Seminars in Fetal & Neonatal Medicine
Volume 17, Issue 1 , Pages 2-11, February 2012

Diversity of microbes in amniotic fluid

  • Daniel B. DiGiulio

      Affiliations

    • Department of Medicine, Stanford University School of Medicine, Stanford, CA, USA
    • Veterans Affairs Palo Alto Health Care System, Palo Alto, CA, USA
    • Corresponding Author InformationCorresponding author. Address: Division of Infectious Diseases and Geographic Medicine, Department of Medicine, Stanford University School of Medicine, 300 Pasteur Drive, Room S-101, Stanford, CA 94305-5107, USA. Tel.: +1 650 723 6661; fax: +1 650 433 8887.

published online 05 December 2011.

Article Outline

Summary 

Recent polymerase chain reaction (PCR)-based studies estimate the prevalence of microbial invasion of the amniotic cavity (MIAC) to be ≥30–50% higher than that detected by cultivation-based methods. Some species that have been long implicated in causing MIAC remain among the common invaders (e.g. Ureaplasma spp., Mycoplasma spp., Fusobacterium spp. Streptococcus spp., Bacteroides spp. and Prevotella spp.). Yet we now know from studies based on PCR of the 16S ribosomal DNA that cultivation-resistant anaerobes belonging to the family Fusobacteriaceae (particularly Sneathia sanguinegens, and Leptotrichia spp.) are also commonly found in amniotic fluid. Other diverse microbes detected by PCR of amniotic fluid include as-yet uncultivated and uncharacterized species. The presence of some microbial taxa is associated with specific host factors (e.g. Candida spp. and an indwelling intrauterine device). It appears that MIAC is polymicrobial in 24–67% of cases, but the potential role of pathogen synergy is poorly understood. A causal relationship between diverse microbes, as detected by PCR, and preterm birth is supported by types of association (e.g. space, time and dose) proposed as alternatives to Koch's postulates for inferring causality from molecular findings. The microbial census of the amniotic cavity remains unfinished. A more complete understanding may inform future research directions leading to improved strategies for preventing, diagnosing and treating MIAC.

keywords: Chorioamnionitis, Intra-amniotic infection, Leptotrichia species, Microbial diversity, Preterm birth, Sneathia sanguinegens

 

Back to Article Outline

1. Introduction 

Intra-amniotic infection resulting from microbial invasion of the amniotic cavity (MIAC) is considered a leading etiology of preterm birth,1, 2, 3 which, in turn, is the chief cause of neonatal mortality worldwide.4 Yet our knowledge of MIAC remains incomplete, including its prevalence, optimal diagnosis, pathogenic mechanisms, and host susceptibilities. An improved understanding of the diverse microbial taxa involved in MIAC is one potentially critical step in addressing these knowledge gaps. For example, studies that characterize the microbial causes of MIAC may pinpoint species that warrant further investigations to elucidate clinically relevant traits such as virulence mechanisms, evasion of the host immune system, and antimicrobial susceptibilities. These types of insights, in conjunction with the application of improved microbial detection methods in the clinical setting may facilitate the development of new prevention, diagnosis and treatment strategies.

The primary aim of this review is to present current knowledge of the diversity of microbial taxa that invade the human amniotic cavity. Particular emphasis is placed on emerging knowledge from recent molecular studies, including evidence for causal associations with preterm birth and its adverse clinical sequelae. Related topics beyond this review's scope that are addressed elsewhere in this issue include the microbial diversity of the genital tract, and mechanisms of intrauterine infection and preterm labor.

Back to Article Outline

2. Microbial detection methods for investigating MIAC: strengths and limitations 

The question of whether the amniotic cavity is sterile during healthy pregnancy, and remains so throughout gestation, has been considered for at least a century.5, 6 Natural corollaries to this query are: what is the prevalence of MIAC, and what are the microbial species that cause it? Our ability to answer these questions is a function of the methods available for their investigation; as methods improve so too does our knowledge. Increasingly, molecular methods have been used to supplement culture-based methods in recent studies of MIAC.

2.1. Cultivation-based microbiologic approaches 

Pure culture methods provide a powerful approach to microbial analysis based on morphologic, biochemical, and growth characteristics, among others. However, it has been long recognized from studies of complex microbial communities that up to 99% or more of microbial species visualized by microscopy are uncultivable, a phenomenon termed the ‘great plate count anomaly’. Microbes readily recovered in the laboratory have been likened to the ‘weeds’ of the microbial world: species that grow rapidly on generic media, under aerobic conditions, and at moderate temperatures. Due to this systematic bias of culture methods, it follows that species recovered most frequently in pure culture may not be numerically dominant nor of foremost clinical significance among the microbes present in a sampled environment.

2.2. Molecular microbiologic approaches 

Molecular approaches such as polymerase chain reaction (PCR) enable the detection of microbes without the need to isolate them in pure culture, thereby overcoming inherent biases of culture-based methods. Initial PCR assays for microbial detection were taxon-specific (i.e. they targeted microbes at narrow taxonomic levels, such as species or genus). However, the development of broad-range PCR enabled microbial detection at broad taxonomic levels such as the domain Bacteria (which is comprised of all bacterial species, including as-yet uncultivated and unknown members). This is achieved by means of ‘universal’ PCR primers that bind to conserved sequence regions of genes that are ubiquitous in microbes, such as the 16S ribosomal DNA (rDNA). These conserved regions are flanked by variable regions that are also amplified during the PCR reaction. Sequencing these variable regions permits sequence comparisons to a reference database, thus enabling microbial identification.

Back to Article Outline

3. Bacterial causes of MIAC 

3.1. Initial investigations of MIAC: early insights from indirect evidence 

In a seminal study published in 1927 of the presence of bacteria in amniotic fluid of women undergoing caesarean section, Harris and Brown found that all 28 subjects who were in labor for <6 h had negative cultures, whereas all 22 subjects who were in labor for >6 h had positive cultures.5 These findings bolstered the prevailing presumption for much of the 20th century that the amniotic cavity invariably remains sterile prior to labor onset.7, 8 Yet, some authors recognized that early onset neonatal pneumonia and sepsis occurring within three days after delivery implicated the prepartum intra-amniotic environment as a potential source of infection,9, 10, 11 and that additional studies were warranted.8, 9

3.2. Cultivation-based investigations of MIAC: direct evidence from bacteriologic studies 

The advent of techniques for direct sampling of the amniotic cavity – including amniocentesis via the transabdominal and transcervical approaches, and intrauterine catheter placement – facilitated the study of the intra-amniotic milieu prior to labor onset. Harwick et al. analyzed 50 women, most of whom were undergoing elective pregnancy termination at 14–20 gestational weeks. The authors recovered Mycoplasma hominis and Staphylococcus epidermidis from four and two cases, respectively; they hypothesized that M. hominis accessed the intrauterine environment during vaginal examination and that S. epidermidis represented skin-associated contamination.12 Bobitt and Ledger used quantitative cultivation methods to recover bacteria from amniotic fluid in seven of 10 cases of spontaneous preterm labor with intact membranes. This study provided bacteriologic evidence linking clinically silent amnionitis with preterm labor, and highlighted the importance of anaerobic bacteria including Bacteroides species.13

Subsequent studies extended the findings of clinically silent bacterial amnionitis. Cassell et al. recovered M. hominis or U. urealyticum from four of 33 cases of discolored amniotic fluid collected in early pregnancy in the setting of intact membranes. U. urealyticum was associated with chronic and clinically silent intrauterine infections characterized by an intense inflammatory response, and complications in three of four patients suggested an adverse effect on pregnancy outcome.14 A number of other studies that investigated MIAC by means of culture-based methods have been reviewed elsewhere.3, 15 From these studies in aggregate, the emergent picture of MIAC was of a clinically silent infection most frequently caused by one of several bacterial genera (e.g. Ureaplasma, Mycoplasma, or Fusobacterium; but also Streptococcus, Bacteroides, Gardnerella, or Candida spp.) and disproportionately associated with early (<32 gestational weeks) preterm births.

