Cervicovaginal Microbiota Composition in Chlamydia trachomatis Infection: A Systematic Review and Meta-Analysis

In healthy women, the cervicovaginal microbiota is characterized by the predominance of Lactobacillus spp., whereas the overgrowth of anaerobic bacteria leads to dysbiosis, known to increase the risk of acquiring genital infections like Chlamydia trachomatis. In the last decade, a growing body of research has investigated the composition of the cervicovaginal microbiota associated with chlamydial infection via 16s rDNA sequencing, with contrasting results. A systematic review and a meta-analysis, performed on the alpha-diversity indices, were conducted to summarize the scientific evidence on the cervicovaginal microbiota composition in C. trachomatis infection. Databases PubMed, Scopus and Web of Science were searched with the following strategy: “Chlamydia trachomatis” AND “micro*”. The diversity indices considered for the meta-analysis were Operational Taxonomic Unit (OTU) number, Chao1, phylogenetic diversity whole tree, Shannon’s, Pielou’s and Simpson’s diversity indexes. The search yielded 425 abstracts for initial review, of which 16 met the inclusion criteria. The results suggested that the cervicovaginal microbiota in C. trachomatis-positive women was characterized by Lactobacillus iners dominance, or by a diverse mix of facultative or strict anaerobes. The meta-analysis, instead, did not show any difference in the microbial biodiversity between Chlamydia-positive and healthy women. Additional research is clearly required to deepen our knowledge on the interplay between the resident microflora and C. trachomatis in the genital microenvironment.


Introduction
The cervicovaginal micro-environment harbours a variety of resident microorganisms, named microbiota, whose interplay with the host is involved in either health or disease [1]. The cervicovaginal microbiota is the simplest microbial community in the human organism since, in healthy reproductive-age women, it is characterized by the predominance of one or a few bacterial species, mainly Lactobacillus spp., alongside other bacterial species present in much lower amounts, including Staphylococcus spp., Streptococcus spp., Bifidobacterium spp., Ureaplasma spp., Mycoplasma spp., Veillonella spp., etc. [1,2] Indeed, Lactobacillus spp. is generally regarded as the main host-defence factor within the cervicovaginal ecosystem against potential pathogens via several mechanisms, including competitive exclusion, anti-microbial compound production (e.g., lactic acid), immune system activation as well as the maintenance of a low vaginal pH [3][4][5]. In certain conditions, the depletion of Lactobacilli and the overgrowth of anaerobic bacteria, like Gardnerella spp., Prevotella spp., Atopobium spp., etc., has often been associated with cervicovaginal dysbiosis, known to increase the risk of acquiring genital infections like Chlamydia trachomatis [6,7].

Methods
This systematic review and metanalysis has been performed in accordance with the latest version of the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines [30]. The Review protocol was registered on the International prospective register of systematic reviews (PROSPERO, reference number CRD42022341268). Zotero citation management software (RRID:SCR_013784) was used to identify any duplicates and to manage and screen the selected literature records.

Literature Research
The Review included articles published up to 30 June 2022, on the databases PubMed, Scopus, and Web of Science, with the following search strategy and Boolean operator: "Chlamydia trachomatis" AND "micro*". Truncation filters (*) were used to represent any combination of letters. Three independent reviewers (MDP, SF, and IS) performed the search, reading the titles and abstracts of the articles identified by the search strategy. During the multi-step exclusion process, any disagreement on the studies was discussed until reaching a consensus. The process was supervised by other investigators (PP and RS). Figure 1 shows the PRISMA flow chart diagram summarizing the selection steps for the present Systematic Review.
During the multi-step exclusion process, any disagreement on the studies was discussed until reaching a consensus. The process was supervised by other investigators (PP and RS). Figure 1 shows the PRISMA flow chart diagram summarizing the selection steps for the present Systematic Review.

Inclusion and Exclusion Criteria
The review included only studies investigating the composition of the cervicovaginal microbiota via 16s rRNA gene sequencing, comparing reproductive-age women positive to C. trachomatis infections to age-matched healthy controls. We did not consider studies analysing the cervicovaginal microbiota in women positive to other genital pathogens,

Inclusion and Exclusion Criteria
The review included only studies investigating the composition of the cervicovaginal microbiota via 16s rRNA gene sequencing, comparing reproductive-age women positive to C. trachomatis infections to age-matched healthy controls. We did not consider studies analysing the cervicovaginal microbiota in women positive to other genital pathogens, including Neisseria gonorrhoeae, Trichomonas vaginalis, Mycoplasma spp., Candida spp., HPV, and herpes simplex virus 2 (HSV-2).
Only articles presenting controlled trials published in peer-reviewed journals were considered eligible. Reviews, meta-analysis, editorials, commentaries, case reports, case series, semi-experimental and experimental studies, proceedings, individual contributions (e.g., conference speeches), and purely descriptive studies published in scientific conferences without any quantitative or qualitative findings, were excluded from the review. Finally, articles published in languages other than English were also excluded. The search was performed until 30 June 2022 on the databases.

