Bacterial Vaginosis: What Do We Currently Know?

1The vaginal microbial community is complex and dynamic, consisting of a group of bacteria typically characterized by abundant Lactobacilli that evolve during the life of the woman, depending on age, hormonal estrogen levels, sexual practices and the environment. The vaginal microbiota plays a crucial role in women’s health (infection, reproduction…), and that of their fetuses.

2BV is a dysbiosis of the vaginal microbiota characterized by a shift from Lactobacilli dominance to that of a mixture of various anaerobic bacteria. It is the most common vaginal disorder worldwide in women of childbearing age. BV is associated with significant adverse healthcare outcomes, including increased susceptibility to sexually transmitted infections, urogenital infections, pelvic inflammatory disease, and an increased risk of abnormal pregnancy. The etiology of BV is still unknown. Standard antibiotic therapy often fails, with an estimated relapse rate of 50% at six months follow-up.

3The vaginal ecosystem is colonized from the very first hours of the birth of a female and remains throughout her life until death. Women of childbearing age produce about 1 to 4 mL of vaginal fluid, containing 108 to 109 bacterial cells per mL.

4The vaginal flora was first described by the German gynecologist Albert Döderlein in 1892, who reported a homogeneous vaginal flora of gram-positive bacilli in healthy women. They were named “Döderlein’s bacilli” and were later identified as members of the Lactobacillus genus by Beijerink in 1901. Under normal conditions, 70-90% of the vaginal bacterial species in healthy premenopausal women are Lactobacilli. As molecular techniques have advanced, our understanding of the diversity and complexity of the vaginal bacterial community has broadened. Among more than 200 Lactobacillus species with standing in the nomenclature, over 20 species have been found in the vaginal flora. Sequencing of the 16 rRNA gene revealed that the vaginal bacterial community, mainly composed of Lactobacilli, is classified into five groups named community state types, namely I, II, III, IV and V. Four of these groups are dominated by Lactobacillus. The first is dominated by L. crispatus, the second by L. gasseri, the third by L. iners, and the fifth by L. jensenii, while the fourth contains a smaller proportion of Lactobacilli but is composed of a polymicrobial mixture of strict and facultative anaerobes (Gardnerella, Atopobium, Mobiluncus, Prevotella…). Although there is always a temporal transition between vaginal bacterial communities ( Figure 1 ).

5Thus, many other bacteria are present at lower concentrations in healthy vaginal flora, such as Peptostreptococcus, Bacteroides, Corynebacterium, Streptococcus, and Peptococcus.

6The composition of the vaginal microbiota evolves throughout a woman’s lifespan. Various physical and hormonal changes occur in the vagina biotope during these different stages of a woman’s life ( Figure 2 ).

7Vaginal bacterial communities of women of childbearing age may vary between women from different regions, but also between women of different ethnicities living in the same geographical area. In 2011, a study by Ravel et al. characterized the vaginal microbiota of asymptomatic North American women with pyrosequencing, showing that the vaginal flora of Asian and white American women was dominated by Lactobacilli, unlike Hispanic and African-American women, of whom only 60% had a Lactobacillus -dominated vaginal flora. In addition, Caucasian and Asian women tend to have high levels of L. crispatus and lower levels of L. iners compared to African women. In another study using 16S rRNA gene sequencing, Fettweis et al. demonstrated that the vaginal microbiota of European-ancestry women was dominated by Lactobacilli, as opposed to African-American women, who presented a mixed vaginal community containing, among others, Mycoplasma hominis, Aerococcus, L. iners and numerous strict anaerobes, including gram-positive anaerobic cocci, BV-associated bacteria, Sneathia, Prevotella amnii, Megasphaera, Atopobium, and Gardnerella vaginalis. The vaginal pH also differs between racial groups. African-American and Hispanic women had a vaginal pH (4.7 and 5.0, respectively) higher than what is considered the norm (<4.5).

8The vaginal flora presents one of the most important defense mechanisms for reproductive function and maintaining a healthy environment. The stability of this flora prevents the proliferation of commensal microorganisms and colonization by pathogens, thereby preventing infection. Bacteria form a adhered monolayer on the vaginal mucosa and produce antimicrobial compounds that maintain this health equilibrium, such as hydrogen peroxide (antimicrobial product protecting against deleterious microorganisms), lactic acid (which maintains the normal vaginal pH between 3.5 to 4.5), bacteriocins (antibiotics that inhibit the growth of harmful microorganisms within the vagina), and arginine deaminase enzyme (metabolizes arginine into citrulline and ammonia (NH3), depriving anaerobic pathogens of this amino acid necessary for their growth).

9Notably, L. crispatus and L. jensenii may produce hydrogen peroxide, an oxidizing agent, toxic for catalase-negative bacteria and also capable in vitro of inhibiting HIV-1 and herpes simplex virus type 2. The vaginal acids produced can, in the presence of viral RNA, stimulate the maturation of dendritic cells, activation of 17 subclasses of T helper lymphocytes, and the production of protective inflammatory cytokines and interferon-γ.

10In addition to the role of Lactobacilli, cervical mucus is mainly composed of mucin, which protects the vaginal mucosa and optimizes its barrier role against microbial colonization. Analyses of the composition of cervical mucus and vaginal secretions have demonstrated the presence of several proteins with antimicrobial activities that act independently of the presence of antibodies, such as lactoferrin, lysosyme, calprotectin [also known as MRP8/MRP14 (“myeloid related protein”)], cathelicidin LL-37.

11Formerly known as non-specific vaginitis, BV is characterized by a change in the vaginal flora composition, with a dramatic depletion of Lactobacilli due to a significant overgrowth of obligate or facultative anaerobes previously a minority in the vagina, such as Gardnerella vaginalis, Atopobium vaginae, Ureaplasma urealyticum, Mycoplasma hominis, Prevotella, Peptoniphilus, Megasphaera, Mobiluncus, and several fastidious and uncultured bacteria, including BV-associated bacteria (BVAB-1 to 3). The factor triggering this overgrowth of anaerobic bacteria is unknown. It is linked to an alkaline vaginal ecosystem due to an increase of vaginal pH following the loss of Lactobacilli protective effects ( Figure 3 ).

12The diversity of the vaginal flora in patients with BV was described in 1921 by Schröder. In 1955, Gardner and Dukes claimed that the etiological agent of BV was Haemophilus vaginalis, a gram-negative rod later renamed Gardnerella vaginalis. The bacteria present in the microbiota of BV form a biofilm on the vaginal epithelium and secrete a cytotoxin capable of killing the epithelial cells. In addition, G.vaginalis produce proteolytic enzymes able to degrade proteins and decarboxylases that convert amino acids. Not being degraded, the amine compounds become malodorous (fishy odor: “Whiff test”) thank to an increase of the vaginal pH. Subsequently, cytotoxicity resulting from the combination of organic acids present in the vagina during BV and bacterial polyamines leads to the production of a vaginal discharge caused by the exfoliation of vaginal epithelial cells. Furthermore, this bacterium, cover the vaginal epithelial cells, causing the formation of “clue-cells”, a specific characteristic of BV.

13A few years later, G. vaginalis was found in 40% of healthy women, and its role was disputed. Thus, colonization of G. vaginalis does not always promote BV, suggesting that this bacterium alone may be essential but not sufficient for BV development. In addition to G. vaginalis, some anaerobic bacteria are highly associated with BV, indicating that BV is a polymicrobial syndrome, which does not follow Koch’s postulates.

14Recent progress in research of BV pathogenesis have determined the existence of 13 different species within the genus Gardnerella. Although these Gardnerella species may be closely linked genetically, only a few of them may be implicated in disease such as BV. Healthy women may be colonized by non-pathogenic Gardnerella species, whereas virulent strains are involved in the development of BV. Advances in technology, particularly next-generation sequencing, have clarified much of this issue. Arguably, all papers dealing with G. vaginalis prior to that of Vaneechoutte et al. are not specifically about G. vaginalis but rather about Gardnerella spp.

15Based on a recent prospective study, an updated conceptual model on the pathogenesis of BV was outlined. The potential synergistic relationship between G. vaginalis, P. bivia, A. vaginae was studied. After sexual exposure to virulent strains of G. vaginalis, these strains displace vaginal Lactobacilli and begin to form a BV biofilm on the vaginal epithelium. Subsequently, proteolysis by G. vaginalis occurs which promotes the growth of P. bivia. This bacterium produces an ammonia product which in turn promotes the growth of G. vaginalis and the biofilm develops. These 2 bacteria then produce sialidase that degrades the biofilm and P. bivia may thus degrade the mucin layer of the vaginal epithelium. After the loss of the protective mucus layer, increased adhesion of other BV-associated bacteria, including A. vaginae, to the polymicrobial biofilm will occur. The role of the other bacteria remains unknown. Further research focusing on the complex interactions between bacteria during BV is needed.

