Albanir Laier Bordignon, Felipe Lopes Grillo, André Rochinski Busanello, Newton Sérgio de Carvalho
Mycoplasma genitalium is an increasingly drug-resistant sexually transmitted infection, with widespread macrolide resistance rendering azithromycin frequently ineffective. Moxifloxacin, the main second-line treatment, is now threatened by fluoroquinolone resistance linked to the parC gene. Resistance-guided therapy, which tailors treatment to macrolide-resistance mutations detected after nucleic acid amplification testing, shows promise, but its pooled cure rate has not been quantified. Following a protocol registered in PROSPERO (CRD420251077720), we searched five databases for studies of resistance-guided therapy reporting cure (negative test-of-cure 14-90 days after treatment); 11 studies were eligible. Cure proportions were pooled with a random-effects Freeman-Tukey model with Hartung-Knapp adjustment and, as a co-primary approach, a generalized linear mixed model; we performed sensitivity analyses, a temporal meta-regression of moxifloxacin cure, and stratification by parC mutation. Pooled cure was 94.6% (95% confidence interval [CI] 92.4-96.4) and 93.9% (95% CI 91.7-95.5) across the two models, with no heterogeneity (I² = 0%) and robust in sensitivity analyses. However, moxifloxacin cure declined over time (odds ratio per calendar year 0.92; 95% CI 0.85-0.99; p = 0.033), driven by the parC S83I mutation (98% cure in wild-type versus 38% in mutants). In a modeled high-resistance Brazilian scenario (79% macrolide resistance), expected cure with empirical azithromycin fell to approximately 22%, versus approximately 93% with resistance-guided therapy. The evidence derives predominantly from a single center and a short-term surrogate outcome, with low-to-very-low certainty. Resistance-guided therapy achieves high short-term cure and should be prioritized over empirical therapy, yet its efficacy is time-limited and calls for parC-guided regimens as resistance evolves.