Abstract
Objective
Assess risk factors for incident and endometrial Mycoplasma genitalium infection and determine if M. genitalium is associated with histological endometritis, an indicator of pelvic inflammatory disease.
Methods
This study was a secondary data analysis within the T cell Response Against Chlamydia (TRAC) Study, a prospective evaluation of 246 women withor at risk for Chlamydia trachomatis from urban outpatient clinics, who were followed quarterly for12 months. Risk factors for incident M. genitalium infection were determined by Cox regression. Relative risks were estimated by Poisson regression withrobust error measurements for models examining the association between M. genitalium and endometritis (histological evidence of plasma cells in endometrial stroma and neutrophils in the endometrial epithelium) and for models examining risk factors for detection of endometrial M. genitalium infection.
Results
M. genitalium prevalence was 16.7%, incidence was 25.3 per 100 person-years and 23% had repeated positive tests. Black race (non-black HRadj 0.4, 95% CI 0.2 to 0.9), less education (HRadj 2.4, 95% CI 1.2 to 5.1) and a new sexual partner (HRadj 3.1, 95% CI 1.7 to 5.4) were associated with incident M. genitalium. M. genitalium was associated with endometritis (RRadj 2.0, 95% CI 1.1 to 3.7). Trichomonas vaginalis (RRadj 2.0, 95% CI 1.2 to 3.4) and endometrial C. trachomatis (RRadj 1.7, 95% CI 1.1 to 2.8) were associated with endometrial M. genitalium.
Conclusions
M. genitalium is prevalent in women at high risk for C. trachomatis, persists over multiple follow-up visits and is associated with histological endometritis. Studies are needed to determine if screening for M. genitalium will improve reproductive outcomes.
INTRODUCTION
Mycoplasma genitalium is an emerging STI linked to reproductive morbidities.1 Urethral M. genitalium is a cause of non-gonococcal urethritis in men and 30% of persistent or recurrent male urethritis.2 However, the role of M. genitalium in female reproductive tract disease needs further research.
Inoculation of female mice3 and non-human primates4 with M. genitalium leads to endometritis and salpingitis. A recent meta-analysis reported that M. genitalium was associated with a twofold increased risk of clinical pelvic inflammatory disease (PID).1 However, coinfection with Chlamydia trachomatis and Neisseria gonorrhoeae is common1 complicating interpretation of the independent role of M. genitalium in PID. Of the 10 studies that examined PID in the meta-analysis, four adjusted for coinfection.5–8 Only one of four studies examined the endometrium for evidence of infection and inflammation.7 Thus, unanswered questions regarding the epidemiology and pathogenesis of M. genitalium in women limit formulation of clear guidelines for screening and treatment.9
From a public health standpoint, it is important to understand if M. genitalium causes upper genital tract pathology. M. genitalium is not susceptible to standard antibiotics for syndromic treatment of cervicitis and clinical PID.7,10 Macrolide resistance rates for M. genitalium are high (60%–80%).11,12 Thus, M. genitalium may persist in the genital tract after treatment and clearance of C. trachomatis and N. gonorrhoeae. We examined the association between M. genitalium and histologically confirmed endometritis controlling for other STIs. Additionally, we identified factors associated with endome-trial and incident M. genitalium.
MATERIALS AND METHODS
Patient population
This study was a secondary data analysis within the T cell Response Against Chlamydia (TRAC) Study. TRAC identified risk factors for C. trachomatis and evaluated serum antichlamydia antibodies in association with chlamydial bacterial burden and incident infection.13 The data from the current study evaluated risk factors for M. genitalium and have not been presented previously. TRAC was a longitudinal cohort study of 246 women aged 15–35 years with STIs or who were at risk for STIs. All 246 women were included in the current analysis. Women were enrolled in the study from three urban sites in Pittsburgh, Pennsylvania, during 2011–2015. Eligibility criteria included mucopurulent cervicitis, diagnosis of chlamydial infection prior to treatment or reported sexual contact with an individual with a recent diagnosis of chlamydial urethritis or non-gonococcal urethritis. Women with PID according to the Centers for Disease Control and Prevention guidelines9 were excluded. All study participants provided informed written consent for infection testing and genetic analysis at enrolment and for follow-up visits. The University of Pittsburgh Institutional Review Board (MOD10010159–12/PRO10010159) and the University of North Carolina Institutional Review Board (13–3074) approved the study.
Data collection
At enrolment, study personnel obtained demographic and medical history data, participants completed a self-reported questionnaire and cervical, vaginal and endometrial swab samples were obtained.13 Gram staining was used for bacterial vaginosis (BV) scoring by Amsel criteria.14 Molecular testing for Trichomonas vaginalis (APTIMA TV, Gen-Probe, San Diego, California, USA). Nucleic acid amplification tests (NAAT) detected C. trachomatis, N. gonorrhoeae (APTIMA Combo 2, Gen-Probe) and M. genitalium (APTIMA MG, Gen-Probe).
