Mycoplasma genitalium among gay, bisexual and other men who have sex with men in Montréal

Abstract

Background: The bacteria Mycoplasma genitalium has been identified as a causative agent of urethritis in men, especially in gay, bisexual and other men who have sex with men (gbMSM). Canadian clinic-based data have identified a high prevalence of M. genitalium and resistance to antibiotic treatments. This article estimates the prevalence of M. genitalium infections among Montréal gbMSM, explores correlates for M. genitalium infection and estimates the prevalence of mutations associated with antimicrobial resistance (AMR).

Methods: Engage Cohort Study is a multi-site longitudinal study on sexually active gbMSM, aged 16 years and older, recruited via respondent-driven sampling in Montréal, Toronto and Vancouver. Participants completed a questionnaire on behaviour and were tested for sexually transmitted and blood-borne infections at each visit. For this sub-study, Montréal participants with a follow-up visit that occurred between November 2018 and November 2019 were included.

Results: A total of 2,064 samples were provided by 716 participants. Prevalence of M. genitalium infection was 5.7% at rectal and/or urethral sites, 4.0% at rectal site and 2.2% at urethral site. Correlates for M. genitalium infection were younger age and reporting six or more sexual partners in the past six months. Prevalence of macrolide resistance associated mutations (MRAM), quinolone resistance associated mutations (QRAM) and either MRAM or QRAM, was 82%, 29% and 85%, respectively.

Conclusion: This first population-based study among gbMSM in Canada documents a high prevalence of urethral and rectal M. genitalium infection and high levels of AMR. Our results highlight the importance of access to testing and AMR detection when indicated.

Introduction

Mycoplasma genitalium has been identified as a growing health concern for sexually active gay, bisexual and other men who have sex with men (gbMSM) by causing acute, persistent or recurrent urethritis^{Footnote 1Footnote 2Footnote 3Footnote 4Footnote 5Footnote 6}. The data concerning M. genitalium as a causative agent of clinical proctitis are conflicting^{Footnote 4Footnote 5Footnote 6Footnote 7Footnote 8}. Mycoplasma genitalium co-infection with other bacterial sexually transmitted infections (STIs) has been frequently reported in gbMSM^{Footnote 7Footnote 9}.

Mycoplasma genitalium infection is not a notifiable condition in Canada^{Footnote 10Footnote 11} yet there are no published Canadian community-based studies concerning M. genitalium infection. Studies conducted in 2013 (Ontario), 2016 (Alberta) and 2019 (Saskatchewan), among men and women who had STI symptoms or sought medical attention for STI screening, have shown high rates of M. genitalium infection and macrolide resistance associated mutations (MRAM) and a significant presence of quinolone resistance associated mutations (QRAM)^{Footnote 12Footnote 13Footnote 14}.

More detailed Canadian data are required to guide testing and treatment of M. genitalium infections in gbMSM. The objectives of this study are to 1) estimate the prevalence of M. genitalium infection and other selected bacterial STIs by anatomical site among Montréal gbMSM, 2) explore correlates of M. genitalium infection and 3) estimate the prevalence of MRAM and QRAM.

Methods

Engage Cohort Study

Engage Cohort Study is a collaboration between researchers and community-based organizations to study the sexual health, including human immunodeficiency virus (HIV) and sexually transmitted and blood-borne infections (STBBIs), of gbMSM in Montréal, Toronto and Vancouver. Details for this cohort study were described elsewhere ^{Footnote 15Footnote 16Footnote 17}. In brief, participants were recruited using respondent-driven sampling (RDS), a survey method for sampling hard-to-reach populations deriving from chain referral sampling ^{Footnote 18}. Thus, enrolled participants recruited other eligible participants through their social networks. Eligibility criteria were as follows: French or English-speaking cisgender or transgender men; 16 years of age or older; and reporting at least one sexual encounter with a man in the prior six months. After recruitment, participants were invited every 6–12 months for subsequent visits at the community study site. At each visit, participants completed a self-administered computer-assisted questionnaire and provided biological samples, including first-pass urine, a pharyngeal and a rectal swab and a blood sample.

Sub-study in Montréal

Montréal recruitment into the Engage Cohort Study started in February 2017. For this one-time point sub-study, participants with a follow-up visit that occurred between November 2018 and November 2019 were included.

