Mycoplasma genitalium is an emerging sexually transmitted infection (STI) associated with reproductive tract syndromes in both men and women. In men, infection with M. genitalium causes urethritis, and in women it is associated with cervicitis and pelvic inflammatory disease [1–4]. Mycoplasma genitalium infection is often asymptomatic, and is able to persist for almost 2 years [4–7]. The optimal treatment for eradication of M. genitalium is still under investigation, but to date it would appear that azithromycin shows the best results in treating M. genitalium, with Moxifloxacin reserved for cases of treatment failure [8–10]. Mycoplasma genitalium is a common infection among HIV-infected individuals [1,5,11–14], and as with other STIs, may enhance HIV acquisition. Introduction of M. genitalium into the genital tract has been shown to cause mucosal disruption and an inflammatory response [15–17]. The associated inflammatory response triggers production of cytokines , which in turn initiate recruitment and activation of HIV susceptible cells [19,20]. This creates an opportunity for HIV to cross the genital mucosal barrier, accessing the subepithelium and attaching to these susceptible cells where HIV replication can begin. In-vitro studies have also shown that M. genitalium can enhance HIV viral replication, thus potentially increasing transmissibility and accelerating disease progression [21,22].
A systematic review and meta-analysis showed a strong and consistent association between M. genitalium and HIV infection [odds ratio (OR) = 2.01; 95% confidence interval (CI) 1.44–2.79] . However, studies in this review were almost exclusively cross-sectional, and did not provide evidence of a causal role of M. genitalium on either HIV acquisition or transmission. A longitudinal study in Kenya found that prevalent HIV infection was independently associated with incident M. genitalium. A study among HIV seroconcordant and discordant couples in the United States found that M. genitalium infection was independently associated with a two-fold to three-fold risk of HIV seroconcordance, suggesting a role for M. genitalium on HIV transmission . Two cross-sectional studies have looked at the association between M. genitalium and detection of genital HIV. Among women in Kenya, a high M. genitalium organism burden was independently associated with a three-fold risk of detectable genital HIV DNA , whereas among women in the United States no association was found between M. genitalium and detection of vaginal HIV RNA [adjusted odd ratio (AOR) = 0.93, 95% CI 0.31–2.79)] .
To date, no studies have explored the association of M. genitalium with HIV acquisition. We analyzed a nested case–control study to evaluate the longitudinal association between M. genitalium and HIV-1 acquisition among women from Zimbabwe and Uganda.
Between 1999 and 2004 women from Zimbabwe and Uganda enrolled in a cohort study to assess the effect of hormonal contraception use on HIV-1 acquisition. In Zimbabwe, 2296 women were recruited from family planning clinics, and in Uganda 1837 women were recruited from family planning clinics, and 398 from STI clinics and sex worker networks. Details of this study have been published previously . Briefly, participants were 18–35 years of age and HIV-negative. They were seen once every 12 weeks for a median of 22 months. Demographic information was collected at baseline, and behavioral and reproductive tract infection (RTI) and STI data were collected at each visit. A speculum examination was conducted at each visit. Cervical swabs were collected for Neisseria gonorrhoeae and Chlamydia trachomatis testing. Immediately after collection swabs were placed in 1 ml Amplicor transport medium, with 1 ml diluent added and stored at −5°C, and were frozen at −20°C within 48 h of collection. All participants provided written informed consent and the study was approved by ethical review boards in the United States, Zimbabwe, and Uganda prior to initiating study procedures.
During follow-up, 218 women acquired HIV-1, including 155 in Zimbabwe and 63 in Uganda. Of these, samples from 152 women in Zimbabwe and 38 in Uganda were available from the sample repository, and were classified as cases for this research. Each case was matched with up to two controls (353 control samples in total). Controls were women who remained HIV-negative throughout the period of study follow-up. Matching criteria were based on: study site, age, follow-up time, and a composite RTI indicator variable. The RTI indicator variable was designed to control for risk of exposure rather than pathogen interaction. It was equal to one if women were infected with C. trachomatis, N. gonorrhoeae, or bacterial vaginosis (BV) at either of the two visits being evaluated, and to zero if negative for all three at both visits .
