight. Inoculated plates were incubated at 37°C without shaking for 24 hours. After incubation, the growth was confirmed at the optical density of 600 nm using a microplate reader (spectramMax M2; Molecular Devices, Downington, PA), after which the LB was removed from each well and the wells rinsed with phosphate-buffered saline (PBS) to remove planktonic cells. Crystal violet (CV) solution (0.1%, wt/vol) was applied and the microtiter plate was incubated at room temperature for 10 minutes, followed by PBS washing to remove the excess CV. Thirty percent acetic acid was added to each well to extract the CV stain. The absorbance of CV staining was read at 595 nm using the microplate reader.

Determination of the Minimum Inhibitory Concentration

The minimum inhibitory concentration (MIC) of the treatments were determined against laboratory strains of PA01 and SA 25923, 30 clinical CSOM isolates of PA, and 30 clinical CSOM isolates of SA using the broth microdilution method according to Clinical and Laboratory Standards Institute guidelines. In brief, the bacteria were grown overnight at 37°C in LB medium. Subsequently, the drug was mixed in LB using the serial dilution method (twofold), in a 96-well polypropylene microplate. The bacteria inoculum (10⁵ colony-forming units [CFU]/ml) was prepared using a 0.5 McFarland standard and was then added to each well and incubated at 37°C for 24 hours. The bacteria growth in the presence of the treatments was evaluated by visual observation of the solution (clear or turbid) in the wells. The MIC was obtained from the lowest concentration of the drug, which showed a clear solution.

Determination of Minimum Biofilm Eradication Concentration

Overnight cultures of PA01, SA 25923, PA clinical isolates, and SA clinical isolates were diluted 1:1,000 in fresh LB medium and 150 μl of the dilution per well were added in a Minimum Biofilm Eradication Concentration (MBEC) Assay Biofilm Inoculator (Innovotech, Edmonton, Canada) with 96 wells. Biofilms were allowed to develop onto peg lids for 48 hours without shaking. The peg lids were gently rinsed to remove planktonic bacteria and incubated in a new MBEC Assay Biofilm Inoculator with 96 wells containing a serial dilution (twofold) of tested drug in PBS. These MBEC Assay Biofilm Inoculator with 96 wells were incubated for 24 hours. Then the peg lids were washed and sonicated for 15 minutes in a new MBEC Assay Biofilm Inoculator with 96 wells containing fresh medium (recovery media). The recovery media was allowed to incubate for 48 hours. The MBEC value was defined as the lowest concentration of drug where the optical density at 600 nm is less than 0.1.

In Vivo Bacterial Inoculation

Our CSOM model has been validated and previously described (19). Briefly, 10 μl of bacterial persister cells was inoculated through a subtotal tympanic membrane perforation. Stationary phase culture (30 h old) of PA01 was treated with 0.3% ofloxacin otic solution (final concentration of 5 μg/ml) for 5 hours. After which, the bacteria were washed three times in PBS and the CFU/ml was determined. For PA01, reproducible infections were obtained with an inoculum of 1.0 × 10⁷ CFU/ml. Control mice received equal volumes of sterile PBS.

Serial Passage Resistance Induction Studies

Serial passage MICs were performed in 96-well microtiter panels. First, an aliquot of the well with the highest concentration permitting growth was taken and back diluted in fresh media to a turbidity of a 0.5 McFarland standard from inoculated microtiter panels. After overnight incubation at 37°C, this suspension was then further diluted (1 of 100) and used to inoculate a fresh MIC panel resulting in a final concentration of 1.5 × 10⁶ CFU/ml. Panels were incubated according to CLSI guidelines, MICs were recorded, and the next inoculum was prepared from the well containing the highest concentration of drug that allowed growth in an identical fashion as described previously. Fourteen repeat passages were performed.

Statistical Analysis

Statistics were performed using GraphPad Prism 9.0 (GraphPad Software, Inc., La Jolla, CA). p Values were calculated using two-tailed t test with Welch’s correction, a two-tailed Mann–Whitney test, or an analysis of variance with data considered significant when values were less than 0.05 as indicated: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001.

See supplemental methods for discussion of in vivo techniques, http://links.lww.com/MAO/B505.