Endometrial biopsies from study participants were obtained as follows: after cleansing the cervix with betadine, a sterile sheath-enveloped endometrial sampler (Unimar Pipelle de Cornier, CooperSurgical, Shelton, Connecticut, USA) was placed into the endometrial cavity, a tissue sample was aspirated into the cannula and the cannula was pulled back into the sheath prior to removal. Tissue was discharged into a sterile petri dish. Tissue proximal to the sampling portal of the cannula was placed into 10% formalin fixative for histology, and distal tissue was swabbed for molecular testing. Endometritis was determined by two pathologists who independently scored H&E stained sections of endometrial biopsy tissue according to published Kiviat criteria15 (five or more neutrophils per ×400 field in surface epithelium and/or one or more plasma cells per ×120 field in endometrial stroma). Discrepant evaluations were resolved after rereview and discussion.
A standardised self-reported questionnaire was administered at 1, 4, 8 and 12 months, addressing the participant’s interim medical and sexual history. Vaginal fluid and vaginal and cervical swabs were collected at each follow-up visit for STI and BV diagnosis.
All participants at enrolment received single-dose agents for gonorrhoea (ceftriaxone, 250 mg intramuscularly) and chlamydia (azithromycin, 1 g orally). At enrolment, women diagnosed with BV and/or T. vaginalis received metronidazole, 2 g orally. No treatment was administered to women with M. genitalium infection alone, since none of the women in the cohort had clinical PID.9
Among women positive for M. genitalium by Aptima testing, DNA was extracted using a Quick-DNA Universal Kit (Zymo Research, Irvine, California, USA) from 100 μL to 300 μL of reserved cervical swab eluates to quantify M. genitalium burden, strain typing and detection of 23S rRNA alleles. The abundance of M. genitalium DNA was determined by quantitative PCR with SsoAdvanced SYBR mix (Bio-Rad) using primers directed against the highly conserved MG190 locus16 using a CFX iCycler (Bio-Rad) in duplicate. Primer design (M. genitalium 190 Forward: GAGA TCAG GATA GAAC TGAG GAGTA, M. genitalium 190 Reverse: TGAT CGCT CCAC TTGC ATTAT) was based on the genome sequence of M. genitalium strain G37, GenBank Accession #NC_000908.2. Nested PCR amplified a fragment of mgpB for genotyping and region V of the 23S rRNA locus where macrolide resistance mutations accumulate.17 PCR cycling parameters were: one cycle of 98°C for 3 min, 38 cycles of 98°C for 30 s, 58°C for 30 s and 72°C for 30 s, followed by one cycle of 72°C for 3 min using Phusion Flash High-Fidelity PCR master mix (Thermo Scientific, Waltham, Massachusetts, USA). The resultant PCR products were cleaned and concentrated using a ‘DNA Clean and Concentrator-5’ kit (Zymo Research) prior to sequencing. Sequencing results were analysed using Clone Manager (Scientific & Educational Software, Denver, Colorado, USA) and Geneious (Auckland, New Zealand).
Statistical analyses
Incident M. genitalium infection was any positive M. genitalium test during follow-up in women who were negative for M. genitalium at their prior visit. The incidence rate was the number of incident infections per person-years at risk. Risk factors for incident infection were determined with Cox regression, with the Wei, Lin and Weissfeld method for within patient correlation,18 to calculate hazard ratios (HR) and 95% CIs.
Associations between baseline M. genitalium (positive at the cervix and/or endometrium) and endometritis were examined. Among M. genitalium-positive women with data on endometrial M. genitalium infection (39/41), we compared cervical burden between women with cervical M. genitalium infection only and women with cervical and endometrial M. genitalium infection. We also identified demographic (age, race, ethnicity, marital status, education and insurance), reproductive (pregnancies, sexual behaviour and history of STIs), contraceptive, behavioural and microbial factors associated with endometrial M. genitalium infection. Modified Poisson regression with robust error estimates was used to calculate relative risk (RR) and 95% CIs.19 A stepwise approach was used to build all multivariable models. Variables with P≤0.2 in univariable analyses were considered for multivariable analysis, whereas those with P≤0.1 were retained in the final model. All analyses were conducted using SAS V.9.2.
RESULTS
M. genitalium prevalence was 16.7% (41/246; 95% CI 13.1% to 20.9%), and 39/41 women were tested for M. genitalium endometrial infection with 48.7% testing positive (19/39). Among M. genitalium-positive women, 70.7% (29/41; 95% CI 58.1% to 80.8%) had C. trachomatis; 73.2% (30/41; 95% CI 60.6% to 82.9%) had BV; 26.8% (11/41; 95% CI 17.2% to 39.4%) had T. vaginalis; and 12.2% (5/41; 95% CI 6.1% to 23.0%) had N. gonorrhoeae. Women with M. genitalium had a mean age of 21±2.4 years, the majority were of black race (70.7%), single (90.2%) and on Medicaid (57.5%). Thirty per cent were uninsured, 41.4% had at least a high school education, 63.4% smoked and the majority denied drug use (65.7%). These characteristics represent TRAC overall, where the median age was 21, 89% were single, 66% were of black race, 36% had a high school education and 47% were on Medicaid.
