High prevalence of circulating dual-class resistant Mycoplasma genitalium in asymptomatic MSM in Tokyo, Japan

Abstract

Objectives

To assess the prevalence and antibiotic resistance profile of Mycoplasma genitalium detected from urogenital/rectal swab samples obtained from MSM in Tokyo, Japan.

Methods

We performed PCR-based screening for M. genitalium urogenital/rectal infection in 982 asymptomatic MSM between 1 January 2019 and 5 November 2020. Mutations in the antibiotic resistance-associated genes gyrA and parC and the 23S rRNA of M. genitalium were analysed.

Results

The prevalence of M. genitalium infection was 6.1%: the prevalence of rectal and urogenital infection was 4.7% and 1.4%, respectively. Among the cases, 48 were successfully analysed for 23S rRNA, 41 for parC mutations and 37 for gyrA mutations. Macrolide- and quinolone-resistance associated mutations (23S rRNA and parC mutations) were observed in 43 (89.6%) and 28 (68.3%) cases, respectively. The quinolone-resistance associated mutation-harbouring variants also harboured macrolide-resistance associated mutations. The S83I mutation in the parC gene was most commonly identified (24 cases, 58.5%), and its combination with M95I or D99N mutation in the gyrA gene was observed in 9 of 36 successfully analysed cases (25.0%). No significant association was observed between the presence of antibiotic resistance and antibiotic exposure for either macrolides or fluoroquinolones (P = 0.785 and 0.402, respectively).

Conclusions

In Tokyo, there is an alarmingly high prevalence of M. genitalium harbouring macrolide and/or quinolone resistance-associated mutations in MSM, irrespective of antibiotic exposure. The high prevalence of M. genitalium strains with both parC and gyrA mutations limits the efficacy of sitafloxacin. Therefore, suitable alternatives are required to treat such M. genitalium infections.

Introduction

Mycoplasma genitalium causes a sexually transmitted infection (STI). This pathogen was initially identified in a case of non-gonococcal urethritis and it is associated with other urogenital conditions in both men and women.^1,2 Frequent reports of asymptomatic M. genitalium infections^1,3 have led to speculations that the rectum may be an important reservoir of M. genitalium.³ However, routine screening for M. genitalium is not recommended in asymptomatic individuals, because screening has not been shown to prevent the infection or related complications.⁴ Moreover, empirical antibiotic treatment for M. genitalium infection or concurrent STIs, like those caused by Chlamydia trachomatis, raises concerns about the spread of M. genitalium strains harbouring macrolide and quinolone resistance-associated mutations.^3,5 Azithromycin failure is associated with mutations at positions 2071 or 2072 of the 23S rRNA gene;⁶ fluoroquinolone failure, especially that of moxifloxacin, is associated with the S83I, S83R, D87N or D87Y mutations in parC and might be associated with the M95I or D99N mutations in gyrA.^7,8 Combined mutations in the parC and gyrA genes, such as S83I plus D99N, G93C or M95I, are associated with increased MIC values for quinolones⁹ and S83I combined with M95I or D99 are more strongly associated with quinolone failure, compared with the S83I mono-mutations.¹⁰ In fact, multidrug resistance in M. genitalium has become a global threat. A recent meta-analysis by Machalek et al.¹¹ reported that the rate of macrolide resistance-associated mutations (MRM; in the 23S rRNA gene) in M. genitalium was approximately 10.0%, while that of quinolone resistance-associated mutations (QRM; in parC) was 4.8% before 2010; these rates increased to 51.4% and 9.3% for macrolide and quinolone resistance, respectively, in 2016–17.¹¹ The prevalence of MRM and QRM is higher in the Western Pacific region, including Australia and Japan, compared with that in European countries.¹¹ A more recent study from Belgium on samples collected in 2015–18 showed higher frequencies of 74.3% and 39.7% for MRM and QRM, respectively; 17.8% of the samples harboured S83I mutations.¹² However, in individuals with asymptomatic rectal infections, the prevalence of antibiotic resistance-associated mutations remains unclear,³ and rectal infections, rather than urethral infections, may help to better detect circulating strains due to higher prevalence.³ Here, we report the current epidemiology of M. genitalium rectal and urethral infections among MSM in Tokyo, Japan, an endemic area of MDR M. genitalium.

