Effectiveness and Safety of Ceftriaxone Compared to Standard of Care for Treatment of Bloodstream Infections Due to Methicillin-Susceptible Staphylococcus aureus: A Systematic Review and Meta-Analysis

(1) Background: Ceftriaxone is a potential alternative for the treatment of methicillin-susceptible Staphylococcus aureus (MSSA) bloodstream infections (BSIs) in acute care and outpatient parenteral antimicrobial therapy (OPAT) settings. We evaluated the effectiveness and safety of ceftriaxone for the treatment of MSSA BSIs. (2) Method: We searched PubMed, Embase, and Cochrane Library from their inception to October 30th 2021. Our outcomes included clinical cure, microbiological cure, 30- and 90-day mortality, 90-day hospital readmission, and adverse drug reactions (ADRs). We compared ceftriaxone against standard of care (SOC) therapy. We used the random-effects model for the meta-analysis, and our estimated effects were reported as odds ratios (ORs) with 95% confidence intervals (CI). (3) Results: Twelve retrospective cohort studies were included, comprising 1037 patients in the ceftriaxone arms and 2088 patients in the SOC arms. The clinical cure rate of the ceftriaxone regimen was not statistically different from SOC: OR 0.65 (95% CI: 0.29–1.45). Ceftriaxone was also not statistically different from SOC in microbiological cure: OR 1.48 (95% CI: 0.29–7.51); 30-day mortality: OR 0.79 (95% CI: 0.14–4.65); 90-day mortality: OR 0.82 (95% CI: 0.38–1.80); 90-day hospital readmission: OR 1.20 (95% CI: 0.92–1.56); and ADRs: OR 0.92 (95% CI: 0.39–2.18). (4) Conclusion: Ceftriaxone could provide an alternative for the treatment of MSSA BSIs in acute care and OPAT settings (except in patients whose BSIs were due to infective endocarditis).


Introduction
Staphylococcus aureus is one of the most common pathogenic Gram-positive bacteria, and it causes a wide range of community-acquired and hospital-acquired infections [1,2]. S. aureus expresses several virulence mechanisms and can cause serious infections associated with mortality that can be as high as 22-48% [3,4]. In the United States (U.S.), S. aureus accounts for 23% of bloodstream infections (BSIs) [1,5]. Patient groups at increased risk for S. aureus BSIs include individuals aged 70 years and older, individuals with HIV, those who inject drugs, and patients on hemodialysis [3].
The usual length of treatment for methicillin-susceptible S. aureus (MSSA) BSI is a minimum of 14 days for an uncomplicated infection and up to 4-6 weeks for complicated infection [1,3]. Specifically, parenteral anti-staphylococcal penicillin and firstgeneration cephalosporin antibiotics are the mainstay of therapy for MSSA BSI. While nafcillin, oxacillin, and cefazolin are the standard of care (SOC) for MSSA BSI, ceftriaxone can be administered once daily compared to nafcillin or oxacillin (6 times daily) and cefazolin (3 times daily), and it can be given peripherally by intravenous (IV) bolus injection [6]. Thus, ceftriaxone is convenient to use, especially in outpatient parenteral antimicrobial therapy (OPAT) settings without hospitalization [7]. Delivery options in OPAT settings include home-based, infusion center-based, and skilled nursing facility-based.
Several studies have evaluated ceftriaxone for the treatment of MSSA BSIs [8][9][10]. However, these studies have conflicting results and limited sample sizes; thus, robust evidence is inconclusive. A recent systematic review and meta-analysis and another systematic review evaluated ceftriaxone use for MSSA infections [11,12]. However, these studies evaluated mixed infections and did not include all of the available studies. The objective of this study is to evaluate the evidence for the effectiveness and safety of ceftriaxone for MSSA BSI in acute care and OPAT settings qualitatively utilizing a systematic review and quantitatively by meta-analysis.

Materials and Methods
We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and the Meta-Analyses of Observational Studies in Epidemiology (MOOSE) guidelines [13,14]. PRISMA and MOOSE checklists are available in Tables S1 and S2, respectively.

Literature Source
We performed a systematic search in PubMed, Embase, and Cochrane Library from their inception to 30 October 2021. We executed the search with the following concepts: ceftriaxone, BSI, and MSSA. The complete search strategy is available in Table S3. In addition, we manually searched the citations of key studies.