Women with M. genitalium monoinfection were more likely to report pain during intercourse (5/10; 50%) compared with women with M. genitalium coinfection (6/30; 20%), women with chlamydia and/or gonorrhoea (15/109; 13.8%) and uninfected women (11/69; 15.9%), χ2 P=0.02). Urinary symptoms and abnormal vaginal discharge did not significantly differ between groups. There were no differences in reported symptoms between women with cervical M. genitalium only compared with those with M. genitalium endometrial infection.
Incidence of M. genitalium was 25.3 per 100 person-years (95% CI 15.3 to 28.4). Among these women, 43.6% had BV, 13.6% had T. vaginalis, 6.8% had C. trachomatis and none had N. gonorrhoeae. Incident M. genitalium monoinfection was accompanied by a similar frequency of urinary symptoms (12%) compared with M. genitalium coinfection (10.5%) and no infection (7%; P=0.45). Results were similar for urinary frequency (12%, 10.5% and 6%; P=0.30), abnormal vaginal discharge (8%, 5% and 8%; P=0.70) and vaginal bleeding (12%, 11% and 9%; P=0.77). Overall, 28% of M. genitalium monoinfections were accompanied by any symptom, compared with 31.6% for coinfection and 28% for women without STIs or BV.
In multivariable models, non-black race (HR 0.4, 95% CI 0.2 to 0.9), less than a high school education (HR 2.4, 95% CI 1.2 to 5.1) and reporting a new sexual partner at follow-up (HR 3.1, 95% CI 1.7 to 5.4) were associated with incident M. genitalium infection (table 1).
Table 1.
Factors associated with incident Mycoplasma genitalium infection
| Variable | HR (95%CI) | P values | HRadj.(95%CI)* | P values |
|---|---|---|---|---|
| Demographics and behaviour | ||||
| Age | 0.9 (0.7 to 0.9) | 0.04 | 0.9 (0.8 to 1.0)† | 0.05 |
| Non-black race | 0.4 (0.2 to 0.8) | 0.007 | 0.4 (0.2 to 0.9)‡ | 0.03 |
| Uninsured | Reference | Reference | ||
| Private insurance | 1.1 (0.5 to 2.6) | 0.80 | NA | NA |
| Medicaid | 1.4 (0.6 to 2.9) | 0.39 | NA | NA |
| Posthigh school education | Reference | Reference | ||
| Less than high school education | 2.7 (1.3 to 5.8) | 0.009 | 1.6 (0.8 to 5.1)‡ | 0.02 |
| High school education | 2.0 (1.0 to 3.7) | 0.05 | 1.6 (0.8 to 3.1)‡ | 0.16 |
| Married | 0.6 (0.2 to 1.7) | 0.31 | NA | NA |
| Non-smoker | 0.6 (0.3 to 1.0) | 0.06 | 0.7 (0.4 to 1.4)† | 0.38 |
| Drug use | 1.5 (0.9 to 2.7) | 0.15 | 1.0 (0.6 to 1.9)† | 0.90 |
| Alcohol use | 1.2 (0.7 to 2.0) | 0.62 | NA | NA |
| Douching past 30 days | 1.1 (0.4 to 3.3) | 0.81 | NA | NA |
| Reproductive history at enrolment, sexual behaviour and contraception use at follow-up | ||||
| History of Chlamydia trachomatis | 2.1 (1.1 to 3.9) | 0.03 | 1.7 (0.9 to 3.3)‡ | 0.09 |
| History of Neisseria gonorrhoeae | 1.4 (0.7 to 2.6) | 0.33 | NA | NA |
| History of Trichomonas vaginalis | 1.1 (0.6 to 2.1) | 0.71 | NA | NA |
| History of BV | 1.3 (0.7 to 2.2) | 0.40 | NA | NA |
| Infected partner | 1.5 (0.4 to 5.4) | 0.53 | NA | NA |
| Multiple partners | 2.4 (1.2 to 4.6) | 0.009 | 1.1 (0.5 to 2.6)† | 0.78 |
| New sexual partner | 2.9 (1.6 to 5.3) | 0.0003 | 3.1 (1.7 to 5.4)‡ | 0.0001 |
| Uncircumcised partner | 1.1 (0.3 to 4.2) | 0.93 | NA | NA |
| Oral sex | 1.7 (0.9 to 3.1) | 0.08 | 1.4 (0.7 to 2.6)† | 0.29 |
| Anal sex | 1.5 (0.5 to 4.6) | 0.51 | NA | NA |
| Consistent condom use | Reference | Reference | Reference | |
| Rarely or never use condoms | 1.6 (0.9 to 3.3) | 0.11 | 1.7 (0.9 to 3.1)† | NA |
| Inconsistent condom use | 1.1 (0.5 to 2.4) | 0.80 | NA | NA |
| Pills | 0.8 (0.4 to 1.9) | 0.65 | NA | NA |
| Depo-Provera injections | 0.9 (0.4 to 2.1) | 0.77 | NA | NA |
| Withdrawal | 1.4 (0.7 to 2.8) | 0.33 | NA | NA |
| None | 0.5 (0.2 to 1.6) | 0.23 | NA | NA |
| Intrauterine device | 1.1 (0.5 to 2.4) | 0.74 | NA | NA |
| Microbiology at enrolment and follow-up | ||||
| Baseline C. trachomatis | 1.5 (0.8 to 3.1) | 0.20 | 1.5 (0.8 to 2.8)† | 0.26 |
