The Lancet July 11, 2026

Benzylpenicillin versus flucloxacillin or cloxacillin for the treatment of penicillin-susceptible Staphylococcus aureus bacteraemia (SNAP): an international, multicentre, open-label, non-inferiority randomised controlled trial

Joshua S Davis et al. (Staphylococcus aureus Network Adaptive Platform (SNAP) Trial Group)

Bottom Line

The SNAP trial demonstrated that benzylpenicillin halved the risk of acute kidney injury and showed a high probability of non-inferiority for mortality compared to flucloxacillin or cloxacillin in adults with penicillin-susceptible Staphylococcus aureus bacteraemia.

Key Findings

1. 90-day all-cause mortality occurred in 14% (21/152) of patients in the benzylpenicillin group compared to 22% (26/121) in the flucloxacillin or cloxacillin group (adjusted OR 0.67, 95% CrI 0.35–1.28).
2. The 96.1% posterior probability of non-inferiority for mortality did not formally meet the prespecified stopping threshold (>99%) due to early study closure, though it showed an 88.9% posterior probability of superiority.
3. Acute kidney injury (AKI) occurred in 11% (17/153) of the benzylpenicillin group versus 22% (27/124) in the flucloxacillin/cloxacillin group (adjusted OR 0.50, 95% CrI 0.26–0.94), demonstrating a 98.4% posterior probability of superiority for safety.
4. 28-day mortality numerically favored benzylpenicillin at 8% (13/156) versus 13% (16/124) for the comparator arm (adjusted OR 0.77, 95% CrI 0.36–1.62).

Study Design

Design
RCT
Open-Label
Sample
281
Patients
Duration
90 days
Median
Setting
8 countries
Population Adults (≥18 years) with penicillin-susceptible Staphylococcus aureus (PSSA) bacteraemia.
Intervention Benzylpenicillin (intravenous).
Comparator Flucloxacillin or cloxacillin (intravenous).
Outcome All-cause mortality at 90 days.

Study Limitations

The open-label design may have influenced subjective clinical decisions; for example, the change of antibiotics due to perceived inefficacy was higher in the benzylpenicillin group (7%) versus the comparator group (1%).
Early termination of this randomization domain meant the primary mortality endpoint narrowly missed the formal >99% posterior probability threshold required to officially declare non-inferiority.
A small amount of missing 90-day primary outcome data (8 patients total) necessitated sensitivity analyses, although these supported the main findings.

Clinical Significance

Although the strict prespecified non-inferiority criterion for mortality was not formally met due to early trial closure, the highly probable comparable efficacy—combined with a halved risk of acute kidney injury—positions benzylpenicillin as the preferred and safer therapy over flucloxacillin or cloxacillin for adults with penicillin-susceptible S. aureus (PSSA) bacteraemia.

Historical Context

For decades, the vast majority of S. aureus isolates produced beta-lactamase, rendering them resistant to native penicillin. Consequently, anti-staphylococcal penicillins (e.g., flucloxacillin, cloxacillin, nafcillin) became the empiric and definitive standard of care. Recently, however, penicillin-susceptible S. aureus (PSSA) has re-emerged globally, accounting for up to 25% of S. aureus bacteraemia isolates in some regions. With the advent of reliable phenotypic susceptibility testing, clinicians sought to return to native benzylpenicillin due to its theoretically superior pharmacokinetic profile and lower expected nephrotoxicity. The platform SNAP trial was designed to replace historical empiricism with rigorous evidence for exactly these types of critical S. aureus infections.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

What is the difference in mechanism and spectrum between benzylpenicillin and flucloxacillin, and why might flucloxacillin be more nephrotoxic?

Key Response

Both are beta-lactams that inhibit cell wall synthesis. Benzylpenicillin (Penicillin G) is susceptible to beta-lactamase, hence only effective for penicillin-susceptible S. aureus (PSSA). Flucloxacillin is an antistaphylococcal penicillin resistant to staphylococcal beta-lactamase, used for MSSA. Antistaphylococcal penicillins are associated with higher rates of acute interstitial nephritis (AIN) compared to natural penicillins, which explains the higher AKI risk observed in the trial.

Resident
Resident

When managing a patient with MSSA bacteremia whose isolate is later found to be penicillin-susceptible, how should the SNAP trial findings influence your management regarding organ toxicity?

Key Response

The SNAP trial supports actively de-escalating from flucloxacillin or cloxacillin to benzylpenicillin for PSSA bacteremia. Residents should recognize that de-escalation in this context not only narrows the antimicrobial spectrum but also actively reduces patient harm by halving the risk of acute kidney injury without compromising survival.

