Introduction

Epidemiological and laboratory evidence have shown that human papillomavirus (HPV) infections play an important role in the progression of uterine cervical dysplasia and squamous cell cervical carcinoma (SCC) [1, 2]. However, among all HPV infections, only a small fraction of them would progress to uterine cervical dysplasia (10% of HPV infections) and SCC (< 1% of HPV infections) [1,2,3,4], which suggests that other factors may participate in this variable natural history. Among various risk factors associated with HPV infections and progression to abnormal cervical cytopathology, sexually transmitted infections (STIs) other than HPV, commonly presented as genital mycoplasmas, are gradually becoming more important. Evidence accumulated that genital mycoplasmas infections could cause serious long-term adverse sequelae in women [5,6,7].

Genital mycoplasmas are groups of small free-living pathogenic bacterium belonging to class Mollicutes that lives on the ciliated epithelial cells of the urinary and genital tracts in humans [8], generally transmitted through direct interaction between hosts—venereally through genitogenital or orogenital contact and vertically from mother to child (either in utero or at birth) [9]. Genital mycoplasmas often refer to six species that colonize in the genital tract, involving Ureaplasma urealyticum, Ureaplasma parvum, Mycoplasma hominis, Mycoplasma genitalium, Mycoplasma primatum, and Mycoplasma spermatophilum, the latter two of which are considered non-pathogenic for humans [10]. There is growing evidence suggesting that genital mycoplasmas have now become a group of common etiological pathogen in women and exerted adverse effects on female reproductive health [11]. Specifically, in a Britain national survey, M. genitalium was found in 1.3% (0.9–1.9%) women aged 16–44 years and 2.4% (1.2–4.8%) women aged 16–19 years [12]. In addition, a recent meta-analysis also demonstrated that M. genitalium infection was significantly associated with significantly increased risk of cervicitis, pelvic inflammatory disease, preterm birth, and spontaneous abortion [7]. Current studies indicated that mycoplasma infections may arouse those changes through inducing chromosomal alterations that may lead to transformation of mammalian cells, especially in a chronic pattern [13, 14].

In recent decades, a substantial number of epidemiological studies investigated the association between genital mycoplasmas infection and HPV infections, abnormal cervical cytopathology, and cervical cancer [15,16,17]. However, their results remained inconclusive. Therefore, we aimed to perform the first systemic review and meta-analysis to assess the association between genital mycoplasmas infection and HPV infections, abnormal cervical cytopathology, and cervical cancer.

Methods

Literature search

We searched all published studies regarding the association between the genital mycoplasmas infection and HPV infections, abnormal cervical cytopathology, and cervical cancer using computerized databases (Embase and PubMed) and by manually searching references of included studies. The databases were searched before 26 January 2017. The following strategy was employed: (Uterine Cervical Neoplasms OR cervical cancer OR uterine cervical dysplasia OR cervical precancerous lesions OR abnormal cervical cytology OR abnormal cervical cytopathology OR cervical intraepithelial neoplasia OR squamous intraepithelial lesion OR atypical squamous cells OR human papillomavirus) AND (ureaplasma OR mycoplasma).

Study selection

Two reviewers (HY and TS) independently selected the included studies based on previously established standards. The included studies must meet the following criteria: (1) original peer-reviewed English articles concerning the association between genital mycoplasmas and HPV infections and related disorders; (2) cross section, case–control, or cohort studies; and (3) studies contain adequate methods to assay genital mycoplasmas. Studies meeting the following criteria were excluded: (1) animal studies, genomic studies, and in-vitro experiments; (2) conference, review, guideline, letter, and editorials; and (3) non-sufficient data to assess the associations between genital mycoplasmas and above-mentioned disorders.

Data extraction and quality assessment

Two reviewers (HY and XZ) independently extracted the following items: first author, year of publication, study region, study design, sample size, recruitment period, source of study population, age range, pregnancy or not, genital mycoplasmas assay, numbers of patients presented with genital mycoplasma, or/and HPV infections. Discrepancies were solved by discussing among reviewers to reach consensus. We also performed quality assessment of the included studies, adopting the Newcastle–Ottawa Scale (NOS) for case–control or cohort studies and the Agency for Healthcare Research and Quality (AHRQ) for cross-sectional studies [18]. The NOS consists of nine items with three components: selection (4 scores); comparability (2 scores); and exposure belong to case–control or outcome belong to cohort (3 scores). A study is considered as high quality if it has a score ≥ 7 [19]. The AHRQ is composed of 11 items with yes/no/unclear options: the “yes” response is scored 1 and, otherwise, is scored 0, and 8–11 scores are regarded as high quality [20].