Biological specimen collection and laboratory testing

To detect Neisseria gonorrhoeae and Chlamydia trachomatis, nucleic acid amplification tests were used (cobas^® 4800; Roche Diagnostics, Branchburg, New Jersey). For M. genitalium detection, samples were kept at room temperature in the cobas^® PCR Media (Roche Diagnostics) for a maximum of one year or as frozen eluates. Specimens were analyzed using the Allplex^TM CT/NG/MG/TV assay (Seegene Inc.). Mycoplasma genitalium-positive samples were subsequently analyzed by real-time PCR to detect MRAM and QRAM by using the Allplex^TM MG & AziR and Allplex^TM MG & MoxiR assays, respectively.

Outcomes and correlates

Using current knowledge based on existing literature, variables were selected from the Engage Cohort Study questionnaire ^{Footnote 19Footnote 20}. Variables were grouped into the following categories: sociodemographic; sexual partners in the past six months (P6M); methods of finding sexual partners in the P6M; substance use in the P6M; and STBBIs in the P6M. The variable "chemsex" was defined as crystal methamphetamine, gamma-hydroxybutyrate (GBH), ecstasy/3,4-methylenedioxymethamphetamine (MDMA), ketamine, or poppers (i.e. alkyl nitrites) consumption in the two hours before or during sex with at least one of the last five sexual partners in the P6M. The variable "self-reported STI diagnosis" refers to a diagnosis by a healthcare professional in the P6M of C. trachomatis, N. gonorrhoeae, lymphogranuloma venereum (LGV) or syphilis. An individual was considered to have an M. genitalium infection if either their urine or their rectal sample was positive. Key mutations associated with azithromycin resistance (positions 2058 or 2059 in region V of the 23S ribosomal ribonucleic acid gene) and moxifloxacin resistance (S83I, S83R, S83N, D87N, or D87Y in parC) were used to define MRAM and QRAM, respectively.

Statistical analyses

Prevalence and odds ratios (OR) were estimated and adjusted for the recruitment method as well as censoring, using a combination of RDS-II weights ^{Footnote 21} and inverse-probability-of-censoring weights ^{Footnote 22}. The RDS-II weights are inversely proportional to the participants' network size, meaning that data for individuals with large networks were weighted less. The 95% confidence intervals (CI) were calculated using robust (sandwich) variance estimation to account for the within-subject correlation induced by weighing ^{Footnote 23}. Prevalence data was not adjusted MRAM and QRAM since one individual with a larger weight could easily dominate the subsample within small subsamples (each MRAM and QRAM subsample had n fewer than 100 positive specimens). Logistic regression was used to predict M. genitalium infection among gbMSM. Since the aim was prediction, there was no need to consider confounding or effect modification. Predictive performance was assessed using Akaike information criterion (AIC).

Ethics

Ethics approval was received from the Research Institute of the McGill University Health Centre.

Results

Between February 2017 and June 2018, 1,179 participants were recruited in Montréal. A follow-up study visit, during which samples were collected for M. genitalium testing, occurred for 717 participants. One participant was excluded from M. genitalium prevalence analyses because only a pharyngeal sample was provided. Overall, 716 participants provided a total of 2,064 samples (Figure 1).

Figure 1 - Text description

Figure 1 is a basic flow diagram that identifies how participants were selected for the purpose of this study. 1,179 individuals were recruited at baseline in the Engage Cohort Study in Montréal. The 717 participants had a follow-up visit between November 2018 and November 2019 during which samples were collected for Mycoplasma genitalium testing. One participant was excluded from M. genitalium prevalence analysis because only a pharyngeal sample was provided. All of the 716 participants who provided at least one rectal or one urine sample were included in the analysis from which a total of 2,064 samples were collected.

Most participants identified their ethnocultural identity as French or English Canadian (53.5%) and their sexual orientation as gay (82.2%). The majority reported having an education level higher than high school level (79.1%), a gross annual income of $30,000 or less (60.1%), being HIV-negative (84.9%) and having five or fewer male sexual partners in the P6M (67.6%) (Table 1).

Prevalence of Mycoplasma genitalium infection and other sexually transmitted infections

Mycoplasma genitalium prevalence was 5.7% (95% CI: 4.0–8.1) (rectal or urethral site) with anatomical site-specific prevalence being 4.0% (95% CI: 2.6–6.0) at the rectal site and 2.2% (95% CI: 1.2–4.0) at the urethral site (Table 2). The overall prevalence of M. genitalium was detected at the pharyngeal site in only two individuals (0.2%, 95% CI: 0.1–0.9). Prevalences of C. trachomatis and N. gonorrhoeae are detailed in Table 2. Among the individuals with urethral C. trachomatis infection, one of five were co-infected with M. genitalium (20%); among those with rectal C. trachomatis infection, two of 22 were co-infected with M. genitalium (9.1%) (Table 3). Among those with rectal N. gonorrhoeae infection, two of 12 were co-infected with M. genitalium (16.7%); no urethral N. gonorrhoeae infection was observed.