Fellow
Fellow

How does the historical concern for the 'inoculum effect' and cryptic beta-lactamase production in PSSA isolates influence the interpretation of the SNAP trial's non-inferiority results for mortality?

Key Response

Historically, ID specialists avoided penicillin for PSSA due to the 'inoculum effect' (high bacterial burdens overcoming the drug) and the risk of undetected weak beta-lactamase producers leading to false susceptibility. The high probability of non-inferiority in mortality suggests that modern standard phenotypic testing is sufficient and these theoretical concerns do not translate to worse clinical outcomes in appropriately identified PSSA bacteremia.

Attending
Attending

Given the clear safety benefit of reduced AKI and non-inferior efficacy, should benzylpenicillin become the standard of care for all PSSA bacteremia, and what logistical barriers exist for implementation?

Key Response

The trial strongly argues for benzylpenicillin as the standard of care for PSSA to prevent nephrotoxicity. However, attendings must navigate operational barriers: many microbiology labs do not routinely test or report penicillin susceptibility for S. aureus, assuming MSSA treatment is sufficient, and benzylpenicillin requires frequent dosing or continuous infusion, complicating outpatient parenteral antimicrobial therapy (OPAT).

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The SNAP trial utilized a non-inferiority design for mortality. How does the choice of the non-inferiority margin and the intention-to-treat versus per-protocol analysis impact the validity of the conclusions?

Key Response

In non-inferiority trials, setting an appropriately narrow margin is critical to avoid accepting an inferior treatment. Furthermore, unlike superiority trials where intention-to-treat (ITT) is conservative, ITT in non-inferiority trials can bias results toward the null (non-inferiority) due to non-adherence or crossovers. A rigorous critique must verify that both ITT and per-protocol analyses demonstrate non-inferiority to ensure the mortality claim is robust.

Journal Editor
Journal Editor

As a peer reviewer, how do you critically evaluate the open-label nature of this trial when assessing the secondary endpoint of acute kidney injury (AKI)?

Key Response

While mortality is a hard, objective endpoint, AKI is subject to diagnostic and surveillance bias. In an open-label trial, clinicians aware that a patient is on a potentially nephrotoxic drug (flucloxacillin) might check serum creatinine more frequently or manage fluids differently, potentially skewing AKI detection rates. Reviewers must scrutinize the protocol's standardization of lab draws to ensure the AKI difference is a true pharmacological effect.

Guideline Committee
Guideline Committee

Current IDSA and ESCMID guidelines for S. aureus bacteremia focus on antistaphylococcal penicillins or cefazolin for MSSA. How should the SNAP trial data shift the GRADE recommendations for PSSA specifically?

Key Response

Current guidelines mention penicillin for PSSA but lack strong RCT evidence advocating its active use over antistaphylococcal penicillins. The SNAP trial provides high-quality (Level 1) evidence that de-escalation to benzylpenicillin reduces AKI by 50% without compromising efficacy. Guidelines should be updated with a strong recommendation for routine penicillin susceptibility testing in MSSA and subsequent de-escalation to benzylpenicillin to minimize drug-induced nephrotoxicity.

Clinical Landscape

Noteworthy Related Trials

2018

ARREST Trial

n = 758 · Lancet

Tested

Adjunctive rifampicin alongside standard antibiotic therapy

Population

Adults with Staphylococcus aureus bacteraemia

Comparator

Standard antibiotic therapy plus placebo

Endpoint

Bacteriologically confirmed treatment failure, recurrence, or death within 12 weeks

Key result: Adjunctive rifampicin provided no overall benefit over standard therapy and increased the frequency of adverse drug interactions and events.
2020

CAMERA2 Trial

n = 352 · JAMA

Tested

Standard MRSA therapy plus an antistaphylococcal beta-lactam (e.g., flucloxacillin)

Population

Adults with methicillin-resistant Staphylococcus aureus (MRSA) bacteraemia

Comparator

Standard MRSA therapy (vancomycin or daptomycin) alone

Endpoint

Composite of mortality, persistent bacteremia, relapse, or treatment failure at 90 days

Key result: Adding a beta-lactam did not improve composite outcomes and significantly increased the risk of acute kidney injury.
2024

SABATO Trial

n = 213 · Lancet Infect Dis

Tested

Early switch to oral antimicrobial therapy after 5 to 7 days of intravenous treatment

Population

Patients with low-risk Staphylococcus aureus bloodstream infection

Comparator

Standard continuous intravenous therapy

Endpoint

Composite of SAB-related complications or death within 90 days

Key result: Early oral switch was safely non-inferior to prolonged standard intravenous therapy for low-risk SAB patients.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis