Mycoplasma genitalium and fluoroquinolone resistance: From genotype to phenotype

Dear Editor,

We have read with great interest the abridged original “Targeted antibiotic therapy in Mycoplasma genitalium (MG) infections: analysis of mutations associated with resistance to macrolides and fluoroquinolones”, published by Piñeiro et al.,¹ in which the prevalence of resistance to macrolides (16.3%) and fluoroquinolones (7.9%) is documented in a well-characterised cohort of 313 patients with MG infection. The work provides relevant information on the management of MG infection, even so, we would like to make some considerations.

On the one hand, in relation to the results obtained from the targeted therapy, a significant number of MG infections without initial mutations (13.9%) end up developing resistance after treatment with the macrolide azithromycin. This fact, in addition, is much more marked when azithromycin is used in single doses with respect to the extended regimen (31.3% vs. 7.0%; p < 0.001), which has already been described previously although with less consistency.² In this line, we would like to highlight a recent study conducted by Read et al., where a sequential strategy with doxycycline followed by the targeted antibiotic for MG, either azithromycin or fluoroquinolone moxifloxacin, is used for the treatment of non-gonococcal urethritis.³ Thus, the previous use of doxycycline, together with higher doses of azithromycin, achieved resistance selection in only 2.6% of MG infections. It is also worth mentioning that the use of doxycycline in dual therapy could reduce the occurrence of resistance in MG.⁴

Finally, we would like to expand the study of resistance to fluoroquinolones exposed by Piñeiro et al. In a research study published by our group,⁵ resistance-associated mutations were detected in 8.3% of infections, similar to the few studies carried out in other parts of Europe and presented in the supplementary table of the article by Piñeiro et al. (7.1%). However, as the authors point out, the presence of missense mutations in parC is not always associated with a therapeutic failure with fourth-generation fluoroquinolones, such as moxifloxacin (remember that other fluoroquinolones such as levofloxacin or ciprofloxacin are ineffective against MG).^4,6 There are few studies that correlate the presence of these mutations with the in vivo and in vitro activity of moxifloxacin. In 2013, Couldwell et al. recorded the first cases of therapeutic failure associated with the S83I mutation in ParC (MG numbering).⁷ Later, Murray et al. observed similar events associated with the S83I and S83R mutations.⁸ Finally, in a recent in vitro susceptibility study, Hamasuna et al. relate the S83I mutation, and subtly the D87Y mutation, with an increase in the minimum inhibitory concentration (MIC) of moxifloxacin.⁶ In addition, from an enzymatic-structural point of view, mutations in the S83 and D87 amino acids of ParC in MG can affect the antibiotic–enzyme interaction and thus confer resistance to moxifloxacin.⁹ In this sense, we consider that only the mutations in S83 and D87, and perhaps the D82N mutation due to its proximity to the antibiotic target region, should be considered as associated with resistance to fluoroquinolones in the study by Piñeiro et al. In contrast, the A69T mutations (previously described,⁶ and not associated with an increase in the MIC of moxifloxacin), S95N (conservative change) and T101I (far from the target region) should not a priori be included in the above estimate. This would give a prevalence of mutations associated with fluoroquinolone resistance of 4.9–5.6%, and not 7.9%, as stated in Piñeiro et al. It would also be interesting if the authors could clarify which mutation in ParC was associated with the only therapeutic failure to moxifloxacin that they recorded (first well documented case in Spain).

On the other hand, the presence of mutations in GyrA (especially in M95 and D99), although it has been associated with therapeutic failure,⁷ could have a less important effect associated with a reduction in the susceptibility of the bacterium to fluoroquinolone.^6,9

Currently, in Spain, due precisely to the lack of therapeutic alternatives, infections that do not respond to macrolides should probably be treated with moxifloxacin, regardless of mutations in parC or gyrA. Doxycycline, pristinamycin (not marketed in Spain) or the fourth generation fluoroquinolone sitafloxacin (not marketed in Spain, with increased activity against MG) could be some alternatives.^2,4,6,10

Conflicts of interest

None.