Research and analysis

Mycoplasma genitalium Antimicrobial Resistance Surveillance (MARS) report: 2024

Updated 13 November 2025

Applies to England

This report includes sentinel Mycoplasma genitalium Antimicrobial Resistance Surveillance (MARS) data from March to June 2024.

Main findings

Mycoplasma genitalium (M. genitalium) is a sexually transmitted pathogen that is often asymptomatic but is an established cause of non-gonococcal urethritis (NGU) and is associated with cervicitis and pelvic inflammatory disease (PID). In 2024, the MARS programme included 1,049 M. genitalium specimens with linked demographic and clinical data. These showed:

• macrolide resistance increased slightly from 62.2% in 2023 to 63.6% in 2024
• fluoroquinolone resistance increased from 12.1% in 2023 to 16.7% in 2024
• dual macrolide and fluroquinolone resistance increased from 10.1% in 2023 to 13.7% in 2024
• most (87.3%) individuals were prescribed the recommended first- or second-line treatments of azithromycin (a macrolide) or moxifloxacin (a fluoroquinolone)
• over three-quarters of individuals received doxycycline (a tetracycline) as a pre-treatment
• among individuals who had a positive test of cure (TOC) after treatment with azithromycin, 90.9% were infected with a macrolide-resistant M. genitalium strain
• among individuals who had a positive TOC after treatment with moxifloxacin, 54.5% were infected with a fluoroquinolone-resistant M. genitalium strain

Recommendations

Possible cases of treatment failure with third-line treatments should be reported to the UK Health Security Agency (UKHSA) via the HIV and STI Data Exchange.

Healthcare practitioners should ensure that individuals are tested appropriately, and those diagnosed with M. genitalium are treated and managed according to national guidelines and should be alert to changes in recommended first-line therapies.

Introduction

Background

M. genitalium is a sexually transmitted pathogen that is often asymptomatic, but is an established cause of NGU and is associated with cervicitis and PID. A meta-analysis of studies conducted between 1991 and 2016 found that the prevalence of M. genitalium among individuals aged 16 to 44 years was 1.3% (95% confidence interval (CI), 1.0% to 1.8%) in high-income countries and 3.9% (95% CI, 2.2% to 6.7%), in countries with lower human development indices. In the UK, the prevalence of M. genitalium infection has been reported as 1% in the general population (aged 16 to 44 years), and up to 38% in individuals attending sexual health services (SHSs).

M. genitalium diagnoses reported to GUMCAD STI Surveillance System (GUMCAD) have increased over recent years (see Figure 1), with 9,058 cases reported in 2024. An analysis has shown that the increasing trend between 2021 and 2023 is likely genuine as testing rates have remained stable, although may in part be due to an increasing number of SHSs starting to test for and report M. genitalium diagnoses to GUMCAD over this time. There were more men, most of whom were heterosexual, diagnosed than women from 2019 to 2024, consistent with men being more likely to experience symptoms. Compared with a recent analysis of the M. genitalium diagnoses in GUMCAD in 2023, the MARS 2023 sample was biased towards London residents and gay, bisexual, and other men who have sex with men (GBMSM).

M. genitalium is diagnosed exclusively by nucleic acid amplification tests due to difficulties in establishing in vitro culture. The limited availability of these molecular diagnostics leading to syndromic management in some settings, along with empirical treatment of NGU and other sexually transmitted infections (STIs) with azithromycin, has led to the emergence of resistant M. genitalium infections worldwide.

M. genitalium lacks a cell wall and is therefore inherently resistant to antibiotics that target cell wall synthesis, such as penicillin and other beta-lactams. Additionally, the organism has a small genome and lacks several proteins involved in metabolic pathways that are targeted by other antimicrobial classes, which further limits treatment options.

The macrolide antibiotic azithromycin is recommended as the first-line treatment, ideally when genotypic susceptibility has been confirmed. Moxifloxacin, a broad-spectrum fluoroquinolone, is the recommended second-line treatment. Doxycycline, a second-generation tetracycline, is recommended as pre-treatment to reduce bacterial load and to increase the effectiveness of the subsequently prescribed antibiotic. Whilst genomic markers predictive of resistance to doxycycline have not been reported in M. genitalium, it is established that this bacterial species is naturally less susceptible to tetracyclines.

Macrolide resistance is conferred by single base mutations, which inhibit antimicrobial binding, primarily at position A2058 or A2059 in region V of the 23S ribosomal ribonucleic acid (rRNA) gene (Escherichia coli numbering). The association between 23S rRNA gene mutations and high azithromycin minimum inhibitory concentrations (MICs) and clinical treatment failure is well documented.

Conversely, the effect of specific mutations on fluoroquinolone resistance in M. genitalium is not fully understood. Amino acid substitutions in the quinolone resistance-determining region (QRDR) of the deoxyribonucleic acid (DNA) Topoisomerase IV subunit C (parC) gene, particularly at positions S83 and D87 (M. genitalium numbering), are associated with elevated moxifloxacin MICs and clinical treatment failure (1 to 3). It is unclear whether ParC substitutions beyond positions S83 and D87 confer moxifloxacin resistance, due to their rarity and insufficient phenotypic MIC and associated clinical outcome data.

Substitutions in the QRDR of the DNA Gyrase Subunit A (gyrA) gene are less common and usually occur alongside a ParC substitution, but their contribution to resistance is not well established. However, the combined presence of the ParC S83I and gyrA M95I substitutions appears to be of particular importance, as individuals infected by ‘double mutation’ strains were twice as likely to fail treatment with moxifloxacin than those with ParC substitutions alone.

The UKHSA performed 2 pilot collections of the MARS programme in 2019 and 2020, before MARS was launched as an annual surveillance programme at a larger scale in 2023.

Figure 1. Number of M. genitalium diagnoses in GUMCAD by gender identity and sexual orientation in England between 2019 and 2024 [note 1]

Source: Data from routine returns to the GUMCAD STI Surveillance Systems.