PICOS Criteria and Study Selection
We searched using the following modified Population, Intervention, Comparator, Outcomes, and Studies (PICOS) criteria [15]. (1) Population: adult patients aged 18 years and older with MSSA BSIs; (2) intervention: ceftriaxone antibiotic; (3) comparator: antistaphylococcal antibiotics (nafcillin, oxacillin, cloxacillin, and cefazolin); (4) outcomes: clinical cure; microbiological cure; 30-day mortality; 90-day mortality; 90-day hospital readmission; and adverse drug reactions (ADRs); (5) studies: randomized controlled trials and observational studies. Study investigators (Y.S.A. and G.B.) reviewed the titles and abstracts of all identified studies, and potentially eligible studies were retrieved for full-text review. Studies that met the PICOS criteria were included in the systematic review and meta-analysis. We excluded case reports, case series, editorials, commentaries, letters, studies with BSIs caused by pathogens other than MSSA, and publications in a language other than English.

Data Extraction
Two reviewers (Y.S.A. and G.B.) independently screened the titles and abstracts of all the identified studies in the databases for eligibility. We resolved discrepancies through discussion and consensus; unresolved matters were reviewed by a third person (F.S.). We extracted the following information for the included studies using a predeveloped worksheet: author and publication date, location and date of the study, sample size, sources of BSIs, and antibiotic regimens used. For studies that reported clinical/microbiological failure, we calculated the clinical/microbiological cure as the total number of patients minus the number of patients who failed.

Quality Assessment
The quality evaluation of the included studies was performed using the star system of the Newcastle-Ottawa Scale (NOS) for observational studies [16]. The studies were categorized based on their individual score: low quality (1-3 stars), medium quality (4-6 stars), and high quality (7-9 stars).

Summary Measures and Statistical Analysis
The meta-analytic computations were performed using STATA 17, StataCorp LLC, College Station, TX, USA. The pooled estimated effects were reported as odds ratios (ORs) with 95% confidence intervals (CI). We used the restricted maximum likelihood random-effects model to account for heterogeneity in the effect sizes of the studies [17]. We used Cochrane's Q to calculate the heterogeneity by the weighted sum of squares, and the I 2 statistic was used to express the percentage of variation due to heterogeneity [18]. Statistically significant heterogeneity was set at a threshold of p < 0.05 for Cochrane's Q statistic and > 30% for I 2 statistic. We assessed publication bias with funnel plots of standard error against the estimated effect. We used the Egger linear regression test method to evaluate the asymmetry of the funnel plot [19].

Subgroup Analysis
To evaluate the clinical cure rates based on the acuity level of the patients, we performed a subgroup analysis based on the treatment setting of either acute care or OPAT.

Study Selection
The initial search yielded 54 relevant abstracts in PubMed, 160 abstracts in Embase, and 0 abstracts in Cochrane Library. There were three abstracts identified from the review of citations and conferences. After removing duplicates, 50 abstracts were screened. We further excluded three abstracts that were deemed irrelevant based on the review of the titles. In total, 47 abstracts were reviewed. After removing 26 irrelevant studies and 1 in-vitro study, 20 articles underwent a full-text review for eligibility. Based on the full-text review, eight articles were excluded and a total of 12 studies were included in the systematic review, while 11 studies were included in the meta-analysis ( Figure 1).

Clinical Cure
Seven studies [8,10,[22][23][24]26,28] reported data on clinical outcomes and cure rate for patients with MSSA BSIs that included 235 patients in the ceftriaxone arms and 427 patients in SOC arms. The pooled clinical cure rate when using a ceftriaxone regimen was not statistically different from SOC: OR 0.65 (95% CI: 0.29-1.45; Figure 2A). In the subgroup analysis for acute care settings, five studies [8,10,22,23,26] were included that comprised 198 patients in the ceftriaxone arms and 364 patients in SOC arms. Similarly, the clinical cure when using a ceftriaxone regimen was not statistically different from SOC: OR 0.75 (95% CI: 0.30-1.91; Figure 2A).

Microbiological Cure
Three studies [8,9,22] reported data on microbiological cures for patients with MSSA BSIs. The studies included 210 patients in the ceftriaxone arms and 169 patients in the SOC arms. The pooled microbiological cure rate when using a ceftriaxone regimen was not statistically different from SOC: OR 1.48 (95% CI: 0.29-7.51; Figure 2B).

Mortality Due to Treatment Failure
Five studies [8-10,20,22] reported data on mortality secondary to the treatment failure when using a ceftriaxone regimen for the treatment of MSSA BSI. For the assessment

Mortality Due to Treatment Failure
Five studies [8][9][10]20,22] reported data on mortality secondary to the treatment failure when using a ceftriaxone regimen for the treatment of MSSA BSI. For the assessment of 30-day mortality, three studies [8,20,22] were included with 256 patients in the ceftriaxone arms and 205 patients in SOC arms. The pooled OR when using a ceftriaxone regimen was not statistically different from SOC: OR 0.79 (95% CI: 0.14-4.65; Figure 3A). Moreover, for the assessment of 90-day mortality, we included three studies [9,10,22] with 223 patients in the ceftriaxone arms and 184 patients in the SOC arms. In the assessment of 90-day mortality, the pooled OR when using a ceftriaxone regimen was not statistically different from SOC: OR 0.82 (95% CI: 0.38-1.80; Figure 3B).

