Biofilm in Bacterial Vaginosis: A Legitimate Therapeutic Challenge?

Found worldwide, bacterial vaginosis (BV) remains the most common vaginal infection in reproductive-age women [1]. BV is polymicrobial in nature, and understanding the pathogenesis of BV remains incomplete, though evidence supporting the role of sexual transmission has increased [2]. BV was originally considered a nuisance-type syndrome, but the last 2 decades have revealed an ever-growing list of serious complications with profound consequences: preterm delivery of low birth weight infants, fetal loss, pelvic inflammatory disease, and increased risk of sexually transmitted infections, the most important of which is HIV [1]. Underappreciated is the risk of recurrent symptomatic vaginitis, often mixed or followed by recurrent vulvovaginal candidiasis, with both infections incurring a considerable personal and national economic cost. Two antibiotic classes—nitroimidazoles and lincomycins, administered orally or topically with equal efficacy—have been available for 4 decades. While these achieve fairly rapid moderate rates of short-term symptom relief, unfortunately, high rates of symptomatic recurrent disease as well as chronic refractory BV are widely reported [3]. Faced with an empty treatment pipeline, patients and practitioners remain frustrated and helpless in confronting BV recurrence [4–6].

BV recurrence is frequently the consequence of sexual reinfection, clinically indistinguishable from BV relapse, except in sexually inactive women, although most episodes are thought to be the result of relapse. These cases of relapse are firstly characterized by unexplained failure of vaginal recovery of protective Lactobacillus species following antibiotic therapy, thus fostering the rationale for probiotic therapy concurrent with or immediately following antimicrobials. The failure of lactobacillus recovery is of complex etiology, in part reflecting genetic influences on vaginal microbiota, but it mostly indicates an unfavorable vaginal environment likely due to the continued presence of BV-associated bacterial pathogens or their metabolic or mucosal inflammatory consequences. All this frequently occurs in the absence of symptoms and with a negative result upon an Amsel diagnostic test, which was never designed to serve as a prognostic test. Unfortunately, practitioners entirely lack clinically useful prognostic tests. Considerable progress has been made in readily identifying the many vaginal bacterial species quantified and causally linked to BV, by using polymerase chain reaction and next-generational sequencing [7]. In the complex environment of a polymicrobial infection, the suspected primary leader pathogen, Gardnerella vaginalis, interacts with multiple secondary BV-associated bacteria (BVAB; Prevotella bivia, Fannyhessea vaginae, and more), enhancing their numbers, virulence, and most importantly persistence [4, 8]. Information has expanded about these collaborating pathogens, mainly anaerobes, often found independently and without consequence in healthy women. We are left with the original unanswered question of underlying failure of healthy microbiota recovery in spite of postantibiotic symptom resolution and favoring the hypothesis of BVAB persistence with multiple biologic and therapeutic consequences.

BVAB persistence, evident as dysbiosis, is not synonymous with antimicrobial resistance, although likely linked. In the situation of an infection of monobacterial etiology, one can easily confirm resistance by performing in vitro susceptibility tests. Such tests are not possible when dealing with a polymicrobial infection where several participant pathogens are not cultivatable, thereby defying evaluation. Identifying bacterial resistance genes is a future likely solution [9]. Significant BVAB antimicrobial resistance in women with recurrent BV has been documented in vitro and clinically. Several investigators reported improved BV eradication with use of higher doses of antibiotics topically [10–12]. The continued absence of a conventional experimental animal model has markedly impaired progress in answering many questions in BV. Nevertheless, data correlating antimicrobial resistance and clinical failure remain incomplete and inadequate.

Progress was, however, evident with the demonstration by Swidsinski et al almost 2 decades ago, using peptide nucleic acid fluorescent in situ hybridization (FISH) probes of an epithelial surface or mucosal biofilm in vaginal biopsy specimens obtained from women with BV [13]. Clearly evident within the biofilm were abundant G vaginalis microorganisms. In addition, studies of urethral swabs obtained from male partners of women with BV demonstrated identical biofilm covering urethral cell clusters incorporating G vaginalis and BVAB, thus confirming the role of heterosexual BV transmission [14]. Biofilms are defined as highly organized sessile microbial communities of bacteria, fungi, or both, in which, following adherence of these microbiota to an interface, there follows the self-production of an encasing extracellular matrix further enhancing attachment [15, 16]. The environment within the biofilm may contain 1 and frequently multiple microbial species, thereby facilitating phenotypic and functional microbiota change, including gene transcription [16]. Antibiotic penetration of biofilm communities is frequently impaired, resulting in reduced access to microbial targets creating a refuge or safe haven for pathogens and facilitating microbial persistence with or without confirmed in vitro resistance. Transcriptional genetic changes are enhanced in this new environment, further playing a major role explaining decreasing BVAB antibiotic susceptibility [17]. An inaccessible vaginal reservoir of resistant BVAB is created. Biofilms have increasingly been recognized in diverse infection sites, especially in relation to medical and prosthetic devices [15].