Data Extraction Process and Quality Assessment
From each study included in the review, the following data were extracted: bibliographic information, study design, study population size, age, and ethnicity, 16s rRNA gene primers and sequencing platform used, and main conclusions.
Three different reviewers (MDP, SF, and IS) assessed the methodological quality of the selected studies with the Newcastle-Ottawa Scale (NOS) rating tool, adapted for evaluating case-control, cross-sectional and cohort studies [31]. The NOS is divided in eight categories evaluating three different quality aspects: selection, comparability, and outcome; scores range from 0 to 9 and the quality of a study was considered to be high if the NOS score was 7 to 9, intermediate if the NOS score was 4 to 6, and low if it was 0 to 3 [32].

Data Items
The outcomes of interest were the OTU, Chao1, Phylogenetic Diversity whole tree, Shannon's index, Pielou's evenness index, and Simpson's index mean levels in C. trachomatis positive (POS) as compared to C. trachomatis negative (CTRL) women.

Effect Measure
For each outcome, an effect measure was calculated as the Mean Difference (MD) between the mean values observed in the POS and CTRL groups.

Synthesis Methods
In each synthesis, only the studies that reported the necessary data for MD calculation (mean and SD in each group of interest) were included.
Each parameter of interest showed a high variability; therefore, before the synthesis, difference in means between POS and CTRL groups and the corresponding standard error (SE), calculated on the raw scale, were converted to an approximate difference and standard error on the logarithmic scale following method 3 proposed by Higgins et al. (2008) [33]. To obtain the pooled MD, it was decided to apply the random effects model. MD was reported with the corresponding 95% Confidence Interval (95% CI).
The results were represented by the forest plot that is a graphical representation of the estimated results from each study included in the analysis along with the pooled result. The heterogeneity between the studies was assessed through the visual inspection of the forest plot, by the overlap of the CIs and possible outliers, and quantified through the I 2 index. The I 2 statistic describes the percentage of variation across studies that is due to heterogeneity rather than chance. We considered that an I 2 of 50% indicates moderate heterogeneity and 75% or greater indicates substantial heterogeneity. The p value of the chi-squared (χ 2 ) test, a statistical test for heterogeneity, was included in the forest plots.
Publication bias was assessed with visual inspection of the funnel plot, that is a scatter plot of the effect estimates from individual studies against standard error (SE). Egger's test was not performed because in each analysis the number of studies was less than 10. We considered a p value < 0.05 to be statistically significant. The data was entered and analysed with the statistical software STATA v16.

Study Selection Process
In total, we recovered 425 studies from all searched databases (n = 162 from Scopus, n = 165 from Web of Science, and n = 98 from PubMed), and after applying filters by automation tools, 317 articles remained. Out of the remaining 317 papers, 183 were excluded after removing duplicates, and 134 were subjected to further screening and evaluated for inclusion in the systematic review after considering inclusion and exclusion criteria. A total of 117 papers was then excluded because they did not fit inclusion criteria. One more article was not considered since it did not included controls. At the end of the process, 16 articles were included in the systematic review. Due to difficulties in extrapolating alpha-diversity data, only 7 of them were considered for the meta-analysis.
From the included studies has also emerged the need to restore a protective vaginal microbiota after antibiotic treatment for C. trachomatis, via, for example, the use of probiotics [15,20,22]. In this regard, besides comparing the resident microflora of the genital ecosystem between C. trachomatis-positive and healthy women at baseline, some authors have also investigated the effects of the standard treatment with azithromycin. In particular, Ziklo et al. (2018) have observed an increased prevalence of indole-producing bacteria in the vaginal microbiota post-treatment, suggesting a destructive effect of azithromycin on vaginal health [15]. On a similar note, the study from Tamarelle et al., 2020, has described a robust increase in L. iners' abundance, suggesting that after antibiotic treatment, the individual risk for C. trachomatis remains high, as evidenced by the increased rate of reinfections observed in their study [22]. CT women showed a marked increase in alpha-diversity indices (Shannon's and Shannon-weaver's) and an overall decrease in Lactobacillus spp., alongside an increase in anaerobic bacterial species, including G. vaginalis, A. vaginae, P. amnii, P. timonensis, and L. amnionii.
van der Veer et al. [14] 2013-2014 Women notified for C. trachomatis The endocervical microbiota diversity is not grossly altered in CT+ women, although CT+ women had higher relative abundance of G. vaginalis and other anaerobes, like Megasphaera spp., A. vaginae, Dialister spp., and Prevotella spp., all BV-associated bacteria, as compared to CT-women.
Ziklo et al. [15] Not reported Women with diagnosed C. CSTs were not significantly associated with C. trachomatis status but higher proportions of CT+ women were found in CST-III (L. iners) and CST-IV (anaerobic bacteria), rather than in CST-I (L. crispatus).
van Houdt et al. [19] 2008-2012 Women screened for C. trachomatis who tested negative at the first visit (n = 115). At subsequent yearly screening, n = 60 women tested CT-positive, and n = 55 women tested CT-negative (controls).

