Background When choosing empirical anti-microbial therapy for non-gonococcal urethritis (NGU) and cervicitis, efficacy against Chlamydia trachomatis over Mycoplasma genitalium is prioritised. However, M.genitalium is associated with reproductive sequelae in women and first-line recommended therapy with macrolide antibiotics differs to that for Chlamydia infection. We determined: M.genitalium and C.trachomatis frequency among symptomatic men with and without urethritis; prevalence of macrolide and fluoroquinolone associated genotypic resistance in M.genitalium; and the phylogenetic spread of genotypic M.genitalium antimicrobial resistance using a validated dual-locus typing system.

Methods Urethritis was diagnosed by combining urethral smear and clinical criteria. Nucleic acid amplification was used for detecting M.genitalium and C.trachomatis. Single nucleotide polymorphisms (SNPs) in the 23S ribosomal RNA gene (23S rRNA), in gyrA, gyrB, and parC were detected by DNA sequencing. MG191 SNP typing and MG309 variable number tandem analysis were utilised to assess M.genitalium strain diversity.

Results 217 men were recruited. C.trachomatis and M.genitalium prevalence was 14.7% (95% CI: 7.8–21.6) and 16.7% (95% CI: 9.5–24.0) respectively in NGU cases and both significantly higher than in those with no urethritis. 9/22 (41%; 95% CI: 20%–62%) of M.genitalium strains had markers of macrolide associated genotypic resistance. Of 15 M.genitalium strains analysed only one possessed a parC mutation, associated with fluoroquinolone resistance. Dual-locus typing assigned all M.genitalium strains to two major clusters, both of which contained diverse strains carrying resistant mutations. All strains were phylogenetically dispersed among international reference controls, typed using MG191.

Conclusion Frequency of M.genitalium was as high as C.trachomatis in NGU patients. More than 40% of M.genitalium strains had SNPs associated with macrolide resistance but fluoroquinolone associated genotypic resistance was rare. All strains were phylogenetically dispersed indicating that these infections were probably not part of a local clonal expansion. Treatment guidelines for NGU require re-appraisal in light of these findings.