Mycoplasma genitalium infection correlates

Younger age (29 years or younger) and the following factors (all reported in the past six months) were significantly associated in univariate analysis having more male sexual partners (6–10 partners and 11 or more partners compared to five or fewer); having at least one new sexual partner; reporting at least one condomless anal sex act (insertive or receptive) with another man; engaging in chemsex; and having received a diagnosis of an STI (Table 4). Living with HIV was not associated with M. genitalium infection. The best predictive regression model of M. genitalium infection included the following factors: younger age (29 years or younger) (OR: 2.5, 95% CI: 1.2–5.5); and declaring more male sexual partners P6M (6–10 partners and 11 or more partners) (respective OR: 3.3, 95% CI: 1.3–8.5, and OR: 5.7, 95% CI: 2.3–14.1) (Table 5).

Antimicrobial resistance of Mycoplasma genitalium

For the three participants who were infected at both the urethral and rectal sites, the results obtained from the urethral site were used to calculate the prevalence of antimicrobial resistance (AMR). Prevalence of MRAM was 82% (n=46/56) and prevalence of QRAM was 29% (n=16/55) (Table 6). Prevalence of either MRAM or QRAM was 85% (n=46/54), while prevalence of both MRAM and QRAM was 28% (n=15/54).

Discussion

This first Canadian community-based study estimates the prevalence of M. genitalium infection at 5.7% (urethral or rectal infection) among gbMSM. It is challenging to contextualize our data since population-based prevalence studies are lacking. Compared to Canadian STI clinic-based studies, the urethral M. genitalium prevalence in our study (2.2%) was lower than previous estimates among men in Ontario (4.5%, 2013), Alberta (5.3%, 2016) and Saskatchewan (6.2%, 2019) ^{Footnote 12Footnote 13Footnote 14}. In Australia, urethral M. genitalium prevalence among men who have sex with men (MSM) recruited in STI clinics ranged from 2.7–4.7% and prevalence of rectal infections (7.0%–8.9%) was higher than in our study (4.0%) ^{Footnote 24Footnote 25}. Consistent with our results (n=3/689; 0.4%), a very low number of pharyngeal M. genitalium infections among MSM were reported in Australia (n=0/508 to n=8/464; 2.0%) ^{Footnote 9Footnote 25}. We hypothesize that oral transmission is negligible, and we excluded M. genitalium-positive pharyngeal samples from our prevalence estimation. Rectal M. genitalium infection (4.0%) was more common than rectal C. trachomatis (2.6%) and N. gonorrhoeae (1.4%) infections. Urethral M. genitalium prevalence estimates were more similar to those of C. trachomatis infection (M. genitalium, 2.2%; C. trachomatis, 1,9%; N. gonorrhoeae, 0%). A United States cohort study conducted in 2018–2019 among young gbMSM and transgender women found that M. genitalium was more prevalent than other STIs in both rectal (M. genitalium, 21.7%; C. trachomatis, 8.8%; N. gonorrhoeae, 6.8%) and urine samples (M. genitalium, 8.9%; C. trachomatis, 1.6%; N. gonorrhoeae, 0.8%) ^{Footnote 26}. A 2017–2018 Australian study found that among asymptomatic MSM, C. trachomatis prevalence was comparable to M. genitalium in rectal samples (M. genitalium, 7.0%; C. trachomatis, 8.5%; N. gonorrhoeae, 6.2%) and urine samples (M. genitalium, 2.7%; C. trachomatis, 1.7%; N. gonorrhoeae, 0.7%). It also found that 9.2% of MSM that tested positive for rectal C. trachomatis were co-infected with M. genitalium while 6.1% of positive rectal N. gonorrhoeae samples demonstrated co-infection with M. genitalium ^{Footnote 24}. In our study, 9.1% of gbMSM that tested positive for C. trachomatis at the rectum were co-infected with M. genitalium, 16.7% of rectal N. gonorrhoeae infections showcased M. genitalium co-infection.

In univariate analyses, multiple risk factors for STI transmission, such as chemsex P6M, new sexual partners P6M and a STI diagnosis P6M, were identified. Younger age and having multiple male sexual partners were retained in our predictive model. These findings are consistent with studies that identified younger age^{Footnote 24 Footnote 27 Footnote 28} and multiple sexual partners ^{Footnote 19Footnote 20Footnote 29 Footnote 30} as correlates of M. genitalium infection. While a United Kingdom study documented a higher prevalence of M. genitalium among gbMSM living with HIV ^{Footnote 31}, HIV infection was not associated with M. genitalium infection in our study. More studies are needed to clarify the role of M. genitalium in HIV acquisition or transmission among gbMSM as it has been identified as a risk factor of HIV infection, especially in MSM ^{Footnote 32 Footnote 33}.