| Baseline N. gonorrhoeae | 1.8 (0.8 to 4.3) | 0.18 | 1.2 (0.5 to 2.9)† | 0.62 |
| Baseline M. genitalium | 1.9 (1.0 to 3.8) | 0.07 | 1.8 (0.8 to 3.7)† | 0.14 |
| Baseline T. vaginalis | 1.8 (1.0 to 3.4) | 0.05 | 1.5 (0.8 to 2.9)† | 0.22 |
| Baseline bacterial vaginosis | 1.3 (0.6 to 2.6) | 0.45 | NA | NA |
| C. trachomatis at follow-up | 0.8 (0.3 to 2.6) | 0.72 | NA | NA |
| T. vaginalis at follow-up | 1.3 (0.6 to 2.9) | 0.51 | NA | NA |
| Bacterial vaginosis at follow-up | 1.0 (0.7 to 1.4) | 0.99 | NA | NA |
P≤0.2 in univariable analyses were tested in the multivariable analysis, where variables with a P≤0.1 were retained.
Not retained in the final model.
Included in final model.
BV, bacterial vaginosis.
A total of 72 women tested positive for M. genitalium at least once during the study. Among these women, 31 (43%) had two or more positive M. genitalium tests and 22 (31%) tested positive for M. genitalium during two or more consecutive visits. This is higher than for C. trachomatis, where 23% (40/171) had two or more positive tests and 6% (11/171) tested positive during consecutive visits and for N. gonorrhoeae where 25% (7/28) had two or more positive tests and 14% (4/28) tested positive at consecutive visits. Results for T. vaginalis were similar to M. genitalium, 45% (29/65) had two or more positive tests and 34% (22/65) tested positive at consecutive visits.
We investigated azithromycin treatment and resolution of infection retrospectively. Of the 41 women with M. genitalium at enrolment, eight were lost to follow-up, 23 were M. genitalium negative by their first follow-up visit and 10 tested positive. We were unable to evaluate the genotype or 23S rRNA for 5 of 10 positive women. We identified mutant alleles associated with azithromycin resistance (2058G or 2059G) for three women. One additional individual acquired a new isolate, as their mgpB gene sequences amplified from enrolment and follow-up specimens differed. Another carried an identical azithromycin-sensitive wild-type 23S rRNA allele at enrolment and follow-up, consistent with reinfection. Sequence analysis of 9 of 10 samples from M. genitalium negative women at follow-up confirmed a wild type, drug-sensitive region V, while one specimen carried the 2058G mutation, indicating this patient’s immune response led to effective clearance or follow-up infection was missed.
A total of 135 participants had evaluable endometrial histo-logical data, and 46 were positive for endometritis. The frequency of missing endometritis data did not differ between M. genitalium positive (46.3%) and negative women (44.8%). A total of 15 M. genitalium positive women had endometritis where 10 had M. genitalium endometrial infection and 5 had cervical M. genitalium only. Thirty-one women with endometritis were M. genitalium negative, 23 were positive for BV, 18 had C. trachomatis, 7 had T. vaginalis and 2 had N. gonorrhoeae. M. genitalium positivity (cervical or endometrial) was associated with endometritis (RR 2.0, 95% CI 1.1 to 3.7; table 2) after adjustment for race, C. trachomatis, N. gonorrhoeae and BV. A total of 90% of women with M. genitalium endometrial infection had histological evidence of endometritis compared with only 50% of women with cervical infection only (RR 1.8, 95% CI 1.0 to 5.6). Among women with M. genitalium monoinfection, 3 of 5 had endometritis. Among those with M. genitalium who also had chlamydia or gonorrhoea, 70.6% had endometritis. Two burden distributions were found, 26 out of 40 women had high cervical burden (>104 genome copies per swab) and 12 had low cervical burden (<500 genome copies per swab). High M. genitalium cervical burden (>104 genome copies per swab) was not significantly associated with endometritis (RR 1.4, 95% CI 0.6 to 3.2).
Table 2.
