Infectious Disease · Medical Literature
Cefazolin for Methicillin-Susceptible Staphylococcus aureus Bacteremia.
Abstract
Staphylococcus aureus bacteremia is associated with high mortality. Whether cefazolin or an antistaphylococcal penicillin should be preferred for the treatment of methicillin-susceptible S. aureus bacteremia is unclear. In an ongoing international Bayesian adaptive platform trial, we conducted an open-label, randomized comparison of cefazolin with an antistaphylococcal penicillin (flucloxacillin or cloxacillin) in adult patients with penicillin-resistant, methicillin-susceptible S. aureus bacteremia. The primary outcome, which was evaluated with a hierarchical Bayesian logistic-regression model, was death from any cause within 90 days after enrollment in the platform. We assessed the posterior probability of the noninferiority of cefazolin to flucloxacillin or cloxacillin (with the criterion for noninferiority prespecified as an adjusted odds ratio of <1.2, which approximates an absolute difference in mortality of <2.5 percentage points if mortality in the antistaphylococcal-penicillin group is 15%), as well as the posterior probability of superiority (with the criterion of an adjusted odds ratio of <1.0). Secondary safety outcomes included the development of acute kidney injury within 14 days. This domain of the ongoing trial was conducted between February 17, 2022, and August 7, 2024, by which time the criterion for noninferiority had been met. Mortality at 90 days among adults who could be evaluated was 15.0% (97 deaths among 645 patients) in the cefazolin group and 17.0% (109 deaths among 642 patients) in the antistaphylococcal-penicillin group (adjusted odds ratio, 0.81; 95% credible interval, 0.59 to 1.12; probability of noninferiority, 99.2%; probability of superiority, 89.8%). Acute kidney injury occurred in 92 of 660 patients (13.9%) in the cefazolin group, as compared with 127 of 648 (19.6%) in the antistaphylococcal-penicillin group (adjusted odds ratio, 0.67; 95% credible interval, 0.50 to 0.89; probability of superiority, 99.7%). In patients with methicillin-susceptible S. aureus bacteremia, cefazolin was noninferior to flucloxacillin or cloxacillin with respect to 90-day mortality and was associated with a lower incidence of acute kidney injury. (Funded by the National Health and Medical Research Council and others; SNAP ClinicalTrials.gov number, NCT05137119.).
Clinical relevance
In adults with penicillin-resistant, methicillin-susceptible Staphylococcus aureus (MSSA) bacteremia, is cefazolin noninferior — and potentially superior — to antistaphylococcal penicillins (flucloxacillin or cloxacillin) with respect to 90-day all-cause mortality, and does it carry a more favorable safety profile? __SNAPSHOT__ {"studyDesign":"Open-label, randomized controlled trial embedded within an international Bayesian adaptive platform trial (Staphylococcus aureus Network Adaptive Platform [SNAP] Trial); noninferiority design with prespecified superiority analysis","setting":"91 sites across 8 countries (Australia, Netherlands, Canada, United Kingdom, New Zealand, Singapore, Israel, South Africa); international, multicenter, academic and tertiary hospital settings","sampleSize":"n=1,341 (cefazolin: n=660 evaluable for AKI / 645 evaluable for mortality; antistaphylococcal penicillin: n=648 evaluable for AKI / 642 evaluable for mortality); trial was stopped early on August 7, 2024 after the prespecified noninferiority threshold was met at the fourth planned interim analysis","population":"Adult inpatients with penicillin-resistant, methicillin-susceptible S. aureus bacteremia; enrolled within 72 hours of index blood culture collection; median age 66 years; 31.4% women","intervention":"Cefazolin 2 g IV every 8 hours (or every 6 hours for critical illness or endocarditis), adjusted for renal function","comparator":"Flucloxacillin 2 g IV every 6 hours (or every 4 hours for critical illness or endocarditis) OR cloxacillin 2 g IV every 4 hours, each adjusted for renal function; allocation by country per local availability","primaryOutcome":"All-cause mortality within 90 days after enrollment, evaluated via hierarchical Bayesian logistic-regression model; noninferiority criterion: adjusted odds ratio (aOR) <1.2 (approximating <2.5 percentage-point absolute mortality difference if comparator-arm mortality is 15%); superiority criterion: aOR <1.0","followUp":"90 days","fundingSource":"Government / non-profit: National Health and Medical Research Council (Australia) and other non-industry funders (ClinicalTrials.gov NCT05137119)"}
Key findings
PRIMARY OUTCOME — 90-day all-cause mortality: 15.0% (97/645) in the cefazolin group vs. 17.0% (109/642) in the antistaphylococcal penicillin group (aOR 0.81; 95% credible interval [CrI] 0.59–1.12; posterior probability of noninferiority 99.2%; posterior probability of superiority 89.8%). The prespecified noninferiority threshold (aOR <1.2) was met. Superiority was not formally declared per the Bayesian framework (threshold of aOR <1.0 not reached with ≥95% posterior probability), though the 89.8% probability of superiority is clinically noteworthy. The absolute mortality difference was 2.0 percentage points favoring cefazolin (calculated). NNT for one prevented death at 90 days: approximately 50 (calculated), assuming the observed absolute risk reduction of 2.0%. SECONDARY / SAFETY OUTCOMES — Acute kidney injury (AKI) within 14 days: 13.9% (92/660) with cefazolin vs. 19.6% (127/648) with antistaphylococcal penicillin (aOR 0.67; 95% CrI 0.50–0.89; posterior probability of superiority 99.7%); absolute risk reduction 5.7 percentage points (calculated); NNH for one prevented AKI: approximately 18 (calculated). Antibiotic discontinuation due to adverse events: 1.6% (cefazolin) vs. 9.1% (antistaphylococcal penicillin), an absolute difference of 7.5 percentage points favoring cefazolin (calculated from reported data). Posterior probability of a lower rate of renal replacement therapy initiation within 90 days: 94.6% with cefazolin. Acute liver injury within 14 days was reported as similar between groups (precise values not fully reported in available abstract/secondary sources). No significant between-group differences were observed in 14-, 28-, or 42-day mortality, microbiologic treatment failure, new foci of infection between days 15–90, or Clostridioides difficile infection within 90 days. Protocol-adherent (per-protocol) analysis: aOR 0.88 (95% CrI 0.61–1.26); the upper bound of the 95% CrI (1.26) exceeded the noninferiority margin of 1.20, representing a methodologic sensitivity concern. Endocarditis subgroup: directionally favorable for cefazolin (point estimate aOR 0.54), but credible intervals were wide (0.20–1.43) and the subgroup was underpowered; the probability of noninferiority in this subgroup dropped to 94%, insufficient to resolve the cefazolin inoculum effect question.
