Preterm delivery, defined as a birth before 37 weeks of gestation, occurs in 5–13% of all pregnancies.1 It is the leading cause of neonatal mortality and severe morbidity, and its economic and human costs are great.2 Bacterial vaginosis is a bacterial imbalance of vaginal flora, initially defined according to the clinical criteria of Spiegel et al.3 Its prevalence during pregnancy ranges from 6% to 55%,4,5 and it is associated with a doubled risk of preterm delivery.6,7 The earlier bacterial vaginosis is diagnosed, the higher the risk of preterm delivery.6 Although the pathogenesis and pathways relating bacterial vaginosis to preterm birth remain largely unelucidated,8 intrauterine infection or inflammation is thought to be one of these pathways.1
Recent advances in molecular techniques have increased our knowledge of the microbial ecosystem of bacterial vaginosis, associating some bacteria with bacterial vaginosis for the first time.9,10 Of these, one of the most interesting may be Atopobium vaginae, an anaerobic and fastidious bacteria recently reported to be highly associated with bacterial vaginosis.9,11,12 Specific proportions of different Lactobacillus species are related to the overgrowth of particular bacterial vaginosis–associated bacteria, with an inverse relationship between the disappearance of Lactobacillus species and the development of Gardnerella vaginalis and A vaginae in particular.11 Although G vaginalis and A vaginae are both commonly present in normal flora, high vaginal concentrations of both are highly specific for bacterial vaginosis11–13 and are associated with recurrence after treatment.12 In addition, resistance to antimicrobial agents may be due to the survival of both microorganisms as biofilm on the vaginal epithelium after therapy.14 These findings support the hypothesis that G vaginalis and A vaginae are the key species in the pathologic pathway of bacterial vaginosis. Their specific effects on pregnancy outcome are unknown.
The purpose of the study was to estimate the relationship between molecular quantification of the main bacterial vaginosis–associated bacteria and the risk of preterm delivery among women with preterm labor.
MATERIALS AND METHODS
This prospective study of pregnant women admitted for preterm labor took place from July 2007 through July 2008 at the Department of Obstetrics and Gynecology of Marseille Public Hospital System (France), a tertiary referral maternity unit with an onsite neonatal intensive care unit. Women were included if they had a singleton pregnancy and preterm labor at a gestational age of 24 to 33 6/7 weeks. Preterm labor was defined by the presence of regular and painful uterine contractions occurring at a minimum frequency of two every 10 minutes and a cervical length of 25 mm or less measured by transvaginal sonography before 34 weeks of gestation. Gestational age was determined by a combination of the last menstrual period and the first trimester ultrasonographic evaluation. The following criteria excluded women from the study: ruptured membranes, fetal malformation, maternal complications (preeclampsia or eclampsia), intrauterine growth restriction requiring early delivery, or treatment for a urinary tract or vaginal infection in the 7 days before admission. At admission, intravenous tocolysis was given to delay spontaneous delivery by 48 hours. Corticosteroid therapy (12 mg) was systemically given intramuscularly and repeated 24 hours later to promote fetal pulmonary maturation.
Vaginal samples were collected from each woman at admission, after she provided written informed consent. The institutional review board (ethics committee of the University Hospital of Nice, France) approved this study (in a decision dated May 21, 2007, reference 07.019). After placement of a nonlubricated speculum into the vaginal vault, sampling was performed with two sterile cytobrushes rotated against the vaginal wall (Scrinet, 5.5 mm; Laboratory CCD International, Paris, France). A sample from one cytobrush was rolled onto a glass slide for Gram staining. The second cytobrush was transferred to a sterile tube containing 500 microliters of BME Baral Medium (Invitrogen, Carlsbad, CA) for DNA extraction and molecular quantification, and was stored at −80°C until use.
Gram staining of the vaginal smear samples was performed with an automated stainer (model 7320 Aerospray Gram Slide Stainer; Wescor, Logan, UT); and the samples were graded with the Nugent score15 to determine the presence of normal (score of 0–3) or intermediate flora (score of 4–6) or bacterial vaginosis (score of 7–10). Two investigators, who were unaware of each other's results and clinical outcomes, graded the samples independently.