Association between Mycoplasma genitalium and endometritis
| Endometritis* | Univariable models | Multivariable models† | ||||
|---|---|---|---|---|---|---|
| M. genitalium diagnostic test result | No n=89 |
Yes n=46 |
RR (95%CI) | P values | RR (95%CI) | P values |
| Results based on a positive M. genitalium test‡ | ||||||
| M. genitalium negative | 82 (92.1) | 31 (67.4) | Reference | Reference | ||
| M. genitalium positive‡ | 7 (7.9) | 15 (32.6) | 2.5 (1.3 to 4.6) | 0.004 | 2.0 (1.1 to 3.7) | 0.04 |
| Results based on M. genitalium monoinfection or coinfection§ | ||||||
| M. genitalium negative | 82 (92.1) | 31 (67.4) | Reference | Reference | ||
| M. genitalium monoinfection | 2 (2.2) | 3 (6.5) | 2.2 (1.0 to 4.8) | 0.05 | 1.8 (0.9 to 3.5) | 0.07 |
| M. genitalium coinfectiorrt‡ | 5 (5.6) | 12 (26.1) | 2.5 (1.2 to 4.6) | 0.004 | 1.9 (1.1 to 3.6) | 0.04 |
A total of 135 women had evaluable endometrial histological data.
Model for M. genitalium positive women was adjusted for race, chlamydia, gonorrhoea and bacterial vaginosis. The models for M. genitalium monoinfection or coinfection were adjusted for race and bacterial vaginosis.
M. genitalium negative includes all women who tested negative for M. genitalium at the cervix or endometrium. M. genitalium positive includes all women who tested positive for M. genitalium at the cervix or endometrium.
M. genitalium mono (no coinfection with chlamydia or gonorrhoea) or coinfection with chlamydia or gonorrhoea.
Table 3 displays factors associated with M. genitalium endometrial infection compared with cervical infection only. In the multivariable model, oral sex was associated with reduced risk of endometrial M. genitalium (RR 0.6, 95% CI 0.3 to 0.9) and vaginal T. vaginalis with enhanced risk (RR 2.0, 95% CI 1.2 to 3.4). Women who had C. trachomatis detected at the cervix and endometrium had an increased risk of endometrial M. genitalium (RR 1.7, 95% CI 1.1 to 2.8). High cervical burden (>104 genome copies per swab) was marginally associated with M. genitalium endometrial infection (RR 2.3, 95% CI 1.0 to 5.2).
Table 3.
Factors associated with endometrial Mycloplasma genitalium infection
| Risk factor* | Cervix only n=20 |
Endometrial n=19 |
RR (95%CI) | P values | RRadj (95%CI)† | P values |
|---|---|---|---|---|---|---|
| Demographics and behaviour | ||||||
| Age‡ | 21±3.1 | 21±2.4 | 1.0 (0.9 to 1.1) | 0.85 | NA | NA |
| Non-black race | 9 (45.0) | 3 (15.8) | 0.4 (0.2 to 1.2) | 0.10 | 0.7 (0.3 to 1.4)§ | 0.31 |
| Uninsured | Reference | Reference | Reference | |||
| Private insurance | 2 (10.0) | 3 (15.7) | 0.9 (0.4 to 2.1) | 0.89 | NA | NA |
| Medicaid | 12 (60.0) | 11 (57.9) | 1.2 (0.5 to 2.8) | 0.57 | NA | NA |
| Posthigh school education | Reference | Reference | Reference | NA | NA | |
| <High school education | 4 (20.0) | 5 (26.3) | 1.0 (0.5 to 2.6) | 0.96 | NA | NA |
| High school education | 9 (45.0) | 8 (42.1) | 1.2 (0.5 to 2.6) | 0.66 | ||
| Married | 2 (10.0) | 2 (10.5) | 1.0 (0.4 to 2.9) | 0.97 | NA | NA |
| Non-smoker | 10 (50.0) | 5 (26.3) | 0.6 (0.3 to 1.3) | 0.16 | 0.7 (0.4 to 1.4)§ | 0.34 |
| Drug user | 5 (25.0) | 9 (47.4) | 1.6 (0.9 to 3.0) | 0.13 | 1.0 (0.4 to 2.5)¶ | 0.97 |
| Douching | 9 (45.0) | 9 (47.3) | 1.1 (0.5 to 1.9) | 0.88 | NA | NA |
| Reproductive history, sexual activity and contraception | ||||||
| History of Chlamydia trachomatis | 11 (55.0) | 5 (26.3) | 0.5 (0.2 to 1.1) | 0.10 | 0.8(0.4 to 1.7)¶ | 0.59 |
| History of Neisseria gonorrhoeae | 7 (35.0) | 4 (21.0) | 0.7 (0.3 to 1.6) | 0.37 | NA | NA |
| History of Trichomonas vaginalis | 4 (20.0) | 2 (10.5) | 0.5 (0.1 to 3.1) | 0.48 | NA | NA |
| History of BV | 10 (50.0) | 7 (36.8) | 0.6 (0.2 to 2.2) | 0.43 | NA | NA |
| Female partner | 3 (15.0) | 6 (31.5) | 1.6 (0.9 to 3.0) | 0.11 | 1.2 (0.6 to 2.2)¶ | 0.60 |
| Infected partner | 6 (30.0) | 5 (26.3) | 0.8 (0.2 to 3.4) | 0.81 | NA | NA |
| Multiple partners | 3 (15.0) | 3 (16.7) | 1.1 (0.2 to 6.0) | 0.94 | NA | NA |
| New partner 30d | 5 (25.0) | 4 (21.1) | 0.8 (0.2 to 3.7) | 0.79 | NA | NA |
