Background Current recommended treatment for Mycoplasma genitalium (Mg) is azithromycin. Macrolide resistance mutations (MRM), predominantly on the 23SrRNA gene of Mg, have been found to be associated with failure of azithromycin. We aimed to determine the efficacy of 1g-azithromycin in a prospective cohort of Mg-infected STI clinic attendees, and to determine the contribution of MRM to treatment failure.

Method We commenced an observational study in July 2012 in which symptomatic patients diagnosed with Mg by PCR at Melbourne Sexual Health Centre are retested for Mg 14 and 28 days following treatment with 1g-azithromycin. Testing for MRM using high-resolution melt analysis (HRM) is conducted on day 0 and on positive samples at days 14 and 28. Participants are managed on the basis of clinical symptoms and not detection of MRM. Study will complete, May 2013.

Results 105 participants have been recruited; 89 have completed all study requirements. 48/89 (54%; 95% CIs 44–64%) participants were Mg PCR negative at day 28. 41/89 (46%; 95% CIs 36–56%) did not respond to 1g azithromycin: 11/41 (27%) had a persistently positive Mg PCR on day 28 without reported risk of re-exposure (presumptive failures) and 30 of 41 (73%) had persistent symptoms of MG prior to day 28 and required interim treatment with moxifloxacin (probable failures). Of the 41 failures, 40 (98%) had MRM detected: 30 (75%) at baseline and 10 (25%) at day 14 only. Of the 48 azithromycin-responders 4(8%) had MRM detected at baseline.

Conclusion The azithromycin cure rate for Mg in this clinic cohort was only 54%. MRM were detected in virtually all cases of azithromycin-failure, and were uncommon in azithromycin-responders. The majority of MRM were detected prior to treatment. These findings have implications for the use of macrolides as current recommended treatment for M.genitalium, and highlight the need for evaluation of alternative treatment approaches.