Blood was collected for HIV diagnosis using enzyme-linked immunosorbent assays (ELISA) [Recombigen HIV-1/HIV-2 (Cambridge Biotech, Galway, Ireland), Organon Vironostika (Organon Teknika, Durham, North Carolina, USA), Abbott Murex (Abbott Park, Illinois, USA), Sanofi (Sanofi Diagnostics Pasteur, Redmond, Washington, USA)] and HIV rapid testing [HIV SAV1 or SAV2 (Savyon Diagnostics, Ashdod, Israel), Capillus HIV-1/HIV-2 (Trinity Biotech USA, Jamestown, New York, USA), or Determine (Abbott)], with confirmatory testing using Western blot (BioRad, Hercules, California, USA) and/or HIV PCR (Amplicor HIV-1 DNA test, version 1.5; Roche Diagnostics, Branchburg, New Jersey, USA). After confirmation of infection, HIV PCR was performed on samples from previous visits and the HIV detection visit was defined as the first visit with a positive HIV PCR result.
Serological diagnosis of HSV-2 was conducted using ELISA (Focus Diagnostics, Cypress, California, USA). C. trachomatis and N. gonorrhoeae were diagnosed by PCR using Amplicor CT/NG (Roche Diagnostics), according to U.S. package insert instructions with one modification: due to limitations of local plate readers the cutoff for N. gonorrhoeae positivity was set at an optical density (OD) of greater than or equal to 2.500 at 450 nm absorbance. PCR testing for Trichomonas vaginalis was conducted retrospectively on samples at follow-up visits using the Amplicor CT/NG master mix with a T. vaginalis-specific primer added . Vaginal swabs were collected for wet mount microscopy, as well as BV diagnosis using Nugent scoring.
Mycoplasma genitalium testing was performed on samples at the HIV detection visit (t0) and the last HIV-negative visit (t-1) for cases, and equivalent visits in follow-up time for controls. Presence of M. genitalium DNA was assessed by in-house PCR-ELISA, using the archived cervical swab samples which had been stored at −20°C since collection. Specific primers used in this assay targeted the MgPa 1 and MgPa 3 genes. This assay has been described in detail elsewhere . All samples with an OD of greater than 0.100 at 405 nm were repeated, and a positive result was defined as an OD upon repeat testing of greater than 0.100 at 405 nm after background subtraction.
We did not have access to real-time PCR testing for all samples in this study, and funding precluded shipping all study samples to another laboratory. However, 63 (6.8%) samples were sent to the Sexually Transmitted Infections Reference Centre in South Africa for validation using the commercial Mycoplasma genitalium Real-TM assay (Sacace Biotechnologies, Como, Italy). Using results from this validation, we performed a receiver operating characteristics analysis to verify the optimal cutoff point for the PCR-ELISA, which confirmed an OD of greater than 0.100, as per the published protocol.
STI symptoms and sexual behaviors were assessed by participant self-report in interview-administered questionnaires. Frequency of condom use in the previous 3 months was categorized as always/sometimes/never. Participants reporting either sex work or multiple sexual partners or having a new sex partner since the previous visit were combined to form a composite high-risk behavior variable. Participant-reported primary partner risk, including a HIV-infected partner, history of sex with a sex worker, urethral discharge, weight loss, or spending nights away from home, was similarly combined into a composite variable.
Baseline characteristics of cases and controls were compared using chi-squared and t tests. To explore risk factors for HIV-1 acquisition we used time-dependent variables from data at the t-1 and t0 visits, respectively, and time-independent variables using baseline data. We analyzed risk factors for HIV-1 acquisition using conditional logistic regression to estimate ORs and 95% CIs, with M. genitalium as the primary exposure of interest. As there was no difference in HIV-1 incidence by hormonal contraceptive use group in the main trial , a pooled analysis was conducted combining all three contraceptive groups. We assessed risk factors for HIV-1 acquisition using likelihood ratio tests to determine the strength of associations. Variables associated with HIV-1 at the P ≤ 0.20 level in univariable analysis were entered into a multivariable conditional logistic regression model. A hierarchical model was used, first fitted on sociodemographic and behavioral factors and then on STI-related variables, as they may be on the causal pathway between sociodemographic and behavioral factors and HIV . Variables were retained in the final model if they were associated with HIV-1 at P ≤ 0.10, or if they modified the effect estimate of our primary exposure, M. genitalium, by greater than 10%. The population attributable fractions (PAFs) of M. genitalium and other factors on HIV-1 acquisition were estimated using AORs . Statistical analyses were conducted using STATA 11 (StataCorp LP, College Station, Texas, USA).
