Objectives To determine the prevalence of Gardnerella vaginalis in women with normal vaginal flora.
Methods Women without symptoms or signs of vaginal infection and five or fewer lifetime sexual partners were recruited for a longitudinal study of vaginal flora. Negative Amsel criteria and a Nugent score of 0–3 were required for enrolment. Vaginal specimens were self-collected daily for Gram stain and every 3 days for PCR for G vaginalis for 30 days. Women completed daily diaries recording sexual activity, symptoms and menses.
Results Twenty women were recruited for the study with 19 completing all specimens and 1 lost to follow-up. During the 30-day study period, 13/19 (68.4%) of women had normal Nugent scores (0–3) whereas 6/19 (31.6%) of women had at least 2 days of Nugent scores in the intermediate range (p=0.09). Among the 19 women, 9 (47%) were negative for G vaginalis by PCR throughout the study period whereas 10 (53%) had at least one specimen that demonstrated the presence of G vaginalis by PCR. Of those women with intermediate flora on Gram stain during the course of the study 5/6 (83.3%) were positive for G vaginalis while 5/13 (38.5%) of those women with only normal Nugent scores were positive for G vaginalis. Thus, 61.5% of women with normal Nugent scores had no evidence of G vaginalis by serial PCR.
Conclusions Gardnerella may not be part of the normal flora in women with optimal vaginal health.
- BACTERIAL VAGINOSIS
- VAGINAL MICROBIOLOGY
Statistics from Altmetric.com
The composition of the vaginal flora in reproductive-aged women has been well described and in broad terms can be classified as normal, intermediate and consistent with bacterial vaginosis (BV) as defined by the Nugent Gram stain criteria.1 The latter two categories of vaginal flora have been clearly shown to be associated with adverse outcomes such as preterm birth and acquisition of sexually transmitted disease (STD),2 ,3 perhaps partly due to local production of cytokines associated with these changes. It has been shown that increases in Gardnerella vaginalis alone are associated with acquisition of STD and cytokine production.3 ,4 Thus, it could be argued that optimal vaginal health should be strictly defined as the presence of lactobacilli without evidence of G vaginalis. We sought to determine the percentage of women without clinical evidence of altered vaginal flora who were colonised with G vaginalis.
Women without clinical evidence of vaginitis or STD were invited to participate. Women were recruited with the use of advertisements posted across the University campus. The study was approved by the University of Alabama at Birmingham Institutional Review Board. To be eligible for enrolment, women were required to be heterosexual, premenopausal, have had five or fewer lifetime sex partners, have no symptoms of a genital tract infection, have a vaginal pH of <4.7, absence of clue cells, negative whiff test and a vaginal Gram stain showing predominantly lactobacilli (Nugent score 0–3 without the presence of Gardnerella morphotypes). During the initial pelvic examination, vaginal specimens were collected for PCR for G vaginalis. Women completed baseline questionnaires. In addition, women self-collected daily vaginal specimens placed onto prelabeled glass slides for Gram staining and every third day collected vaginal specimens for PCR. The latter were frozen in prelabeled tubes in special insulated ‘lunchbox’ carriers containing chemical ice packs in the participant's home freezer until returned to the clinic. Once received by the laboratory, swabs for PCR testing were frozen at −80°C until DNA extraction. Women returned the samples at their follow-up visit in 30 days. Women also completed daily diaries recording sexual activity, symptoms and menses, which were collected at the follow-up visit.
Gram stains were performed and interpreted using the method of Nugent et al.1 Total DNA from specimens was extracted via an inorganic column-based method (DNAeasy Blood and Tissue Kit, Qiagen, Valencia, California, USA). Primers designed at our site using sequencing information for the 16S and 23S ribosome RNA of G vaginalis (GenBank Accession L08167)5 were used to amplify a 170 bp region within this internal spacer region. Primers were designed using one sequence. The sequences were then ‘BLASTED’ against GenBank. All G vaginalis sequences in GenBank at the time with sequence information covering this area were detected. Reactions were carried out in a total volume of 25 µL composed of 1X Coral Load Buffer (Qiagen, Valencia, California, USA), 0.2. M forward primer (Gvag16SitsF: 5′ –ccggctccattttggtggag-3′), 0.2 M reverse primer (Gvag16SitsR: 5′-atatcgcagcccgtcacgtc-3′), 200 M each dNTP and template DNA. Amplification conditions were: 95° C×3′/(95°C×10″; 67.3°C×30″)×40/4° C Pause. Amplicons were run in 1 X Tris-acetate EDTA (TAE) buffer on 2% agarose gels prestained with ethidium bromide. All gels contained an ATCC control DNA from G vaginalis. PCR was performed on specimens collected on days 1, 4, 7, 10, 14, 17, 21, 24, 28 and 30.
Twenty women were recruited for the study with 19 completing all specimens and 1 lost to follow-up. The mean age of the study population was 31.4 years. Seventy-five per cent of the women were African–American and 25% Caucasian. The mean number of lifetime sex partners was 2.7 with one subject reporting none. Only 17% of the subjects reported always using condoms with sexual activity. During the 30-day study period, 13/19 (68.4%) of women had normal Nugent scores (0–3) whereas 6/19 (31.6%) of women had at least 2 days of Nugent scores in the intermediate range (p=0.09). Among the 19 women, 9 (47%) were negative for G vaginalis by PCR throughout the study period whereas 10 (53%) had at least one specimen that demonstrated the presence of G vaginalis by PCR. The woman who denied any history of sexual activity was negative for G vaginalis. Of those women with intermediate flora on Gram stain during the course of the study, 5/6 (83.3%) were positive for G vaginalis while 5/13 (38.5%) of those women with only normal Nugent scores were positive for G vaginalis. Thus, 61.5% of women with normal Nugent scores had no evidence of G vaginalis by serial PCR. Seventeen women completed their diaries. Eleven of the 17 women (65%) participated in sexual intercourse during the study and none used condoms. All of these women were monogamous with the exception of one woman who had two partners during the month. Only one woman douched during the study. Of the women who were negative for Gardnerella 4/8 (50%) reported sexual intercourse versus 7/9 (78%) of those who had Gardnerella (p=0.5, 95% CI 0.3 to 1.4). There was no association between age, race and number of sexual partners with the presence of G vaginalis.
BV remains an enigma; however, recent data have been helpful in refining the hypotheses related to its pathogenesis. Gardnerella is always present in the vaginal flora of women with BV and this has been confirmed by molecular techniques.6 Studies have confirmed that more virulent factors, such as adherence and cytotoxicity, are associated with G vaginalis than with other BV-associated bacteria.7 Evaluation of the biofilm community of BV also confirms that G vaginalis is present in significantly higher concentrations than anaerobic bacteria.8
Studies that have demonstrated G vaginalis in women without BV have been long cited as evidence that Gardnerella is not the key to the aetiology of BV. However, our conceptual framework for what constitutes optimal or normal vaginal flora has evolved since the first of these studies was undertaken nearly 30 years ago. Totten et al were able to culture GV from up to 68% of women who did not meet the clinical criteria for BV.9 However, their definition of ‘normal’ flora was insufficiently strict to exclude women with intermediate flora. Recently, using molecular probes, Fredericks et al found G vaginalis in 59% of women without BV but again women were simply classified as having or not having BV.6 Using similar techniques, Menard et al differentiated between normal and intermediate flora by Nugent score and found a significant difference in the number of women who had G vaginalis, with 47% being positive in the normal group and 70% in the intermediate group (p=0.007). Further, they found that nearly 50% of the samples designated as intermediate flora by Gram stain were classified as BV using molecular techniques.10
Our study of women with lactobacillus-predominant, that is, optimal vaginal flora, is in agreement with the findings of Menard et al in that 45% of women had no evidence of G vaginalis by PCR from multiple samples throughout the reproductive cycle. Further, 61.5% of women with normal Nugent scores throughout the course of the study had no evidence of G vaginalis by serial PCR. The fact that nearly half of all the women studied and 61.5% of women with normal Nugent scores had no evidence of G vaginalis by PCR speaks against the hypothesis that Gardnerella is part of the normal or optimal vaginal flora.
The hypothesis that Gardnerella is the pathogen necessary for initiation of alteration of the vaginal flora does not negate the concept that BV, especially symptomatic BV, is caused by a polymicrobial community of bacteria. It is likely that G vaginalis is necessary but not sufficient for the production of symptoms that characterises BV. However, without acquisition of Gardnerella, the biofilm characteristic of BV would not form and cause BV. Recent studies have also suggested that there may be differences in virulence properties between different strains of Gardnerella.11
Our study did not find an association between the presence of G vaginalis and number of sexual partners; however, this is likely due to the small sample size. There was a trend towards the presence of Gardnerella and unprotected sex. Another limitation of our study is the reliance on self-reported sexual histories. Although PCR was performed every third day rather than every day, if G vaginalis were present on other days, DNA should still have been present in the samples for which PCR was performed.
In summary, nearly half of women with predominantly normal flora at enrolment had no evidence of Gardnerella. The presence of G vaginalis was more common among women with intermediate flora during the course of the study than those with normal Nugent scores. Larger studies are needed to confirm these findings; however, data such as these bring the dogma that G vaginalis is part of the ‘normal’ vaginal flora into question. In addition, one Nugent score of 0–3 is insufficient to determine characteristics of the vaginal flora throughout the menstrual cycle.
Gardnerella vaginalis is not present in all women with optimal vaginal flora.
One cross-sectional Nugent score is insufficient to determine if women consistently have optimal vaginal flora.
47% Of women had no evidence of Gardnerella vaginalis.
Handling editor Jackie A Cassell
Contributors JRS: Conception and design, analysis and interpretation of the data, preparing the manuscript and final approval of the version to be published. MSF: Acquisition of data, input into manuscript content and final approval of the version to be published. CAR: Acquisition and interpretation of the data, input into manuscript preparation and final approval of the version to be published.
Competing interests None.
Ethics approval University of Alabama at Birmingham.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.