3.3. Investigations of MIAC based on taxon-specific PCR: emphasis on Ureaplasma species and Mycoplasma species 

Initial PCR-based studies of MIAC emphasized assays targeting Ureaplasma species and/or Mycoplasma species because prior cultivation-based studies implicated these taxa as the most common invaders of the amniotic cavity. Two early studies by Blanchard et al. demonstrated the high sensitivity of PCR for detecting U. urealyticum and Mycoplasma species in amniotic fluid.16, 17 Subsequent investigators leveraged these findings to study the association of MIAC with various clinical phenotypes.

Yoon et al. found the prevalence of U. urealyticum in women with preterm PROM was 28% (43/154) based on PCR and 16% (25/154) based on culture.18 In a separate study of preterm labor with intact membranes, the same authors found U. urealyticum prevalence to be 6% (15/254) by PCR; 40% (6/15) of these were negative by culture.19 In two PCR-based studies of MIAC during the midtrimester, Gerber et al. found the prevalence of U. urealyticum to be 11.4% (29/254),20 and Perni et al. found the prevalences of U. urealyticum and M. hominis to be 12.8% (22/172) and 6.1% (11/179), respectively.21 In women with suspected cervical insufficiency, Bujold et al. found the rate of MIAC caused by Ureaplasma spp. or Mycoplasma spp. to be 47% (7/15) overall, and 33% (5/15) by culture methods.22 Oyarzún et al. used a panel of 16 taxon-specific PCR assays to study 50 women with preterm labor and intact membranes, and found the prevalence of MIAC to be 46% (23/50) by means of PCR and 12% (6/50) by means of culture (P < 0.0001).23 Leon et al. used a similar approach (but included a panel of eight PCR assays targeting oral cavity pathogens) to analyze women in preterm labor with intact membranes or with preterm PROM, and detected Porphyromonas gingivalis in both amniotic fluid and subgingival material in 30.8% (8/26) of subjects.24

These studies based on taxon-specific PCR demonstrated in aggregate that the true rate of MIAC caused by Ureaplasma spp. and Mycoplasma spp. was higher than that indicated by culture methods, and that this finding held true across various clinical phenotypes. What remained largely unanswered was the degree to which other diverse microbial species were involved in MIAC.

3.4. Investigations of MIAC based on broad-range PCR: profiling microbial diversity with discovery methods 

Broad-range PCR has been used more recently to assay amniotic fluid in studies of various clinical phenotypes. Table 1 summarizes studies25, 26, 27, 28, 29, 30, 31, 32 of two phenotypes of chief interest – preterm labor with intact membranes and preterm PROM – that reported positive findings from more than one subject. For each clinical phenotype, most studies found an overall prevalence that was ∼30–50% higher than that detected by culture-based methods alone (Table 1). Even this higher overall prevalence is likely an underestimation because PCR is an imperfect method and some studies were limited by sample volumes27, 31 or by sample age (e.g. see Fig. 7 of DiGiulio et al.31). Species richness is another parameter of MIAC that culture-based methods have consistently underestimated in studies. The number of detected taxa overall ranged from about 1.5–3.5 times the number recovered by culture methods alone (Table 1).

Table 1. Studies of preterm labor or preterm PROM that included broad-range PCR methods.
Broad-range PCR study, by clinical phenotypeNo. of study subjects with reported phenotypeSequencing depth (clones/PCR reaction)MIAC prevalence as detected by:Prevalence (total to culture-based)No. of microbial taxa detected by:Polymicrobial rate as detected by:
CulturePCRCulture and PCR combinedCulturePCRCulture and PCR combinedCulturePCRCulture and PCR combined
Preterm labor
Hitti et al.25,a69423.2%30.4%31.9%137.5%NAbNAbNAb37.5%NAbNAb
Markenson et al.26,c54(Not done)9.3%55.6%55.6%600.0%6NAbNAb3.3%NAbNAb
DiGiulio et al.27166249.6%11.4%15.1%156%1114188.0%16.0%24.0%
Han et al.28,d261038.5%50.0%50.0%130%7121615.4%46.2%61.5%
Marconi et al.29,c2010NA40.0%40.0%NAbNAb5NAbNAb62.5%NAb
Preterm PROM
Jalava et al.3017111.8%17.6%17.6%150%235033.3%66.7%
Han et al.28,d201030.0%40.0%40.0%133%6e711e25.0%062.5%
DiGiulio et al.31,c2041034.3%45.1%49.5%144%14444811.9%21.8%33.7%

PCR, polymerase chain reaction; MIAC, microbial invasion of the amniotic cavity; PROM, premature rupture of the membranes.

aThe PCR-based findings from this study that are included in this table were presented in a follow-up report.32

bResult listed as not applicable (NA) because the study either did not perform these analyses, or reported the results in insufficient detail.

cThese studies included in their approach taxon-specific PCR assays, in addition to broad-range PCR assays.

dThis single study included both clinical phenotypes in the study population, thus the study appears twice in this table.

eThis study reported some results as ‘mixed anaerobes’ without additional details; these are counted in this table as a single taxon.

From these studies based on broad-range PCR, an early picture of the microbial diversity of the amniotic cavity is beginning to emerge. Fig. 1 presents the phylum-level taxonomic distribution of bacteria associated with MIAC based on 212 subjects from three studies of preterm labor (Fig. 1A), and 267 subjects from three studies of preterm PROM (Fig. 1B) (the numbers of subjects and studies vary from Table 1 owing to study methodologies that impact the subsequent analyzability of subjects for generating Fig. 1). This broad overview demonstrates that the ‘big four’ phyla that predominate in most human-associated habitats (i.e. Firmicutes, Bacteroidetes, Actinobacteria, and Proteobacteria)33 are well represented in both preterm parturition phenotypes. Yet, the relatively high proportion of members of a fifth phylum, Fusobacteria (31% in preterm labor, and 10.5% in preterm PROM), is a clear difference from most other human body habitats that have been studied.33 This suggests that significant habitat filtering is contributing to a distinct microbial profile of the intra-amniotic space, which appears to be discernible at even very broad taxonomic levels.

  • View full-size image.
  • Fig. 1 

    Phylum-level distribution of the bacterial taxa detected in studies that used broad-range polymerase chain reaction methods to assay amniotic fluid in the setting of either preterm labor with intact membranes or of preterm premature rupture of the membranes (PROM). Only studies that reported findings in a manner that enabled taxonomic analysis of all subjects are included.27, 28, 29, 30, 31

In addition to the potential for a distinct microbial profile that is specific to the amniotic cavity, it is possible that stereotypic differences exist in this profile with respect to clinical phenotype (e.g. preterm labor and intact membranes versus preterm PROM). For example, although the Fusobacteria and the Tenericutes together comprise about half of the taxonomic distribution found in each setting, their relative proportions appear to be quite different across phenotypes. Similarly, the Bacteroidetes and the Firmicutes have been detected to date at much different relative frequencies in each phenotype (Fig. 1A vs. B).

