Treatment Outcomes with Nafcillin versus Cefazolin for Methicillin-Susceptible Staphylococcus aureus Bloodstream Infections

Background: Historically, nafcillin (NAF) and oxacillin have been drugs of choice for the treatment of methicillin-susceptible Staphylococcus aureus (MSSA) infections. Cefazolin (CFZ) is an alternative in patients unable to tolerate antistaphylococcal penicillins, and data from small retrospective studies suggest that CFZ may have similar efficacy in the treatment of MSSA bloodstream infections (BSI). We compared rates of treatment failure and adverse effects associated with the use of NAF and CFZ for MSSA BSI at a large academic medical center.

Methods: Adult inpatients with ≥ 1 blood culture positive for MSSA between 7/1/2011 and 8/31/2015 were candidates for inclusion in this retrospective cohort study. Patients were included if they received ≥ 72 hours of NAF or CFZ as directed therapy after receiving ≤ 72 hours of any empiric therapy. Patients were excluded if they received antibiotics active against MSSA other than NAF or CFZ, had a polymicrobial index BSI, or had a BSI due to any pathogen within the previous 30 days. The primary endpoint was treatment failure, defined as clinician documentation of treatment failure for any reason, recurrence of MSSA infection within 30 days, or 30-day, all-cause, in-hospital mortality. Secondary outcomes included antibiotic-related acute kidney injury (AKI), acute interstitial nephritis (AIN), hepatotoxicity, and rash.

Results: Of 100 patients with MSSA BSI, 73 (73%) received NAF and 27 (27%) received CFZ. Baseline characteristics were similar among groups except CFZ patients had higher APACHE II scores (median 17 vs. 10; p < 0.01), a higher frequency of renal dysfunction (40.7% vs. 6.8%; p < 0.01), and a higher frequency of hemodialysis dependence (33.3% vs. 1.4%; p < 0.01). No differences in treatment failure (2.7% vs. 3.7%; p = 1.00) or clinical cure (76.7% vs. 70.4%; p = 0.60) rates were detected between the NAF and CFZ groups, respectively. Acute kidney injury was numerically more frequent in NAF patients (24.7% vs. 7.4%; p = 0.09).

Conclusion: No difference in treatment failure was observed between patients who received NAF or CFZ for MSSA BSI among this cohort. In light of the higher incidence of AKI among NAF patients, future studies evaluating CFZ as a preferred agent are warranted.