Abstract
OBJECTIVES: Benzylpenicillin with ceftriaxone is used for outpatient antimicrobial therapy of Enterococcus faecalis infective endocarditis (EFIE) due to poor stability of aminopenicillins. The aim was to correlate the impact of benzylpenicillin-ceftriaxone synergy on EFIE treated with benzylpenicillin-ceftriaxone, and investigate whether simpler phenotypic methods can predict synergy. METHODS: Clinical outcomes of a retrospective cohort of EFIE patients were correlated with treatment and synergy. Isolates were assessed for synergy using the checkerboard method, compared with double disc diffusion (DDD) and layered and crossed gradient diffusion strip (GDS) tests, with sensitivity and specificity of these methods calculated. RESULTS: Thirty-eight episodes of EFIE in 34 patients were included; the majority received benzylpenicillin-ceftriaxone alone (n = 16) or in sequence (n = 12; any benzylpenicillin-ceftriaxone n = 28), and 10 received other regimens. There was no statistical difference between any benzylpenicillin-ceftriaxone versus other therapies on outcomes, nor between benzylpenicillin-ceftriaxone synergy and outcome. GDS was an unreliable predictor of checkerboard synergy; DDD was reasonable. Five isolates lacked benzylpenicillin-ceftriaxone synergy (one also lacking ampicillin-ceftriaxone synergy); these all had high-level ceftriaxone resistance [zone diameter 6 mm, or broth microdilution (BMD) MIC >512 mg/L]. Most isolates from relapse cases developed reduced zone diameters to ceftriaxone. Two isolates lacking synergy had the same mutation near the ceftriaxone response regulator (CroR) binding site in the pbp4 promoter region. CONCLUSIONS: Ceftriaxone susceptibility, either MIC <512 mg/L by BMD or disc zone diameter >6 mm, is the best predictor of ampicillin-ceftriaxone and particularly benzylpenicillin-ceftriaxone synergy in E. faecalis. There was no clear relationship between the absence of benzylpenicillin-ceftriaxone synergy and outcome in this highly comorbid cohort.