‘Mycoplasma hominis does not share common risk factors with other genital pathogens’: Findings from a South African pregnant cohort

Background The role of Mycoplasma hominis (M. hominis) as a genital tract pathogen was still debatable. This study identified the risk factors associated with the prevalence of M. hominis in South African pregnant women. Methods This was a cross-sectional analysis of n = 221 prenatal patients attending a Durban hospital during November 2017 to April 2018. M. hominis was detected from urine samples using the quantitative polymerase chain reaction. The population characteristics were described using frequencies stratified by the infection status of M. hominis. In addition, a univariate analysis was used to assess the relationship between each risk factor and infection status. The analysis further considered logistic regression to assess the influence of these risk factors univariately and in the presence of other factors. The coinfection rate between M. hominis and bacterial vaginosis (BV), Trichomonas vaginalis (T. vaginalis), Mycoplasma genitalium (M. genitalium) and Candida species was also determined. All the tests were conducted at 5% level of significance. Results The prevalence of M. hominis in this study population was 48% (106/221). In the univariate analysis, factors significantly associated with M. hominis positivity included having past abnormal vaginal discharge (p = 0.037), having current abnormal vaginal discharge (p = 0.010) and a borderline significance (p = 0.052), which were noted for previous pre-term delivery. However, none of these factors were sustained in the multivariate analysis. There was a statistically significant association between M. hominis and BV positivity (p < 0.001). Similarly, M. hominis and M. genitalium positivity was significant (p = 0.006). Conclusion This study showed that M. hominis does not share common risk factors with known genital tract pathogens in a population of pregnant women and therefore cannot be considered a genital tract pathogen.

http://www.sajid.co.za Open Access spontaneous abortion, stillbirth, preterm birth, low birth weight and perinatal mortality. 7,8,9,10,11,12 An earlier study had provided evidence stating that M. hominis is not a vaginal pathogen in adults. 13 However, a study published in the same year by Arya et al. referred to M. hominis as a vaginal pathogen because of its association with Trichomonas vaginalis (T. vaginalis). 14 There have been limited studies since 2001, which have contributed to resolving the discordance regarding the role of M. hominis as a genital tract pathogen. The aim of the current study was to describe the prevalence and factors associated with M. hominis in pregnant women as well as its likelihood of being considered as a genital tract pathogen.

Study setting and population
This study was a sub-study of a larger study that investigated the laboratory-based diagnosis of vaginal infections in pregnant women from Durban. The larger study included n = 273 women, 18 years and older from all gestational ages who are willing to provide written informed consent.
The study population was recruited from the King Edward VIII hospital in Durban, KwaZulu-Natal between October 2017 and April 2018. All patients were outpatients. There was a 10% refusal rate by the women during screening. A questionnaire was administered to collect data on the women's demographics, sexual behaviour and clinical information. All interviews were conducted in private, and all study-related information was stored securely. All records and specimens had been identified by study identification numbers only to maintain participant confidentiality.
Only participants who had given written informed consent were included in the study. The study did not collect data on whether the women were experiencing any complications during pregnancy. During the study visit, women were asked to provide self-collected vaginal swab and urine samples. The women were tested for HIV at the clinic as part of routine care. Permission to obtain data on HIV status was obtained from the women. Because of the cross-sectional design of the study, the women were not followed up to collect information on pregnancy outcomes.
For the sub-study, n = 221 urine DNA extracts were available for analysis. Laboratory testing and analyses were performed at the School of Clinical Medicine Laboratory, University of KwaZulu-Natal.

Data collection
At enrolment, a face-to-face questionnaire was administered to collect data on the women's demographics (age, level of education and marital status), sexual behaviour (condom use, number of lifetime sex partners, age of sexual debut, partner having other partners, intravaginal practices, cohabitation status and recreational habits such as smoking and consuming alcohol) and clinical information (gestational age, history of previous pregnancies and history of STIs).

Detection of Mycoplasma hominis from urine
A sensitivity detection assay was performed on the urine DNA extracts. DNA extraction involved a starting volume of 10 ml of urine. A standardised starting volume was used across all samples. The urine was centrifuged for 45 min at 14 000 × g and the supernatant was discarded. Non-template and positive controls (TaqMan™ Vaginal Microbiota Extraction Control; cat no. A32039) were also included. Amplification was performed at 95° C for 30 s followed by 45 cycles comprising denaturation at 95° C for 3 s and annealing at 60° C for 30 s. Detection of amplified fluorescent products was carried out at the end of the annealing phase. The raw fluorescence data that included the C T mean values were automatically generated using the Quant Studio 5 Real-time PCR system software.

Detection of Mycoplasma genitalium from urine
Mycoplasma genitalium was detected using the TaqMan Realtime PCR (sensitivity) assay (Thermo Fisher Scientific, United States) using commercially available primers and probes specific for M. hominis (Ba04646249_sl). The reaction and cycling conditions were as per the M. hominis assay conditions.