Materials and methods

Study population

We conducted two prospective cohort studies at two outpatient clinics at the AIDS Clinical Center (ACC) and the Sexual Health Clinic (SHC), National Centre for Global Health and Medicine (NCGM), Tokyo, Japan, between 1 January 2019 and 5 November 2020. MSM who regularly visited the ACC (age >20 years) or the SHC (age >16 years) were eligible to participate. The ACC cohort comprised HIV-positive MSM, while the SHC cohort comprised HIV-negative but high-risk MSM who regularly tested for HIV and other STIs. All participants provided written informed consent and completed a questionnaire regarding their recent sexual behaviour. Data on participant demographics were obtained from the medical records. This study was approved by the Human Research Ethics Committee of NCGM (NCGM-G-003350-00; NCGM-G-002091-00) and conducted according to the principles expressed in the Declaration of Helsinki of 1964 and the later amendments.

Detection of M. genitalium and macrolide and quinolone resistance

Ten millilitres of first void urine and/or a rectal swab were collected from each participant and screened for M. genitalium using a commercial PCR assay kit (LSI Medience Co., Tokyo, Japan).¹³M. genitalium-positive samples were used for resistance detection. Resistance-associated mutations were detected by targeted amplification of the 23S rRNA V domain and of the relevant regions of the gyrA and parC genes, as previously described,¹⁰ followed by Sanger sequencing, performed using the ABI PRISM BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Japan) on the Applied Biosystems 3500 Genetic Analyzer (Thermo Fisher Scientific). The sequences were compared with the corresponding target regions of the M. genitalium G37 complete genome (GenBank: L43967.2) to detect and identify mutations.

Statistical analysis

The prevalence of M. genitalium infection was estimated from the number of infected individuals among the study population. Statistical analyses were conducted using STATA V.16 (StataCorp, College Station, TX, USA) and GraphPad Prism V9 (GraphPad Software, CA, USA). The χ² or Mann–Whitney U tests and the Kruskal–Wallis test were used for the categorical and continuous variables, respectively; P < 0.05 was considered statistically significant.

Results

Participants’ demographic and clinical characteristics

A total of 982 MSM were analysed. The median age was 39 (31–47) years. Among them, 49.2% (483/982) and 50.8% (499/982) were HIV-positive and -negative, respectively. All study participants were asymptomatic, as per the questionnaires or direct examination. The median CD4+ T cell count in patients with HIV (PLWH) was 585.5 (452.5–758) cells/mm³; 94.4% of PLWH showed well-controlled viral replication (HIV-1 RNA < 200 copies/mL).

Prevalence of M. genitalium infection

The prevalence of M. genitalium infection among the study participants was 6.1% (60/982). M. genitalium was significantly more prevalent in rectal samples compared with that in urogenital samples (4.7% [46/977] versus 1.4% [14/980], P < 0.0001; Figure 1). No cases of urogenital and anorectal coinfection were observed. The difference in the prevalence of M. genitalium infection between HIV-positive (7.5% [36/483]) and -negative MSM (4.8% [24/499]) was not statistically significant (P = 0.083). The prevalence of rectal and urogenital M. genitalium infections in PLWH was 6.0% (29/482) and 1.4% (7/483), respectively, while that in HIV-negative participants was 3.4% (17/495) and 1.4% (7/497), respectively; the differences were not significant (P = 0.057 and 0.957, respectively).

Figure 1.

(a) Asymptomatic M. genitalium infection rate; ****P < 0.0001. (b) The rate of mutations associated with macrolide and/or fluoroquinolone resistance (MRM or QRM, respectively), defined as parC (excluding D82N) or gyrA mutations. MG, Mycoplasma genitalium.