The very high prevalence of MRAM (82%; n=46/56) and QRAM (29%; n=16/55) found among the Engage Cohort Study Montréal's gbMSM is a worrisome finding. This prevalence is higher than previous Canadian MRAM estimates (men in Alberta in 2016, 64%; women and men in Saskatchewan in 2019, 63% and men in Ontario in 2013, 63%) ^{Footnote 12Footnote 13Footnote 14}. Treatment failure with azithromycin has been well described with single nucleotide polymorphisms at positions 2058 and 2059 in region V of the 23S ribosomal ribonucleic acid ^{Footnote 34}. For QRAM, S83 in the parC gene is significantly associated with moxifloxacin resistance ^{Footnote 34}. While several single nucleotide polymorphisms contribute to quinolone resistance, none are as strong predictors of treatment failure than macrolide resistance with 23S ribosomal ribonucleic acid single nucleotide polymorphisms ^{Footnote 34Footnote 35}. Previous Canadian studies found a QRAM prevalence of 11%–20% among men and women ^{Footnote 12Footnote 13Footnote 14}. A meta-analysis compiling studies from 2010–2019 estimated MRAM and QRAM prevalence at 52% and 10%, respectively, in the Americas region ^{Footnote 2}. A 2017–2018 United States clinic-based study among men with urethritis found MRAM and parC QRAM prevalence levels of 64% and 12%, respectively ^{Footnote 28}. Being infected with a macrolide-resistant M. genitalium is more likely in gbMSM than in women and men with female partners only ^{Footnote 1 Footnote 36 Footnote 37}. This could be explained by transmission in closely-knit sexual networks and increased exposure to antibiotics ^{Footnote 37}. The increasing azithromycin resistance could be explained by its widespread use for the treatment of certain STIs ^{Footnote 2Footnote 7Footnote 38 Footnote 39 Footnote 40}. In our study, 28% of M. genitalium-positive samples had both MRAM and QRAM. Dual resistance has already been reported in gbMSM on HIV PrEP and those living with HIV ^{Footnote 36 Footnote 41}.

Implications for research and practice

In our study, we identified a high prevalence of M. genitalium infections among gbMSM, especially among younger individuals and those reporting multiple male sexual partners. Although most current guidelines state that routine screening for M. genitalium infection is not recommended (as it would contribute to selection pressure of resistant strains), they vary in terms of testing indications and timing in symptomatic individuals^{Footnote 42Footnote 43Footnote 44}: at the time of initial presentation of urethritis (concomitantly with N. gonorrhoeae and C. trachomatis testing)^{Footnote 42Footnote 43Footnote 44}, only for recurrent non-gonococcal urethritis ^{Footnote 4} or only for non-chlamydial non-gonococcal persistent or recurrent urethritis, following empiric treatment for N. gonorrhoeae and C. trachomatis and when pretreatment nucleic acid amplification tests or follow-up test of cure are negative for C. trachomatis and N. gonorrhoeae^{Footnote 45}. Regarding rectal screening, some clearly state it is not recommended ^{Footnote 4} or do not mention extra genital testing ^{Footnote 46}. The high prevalence of M. genitalium infection among gbMSM with C. trachomatis or N. gonorrhoeae infection demonstrates the need for clinicians to remain highly vigilant of a possible co-infection in the case of persistent symptoms after adequate treatment. Our findings of 4.0% prevalence of rectal M. genitalium among gbMSM in Montréal, being almost two-fold the prevalence of urethral M. genitalium infection (2.2%), and much higher than N. gonorrhoeae rectal infection (1.4%) or C. trachomatis infection (2.6%), may add to epidemiologic evidence in the process of updating the Canadian guidelines ^{Footnote 45}. Finally, the most recent guidelines touching upon the management of M. genitalium infection recommend AMR-guided therapy ^{Footnote 4Footnote 42Footnote 44}. This approach has demonstrated potential in reducing treatment failures^{Footnote 47Footnote 48}. Based on the identified susceptibility profile, doxycycline is used as initial empiric treatment and is followed by either azithromycin or moxifloxacin ^{Footnote 49}. Because of limited availability of tests in Canada and according to the current Canadian guidelines, treatment initiation for M. genitalium should occur in the context of syndromic management of persistent or recurrent urethritis ^{Footnote 10}. Recommended treatment consists of azithromycin and moxifloxacin as first and second lines of treatment ^{Footnote 45}. The high AMR observed in our study supports the need for M. genitalium detection and AMR testing in a short turnaround time ^{Footnote 42Footnote 44Footnote 47}. It also highlights the need, when both QRAM and MRAM are detected, for an easier and quicker access to alternative treatments such as pristinamycin, which can currently be requested through the Health Canada's special access program^{Footnote 42Footnote 46Footnote 50}.

