The development of mouthwashes without anti-gonococcal activity for controlled clinical trials: an in vitro study

Introduction

The importance of antimicrobial resistance (AMR) in Neisseria gonorrhoeae cannot be overstated. The bacterium is renowned for its capability to acquire AMR and has developed resistance to all classes of antimicrobials used for its treatment¹. AMR frequently emerges in core groups, such as men who have sex with men (MSM)². The pharmaco-ecological theory of AMR states that this resistance is driven by two main factors: (a) frequent transmission of gonococci between individuals within a densely interconnected sexual network, and (b) excessive antimicrobial use which acts as a selection pressure on circulating gonococci to acquire AMR^3–5. If this theory is correct, current efforts to reduce sexually transmitted infection (STI) prevalence via expanded screening and antimicrobial therapy in MSM may paradoxically be playing an important role in the promotion of gonococcal AMR^5,6.

These considerations have led to efforts to reduce the prevalence of gonococci in MSM and other core groups with non-antimicrobial products. One option is the use of an antiseptic mouthwash to decrease the oropharyngeal prevalence of gonococci (and other STIs). A modeling study showed that regular use of a mouthwash by MSM could reduce the prevalence of gonococci at different body sites⁷. A further consideration is that the oropharynx plays a central role in the emergence and spread of gonococcal AMR among MSM because of multiple reasons, which are reviewed elsewhere⁸. If a mouthwash can reduce the prevalence of oropharyngeal gonorrhoea without selecting for AMR, this may have the added benefit of reducing the probability of AMR emerging at this site⁴. Two randomized controlled trials (RCTs) are currently underway to assess whether regular mouth washing and gargling in MSM is able to reduce the cumulative incidence of gonorrhoea and other STIs. The OMEGA (Oral Mouthwash use to Eradicate GonorrhoeA) study is an RCT that assesses whether daily use of Listerine Zero^® can reduce the incidence of pharyngeal gonorrhoea in a population of Australian MSM (ACTRN12616000247471)⁹. We are conducting a second RCT to assess if the use of Listerine Cool Mint^® (LCM) is able to reduce the cumulative incidence of gonorrhoea (PReGo – Preventing Resistance in Gonorrhoea Study; registered at ClinicalTrials.gov with the identifier NCT03881007).

The choice of an optimal placebo is critical to the success of these RCTs. It is particularly important that a placebo is inert and has no bactericidal or bacteriostatic effect on gonococci. If it did, it would increase the probability of a false negative study outcome.

So far, no study has assessed placebo mouthwashes for this purpose. In this paper, we describe the process of developing and testing a series of candidate placebo mouthwashes. Our aim was to find the most suitable formulation for use as a placebo in the PReGo study. The major criterion we used to assess the mouthwash was its anti-gonococcal activity.

Methods

Assessment of antibacterial effect

Each gonococcal isolate was brought into suspension in 3mL PBS at a 0.5 to 0.8 McFarland turbidity, corresponding to a concentration of 10⁸ CFU/mL. From these suspensions, 100μl was then added to 900µL of each mouthwash, resulting in a concentration of 10⁷ CFU/mL. After 30 seconds, 60 seconds, five minutes, 30 minutes and 60 minutes at ambient temperature (20 ± 5°C), 10µL aliquots were plated onto blood agar (5% horse blood) and incubated for 24 hours in a 6 ± 1% CO₂ environment at 35 ± 2°C. Bacterial growth was visually scored on a semi-quantitative five-point scale, as described in Figure 1. Plating was conducted in duplicate for each isolate and all bacterial growth assessments were made by a single observer.

Figure 1. Five-point scale for scoring Neisseria gonorrhoeae growth on blood agar (0 = no growth; 1 = some colonies; 2 = numerous colonies; 3 = entire agar plate covered with colonies; 4 = confluent growth of colonies).

No statistical analysis was performed. This study did not involve any experiments on humans or animals and thus no ethical clearance was required.

Results

All three isolates were susceptible to ceftriaxone and spectinomycin; isolates B and C had a slightly increased minimum inhibitory concentration (MIC) for azithromycin and isolate A had a high MIC for ciprofloxacin and cefixime. None of the strains produced penicillinase (Table 3).

All isolates were fully susceptible to LCM and Corsodyl^®; a full bactericidal effect was observed after an exposure of 30 seconds or longer (Table 4)¹¹.

Exposure to Biotene^® for 30 minutes or longer was found to inhibit gonococcal growth considerably (Table 4).