RESULTS

MICs of Povidone-Iodine for Clinical Bacterial Strains

The MICs of povidone-iodine for a large number of PA and SA strains isolated from distinct CSOM patients are shown in Table 1. These values are consistent with our previous work looking at the MIC value of a single laboratory and clinical strain of PA and SA (13). Once again, clinical strains had a higher MIC than the laboratory strain PA01. No strains are able to continue to grow in the presence of povidone-iodine. The highest MIC of a PA or SA strain tested was 9,600 μg/ml, which is well below the topically applied concentration of povidone-iodine of 100,000 μg/ml.

TABLE 1.

Minimum inhibitory concentration of chronic suppurative otitis media clinical isolates against povidone iodine (n = 3 for each strain)

Species	Strain Description	Povidone-Iodine MIC (μg/ml)
P. aeruginosa	PA-CSOM-1	9,600
P. aeruginosa	PA-CSOM-2	9,600
P. aeruginosa	PA-CSOM-3	9,600
P. aeruginosa	PA-CSOM-4	9,600
P. aeruginosa	PA-CSOM-5	9,600
P. aeruginosa	PA-CSOM-6	9,600
P. aeruginosa	PA-CSOM-7	9,600
P. aeruginosa	PA-CSOM-8	9,600
P. aeruginosa	PA-CSOM-9	9,600
P. aeruginosa	PA-CSOM-10	9,600
P. aeruginosa	PA-CSOM-11	9,600
P. aeruginosa	PA-CSOM-12	4,800
P. aeruginosa	PA-CSOM-13	4,800
P. aeruginosa	PA-CSOM-14	9,600
S. aureus	SA-CSOM-1	4,800
S. aureus	SA-CSOM-2	4,800
S. aureus	SA-CSOM-3	9,600
S. aureus	SA-CSOM-4	9,600
S. aureus	SA-CSOM-5	9,600
S. aureus	SA-CSOM-6	9,600
S. aureus	SA-CSOM-7	9,600
S. aureus	SA-CSOM-8	4,800
S. aureus	SA-CSOM-9	9,600
S. aureus	SA-CSOM-10	9,600
S. aureus	SA-CSOM-11	4,800
S. aureus	SA-CSOM-12	4,800
S. aureus	SA-CSOM-13	4,800
S. aureus	SA-CSOM-14	4,800
S. aureus	SA-CSOM-15	4,800
S. aureus	SA-CSOM-16	4,800
S. aureus	SA-CSOM-17	4,800
S. aureus	SA-CSOM-18	4,800
S. aureus	SA-CSOM-19	4,800
S. aureus	SA-CSOM-20	4,800
S. aureus	SA-CSOM-21	9,600
S. aureus	SA-CSOM-22	4,800
S. aureus	SA-CSOM-23	4,800
S. aureus	SA-CSOM-24	4,800
S. aureus	SA-CSOM-25	4,800
S. aureus	SA-CSOM-26	4,800
S. aureus	SA-CSOM-27	4,800
S. aureus	SA-CSOM-28	4,800

CSOM, chronic suppurative otitis media; MIC, minimum inhibitory concentration; PA, P. aeruginosa; SA, S. aureus.

MBECs and Minimum Duration of Killing of Povidone-Iodine for Clinical Bacterial Strains

The MBECs of povidone-iodine for a large number of PA and SA strains isolated from distinct CSOM patients are shown in Table 2. Traditionally, this experiment is performed with 24-hour exposure to the drug of interest. The biofilm was eradicated in all strains tested and was at the lowest concentration we could test in nearly all the strains.

TABLE 2.

Minimum biofilm eradication concentration of chronic suppurative otitis media clinical isolates against povidone-iodine (n = 3 for each strain)

Species	Strain Description	Povidone-Iodine MBEC (μg/ml)
		10-min Exposure	24-h Exposure
P. aeruginosa	PA-CSOM-1	6,250	1,500
P. aeruginosa	PA-CSOM-2	3,000	<800
P. aeruginosa	PA-CSOM-3	5,000	1,500
P. aeruginosa	PA-CSOM-4	<800	<800
P. aeruginosa	PA-CSOM-5	<800	<800
P. aeruginosa	PA-CSOM-6	<800	<800
P. aeruginosa	PA-CSOM-7	<800	<800
P. aeruginosa	PA-CSOM-8	<800	<800
P. aeruginosa	PA-CSOM-9	<800	<800
P. aeruginosa	PA-CSOM-10	<800	<800

CSOM, chronic suppurative otitis media; MBEC, minimum biofilm eradication concentration; PA, P. aeruginosa; SA, S. aureus.

The Minimum Duration of Killing is a framework for understanding bacterial killing achieved by the relationship between drug concentration and exposure time (20). It is not clinically viable to expose a patient to a topical substance for 24 continuous hours. Therefore, we attempted to see if the time frame could be shortened to match what can be reasonably achieved in an office visit. Ten-minute exposure was able to eradicate biofilms in all strains with the majority still being at the lowest concentration possible (Table 2). The few strains that did require higher concentrations for biofilm eradication at this shortened time frame still was more than a magnitude lower than the achievable concentration with povidone-iodine solution.

Povidone-Iodine Does Not Generate Resistance by Serial Passaging In Vitro

Figure 1 displays serial passage of povidone-iodine compared with ofloxacin comparing the MIC at each serial passage. Compared with ofloxacin, where serial passage induces resistance, no resistance was shown in the povidone-iodine group after 14 passages.

FIG. 1.

Povidone-iodine does not induce resistance in PAO1 after repeated serial passage, whereas, in comparison, ofloxacin rapidly induces resistance (n = 3 replicates).

Povidone-Iodine Has Manageable In Vivo Toxicity

To investigate the cytotoxicity of povidone-iodine, we exposed RAW 264.7 cells to varying concentrations of povidone-iodine to determine cell viability. Concentrations as low as 10% (10,000 μg/ml) reduced cell viability 50% compared with the control solution (Fig. 2). This reduction was found to be a significant difference (p < 0.0001). There was no dose–response relationship, with higher concentrations having the same effect on cell viability.

FIG. 2.

Comparison of cellular viability after exposure to varying concentrations of povidone-iodine. There was a significant reduction cellular viability after exposure to povidone-iodine compared with the control (****p < 0.0001). There was no difference in the cellular viability between any of the tested concentrations of povidone-iodine (n.s., p > 0.05). N = 8 replicates.

We further examined for in vivo ototoxicity, using distortion product otoacoustic emission and auditory brainstem response, in mice who received a single transtympanic injection of a commercially available povidone-iodine solution, a 1:10 diluted version of that solution, a povidone-iodine scrub, or saline (n = 5 for each group; Fig. 3). Povidone-iodine scrub was utilized because it contains a high concentration of detergent, and there are earlier reports describing a high toxicity from the combination of povidone-iodine with detergents (21). The average threshold for the undiluted povidone-iodine solution was elevated compared with the saline and reached significance at one frequency on the distortion product otoacoustic emission (p < 0.05), whereas povidone-iodine scrub showed threshold elevation compared with saline (p < 0.05). However, diluted povidone-iodine solution was not significantly different and more closely overlapped with the saline thresholds (p > 0.05).

FIG. 3.

DPOAE and ABR for mice (n = 5 per group) treated with an intratympanic injection of saline, full-strength povidone-solution, 1:10 dilution povidone-iodine solution, and povidone-iodine scrub. Thresholds differed across three frequencies, 11.3, 16, and 23, by treatment for the DPOAE (ANOVA, *p < 0.05). By Tukey post hoc pairwise comparison across frequencies, povidone-iodine scrub differed from the other group at all three frequencies and the full-strength povidone-iodine solution at one frequency; 16 (*p < 0.05) thresholds are plotted as means ± standard error. Thresholds did not differ significantly by treatment for ABR (ANOVA, p < 0.05). ABR indicates auditory brainstem response; ANOVA, analysis of variance; DPOAE, distortion product otoacoustic emission.

Povidone-Iodine Fails to Eradicate PA in an In Vivo CSOM Animal Model

Finally, we examined the in vivo efficacy of povidone-iodine solution in our CSOM animal model. Fifteen mice had their middle ears inoculated with PAO1, and after the infection had been established, five mice were each treated with saline, full strength povidone-iodine, or 1:10 dilute povidone-iodine solution twice a day for 14 days. During the course of treatment, the otorrhea resolved in the povidone-iodine groups; however, 28 days after cessation of treatment the otorrhea returned in every mouse. To confirm that this was not a sterile otorrhea, we used in vivo imaging system, which capitalizes on the bioluminesce of our PAO1, which showed the presence of PA in the middle ears of all the treated mice (Fig. 4). In addition, we measured the number of colony-forming units from the otorrhea and there was no significant difference between the amount from the saline-treated group (5.7 log₁₀ CFU) compared with both povidone-iodine groups (5.8 log₁₀ and 5.6 log₁₀ CFU; p > 0.05; Fig. 4). This demonstrates that in this in vivo animal model, povidone-iodine fails to eradicate CSOM.