Data synthesis and statistical analysis

We calculated overall odds ratio (OR) and correlative 95% confidence interval (CI) to assess the strength of the association between the genital mycoplasma infection and HPV infections and related disorders. We utilized Q test and I2 statistic to evaluate the heterogeneity between each study, and heterogeneity was considered significant if p < 0.1 or I2 > 50% [21]. We used random effects model to generate the pooled OR if significant heterogeneity was found; otherwise, the fixed-effects model was used. In addition, we performed subgroup analysis based on different types of cervical cytopathology (LSIL group included low-grade squamous intraepithelial lesion, cervical intraepithelial neoplasia I; HSIL group included high-grade squamous intraepithelial lesion, cervical intraepithelial neoplasia II or III.). We also performed Egger’s asymmetry tests and visually examined funnel plots to evaluate potential publication bias [22, 23]. Besides, we performed sensitivity analysis by excluding individual studies separately from analysis. All data analyses were performed by Revman software (Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014.).

Results

Literature search

After removing the duplicate records from Embase and PubMed databases, a total of 225 records were recognized. After screening for title and abstract, we excluded 199 records due to irrelevant content. The remained 26 records were screened by reviewing the full text. We excluded two studies [24, 25] with duplicate recruited population and two studies [26, 27] without detailed data on mycoplasmas. Eventually, 22 studies with a total of 16,181 participants were included in this meta-analysis. The screening process is shown in Fig. 1. Most of these studies were case–control or cross-sectional studies, except McNicol’s study, which was cohort design [28]. Among these studies, 12 studies [5, 6, 17, 29,30,31,32,33,34,35,36,37] reported U. urealyticum, 5 studies [5, 15, 32, 38, 39] reported U. parvum, 9 studies [5, 17, 28, 30, 35,36,37,38,39] reported M. hominis, and 8 studies [5, 16, 39,40,41,42,43,44] reported M. genitalium. Recruited control group included community-based women (n = 6), hospital-based patients (n = 15), and female sex workers (n = 2). Among those 22 studies, 9 studies investigated non-pregnant women, 2 studies involved pregnant women, and the rest of studies (n = 11) did not describe the status of pregnancy. The characteristics of eligible studies are presented in Table 1. The pooled results are shown in Table 2.

Fig. 1
figure 1

Flow chart of study selection process

Full size image
Table 1 Summary of eligible studies in the meta-analysis
Full size table
Table 2 Summary of pooled results
Full size table

Association between genital mycoplasmas infection and HPV infections (Fig. 2)

U. urealyticum and HPV

Eight studies with a total of 7006 participants were included, and the pooled analysis showed that women infected with U. urealyticum had a significantly increased risk of HPV infections compared with U. urealyticum-negative women. Pooled OR was 1.57 (95% CI 1.05–2.34, p = 0.03), with I2 = 80% for heterogeneity.

Fig. 2
figure 2

Forest plot of association between genital mycoplasmas infections and HPV infection. a U. urealyticum and HPV infection. b U. parvum and HPV infection. c M. hominis and HPV infection

Full size image

U. parvum and HPV

A total of three studies with 692 participants were included, and the pooled analysis demonstrated that U. parvum was associated with a significantly increased risk of HPV infections (OR 3.02, 95% CI 2.10–4.33, p < 0.01), and I2 = 55% for heterogeneity.

M. hominis and HPV

A total of three studies with 549 participants were included, and we failed to find a significant statistical association of M. hominis infection with HPV infections. Pooled OR was 1.64 (95% CI 0.93–2.91, p = 0.09), with I2 = 0% for heterogeneity.

Association between genital mycoplasmas and high-risk HPV infections (Fig. 3)

U. urealyticum and high-risk HPV

Pooled analysis of four studies with 6197 participants showed that U. urealyticum infection was associated with a significantly increased risk of high-risk HPV infections. Pooled OR was 1.37 (95% CI 1.05–1.80, p = 0.02), with I2 = 54% for heterogeneity.

Fig. 3
figure 3

Forest plot of association between genital mycoplasmas infections and high-risk HPV infection. a U. urealyticum and high-risk HPV infection. b M. genitalium and high-risk HPV infection

Full size image

M. genitalium and high-risk HPV

A total of five studies with 3336 participants were included, and analysis showed that women infected with M. genitalium had a significantly increased risk of high-risk HPV infections. Pooled OR was 1.50 (95% CI 1.11–2.02, p < 0.01), with I2 = 6% for heterogeneity.