European Vaginal swabs V3-4 Illumina
Five CSTs were identified, four CSTs were dominated by Lactobacillus spp., of which L. crispatus (CST-I) and L. iners (CST-III) were the most common, and one CST was characterized by an array of strict and facultative anaerobes (CST-IV). Women with L. iners dominated CST-III had increased risk of CT infection.

European Endocervical swabs V3-4 Illumina
Alpha diversity indices (Shannon's and Shannon-weaver's) were higher in either CT or HPV/CT co-infected women as compared to healthy controls. The cervical microbiota of CT positive and HPV/CT co-infected women was characterized by decreased Lactobacillus spp. and increased anaerobic bacterial species, like G. vaginalis, A. vaginae and A. christensenii. L. iners were also more frequently found in CT positive and HPV/CT co-infected women. By contrast, HPV positive women showed a similar microbiota to those in healthy controls.

European
Endocervical swabs V4 Illumina Alpha-diversity indices (Shannon's and Shannon-Weaver's) were significantly higher in CT women as compared to HC. CT microbiota was dominated by anaerobes (CTS-IV), and a specific network of G. vaginalis, P. amnii, P. buccalis, P. timonensis, A. christensenii and V. guangxiensis was identified as potential biomarker of CT infection. CT was also significantly correlated with increased levels of lactoferrin, IL-6, IL-1α, IFN-α, and IFN-β, whereas very low levels of IFN-γ were observed.

African-American Vaginal swabs V3-4 Illumina
Women with CT infection or coinfection CT/MG were associated with a CST-IV microbiota, characterized by decreased Lactobacillus spp. Significant differences in vaginal metabolites were identified in CT+ or CT+/MG+ women as compared to uninfected women, before and after adjustment for CSTs, with significant overlap between CT+ and CT+/MG+ women.
Tamarelle et al. [20] Not reported Women with confirmed C. trachomatis infection at baseline and after azithromycin treatment at 3, 6 and 9 months (n = 149), and CT negative controls (n = 99).

African-American Vaginal swabs V3-4 Illumina
CT women microbiota was dominated, at the time of diagnosis, by L. iners or a diverse array of BV or CST-IV associated bacteria, such as G. vaginalis, A. vaginae and M. curtisii. L. iners-dominated communities were most common after azithromycin treatment (1 g monodose), consistent to the observed relative resistance of this bacterium to azithromycin. Chen et al. [22] 2019-2020 Women with tubal infertility and C. trachomatis infection before (CT-P, n = 6) and after treatment (CT-PT, n = 4), as compared to infertile (CT-N, n = 8) or healthy women (CT-C, n = 7) without chlamydial infection (controls)

Asian Vaginal swabs V3-4 Illumina
Women with tubal infertility and CT presented a L. iners dominated microbiota with a decrease in Lactobacillus, Bifidobacterium, Enterobacter, Atopobium, and Streptococcus spp., which could be restored with varying degrees by azythromycin treatment. C. trachomatis-positive women also had increased levels of IFNγ and IL-10.

European
Vaginal and anal swabs V3-4 Illumina Alpha-diversity via Pielou's index was higher in the vaginal microbiota of CT-positive women than CT-negative women. In CT-positive women, the vaginal microbiota was depleted of Lactobacillus spp., with a significant increase in anaerobes, like Sneathia spp., Parvimonas spp., and Megasphaera spp. CT positively correlated with Ezakiella spp. The predicted metabolic functions showed increased chorismate and aromatic amino-acid biosynthesis, as well as mixed acid fermentation, in the vaginal microbiota of CT-positive women.