FIG. 4.

Povidone-iodine does not reduce bacterial load in the long term. A, Experimental schema. Fourteen days after the establishment of infection, mice (n = 5 per group) were treated with saline, povidone-iodine solution, or povidone-iodine scrub. B, At day 42, there is no significant difference in bacterial counts in effusions (p > 0.05). C, Representative imaging of IVIS confirming active bacterial infection. IVIS indicates in vivo imaging system.

DISCUSSION

CSOM continues to be a difficult disease to manage successfully with ototopical drugs (22). This is related to the complex interaction of bacterial resistance, biofilms, and persister cells (6). Our previous work raised the possibility that povidone-iodine could be an effective treatment (13).

Many clinicians use in vitro MIC to direct in vivo therapy (23). MIC is defined as the lowest concentration where a compound is able to inhibit visible growth of planktonic bacteria, before being then compared with guidelines to determine if the bacteria are sensitive, intermediately sensitive, or resistant (24,25). The guidelines are based on achievable drug concentrations within the serum without consideration for higher drug concentrations that can be reached locally. This has driven clinicians to treat CSOM with ototopicals drugs that the bacteria are technically resistant to which is a reasonable consideration if the objective is to inhibit growth; however, chronic bacterial infections require eradication, not temporary inhibition. MIC fails to explicitly look at the ability of a drug to eradicate persister cells within biofilms that are driving the recurrence in CSOM (9,13). MBEC is a validated test used to determine the minimum concentration necessary to eliminate biofilms, including persister cells (26). Interestingly, the MBECs were lower than the MIC for nearly all the strains for povidone-iodine, which is different from that seen with fluoroquinolones (13).

Povidone-iodine continues to be an excellent antimicrobial when tested in vitro. The MIC and MBEC for all the clinically derived strains of PA and SA were low and readily achievable. There is abundant evidence that repeated exposure to antibiotics can lead to the development of resistance to the drug in question, other members of the class, and even unrelated drugs (27). Unlike fluoroquinolones, which are well known to have resistance develop to the class, there was no evidence of resistance developing in the face of repeated exposures of povidone-iodine (28). This would imply that the bacterial target that povidone-iodine acts through to exert its cell killing effect is one that cannot be modified. These are the exact characteristics one would look for in developing a treatment for CSOM.

Unfortunately, this study has also demonstrated significant limitations to povidone-iodine being utilized regularly to manage CSOM in the clinic. Our study raises the safety concern for middle ear exposure for povidone-iodine with the knowledge that the safety of topical antiseptic agents has been in dispute for decades (29). We showed changes in cell viability and hearing after exposure to povidone-iodine. The hearing changes may be able to be managed with using a low concentration, 1% or less, and avoiding any detergents in the solution; however, there did not seem to be a concentration that did not lead to a change in cell viability. This is distinctly different from that seen in fluoroquinolones, which have a more classical dose–response effect to exposure (30). Povidone-iodine induces cell death by indiscriminately oxidizing proteins, nucleic acids, and fatty acids (31). Because this is not a specifically prokaryotic targeted mechanism, in contrast to the action of most antibiotics, it is somewhat unsurprising that eukaryotic cells may also be impacted by povidone-iodine.

The precise impact of povidone-iodine on the human inner ear remains ambiguous. There is only one case report in the literature of a directly attributed ototoxic event in humans related to povidone-iodine usage, whereas a number of surgeons do use povidone-iodine as their preferred preoperative scrub for otologic surgery (17). In contrast, there are a large number of animal studies reporting variable degrees of ototoxicity similar to what we have reported (18,21). Whether routine povidone-iodine usage is safe has not been settled and likely never will because of issues of clinical equipoise and the natural differences between humans and animal models.