Note 1: the rise in diagnoses shown in this graph is partly due to improved reporting completeness of M. genitalium diagnoses by SHSs in England; between 2019 and 2020 only approximately 80% of SHSs were reporting diagnoses. Disruption to SHSs in 2020 and 2021 during the COVID-19 pandemic also contributed to the lower number of diagnoses reported in these years.

Objectives

The MARS sentinel surveillance programme aims to:

• estimate the prevalence of genotypic macrolide and fluoroquinolone resistance in M. genitalium infections in individuals attending sentinel SHSs in England
• determine the demographic, behavioural and clinical factors associated with resistance
• assess the treatment outcomes of individuals diagnosed with M. genitalium and relate them to the presence of markers of resistance

Methods

MARS sentinel surveillance programme

MARS sentinel surveillance data and M. genitalium specimens are obtained annually from a network of SHSs and their associated laboratories. In MARS 2024, 32 SHSs and 12 laboratories in England participated in the programme. This included 3 additional non-London SHSs since 2023, bringing the total to 21 non-London SHSs and 11 London SHSs, covering 6 of the 9 UKHSA public health regions (see Figure 2). In 2024, the rate of M. genitalium infections per 100,000 of the population was highest in upper tier local authorities in London. Outside of London, the upper tier local authorities with the highest rates of M. genitalium diagnoses were Manchester, Brighton and Hove, and Liverpool (Figure 2). Testing for M. genitalium is only recommended in individuals who present with specific syndromes or those identified as contacts, in line with the British Association for Sexual Health and HIV (BASHH) guidelines.

Participating laboratories were requested to store aliquots of residual M. genitalium-positive specimens collected as part of routine patient care during the MARS collection period (1 March to 30 June 2024). Aliquots were assigned a unique MARS ID and were stored at -20 degrees Celsius prior to shipment to UKHSA via courier at the end of the collection period. Five participating laboratories routinely refer specimens to the UKHSA’s Sexually Transmitted Infections Reference Laboratory (STIRL) for M. genitalium detection and/or antimicrobial resistance (AMR) testing. As such, positive specimens received during the collection period were automatically included in MARS by STIRL.

M. genitalium-positivity was confirmed using the STIRL in-house real-time polymerase chain reaction (PCR) assay which targets 2 intrinsic M. genitalium genes: M. genitalium adhesin P1 (MgPa) and glyceraldehyde-3-phosphate dehydrogenase (gap).

For specimens confirmed to be positive for M. genitalium, molecular antimicrobial susceptibility testing was performed via real-time-PCR followed by Sanger sequencing for:

• macrolide resistance: Region V of the 23S rRNA gene
• fluoroquinolone resistance: the QRDR of the parC gene
• fluoroquinolone resistance in specimens where the parC genotype was non-wildtype: QRDR of the gyrA gene

Predicted antibiotic resistance was inferred from the detection of known M. genitalium mutations in these genes, acknowledging the limited data for correlating some QRDR mutations with fluoroquinolone resistance (see Table 2b and Table A1). Sequence data was analysed using BioNumerics software (version 6.1) developed by Applied Maths.

Antimicrobial susceptibility results were securely linked to the pseudonymised GUMCAD data to obtain demographic and clinical details. GUMCAD is a disaggregated, patient-level dataset of all STI tests and diagnoses at SHSs in England. This data was further enriched with additional demographic, clinical, and behavioural information provided by SHSs through the HIV and STI Data Exchange.

The Sankey diagram (Figure 8) was generated using the SankeyMATIC tool.

Figure 2. Map showing the 32 sentinel SHSs participating in MARS 2024 across England and London (shown at larger scale), as well as the rate of M. genitalium diagnoses per 100,000 of the population by upper tier local authority in 2024

Source: Data from MARS sentinel surveillance programme.

Note 2: London SHSs are omitted from the England map as clinic markers overlap at this scale.

Results

Sampling frame

A total of 1,327 specimens were collected from patients at participating SHSs and sent to STIRL (see Figure 3). Of these, 90.9% (1,206 specimens from 1,049 individuals) were confirmed to be M. genitalium-positive using the UKHSA in-house real-time-PCR assay. Discrepancies between local and STIRL M. genitalium detection results were most likely due to DNA degradation of specimens on storage, and/or disparities in assay sensitivity between commercial platforms and the STIRL in-house assay.

Of the 1,049 patients, 86.3% (number (n)=905) of the MARS sample were linked to a GUMCAD record, providing some of the demographic data. Records for all 1,049 individuals were successfully enhanced with additional demographic, behavioural and clinical information provided by SHS clinicians (see Tables 1a to 1c). Thus, the final MARS 2024 sample included all 1,049 individuals.

Macrolide susceptibility data was available for 89.9% (n=943) of specimens; fluoroquinolone susceptibility data was available for 83.3% (n=874) of specimens; and data for both were available for 81.1% (n=851) of specimens. The gyrA genotype was successfully investigated in 93.2% (n=151 out of 162) of eligible specimens.

Figure 3. The sentinel surveillance sampling frame flowchart in MARS 2024

Source: Data from MARS sentinel surveillance programme.

Sentinel surveillance sample

The MARS sample includes specimens from a subset of all M. genitalium diagnoses in England from 1 March to 30 June. Data and demographic breakdowns of all reported M. genitalium diagnoses in England are presented in the UKHSA STI annual data tables.

The MARS sentinel sample included 1,049 individuals who were diagnosed with M. genitalium infections at sentinel SHSs between 1 March and 30 June 2024. During the same period, 2,867 M. genitalium diagnoses were reported to GUMCAD across England, meaning the MARS 2024 sample represents 36.6% of the total national diagnoses, compared with 45.4% in MARS 2023. The sentinel nature of the MARS programme may mean the results are not fully representative of all M. genitalium diagnoses and AMR in M. genitalium in England.

Tables 1a to 1c show the demographic summary of 1,049 patients included in the MARS 2024 sample.

