For roughly two decades, a single 1-gram dose of azithromycin was the standard treatment for many cases of urethritis and cervicitis. It was simple and reliable for chlamydia, and it was assumed to work well against a less familiar organism, Mycoplasma genitalium. That assumption no longer holds. This article explains what M. genitalium is, why cure rates with single-dose azithromycin fell from most patients to roughly two-thirds or fewer, what the genetic basis of that resistance is, and how the newer approach called resistance-guided sequential therapy restores cure rates above 90 percent in published studies. It also explains where the evidence is strong, where it is still developing, and what an informed patient can reasonably discuss with a physician.

An unusual organism with an unusual weakness

M. genitalium was first isolated in the early 1980s by researchers in London working with samples from men with non-gonococcal urethritis. It is one of the smallest self-replicating organisms known, with a genome of only about 580,000 base pairs. It grows extremely slowly in the laboratory, sometimes taking weeks or months, which is why it went largely undiagnosed for years. Routine detection only became practical with nucleic acid amplification tests, and the first such test was cleared by the US Food and Drug Administration in 2019.

Two features of the organism matter for treatment. First, it has no cell wall, so penicillins and cephalosporins, which attack cell-wall construction, are useless against it. Second, it carries only a single copy of the ribosomal RNA operon. Many bacteria carry several copies, so a resistance mutation in one copy is diluted by the others. In M. genitalium, one mutation is enough to make the entire organism resistant. That single fact goes a long way toward explaining what happened next.

The infection matters because it is a recognised cause of persistent urethritis in men and is associated with cervicitis and pelvic inflammatory disease in women. A 2015 meta-analysis led by researchers at the University of Washington and published in Clinical Infectious Diseases found associations with preterm birth and spontaneous miscarriage, as well as with infertility. These are associations from observational studies, not proof that the organism causes each outcome, and the size of the effect varied between studies. Even so, for families hoping to welcome children, and for the unborn children themselves, they are a sound reason to take accurate diagnosis seriously.

How azithromycin works, and how the organism escapes it

Azithromycin belongs to the macrolide family. It descends from erythromycin, first isolated in 1952 from a soil bacterium now called Saccharopolyspora erythraea, collected in the Philippines. Azithromycin itself was developed in the early 1980s by chemists at the Pliva company in Croatia, who modified erythromycin's ring structure to make it more stable and longer-lasting in tissue. It is a fair example of a pattern seen throughout medicine: some of our most useful tools began as chemistry already at work in the created world, waiting to be noticed and carefully refined.

Macrolides bind to the bacterial ribosome, the cell's protein-building machinery, at a site formed by a region of the 23S ribosomal RNA known as domain V. Blocking that site stops the organism from making proteins. Macrolide resistance in M. genitalium comes almost entirely from single-letter changes at two positions in that region, conventionally labelled A2058 and A2059 (using the reference numbering from E. coli). The most common substitutions are A2058G and A2059G. Each one changes the shape of the binding pocket enough that azithromycin no longer attaches effectively. Because the organism has only one copy of the gene, a single such change produces a fully resistant infection.

Why single-dose failures rose

Warning signs appeared early. In 2006, clinicians at the Melbourne Sexual Health Centre in Australia reported in Emerging Infectious Diseases that a single 1-gram dose failed in a substantial share of men with M. genitalium urethritis, close to three in ten. A 2008 study by Jørgen Skov Jensen of the Statens Serum Institut in Copenhagen and colleagues, published in Clinical Infectious Diseases, showed something more troubling: some infections that were susceptible before treatment carried resistance mutations afterward. The single dose was not merely failing; it was selecting for resistant organisms in the very patients being treated. Later studies suggested that roughly one in ten initially susceptible infections may acquire resistance this way, though estimates vary.