A quantitative molecular tool targeted Lactobacillus species, A vaginae, G vaginalis, Mycoplasma hominis, and a human albumin gene; it was based on a specific quantitative real-time polymerase chain reaction (PCR) assay and serial dilutions of a plasmid suspension, as previously described.13 Briefly, after DNA extraction, a quantitative real-time PCR assay was performed with a Stratagene MX 3000P (Agilent, La Jolla, CA). The amplification program was run at 50°C for 2 minutes and at 95°C for 15 minutes, followed by 45 cycles at 95°C for 30 s and at 60°C for 1 minute. Five microliters of 1) a pure undiluted DNA sample, 2) a DNA sample diluted to 1×10 microliters, 3) a DNA sample diluted to 1×100 microliters, or 4) the serially diluted plasmid suspension was added to the 20-microliter PCR mix that contained the Quantitect Probe PCR Kit mix (Qiagen, Courtaboeuf, France), the two pairs of primers, the two probes, and 100 units Uracil DNA glycosylase (Sigma-Aldrich, Saint Quentin Fallavier, France). All plasmid scale solutions were tested in duplicate. Negative controls were introduced in each reaction plate. The microbial quantification was interpreted for a sample if, for each reaction plate, 1) the standard curve remained linear and reproducible, 2) the cycle threshold values for all microorganisms tested in undiluted and diluted samples were reproducible and linear, and 3) the range of values for the number of albumin copies in the vaginal samples used as an internal control was narrow. The quality and reproducibility of each PCR result were validated, and not one of the 90 samples used in this present study had to be excluded. The quantitative molecular tool made it possible to reduce the detection limit for all microorganisms to 10 copies per 5 microliters (103 copies per milliliter). The final results were expressed as copies of microorganism DNA per milliliter of vaginal suspension.
All data were collected prospectively. The information recorded included gestational age at admission and delivery, neonatal weight, intrauterine infection, and neonatal infection. Intrauterine infection was defined as any three of the following signs: maternal fever of 38°C or higher, maternal tachycardia (more than 90 beats per minute), peripheral-blood leukocytosis (more than 15,000/mL), C-reactive protein more than 15 ng/mL, or fetal tachycardia (more than 160 beats per minute). Neonatal infection was defined as a C-reactive protein more than 20 ng/mL and at least two peripheral samples with positive bacterial culture tests. The primary outcome measure was the relationship between bacterial concentration at admission and preterm delivery, before 37 weeks of gestation. The pregnant women were subsequently classified in two groups: those who delivered before 37 weeks of gestation (preterm) and those who delivered at term (term).
The following tests were performed for statistical analysis. Group testing of categorical data are presented as n (%), and groups were compared with χ2 or Fisher exact tests, as appropriate. The sensitivity and specificity of molecular cutoff values and 95% confidence intervals (CIs) were calculated using the University of British Columbia Bayesian Calculator type 2 Kaplan-Meier survival analysis and Cox proportional hazard models were applied to examine the interval from preterm labor to delivery according to the concentrations of various microorganisms in the vaginal fluid. The groups were compared with the log-rank test.
The study included all 90 pregnant women admitted for preterm labor during the study period who met all the inclusion and none of the exclusion criteria. Table 1 summarizes their obstetric and neonatal characteristics. Gestational age at admission ranged from 24 to 33 weeks (mean 28 weeks). Gestational age at delivery ranged from 24 to 41 weeks (mean 36 weeks). The preterm group included 36 women (40%). The causes of preterm delivery were preterm premature rupture of membranes (58%), failure of tocolysis to suppress uterine activity and cervical modification (64%), and intrauterine infection (22%).
The Nugent score and results of the qualitative PCR assays are reported in Table 2. Of the 90 vaginal samples, two (2.2%) had bacterial vaginosis, 12 (13.3%) had intermediate flora, and 76 (84.5%) had normal flora according to their Nugent scores. Both women with bacterial vaginosis gave birth prematurely, the first at 35 weeks of gestation because of premature labor and the second at 36 weeks of gestation because of premature rupture of membranes. There was no association between Nugent score and preterm delivery. Preterm delivery was not associated with the presence of Lactobacillus species, G vaginalis, A vaginae, and M hominis in vaginal samples.