Table 1 presents baseline characteristics for cases (HIV seroconverters) and controls (nonseroconverters), respectively. The median number of days between the two evaluation visits was 83 and 82 for cases and controls, respectively. Cases and controls were similar in terms of age at enrolment (mean 25 years), parity (mean 2.2), and education (mean 9.5 years of school).
At baseline, cases had a higher risk profile, being less likely than controls to be living with their partner (31.8 vs. 68.2%; P < 0.001), and more likely to have consumed alcohol in the last 30 days (21.1 vs. 13.7%; P = 0.03), to have more lifetime partners (median 2 vs. 1; P = 0.03), to report having had more than one partner in the last 12 months (12.6 vs. 5.6%; P = 0.004), to have ever engaged in sex work (3.2 vs. 0.3%; P = 0.004), to report never using condoms (46.8 vs. 58.8%; P = 0.01), and to report high participant behavioral (7.2 vs. 2.2%; P = 0.004) and partner risk (52.7 vs. 37.8%; P = 0.001). Cases were also more likely than controls to be infected with HSV-2 (72.5 vs. 51.1%: P < 0.001), N. gonorrhoeae (8.5 vs. 2.0%; P < 0.001), trichomoniasis (8.5 vs. 2.3%; P = 0.001) and candidiasis (18.5 vs. 12.9%; P = 0.08).
A total of 183 case and 337 control samples were available from the sample repository and were tested for M. genitalium at the t-1 visit (prior to HIV-1 detection). Mycoplasma genitalium was a common infection in these populations (Fig. 1), with a prevalence of 9.4% at this visit (14.8% in cases, 6.5% in controls), higher than both N. gonorrhoeae (5.8%) and C. trachomatis (4.7%). A total of 181 case and 302 control samples were tested at the t0 HIV-1 detection visit. Prevalence of M. genitalium was 9.7% at this visit (14.9% in cases, 6.6% in controls), and was again more prevalent than N. gonorrhoeae (6.7%) and C. trachomatis (4.4%). Of 460 women with M. genitalium data available for both visits 84 (18.3%) women had at least 1 M. genitalium infection, and of these 9 (10.7%) were infected at both visits.
In univariable analysis, M. genitalium at the t-1 visit was strongly associated with subsequent HIV-1 acquisition (OR = 2.23; 95% CI 1.21–4.12; Table 2). In multivariable analysis, after adjusting for potential confounding, the association between M. genitalium and HIV-1 acquisition persisted (AOR = 2.42; 95% CI 1.01–5.80). Other factors measured at the t-1 visit that were independently associated with subsequent HIV-1 acquisition included partner risk (AOR = 2.02, 95% CI 1.17–3.50), HSV-2 infection (AOR = 7.43, 95% CI 3.83–14.41), and T. vaginalis infection (AOR = 3.18, 95% CI 1.25–8.11) (Table 2).
At the t-1 visit, it was estimated that 8.7% (95% CI 0.1–12.2) of incident HIV-1 infections were attributable to M. genitalium. By comparison, 9.1% of incident HIV-1 infections were estimated to be attributable to T. vaginalis, 72.6% to HSV-2, and 23.6% to partner risk.