90-day Hospital Readmission Due to Treatment Failure
Four studies [10,21,26,28] comprising 603 patients in the ceftriaxone arms and 1657 patients in the SOC arms reported data on hospital readmission due to treatment failure for patients with MSSA BSIs. The pooled OR when using a ceftriaxone regimen was not statistically different from SOC: OR 1.20 (95% CI: 0.92-1.56; Figure 4A). for the assessment of 90-day mortality, we included three studies [9,10,22] with 223 patients in the ceftriaxone arms and 184 patients in the SOC arms. In the assessment of 90day mortality, the pooled OR when using a ceftriaxone regimen was not statistically different from SOC: OR 0.82 (95% CI: 0.38-1.80; Figure 3B).

90-day Hospital Readmission Due to Treatment Failure
Four studies [10,21,26,28] comprising 603 patients in the ceftriaxone arms and 1657 patients in the SOC arms reported data on hospital readmission due to treatment failure for patients with MSSA BSIs. The pooled OR when using a ceftriaxone regimen was not statistically different from SOC: OR 1.20 (95% CI: 0.92-1.56; Figure 4A).

Adverse Drug Reactions
Five studies [9,10,22,26,28] comprising 333 patients in the ceftriaxone arms and 368 patients in the SOC arms reported data on ADRs for patients with MSSA BSIs. The rate of ADRs when using a ceftriaxone regimen was not statistically different from SOC: OR 0.92 (95% CI: 0.39-2.18; Figure 4B).

Publication Bias for the Clinical Cure
The funnel plot of standard error against the effect estimate reveals the overall symmetry of the included studies ( Figure S1). Additionally, the Egger test did not reveal statistically significant publication bias (p = 0.77).

Publication Bias for the Clinical Cure
The funnel plot of standard error against the effect estimate reveals the overall symmetry of the included studies ( Figure S1). Additionally, the Egger test did not reveal statistically significant publication bias (p = 0.77).

Discussion
Administration of ceftriaxone for MSSA BSIs is convenient because it is a once-daily antibiotic regimen [9]. In this systematic review and meta-analysis, we evaluated the use of ceftriaxone for the treatment of MSSA BSI. We evaluated clinical and microbiological cures, 30-and 90-day mortality, 90-day hospital readmission, and ADRs and found that

