Answers are grounded in this paper and cite their sources.
(See the Major Article by McDanel et al on pages 100–6.) Antistaphylococcal penicillins (ASPs) such as nafcillin, oxacillin, cloxacillin, flucloxacillin, and cefazolin have been long available to treat methicillin-susceptible Staphylococcus aureus (MSSA) bacteremia; nevertheless, optimal therapy remains unresolved. MSSA strains, of which >90% produce β-lactamase (Bla), continue to be an important cause of bacteremia, which is associated with 30-day mortality rates of 20%–22% [1–4]. The therapy debate began in the 1970s when investigators comparing cefazolin to methicillin and nafcillin against MSSA found the latter to be more resistant to hydrolysis by staphylococcal Bla and devoid of an inoculum effect (a marked increase in minimum inhibitory concentrations [MICs] determined with a high vs a lower inoculum of MSSA) [5]. Thereafter, cefazolin failures in the treatment of MSSA endocarditis were associated with MSSA strains producing Bla that efficiently degraded cefazolin [6–9]. These strains demonstrated an inoculum effect wherein MICs determined with an inoculum of 107 colony-forming units (CFU)/mL were multiple-fold higher than those obtained with a standard 105 CFU/mL inoculum – often exceeding 32 µg/mL and in the nonsusceptible range [5, 8]. Animal model studies did not resolve the efficacy question. In the rabbit endocarditis model, Goldman and Petersdorf found cefazolin to be ineffective against a MSSA strain that exhibited a marked inoculum effect and hydrolyzed cefazolin [10]. In contrast, in this model Carrizosa and colleagues found cefazolin to be effective and comparable to Bla-stable drugs, cephalothin and methicillin, against several MSSA strains that rapidly degraded cefazolin [11, 12]. Using this model with a MSSA strain devoid of an inoculum effect or Bla hydrolysis of cefazolin, Steckelberg et al, with doses chosen to mimic human therapy, found nafcillin to be more effective than cefazolin [13]. In a mouse intraperitoneal infection model with a Bla-producing MSSA strain that displayed a cefazolin inoculum effect, Chapman and Steigbigel found cefazolin to be ineffective at low doses, but at higher doses it was effective and comparable to nafcillin [14]. Recently the potential in vivo consequence of Bla hydrolysis of cefazolin and the inoculum effect has been suggested more strongly. An MSSA strain, which had been recovered from a patient with native valve endocarditis who failed cefazolin therapy, produced a type A Bla that rapidly degraded cefazolin and produced a striking inoculum effect (an MIC with an inoculum of 5 × 105 CFU/mL of 0.5 µg/mL increased to 128 µg/mL with an inoculum of 5 × 107 CFU/mL). Notably, this strain did not degrade nafcillin nor demonstrate an inoculum effect [8]. Curing this MSSA strain of its Bla gene produced an isogenic derivative that no longer produced Bla nor a cefazolin inoculum effect but that retained infectivity comparable to the parent strain in the rabbit endocarditis model [15]. In that model compared to nafcillin, cefazolin treatment with doses chosen to simulate human dosing was ineffective against infection caused by the Bla-producing parent strain. In contrast, cefazolin treatment was effective and comparable to nafcillin against infection caused by the Bla-negative derivative MSSA strain. These findings suggest that production of type A Bla and the resultant inoculum effect may be associated with reduced cefazolin efficacy. Among MSSA clinical isolates collected from multiple continents, 11%–36% display a cefazolin inoculum effect [16–20]. It is most common in isolates producing type A and C Bla; however, type A Bla is more often associated with cefazolin MIC increases to nonsusceptible levels [18, 19]. While MSSA isolates exhibiting the cefazolin inoculum effect have been associated with failure of cefazolin treatment of MSSA bacteremia, the findings have not been statistically significant [19, 21]. Nevertheless, the vulnerability of cefazolin to type A Bla from MSSA producing a cefazolin inoculum effect and the frequency of clinical isolates with this phenotype engender concern regarding cefazolin treatment of patients with MSSA bacteremia, particularly bacteremia from infections with high bacterial loads. Concern regarding the potential therapeutic inferiority of cefazolin vs ASP in the treatment of severe MSSA infections remains apparent in current expert guidelines [22, 23]. Comparative efficacy studies have attempted to resolve the cefazolin vs ASP treatment of MSSA bacteremia controversy. These studies have been propelled by cefazolin’s favorable pharmacokinetics, convenient posthemodialysis dosing, cost savings, and, in comparison to nafcillin, significantly fewer adverse drug events with resultant premature discontinuation of therapy [24]. Paul et al found 90-day mortality after definitive cefazolin treatment (n = 72) of MSSA bacteremia not to be significantly different from that