16BV ranges from asymptomatic to an increase in vaginal discharge with or without a fishy odor. The collection of a specimen for diagnosis can be performed using a speculum during the pelvic exam. When there is no reason for a pelvic exam as part of the clinical evaluation, a self-collected vaginal swab may also be provided.

17Two main categories of diagnostic strategies for BV exist: the “bedside” method introduced in 1983, mainly based on real-time clinical criteria –”Amsel’s criteria”, and laboratory-based testing developed in 1991, relying on the evaluation of morphotypes on gram staining – “ Nugent’s score”. Amsel’s criteria and Nugent’s score are the most common diagnostic methods used for BV. Furthermore, the World Health Organization (WHO) has considered the Nugent’s score as a gold standard for studies. However, the current recommended best clinical practice for diagnosing BV in women is gram staining microscopy according to the Hay-Ison criteria, as it is easier and faster to use. The Hay-Ison criteria were analogous to the Nugent’s scores. Hay’s grades I, II, and III were similar to Nugent’s scores 0-3, 4-6, and 7-10.

18Even if Nugent’s score is considered as a gold standard by the WHO, it has some pitfalls. In fact, intermediate flora is so far an uncharacterized category and is a challenge in the diagnosis of BV. In addition, the identification of morphotypes is subjective and technician-dependent, thus the diagnosis may be influenced by individual skills and experience.

19Recently, a study made by Wang et al. provided a proof of concept for a deep learning-based model to quantify Gram staining and, consequently, automated Nugent score classification. Deep learning methods, particularly convolutional neural network (CNN) models, have demonstrated excellent performance in computer vision tasks. This model outperformed human healthcare professionals in terms of accuracy and stability for three diagnostic categories of Nugent scores. The deep learning model may offer translational applications in automating the diagnosis of BV with appropriate supporting hardware.

20The diagnosis of BV is problematic and challenging because of its intricate polymicrobial features and a wide range of clinical features. In order to overcome these diagnostic problems, alternative diagnostic strategies have been attempted, such as molecular, enzymatic and chromatographic techniques.

21A molecular technique used in the diagnosis of BV is specific quantitative real-time PCR (qPCR) test. It is a quantitative, reproducible and reliable molecular biology tool that measures the presence of bacteria presents in BV, such as Atopobium vaginae, BVAB2, Gardnerella vaginalis, Leptotrichia/Sneathia spp., Megasphaera spp., and Mobiluncus spp….. Many studies proposed objective molecular cut-off values from bacterial load to predict BV.

22Several commercially molecular diagnostic assays have been reported for the diagnosis of BV in women including the NuSwab R multiplex quantitative PCR, SureSwab BV DNA real-time quantitative assay, BD Max vaginal panel and BV multiplex assay. The quantification of these bacteria therefore makes it possible to establish a precise diagnosis of BV, with a sensitivity ranging from 90.5% to 96.7% and a specificity ranging from 85.8% to 95% compared to Amsel criteria and Nugent score.

23Even though these tests have a higher sensitivity and specificity than the currently available diagnostic tools, they are not point-of-care tests and are more expensive. However, Dessai et al. are the first to report on the performance of the BD AffirmTM VPIII test as a POCT in a prenatal population. But the test has been shown to be inadequate as a screening test for vaginal infections in pregnancy.

24Finally, POC tests for BV are not available or are simply too expensive to be used routinely. It is therefore mandatory that the development and evaluation of new diagnostic tests include a cost analysis.

25As the presence of sialidase is currently considered a key indicator of BV in the clinical examination, an enzymatic approach has been developed: The OSOM R BVBlue R test as a POC diagnostic test for BV. It is based on the qualitative detection of a high level of sialidase produced by anaerobic pathogens in vaginal fluid samples. It has been shown to be reliable compared to conventional methods such as Amsel criteria and Nugent score

26In addition, a recent study by Liu et al., showed the feasibility of turn-on tetravalent sialic acid-coated tetraphenylethene luminogen (TPE4S) as a powerful diagnostic tool for high-throughput fluorescence-guided diagnosis of BV. This study uses light signal intensity to detect and measure the relative concentration of sialidase in a vaginal sample. All reagents are present in a reagent bead and sample buffer, essentially allowing for a one-step test. The test is highly sensitive and quantitative, with a sensitivity and specificity of 95.40% and 94.94%, respectively, compared to the Amsel method and 92.5% and 91.8% compared to the BVBlue diagnostic results. Notably, this method gives a more accurate classification and quantification of BV severity based on relative fluorescence intensity (I/I0). Thus this test can be a potential tool for diagnosis of BV, and risk assessment of patients with BV based on sialidase activity levels and monitoring of antibiotic therapy.