Placebo 1, an ethanol-containing mouthwash was designed to have a similar color and taste as LCM^® but led to almost complete inhibition of gonococcal growth even after a short duration of exposure. Placebo 2 contained no ethanol and a lower amount of mint spiritus. Yet, its bacteriostatic effect was comparable to Biotene^®. In order to determine if mint spiritus or malachite green were the inhibiting factors, these ingredients were sequentially omitted in Placebo 3 and 4. Raspberry extract was added to both in order to improve the taste, but this resulted in strong inhibition of gonococcal growth in both cases. Placebo 5 contained elderberry extract instead, but substantial gonococcal growth inhibition was seen here, too. Placebo 6 contained another type of colorant (solutio viridis) and showed the least bacteriostatic effect after 30 and 60 minutes of exposure (Table 4). During every experiment, there was full and confluent gonococcal growth after exposure to the negative control substance (PBS) (Table 4).

We noted a slight difference in susceptibility to the mouthwashes between the three tested gonococcal isolates. Isolate A was more susceptible to placebos 1–6 and Biotene^® compared with isolates B and C. However, all strains showed equivalent susceptibility to LCM and Corsodyl^® (Table 5 and Table 6). These differences were not assessed for statistical significance.

Discussion

The recognition of the oropharynx as a source of gonococcal transmission and the genesis of antimicrobial resistance in groups such as MSM has directed research interest towards novel non-antimicrobial methods to prevent or treat oropharyngeal gonococcal infection. Mouthwashes are one such option. In order to determine the efficacy of an intervention involving the use of a mouthwash, RCTs should be performed, and a non-bactericidal placebo is a prerequisite for these trials.

Commercially available non-alcohol containing mouthwashes (like Biotene^®) are an attractive option, but our experiments suggest that exposure to Biotene^® for longer than five minutes may inhibit the growth of gonococci. Mouthwashes are typically used for 60 seconds but the substantivity of its ingredients may result in the antibacterial activity of mouthwashes persisting for over six hours^12,13. To optimize their STI preventive potential, mouthwashes could be used pre and post sex, which could lead to multiple exposures per day. These considerations triggered the search for a placebo with minimal inhibitory effect for periods of up to 60 minutes.

All three isolates in the experiment were fully susceptible to LCM and Corsodyl^® and isolate A was most susceptible to all placebo mouthwashes. Although this difference in susceptibility may have been the result of random variability, we could speculate that, in the absence of overt resistance to antiseptics, there might be a mechanism that partially protected isolate B and C from the harmful effect of some of the mouthwash constituents. Isolates B and C had a reduced susceptibility to azithromycin. This might have been due to the increased expression of an efflux mechanism such as the Mtr (multiple transferable resistance) efflux pump, which is linked to resistance to macrolides, as well as to many other substances like dyes and detergents¹⁴. We did, however, not perform genotypic assessment of the isolates used in this experiment.

After testing multiple combinations of ingredients, we found that Placebo 6 had the least bacteriostatic effect in vitro.

The limitations of this study include the following. First, budget and time constraints did not allow to perform any in vivo evaluation of the mouthwashes. It is of particular interest to know whether participants can distinguish Placebo 6 from an antibacterial mouthwash. A formal head-to-head comparison with LCM is planned as part of the PReGo study. Second, the sample size of the current experiment was too small to statistically assess differences between isolates and between mouthwashes. We may have over- or underestimated the true effect of the mouthwashes. Additionally, the experiments were performed sequentially, which may have introduced some inter-run variation. In each experiment we did however include a PBS exposed control. and plating was done in duplicate. Third, the observer who assessed bacterial growth was not blinded to the ingredients of the mouthwashes, we did not use a validated quantitative assessment method and we did no further in vivo or in vitro fitness testing of the isolates after exposure to the mouthwashes. Fourth, we used isolates from women with pharyngeal gonococcal infection, which are possibly not representative of the gonococci circulating among MSM. Their susceptibility pattern was, however, similar to that observed in most gonococcal isolates from MSM. Fifth, our in vitro findings are not necessarily representative of the in vivo setting as anatomical and biological properties may influence the effect of a mouthwash against gonococci in the throat. Bioactive molecules in saliva may, for example, have synergistic or antagonistic effects on the mouthwash’s active ingredients. Finally, we did not assess the effect of the placebo on the oropharyngeal microbiome. An increased or decreased growth of other oropharyngeal commensals might theoretically compete with gonococcal proliferation in the throat and influence gonococcal infectivity as well.

Conclusion

This experiment has shown that it is hard to develop an ideal placebo mouthwash as a range of frequently used ingredients inhibit gonococcal growth. A commercially available mouthwash like Biotene^® seemed the perfect option at first but it had a bacteriostatic effect. A process of serial testing of various placebos resulted in a placebo mouthwash, which we believe demonstrates a good trade-off between anti-gonococcal properties and taste.

Data availability