The cheaper staph antibiotic just won the biggest trial ever run

The cheaper drug won.

The largest randomized trial ever run for Staphylococcus aureus bacteremia, the bloodstream infection that still kills roughly one in five patients within ninety days, just told infectious-disease doctors that the generic first-generation cephalosporin they have been politely sidelining for decades works as well as the expensive antistaphylococcal penicillin they were taught to reach for first, and is gentler on the kidneys. The trial answering the question is called SNAP. It took public money to run, because no one with a marketing budget had a reason to ask.

The unsettled question has shadowed infectious-disease practice for a generation. A patient comes in with S. aureus growing in the blood, the lab calls back to say it is methicillin-susceptible, and the team has to pick: a cheap, generic first-generation cephalosporin called cefazolin, or one of the antistaphylococcal penicillins like nafcillin and flucloxacillin, drugs rougher on the kidneys and routinely more expensive to deliver. Bench microbiology said the penicillins should be better in principle. Bedside experience said cefazolin seemed to work fine. The randomized trials that would actually settle it were never run.

Until SNAP.

The Staphylococcus aureus Network Adaptive Platform trial, whose first results began rolling out in 2025 and were summarized this week by MedPage Today, is the largest randomized controlled trial ever conducted for S. aureus bacteremia. Investigators across Australia, New Zealand, Singapore, Canada, Israel, South Africa, the United Kingdom and elsewhere in Europe enrolled 1,341 adults with methicillin-susceptible S. aureus bloodstream infections, the strain everyone agrees is curable in principle, and randomized them to either cefazolin or flucloxacillin. They followed everyone for ninety days. At that point, 15.0 percent of cefazolin patients had died, against 17.0 percent of flucloxacillin patients. Cefazolin was non-inferior on the primary outcome. The secondary signals leaned its way: less acute kidney injury, lower early mortality.

That phrase, non-inferior, is the polite trial-talk that masks something less polite once you sit with the rest of the data. Cefazolin did not just hold the line. It did the job with fewer complications, against the drug clinicians have been trained to consider the more rigorous choice.

Flucloxacillin and its American cousin nafcillin have spent decades on the staph-bacteremia top shelf because of an old in-vitro observation called the inoculum effect: at very high bacterial loads, certain cefazolin-degrading enzymes get a chance to express themselves, and cefazolin’s minimum inhibitory concentration creeps up in the lab. The worry was that in an actual patient, whose bloodstream infection is by definition a high-inoculum event, cefazolin might quietly fail. A 2018 cohort study in Clinical Infectious Diseases put numbers behind the worry, finding 30-day all-cause mortality more than doubled when the inoculum effect was present.

SNAP did not isolate inoculum-effect-positive strains as a subgroup; it tested the clinical question the laboratory worry raised, which is the question that actually mattered. With 1,341 randomized patients across a dozen countries and the largest pragmatic sample the field has ever assembled, what does cefazolin do in actual sick people with methicillin-susceptible disease? The answer was the unflashy one. The drug treated as a second-line option for decades worked at least as well as the first-line choice, and was easier on the kidneys.

Worth saying who paid for the work, because the answer is unusual. SNAP was funded by Australia’s National Health and Medical Research Council and an International Clinical Trial Collaborations grant from the Medical Research Future Fund. Public money, no pharmaceutical sponsor. Cefazolin is generic and old; there was no commercial partner who stood to gain from proving an off-patent drug was the right answer. The SNAP team, led from the Doherty Institute in Melbourne and Hunter Medical Research Institute at the University of Newcastle, are antibiotic-stewardship people whose incentives run the opposite direction from the industry’s. They had to build a global adaptive platform spanning eight countries to settle a question that should have been settled before most of the patients in their trial were born.

For American clinicians the practical wrinkle is that flucloxacillin is not routinely used here; the local equivalents are nafcillin and oxacillin, drugs with similar pharmacology and a similar reputation for being rough on the kidneys. SNAP did not test those drugs directly, but the clinical inference is the obvious one, and the same year produced two pieces of corroborating evidence pointed the same direction. CloCeBa, a French multicenter RCT published in The Lancet in 2025, randomized 292 adults with MSSA bacteremia to cefazolin or cloxacillin and found cefazolin non-inferior, with serious adverse events in 15 percent of the cefazolin arm versus 27 percent of the cloxacillin arm, and acute kidney injury in 1 percent of cefazolin patients versus 12 percent of cloxacillin patients. A 2025 systematic review and meta-analysis in Clinical Microbiology and Infection drew the same arrow on the pooled observational data. The evidence is converging from multiple directions on the same answer.

What it amounts to is a quiet, fully public-money course correction on a default that has been in textbooks for as long as most infectious-disease fellows have been alive. The drug that is cheaper, generic, easier on the kidneys, and now has the largest dataset behind it is the same drug clinicians have been politely sidelining for theoretical reasons. The footnote about the inoculum effect can stay in the textbook. The default it has kept in place no longer can.

Sources

1. MedPage Today – Cefazolin Proves Its Mettle for Methicillin-Susceptible Staph Infections (June 17, 2026)
2. NEJM Journal Watch / HIV and ID Observations – SNAP Trial Helps Resolve Long-Running Controversies Over Management of Staph Bacteremia (April 2025)
3. Clinical Infectious Diseases – The Staphylococcus aureus Network Adaptive Platform Trial Protocol: New Tools for an Old Foe (2022)
4. Doherty Institute – SNAP clinical trial 12-month milestone (NHMRC and MRFF funding)
5. Clinical Infectious Diseases – The Cefazolin Inoculum Effect Is Associated With Increased Mortality in MSSA Bacteremia (2018)
6. The Lancet – CloCeBa: Cloxacillin versus cefazolin for methicillin-susceptible Staphylococcus aureus bacteraemia (2025)
7. Clinical Microbiology and Infection – Cefazolin vs antistaphylococcal penicillins for MSSA bacteraemia: systematic review and meta-analysis (2025)
8. CMI Communications – Communicable Episode 26: SNAP out of it, rethinking antistaphylococcal penicillins for S. aureus bacteraemia (2025)