Asian Vaginal swabs V3-4 Illumina
In CT women, the vaginal microbiota was dominated by L. iners, with increased relative abundance of G. vaginalis as compared to TV and non-STI women. In TV women, Lactobacillus spp. was significantly lower, and S. agalactiae, P. bivia, S. sanguinegens and G. asaccharolytica were significantly enriched, as compared to the other patient groups.
Another interesting aspect was the evidence that unique microbial profiles might characterize genital infection with other viral, protozoan, or fungal pathogens, potentially representing distinct markers of infection [16,17,26,28,29]. Indeed, the compositional structure and biodiversity of the cervicovaginal microbiota in C. trachomatis-positive women was also compared to women with different co-infections, or to women with other genital pathogens alone. The main findings showed that mixed infections, with C. trachomatis and HPV, M. genitalium, and N. gonorrhoeae/T. vaginalis, are associated with a CST-IV microbiota, with increased anaerobic bacterial species, or to L. iners-dominated microbiota, and did not present significant differences as compared to women positive to C. trachomatis alone [16,17,29]. By contrast, women with HPV alone showed a cervicovaginal microbiota similar to that observed in healthy controls, whereas women with either T. vaginalis or C. albicans infections presented a microbiota characterized by a lower abundance of Lactobacillus spp. and much higher abundance of anaerobes, such as Prevotella spp., Gardnerella spp., Atopobium spp. and Megasphaera spp., as compared to C. trachomatispositive women [26,28].

Meta-Analysis
Meta-analysis was performed using the following indices: Chao1's diversity index, OTU number, phylogenetic diversity whole tree, Shannon's diversity index, Pielou's evenness diversity index, and Simpson's diversity index.

Chao1's Diversity Index
Four studies were eligible for the synthesis of Chao1 mean levels for a total of 95 subjects, 43 positive and 52 controls. The results of the meta-analysis indicate a nonsignificant difference between positives and controls in Chao1 levels; the pooled MD on a logarithmic scale was equal to −0.06 (95% CI −0.33 to 0.20; p = 0.638) corresponding to an approximate relative reduction in the positives group of about −6%. The heterogeneity between the studies was not significant (I 2 = 51.1%; p = 0.105) (Figure 2A). All the studies presented a NOS score of 8 and only one of 7 [22] ( Table 2). The funnel plot seemed to not show an evidence of publication bias even if there were only 4 studies included in the analysis ( Figure S1A, Table S1).

Scoring Results
The median NOS score of the included studies was 8 (interquartile range, IQR 1.25), indicating a high average quality level. Specifically, most studies (n = 12) were of high to very high quality (score of 7 to 9), 3 studies were of intermediate quality (score of 4 to 6), and 1 study was of low quality (score of 0 to 3). Table 2 shows the results of the scoring

OTU Number
Four studies were included in the meta-analysis of OTU levels. Pooling the studies there was a total of 207 subjects, 72 positive and 135 controls. The results of the metaanalysis showed a non-significant difference between positives and controls in the OTU; the pooled MD on a logarithmic scale was equal to 0.11 (95% CI −0.28 to 0.50; p = 0.582) corresponding to a relative increase in positives of approximately 11% ( Figure 2B). The heterogeneity between the studies was high and significant (I 2 = 86.4%; p < 0.001). In fact, two studies [22,26] reported a negative difference indicating lower values in positive group, while the other two studies reported a positive difference indicating higher values in the positive group [24,25]. All the studies presented a NOS score of 8 and only one of 7 [22] ( Table 2). The funnel plot does not seem to show the presence of publication bias but there were only 4 studies ( Figure S1B, Table S1).

Phylogenetic Diversity Whole Tree
Only two studies reported data about this parameter [24,26] so the analysis was performed considering these two studies. The pooled total number of subjects was of 55, 27 positives and 28 controls. The results of the meta-analysis indicate a non-significant difference between positives and controls; the pooled MD on a logarithmic scale was equal to 0.08 (95% CI −0.12 to 0.28; p = 0.412) corresponding to a relative increase in the positive group of about 8% ( Figure 2C). The heterogeneity between the studies was not significant (I 2 = 26.8%; p = 0.242) ( Figure S1C, Table S1).

Shannon's Diversity Index
Seven studies report data for Shannon's parameter for a total of 345 subjects, comprising 138 positives and 207 controls. The results of the meta-analysis indicated a nonsignificant difference between positives and controls; the pooled MD on a logarithmic scale was equal to 0.21 (95% CI −0.08 to 0.51; p = 0.158) corresponding to a relative increase in the positive group of about 21%. The heterogeneity between the studies was significant (I 2 = 77.1%; p < 0.001) ( Figure 2D). Two studies [24,25] reported a positive and significant difference, while the remaining 5 studies reported non-significant results. The studies included in this meta-analysis were of high quality except for one of intermediate quality [27] ( Table 2). The funnel plot did not seem to show the presence of publication bias ( Figure S1D, Table S1).