Table 3 reports the vaginal concentrations of Lactobacillus species, G vaginalis, A vaginae, and M hominis. High vaginal concentrations of A vaginae (106/mL or more) and G vaginalis (107/mL or more) were detected in significantly more women in the preterm group than in the term group. Accuracy of high vaginal concentrations of G vaginalis and A vaginae to predict preterm birth is presented in Table 4. Overall, high vaginal concentrations of these two bacteria were detected in 17 women (19%), 11 of whom (65%) subsequently gave birth preterm. In contrast, preterm birth occurred in 25 of 73 women (34%) with lower vaginal concentrations of both microorganisms (A vaginae less than 106/mL and G vaginalis less than 107/mL) (P=.02). Five women (5.5%) had high vaginal concentrations of both A vaginae and G vaginalis, and four of them (80%) had preterm delivery.
Survival analysis was used to assess the relationship between high vaginal concentrations of G vaginalis and A vaginae and the interval from preterm labor to delivery (Fig. 1). High vaginal concentrations of A vaginae and G vaginalis were associated with a significantly shorter interval than that in women with lower vaginal concentrations of both (log-rank P=.03). The mean interval between onset of preterm labor and delivery interval was 87 days (95% CI 76–97) for a vaginal concentration of A vaginae less than 106/mL and 58 days (95% CI 37–80) for a vaginal concentration of A vaginae of 106/mL or more (P=.03). The mean time to delivery was 86 days (95% CI 75–96) for a vaginal concentration of G vaginalis less than 107/mL and 53 days (95% CI 35–71) for a vaginal concentration of G vaginalis of 107/mL or more (P=.03). The mean interval between onset of preterm labor and delivery interval was 85 days (95% CI 75–95) with low vaginal concentrations of A vaginae (less than 106/mL) and G vaginalis (less than 107/mL), and it decreased to 46 days (95% CI 30–61) for women with higher vaginal concentrations of A vaginae and G vaginalis together (P=.03). Cox proportional hazard models examined the relationship between the duration of this interval and high vaginal concentrations of A vaginae and G vaginalis (Table 5). The highest hazard ratio for a short preterm labor-to-delivery interval was for high vaginal fluid concentrations of A vaginae and G vaginalis together (hazard ratio 3.3, 95% CI 1.1–9.5, P=.03).
The objective of this study was to assess the relationship between the quantity of the main bacterial vaginosis–associated bacteria, assessed with quantitative real-time PCR, and preterm delivery among 90 pregnant women with spontaneous preterm labor and intact membranes. We found that high vaginal concentrations of A vaginae (106/mL or more) and G vaginalis (107/mL or more) were significantly associated with preterm delivery and with a shorter interval between preterm labor and delivery.
The advantage of our quantitative real-time PCR assay is its ability to quantify microorganisms in vaginal fluid and especially its outstanding reproducibility, as we reported previously.13 The accuracy of the standard quantification was assessed by the linearity of DNA amplification of samples with known concentrations and the reproducibility of the quantification in each PCR run. Human albumin was used as an internal control to control sampling quality and ensure accurate sample comparisons. Based on this method, it was possible to show in a small and selective study that the quantity of A vaginae and G vaginalis in vaginal fluid was associated with preterm delivery contrary to their presence or absence, assessed qualitatively.
High vaginal concentrations of G vaginalis and A vaginae, assessed by molecular quantification, have previously been described as common in women with bacterial vaginosis flora.11–13 Moreover, we previously reported that the combination of A vaginae of 108/mL or more and G vaginalis of 109/mL or more had excellent sensitivity (95%), specificity (99%), and positive (95%) and negative (99%) predictive values for bacterial vaginosis diagnosis in a general obstetric population.13 These findings indicate that high vaginal concentrations of A vaginae and G vaginalis might serve as a clear definition of the microbiologic bacterial vaginosis associated with preterm delivery. Accurate identification of bacterial vaginosis microflora linked with preterm delivery should make it possible to reexamine the so-far disappointing results of bacterial vaginosis management during pregnancy.16,17
Numerous microorganisms and infections have been associated with preterm delivery.18 To study the relationship between vaginal colonization by bacterial vaginosis–associated bacteria and preterm delivery, we excluded the women with known current vaginal or urinary tract infection. Only two women (2.2%) with preterm labor had bacterial vaginosis, and both delivered before 37 weeks of gestation. The low prevalence of bacterial vaginosis may be explained by the wide variability of bacterial vaginosis prevalence in Europe.5,19,20 Bacterial vaginosis is said to double the risk of preterm delivery in women with symptoms of preterm labor.6,7 Unfortunately, in our small cohort study the prevalence of bacterial vaginosis was too low to allow us to find a significant association between bacterial vaginosis and preterm delivery.