27Another new approach based on immunodetection also targeting sialidase has been developed for the diagnosis of BV. The nanophotonic operating principle of this biodetection method allows a cheaper, faster and simpler analysis than the indirect enzyme-linked immunosorbent assay (ELISA). This nanotechnology has a high sensitivity and specificity (96,29%, respectively). This method offers an original approach to perform a very rapid diagnosis of BV.

28BV may appear at any reproductive age (between 15 to 44 years-old). Its prevalence rates vary considerably among the geographic regions of the world, within the same country, and even within the same population, depending on ethnic origin and socioeconomic status. Although its exact prevalence remains difficult to determine, BV occurs between 4-75%, depending on the population studied. Intermediate in the USA (29%), the prevalence of BV is estimated to be low in Europe, with a maximum (> 20%) in Poland and Norway. In Africa, the estimated prevalence tends to be high. However, BV prevalence is lowest in west Africa (6-8% in Burkina Faso, 14.2% in Nigeria) than southern and eastern Africa: 32.5% in Zimbabwe, 37% in Kenya, 38% in Botswana, and 68.3% in Mozambique.

29Although the absence of a known causal agent makes it difficult to characterize BV as a sexually transmitted infection (STI), it is strongly associated with sexual activities and has some characteristics of a sexually transmitted disease not by microorganism transfer, but by mechanical or chemical interaction such as contact with highly alkaline semen. Overall, BV is diagnosed in post-pubertal women who have never been sexually active, but at a lower prevalence than those who are sexually active. The prevalence varies with the number of sex partners. It has been evaluated at 18.8% for non-sexually active women, 22.4% for women with one lifetime partner and 43.4% and 58% for women having 2-3 lifetime sex partners and those having ≥ 4-lifetime sex partners, respectively.

30In this dynamic, sex workers had a higher bacterial vaginal diversity but a much lower abundance of Lactobacillus species than women who are not engaged in sex work. Compared with male partners of healthy women, BV-related bacteria can be found in the penile skin, urethra, spermatozoa, and prostatic fluid microbiota of male partners of women with BV. Furthermore, biofilm fragments have been found in their urine and sperm, suggesting that male partners are a reservoir, and also that heterosexual transmission may occur. Nevertheless, there is no corresponding illness in male partners. Use of condoms by male partners also prevents acquisition and recurrence of BV. Also, since the preputial area of some men hosts BV-associated microorganisms, male circumcision may reduce the risk of BV.

31Prevalence rates also depend on the nature of the couple and their sexual practices. In fact, BV prevalence varies between 10-30% in heterosexual women, and is more frequent, 25-50%, in women who have sex with women (WSW). The reasons for this difference in prevalence are not clear, although sexual activities involving the transmission of vaginal fluid increase the risk of BV acquisition. Several studies have indicated that certain sexual behaviors, including non-coital sexual practices such as digital and penile penetration, anal and oral intercourse followed by vaginal penetration, enhance the risk of BV. In WSW, a symptomatic female sexual partner, receptive oral sex, and the use and sharing of unwashed sex toys constitute risk factors for BV. These observations have led some to consider BV as not an infection, but rather a taxonomic change in the vaginal microbiota resulting from translocation of oral or fecal microbiota during non-coital sexual practices.

32Additionally, genital hygiene can also promote disequilibrium in the vaginal microbiota. One study found that patients who did not wash their vaginal region were more susceptible to BV than those who often washed the vaginal region, a prevalence of 53.9% and 40.2%, respectively. Similarly, the prevalence of BV is higher in patients who do not change their underpants frequently compared to those who change it more frequently (57.6% versus 36.9%). In addition, other sexual sanitary habits, including vaginal douching and washing, as well as cigarette smoking, certain contraceptive methods like disposable intra-uterine devices and stress may also enhance the risk of developing BV.

33Women with BV are vulnerable: the presence of BV-related bacteria and/or sexually transmissible microorganisms in the BV microbiota can lead to opportunistic infections. During this imbalance, 10-30% of pregnant women with BV give birth prematurely, a preterm delivery often accompanied by perinatal mortality, up to 70% worldwide. During pregnancy, BV increases the risk of preterm labor, late miscarriage, intrauterine fetal death, preterm rupture of the membranes, amniotic fluid infections, chorioamnionitis, post-abortion and postpartum infections in these women.