A higher-resolution snapshot of the taxonomic distribution in amniotic fluid is illustrated by rank abundance curves based on genus-level classification (Fig. 2). One notable pattern is the relatively high prevalence of the genera Sneathia and Leptotrichia, which is unexpected based on their near-absence in culture-based studies. Indeed, it now appears that Sneathia species are about as prevalent in MIAC as are Mycoplasma species (Fig. 2), which for decades have been considered among the most frequent invaders of the amniotic cavity. Interestingly, the pathogenic potential of Sneathia spp. and Leptotrichia spp. has garnered increasing support from recent case reports, including at least eight maternal cases and two neonatal cases of bacteremia in the peripartum period or shortly after septic abortion.34

  • View full-size image.
  • Fig. 2 

    Genus-level rank abundance curves of bacterial prevalence based on studies that used broad-range polymerase chain reaction (PCR) to assay amniotic fluid. (A) All 21 taxa reported in studies of women in preterm labor with intact membranes. (B) The 18 taxa that were detected in two or more subjects in studies of women with preterm premature rupture of the membranes; 21 additional taxa detected in one subject apiece are not shown (Actinomyces, Atopobium, Brachybacterium, Campylobacter, Clostridium, Coprobacillus, Dialister, Eikenella, Eubacterium, Faecalibacterium, Filifactor, Finegoldia, Kingella, Klebsiella, Mobiluncus, Myroides, Neisseria, Rothia, †uncultured Clostridiaceae (clone 209-b07), †uncultured Clostridiaceae (clone 209-b10), and †TM-7 like). Only studies that reported findings in a manner that enabled taxonomic analysis of all subjects are included.27, 28, 29, 30, 31 *The two subjects with Escherichia species were positive by both PCR and culture; in both cases, Escherichia assignment is based on culture findings because the 16S rDNA gene is unreliable for differentiating between the genera Escherichia and Shigella. †Denotes the taxa that could not be classified to genus-level resolution, owing to current limitations of reference sequence databases.

Another consistent pattern evident from Fig. 2 is the rarity of Lactobacillus species relative to their presumed prevalence in the vagina, thereby providing further evidence for habitat filtering. However, additional studies characterizing the vaginal microbiota during pregnancy are needed to confirm this presumption.

Other features of the genus-level profile appear to vary across the two phenotypes. The taxonomic distribution in preterm labor (Fig. 2A) appears to be characterized by higher evenness and a more gradated ‘step-down’ pattern than in preterm PROM (Fig. 2B), which shows a highly skewed distribution with a long tail due to an overwhelming predominance of Ureaplasma. Furthermore, it is possible that the observed differences in rank order of specific genera between the two phenotypes are meaningful. However, it is important to interpret existing data with caution because some differences may be biased by variations in patient populations, study methods (e.g. DNA extraction methods, and the inclusion of taxon-specific PCR assays in some studies), or other factors. Additional studies are clearly needed to enable a more robust analysis.

Other clinical settings have also been studied by means of broad-range PCR of amniotic fluid. These include pre-eclampsia (n = 62 subjects),35 small-for-gestational-age fetus (n = 52 subjects),36 immediately prior to cesarean section (n = 48 subjects),37 and several reports of single cases of MIAC in various clinical phenotypes.38, 39, 40, 41 Due to the limited sizes of these studies, their findings were primarily descriptive rather than statistical.

Back to Article Outline

4. Non-bacterial taxa associated with MIAC 

4.1. Fungi 

Members of the genus Candida appear to be the only fungi to invade the amniotic cavity with appreciable frequency. Candida albicans is by far the predominant species recovered directly from amniotic fluid, but Candida glabrata has also been isolated from the amniotic cavity,42 and Candida parapsilosis has been reported to cause fetal infection.43 Cases of perinatal candida infections (that presumably originated as MIAC) are often associated with very early preterm births (<28 weeks) and grave perinatal outcomes. Despite the clinical importance of Candida species, they appear to be infrequent invaders of the amniotic cavity based on their low prevalence in amniotic fluid44 relative to their prevalence in the vagina of pregnant women (up to 38%).45 This is consistent with the lifestyle of Candida species as opportunistic pathogens.

Chaim et al. reviewed cultivation-based studies of amniotic fluid retrieved by transabdominal amniocentesis from subjects in preterm labor with intact membranes (13 studies; 773 subjects) and subjects with preterm PROM (eight studies; 625 subjects). The prevalence of Candida spp. was ∼0.65% for each clinical phenotype.44 However, this estimate is likely an underestimation. More recent studies that included PCR-based methods found a higher prevalence overall than that revealed by culture in the settings of preterm labor with intact membranes [1.2% (2/166) vs. 0.6% (1/166), respectively]27 and of preterm PROM [5% (11/221) versus 3.2% (7/221), respectively].31

Compelling evidence now implicates the presence of an IUD during pregnancy as a strong risk factor for MIAC caused by Candida species. Anecdotal reports first suggested Candida albicans as a cause of septic abortion associated with indwelling IUDs46; subsequent reports strengthened the association of intrauterine devices with perinatal infections due to Candida species.42, 44, 47, 48 However, the first statistical support came from Kim et al., who analyzed 196 pregnancies associated with an indwelling IUD and found candida MIAC, as diagnosed by culture, to occur more frequently in pregnancies with, than in those without, an IUD (31.1% vs. 6.3%; P < 0.001).49 A PCR-based study of subjects in preterm PROM found a similar, statistically significant disparity in the rate of MIAC caused by Candida species in those with and without an IUD [28% (5/18) vs. 3% (6/203), respectively; P < 0.01].31

An association of cervical cerclage with MIAC or related infections (e.g. funisitis or chorioamnionitis) caused by Candida species has also been reported42, 44, 47, 48; however, further studies are needed to characterize this association more precisely. Because the primary indication for cerclage (i.e. cervical incompetence) is itself associated with MIAC,50 it is important that such studies analyze MIAC caused by any microbial species (and not solely Candida species).

The propensity of Candida species to form biofilms on foreign bodies, including IUDs51 and cerclage material, is presumed to underpin these associations.

4.2. Archaea 

Archaea is one of three recognized domains of life.52 Archaeal species have been previously detected in the vagina and at sites of tissue inflammation in the oral cavity. These two observations suggest that archaea have both the ‘opportunity’ (i.e. residence in a site near the amniotic cavity) and the ‘means’ (i.e. pathogenic potential) to cause MIAC. Broad-range PCR to detect archaea has been used in four studies of MIAC, each of which investigated a distinct clinical phenotype (preterm labor with intact membranes, preterm PROM, pre-eclampsia, and small-for-gestational-age fetus).27, 31, 35, 36 These studies had a combined subject population of 484 women and found no cases of archaeal MIAC. This suggests that archaea rarely, if ever, invade the amniotic cavity. Other possible explanations are that archaeal abundance in amniotic fluid is consistently below the detection limit of the PCR assays used, or that archaeal DNA extraction from clinical specimens is refractory to one or more steps (e.g. cell lysis) of the molecular protocols used.