Even more importantly, however, is the lack of efficacy of povidone-iodine at curing CSOM in our in vivo animal model. Similar to our findings of in vivo use of fluoroquinolones (13), our mice had their disease temporarily convert to an inactive state before returning to an active draining state with no change to the underlying bacterial load. This mirrors closely what is seen in humans where most patients can achieve short-term disease quiescence, whereas most will also see recurrence when followed up in the medium term (1,16,32).

The exact cause for the failure of povidone-iodine to resolve the in vivo infection in our animal model is unclear. There are a variety of possibilities that could explain the dichotomy between the in vitro and in vivo efficacy. Povidone-iodine has a viscosity as high as 85.9 cP, which is similar to the value seen for olive oil (33). In contrast, fluoroquinolones have such low viscosity at baseline that methylcellulose needs to be added in to allow it to be easily administered from a dropper (34). This higher viscosity may lead to poor distribution of povidone-iodine within the middle ear. A single area not able to be reached by the drug can allow for a reservoir of bacteria to survive that can lead to a repopulation of the entire middle ear causing a recurrence.

In addition, the pharmacokinetics of povidone-iodine when topically applied to the middle ear is unknown. Although there is a report of povidone-iodine being able to penetrate skin in an ex vivo model, the degree and depth of mucosal penetration have not been tested (35). This is of import as there is increasing evidence that recurrence of chronic PA and SA infections is secondary to intracellular persister cells that, in essence, “wait out” the course of treatment only to act as a source for future infections once treatment ends (36,37). Povidone-iodine may not be able to penetrate deep enough to attack this concealed population of bacteria. Finally, antibiotics are known to change the local microenviroment, which can in turn modify the immune response (38). The precise impact of povidone-iodine on the microenviroment of the middle ear has yet to be studied. Nonetheless, our work showed that a monocyte-derived cell line had reduced cell viability, so it is possible that a similar effect is occurring on the resident macrophages potentially limiting the local immune response.

Moreover, there are inherent differences between the disease as modeled in mice and the disease of CSOM as seen in humans. The full extent and spectrum of CSOM are unable to be fully recapitulated. Specifically, the impact of working in an inbred strain compared with the variability within humans and patient-specific factors cannot be underestimated.

Further work will be necessary to clarify which factor or if all of the above are working in concert to lead to the in vivo failure of povidone-iodine. Despite the in vivo failure of povidone-iodine in CSOM, there is still the potential for its use in draining mastoid cavities and otitis externa where the tympanic membrane is intact. These patients, combined with meticulous cleaning, could have longer treatment times (>10 min) combined with protection from ototoxicity. The treatment may still be limited by the penetration of the povidone-iodine into the epithelium or its ability to address intracellular persisters.

Because there is a lack of ototopical options with in vivo eradication potential of biofilm-associated persisters, early surgical debridement in CSOM is advocated to avoid the long-term development of permanent sensorineural hearing loss (1,39). Surgical intervention does not guarantee success in resolving CSOM because it is difficult to cure a microscopic disease with a macroscopic treatment. Adding ototopical treatment intraoperatively may have a synergistic effect to improve outcomes, and it will also likely lower the bacterial load during therapy and therefore could be used in the preoperative weeks to improve conditions for surgery. Using povidone-iodine after exenterating all air cells without the mastoid, removing adhesions, and clearing out biofilm may allow for it to reach more fully the entire middle ear space. This could be a workaround if povidone-iodine primarily failed as an in vivo treatment because of poor distribution related to the physical characteristics of the solution. Moreover, it would allow the oval and round window to be completely blocked from the povidone-iodine solution to obviate any concerns about ototoxicity. Nonetheless, new treatment options are desperately needed to help the millions of people dealing with the scourge of CSOM.

The findings of our study are limited by our analysis of solely PA and SA. Although these are the predominant bacterial species involved in CSOM (40), the efficacy of povidone-iodine on CSOM related to other bacteria such as Proteus or Enterococcus is unknown. In addition, our CSOM model utilize mice that have never had a previous infection, which is decidedly different from what is seen in humans. Lastly, these mice received repeated courses of anesthesia to receive transtympanic treatment and to test hearing function, which, although unlikely, could have an impact on the immune system’s ability to clear the CSOM infection.

CONCLUSION

Povidone-iodine, like all currently used ototopicals, has limited utility for treating CSOM because it is unable to stop recurrence of infection once treatment has been completed even when using potentially an ototoxic concentration. There is a tremendous need for the development of novel ototopical therapies to solve this pressing global problem.