34In non-pregnant women, bacteria involved in BV can initially cause cervicitis, endometritis, salpingitis, and urinary tract infections. After damage of the cervix, bacteria can migrate from the lower to upper genital tract, reaching the uterus and fallopian tubes and causing illnesses such as pelvic inflammatory disease, post-hysterectomy infections, and even cervical cancer or tubal infertility (van; ). Likewise, BV is associated with significantly increased rates of acquiring herpes simplex virus, human immunodeficiency virus, papillomavirus and transmission of the pathogens causing syphilis, chancroid, gonorrhea, trichomoniasis, and chlamydia.

35Considering that clinical cure corresponds to the disappearance of all symptoms, the treatment of BV is currently focused on stopping the proliferation of BV-associated microorganisms and restoring the normal vaginal flora. Typically, clinical therapies include the use of antibiotics with broad activity against anaerobic microbes and protozoa: clindamycin and the nitroimidazoles (metronidazole and tinidazole) and/or use of probiotics.

36The first line of therapy recommended by the World Health Organization (WHO) is 500 mg oral metronidazole twice daily for one week. However, treatment with metronidazole may cause side effects such as gastrointestinal pain, nausea, and vomiting. Other proposed therapeutic regimens include 300 mg oral clindamycin twice daily for one week, 100 mg of intravaginal clindamycin ovule daily for 5 days and an application of 0.75% intravaginal metronidazole gel for 5 days or 2% of intravaginal clindamycin cream at bedtime for one week. However, it should be noted that local application of clindamycin may damage latex-based products such as condoms and may also trigger pseudomembranous colitis.

37In addition, the use of tinidazole, a drug similar to metronidazole, has been approved and proposed as an alternative therapy in an oral regimen (either 2 g/day for 2 days or 1g/day for 5 days) if metronidazole and clindamycin are not tolerated.

38Some researchers have evaluated the efficacy of other antimicrobial agents, such as azithromycin, secnidazole or ornidazole and rifaximin. Secnidazole has shown activity similar to that of the recommended nitroimidazoles and also spares Lactobacilli, a beneficial characteristic in BV treatment. Even for Rifaximin, it acts on BV by restoring Lactobacilli and increasing lactic acid in patients.

39Taken locally or orally, these antimicrobial agents have almost similar efficacy, with cure rates around 58% to 92% after 1 month of treatment. Nevertheless, these results are temporary, leading to recurrence or re-infection at rates above 50% within 6-12 months of treatment. The reasons for this high relapse rate remain unclear. However, it appears that, with the formation of bacterial biofilms, these recommended therapies only temporarily eradicated BV-associated microorganisms, or these bacteria are reintroduced in the vagina by their sex partners. Further, the presence of some BV-associated bacteria such as Peptoniphilus lacrimalis, Megasphaera type 2 and BVAB-1 to 3 at the beginning of treatment is strongly related to BV recurrence, thus causing antibiotic failure. To this end, two recent studies have examined the acceptability, tolerability, and especially the efficacy of concomitant partner treatment to improve the cure of BV. The first study by Schwebke et al, showed no significant reduction in BV recurrence in female partners after treatment of the male partner with multidose metronidazole, although women whose partners adhered to multidose metronidazole were less likely to fail treatment. The second study showed that simultaneous partner treatment had a significant change in the overall composition of genital microbiota in both partners immediately after treatment.

40As antibiotic treatments can have a negative impact on the stability of the vaginal flora, Lactobacillus probiotics, an alternative and complementary therapy to antibiotic treatment, has been developed to help restore and maintain the healthy vaginal flora. Probiotics consist of living microorganisms that confer a health benefit on the host when they are administered in an appropriate quantity. Nine studies from 1989 to 2014 tested the effectiveness of vaginal or oral Lactobacillus probiotics. The results showed that both oral and vaginal Lactobacillus treatments were effective in curing acute BV. On balance, the application of Lactobacillus in the form of vaginal capsules (containing≥108 CFU of Lactobacillus strains per dose) or a fermented milk product (containing≥5×109 CFU of Lactobacillus strains per dose) may be an equally effective alternative to standard antibiotic capsules. Only the strains L. reuteri RC-14 and L. rhamnonus GR-1 have positive clinical effects (; Ouarabi et al., 2017). Administrated orally (twice daily) or vaginally (once weekly), probiotics may restore the normal Lactobacillus-dominated microbiota and reduce BV recurrence. Nevertheless, probiotics have had minimal success in African women.