4.3. Viruses 

Few systematic studies of viral prevalence and diversity in the amniotic cavity have been undertaken. Wenstrom et al. studied 62 cases of spontaneous pregnancy loss occurring within 30 days of amniocentesis (out of 11 971 procedures) plus matched controls. Using a panel of PCR assays targeting seven viral taxa, the authors found viral nucleic acids in 8% (5/62) of cases and 15% (9/60) of controls. Of positive cases, adenovirus accounted for most (64%; 9/14), followed by cytomegalovirus (21%; 3/14), herpes simplex virus (7%; 1/14), and human parvovirus B19 (7%; 1/14); they found no cases of enteroviruses, Epstein–Barr virus or influenza A virus. The authors concluded that viral presence in midtrimester amniotic fluid was not significantly associated with early post-amniocentesis pregnancy loss.53

Baschat et al. used a near-identical panel of PCR assays as that of Wenstrom et al. (but with respiratory syncytial virus substituted for influenza A virus) to analyze 686 midtrimester amniotic fluid samples. They found viral nucleic acids in 6.4% (44/686). Of positive samples, adenovirus accounted for most (84%; 37/44), followed by cytomegalovirus (11%; 5/44), enterovirus (4.5%; 2/44), Epstein–Barr virus (4.5%; 2/44), and respiratory syncytial virus (2%; 1/44).54 A follow-up study found no association of a positive PCR with adverse perinatal outcome.55

Other viral taxa that have been detected by culture or PCR of amniotic fluid include human immunodeficiency virus, influenza A virus, rubella virus, and varicella-zoster virus. Although some viral taxa are well-described causes of congenital infections, birth defects, miscarriage, and fetal demise, their association with MIAC leading to preterm birth warrants additional systematic investigations based on viral discovery methods, such as microarrays or high-throughput nucleic acid sequencing.

4.4. Protozoa 

Protozoa are important causes of congenital infections that typically arise as fetal infections in utero. In contrast with models of pathogenesis of bacterial MIAC (in which microbial replication in amniotic fluid may be critical for spread of infection to the fetus or other tissues56), the propagation of protozoa in amniotic fluid is not an apparent prerequisite for fetal infection. Rather, maternal parasite transfer to the fetus occurs primarily through placental membranes (e.g. Trypanosoma cruzi) or by crossing villous trophoblast (e.g. Toxoplasma gondii). Thus protozoal presence in amniotic fluid may not be critical for pathogenesis, and is likely a consequence of fetal infection rather than a cause.

With the notable exception of T. gondii, for which a large body of literature has established PCR of amniotic fluid as a mainstay of diagnosis, studies involving the detection of protozoa in amniotic fluid have been limited and unremarkable. One study that included cultivation methods for Trichomonas spp. in its methodology failed to detect this parasite in amniotic fluid from 26 subjects with preterm PROM.57 In a PCR-based study of congenital Trypanosoma cruzi infection, the authors concluded that the release of parasites in amniotic fluid is probably a rare event that is unhelpful for the routine diagnosis of congenital Chagas' disease.58 Finally, placental malaria is a well-known condition but studies regarding the detection of Plasmodium spp. in the amniotic cavity are lacking.

4.5. Microbial co-occurrence and pathogen synergy 

Combined PCR and culture-based studies of women with preterm labor or preterm PROM have demonstrated that the percentage of MIAC cases that are polymicrobial range from 24% to 67%, which is higher than the rate found in these same studies by either culture-based methods (0–38%) or PCR-based methods (0–63%) alone (Table 1). Studies remain too limited to date to enable robust statistical analyses of co-occurrence of particular species, but such analyses may be important. Animal models reveal polymicrobial infections to be dynamic and interactive processes in which avirulent and even beneficial59 strains may synergistically enhance pathogenicity. Pathogen synergy is orchestrated by elaborate interactions between microbes possessing diverse capabilities,59 including biofilm formation, phagocytosis inhibition, toxin production, nutrient provision, and transcriptional modulation of virulence factors in co-located species, among others. Unravelling the potential role of these various interactions in contributing to the pathogenesis of MIAC seems a daunting task; however, a more complete understanding of the microbial diversity in amniotic fluid, including co-occurrence of specific taxa, is a logical first step toward this aim, and may identify microbial groups of interest for more detailed studies of pathogenesis.

Back to Article Outline

5. Clinical significance of the diverse microbes detected by PCR of amniotic fluid 

The discovery potential of broad-range PCR is highlighted by the fact that at least six taxa in Fig. 2 are as yet uncultivated, and some have highly divergent 16S rDNA sequences from their nearest database relative (indicating that they represent previously uncharacterized species). Although a large body of compelling evidence indicates that MIAC, as detected by cultivation methods, plays a causal role in preterm birth,1, 2, 3, 60 there are legitimate reasons to question whether the detection of diverse microbes by PCR-based methods have equivalent clinical relevance. First, PCR assays are capable of detecting DNA from non-viable microbes. Second, the high sensitivity of PCR makes this technique susceptible to false-positive results from even low levels of contaminating DNA. Investigators have addressed these issues by correlating PCR-generated findings with clinically relevant parameters, such as host inflammation, and pregnancy and neonatal outcomes. These correlations are informative because, for example, potential DNA contamination would presumably affect amniotic fluid specimens randomly and therefore exhibit no significant correlations with outcomes.

5.1. Associations of MIAC with host inflammation, preterm delivery, and neonatal morbidity 

Associations between the PCR-based diagnosis of MIAC and either host inflammation or perinatal outcomes have been investigated primarily in the settings of preterm labor with intact membranes,19, 25, 26, 27, 28, 29 and preterm PROM.18, 28, 31 All studies that were sufficiently powered (e.g. comprised of >50 subjects) found statistically significant associations between MIAC and one or more of four categories of clinically relevant measures: (i) elevated concentrations of inflammatory markers in amniotic fluid, including white blood cells,18, 19, 27, 31 interleukin (IL)-1β,29 IL-6,18, 19, 25, 26, 27, 28, 29 tumor necrosis factor (TNF)-α,25 and protein biomarkers measured by mass spectrometry28; (ii) increased rates of host tissue inflammation, including histologic chorioamnionitis18, 27, 28, 31 and funisitis18, 19, 27, 28, 31; (iii) the outcome of preterm delivery, based on lower median18, 19, 25, 31 or mean26 gestational age at delivery, or a high positive predictive value for the outcome of preterm delivery (e.g. 100%)27; (iv) neonatal morbidity, as defined by a composite of various adverse neonatal outcomes,18, 19, 27 or by specific outcomes, including the respiratory distress syndrome,31 necrotizing enterocolitis,31 and early onset neonatal sepsis.28

Interestingly, the frequency of a positive PCR assay (Fig. 4A of DiGiulio et al.27) has been shown to have an inverse correlation with gestational age that is remarkably similar to that shown for a positive culture result (Fig. 1 of Watts et al.61). This is important because mortality also exhibits an inverse relationship with gestational age, such that early preterm neonates (<32 gestational weeks) account for the vast majority of deaths.62

Studies of the predictive value of PCR in asymptomatic women undergoing midtrimester amniocentesis found that a positive PCR was associated with elevated intra-amniotic concentrations of IL-4,21 and with the outcomes of preterm labor and preterm birth,20 and of preterm PROM.21

Other reports of various clinical phenotypes found associations between MIAC and either host inflammation or perinatal outcomes, but these relationships were primarily descriptive due to limited study sizes. These phenotypes included preterm labor with intact membranes,23, 39 pre-labor rupture of membranes,30 pre-eclampsia,35 suspected cervical insufficiency,22 and midtrimester septic abortion.38

5.2. Associations of MIAC with time interval from amniocentesis to delivery 

A statistically significant temporal association between an exposure and a subsequent event suggests a causal relationship. This has prompted investigators to correlate the finding of a positive PCR of amniotic fluid with a shortened time interval to delivery. Four studies that used survival analysis methods found an association between a positive PCR and timing of delivery that was striking in its consistency across all studies (totalling 778 analyzable subjects with either preterm labor and intact membranes,19, 27 or preterm PROM18, 31) (Fig. 3). Other studies also found significant associations between a positive PCR of amniotic fluid and either the median25 or mean26 time from amniocentesis to delivery.