Among 1,049 individuals with a M. genitalium positive specimen, 65.6% (n=688) were male, most of whom (68.1%) were heterosexual. Individuals most frequently belonged to the White ethnic group (48.0%) and the modal age group was aged 25 to 34 years (43.9%), with ages ranging from 15 to 73 years. The majority of individuals were born in the UK (66.5%; n=660) and just below half (47.7%; n=500) of individuals attended an SHS in London.

Among all individuals, 2.8% were living with HIV. Among those who were HIV negative or where HIV status was unknown, 12.6% (n=117) were using HIV pre-exposure prophylaxis (PrEP) at the time of M. genitalium diagnosis. At diagnosis, 66.5% of individuals displayed signs and symptoms indicative of M. genitalium infection, with 83.5% of GBMSM and 77.6% of heterosexual men presenting with symptoms, compared to 42.2% of women. Of those who were asymptomatic or had symptoms not indicative of M. genitalium, 50.4% were tested because they were a contact (n=177 out of 351).

At the time of M. genitalium diagnosis, 17.5% of individuals in the MARS sample were diagnosed with a concurrent STI, most frequently chlamydia (9.2%) or gonorrhoea (5.4%). Over one fifth of individuals had been diagnosed with another STI in the past 12 months (22.2%), again most commonly chlamydia (8.4%) or gonorrhoea (8.4%). Individuals most frequently reported having 0 to 1 (60.0%) or 2 to 5 (32.7%) sexual partners in the 3 months prior to their M. genitalium diagnosis, and 9.3% of individuals reported having a sexual partner outside of the UK in the same time interval. The majority (97.7%) of specimens included in the sentinel surveillance sample were genital samples (including urine samples), however, specimen type was not reported for 146 individuals.

Table 1a. Age, ethnic group, country of birth and location of SHS attended among individuals included in the MARS 2024 sample, by gender identity and sexual orientation, England [note 3]

Characteristic	GBMSM	Heterosexual men	Women [note 4]	Other [note 5]	Total
Metric	n (% of total number (N)	n (% of N)	n (% of N)	n (% of N)	n (% of N)
Number of individuals	212 (20.2%)	469 (44.7%)	358 (34.1%)	10 (1.0%)	1,049 (100.0%)
Aged 15 to 24 years	24 (11.3%)	140 (29.9%)	159 (44.4%)	1 (10.0%)	324 (30.9%)
Aged 25 to 34 years	101 (47.6%)	209 (44.6%)	142 (39.7%)	8 (80.0%)	460 (43.9%)
Aged 35 to 44 years	71 (33.5%)	84 (17.9%)	46 (12.9%)	1 (10.0%)	202 (19.3%)
Aged 45 years and over	16 (7.6%)	36 (7.7%)	11 (3.1%)	0 (0.0%)	63 (6.0%)
Ethnic group: Asian	19 (9.0%)	35 (7.5%)	39 (10.9%)	0 (0.0%)	93 (8.9%)
Ethnic group: Black	33 (15.6%)	158 (33.7%)	51 (14.3%)	3 (30.0%)	245 (23.4%)
Ethnic group: Mixed	20 (9.4%)	27 (5.8%	14 (3.9%)	3 (30.0%)	54 (5.2%)
Ethnic group: Other	8 (3.8%)	27 (5.8%)	13 (3.6%)	1 (10.0%)	49 (4.7%)
Ethnic group: White	122 (57.6%)	171 (36.5%)	208 (58.1%)	2 (20.0%)	503 (48.0%)
Ethnic group: not reported [note 7]	10	27	14	3	54
Country of birth: UK	114 (56.2%)	297 (67.4%)	243 (71.1%)	6 (85.7%)	660 (66.5%)
Country of birth: outside of the UK	89 (43.8%)	144 (32.7%)	99 (29.0%)	1 (14.3%)	333 (33.5%)
World region of birth: not reported [note 7]	9	28	16	3	56
SHS location: London	138 (65.1%)	216 (46.1%)	142 (39.7%)	4 (40.0%)	500 (47.7%)
SHS location: outside of London	74 (34.9%)	253 (53.9%)	216 (60.3%)	6 (60.0%)	549 (52.3%)

Source: Data from MARS sentinel surveillance programme.

Note 3: these percentages reflect the distribution of these characteristics among people attending the 32 sentinel SHSs participating in MARS.

Note 4: includes women who have sex with men and women who have sex with women.

Note 5: includes individuals who are gender diverse and males where sexual orientation was not reported.

Note 6: the ethnic categories above are as specified by the Office for National Statistics (ONS).

Note 7: individuals where demographic data was not reported are not included in the percentage calculations.

Table 1b. HIV status and use of HIV PrEP among individuals included in the MARS 2024 sample, by gender identity and sexual orientation, England [note 8]

Characteristic	GBMSM	Heterosexual men	Women [note 9]	Other [note 10]	Total
HIV status: living with HIV	21 (10.1%)	4 (0.9%)	2 (0.7%)	0 (0.0%)	27 (2.8%)
HIV status: negative	186 (89.9%)	425 (99.1%)	307 (99.4%)	9 (100.0%)	927 (97.2%)
HIV status: not reported [note 11]	5	40	49	1	95
Use of HIV PrEP [note 12]: yes	111 (62.4%)	4 (1.0%)	2 (0.6%)	0 (0.0%)	117 (12.6%)
Use of HIV PrEP [note 12]: no	67 (37.6%)	418 (99.1%)	322 (99.4%)	8 (100.0%)	815 (87.5%)
Use of HIV PrEP [note 9]: not reported [note 11]	13	43	32	2	90

Source: Data from MARS sentinel surveillance programme.

Note 8: these percentages reflect the distribution of these characteristics among people attending the 32 sentinel SHSs participating in MARS.

Note 9: includes women who have sex with men and women who have sex with women.

Note 10: includes individuals who are gender diverse and males where sexual orientation was not reported.

Note 11: individuals where demographic data was not reported are not included in the percentage calculations.

Note 12: HIV PrEP use is only among individuals who are HIV negative or where HIV status is unknown.