Preterm delivery was not associated with PCR quantification of M hominis in vaginal fluid. A recent literature review reports a discordant relationship between this microbe and pregnancy outcome.21 We also found that the quantity of Lactobacillus species, measured by quantitative real-time PCR, was not associated with pregnancy outcome. This contradicts the results of several studies that have reported a lower probability of chorioamnionitis and preterm delivery and of second-trimester pregnancy loss in women colonized by normal vaginal flora and H2O2-producing lactobacilli.22–24 However, our specific real-time PCR targeted limited numbers of important Lactobacillus species, including L gasseri, L crispatus, L jensenii, L iners, and L acidophilus. It was unable to differentiate H2O2-producing lactobacilli and the Lactobacillus species that are closely related to the normal vaginal flora from those associated with bacterial vaginosis flora. Others have reported that L crispatus is predominant in normal vaginal flora, and L iners is predominant in bacterial vaginosis flora.11,25
Some points must be discussed as potential limitations of our study. First, the significant limitation of our study is the small number of women included. The study population was restricted to pregnant women with preterm labor to ensure that we obtained a high prevalence of preterm delivery. Unfortunately, the small number of selective preterm labor women may explain the absence of association of some high levels of bacterial concentrations with preterm delivery. As a consequence, caution is thus needed in generalizing our results regarding whether the presence of vaginal bacteria is predictive of preterm delivery. A larger study is needed to determine whether the presence of vaginal microorganism is predictive of preterm delivery. Second, we did not attempt to test for each bacterial species known to be found in the vagina, but the spectrum of targeted microorganisms chosen was significant in bacterial vaginosis. It is possible that quantification of additional vaginal microorganisms would modify the results. Finally, it is a preliminary study of unique value (vaginal quantification of bacterial colonization) and supports the need for further testing of associations, such as vaginal symptoms, vaginal pH, and H2O2-producing lactobacilli.
1. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet 2008;371:75–84.
2. Gilbert WM. The cost of preterm birth: the low cost versus high value of tocolysis [published erratum appears in BJOG 2008;115:674–5]. BJOG 2006;113(suppl 3):4–9.
3. Spiegel CA, Amsel R, Holmes KK. Diagnosis of bacterial vaginosis by direct gram stain of vaginal fluid. J Clin Microbiol 1983;18:170–7.
4. Guaschino S, Ricci E, Franchi M, Frate GD, Tibaldi C, Santo DD, et al. Treatment of asymptomatic bacterial vaginosis to prevent pre-term delivery: a randomised trial. Eur J Obstet Gynecol Reprod Biol 2003;110:149–52.
5. Hogan VK, Culhane JF, Hitti J, Rauh VA, McCollum KF, Agnew KJ. Relative performance of three methods for diagnosing bacterial vaginosis during pregnancy. Matern Child Health J 2007;11:532–9.
6. Leitich H, Bodner-Adler B, Brunbauer M, Kaider A, Egarter C, Husslein P. Bacterial vaginosis as a risk factor for preterm delivery: a meta-analysis. Am J Obstet Gynecol 2003;189:139–47.
7. Leitich H, Kiss H. Asymptomatic bacterial vaginosis and intermediate flora as risk factors for adverse pregnancy outcome. Best Pract Res Clin Obstet Gynaecol 2007;21:375–90.