  • View full-size image.
  • Fig. 3 

    Kaplan–Meier survival analysis of the amniocentesis-to-delivery interval according to results of polymerase chain reaction (PCR) and culture of amniotic fluid from four separate studies of two clinical phenotypes [preterm labor with intact membranes, or preterm premature rupture of the membranes (PROM)]. (A and B) Studies that used taxon-specific PCR assays to detect Ureaplasma spp. (C and D) Studies that used broad-range PCR assays. Sources: (A) Yoon et al.18; Courtesy of Elsevier (B) Yoon et al.19; Courtesy of Elsevier (C) DiGiulio et al.27; (D) DiGiulio et al.31 Courtesy of John Wiley.

5.3. Associations of MIAC with microbial abundance 

A change in the magnitude of an exposure (i.e. dose) that is associated with an increase in the rate or severity of a disease (i.e. response) represents a biological gradient and suggests a causal association. The two studies that analyzed microbial abundance in amniotic fluid by means of real-time quantitative PCR (qPCR) for bacterial rDNA found a significant linear correlation (i.e. dose–response) with early gestational age at delivery in women in preterm labor with intact membranes (r2 = 0.42; P < 0.002)27 and in women with preterm PROM (r2 = 0.26; P < 0.01).31

Back to Article Outline

6. Evaluating causality: molecular Koch's postulates as applied to the molecular detection of microbes associated with MIAC 

The gold standard criteria for establishing a causal role for microbes are the postulates of Robert Koch. Yet Koch himself realized after proposing the postulates that they could not be fulfilled for pathogens that were uncultivable by then-current methods (e.g. Vibrio cholerae, and Mycobacterium leprae).63 Because this intractable problem of uncultivable pathogens persists, an alternative causal framework based on epidemiologic guidelines suggested by Sir Austin Bradford Hill64 has been proposed for molecular investigations.63 Romero et al. applied this framework in an informative review of infection and preterm birth3; however, it was written before many of the PCR-based studies in the present review were conducted.

Based on Hill's criteria, the body of evidence as reviewed above provides support for a causal role for diverse microbes detected in the amniotic cavity by molecular methods. This evidence includes the demonstration of statistical significance, across multiple studies, for types of associations that address three of Hill's criteria in particular: (i) strength of association of MIAC with inflammation, preterm birth and adverse neonatal outcomes (Section 5.1); (ii) temporal association of MIAC with a shortened time interval to delivery (Section 5.2); and (iii) biological gradient as evidenced by a dose–response association of microbial abundance in MIAC with clinical outcomes (Section 5.3);. In addition, general support exists for almost all of the remaining six criteria (many of which are not as amenable to quantitative analysis and statistical proof as the three criteria above), including: (iv) consistency of association (similar findings from studies of different populations conducted at different times); (v) plausibility (a potential biological mechanism exists1, 3 to explain the relationship); (vi) analogy (relationship conforms to a previously described relationship3); (vii) experimentation (manipulation of the exposure changes the outcome; in the case of MIAC, experimentation is necessarily limited to animal models such as non-human primates65); and, (viii) coherence (causal association is compatible with present knowledge3 of the natural history of the disease). Hill's ninth criterion, specificity of association (outcome is unique to the exposure), is not applicable to diseases of multifactorial etiology, such as preterm labor and delivery.

Back to Article Outline

7. Pathways and anatomic sources of diverse microbes that invade the amniotic cavity 

Four potential routes of MIAC have been postulated: (i) ascending migration from the vagina and cervix; (ii) hematogenous dissemination followed by transplacental invasion; (iii) retrograde seeding from the peritoneal cavity via the Fallopian tubes; (iv) iatrogenic inoculation at the time of an intrauterine procedure.1

The vast majority of the bacteria and fungi that cause MIAC are thought to originate from the indigenous human microbiota (although notable exceptions exist; e.g. exogenously acquired Listeria monocytogenes). Microbes that inhabit body surfaces and cavities outnumber human cells by a factor of 10 (or more), and are a rich source of potential intra-amniotic pathogens. Because the first two routes of invasion (ascending migration and hematogenous dissemination) are considered most important, the microbial residents of sites that access these routes, in particular, influence the diversity of microbes in amniotic fluid.

7.1. The vagina 

A large and compelling body of evidence implicates the vaginal microbiota as the dominant source of microbes responsible for MIAC.2 Current understanding of what constitutes a ‘healthy’ vaginal microbiota, including how it varies over time or across heterogeneous patient populations (e.g. by race, ethnicity, or pregnancy status) is evolving.

Historically, the prevailing conventional view was that a predominance in the vagina of virtually any species of the genus Lactobacillus, and particularly the species L. acidophilus, indicated a healthy vaginal ecology, presumably owing to lactic acid production. However, subsequent studies revealed that species other than L. acidophilus comprise the numerically dominant lactobacilli in the vagina of most healthy women.66 Ravel et al. profiled the vaginal microbiota of reproductive-age women by means of high-throughput DNA sequencing and found that the communities clustered into five groups: four were dominated by species of Lactobacillus, whereas the fifth (designated type IV) had lower proportions of lactic acid bacteria and higher proportions of strict anaerobes.67 Many taxa comprising the type IV community have been associated with bacterial vaginosis and belong to the same genera as that of taxa recently detected in the amniotic cavity (e.g. Atopobium, Dialister, Finegoldia, Leptotrichia, Peptoniphilus, and Sneathia spp.) (Fig. 2).

Bacterial vaginosis is a pathologic state of altered microbial community composition characterized by a marked decrease in the abundance of Lactobacillus spp. and a commensurate increase in diverse anaerobic and facultative bacteria. Bacterial vaginosis, which fits the paradigm of ‘community as pathogen’, owing to its association with an approximately two-fold increased risk68 of preterm birth, is more common in women of African or Hispanic descent. Interestingly, in the study of apparently healthy women by Ravel et al., black and Hispanic women had higher rates of the type IV community (40.4% and 38.1%, respectively) than did Asian women or white women (19.8% and 10.3%, respectively); the Type IV community was associated with higher Nugent scores and higher pH than were the other communities.67 Functional studies suggest that H2O2 production promotes a healthy microbiota and prevents bacterial vaginosis,66 but that it varies by Lactobacillus species.66, 69 Systematic molecular studies of the vaginal microbiota during pregnancy are lacking; however, Wilks et al. found that the presence of H2O2-producing lactobacilli in the vagina at 20 weeks' gestation was associated with reduced risk of bacterial vaginosis and chorioamnionitis.70 Precisely how the diversity of microbes in amniotic fluid is impacted by functional aspects of the vaginal microbiota is of potentially great importance and deserves further study.

7.2. The oropharynx 

Growing evidence suggests that the microbiota of the oral cavity is a source of diverse microbes causing MIAC.24, 31, 32, 37, 39 The primary route hypothesized is hematogenous dissemination, particularly after seeding of the bloodstream in association with gingivitis or periodontitis37; however, colonization of the vaginal tract with microbes from the oral cavity during receptive oral sex has also been proposed.38

Particular species belonging to the genera Fusobacterium or Streptococcus are most frequently invoked to implicate the oral cavity as a source of microbes causing MIAC,37 but additional support comes from recent findings involving Bergeyella spp.39 and P. gingivalis.24 Rothia dentocariosa and Filfactor alocis are other species typically associated with the oral cavity (the latter often in the setting of endodontic infections) that have been detected by PCR of amniotic fluid in the setting of preterm PROM.31 Further studies are needed to better define the role of various oral microbes in contributing to MIAC and preterm birth.