Furthermore, repeated study of adherent polymicrobial vaginal biofilms in BV revealed abundant G vaginalis and a smaller number of BVAB, including F vaginae, in vaginal biopsies obtained from asymptomatic women immediately following administration of conventional doses of metronidazole gel [18]. The FISH technique repeatedly showed the dominance of G vaginalis in BV biopsy specimens, which, with multiple in vitro virulence studies, established the primary importance of certain clades of G vaginalis as the principal but not the only pathogen in BV [19]. In summary, FISH probe studies have provided considerable evidence implicating biofilm in the development of secondary antibiotic drug resistance in women with BV in addition to the universal principle of repeated antimicrobial drug exposure [17, 20]. Biofilm-linked resistance in no way excludes resistance in vaginal planktonic microbiota, and the relative contribution of each subpopulation is unknown. Nevertheless, in the presence of an empty antibiotic pipeline, what better basis to initiate a new phase of BV treatment via the introduction of clinically tolerated biofilm “busters.” At the initiation of this therapeutic phase, no known or well-studied agents, either topical or systemic, were available [21]. The challenge to developing BV biofilm destructive agents and those capable of preventing biofilm formation was again made more difficult by the absence of an animal model, as well as the absence of biofilm markers, chemical or biologic, for confirmation of their presence, composition, and quantity.

To date, few clinical studies to evaluate use of biofilm-destroying agents have been attempted or proven to be effective in women with BV. After demonstrating significant biofilm destruction in an established in vitro non-BV model, boric acid in combination with EDTA was used to treat 53 women with BV, without concomitant antibiotic therapy [22]. Only modest clinical benefit was observed in the small number treated, and none were studied with resistant or refractory BV. More encouraging, however, was the use of boric acid (600 mg daily per vagina) when used consecutively after or simultaneously with metronidazole, suggesting a synergistic effect [23, 24]. Boric acid was well tolerated and safe in both studies. Unfortunately, both published studies were uncontrolled and not compared with antibiotic use alone. It seems apparent that antibiofilm agents, when used alone and without proposed antibacterial synergy, are not an immediate solution and are without evidence of significant clinical efficacy. Once more using the FISH technique and utilizing posttreatment vaginal biopsies, Swidsinski et al reported the frequent failure of several topical agents, predominantly antiseptics, probiotics, and antibiotics alone to eradicate biofilm in women with BV, and studies also revealed biofilm-acquired or secondary resistance [25]. Correlating successful clinical response with biofilm status following therapy has been limited and disappointing.

In the current journal, Gao et al report a thorough review of potential clinically useful antibiofilm agents [26]. Most of the investigated agents are not available commercially, certainly not as biofilm destroyers, and unfortunately are rarely subjected to clinical efficacy evaluation in controlled studies. Almost all the information presented is based on preclinical in vitro testing, with only a few evaluated for efficacy against BVAB. Apart from preliminary, mainly negative data associated with probiotics, little evidence of BV prevention is presented. Accordingly, clinical studies with a variety of antibiofilm agents are urgently needed and, if undertaken, need to withstand scientific scrutiny. A major consideration with the current list of potential antibiofilm agents reported is recognition of patient tolerance of “destructive” chemical or biologic agents deemed “nontoxic,” given the clinical sensitivity of vulvovaginal mucosa.

Against this background of pessimism, a new area of optimism recently emerged with the publication of a vaginal organ version, as part of the “organ on chip” series, conceived to replace experimental animal models [27]. This vaginal organ version is more than a traditional tissue culture model; instead, it has viable, functional vaginal stratified squamous epithelial cells in tissue culture, which is housed in a microfluidic format that closely resembles the natural epithelial surface. This surface is based on a foundation of stromal fibroblasts and is covered by a physiologic supernatant that is readily modifiable. A BV model has now been published and is amenable to reasonable therapeutic intervention [27]. The only component missing from the “vagina on chip” BV version is the absence of biofilm, but this is clearly possible to create and study. This model may well serve as the ideal preclinical measure to study antibiofilm agents and new BV antibiotics.

Notes

Financial support. No financial support was received for this work.

References

Author notes