The pooled evidence then confirmed the trend. A 2015 systematic review and meta-analysis by Lau, Bradshaw and colleagues in Clinical Infectious Diseases found that the cure rate of the 1-gram dose fell from about 85 percent in studies before 2009 to about 67 percent in studies from 2009 onward. A 2020 systematic review in The Lancet Infectious Diseases by Machalek and colleagues estimated that the global prevalence of macrolide resistance rose from about 10 percent before 2010 to about half of infections by 2016–2017, with marked regional variation. Several factors likely contributed:

Some Scandinavian clinicians observed less resistance selection with an extended five-day azithromycin course. That evidence is observational and not settled, and the extended course does nothing for an infection that is already resistant.

How resistance-guided sequential therapy works

The answer developed by researchers, again prominently at the Melbourne Sexual Health Centre, was to stop guessing. Resistance-guided sequential therapy combines two steps: first lower the bacterial load, then choose the finishing drug according to the organism's actual genetic profile.

A prospective evaluation by Read and colleagues, published in Clinical Infectious Diseases in 2019, reported cure rates above 90 percent in both the azithromycin and moxifloxacin pathways, with far less new macrolide resistance than had been seen with single-dose treatment. A larger follow-up cohort from the same centre, reported by Durukan and colleagues in 2020, found similar results. These are well-conducted cohort studies rather than randomised trials, which is an honest limitation, but the consistency of the results and the clear biological rationale make the evidence reasonably strong. The 2021 sexually transmitted infection treatment guidelines from the US Centers for Disease Control and Prevention adopted this approach, and British and European guidelines use closely related strategies.

Patients should know that none of these drugs carries a specific FDA-approved indication for M. genitalium; their use here is off-label, though supported by guidelines and clinical evidence. Moxifloxacin also carries FDA boxed warnings for serious side effects, including tendon rupture, peripheral nerve damage, central nervous system effects and worsening of myasthenia gravis, along with cautions about heart rhythm and aortic aneurysm. It is generally avoided in pregnancy, which makes treatment during pregnancy genuinely difficult and a matter for careful, individual discussion with an obstetrician.

Limits, new resistance and the road ahead

Resistance-guided therapy is not the end of the story. Fluoroquinolone resistance, driven mainly by mutations in the parC gene and sometimes combined with gyrA changes, is rising, and is especially common in parts of the Asia-Pacific region, including Japan and China. Infections resistant to both macrolides and fluoroquinolones remain uncommon in most Western countries, but they are increasingly reported and hard to treat. Options for such cases include pristinamycin (not available in the United States), minocycline and, in Japan, sitafloxacin; evidence for these comes largely from small case series. Lefamulin, approved in the United States only for community-acquired pneumonia, shows activity against M. genitalium in laboratory studies and isolated case reports, but its use for this infection remains investigational.

Testing policy is equally important. Current guidelines generally recommend testing people with symptoms, such as persistent urethritis, cervicitis or pelvic inflammatory disease, rather than screening everyone without symptoms. The reasoning is that untargeted treatment of asymptomatic infections may drive further resistance without clear benefit.

What informed patients can do

Good stewardship of one's own health, and of the people one loves, includes understanding a diagnosis well enough to take part in decisions about it. Reasonable questions to raise with a physician include whether M. genitalium testing is appropriate for persistent symptoms, whether resistance testing is available, and what the risks and benefits of each drug are in one's own circumstances. Completing the full course, abstaining from sexual contact until both partners have finished treatment, and making sure a spouse or partner is evaluated protect the household from reinfection. Faithfulness within marriage remains the most dependable protection of all. No regimen guarantees cure, and some cases need retesting or second-line treatment, but an informed patient working closely with a trusted doctor is well placed to get the best available outcome.

Key takeaway: single-dose azithromycin lost its effectiveness against Mycoplasma genitalium because one ribosomal mutation is enough to confer resistance and the short course selected for it, and doxycycline followed by a drug chosen according to resistance testing now achieves cure rates above 90 percent in published studies.