| Oral sex | 15 (65.2) | 10 (52.6) | 0.6 (0.3 to 1.2) | 0.13 | 0.6 (0.3 to 0.9)§ | 0.04 |
| Consistent condom use | Reference | Reference | Reference | NA | NA | |
| Rarely/never use condoms | 5 (25.0) | 3 (16.7) | 0.5 (0.1 to 2.5) | 0.38 | ||
| Inconsistent condom use | 7 (35.0) | 5 (26.3) | 0.5 (0.1 to 2.3) | 0.41 | ||
| Depo-Provera injections | 3 (15.0) | 4 (21.1) | 1.2 (0.6 to 2.5) | 0.59 | NA | NA |
| Withdrawal | 8 (40.0) | 6 (31.5) | 0.7 (0.2 to 2.6) | 0.61 | NA | NA |
| None | 3 (15.0) | 3 (16.7) | 1.1 (0.2 to 6.0) | 0.95 | NA | NA |
| Microbiology | ||||||
| C. trachomatis | ||||||
| None | Reference | Reference | Reference | |||
| Cervical | 16 (80.0) | 13 (68.4) | 0.7 (0.4 to 1.4) | 0.38 | NA | NA |
| Endometrial | 2 (10.0) | 8 (42.1) | 2.1 (1.2 to 3.7) | 0.009 | 1.7(1.1 to 2.8)§ | 0.02 |
| N. gonorrhoeae | 3 (15.0) | 2 (11.1) | 0.7 (0.1 to 4.8) | 0.73 | NA | NA |
| BV | 15 (75.0) | 13 (68.4) | 2.8 (0.6 to 12.3) | 0.17 | 1.5 (0.7 to 2.8)¶ | 0.27 |
| T. vaginalis | 2 (10.0) | 8 (42.1) | 2.1 (1.2 to 3.7) | 0.009 | 2.0 (1.2 to 3.4)§ | 0.01 |
| High bacterial burden** | 10 (50.0) | 16 (84.2) | 2.4 (0.9 to 6.8) | 0.08 | 2.3 (1.0 to 5.2)§ | 0.05 |
Factors with cell size <2 were not analysed. This included some contraception methods (pills and IUD) and sexual activity variables (uncircumcised partner, anal sex, etc and so on).
P≤0.2 in univariable analyses were tested in the multivariable analysis, where variables with a P≤0.1 were retained.
Age is a continuous variable, values listed in table are mean±SD. All others are binary or categorical and listed as n (%).
Included in final model.
Not retained in the final model.
High burden >104 DNA copies/swab.
BV, bacterial vaginosis; IUD, intrauterine device.
DISCUSSION
Among women at high risk for chlamydia, M. genitalium prevalence was 16.7% and incidence was 25.3 per 100 person-years. This is higher than a population of female university students in the UK20 but comparable with populations of women at risk for STIs.10,21 Baseline C. trachomatis coinfection with M. genitalium was common in our study. This is due to recruitment of women with or at high risk for C. trachomatis within a single geographic location. In a general population of low-risk and high-risk clinics/hospitals, Getman et al10 reported that 68.7% of females with M. genitalium had monoinfection (absence of C. trachomatis, N. gonorrhoeae and T. vaginalis). After treatment, the C. trachomatis coinfection rate for incident M. genitalium was 6.8%.
Examination of data at follow-up revealed that non-black race reduced the risk for M. genitalium, while reporting a new sexual partner and lesser education increased risk. Sexual behaviour is linked to M. genitalium.22,23 Race is also a risk factor for M. genitalium.22 Age was not associated with M. genitalium, but 80% of cohort was <25 years of age.
Only 28% of women with M. genitalium incident infections reported symptoms, which is consistent with reports that M. genitalium is asymptomatic in 75% of infected females.1 Despite the asymptomatic nature, 48.7% of M. genitalium-positive women had endometrial infection. Furthermore, 23% of M. genitalium-positive women had positive tests at two or more consecutive visits. Women who tested positive for M. genitalium following receipt of 1 g orally of azithromycin were either rein-fected by their original partner or from a new partner or failed therapy. Azithromycin treatment failure is common with M. genitalium,7,10 and macrolide resistance rates are high, 60%–80% in men and women.11,12 Although we cannot confirm azithromycin treatment failure, we detected the 2058G mutation in three of five women with testable DNA who were positive for infection 1-month post-treatment. Symptoms between women with repeated positive tests compared with those who cleared infection did not differ. We do not know if women testing positive at follow-up had endometritis, because endometrial biopsies were not obtained at follow-up.