Mycoplasma genitalium at the t0 visit was also strongly associated with HIV-1 acquisition between t-1 and t0 (OR = 2.34; 95% CI 1.23–4.46), and this association persisted after adjustment for the potential confounding variables in Table 3 (AOR = 2.18; 95% CI 0.98–4.85). Other factors measured at the t0 visit that were independently associated with recent HIV-1 acquisition included living with a partner (AOR = 0.39, 95% CI 0.18–0.84), hormonal contraceptive use (AOR = 1.73, 95% CI 0.96–1.31), HSV-2 infection (AOR = 4.66, 95% CI 2.66–8.16), N. gonorrhoeae (AOR = 4.76, 95% CI 1.79–12.67), candidiasis (AOR = 2.03, 95% CI 0.99–4.13), and reported abnormal vaginal discharge since last visit (AOR = 2.24, 95% CI 1.00–5.02) (Table 3).
This is the first study to longitudinally assess the relationship between M. genitalium and HIV-1 acquisition. Mycoplasma genitalium was a common infection in these populations, and was more prevalent than other nonviral STIs. We found a greater than two-fold risk of HIV-1 acquisition among women infected with M. genitalium at both the visits prior to and at the time of HIV-1 acquisition.
Mycoplasma genitalium was the only nonviral STI independently associated with incident HIV-1 at both visits. The association at the time of HIV acquisition may be due to simultaneous infection of M. genitalium and HIV-1, or even subsequent M. genitalium infection, as HIV-1 has been shown to be a risk factor for M. genitalium acquisition . However, the association at the visit prior to HIV acquisition is likely to reflect an increased susceptibility to HIV-1 acquisition among women infected with M. genitalium, possibly due to an acute or chronic host cell response.
There is good biological plausibility for an association between M. genitalium and HIV acquisition. In-vitro studies have shown that the introduction of M. genitalium into the urogenital tract causes tissue damage, primarily due to the host cell inflammatory response and cytokine secretion . Proinflammatory cytokines can initiate recruitment of susceptible inflammatory cells to the endocervix, and increase presence and activation of other HIV susceptible cells [19,20]. This host response plausibly increases susceptibility to HIV infection.
Mycoplasma genitalium is capable of persisting for long periods [7,32], however, it is possible that the impact of M. genitalium on HIV susceptibility may vary by stage of infection. Mycoplasma genitalium is capable of intracellular infection, and in mouse models intracellular M. genitalium is metabolically active with higher organism burden reported as compared to extracellular infection . Further research is required to explore the impact of stage of M. genitalium infection, and intracellular versus extracellular infection, on HIV susceptibility.
A major strength of this study was the prospective design, which allowed us to assess the effect of preexisting M. genitalium infection on subsequent HIV-1 acquisition. Further, the study was conducted among two general populations of African women, and therefore results are more likely generalizable to other populations. Highly sensitive laboratory techniques were used to diagnose other STIs, and to confirm the timing of the HIV-1 detection visit, thus reducing bias due to misclassification of potential confounders or our outcome.
A limitation of this research is the relatively low sensitivity of PCR-ELISA assays for detection of M. genitalium as compared to real-time PCR. Validation using real-time PCR as the gold standard showed a sensitivity and specificity of 68% and 89%, in 37 positive and 26 negative samples, respectively. The numbers precluded analysis of sensitivity and specificity by HIV status, but if the sensitivity and specificity was similar by HIV-1 status this is likely to have diluted the association detected between M. genitalium and HIV-1 acquisition. The numbers also precluded analysis stratified by country, although in both countries women who seroconverted to HIV-1 were more often infected with M. genitalium. Another limitation was that we were unable to measure M. genitalium quantitatively. Future research should investigate the effect of bacterial load on HIV acquisition. As with any research estimating the impact of STIs on HIV acquisition, there is the possibility of residual confounding due to the shared transmission pathways. While the parent study did not measure actual exposure to HIV through partner testing, we have made an effort to control for partner risk through participant's perceptions of their risk, though this is an imprecise measure.
Mycoplasma genitalium is not currently a part of standard STI screening programs, and in the absence of rapid diagnostic tests, M. genitalium would be treated syndromically. The failure of tetracyclines to eradicate M. genitalium is noteworthy, as particularly in developing countries this is also the most common treatment for non-gonococcal urethritis, an infection associated with M. genitalium. With the exception of azythromicin common treatments for N. gonorrhea and C. trachomatis, which many syndromic treatments target, are less effective in treating M. genitalium. Further research on the prevalence and natural history of M. genitalium is required, but given its high prevalence in these populations, inclusion of M. genitalium as part of routine syndromic management may be worth considering. This would, however, require revision to current guidelines.