Table 1c. M. genitalium symptom status, previous STI diagnosis, number of sexual partners, and sexual partners whilst abroad among individuals included in the MARS 2024 sample, by gender identity and sexual orientation, England [note 13]

Characteristic	GBMSM	Heterosexual men	Women [note 14]	Other [note 15]	Total
Symptomatic at diagnosis: no	35 (16.5%)	105 (22.4%)	207 (57.8%)	4 (40.0%)	351 (33.5%)
Symptomatic at diagnosis: yes	177 (83.5%)	364 (77.6%)	151 (42.2%)	6 (60.0%)	698 (66.5%)
Any concurrent STI diagnosis: no	156 (78.0%)	389 (85.1%)	280 (81.6%)	8 (80.0%)	833 (82.5%)
Any concurrent STI diagnosis: yes	44 (22.0%)	68 (14.9%)	63 (18.4%)	2 (20.0%)	177 (17.5%)
Any concurrent STI diagnosis: not reported [note 16]	12	12	15	0	39
Concurrent STI diagnosis [note 17]: gonorrhoea	25 (11.8%)	13 (2.8%)	18 (5.0%)	1 (10.0%)	57 (5.4%)
Concurrent STI diagnosis [note 17]: chlamydia	17 (8.0%)	50 (10.7%)	29 (8.1%)	0 (0.0%)	96 (9.2%)
Concurrent STI diagnosis [note 17]: syphilis	6 (2.8%)	4 (0.9%)	0 (0.0%)	0 (0.0%)	10 (1.0%)
Concurrent STI diagnosis [note 17]: other	6 (2.8%)	10 (2.1%)	10 (2.8%)	1 (10.0%)	26 (2.5%)
Any previous STI diagnosis (last 12 months): no	100 (57.5%)	322 (81.5%)	256 (84.8%)	5 (71.4%)	683 (77.8%)
Any previous STI diagnosis (last 12 months): yes	74 (42.5%)	73 (18.5%)	46 (15.2%)	2 (28.6%)	195 (22.2%)
Any previous STI diagnosis (last 12 months): not reported [note 16]	38	74	56	3	171
Previous STI diagnosis (last 12 months) [note 17]: gonorrhoea	48 (22.6%)	23 (4.9%)	16 (4.5%)	1 (10.0%)	88 (8.4%)
Previous STI diagnosis (last 12 months) [note 17]: chlamydia	28 (13.2%)	38 (8.1%)	22 (6.2%)	0 (0.0%)	88 (8.4%)
Previous STI diagnosis (last 12 months) [note 17]: syphilis	6 (2.8%)	0 (0.0%)	1 (0.3%)	0 (0.0%)	7 (0.7%)
Previous STI diagnosis (last 12 months) [note 17]: other	11 (5.2%)	26 (5.5%)	11 (3.1%)	1 (10.0%)	49 (4.7%)
Total sexual partners (past 3 months): 0 to 1	62 (32.1%)	269 (61.3%)	249 (76.4%)	6 (66.7%)	586 (60.6%)
Total sexual partners (past 3 months): 2 to 5	88 (45.6%)	160 (36.5%)	66 (20.3%)	2 (22.2%)	316 (32.7%)
Total sexual partners (past 3 months): 6 or more	43 (22.3%)	10 (2.3%)	11 (3.4%)	1 (11.1%)	65 (6.7%)
Total sexual partners (past 3 months): not reported [note 16]	19	30	32	1	82
Sex partners whilst abroad (past 3 months): no	78 (88.6%)	264 (90.4%)	210 (92.1%)	7 (87.5%)	559 (90.8%)
Sex partners whilst abroad (past 3 months): yes	10 (11.4%)	28 (9.6%)	18 (7.9%)	1 (12.5%)	57 (9.3%)
Sex partners whilst abroad (past 3 months): not reported [note 16]	124	177	130	2	433

Source: Data from MARS sentinel surveillance programme.

Note 13: these percentages reflect the distribution of these characteristics among people attending the 32 sentinel SHSs participating in MARS.

Note 14: includes women who have sex with men and women who have sex with women.

Note 15: includes individuals who are gender diverse and males where sexual orientation was not reported.

Note 16: individuals where demographic data was not reported are not included in the percentage calculations.

Note 17: some individuals had more than one concurrent or previous STI diagnosis, so appear in more than one total.

Antimicrobial resistance

Only clinical specimens that were confirmed to be M. genitalium-positive using the STIRL in-house assay and had sufficient DNA yield underwent antimicrobial susceptibility analysis. Of all sequenced specimens, 63.6% (n=600 out of 943) had a mutation associated with macrolide resistance, most commonly A2058G (50.0%) or A2059G (37.7%) in the 23S rRNA gene (see Table 2a). Two specimens carried the A2058C mutation.

Of all specimens, 16.7% (n=146 out of 874) had an amino acid substitution predictive of fluoroquinolone resistance, consisting of serine to isoleucine83 (S83I) (66.4%), aspartic acid to asparagine87 (D87N) (19.9%), serine to arginine83 (S83R) (7.5%), D87Y (5.5%) or D87N and S83I (mixed; 0.7%; see Table 2b). ParC substitutions other than these were detected in 1.8% (n=16) of specimens (see Table A1); these have unknown clinical significance and are defined as non-wild-type in this report.

Among the 162 ParC mutant specimens, sequencing of the gyrA QRDR was successful for 93.2% (n=151). Analysis revealed that 12 specimens harboured gyrA substitutions that may act synergistically with ParC substitutions to confer fluoroquinolone resistance. Specifically, 4 harboured an aspartic acid to glycine99 (D99G) substitution, 3 harboured an aspartic acid to asparagine99 (D99N) substitution, 3 harboured a methionine to isoleucine95 (M95I) substitution, and 2 harboured an aspartic acid to tyrosine99 (D99Y) substitution. Of these 12 specimens, 10 harboured a S83I ParC substitution, 1 harboured a D87Y ParC substitution, and 1 harboured a S83R ParC substitution, all of which have been reported previously. All 12 specimens were also genotypically resistant to macrolides.

Predicted dual resistance to both macrolides and fluoroquinolones was detected in 13.7% (n=117 out of 851) of specimens. Of the 146 specimens with predicted fluoroquinolone resistance, 83.0% (n=117 out of 141) of these also contained mutations predictive of macrolide resistance (where 23S rRNA testing was successful).

Table 2a. The frequency of mutations in the 23S rRNA gene and their association with macrolide resistance in M. genitalium specimens included the MARS 2024 sample, England (n=943)

Genotype	Frequency	Percentage
Susceptible [note 18]	343	36.4%
Resistant total	600	63.6%
A2058G	300	50.0%
A2059G	226	37.7%
A2058T	56	9.3%
A2059C	14	2.3%
A2058C	2	0.3%
A2058G and A2059G (mixed)	2	0.3%

Source: Data from MARS sentinel surveillance programme.

Note 18: susceptible indicates wild-type sequence.

Table 2b. The frequency of amino acid substitutions in the ParC (n=874) and gyrA (n=151) genes and their association with fluoroquinolone resistance in M. genitalium specimens included in the MARS 2024 sample, England

Genotype	Frequency	Percentage
ParC: susceptible [note 19]	712	81.5%
ParC: ­­­­unknown [note 20]	16	1.8%
ParC: resistant total	146	16.7%
ParC: S83I	97	66.4%
ParC: D87N	29	19.9%
ParC: S83R	11	7.5%
ParC: D87Y	8	5.5%
ParC: D87N and S83I (mixed)	1	0.7%
GyrA: susceptible [note 19]	139	92.1%
GyrA: resistant total	12	7.9%
GyrA: D99G	4	33.3%
GyrA: D99N	3	25.0%
GyrA: M95I	3	25.0%
GyrA: D99Y	2	16.7%

Source: Data from MARS sentinel surveillance programme.

Note 19: susceptible indicates wild-type sequence.

Note 20: unknown includes N91I, D82N, G81C, P62S and G81C (mixed), S83N, S83I and S83N (mixed), and S84P amino acid substitutions (see Appendix: Table A1).

The percentage of specimens included in the MARS 2019, 2020, 2023 and 2024 collection that had molecular markers predictive of macrolide, fluoroquinolone, and dual resistance are presented in Figure 4. In MARS 2024, macrolide resistance increased slightly compared with 2023 (63.6% vs 62.2%), though levels in both years remained lower than those observed in the pilot years (approximately 69%). Fluoroquinolone resistance increased to 16.7% in 2024, rising from 12.1% in 2023 and 11.1% in 2020. Similarly, dual resistance increased to 13.7% in 2024, predominantly driven by the rise in fluoroquinolone resistance. This was an increase compared with around 10% in both 2020 and 2023.

Figure 4. The percentage of M. genitalium specimens with genetic markers predictive of macrolide, fluoroquinolone, and dual resistance in the MARS pilots (2019 and 2020), MARS 2023 and 2024

Source: Data from MARS sentinel surveillance programme.

Resistance profiles by patient and demographic factors

The percentage of M. genitalium specimens with genotypic macrolide or fluoroquinolone resistance by selected demographics are presented in Figures 5 to 7. The following data only include specimens where macrolide (89.9%; n=943 out of 1,049) or fluoroquinolone (83.3%; n=874 out of 1,049) resistance data was available.

Gender and sexual orientation

The percentage of specimens with macrolide resistance-associated mutations was higher in GBMSM (79.6%) compared with heterosexual men (56.6%) and women (63.9%; Figure 5). Compared to MARS 2023, resistance increased slightly in GBMSM and women (from 77.4% and 59.5%, respectively) and decreased slightly in heterosexual men (from 57.6% in 2023).

Although fluoroquinolone resistance was less common overall, it was also higher in specimens from GBMSM (29.8%) compared with heterosexual men (12.9%) and women (14.1%; Figure 5). Relative to MARS 2023, increases were seen in all 3 groups, with the steepest rises in GBMSM and women (rising from 21.6% and 9.5%, respectively).

Figure 5. The percentage of macrolide- and fluoroquinolone-resistant M. genitalium specimens in the MARS 2023 and 2024 sentinel surveillance sample by gender and sexual orientation, England

Source: Data from MARS sentinel surveillance programme.

Age group

In 2024, the percentage of specimens with macrolide resistance-associated mutations was similar across the younger age groups (ranging from 64.0% to 64.7% among those aged 15 to 44; Figure 6. Macrolide resistance was lower among those aged 45 years or over (50.0%), dropping from 64.7% in 2023. In 2023, macrolide resistance was similar across all 4 age groups.

The percentage of specimens with fluoroquinolone resistance-associated mutations in 2024 varied by age group, with a greater percentage of resistant specimens in older age groups (13.9% in those aged 15 to 24 years compared with 20.8% in those aged 45 years and over; Figure 6). Fluoroquinolone resistance increased in all age groups compared to 2023.

Figure 6. The percentage of macrolide- and fluoroquinolone-resistant M. genitalium specimens in the MARS 2023 and 2024 sentinel surveillance sample by age group (years), England

Source: Data from MARS sentinel surveillance programme.

Previous STI diagnoses

In 2024, specimens from individuals who had reported a previous STI diagnosis in the last 12 months were more frequently resistant (76.0%) than those from individuals without a previous STI diagnosis in the last 12 months (61.2%; Figure 7). Previous STIs included gonorrhoea (n=88), chlamydia (n=88), and other (n=49). Data regarding antibiotics prescribed for STIs other than M. genitalium is not available within GUMCAD. The percentage of macrolide resistant specimens among those who reported an STI diagnosis in the last 12 months increased sharply compared with the MARS 2023 data, where the percentage of resistant specimens was 62.4% and similar to those who did not report a previous STI diagnosis.

The percentage of specimens with fluoroquinolone resistance-associated mutations was higher in individuals who reported a previous STI diagnosis in the last 12 months (20.9%) compared with those who did not (14.6%; Figure 7). The percentage of fluoroquinolone resistant specimens increased in both groups compared with the MARS 2023 data (rising from 10.8% and 13.4%, respectively).

Figure 7. The percentage of macrolide- and fluoroquinolone-resistant M. genitalium specimens in the MARS 2023 and 2024 sentinel surveillance sample by previous STI diagnosis (last 12 months), England

Source: Data from MARS sentinel surveillance programme.

Prescribing practices

Antimicrobial prescribing data was available for all individuals in the MARS 2024 sample. The Sankey diagram (Figure 8) shows the treatment pathway for 1,049 patients, up to the second treatment prescribed. For this analysis, TOC outcomes were categorised as positive or negative following treatment, with a positive TOC indicating persistence of M. genitalium infection and a negative TOC suggesting microbiological clearance. Some patients did not have a TOC, in some cases because they did not respond to recall or if the SHS did not offer them. Patients whose treatment journey was incomplete within the MARS sampling period and who would not yet have been offered a TOC were excluded from the treatment outcome analysis.

First treatment and pre-treatment

Most individuals (87.3%; n=916 out of 1,049) were prescribed the recommended first- or second-line treatments of azithromycin (n=546) or moxifloxacin (n=374), either as monotherapy or in combination with other antibiotics. 42.8% (n=213 out of 498) of individuals prescribed azithromycin as part of their first treatment were infected with a macrolide-resistant strain, whereas 92.7% (n=317 out of 342) of individuals prescribed moxifloxacin as part of their first treatment were infected with a macrolide-resistant strain (where sequence data was available).

Over three-quarters of all patients were initially pre-treated with doxycycline (80.8%; n=835 out of 1,033). Of the patients that received azithromycin monotherapy as their first treatment, 95.4% (n=514 out of 539) also received doxycycline pre-treatment. 60.4% (n=223 out of 369) of patients that received moxifloxacin monotherapy as their first treatment also received doxycycline pre-treatment. Of all patients, 7.5% (n=77 out of 1,033) received only doxycycline with no further treatment. Doxycycline pre-treatment was not given to 19.2% (n=198 out of 1,033) of individuals. Of these, 9.5% (n=19 out of 198) were prescribed azithromycin, 69.2% (n=137 out of 198) were prescribed moxifloxacin, and 3% (n=6 out of 198) were prescribed other antibiotics or antibiotic combinations at their first SHS attendance noted within MARS 2024. Additionally, there were 36 patients (3.4%) who did not receive any treatment.

A small number of patients were not prescribed the recommended first- or second-line therapy or were prescribed them in combination with another antibiotic (2.2%; n=23 out of 1,049). Of these, 13 received minocycline, 6 received metronidazole, and 4 received pristinamycin. Seventeen of the individuals received doxycycline pre-treatment and 6 did not. Of the 23 patients who received non-first- or second-line treatments, 34.8% (n=8) had infections that were dual resistant to macrolides and fluoroquinolones. Of these 8, 4 received pristinamycin, 3 received minocycline, and 1 received metronidazole.

Second treatment

Among patients who received azithromycin as their first treatment, the majority had no further treatment (69.0%; n=372 out of 539). A small number of patients (1.7%; n=9 out of 539) received a second course of azithromycin, while the majority were switched to moxifloxacin (27.6%; n=149 out of 539). Of the patients who were switched to moxifloxacin, most (95.1%; n=137 out of 144) had a macrolide-resistant specimen (where macrolide resistance data was available).

Similarly, among those who received moxifloxacin as their first treatment, the majority also had no further treatment (94.3%, n=348 out of 369), and the others either received a second course of moxifloxacin (3.0%; n=11 out of 369), were switched to azithromycin (0.5%; n=2 out of 369) or were switched to another antibiotic (2.2%; n=8 out of 369). Of individuals who received a second treatment after moxifloxacin, 65.0% (n=13 out of 20) had a fluoroquinolone-resistant infection (1 D87N, 1 D87N and S83I mixed and 11 S83I). Of the 13 individuals with a fluoroquinolone-resistant infection, 4 of the individuals with an S83I substitution also had either a M95I (n=2), D99G (n=1) or a D99Y (n=1) substitution in gyrA (where sequence data was available).

Subsequent treatments

Further courses of treatment (third and beyond) were rare and involved a small number of individuals: 23 patients received a third treatment, 8 patients received a fourth, 3 patients received a fifth treatment, 2 patients received a sixth treatment, and 1 patient received a seventh treatment. These treatments were predominantly azithromycin, moxifloxacin, minocycline and pristinamycin. Following these additional treatments, 13 individuals returned a positive TOC, and 10 individuals returned a negative TOC. A TOC was not performed for 10 individuals. No TOC information was available for 4 individuals.

Figure 8. Sankey diagram showing the number of individuals prescribed each antimicrobial (combination) as their first or second treatment for M. genitalium infection in the MARS 2024 sentinel surveillance sample, England

Source: Data from MARS sentinel surveillance programme.

Resistance-associated mutations and treatment outcomes

Azithromycin

Macrolide resistance genotype by TOC outcome for individuals prescribed azithromycin as their first treatment (where macrolide resistance data was available) is shown in Figure 9. A TOC result was available for 54.5% (n=224 out of 411) of patients. A positive TOC was uncommon, occurring in 16.1% (n=66) of patients following azithromycin treatment. Among those with a positive TOC result, almost all (90.9%; n=60 out of 66) had a macrolide resistant infection. Where data was available, 71.0% (n=39 out of 55) of patients with a positive TOC experienced clinical failure, in other words, remained symptomatic. However, macrolide resistance mutations were not unequivocally associated with treatment failure; 14.6% (n=23 out of 158) of individuals with a known negative TOC also had infections with these mutations.

Among patients where a TOC result was not available, 26.7% (n=50 out of 187) had a macrolide resistant infection; similar to the percentage among those with a negative TOC. The most common reason that a TOC was not conducted among these patients was because they did not respond to recall (76.9%; n=30 out of 39).

Figure 9. M. genitalium macrolide resistance genotype by TOC result following first treatment with azithromycin in the MARS 2024 sentinel surveillance sample (n=411), England

Source: Data from MARS sentinel surveillance programme.

Moxifloxacin

Fluoroquinolone resistance genotype by TOC outcome for patients treated with moxifloxacin (including those treated with moxifloxacin as their first treatment or following azithromycin treatment) is shown in Figure 10 (where fluoroquinolone resistance data was available). Of these individuals, a TOC result following treatment was available for 58.9% (n=242 out of 411) and a positive TOC following moxifloxacin treatment was rare, occurring in only 7.5% of patients (n=31 out of 411).

The highest percentage of resistant specimens was observed among patients with a positive TOC, with 51.6% (n=16 out of 31) carrying a resistant infection. Of these 16 individuals, 35.7% (n=5 out of 14) of infections also carried a mutated gyrA gene (where sequencing was successful).

ParC mutations were not fully predictive of TOC outcomes. Among those with a negative TOC, 16.6% (n=35 out of 211) had a resistant specimen. Where a TOC result was not available, the percentage of resistant specimens was similar to those with a negative TOC (18.3%; n=31 out of 169). In most of these patients (60.9%; n=14 out of 23), the reason a TOC was not conducted was because the patient didn’t respond to recall, whereas symptom resolution accounted for only 4.3% (n=1) of these patients. A TOC was also not conducted among 34.8% of patients (n=8) because the SHS did not routinely offer them.

Figure 10. M. genitalium fluoroquinolone resistance genotype by TOC result following first treatment with moxifloxacin in the MARS 2024 sentinel surveillance sample (n=411), England

Source: Data from MARS sentinel surveillance programme.

Discussion

The 2024 MARS programme has shown that macrolide resistance in M. genitalium remains high, with 63.6% of sequenced specimens carrying mutations associated with macrolide resistance. While macrolide resistance decreased slightly in MARS 2023 compared with the 2019 and 2020 pilots, resistance increased in MARS 2024.

Fluoroquinolone resistance, although lower than macrolide resistance at 16.7%, showed a substantial increase from 2023, and a year-on-year rise since 2019. Dual resistance to both antimicrobial classes also increased to 13.7% in 2024, predominantly due to the rise in fluoroquinolone resistance. Although macrolide resistance was universally high, specimens from GBMSM (77.4%) had higher rates of resistance compared with other gender and sexual orientation groups. Resistance to fluoroquinolones was also higher in specimens from GBMSM (29.8%) compared with the overall rate of 16.7%. Increased resistance in specimens from GBMSM was also observed in MARS 2023.

Most (87.3%) individuals were prescribed the recommended first- or second-line treatments recommended by BASHH, azithromycin and moxifloxacin. While the prescribing data may not be complete for all individuals due to the sentinel time period of MARS, they provide useful insight into how AMR is shaping real-world treatment decisions. The findings from MARS suggest that AMR results were used to guide treatment decisions: 92.7% of patients whose first recorded treatment in MARS 2024 was moxifloxacin were infected with a macrolide-resistant strain. This aligns with current guidance recommending fluoroquinolones as a first-line option when macrolide resistance is detected or when a patient has previously experienced treatment failure.

Among the 23 patients who were prescribed non first or second-line treatments, 34.8% had infections that were dual resistant to macrolides and fluroquinolones. However, as MARS only captures data within a 4-month period, it is unclear what previous treatments these patients may have received, or if antimicrobial susceptibility results were available to clinicians to inform treatment decisions.

Almost all (90.9%) patients who had a positive TOC following azithromycin treatment had a macrolide resistant infection, indicating a strong association between 23S rRNA mutations and a positive TOC result. However, macrolide resistance mutations were not invariably predictive of TOC outcomes as 14.6% of individuals with a negative TOC were also infected with a macrolide-resistant strain. One limitation of this analysis is that the effect of other antibiotics was not considered. Notably, 95.4% of individuals who were treated with azithromycin also received doxycycline, suggesting that doxycycline or another antibiotic could have contributed toward azithromycin treatment success, potentially by reducing bacterial load. The possibility of spontaneous self-clearance was also not considered, with a clearance rate of 55% at 3 months after M. genitalium infection reported in a cohort study in Uganda. A recent Austrian study also reported high cure rates with empiric azithromycin therapy (68%), despite a high prevalence of azithromycin resistance (68%), supporting the view that 23S rRNA mutations are sensitive but not entirely specific predictors of microbiological clearance.

For moxifloxacin, the correlation between parC mutations and TOC outcome was less well defined. Of the patients with a positive TOC result, 51.6% had mutations in the parC gene and 35.7% of the parC mutants also had mutations in the gyrA gene. This is consistent with studies that have shown that the presence of mutations in both genes is a strong predictor of moxifloxacin treatment failure. However, almost half (49.4%) of patients with a positive TOC result had a fluoroquinolone-sensitive genotype, which could reflect incomplete adherence to treatment, reinfection, or a TOC performed too soon after treatment. Despite this, the percentage of parC-mutated specimens was lower among those with a negative TOC (16.6%), suggesting that fluoroquinolone resistance-associated mutations likely contribute to moxifloxacin treatment failure, but are not sufficient on their own to predict it, consistent with findings from clinical studies.

Treatment outcomes were difficult to accurately ascertain as a TOC result was only available for 45.0% of azithromycin treatments and 54.5% of moxifloxacin treatments. For both antibiotics, the most common reason a TOC wasn’t performed was because the patient did not respond to recall (76.9% and 60.9%, respectively). Although symptom resolution was only recorded as the reason for non-attendance in one patient by the clinician, it is likely that additional patients did not return as their symptoms had resolved and clinical cure had been achieved. In addition, for the 8 patients where the SHS did not routinely perform TOCs for M. genitalium, may reflect early adoption of the 2025 BASHH guidelines, which no longer recommend a TOC in cases of symptom resolution. Supporting this, the percentages of macrolide and fluoroquinolone resistant specimens were similar between patients with a negative TOC and in those with an unknown TOC outcome. This suggests that assuming treatment success in the absence of a TOC result provides a reasonable proxy for treatment success.

Our analyses are limited as we can only consider treatments reported in our enhanced data collection for a patient. In some cases, the details of treatment prior to the date of specimen collection are not complete. Additionally, further treatments beyond the MARS collection window are not recorded. This means for some cases we cannot explain the relationship between resistance and treatment failure.

The MARS 2024 collection is the largest dataset to examine the pattern of AMR in M. genitalium in England, comprising 1,049 individuals from 32 different SHSs. This data builds on the 2023 MARS Report, and with 2 consecutive years of data we can now begin to observe trends over time. The demographic completeness was excellent, providing rich information for detailed analysis.

Although the sample size was large and spanned 6 of the 9 UKHSA regions, some selection bias is possible, as SHS participation was largely determined by the availability of local M. genitalium testing and subsequent reporting to GUMCAD. Asymptomatic screening for M. genitalium is not recommended, with only those who are symptomatic or contacts recommended to be tested. Given that M. genitalium infections can often be asymptomatic, MARS is not representative of all M. genitalium infections in England. However, comparison with GUMCAD data suggests the MARS sample is broadly representative of national M. genitalium diagnoses, and coverage improved between 2023 and 2024 with more non-London SHSs included.

Conclusions

The MARS 2024 sentinel surveillance programme provides the most comprehensive picture to date of the AMR landscape in M. genitalium in England, building on the findings from MARS 2023. Across both years, results confirm high levels of macrolide resistance, with evidence of increasing fluoroquinolone and dual resistance. Despite this, some genotypically resistant infections appear to respond to the corresponding antimicrobial, suggesting that resistance-associated mutations do not invariably predict treatment failure.

The inclusion of detailed demographic data enables analysis of AMR patterns across different population groups, offering valuable insight into disparities in resistance. With continued annual data collection, MARS will allow longitudinal monitoring of AMR in M. genitalium, improving understanding of trends over time and supporting the development of effective treatment strategies and public health interventions.

Appendix

Table A1. The frequency of amino acid substitutions in the ParC (n=16) gene with an unknown effect on fluoroquinolone resistance in M. genitalium specimens included the MARS 2024 sample, England

Amino acid substitution	Frequency	Percentage
ParC: S83N	6	37.5%
ParC: G81C	3	18.8%
ParC: S84P	3	18.8%
ParC: D82N	1	6.2%
ParC: N91I	1	6.2%
ParC: P62S (susceptible) and G81C (mixed)	1	6.2%
ParC: S83I and S83N (mixed)	1	6.2%

Source: Data from MARS sentinel surveillance programme.

References

1. Deguchi T, Maeda S I, Tamaki M, Yoshida T, Ishiko H, Ito M, Yokoi S, Takahashi Y,  Ishihara S. ‘Analysis of the gyrA and parC genes of mycoplasma genitalium detected in first-pass urine of men with non-gonococcal urethritis before and after fluoroquinolone treatment’. Journal of Antimicrobial Chemotherapy. Volume 48, Issue 5, pages 742 to 744. 2001

2. Murray G L, Bradshaw C S, Bissessor M, Danielewski J A, Garland S M, Jensen J S, Fairley C K , Tabrizi S N. ‘Increasing macrolide and fluoroquinolone resistance in mycoplasma genitalium’. Journal of Emerging Infectious Diseases. Volume 23, Issue 5, pages 809 to 812. 2017

3. Hamasuna R, Le P T, Kutsuna S, Furubayashi K, Matsumoto M,  Ohmagari N,  Fujimoto N, Matsumoto T, Jensen J S. ‘Mutations in parC and gyrA of moxifloxacin-resistant and susceptible mycoplasma genitalium strains’ PLOS ONE. Volume 13, Issue 6, article e0179572. 2018

Acknowledgements

The MARS Team would like to thank the collaborating centres for their continued support, SHSs for the prompt submission of clinical data and laboratories for sending isolates to the national STIRL at the UKHSA, Colindale.

Collaborating centres

• 10 Hammersmith Broadway (S Harish, O Perry, M Rayment, M Varadarajan)
• Archway Sexual Health Clinic and Mortimer Market Centre (R Browne, L Lau, M Pearson, C Thorpe)
• Barking Community Hospital (A Umaipalan)
• Burrell Street Sexual Health Clinic (S Keegan, A Nori)
• Bristol Royal Infirmary (J Lim, M Molosiwa)
• Cambridgeshire Community Services NHS Trust integrated Contraception and Sexual Health (R Acosta, S Basavaraj, S Clapp, N David, I Hawkins, E Hodges, M Gupta, D Johnson, K Mariyappa, H Pintilie, D Raha, C Sewell, P Williams)
• Dean Street Clinic (F Lander, S Patel, G Whitlock)
• Derwent and Solway Clinics (A Hodgson)
• Homerton Hospital (P Horne)
• Jefferiss Wing Centre for Sexual Health (G Garcia, J Husband)
• Royal Liverpool University Hospital (E Clarke)
• John Hunter Clinic (E Brunicki, J Lopez)
• Luton and Dunstable Hospital (K Zyla)
• Royal Berkshire Hospital (R Smart)
• Royal Sussex County Hospital (S Soni)
• The Garden Clinic (J Pakia)
• The Hathersage Centre (S Buckley)
• Ambrose King and Sir Ludwig Guttman Centres (E Chung, L Dufaur, J Gaddie, R Sheriff, C Sood, N Uthayakumar)
• University Hospital Birmingham (V Marshall, J Phattey)

Authors: Kirsty Bennet, James Johnson, Sandhya Vivekanand, Sandra David, Penny Cliff, Anna Vickers, Rachel Pitt-Kendall, Michelle Cole, Hamish Mohammed, Katy Sinka, Sarah Alexander, Helen Fifer.

Suggested citation

Bennet KF, Johnson JEC, Vivekanand, S, David S, Vickers A, Cliff P, Pitt-Kendall R, Cole MJ, Mohammed H, Sinka K, Alexander S, Fifer H, and contributors. Mycoplasma genitalium Antimicrobial Resistance Surveillance (MARS) report: 2024 November 2025, UKHSA, London.