Discussion
Administration of ceftriaxone for MSSA BSIs is convenient because it is a once-daily antibiotic regimen [9]. In this systematic review and meta-analysis, we evaluated the use of ceftriaxone for the treatment of MSSA BSI. We evaluated clinical and microbiological cures, 30-and 90-day mortality, 90-day hospital readmission, and ADRs and found that ceftriaxone use was non-inferior to SOC in clinical effectiveness and safety outcomes that were evaluated. Our meta-analysis included five studies [8,10,22,23,26] that were conducted in acute medical care settings, and ceftriaxone maintained the clinical effectiveness for those patients based on our subgroup meta-analysis of acute care settings.
Methicillin-susceptible S. aureus BSIs can be associated with high morbidity and mortality, and thus using an effective antibiotic therapy is warranted [1]. Our findings suggest that ceftriaxone could provide an alternative treatment for MSSA BSI, and this is consistent with several studies. Barber [22] evaluated MSSA BSIs), and they concluded that ceftriaxone was effective against MSSA infections [11]. Notably, an ongoing clinical trial is evaluating the efficacy and safety of ceftriaxone home therapy for MSSA infections and coagulate-negative staphylococcal infections [30]. Patients are randomly assigned to ceftriaxone or SOC (cloxacillin, cefazolin, or daptomycin) and evaluated at a 6-month follow-up.
Our findings expand the results of a recent meta-analysis by Yetmar et al. that compared ceftriaxone use with SOC in mixed MSSA infections [12]. Our findings were consistent with the meta-analysis by Yetmar et al., since they found that ceftriaxone was not different from SOC in the assessment of 90-day mortality (OR 0.93, 95% CI: 0.46-1.88) and hospital readmission (OR 0.96, 95% CI: 0.57-1.64). However, Yetmar et al. contradicted our finding of ADRs, since the authors found that ceftriaxone had statistically significantly lower ADRs (OR 0.49, 95% CI: 0.27-0.88). A likely explanation is that our meta-analysis included studies in both in-patient acute care and OPAT settings, while Yetmar et al. included studies only in OPAT settings with a variety of infections. Therefore, the difference in study design may explain the better tolerability of ADRs in the Yetmar et al. study as patients in OPAT care are generally more stable. Lastly, Yetmar et al. evaluated the recurrence of all MSSA infections (not only BSIs), and ceftriaxone was not different from SOC (OR 1.04, 95% CI: 0.63-1.72) [12]. Other methodological differences compared to our study were as follows: Yetmar et al. included studies in OPAT settings, excluded 6 studies published in conferences as abstracts, and did not evaluate clinical and microbiological cures. Overall, Yetmar et al. included seven studies: five studies [8][9][10]22,24] evaluated BSIs, one study [31] evaluated osteoarticular infections, and one study [29] evaluated mixed infections. Our systematic review and meta-analysis included 12 studies restricted to ceftriaxone use only for MSSA BSIs, and we evaluated six outcomes.
Since ceftriaxone exerts its pharmacodynamic action on bacteria by protein unbound form, concerns exist due to the high protein binding of ceftriaxone (85-95%) that may result in treatment failure [6,32]. However, the high protein binding of ceftriaxone does not appear to impact the clinical effectiveness based on our results from the meta-analysis. A recent in vitro study by Heffernan et al. evaluated ceftriaxone pharmacodynamics and optimal dosing regimens against MSSA isolates using an infection model [33]. The authors evaluated multiple ceftriaxone dosing regimens (1 g once daily, 2 g once daily, 1 g twice daily, and 2 g twice daily), and found that only high-dose ceftriaxone (2 g twice daily) achieved either sustained bacterial inhibition or killing in the first 24 h of therapy. Currently, the labeling information recommends using a ceftriaxone dose of 2 g twice daily only for patients treated for bacterial meningitis empirically with other appropriate antibiotics and patients treated for infective endocarditis due to enterococcal species in combination with ampicillin [6]. Therefore, Heffernan et al. concluded that ceftriaxone at the routinely used doses may not be a reliable choice, and alternative antibiotics should be used for MSSA infections [33]. The study by Heffernan et al. should be interpreted with caution since it is an in vitro study and the ceftriaxone dosing regimens have not been validated clinically. Furthermore, several studies [10,21] investigated predictors of treatment failure that may lead to mortality or hospital readmission and found that incomplete antibiotic course, lack of source control, metastatic foci of infections, heart failure, infective endocarditis, critical illness, and obesity are associated with MSSA BSI treatment failure.
Notably, we were unable to perform some additional analyses because of the limited available data. Specifically, in addition to the ADRs analyzed above, Clostridioides difficile infection was reported by two studies [10,26], but it was inadequate to perform a metaanalysis. The rate of C. difficile infection was not statistically different between ceftriaxone and SOC: Carr et al. (ceftriaxone vs. SOC: 3% vs. 5.3%) and Mohamed et al. (ceftriaxone vs. SOC: 5.7% vs. 5%). We also attempted to evaluate MSSA BSI recurrence, but the number of studies was inadequate to perform a meta-analysis. The rate of infection recurrence was not statistically different between ceftriaxone and SOC: Carr et al. [10] (ceftriaxone vs. SOC: 12.1% vs. 5.3%), and Barber et al. [8] (ceftriaxone vs. SOC: 10% vs. 13%).
Regarding study limitations, all of the included studies are observational in nature, so the quality of data may be suboptimal, and there is a risk for bias. However, in our quality assessment using the NOS, 11 of the included studies were deemed of high quality and 1 was deemed of medium quality. There was also heterogeneity in the included studies due to different sources of MSSA BSIs. However, the random-effects model was used in the meta-analysis to adjust for the heterogeneity. Some studies did not report the complete regimen information for ceftriaxone or SOC. We also could not evaluate ceftriaxone efficacy based on patient age groups. Additionally, infectious disease consultation is associated with improved BSI treatment outcomes, reduced in-hospital mortality, and shorter hospitalization [35], but we could not evaluate the impact of infectious disease consultation due to a lack of data. Moreover, we could not evaluate the effectiveness of ceftriaxone based on the source of the MSSA BSI because only two studies [9,22] performed subgroup analyses to evaluate outcomes based on the source of MSSA BSI. Patients with infective endocarditis as a source of BSI were underrepresented in the ceftriaxone arms, and thus our findings do not support ceftriaxone use for those patients. Lastly, we included only studies published in the English language.

Conclusions
Ceftriaxone is a viable option for the treatment of MSSA BSIs in acute care and OPAT settings (except for patients with infective endocarditis as a source of BSI). Administration of ceftriaxone for MSSA BSIs is convenient because it is a once-daily antibiotic administered as a peripheral IV bolus injection. To our knowledge, the present study is the first systematic review and meta-analysis for ceftriaxone use specifically for the treatment of MSSA BSI. The study can inform clinicians that ceftriaxone was non-inferior to SOC for the treatment of MSSA BSI in terms of clinical cure, microbiological cure, mortality, hospital readmission, and ADRs. Clinicians should take into consideration patients' factors that necessitate the dose adjustment of ceftriaxone or using an alternative agent to optimize therapy.