At baseline, M. genitalum was associated with histologically confirmed endometritis independent of C. trachomatis and N. gonorrhoeae. Although, M. genitalium is linked to clinical PID, few studies have examined M. genitalium and endometritis or salpingitis, which are more specific indicators of true upper genital tract infection and inflammation.1 Only one study, of 662 women with clinical PID, has adjusted for C. trachomatis and N. gonorrhoeae7 reporting an adjusted OR for endometritis of 3.0 (95% CI 1.5 to 6.1). Cervical M. genitalium is associated with proinflammatory cytokine production and neutrophil influx.24 Thus, it is perhaps not surprising that endometrial infection with M. genitalium may lead to endometrial inflammation. Although data suggest M. genitalium is associated with infertility, the relationship has not been fully established and is limited primarily by the use of serology.1 Thus, it is important to evaluate the role of M. genitalum in upper genital tract inflammation. Our results coupled with existing data that treatment for PID do not eradicate M. genitalium7 may suggest that the association between M. genitalium and endometritis could be influenced by innappropriate treatment.
Screening for M. genitalium is not currently recommended, although may be potentially beneficial for those at high risk for STIs.1 However, risk factors to identify women at enhanced risk for M. genitalium (particularly ascending infection) are not elucidated. Women with endometrial M. genitalium were more likely to have T. vaginalis and endometrial C. trachomatis. The fact that endometrial C. trachomatis and M. genitalium infections tended to cluster may suggest that those women were immunologically susceptible to ascending infections. Women with endometrial M. genitalium had elevated cervical burden. Few studies have examined M. genitalium burden and reproductive morbidity. In males, M. genitalium burden varies, and high burden may increase treatment failure.25 Oral sex reduced the risk of endometrial infection. Oral sex has been suggested to prime the immune system against infection and decreased endometritis in the PID Evaluation and Clinical Health (PEACH) study.26
We used prospective data to examine risk factors for M. genitalium in a population of women at risk for chlamydia and other STIs. This is one of only a handful of studies that examined the association between M. genitalium and histologically confirmed endometritis while controlling for coinfection. The use of the APTIMA NAAT is a strength over studies that have relied on serology. Our sample size limited our power for some analyses, especially for endometrial infection. Furthermore, not all women had data on endometritis, which may lead to bias. However, the proportion of missing data for endometritis was similar between M. genitalium-positive and M. genitalium-negative women, reducing the potential for bias. The results from this study have narrow generalisability as TRAC recruited women were at high risk for C. trachomatis. This also limited our ability to examine M. genitalium monoinfection. Although we controlled for coinfection in multivariable analysis, we were unable to examine monoinfection and coinfection separately.
We found that M. genitalium was prevalent in women at high risk for STIs. M. genitalium may lead to upper genital tract inflammation independent of C. trachomatis and N. gonorrhoeae. Repeated positive tests for M. genitalium were common during follow-up, indicating possible reinfection or chronic infection. Taken together, these data indicate a need for studies to determine if M. genitalium monoinfection or coinfection leads to upper genital tract infection and inflammation and, if so, clinical trials are needed to identify antibiotics that are effective at eradicating M. genitalium to prevent potential long-term reproductive sequelae.
Key messages.
Mycoplasma genitalium was prevalent in women at high risk for sexually transmitted infections.
M. genitalium was associated with histological endometritis independent of chlamydia and gonorrhoea.
Further studies are needed to determine if screening for M. genitalium will improve reproductive outcomes in women at high risk for STIs.
Acknowledgements
We thank the women who agreed to participate in this study: Ingrid Macio, Melinda Petrina, Carol Priest, Abi Jett and Lorna Rabe for their efforts in the clinic and the microbiology laboratory; and the staff at the Allegheny County Health Department STD Clinic for their efforts.
Funding This work was supported by the National Institute of Allergy and Infectious Diseases (grant numbers U19 AI084024 and R3 AI098660).
Footnotes
Competing interests SLH received funding from Becton Dickinson and Cepheid, which market products for the detection of STIs. All other authors have no conflicts to disclose.
Ethics approval University of Pittsburgh IRB (13–3074) and University of North Carolina IRB (MOD10010159–12/PRO10010159).
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement
Any request for data should be made to TD, University of North Carolina Chapel Hill.