The present study showed an increased susceptibility to HIV-1 infection among M. genitalium-infected women. An estimated 8.7% of incident HIV-1 infections were attributable to M. genitalium, suggesting that M. genitalium may be an important, modifiable risk factor for HIV-1 acquisition among women in Zimbabwe and Uganda. Prevalence was higher than other bacterial STIs, which are a mainstay of STI control programs around the world. Further research will be required to confirm a causal relationship and to identify risk factors for M. genitalium infection in African populations. If findings from this research are confirmed, M. genitalium screening and treatment among women at high risk for HIV may be warranted for HIV prevention.
S.M.N. and B.V.D.P. developed the study design. A.V.D.S., S.N.M., and T.C. drafted the funding application and developed the study protocol. C.M., R.S., and T.P. were Investigators on the main trial from which samples and data have been utilized. S.N.M., B.V.D.P., and F.M. conducted the laboratory testing. S.N.M., H.W., and C.K. developed the analysis plan, and analysis was conducted by S.N.M. S.N.M. wrote the first draft of the manuscript. All authors contributed to manuscript writing and review. We would like to thank the study participants in Zimbabwe and Uganda for their participation in the study. Special thanks to B.V.D.P. for her extraordinary assistance with laboratory testing of specimens. This research was funded by the University of California, San Francisco AIDS Research Institute's Innovative AIDS Pilot Awards Program.
Conflict of interest
B.V.D.P. consults for Roche Molecular Systems. No conflict of interest stated for all other authors.
1. Cohen CR, Manhart LE, Bukusi EA, Astete S, Brunham RC, Holmes KK, et al. Association between Mycoplasma genitalium and acute endometritis
2. Haggerty CL. Evidence for a role of Mycoplasma genitalium in pelvic inflammatory disease
. Curr Opin Infect Dis
3. Short VL, Totten PA, Ness RB, Astete SG, Kelsey SF, Murray P, et al. The demographic, sexual health and behavioural correlates of Mycoplasma genitalium infection among women with clinically suspected pelvic inflammatory disease
. Sex Transm Infect
4. Taylor-Robinson D, Jensen JS. Mycoplasma genitalium: from Chrysalis to Multicolored Butterfly
. Clin Microbiol Rev
5. Kapiga SH, Sam NE, Mlay J, Aboud S, Ballard RC, Shao JF, et al. The epidemiology of HIV-1 infection in northern Tanzania: results from a community-based study
. AIDS Care
6. Jensen JS. Mycoplasma genitalium infections. Diagnosis, clinical aspects, and pathogenesis
. Dan Med Bull
7. Cohen CR, Nosek M, Meier A, Astete SG, Iverson-Cabral S, Mugo NR, et al. Mycoplasma genitalium infection and persistence in a cohort of female sex workers in Nairobi, Kenya
. Sex Transm Dis
8. Bradshaw CS, Jensen JS, Tabrizi SN, Read TR, Garland SM, Hopkins CA, et al. Azithromycin failure in Mycoplasma genitalium urethritis
. Emerg Infect Dis
9. Hamasuna R, Jensen JS, Osada Y. Antimicrobial susceptibilities of Mycoplasma genitalium strains examined by broth dilution and quantitative PCR
. Antimicrob Agents Chemother
10. Jernberg E, Moghaddam A, Moi H. Azithromycin and moxifloxacin for microbiological cure of Mycoplasma genitalium infection: an open study
. Int J STD AIDS
11. Ballard RC, Fehler HG, Htun Y, Radebe F, Jensen JS, Taylor-Robinson D. Coexistence of urethritis with genital ulcer disease in South Africa: influence on provision of syndromic management
. Sex Transm Infect
12. Labbe AC, Frost E, Deslandes S, Mendonca AP, Alves AC, Pepin J. Mycoplasma genitalium is not associated with adverse outcomes of pregnancy in Guinea-Bissau
. Sex Transm Infect