7.3. The gut 

The gastrointestinal tract has been proposed only recently as a possible source of microbial invaders of the amniotic cavity. This stemmed from findings of a study of preterm PROM, in which certain gut-associated taxa, such as Faecalibacterium sp., Coprobacillus sp., and other uncultivated bacteria (e.g. two distinct bacteria belonging to the family Clostridiaceae) were detected in amniotic fluid.31 Ascribing provenance to an individual microbial species based on its taxonomic classification has pitfalls. Yet, as systematic studies of human indigenous microbial communities amass, inferences regarding potential sources of particular species become more compelling and may prove useful for highlighting taxa of interest for future investigations.

Two potential routes of infection from bacteria originating in the gut exist. The first is hematogenous spread from microbes that have translocated from the gut lumen into the bloodstream. A precedent exists for this portal of entry in causing other infections (e.g. Streptococcus bovis endocarditis). The second potential route for gut-associated microbes would be colonization of the vagina followed by ascending migration, as is the case with group B streptococcus.

7.4. The skin and other sites 

The potential for microbes that inhabit the skin or other sites (e.g. the nares) to cause MIAC has received little attention. To date, direct microbiological evidence is lacking. However, the skin plays a vital role in protection from pathogen invasion and harbors a rich diversity of microbes.71 Because the microbial community composition of the skin varies at specific body sites based on topographically distinct physiological characteristics, studies that characterize the skin microbiota at relatively high spatial resolution may be required to illuminate this potential area of inquiry.

Back to Article Outline

8. Differential microbial pathogenicity and the role of microbial and host factors 

A detailed discussion of pathogenic mechanisms is beyond the scope of this review. However, the pathogenesis of MIAC is likely impacted by: (i) host factors (e.g. maternal and fetal immune responses, cervical competency, barrier properties of the chorioamnion, antimicrobial properties of amniotic fluid); (ii) microbial factors (e.g. adhesion, biofilm formation, immune system evasion); and, (iii) combined factors (e.g. protective vaginal colonization). The relative contribution of specific factors may differ with respect to individual microbial species, and thus impact the diversity of microbes in amniotic fluid. Some factors have been relatively well-studied for some microbes, for example, cytadherence of Ureaplasma spp. and Mycoplasma spp.72 Other factors have been less-well studied, such as host immune response heterogeneity with respect to diverse microbial taxa,73 and the role of biofilms in amniotic fluid ‘sludge’.74 For most taxa that invade the amniotic cavity (Fig. 2), the degree to which many of these factors vary, particularly at the species and strain levels, is poorly understood. However, it is possible that differential pathogenicity of microbial species based on the above factors impacts the diversity of microbes that invade the amniotic cavity, as well as patient response to treatment.

Antibiotic treatment trials of various syndromes (e.g. bacterial vaginosis75 and preterm labor with intact membranes76) have yielded divergent results with respect to preventing preterm birth. Proposed reasons for antibiotic failure in clinical trials include inadequate antibiotic regimens, suboptimal selection of the highest-risk patients, delayed timing of antibiotic administration with respect to infection progression (e.g. chorioamnionitis), and patient population heterogeneity with respect to race and genotype, among others.77, 78, 79 An improved understanding of the diverse microbes that invade the amniotic cavity may help to address some of these, and to inform future treatment trials.

Back to Article Outline

9. Is current knowledge of the microbial diversity of the amniotic cavity complete? 

The microbial census of the amniotic cavity, as catalogued by broad-range PCR methods, likely remains unfinished. This shortcoming is evident from two observations. First, some microbial taxa detected in culture-based studies have yet to be detected in PCR-based studies. Second, it is likely that many microbial causes of MIAC have yet to be detected by any method. It has been estimated that 20–80% of human-associated bacterial species (depending on habitat) have yet to be isolated in pure culture.33 With respect to the amniotic cavity, the fact that recent PCR-based studies detected many taxa for the first time in amniotic fluid, with most detected as singletons (Fig. 2), suggests that further PCR-based investigations of MIAC would likely uncover additional diverse, cultivation-resistant species.

Back to Article Outline

10. Conclusions 

The prevalence and diversity of microbes causing MIAC is significantly greater than that indicated by culture methods, and includes as-yet uncharacterized and uncultivated taxa. The role of diverse bacteria belonging to the family Fusobacteriaceae (particularly Sneathia sanguinegens, Leptotrichia spp., and other phylogenetically related species) appears to be especially important and under-recognized. The clinical significance of a positive PCR of amniotic fluid appears to be equivalent to that of a positive culture. Despite recent findings, the microbial census of the amniotic cavity is incomplete. An improved understanding of the diversity of microbes causing MIAC may propel forward important research regarding their sources, pathogenic mechanisms, pathogen synergy, and clinical consequences; these types of potential insights will be important for developing improved prevention, diagnostic and treatment strategies.

Practice points

 


•Routine culture methods significantly underdiagnose MIAC.

•The clinical significance of a positive PCR of amniotic fluid appears to be equivalent to that of a positive culture.

•The retention of an IUD during pregnancy is associated with MIAC caused by Candida spp.

•Optimal prevention and treatment of MIAC caused by diverse microbial species is unknown, and may vary for different species; further research is required

Research directions

 


•Temporal dynamics of the pregnancy-related host microbiome at multiple body sites.

•Mechanisms of amniotic cavity invasion and pathogenesis, including heterogeneous mechanisms of specific microbes.

•Characterization of the maternal and fetal immune response to diverse species by means of genomic, transcriptomic and proteomic approaches.

•Potential role of viruses, including impact of bacteriophage on the pregnancy-related host microbiome.

•Severity of clinical outcomes as a function of particular microbial species, and co-occurrence of species.

•Prevention and treatment trials targeted to specific microbial groups at specific sites (e.g. vagina, amniotic cavity), based on future findings from above areas of inquiry

Back to Article Outline

Conflict of interest statement 

None declared.

Back to Article Outline

Funding sources 

March of Dimes Foundation.

Back to Article Outline

References 

  1. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med. 2000;342:1500–1507
  2. Goncalves LF, Chaiworapongsa T, Romero R. Intrauterine infection and prematurity. Ment Retard Dev Disabil Res Rev. 2002;8:3–13
  3. Romero R, Gomez R, Chaiworapongsa T, Conoscenti G, Kim JC, Kim YM. The role of infection in preterm labor and delivery. Paediatr Perinat Epidemiol. 2001;15(Suppl. 2):41–56
  4. Lawn JE, Cousens S, Zupan J. 4 million neonatal deaths: when? where? why?. Lancet. 2005;365:891–900
  5. Harris JW, Brown H. Bacterial content of the uterus at cesarean section. Am J Obstet Gynecol. 1927;13:133
  6. Kustner O. Beitrag zur Lehre von der puerperalen Infection der Neugeborenen. Arch Gynak. 1877;11:256
  7. Stroup PE. Amniotic fluid infection and the intact fetal membrane. Obstet Gynecol. 1962;19:736–739
  8. Walsh H, Hildebrandt RJ, Prystowsky H. A microbiologic survey at cesarean section. Obstet Gynecol. 1965;26:665–668
  9. Benirschke K. Routes and types of infection in the fetus and the newborn. Am J Dis Child. 1960;99:714–721
  10. Benirschke K, Clifford SH. Intrauterine bacterial infection of the newborn infant: frozen sections of the cord as an aid to early detection. J Pediatr. 1959;54:11–18
  11. Hepner R, Stephens M. Congenital sepsis. Obstet Gynecol. 1955;6:547–549
  12. Harwick HJ, Iuppa JB, Fekety FR. Microorganisms and amniotic fluid. Obstet Gynecol. 1969;33:256–259
  13. Bobitt JR, Ledger WJ. Unrecognized amnionitis and prematurity: a preliminary report. J Reprod Med. 1977;19:8–12
  14. Cassell GH, Davis RO, Waites KB, et al. Isolation of Mycoplasma hominis and Ureaplasma urealyticum from amniotic fluid at 16–20 weeks of gestation: potential effect on outcome of pregnancy. Sex Transm Dis. 1983;10:294–302
  15. Romero R, Quintero R, Oyarzun E, et al. Intraamniotic infection and the onset of labor in preterm premature rupture of the membranes. Am J Obstet Gynecol. 1988;159:661–666
  16. Blanchard A, Hamrick W, Duffy L, Baldus K, Cassell GH. Use of the polymerase chain reaction for detection of Mycoplasma fermentans and Mycoplasma genitalium in the urogenital tract and amniotic fluid. Clin Infect Dis. 1993;17(Suppl. 1):S272–S279
  17. Blanchard A, Hentschel J, Duffy L, Baldus K, Cassell GH. Detection of Ureaplasma urealyticum by polymerase chain reaction in the urogenital tract of adults, in amniotic fluid, and in the respiratory tract of newborns. Clin Infect Dis. 1993;17(Suppl. 1):S148–S153
  18. Yoon BH, Romero R, Kim M, et al. Clinical implications of detection of Ureaplasma urealyticum in the amniotic cavity with the polymerase chain reaction. Am J Obstet Gynecol. 2000;183:1130–1137
  19. Yoon BH, Romero R, Lim JH, et al. The clinical significance of detecting Ureaplasma urealyticum by the polymerase chain reaction in the amniotic fluid of patients with preterm labor. Am J Obstet Gynecol. 2003;189:919–924
  20. Gerber S, Vial Y, Hohlfeld P, Witkin SS. Detection of Ureaplasma urealyticum in second-trimester amniotic fluid by polymerase chain reaction correlates with subsequent preterm labor and delivery. J Infect Dis. 2003;187:518–521
  21. Perni SC, Vardhana S, Korneeva I, et al. Mycoplasma hominis and Ureaplasma urealyticum in midtrimester amniotic fluid: association with amniotic fluid cytokine levels and pregnancy outcome. Am J Obstet Gynecol. 2004;191:1382–1386
  22. Bujold E, Morency AM, Rallu F, et al. Bacteriology of amniotic fluid in women with suspected cervical insufficiency. J Obstet Gynaecol Can. 2008;30:882–887
  23. Oyarzun E, Yamamoto M, Kato S, Gomez R, Lizama L, Moenne A. Specific detection of 16 micro-organisms in amniotic fluid by polymerase chain reaction and its correlation with preterm delivery occurrence. Am J Obstet Gynecol. 1998;179:1115–1119
  24. Leon R, Silva N, Ovalle A, et al. Detection of Porphyromonas gingivalis in the amniotic fluid in pregnant women with a diagnosis of threatened premature labor. J Periodontol. 2007;78:1249–1255
  25. Hitti J, Riley DE, Krohn MA, et al. Broad-spectrum bacterial rDNA polymerase chain reaction assay for detecting amniotic fluid infection among women in premature labor. Clin Infect Dis. 1997;24:1228–1232
  26. Markenson GR, Martin RK, Tillotson-Criss M, Foley KS, Stewart RS, Yancey M. The use of the polymerase chain reaction to detect bacteria in amniotic fluid in pregnancies complicated by preterm labor. Am J Obstet Gynecol. 1997;177:1471–1477
  27. DiGiulio DB, Romero R, Amogan HP, et al. Microbial prevalence, diversity and abundance in amniotic fluid during preterm labor: a molecular and culture-based investigation. PLoS One. 2008;3:e3056
  28. Han YW, Shen T, Chung P, Buhimschi IA, Buhimschi CS. Uncultivated bacteria as etiologic agents of intra-amniotic inflammation leading to preterm birth. J Clin Microbiol. 2009;47:38–47
  29. Marconi C, de Andrade Ramos BR, Peracoli JC, Donders GG, da Silva MG. Amniotic fluid interleukin-1 beta and interleukin-6, but not interleukin-8 correlate with microbial invasion of the amniotic cavity in preterm labor. Am J Reprod Immunol. 2011;65:549–556
  30. Jalava J, Mantymaa ML, Ekblad U, et al. Bacterial 16S rDNA polymerase chain reaction in the detection of intra-amniotic infection. Br J Obstet Gynaecol. 1996;103:664–669
  31. DiGiulio DB, Romero R, Kusanovic JP, et al. Prevalence and diversity of microbes in the amniotic fluid, the fetal inflammatory response, and pregnancy outcome in women with preterm pre-labor rupture of membranes. Am J Reprod Immunol. 2010;64:38–57
  32. Gardella C, Riley DE, Hitti J, Agnew K, Krieger JN, Eschenbach D. Identification and sequencing of bacterial rDNAs in culture-negative amniotic fluid from women in premature labor. Am J Perinatol. 2004;21:319–323
  33. Dethlefsen L, McFall-Ngai M, Relman DA. An ecological and evolutionary perspective on human–microbe mutualism and disease. Nature. 2007;449:811–818
  34. Eribe ER, Olsen I. Leptotrichia species in human infections. Anaerobe. 2008;14:131–137
  35. DiGiulio DB, Gervasi M, Romero R, et al. Microbial invasion of the amniotic cavity in preeclampsia as assessed by cultivation and sequence-based methods. J Perinat Med. 2010;38:503–513
  36. DiGiulio DB, Gervasi MT, Romero R, et al. Microbial invasion of the amniotic cavity in pregnancies with small-for-gestational-age fetuses. J Perinat Med. 2010;38:495–502
  37. Bearfield C, Davenport ES, Sivapathasundaram V, Allaker RP. Possible association between amniotic fluid micro-organism infection and microflora in the mouth. BJOG. 2002;109:527–533
  38. Alanen A, Laurikainen E. Second-trimester abortion caused by Capnocytophaga sputigena: case report. Am J Perinatol. 1999;16:181–183
  39. Han YW, Ikegami A, Bissada NF, Herbst M, Redline RW, Ashmead GG. Transmission of an uncultivated Bergeyella strain from the oral cavity to amniotic fluid in a case of preterm birth. J Clin Microbiol. 2006;44:1475–1483