REFERENCES
- 1.Lis R, Rowhani-Rahbar A, Manhart LE. Mycoplasma genitalium infection and female reproductive tract disease: a meta-analysis. Clin Infect Dis 2015;61:418–26. [DOI] [PubMed] [Google Scholar]
- 2.Taylor-Robinson D, Jensen JS. Mycoplasma genitalium: from Chrysalis to multicolored butterfly. Clin Microbiol Rev 2011;24:498–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.McGowin CL, Spagnuolo RA, Pyles RB. Mycoplasma genitalium rapidly disseminates to the upper reproductive tracts and knees of female mice following vaginal inoculation. Infect Immun 2010;78:726–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Møller BR, Taylor-Robinson D, Furr PM, et al. Acute upper genital-tract disease in female monkeys provoked experimentally by Mycoplasma genitalium. Br J Exp Pathol 1985;66:417–26. [PMC free article] [PubMed] [Google Scholar]
- 5.Bjartling C, Osser S, Persson K. Mycoplasma genitalium in cervicitis and pelvic inflammatory disease among women at a gynecologic outpatient service. Am J Obstet Gynecol 2012;206:476e1–8. [DOI] [PubMed] [Google Scholar]
- 6.Jurstrand M, Jensen JS, Magnuson A, et al. A serological study of the role of Mycoplasma genitalium in pelvic inflammatory disease and ectopic pregnancy. Sex Transm Infect 2007;83:319–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Haggerty CL, Totten PA, Astete SG, et al. Failure of cefoxitin and doxycycline to eradicate endometrial Mycoplasma genitalium and the consequence for clinical cure of pelvic inflammatory disease. Sex Transm Infect 2008;84:338–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bjartling C, Osser S, Persson K. The association between Mycoplasma genitalium and pelvic inflammatory disease after termination of pregnancy. BJOG 2010;117:361–4. [DOI] [PubMed] [Google Scholar]
- 9.Workowski KA, Bolan GA. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines, 2015. MMWR Recomm Rep 2015;64:1–137. [PMC free article] [PubMed] [Google Scholar]
- 10.Getman D, Jiang A, O’Donnell M, et al. Mycoplasma genitalium prevalence, coinfection, and macrolide antibiotic resistance frequency in a multicenter clinical study cohort in the United States. J Clin Microbiol 2016;54:2278–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Trembizki E, Buckley C, Bletchly C, et al. High levels of macrolide-resistant Mycoplasma genitalium in Queensland, Australia. J Med Microbiol 2017;66:1451–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dionne-Odom J, Geisler WM, Aaron KJ, et al. High Prevalence of Multidrug-Resistant Mycoplasma genitalium in Human Immunodeficiency Virus-infected Men who have Sex with Men in Alabama. Clin Infect Dis 2018;66:796–798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Russell AN, Zheng X, O’Connell CM, et al. Analysis of factors driving incident and ascending infection and the role of serum antibody in chlamydia trachomatis genital tract infection. J Infect Dis 2016;213:523–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Amsel R, Totten PA, Spiegel CA, et al. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983;74:14–22. [DOI] [PubMed] [Google Scholar]
- 15.Kiviat NB, Wølner-Hanssen P, Eschenbach DA, et al. Endometrial histopathologyin patients with culture-proved upper genital tract infection and laparoscopically diagnosed acute salpingitis. Am J Surg Pathol 1990;14:167–75. [DOI] [PubMed] [Google Scholar]
- 16.Ma L, Jensen JS, Mancuso M, et al. Genetic variation in the complete MgPa operon and its repetitive chromosomal elements in clinical strains of Mycoplasma genitalium. PLoS One 2010;5:e15660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wold C, Sorthe J, Hartgill U, et al. Identification of macrolide-resistant Mycoplasma genitalium using real-time PCR. J Eur Acad Dermatol Venereol 2015;29:1616–20. [DOI] [PubMed] [Google Scholar]
- 18.Wei LJ, Lin DY, Weissfeld L. Regression-analysis of multivariate incomplete failure time data by modeling marginal distributions. J Am Stat Assoc 1989;84:1065–73. [Google Scholar]
- 19.Zou G A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004;159:702–6. [DOI] [PubMed] [Google Scholar]
- 20.Oakeshott P, Aghaizu A, Hay P, et al. Is Mycoplasma genitalium in women the “New Chlamydia?” A community-based prospective cohort study. Clin Infect Dis 2010;51:1160–6. [DOI] [PubMed] [Google Scholar]
- 21.Gaydos C, Maldeis NE, Hardick A, et al. Mycoplasma genitalium compared to chlamydia, gonorrhoea and trichomonas as an aetiological agent of urethritis in men attending STD clinics. Sex Transm Infect 2009;85:438–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Manhart LE, Holmes KK, Hughes JP, et al. Mycoplasma genitalium among young adults in the United States: an emerging sexually transmitted infection. Am J Public Health 2007;97:1118–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Andersen B, Sokolowski I, Østergaard L, et al. Mycoplasma genitalium: prevalence and behavioural risk factors in the general population. Sex Transm Infect 2007;83:237–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Dehon PM, Hagensee ME, Sutton KJ, et al. Histological evidence of chronic mycoplasma genitalium-induced cervicitis in HIV-infected women: a Retrospective Cohort Study. J Infect Dis 2016;213:1828–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Gossé M, Nordbø SA, Pukstad B. Bacterial load in daily urine samples of patients infected withmycoplasma genitalium, mutation analysis, and response to treatment. Infect Dis Obstet Gynecol 2016;2016:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pittrof R, Sully E, Bass DC, et al. Stimulating an immune response? Oral sex is associated with less endometritis. Int J STD AIDS 2012;23:775–80. [DOI] [PMC free article] [PubMed] [Google Scholar]