41In the same context, a new study on the applicability of three strains of Lactobacillus spp. (Lactobacillus delbrueckii DM8909, Lactiplantibacillus plantarum ATCC14917 and Lactiplantibacillus plantarum ZX27) according to their in vitro probiotic capacities. These three Lactobacillus spp. strains have shown efficacy in the treatment of BV by limiting the growth, adhesion, biofilm formation and virulence properties of G. vaginalis.

42In addition, sucrose-containing products may promote recolonization, since sucrose is metabolized by Lactobacilli. A triple-blind randomized clinical trial compared the efficacy of a sucrose vaginal gel versus a metronidazole vaginal gel for 5 days in 70 women with diagnosed BV. The sucrose vaginal gel was as effective as metronidazole treatment, according to this individual RCT.

43Another option is to combine Lactobacilli with estriol. Two studies, a PC-RCT and a head-to-head RCT, were conducted on the efficacy of vaginal capsules containing L. acidophilus and 0.03 mg estriol (Gynofor®) in BV. Compared to placebo, the cure rate of BV was significantly higher in the treated group, but according to the comparative study, the combination of estriol and Lactobacilli is equivalent, but not better, results than antibiotic treatment.

44In terms of method of application, vaginal suppositories deposit Lactobacillus strains directly on the vaginal mucosa, whereas oral probiotics survive gastrointestinal transit and increase the number of strains in the colon and feces. In turn, this may promote recolonization of the vagina due to the proximity of the rectum and vagina.

45The management of BV urgently requires implementation of new therapeutic strategies. To disrupt BV-associated biofilms, studies are underway to investigate the role of novel agents such as DNases, retrocyclines, antiseptics, and plant-derived compounds in the treatment of BV. Dequalinium chloride, an antiseptic, has reported similar efficacy to clindamycin intravaginal cream. Thymol, a molecule found in thyme essential oil, has shown an inhibitory effect on biofilms in vitro. The application of acidifying agents, such as vitamin C or buffering agents (polycarbophil or boric acid), in combination with a nitroimidazole antibiotic, has been demonstrated to reduce the recurrence of BV, potentially by destructuring the vaginal biofilm.

46In this regard, other promising therapeutic agents for the treatment of BV are under investigation. Including DNase agents that can disrupt vaginal biofilms by targeting extracellular DNA essential for their structural integrity, retrocyclin 101, a synthetic cyclic antimicrobial peptide that inhibits the growth and development of G. vaginalis in vitro, and the amphoteric tenside pessary (WO3191), which disrupts biofilms after metronidazole treatment and promotes the growth of Lactobacillus species.

47Furthermore, given the success demonstrated by fecal microbiota transplant (FMT) in the management of various intestinal disorders and diseases such as recurrent Clostridium difficile infection, pseudomembranous colitis, inflammatory bowel disease (IBD) and irritable bowel syndrome (IBS), the use of vaginal microbiota transplantation (VMT) in BV treatment is a new therapeutic approach that modulates the vaginal microbiota in order to eradicate this scourge and reduce adverse gynecological outcomes.

48Given the high rates of recurrence and relapse, research is needed to identify and evaluate these novel biofilm-disrupting treatment strategies.

49The taxonomic composition and bacterial proportion of the vaginal microbiota are under the influence of intrinsic and external factors over the female lifespan. In the last decades, understanding of the bacterial diversity of this ecosystem has been increased by molecular methods. Dominated by Lactobacilli that protect against infection, the vaginal flora of healthy women is less complex than that of patients with BV, which presents a diverse microbiota containing numerous obligate anaerobic and uncultivable species. This polymicrobial condition is associated with relatively uncomplicated clinical symptoms that do not occur in all affected women, thus complicating the determination of its etiology. Treatment is usually unsuccessful, with a high rate of relapse. Future studies that thoroughly examine the vaginal bacterial community are needed to cultivate the bacteria associated with BV and its treatment failure, in order to study antibiotic resistance and to establish more effective alternative therapeutic strategies that reduce BV symptoms as well as its associated complications. Overall, unlocking the enigma of the pathogenesis of BV is key for the prevention and management of this public health problem.

50LAC, FF, and KD have equally contributed to the preparation of this manuscript. All authors contributed to the article and approved the submitted version.

51This study was funded by the Institut Hospitalo-Universitaire (IHU) Méditerranée Infection, the National Research Agency under the “Investissements d’avenir” programme, reference ANR-10-IAHU-03, the Region Provence Alpes Côte d’Azur and European FEDER PRIMI funding.

52The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

53All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.