β-Lactam Inoculum Effect in Methicillin-Susceptible Staphylococcus aureus Infective Endocarditis.

Importance: Infective endocarditis (IE) caused by Staphylococcus aureus is associated with high mortality, approximately 20% to 30%, mostly in the first month, with no improvement in recent decades. Current opinion is that antistaphylococcal penicillin and cefazolin are equally effective in treating methicillin-susceptible S aureus (MSSA) IE, and both are recommended as possible first-line treatments. Most MSSA strains carry the β-lactamase blaZ gene, and some blaZ-positive strains exhibit an inoculum effect, meaning increased minimum inhibitory concentrations at high inoculum. This reduced susceptibility to an antibiotic at high bacterial inoculum may be particularly relevant in IE, where vegetations have very high bacterial densities. Objective: To evaluate the association between phenotypic characteristics of S aureus isolates, β-lactam used, and outcome in patients with MSSA IE. Design, Settings, and Participants: This retrospective case series included MSSA cases treated at 3 French university hospitals between February 2016 and February 2022. The study included patients who had clinical isolates available and had definite or possible S aureus IE that involved native or prosthetic valves. Data were analyzed from July 2023 to June 2024. Main Outcomes and Measures: MSSA isolates were tested for the presence of blaZ and for inoculum effects to cefazolin and oxacillin. The association between first-month mortality and the β-lactam used, the presence of blaZ, and the presence of an inoculum effect to the treatment received was evaluated. Results: This study included 216 patients with MSSA IE (median [IQR] age, 65 [49-73] years; 152 [70.4%] male) who were treated with antistaphylococcal penicillin (139 [64.4%]) or cefazolin (77 [35.6%]). One-month mortality of left-sided IE was 44 of 180 patients (24.4%), with no overall difference between patients treated with antistaphylococcal penicillin or cefazolin. However, 1-month mortality was higher in patients infected with blaZ-positive strains than with blaZ-negative strains (38 of 129 [29.5%] vs 6 of 51 [11.8%]; P = .01), and with strains with an inoculum effect to the β-lactam received than with strains without an inoculum effect (25 of 62 [40.3%] vs 13 of 67 [19.4%]; P = .005). On multivariable analysis, the presence of an inoculum effect was independently associated with first-month mortality (HR, 2.84; 95% CI, 1.28-6.30; P = .01). Conclusions and Relevance: In this case series of MSSA IE, the presence of an inoculum effect to the β-lactam received was a risk factor for death in the first month. Phenotyping MSSA isolates for inoculum effect may guide β-lactam choice and improve outcomes.