13. Manhart LE, Mostad SB, Baeten JM, Astete SG, Mandaliya K, Totten PA. High Mycoplasma genitalium organism burden is associated with shedding of HIV-1 DNA from the cervix
. J Infect Dis
14. Pepin J, Labbe AC, Khonde N, Deslandes S, Alary M, Dzokoto A, et al. Mycoplasma genitalium: an organism commonly associated with cervicitis among west African sex workers
. Sex Transm Infect
15. Taylor-Robinson D, Tully JG, Barile MF. Urethral infection in male chimpanzees produced experimentally by Mycoplasma genitalium
. Br J Exp Pathol
16. Tully JG, Taylor-Robinson D, Rose DL, Furr PM, Graham CE, Barile MF. Urogenital challenge of primate species with Mycoplasma genitalium and characteristics of infection induced in chimpanzees
. J Infect Dis
17. Zhang S, Wear DJ, Lo S. Mycoplasmal infections alter gene expression in cultured human prostatic and cervical epithelial cells
. FEMS Immunol Med Microbiol
18. Wu Y, Qiu H, Zeng Y, You X, Deng Z, Yu M, et al. Mycoplasma genitalium lipoproteins induce human monocytic cell expression of proinflammatory cytokines and apoptosis by activating nuclear factor kappaB
. Mediators Inflamm
19. Ho JL, He S, Hu A, Geng J, Basile FG, Almeida MG, et al. Neutrophils from human immunodeficiency virus (HIV)-seronegative donors induce HIV replication from HIV-infected patients’ mononuclear cells and cell lines: an in vitro model of HIV transmission facilitated by Chlamydia trachomatis
. J Exp Med
20. Starnbach MN, Roan NR. Conquering sexually transmitted diseases
. Nat Rev Immunol
21. Phillips DM, Pearce-Pratt R, Tan X, Zacharopoulos VR. Association of mycoplasma with HIV-1 and HTLV-I in human T lymphocytes
. AIDS Res Hum Retroviruses
22. Sasaki Y, Honda M, Makino M, Sasaki T. Mycoplasmas stimulate replication of human immunodeficiency virus type 1 through selective activation of CD4+ T lymphocytes
. AIDS Res Hum Retroviruses
23. Napierala Mavedzenge S, Weiss HA. Association of Mycoplasma genitalium and HIV infection: a systematic review and meta-analysis
24. Perez G, Skurnick JH, Denny TN, Stephens R, Kennedy CA, Regivick N, et al. Herpes simplex type II and Mycoplasma genitalium as risk factors for heterosexual HIV transmission: report from the heterosexual HIV transmission study
. Int J Infect Dis
25. Gatski M, Martin DH, Theall K, Amedee A, Clark RA, Dumestre J, et al. Mycoplasma genitalium infection among HIV-positive women: prevalence, risk factors and association with vaginal shedding
. Int J STD AIDS
26. Morrison CS, Richardson BA, Mmiro F, Chipato T, Celentano DD, Luoto J, et al. Hormonal contraception and the risk of HIV acquisition
27. Van Der Pol B, Kwok C, Pierre-Louis B, Rinaldi A, Salata RA, Chen PL, et al. Trichomonas vaginalis infection and human immunodeficiency virus acquisition in African women
. J Infect Dis
28. Dutro SM, Hebb JK, Garin CA, Hughes JP, Kenny GE, Totten PA. Development and performance of a microwell-plate-based polymerase chain reaction assay for Mycoplasma genitalium
. Sex Transm Dis
29. Victora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach
. Int J Epidemiol
30. Bruzzi P, Green SB, Byar DP, Brinton LA, Schairer C. Estimating the population attributable risk for multiple risk factors using case-control data
. Am J Epidemiol
31. McGowin CL, Popov VL, Pyles RB. Intracellular Mycoplasma genitalium infection of human vaginal and cervical epithelial cells elicits distinct patterns of inflammatory cytokine secretion and provides a possible survival niche against macrophage-mediated killing
. BMC Microbiol
32. Blaylock MW, Musatovova O, Baseman JG, Baseman JB. Determination of infectious load of Mycoplasma genitalium in clinical samples of human vaginal cells
. J Clin Microbiol