with cloxacillin (n = 281), with an adjusted odds ratio (aOR) of 0.91 (95% confidence interval [CI], .47–1.77) [25]. However, in this study patients dying within 7 days of the positive blood culture were excluded. Assessing MSSA bacteremia treated with cefazolin 3 g/day (n = 105) or cloxacillin 8–12 g/day (n = 249), Bai et al found a cefazolin adjusted hazard ratio (aHR) for 90-day mortality of 0.60 (95% CI, .36–1.0; P = .056) relative to cloxacillin and of 0.58 (95% CI, .31–1.08; P = .085) when assessing cases matched by propensity scores for predominant antibiotic therapy [26]. In another study comparing cefazolin (n = 70) to nafcillin (n = 30) treatment of MSSA bacteremia, the aOR for 90-day mortality was 0.40 (95% CI, .09–1.74; P = .22) [27]. Three additional studies compared cefazolin to ASP treatment for MSSA bacteremia. Using a composite definition of treatment failure as the primary endpoint, no study found antimicrobial therapy to be significantly associated with treatment failure [28–30]. While these studies do not demonstrate significant differences in outcome of MSSA bacteremia treated with cefazolin vs ASPs, they are retrospective, include relatively small populations (having often excluded large numbers of patients), and are focused on definitive treatment after a variable period of empiric therapy. Despite statistical adjustments, these studies are unavoidably subject to recognized and unrecognized selection biases. Furthermore, endocarditis and deep-seated infections are underrepresented (and often skewed to ASP therapy) in the study populations [25–27, 30], and overall mortality rates are often less than expected [26–28, 30]. Consequently, caution is required in generalizing from these data to patients with more threatening MSSA bacteremia—a caveat specifically noted by some authors [26, 27]. In this issue of Clinical Infectious Diseases, McDanel et al, working with a cohort of 11154 patients with MSSA bacteremia, retrospectively compare definitive treatment (days 4–14 after the initial positive blood culture) with cefazolin (n = 1163) to ASP (n = 2004) [31]. Cefazolin was associated with a 37% and 23% reduction in all-cause mortality at 30 and 90 days, respectively (aHR 0.63 [95% CI, .51–.78] and 0.77 [95% CI, .66–.90], respectively). The authors cautiously suggest that cefazolin be considered for definitive therapy of patients with bacteremic MSSA infection not involving the central nervous system. Other than ASP toxicities that are fatal or truncate therapy, how do we explain the striking and biologically somewhat implausible greater efficacy of cefazolin vs ASPs? Can we generalize from these data to the definitive therapy answer? I hesitate to do so. Despite the authors’ vigorous efforts to adjust statistically for baseline severity of infection, comorbidity, and clinical indications, confounding by recognized and unrecognized variables may persist. Furthermore, this study included only 28% of the available MSSA bacteremia cohort. Potentially important data were unavailable: the efficacy of empiric therapy, source control, the adequacy of antibiotic dosing, and total duration of antibiotic treatment (that after hospitalization). Most significantly, can we generalize from these data to treatment for those infections at highest risk for cefazolin failure, those caused by MSSA with the type A Bla–cefazolin inoculum effect phenotype? Of particular concern are infections with high bacterial burden and limited source control. Notably, endocarditis was infrequent in the study population (cefazolin 52 [4%] and ASP 168 [8%] patients) and other high-bacterial-burden, deep-seated infections were not defined. Aside from dosing β-lactams to insure optimal pharmacodynamics, where do we go from here? Although high bacterial load infection caused by MSSA with the type A Bla–cefazolin inoculum effect phenotype may be infrequent, it seems prudent to consider Bla-stable ASP the therapy of choice for the most threatening infections. Cefazolin, which is clearly effective therapy for many MSSA bacteremic infections and offers important benefits, is appropriate therapy for uncomplicated MSSA bacteremia and for the consolidation/completion phase of treatment for complicated life-threatening MSSA infection after reduction of bacterial burden by ASP and source control. Comparative effectiveness studies of cefazolin and ASP treatment of complicated bacteremia caused by MSSA with the type A Bla–cefazolin inoculum effect phenotype may further refine selection of therapy. Ultimately, prospective randomized studies comparing cefazolin and ASP treatment of complicated, high-bacterial-burden MSSA bacteremic infections are needed for a definitive answer. Potential conflicts of interest. Author certifies no potential conflicts of interest. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
No takes yet. Share an insight, caveat, or question.
Adolf W. Karchmer (2017) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: