Mycoplasma genitalium Antimicrobial Resistance in Community and Sexual Health Clinic Patients, Auckland, New Zealand

Our retrospective study compared genotypic antimicrobial resistance in Mycoplasma genitalium–positive specimens collected from 48 community and 33 sexual health clinic (SHC) patients. Macrolide resistance was similar in community (75%) and SHC (76%) patients. We observed no significant difference in fluoroquinolone resistance between community (19%) and SHC (27%) patients (p = 0.66).


DISPATCHES
Our retrospective study compared genotypic antimicrobial resistance in Mycoplasma genitalium-positive specimens collected from 48 community and 33 sexual health clinic (SHC) patients. Macrolide resistance was similar in community (75%) and SHC (76%) patients. We observed no significant difference in fluoroquinolone resistance between community (19%) and SHC (27%) patients (p = 0.66).
Missense mutations in parC at amino acid positions 81, 83, or 87 conferred fluoroquinolone resistance (Table 2). Mutations in codon 83 of parC are associated with resistance; therefore, the mutation at T249A was presumed to confer fluoroquinolone resistance. The importance of polymorphisms in parC at C184T and C356T is uncertain. Mutations in gyrA at codon 95 were detected in 6 patients, 5 of whom were SHC attendees. All 6 patients harbored concomitant parC mutations that are associated with fluoroquinolone resistance, meaning that mutations in gyrA were not detected in the absence of parC mutations. The 5 patients with mutations in gyrA at G285A harbored a concurrent parC mutation at G248T, suggesting the strains were similar.
We detected dual macrolide and fluoroquinolone resistance in 20% (16/81) of patients (Table 1). Fluoroquinolone-resistant strains were likely to show concurrent macrolide resistance (89%), with the exception of strains from 2 community patients, which harbored only a fluoroquinolone resistance-associated mutation at codon 83 in parC.
Repeat specimens from 2 community patients were suggestive of macrolide resistance developing during the sampling period. For 1 patient, 2 of 3 urine samples received at 2-month intervals were negative for 23S rRNA mutations, with a mutation only detected in the most recent of the 3 samples. Another community patient was infected with a strain that harbored a parC mutation in codon 83 only; however, a 23S rRNA mutation was later detected in 2 subsequent urine samples collected at 1-month intervals, and the parC mutation persisted. A SHC attendee harbored both 23S rRNA and parC resistance-associated mutations initially, and a subsequent sample collected 5  months later harbored a mutation in gyrA at G285T in addition to the 23S rRNA and parC mutations.

Conclusions
Our findings imply that resistance is common in circulating M. genitalium stains in the general population and highlight the limited and declining antimicrobial options for treatment. Few studies have delineated antimicrobial resistance by referrer type, and our results imply that macrolide-and fluoroquinolone-resistant strains are endemic in sexual networks in the general population, rather than confined to, or disproportionately affecting, persons attending SHCs. This information might be useful at local and national levels for informing sexual health treatment guidelines, but a need exists for supranational monitoring and reporting of resistance, given that it varies between countries (13). Infections caused by strains resistant to both macrolides and fluoroquinolones occurred in 20% of patients, who would require alternative treatment options to obtain clinical and microbiological cure. Although pristinamycin has been successfully used to treat patients with multidrug-resistant infections (14), this treatment is not publicly funded in New Zealand and is only available as an imported medicine by special approval, underscoring the importance of exploring new treatment strategies to manage patients with resistant strains.
The presence of fluoroquinolone resistance in macrolide-sensitive strains in the community is concerning and might signify emergence of circulating fluoroquinolone-monoresistant strains in the region, with consequent implications for future treatment strategies. Although our data do not distinguish between descendants of clonal mutant lineages and de novo variants, we speculate that the introduction of a resistant clone from overseas is likely and that the use of fluoroquinolones in the community contributes to selective pressure for resistant strains.
A limitation of this study was that we did not have patient information regarding treatment of infection or clinical outcomes. This information might help establish whether resistance developed during treatment or occurred through reinfection with a more resistant strain. Another limitation was the inability to establish chain of transmission between patients, which was particularly relevant to the 4 SHC attendees and 1 community patient who harbored similar strains with identical mutations in both the gyrA and parC genes.
Mutations in the gyrB and parE genes act synergistically to increase fluoroquinolone resistance when detected with mutations in gyrA, parC, or both and might also warrant consideration when screening strains for markers of resistance (15). We found that in genotypically fluoroquinolone-resistant strains, single-nucleotide polymorphisms in gyrA were only detected with concomitant parC mutations. This finding might support diagnostic laboratories in efforts to target only the 23S rRNA and parC genes during their genotypic-resistance testing.
Funding was provided by an Auckland District Health Board Charitable Trust grant (research project no. A+7436).

About the Author
Ms. Vesty is a scientific officer at Auckland City Hospital and a PhD candidate at the University of Auckland, New Zealand. Her research interests include medical microbiology and the human microbiome.