Practice implications
This trial established with high posterior probability (99.2%) that cefazolin is noninferior to flucloxacillin or cloxacillin for 90-day all-cause mortality in adults with MSSA bacteremia. The safety advantage for cefazolin was statistically robust and clinically meaningful: AKI risk was reduced by approximately 5.7 absolute percentage points (aOR 0.67; 99.7% posterior probability of superiority), and treatment discontinuation rates due to adverse events were markedly lower. These safety findings are likely clinically important and actionable. The trial did NOT establish formal superiority for mortality by prespecified criteria, though the 89.8% posterior probability of superiority provides meaningful directional evidence. Critically, findings apply to the enrolled population — hospitalized adults with confirmed MSSA bacteremia — and do not directly address nafcillin or oxacillin (the antistaphylococcal penicillins commonly used in the United States); the authors noted that available data suggest the adverse event profiles of these agents are broadly similar to flucloxacillin/cloxacillin, supporting cautious extrapolation. The endocarditis and CNS infection subgroups were substantially underpowered: the inoculum effect of cefazolin against high-burden infections remains unresolved and warrants dedicated investigation before cefazolin can be unequivocally recommended in these specific contexts. Patients with penicillin-susceptible isolates were enrolled in a separate SNAP silo (published concurrently in The Lancet) and are not covered by these results. Practice in countries where cefazolin is already first-line (e.g., much of North America) is confirmed by this trial; in countries where flucloxacillin or cloxacillin has been standard (e.g., United Kingdom, Netherlands, Australia), a transition to cefazolin as default therapy appears strongly justified for most adults with MSSA bacteremia. NOTE: This summary is based on the published abstract and high-quality secondary reporting; specific subgroup data, baseline characteristics, and secondary endpoints may differ in the full published manuscript. __LIMITATIONS__ Open-label design: lack of blinding may have introduced ascertainment or management bias, particularly for subjective endpoints and decisions regarding antibiotic discontinuation.,Post-randomization clinical management (source control, adjunctive antibiotics, diagnostic intensity) was not standardized or fully captured, limiting inference about the contribution of these factors.,Early stopping for efficacy: the trial enrolled approximately one-fifth of the maximum planned sample (1,341 of up to 6,000), which, while efficient by Bayesian adaptive design, may lead to overestimation of treatment effects and limits power for subgroup analyses.,Protocol-adherent analysis upper CrI (1.26) exceeded the noninferiority margin (1.20), raising a technical concern about robustness under the most conservative analytical assumption.,Endocarditis and CNS infection subgroups were critically underpowered; the cefazolin inoculum effect question in high-inoculum infections is unresolved.,Comparators were flucloxacillin and cloxacillin only; nafcillin and oxacillin (used in the U.S.) were not studied directly; extrapolation of safety findings to these agents requires caution.,Geographic restriction of the available antistaphylococcal penicillins limits direct applicability to certain health systems.,International practice variation in concomitant care (echocardiography use, infectious disease consultation rates, source control approaches) may confound results and limit generalizability across all sites equally.,Clinical site variation across 91 sites in 8 countries may introduce heterogeneity in outcome ascertainment.,The 90-day mortality endpoint captures deaths unrelated to MSSA bacteremia; a concurrent SNAP adjudication study found that approximately 60.7% of 90-day deaths were likely not SAB-related, which may attenuate the observed treatment effect toward the null. __BOTTOM_LINE_CATEGORY__ practice_changing __BOTTOM_LINE__ The SNAP trial — the largest randomized study ever conducted in S. aureus bacteremia — demonstrates that cefazolin is noninferior to flucloxacillin or cloxacillin for 90-day mortality in adults with MSSA bacteremia, with a clinically important and statistically robust reduction in acute kidney injury (13.9% vs. 19.6%; aOR 0.67; 99.7% posterior probability of superiority) and substantially lower rates of treatment-limiting adverse events. For most hospitalized adults with MSSA bacteremia, cefazolin should now be considered the preferred agent given its equivalent efficacy and superior safety profile. Caution is warranted in endocarditis and other high-inoculum infections pending dedicated evidence, and the safety comparison to nafcillin/oxacillin (used in the U.S.) requires indirect extrapolation.