  40. Miralles R, Hodge R, McParland PC, et al. Relationship between antenatal inflammation and antenatal infection identified by detection of microbial genes by polymerase chain reaction. Pediatr Res. 2005;57:570–577
  41. Rallu F, Morency AM, Laferriere C, Bujold E. Invasion of the amniotic cavity by an uncultured bacterium, a gram-positive coccus. J Matern Fetal Neonatal Med. 2007;20:185–187
  42. Bruner JP, Elliott JP, Kilbride HW, Garite TJ, Knox GE. Candida chorioamnionitis diagnosed by amniocentesis with subsequent fetal infection. Am J Perinatol. 1986;3:213–218
  43. Kellogg SG, Davis C, Benirschke K. Candida parapsilosis: previously unknown cause of fetal infection. A report of two cases. J Reprod Med. 1974;12:159–162
  44. Chaim W, Mazor M, Wiznitzer A. The prevalence and clinical significance of intraamniotic infection with Candida species in women with preterm labor. Arch Gynecol Obstet. 1992;251:9–15
  45. Parveen N, Munir AA, Din I, Majeed R. Frequency of vaginal candidiasis in pregnant women attending routine antenatal clinic. J Coll Physicians Surg Pak. 2008;18:154–157
  46. Misenhimer HR, Garcia-Bunuel R. Failure of intrauterine contraceptive device and fungal infection in the fetus. Obstet Gynecol. 1969;34:368–372
  47. Romero R, Reece EA, Duff GW, Coultrip L, Hobbins JC. Prenatal diagnosis of Candida albicans chorioamnionitis. Am J Perinatol. 1985;2:121–122
  48. Whyte RK, Hussain Z, deSa D. Antenatal infections with Candida species. Arch Dis Child. 1982;57:528–535
  49. Kim SK, Romero R, Kusanovic JP, et al. The prognosis of pregnancy conceived despite the presence of an intrauterine device (IUD). J Perinat Med. 2009;38:45–53
  50. Romero R, Gonzalez R, Sepulveda W, et al. Infection and labor. VIII. Microbial invasion of the amniotic cavity in patients with suspected cervical incompetence: prevalence and clinical significance. Am J Obstet Gynecol. 1992;167:1086–1091
  51. Auler ME, Morreira D, Rodrigues FF, et al. Biofilm formation on intrauterine devices in patients with recurrent vulvovaginal candidiasis. Med Mycol. 2010;48:211–216
  52. Woese CR, Kandler O, Wheelis ML. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc Natl Acad Sci USA. 1990;87:4576–4579
  53. Wenstrom KD, Andrews WW, Bowles NE, Towbin JA, Hauth JC, Goldenberg RL. Intrauterine viral infection at the time of second trimester genetic amniocentesis. Obstet Gynecol. 1998;92:420–424
  54. Baschat AA, Towbin J, Bowles NE, Harman CR, Weiner CP. Prevalence of viral DNA in amniotic fluid of low-risk pregnancies in the second trimester. J Matern Fetal Neonatal Med. 2003;13:381–384
  55. Miller JL, Harman C, Weiner C, Baschat AA. Perinatal outcomes after second trimester detection of amniotic fluid viral genome in asymptomatic patients. J Perinat Med. 2009;37:140–143
  56. Kim MJ, Romero R, Gervasi MT, et al. Widespread microbial invasion of the chorioamniotic membranes is a consequence and not a cause of intra-amniotic infection. Lab Invest. 2009;89:924–936
  57. Papantoniou NE, Antsaklis AJ, Protopapas AG, Vogiatzi AI, Aravantinos DI. Predictive value of amniotic fluid and fetal blood cultures in pregnancy outcome in preterm prelabor rupture of membranes. J Obstet Gynaecol. 1997;17:18–22
  58. Virreira M, Martinez S, Alonso-Vega C, et al. Amniotic fluid is not useful for diagnosis of congenital Trypanosoma cruzi infection. Am J Trop Med Hyg. 2006;75:1082–1084
  59. Sibley CD, Duan K, Fischer C, et al. Discerning the complexity of community interactions using a Drosophila model of polymicrobial infections. PLoS Pathog. 2008;4:e1000184
  60. Romero R, Mazor M. Infection and preterm labor. Clin Obstet Gynecol. 1988;31:553–584
  61. Watts DH, Krohn MA, Hillier SL, Eschenbach DA. The association of occult amniotic fluid infection with gestational age and neonatal outcome among women in preterm labor. Obstet Gynecol. 1992;79:351–357
  62. Bucher HU, Ochsner Y, Fauchere JC. Two years outcome of very pre-term and very low birthweight infants in Switzerland. Swiss Med Wkly. 2003;133:93–99
  63. Fredericks DN, Relman DA. Sequence-based identification of microbial pathogens: a reconsideration of Koch's postulates. Clin Microbiol Rev. 1996;9:18–33
  64. Hill AB. The environment and disease: association or causation?. Proc R Soc Med. 1965;58:295–300
  65. Gravett MG, Adams KM, Sadowsky DW, et al. Immunomodulators plus antibiotics delay preterm delivery after experimental intraamniotic infection in a nonhuman primate model. Am J Obstet Gynecol. 2007;197:518.e1–518.e8
  66. Antonio MA, Hawes SE, Hillier SL. The identification of vaginal Lactobacillus species and the demographic and microbiologic characteristics of women colonized by these species. J Infect Dis. 1999;180:1950–1956
  67. Ravel J, Gajer P, Abdo Z, et al. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci USA. 2010;108(Suppl. 1):4680–4687
  68. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med. 1995;333:1737–1742
  69. Vallor AC, Antonio MA, Hawes SE, Hillier SL. Factors associated with acquisition of, or persistent colonization by, vaginal lactobacilli: role of hydrogen peroxide production. J Infect Dis. 2001;184:1431–1436
  70. Wilks M, Wiggins R, Whiley A, et al. Identification and H(2)O(2) production of vaginal lactobacilli from pregnant women at high risk of preterm birth and relation with outcome. J Clin Microbiol. 2004;42:713–717
  71. Grice EA, Kong HH, Conlan S, et al. Topographical and temporal diversity of the human skin microbiome. Science. 2009;324:1190–1192
  72. Waites KB, Katz B, Schelonka RL. Mycoplasmas and ureaplasmas as neonatal pathogens. Clin Microbiol Rev. 2005;18:757–789
  73. Menon R, Peltier MR, Eckardt J, Fortunato SJ. Diversity in cytokine response to bacteria associated with preterm birth by fetal membranes. Am J Obstet Gynecol. 2009;201:306.e1–306.e6
  74. Romero R, Schaudinn C, Kusanovic JP, et al. Detection of a microbial biofilm in intraamniotic infection. Am J Obstet Gynecol. 2008;198:135.e1–135.e5
  75. McDonald H, Brocklehurst P, Parsons J, Vigneswaran R. Antibiotics for treating bacterial vaginosis in pregnancy. Cochrane Database Syst Rev. 2003;CD000262
  76. Kenyon SL, Taylor DJ, Tarnow-Mordi W. Broad-spectrum antibiotics for spontaneous preterm labor: the ORACLE II randomised trial. ORACLE Collaborative Group. Lancet. 2001;357:989–994
  77. Iams JD, Romero R, Culhane JF, Goldenberg RL. Primary, secondary, and tertiary interventions to reduce the morbidity and mortality of preterm birth. Lancet. 2008;371:164–175
  78. Lamont RF. Can antibiotics prevent preterm birth – the pro and con debate. BJOG. 2005;112(Suppl. 1):67–73
  79. Lamont RF, Nhan-Chang CL, Sobel JD, Workowski K, Conde-Agudelo A, Romero R. Treatment of abnormal vaginal flora in early pregnancy with clindamycin for the prevention of spontaneous preterm birth: a systematic review and metaanalysis. Am J Obstet Gynecol. 2011;205:177–190

PII: S1744-165X(11)00108-9

doi:10.1016/j.siny.2011.10.001

Seminars in Fetal & Neonatal Medicine
Volume 17, Issue 1 , Pages 2-11, February 2012

Article Tools

* Image Usage & Resolution