8. Larsson PG, Forsum U. Bacterial vaginosis: a disturbed bacterial flora and treatment enigma. APMIS 2005;113:305–16.
9. Fredricks DN, Fiedler TL, Marrazzo JM. Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 2005;353:1899–911.
10. Fredricks DN, Fiedler TL, Thomas KK, Oakley BB, Marrazzo JM. Targeted PCR for detection of vaginal bacterial associated with bacterial vaginosis. J Clin Microbiol 2007;45:3270–6.
11. De Backer E, Verhelst R, Verstraelen H, Algumber MA, Burton JP, Tagg JR, et al. Quantitative determination by real-time PCR of four vaginal Lactobacillus species, Gardnerella vaginalis and Atopobium vaginae indicates an inverse relationship between L. gasseri and L. iners. BMC Microbiol 2007;7:115.
12. Bradshaw CS, Tabrizi SN, Fairley CK, Morton AN, Rudland E, Garland SM. The association of Atopobium vaginae and Gardnerella vaginalis with bacterial vaginosis and recurrence after oral metronidazole therapy. J Infect Dis 2006;194:828–36.
13. Menard JP, Fenollar F, Henry M, Bretelle F, Raoult D. Molecular quantification of Gardnerella vaginalis and Atopobium vaginae loads to predict bacterial vaginosis. Clin Infect Dis 2008;47:33–43.
14. Swidsinski A, Mendling W, Loening-Baucke V, Swidsinski S, Dorffel Y, Scholze J, et al. An adherent Gardnerella vaginalis biofilm persists on the vaginal epithelium after standard therapy with oral metronidazole. Am J Obstet Gynecol 2008;198:97.e1–6.
15. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol 1991;29:297–301.
16. Nygren P, Fu R, Freeman M, Bougatsos C, Klebanoff M, Guise JM; U.S. Preventive Services Task Force. Evidence on the benefits and harms of screening and treating pregnant women who are asymptomatic for bacterial vaginosis: an update review for the U.S. Preventive Services Task Force. Ann Intern Med 2008;148:220–33.
17. McDonald HM, Brocklehurst P, Gordon A. Antibiotics for treating bacterial vaginosis in pregnancy. The Cochrane Database of Systematic Reviews 2007, Issue 1. Art. No.: CD000262. DOI: 10.1002/14651858.CD000262.pub3.
18. Klein LL, Gibbs RS. Use of microbial cultures and antibiotics in the prevention of infection-associated preterm birth. Am J Obstet Gynecol 2004;190:1493–502.
19. Hauth JC, Goldenberg RL, Andrews WW, DuBard MB, Copper RL. Reduced incidence of preterm delivery with metronidazole and erythromycin in women with bacterial vaginosis. N Engl J Med 1995;333:1732–6.
20. Goffinet F, Maillard F, Mihoubi N, Kayem G, Papiernik E, Cabrol D, et al. Bacterial vaginosis: prevalence and predictive value for premature delivery and neonatal infection in women with preterm labour and intact membranes. Eur J Obstet Gynecol Reprod Biol 2003;108:146–51.
21. Waites KB, Katz B, Schelonka RL. Mycoplasmas and ureaplasmas as neonatal pathogens. J Clin Microbiol Rev 2005;18:757–89.
22. Wilks M, Wiggins R, Whiley A, Hennessy E, Warwick S, Porter H, et al. Identification and H2
production of vaginal lactobacilli from pregnant women at high risk of preterm birth and relation with outcome. Clin Microbiol 2004;42:713–7.
23. Holst E, Goffeng AR, Andersch B. Bacterial vaginosis and vaginal microorganisms in idiopathic premature labor and association with pregnancy outcome. J Clin Microbiol 1994;32:176–86.
24. Nelson DB, Bellamy S, Nachamkin I, Ness RB, Macones GA, Allen-Taylor L. First trimester bacterial vaginosis, individual microorganism levels, and risk of second trimester pregnancy loss among urban women. Fertil Steril 2007;88:1396–403.
© 2010 by The American College of Obstetricians and Gynecologists.
25. Verhelst R, Verstraelen H, Claeys G, Verschraegen G, Delanghe J, Van Simaey L, et al. Cloning of 16S rRNA genes amplified from normal and disturbed vaginal microflora suggests a strong association between Atopobium vaginae, Gardnerella vaginalis and bacterial vaginosis. BMC Microbiol 2004;4:16.