Simpson's Diversity Index
Four studies were eligible for the meta-analysis of the Simpson parameter [22,24,27,28]; a total number of 140 subjects were considered, comprising 69 positives and 71 controls. The results of the meta-analysis indicate a non-significant difference between positives and controls. The pooled MD on a logarithmic scale was equal to 0.09 (95% CI −0.40 to 0.58; p = 0.719) corresponding to a relative increase in the positives of about 9% ( Figure 2E). The heterogeneity between the studies was significant (I 2 = 70.4%; p = 0.017). In this metaanalysis, the MD observed in Filardo et al., 2017 [24] study was positive and significant, while in the other three studies the results were not significant. Based on NOS score, all the studies were of high quality with the exception of Cheong et al. (2019) [27] of intermediate quality ( Table 2). The funnel plot did not seem to show evidence of publication bias ( Figure S1E, Table S1).

Pielou's Evenness Diversity Index
Three studies report data about the parameter Pielou [23,25,27] for a total of 241 subjects, comprising 91 positives and 150 controls. The results of the meta-analysis showed a nonsignificant difference between positives and controls. the pooled MD on a logarithmic scale was equal to 0.21 (95% CI −0.15 to 0.58; p = 0.250), corresponding to a relative increase in positives of about 21% ( Figure 2F). The heterogeneity between the studies was significantly high (I 2 = 80.9%; p = 0.005). All the studies included in the meta-analysis were of high to very high quality, with the exception of one study [27] of intermediate quality (Table 2). It was difficult to exclude the presence of publication bias by the funnel plot since there were only 3 studies included in the analysis ( Figure S1F, Table S1).

Scoring Results
The median NOS score of the included studies was 8 (interquartile range, IQR 1.25), indicating a high average quality level. Specifically, most studies (n = 12) were of high to very high quality (score of 7 to 9), 3 studies were of intermediate quality (score of 4 to 6), and 1 study was of low quality (score of 0 to 3). Table 2 shows the results of the scoring method applied to each study included in the review, with reference to publication year and study design.

Conclusions
Overall, based on the findings presented in our review, it emerged that women affected by a C. trachomatis infection possessed an altered cervicovaginal microbiota, with increased abundances of a varied population of strict or facultative anaerobic bacterial species as well as of L. iners.
Concerning the quantitative measures of microbial biodiversity within a bacterial community, namely alpha-diversity indices, a non-significant difference was demonstrated, via a meta-analysis, between C. trachomatis positive and healthy women. Indeed, alpha-diversity indices were scarcely used in the reviewed studies; only about half of them reported alpha-diversity data via a plethora of different indices, including Shannon, Shannon-Weaver's or Pielou's evenness, Chao1, Simpson's, etc., leading to inconsistent results [23][24][25][26][27][28][29]. Specifically, some papers did not find any statistically significant difference in the diversity indices between C. trachomatis-positive women and healthy controls [26][27][28], while others did find increased diversity in the cervicovaginal microbiota of Chlamydiainfected women [23][24][25]29].
There are different limitations that undermine the comparability of results in the studies included in our systematic review and meta-analysis. In particular, there are differences in specimen collection, since either vaginal and/or endocervical swabs have been used; indeed, it is known that significant differences have been demonstrated in the microbiota characterizing the two sites. In addition, different primers and hypervariable regions of 16s rDNA were chosen for sequencing, such as the V3-4, the V4 or the V2-4-8, leading to potentially different results for microbiota composition. In fact, it has been already evidenced that different primers may influence the results of the bioinformatic analysis. At the same time, another relevant issue lies in the statistical analysis since a varied mix of statistical algorithms have been used. In this regard, it is very challenging to compare the biodiversity of the genital microbiota between different studies and patient groups, due to the numerous and diverse statistical measures adopted for the calculation of alpha-diversity indices, as well as to missing or incomplete reporting of data in the papers.
In the future, it will be of great interest to reach a consensus on the several parameters involved in designing metagenomic studies, so that all data are reported and rendered available to the scientific community. Indeed, given the importance of the microbiota not only from a pathophysiological but also clinical point of view, streamlining the presentation of metagenomic data will render their interpretation more accessible to a wider audience. Surely, a standardized procedure will help to significantly push the field forward by providing more complete data for a better understanding of the pathophysiological role of the cervicovaginal microbiota in C. trachomatis infection.