Detection of bacterial vaginosis, Trichomonas vaginalis and Candida species from vaginal swabs
The presence of BV, T. vaginalis and Candida species was detected using the BD Max TM Vaginal Panel assay (Becton Dickinson, United States) from a single vaginal swab. The assay was performed as per the manufacturer's recommendations.

Statistical data analyses
The statistical data analysis was conducted in a freely available Statistical Computing Environment, R software, version 3.6.3 using the RStudio platform. Initially, the population characteristics were described using frequencies stratified by the infection status of the pathogens.
In addition to the frequencies, univariate analysis was used to assess the relationship between each risk factor and the pathogen infection status. The available continuous variable had a skewed distribution calling for a nonparametric test involving a rank-sum test. On the other hand, the categorical risk factors were univariately assessed using the Chi-Square test or the Fisher's exact test in the case of smaller expected frequencies. The significant risk factors were used to fit univariate logistic regressions in order to quantify their relationships with the outcome in terms of odds ratios (ORs). The analysis further considered multiple logistic regression to assess the influence of these univariately significant risk factors in the presence of the other factors. All the tests were conducted at 5% level of significance.

Ethical considerations
Ethics approval for this study was granted by the Biomedical Research Ethics Committee (BREC) of the University of KwaZulu-Natal (BE214/17).

Mycoplasma hominis status
The prevalence of M. hominis in the study population was 48% (106/221).

Predicting the risk of Mycoplasma hominis infection in the presence of other genital infections
The results in Table 4

Association between Mycoplasma hominis and HIV infection
For this analysis, the data on HIV status were available for n = 104 women. A large proportion of the women refused to provide these data. For the 104 women for whom data were available, it was shown that the prevalence of coinfection between HIV and M. hominis was 17.3% (11/104), constituting 45% (18/40) of the M. hominis-positive women. However, there was no statistically significant association between prevalent HIV and prevalent M. hominis (p = 0.975) (Figure 1). That is, the prevalence of M. hominis infection amongst the HIV-negative and HIV-positive women was similar at approximately 45%.

Discussion
To the best of our knowledge, this is the first study to provide an estimate on the prevalence of M. hominis in pregnant women from the Durban area in South Africa. We report a prevalence estimate of 48% for M. hominis in this study population. Our data are consistent with a previous study conducted in South Africa where Redelinghuys and colleagues also reported high prevalence data for M. hominis (50.7%) in pregnant women from Gauteng, South Africa. 5 Going back to the overall study aim, which was to identify risk factors associated with M. hominis as well as to determine if M. hominis shared risk factors with other genital infections, the following factors were significantly associated with the    15 Similarly, Naidoo et al. also showed that women with the prevalent STIs reported less than high school education (p < 0.0001). 16 A significant association between low level of education and prevalence of the viral STI, Herpes simplex virus-2 was also found (p = 0.021). 17 In this study, current abnormal vaginal discharge was significantly associated with the prevalence of M. hominis.
The majority of women who tested positive did not report symptoms of discharge This is not in keeping with another published study conducted in pregnant women, which showed the association of abnormal vaginal discharge as a true risk factor for another genital pathogen. 19 The current study also showed that obtaining a tertiary level of education had significantly reduced the women's risk of infection by 71%. This finding is consistent with other previous studies conducted in KwaZulu-Natal. 16, 17 Abbai and co-workers reported a significant association between HSV-2 infection and women who had received a lower level of education. 17 Similarly, Naidoo et al. associated women receiving lower level education with an increased risk of having a prevalent STI. 16 However, unlike the findings described by Abbai et al. 17 and Naidoo et al., 16 level of education was not significant in the multivariable analysis. Therefore, level of education cannot be deemed as a true low risk factor in this study.  22 Testing for the presence of Ureaplasma species in the study cohort is a future research direction.
This study was limited in that samples were collected from pregnant women attending a single antenatal facility. However, the hospital from which the women were sampled in this study serves as a central hospital for women from around the Durban area, thereby making the population more generalised. A second limitation is the lack of data on pregnancy outcomes in relation to the prevalent infections. Because of the cross-sectional nature of this study, the data were not collected; however, this limitation will be addressed in future studies. Finally, a full dataset on HIV status was not available for this study, because of the refusal to provide the data by the study women and therefore, this study was unable to draw sound conclusions regarding the association between M. hominis and HIV infections.

Conclusion
To date, there remains uncertainty regarding the role of M. hominis as a genital tract pathogen. The current study has now provided evidence from a South African-based pregnant population, indicating that M. hominis does not share common predisposing risk factors with that of known genital tract pathogens as well as the causative agents of STIs. Based on these study findings, M. hominis cannot be considered a genital tract pathogen. Previous studies have shown a high prevalence of M. hominis in the vaginal compartment. 23,24 Based on the high prevalence of this pathogen in the vaginal micro-environment, future studies that investigate its explicit role in this environment are needed.