Association between genital mycoplasmas and abnormal cervical cytopathology (Fig. 4)

U. urealyticum and abnormal cervical cytopathology

A total of 9 studies with 2751 participants were included in the pooled analysis, and the results demonstrated that U. urealyticum infection was associated with a significantly increased risk of abnormal cervical cytopathology. Pooled OR was 1.51 (95% CI 1.23–1.85, p < 0.01), with I2 = 36% for heterogeneity. Based on different abnormal cervical cytopathologies, we performed a subgroup analysis by LSIL group and HSIL group. The pooled results from six studies showed that U. urealyticum infection was associated with both increased risk of LSIL (OR 2.02, 95% CI 1.49–2.74, p < 0.01, I2 = 47%) and HISL (OR 1.91, 95% CI 1.38–2.66, p = 0.09, I2 = 48%) with moderate heterogeneity. Besides, as shown in Fig. 5, the symmetric shape of funnel plots suggested that there was no significant publication bias among the studies. The results of sensitivity analysis also remained stable when omitting each individual study.

Fig. 4
figure 4

Forest plot of association between between genital mycoplasmas infections and abnormal cervical cytopathology. a U. urealyticum and abnormal cervical cytopathology. b U. parvum and abnormal cervical cytopathology. c M. hominis and abnormal cervical cytopathology. d M. genitalium and abnormal cervical cytopathology

Full size image
Fig. 5
figure 5

Funnel plot of association between U. urealyticum and abnormal cervical cytopathology. The circles represent the nine included studies about association between U. urealyticum and abnormal cervical cytopathology. The horizontal axis represents the size of association, while the vertical axis represents the standard error. The fixed-effects summary estimate is indicated by the vertical line, and the expected 95% CI of the standard error is indicated by the vertical line

Full size image

U. parvum and abnormal cervical cytopathology

A total of four studies with 1750 participants were included, and pooled analysis showed that U. parvum infection was associated with a significantly increased risk of abnormal cervical cytopathology. Pooled OR was 1.41 (95% CI 1.10–1.80, p = 0.006), with I2 = 2% for heterogeneity. The subgroup analysis showed that U. parvum infection increased risk of HSIL (OR 1.71, 95% CI 1.21–2.43, p = 0.002). Compared with HSIL group, we failed to find any significant result in LSIL group analysis (OR 1.27, 95% CI 0.95–1.70, p = 0.11).

M. hominis and abnormal cervical cytopathology

A total of 6 studies with 2037 participants were included. The results showed that M. hominis was associated with a significantly increased risk of abnormal cervical cytopathology. Pooled OR was 1.48 (95% CI 1.10–1.99, p = 0.009), with I2 = 0% for heterogeneity. In the subgroup analysis, no significant association was found between M. hominis infection and LSIL (OR 1.30, 95% CI 0.77–2.21, p = 0.33) or HSIL (OR 1.28, 95% CI 0.73–2.25, p = 0.39) group.

M. genitalium and abnormal cervical cytopathology

A total of five studies with 2415 participants were included, no significant statistical association was found between M. genitalium infection and abnormal cervical cytopathology. Pooled OR was 0.78 (95% CI 0.48–1.26, p = 0.31), with I2 = 22% for heterogeneity. In the subgroup analysis, we also did not observe that any significant result was found.

Association between genital mycoplasmas and cervical cancer

As the number of studies concerning the association between genital mycoplasma infection and cervical cancer was limited, we did not perform quantitative analysis of the results. Biernat-Sudolska and his colleagues found that the proportion of women infected with U. urealyticum was significantly higher in women with cervical cancer than women with normal cytopathology. However, they did not analyze M. hominis or M. genitalium infections in women with cervical cancer [39]. While in the study conducted by Hare and his colleagues, U. urealyticum and M. hominis were isolated with similar frequency from both cervical cancer group and normal cytopathology group [37].

Discussion

To our knowledge, this study is the first systemic review and meta-analysis investigating the association between genital mycoplasmas infection and HPV infections, abnormal cervical cytopathology, and cervical cancer. Our findings suggested that both U. urealyticum and U. parvum infection may increase the risk of high-risk HPV infections and abnormal cervical cytopathology. While M. genitalium infection may increase the risk of high-risk HPV infections and M. hominis infection may increase the risk of abnormal cervical cytopathology.