Resistance

Among the 60 isolates, 48 were successfully analysed for MRM, 41 for parC mutations, and 37 for gyrA mutations (Figure 1). MRM, including A2071G, A2071T, or A2072G, were detected in 89.6% (43/48) of samples. MRM were detected in 89.5% (34/38) and 90% (9/10) of rectal and urogenital samples, respectively. Regarding QRM, 68.3% (28/41) isolates harboured parC mutations: 63.6% (21/33) and 87.5% (7/8) in rectal and urogenital samples, respectively (Table S1, available as Supplementary data at JAC-AMR Online). Most of them (58.5% [24/41]) harboured the S83I mutation. In contrast, gyrA mutations were detected in 27.0% (10/37) of the isolates: 25% (7/28) and 33.3% (3/9) in rectal and urogenital samples, respectively. The S83I in the parC combined with the M95I or D99N mutations in the gyrA were detected in 25.0% (9/36) of the successfully analysed cases. Among the study participants, 28.3% (17/60) with M. genitalium infections had a history of azithromycin therapy, while 6.7% (4/60) had undergone fluoroquinolone therapy in the 90 days before enrolment. However, no significant associations were observed between antibiotic exposure and resistance to either of these antibiotics (P = 0.785 and 0.402, respectively).

Discussion

We found a high prevalence of M. genitalium carrying MRM (89.6%), QRM (68.3%) and dual-class resistance-associated mutations (68.3%) in asymptomatic MSM in Tokyo, irrespective of antibiotic exposure.

Almost all M. genitalium carried MRM, and 68.3% of the detected parC mutations, including S83I, S83R and D87N, limit moxifloxacin-based therapy.^7–10 The use of sitafloxacin, an alternative antibiotic for the treatment of resistant M. genitalium strains carrying parC mono-mutations, is approved in some countries, including Japan.^9,10,14 However, sitafloxacin-based treatment may not be as effective against the strains co-harbouring the parC and gyrA mutations reported here (25.0%). To the best of our knowledge, this is the highest ever reported prevalence of strains with mutations associated with dual-class resistance. One study has reported a prevalence of 6.1% of dual-class resistance,¹⁵ whereas other studies have reported a prevalence of 7.4%–10.4% in the context of triple mutations in 23S rRNA, parC and gyrA.^5,16 Another alternative is to focus on quinolone resistance-based treatment strategies; however, the availability of commercial diagnostic tools to investigate quinolone resistance-associated mutations is a major challenge.¹⁷ Importantly, both parC and gyrA mutations should be monitored as potential markers of fluoroquinolone treatment failure.

Doxycycline and sitafloxacin combination therapy and pristinamycin exhibit high efficacy in the treatment of resistant M. genitalium.^14,18 Therefore, further research on combination therapy using doxycycline, which is a low-cost antibiotic, and new potential drugs such as zoliflodacin, lefamulin and gepotidacin, which are effective against resistant M. genitalium in vitro,^19–21 is essential for establishing effective alternate treatments for M. genitalium.

Here, we report M. genitalium infection rates of 4.7% and 1.7% in the rectum and urethra of MSM in Japan. These results are in line with reports from Australia^3,22 and Japan,²³ with rectal and urethral infection rates of 7.0%–8.5% and 2.7%–4.0%, respectively. The high prevalence of M. genitalium in asymptomatic MSM necessitates considering routine screening in high-risk MSM. However, there are no consistently effective treatments against resistant strains of M. genitalium. Unlike in females, in whom pelvic inflammatory diseases and infertility are linked to infections, there is no sufficient evidence of sequelae to justify routine screening in MSM. Moreover, evidence that macrolides induce selective pressure for resistance-associated strains is limited. Therefore, at present, routine screening of M. genitalium in asymptomatic MSM is not fully justified.⁴

Urogenital and anorectal M. genitalium infections are more prevalent in HIV patients, compared with those in non-HIV patients;^24,25 this was not observed in this study, probably because of the small sample size in this study. Additionally, in contrast to the results of a previous study on symptomatic patients from China,⁵ no association between resistant strains and antibiotic exposure in asymptomatic patients was observed in this study. Our results imply that MDR M. genitalium strains are already in circulation in Japan. Therefore, there is an urgent need to optimize antibiotic use based on antimicrobial stewardship programmes to prevent the spread of MDR M. genitalium strains.

This study is not without limitations. Not all samples were successfully sequenced; therefore, the reported rates may be slightly different from the actual ones.

In conclusion, we report an alarmingly high prevalence of dual-class resistant M. genitalium in asymptomatic MSM from Tokyo. Moxifloxacin-based therapy may not be sufficient in this context. Therefore, alternative approaches are needed to treat infections caused by MDR M. genitalium.