Limitations

The small sample size limited our ability to identify correlates of infection or AMR. Data regarding STI-related symptoms was not collected in the study questionnaire which was designed prior to the initiation of this sub-study and was focused on societal and community contexts, social relationships and sexual behaviour. Hence, we could not evaluate the prevalence of M. genitalium in association with clinical presentation. Despite using the RDS method for recruitment, some subgroups of the gbMSM population may be over- or under-represented. Potential biases related to RDS were attenuated by adhering to recommended recruitment procedures, having a large sample size with long recruitment chains and adjusting with RDS-II weights. The AMR data were not RDS-adjusted because they were obtained from too small a subsample. Our prevalence findings might not be generalizable to non-urban Canadian gbMSM populations. We did not find comparison studies analyzing the performance of the Allplex CT/NG/MG/TV Assay, which limited our appreciation of potential information biases. Le Roy et al. calculated an overall agreement of 94.6% between in-house real-time PCR and the Allplex MG & AziR Assay ^{Footnote 51}. The assay, however, showed low sensitivity for macrolide resistance compared to sequencing (sensitivity of 74.5%, specificity of 97.6%).

Conclusion

This first population-based study among Canadian gbMSM documented a high prevalence of urethral and rectal M. genitalium infection. The observed levels of AMR, which exceed the 5% threshold at which a change in empirical treatment is recommended by the World Health Organization, supports the need for AMR-guided therapy ^{Footnote 52}. Efforts should be made to facilitate targeted M. genitalium detection and AMR testing when indicated.

Authors' statement

ASL — Methodology, data interpretation, visualization, writing–drafting
ACL — Conceptualization, methodology, data interpretation, visualization, writing–review and editing, supervision
AF — Methodology, data curation, formal analysis, data interpretation, writing–review and editing
MD — Methodology, formal analysis, writing–review and editing
JC — Manuscript review and editing, project administration, funding acquisition
CF — Conceptualization, investigation, writing–review and editing
IM — Writing–review and editing
DG — Writing–review and editing
TAH — Writing–review and editing
DMM — Writing–review and editing
GL — Conceptualization, methodology, investigation, formal analysis, data interpretation, writing–review and editing, project administration, funding acquisition, supervision

All authors revised the manuscript critically for important intellectual content, approved the final version to be published and agreed to be accountable for all aspects of the work.

Competing interests

J Cox and G Lambert report non-financial support from the Direction régionale de santé publique, Centre intégré universitaire de santé et de services sociaux du Centre-Sud-de-l'Île-de-Montréal. J Cox reports grants and personal fees from ViiV Healthcare and Gilead Sciences Canada, and personal fees from Merck Canada, outside the submitted work. DM Moore reports a grant from the Michael Smith Foundation for Health Research. No other competing interests were declared.

Acknowledgements

Data from this work comes from the Engage Cohort Study. The principal investigators of the Engage Cohort Study are J Cox and G Lambert (Montréal), J Jollimore, NJ Lachowsky and DM Moore (Vancouver), and D Grace and TA Hart (Toronto). The authors thank the Engage Cohort Study participants, office staff, and community engagement committee members, as well as their community partner agencies.

Funding

The Engage Cohort Study is funded by grants TE2-138299, FDN-143342, and PJT-153139 from the Canadian Institutes of Health Research (CIHR), grant CTN300 from the CIHR Canadian HIV/AIDS Trials Network, the Canadian Foundation for AIDS Research, grant 1051 from the Ontario HIV Treatment Network (OHTN), grant 4500370314 from the Public Health Agency of Canada and the Ministère de la Santé et des Services sociaux du Québec. DM Moore is supported with Scholar Awards from the Michael Smith Foundation for Health Research (#5209). T Hart is supported by a Chair in Gay and Bisexual Men's Health from the OHTN. D Grace is supported by a Canada Research Chair in Sexual and Gender Minority Health. The funders had no role in the study design, data collection, and analysis, decision to publish, or preparation of the manuscript.