Concurrent co-infection of multiple pathogens was considered to be one of the most important risk factors for progression of HPV infections or cervical dysplasia [30, 38]. Previous studies suggested that genital mycoplasmas, as a common type of vaginal pathogen, may influence the natural history of HPV infections by initiating cellular anomalies [13]. Our meta-analysis further explored their relationships through pooling epidemiological data and found that U. urealyticum and M. genitalum may increase the risk of high-risk HPV infections. Despite the large number of patients included, our results still need to be interpreted with caution, as moderate heterogeneity existed among the included studies. Besides, the included studies were mostly cross-sectional studies, lacking in a longitudinal view of the HPV infections process in patients; thus, it was hard to investigate the relationship between genital mycoplasma infection and persistent HPV infections. Future prospective cohort studies are needed to explore the relationships of multiple co-infected pathogens with persistent HPV infections in a longitudinal perspective.

Abnormal cervical cytopathology, as an advanced pathological change largely caused by high-risk HPV infections, represented moderate or severe cervical intraepithelial neoplasia or cervical carcinoma in situ. Our results showed that U. urealyticum, U. parvum, and M. hominis can increase the risk of abnormal cervical cytopathology transformation. No heterogeneity was observed among studies concerning U. urealyticum, U. parvum, and M. hominis infections. In clinical practice, the management of LSIL and HSIL is quite different, LSIL tends to be observed during follow-up, while HSIL tends to receive further progressive treatment. Therefore, we then performed a subgroups analysis by analyzing LSIL and HSIL separately, and we found that U. urealyticum infection can increase both the risk of LSIL and HSIL, which suggested that U. urealyticum might have a distinct effect on progression of LSIL and HSIL. The possible mechanism of the association between U. urealyticum infection and abnormal cervical cytopathology might be related to the combination of several complex infection-associated inflammatory responses [15], involving production of reactive oxidative metabolites, increased expression of cytokines, chemokines, and growth and angiogenic factors, decreased cell-mediated immunity, and the generation of free radicals [45].

A potential association between cancers and infection with mycoplasma has been suspected since the 1960s [46]. However, early evidence was restricted by difficult culture conditions of these microbes until the widespread of techniques like PCR, immunohistochemistry, and serum antibody status. Recent study by Barykova et al. was the latest one that indicated a strong link between mycoplasma species and prostate cancer [47]. Besides, Baczynska et al. found that M. hominis and M. genitalium infections might play an important role in ovarian cancer [48]. In addition to those findings in other cancers, several a few studies investigated the association between genital mycoplasma and cervical cancer, which found significant increased positive rate of U. urealyticum in women with cervical cancer, compared to women with normal cytology findings [39]. Furthermore, laboratory studies confirmed the ability of mycoplasma to cause or promote oncogenic transformation [49]. Several different species have been proven to transform rodent and human lines of diverse lineages in vitro [50]. However, as limited numbers of studies available, the role of genital mycoplasmas in cervical cancer is still in ambiguity, and further epidemiological studies and prospective prognosis studies are needed in the future.

Some limitations must be addressed when interpreting our results. First of all, our results were based on pooled analysis of crude epidemiological data. Most included studies did not provide adjusted OR for the association between genital mycoplasmas infection and related disorders, and matching of baseline and other risk factors was also not available. Second, there were some innate heterogeneities in our study due to inclusion of different designs of studies, i.e., cross-sectional and case–control studies. Third, description of the details about concurrent co-infection of other pathogens, such as C. trachomatis and T. vaginalis, was not available in some included studies, which might result in possible selected bias of the patients. Fourth, most included studies concerning HPV infections were cross-sectional studies, which limited us to explore the association between genital mycoplasmas infection and persistent HPV infections. Fifth, the association between ureaplasma and mycoplasma infection and HPV infection related disorders may also be resulted from a more promiscuous sexual life which leads to a higher incidence and prevalence of sexually transmitted diseases. This confounding factor should be taken into account when interpreting our results.

Conclusions

Our systemic review and meta-analysis found that U. urealyticum and U. parvum may increase the risk of HPV, U. urealyticum and M. genitalium may increase the risk of high-risk HPV infections, and U. urealyticum, U. parvum, and M. hominis may increase the risk of abnormal cervical cytopathology. More well-designed longitudinal studies investigating the changes of the natural history of concurrent co-infection of genital mycoplasmas and persistent HPV infection are warranted in the future.