Authors: Kathleen P. Davis, Laura A. McDermott, David R. Snydman, Bree B. Aldridge
Categories: Research Article, MRSA, bacteremia, cefazolin, ceftaroline, combination therapy, rifampicin
Source: Microbiology Spectrum
Methicillin-resistant Staphylococcus aureus (MRSA) bacteremia is a serious clinical challenge with high mortality rates. Antibiotic combination therapy is currently used in cases of persistent infection; however, the limited development of new antibiotics will likely increase the need for combination therapy, and better methods are needed for identifying effective combinations for treating persistent bacteremia. To identify pairwise combinations with the most consistent potential for benefit compared to monotherapy with a primary anti-MRSA agent, we conducted a systematic study with an in vitro high-throughput methodology. We tested daptomycin and vancomycin each in combination with gentamicin, rifampicin, cefazolin, and oxacillin, and ceftaroline with daptomycin, gentamicin, and rifampicin. Combining cefazolin with daptomycin lowered the daptomycin concentration required to reach 95% growth inhibition (IC95) for all isolates tested and lowered daptomycin IC95 below the sensitivity breakpoint for five out of six isolates that had daptomycin minimum inhibitory concentrations at or above the sensitivity breakpoint. Similarly, vancomycin IC95s were decreased when vancomycin was combined with cefazolin for 86.7% of the isolates tested. This was a higher percentage than was achieved by adding any other secondary antibiotic to vancomycin. Adding rifampicin to daptomycin or vancomycin did not always reduce IC95s and failed to produce synergistic interaction in any of the isolates tested; the addition of rifampicin to ceftaroline was frequently synergistic and always lowered the amount of ceftaroline required to reach the IC95. These analyses rationalize further in vivo evaluation of three drug pairs for MRSA daptomycin+cefazolin, vancomycin+cefazolin, and ceftaroline+rifampicin.
Bloodstream infections caused by methicillin-resistant Staphylococcus aureus (MRSA) have a high mortality rate despite the availability of vancomycin, daptomycin, and newer antibiotics including ceftaroline. With the slow output of the antibiotic pipeline and the serious clinical challenge posed by persistent MRSA infections, better strategies for utilizing combination therapy are becoming increasingly necessary. We demonstrated the value of a systematic high-throughput approach, adapted from prior work testing antibiotic combinations against tuberculosis and other mycobacteria, by using this approach to test antibiotic pairs against a panel of MRSA isolates with diverse patterns of antibiotic susceptibility. We identified three antibiotic pairs—daptomycin+cefazolin, vancomycin+cefazolin, and ceftaroline+rifampicin—where the addition of the second antibiotic improved the potency of the first antibiotic across all or most isolates tested. Our results indicate that these pairs warrant further evaluation in the clinical setting.
KEYWORDS: MRSA, combination therapy, bacteremia, ceftaroline, cefazolin, rifampicin
Bacteremia caused by methicillin-resistant Staphylococcus aureus (MRSA) is a severe challenge in clinical settings (1). In a 2017 study of hospital-onset and community-onset cases from a large cohort of U.S. hospitals, MRSA was responsible for over half of all drug-resistant infections reported (2). Worldwide, MRSA caused more than 100,000 deaths attributable to AMR in 2019 (3). Despite advances in antibiotic therapy and control measures, mortality associated with MRSA bacteremia remains high, with estimates ranging from 20% to 30% (4).
Vancomycin and daptomycin are used as primary antibiotics for treating MRSA bacteremia, and both are used alone and as part of combination therapy (5). Ceftaroline is currently approved for monotherapy against acute bacterial skin and skin structure infections caused by MRSA, but not for MRSA bacteremia (6). However, ceftaroline salvage therapy for MRSA bacteremia has been reported to have a high success rate and is considered by some to be an attractive alternative (7, 8). Furthermore, a recent pilot study of MRSA bacteremia was terminated early with in-hospital mortality of 0% for the daptomycin+ceftaroline treatment group vs 26% for the daptomycin or vancomycin monotherapy group (0/17 vs 6/23 patients) (9), and a retrospective cohort study found a considerable, but non-statistically significant, decrease in mortality (4.3% vs 20.8%, P = 0.162) for patients with a primary endovascular source who received daptomycin+ceftaroline rather than the standard of care within 72 hours of index culture (10). Both of these studies indicate the potential benefit of starting daptomycin+ceftaroline therapy earlier in the course of infection, rather than as salvage therapy. A multicenter observational study (11) and a multicenter cohort study (12) also suggest a role for ceftaroline monotherapy against MRSA bacteremia, in particular where daptomycin or vancomycin cannot be used; however, randomized control trials would be necessary for further evaluation of this option.
Other antibiotics are used as secondary agents in combination therapy, usually with daptomycin or vancomycin, for MRSA bacteremia. Gentamicin is sometimes added to daptomycin or vancomycin (13) for potential synergy, and rifampicin is also sometimes used in combination therapy for antibiofilm activity (14) or the management of persistent MRSA bacteremia (4). Oxacillin and cefazolin are considered first-line therapies for methicillin-sensitive S. aureus (15). Despite the resistance of MRSA to oxacillin and cefazolin, the use of oxacillin in combination with daptomycin is supported in the case of MRSA bacteremia salvage therapy (16). This is not an exhaustive list of antibiotics for treating MRSA bacteremia. Linezolid is an oral option that is used for MRSA pneumonia and skin and soft tissue infections and has also been successfully used off-label as salvage therapy for MRSA bacteremia (5, 17). Clindamycin is another oral option for MRSA skin infections that can be used as a step-down therapy or in conjunction with other standard-of-care antibiotics for MRSA bacteremia (18, 19). Unlike gentamicin, rifampicin, cefazolin, and oxacillin, linezolid and clindamycin are not restricted to use as adjunctive therapy (5, 17).
Currently, the use of antibiotic combination therapy against MRSA bacteremia or endocarditis is recommended in the case of initial treatment failure (5, 16), including those cases in which vancomycin or daptomycin resistance is developed during treatment. However, studies assessing the use of daptomycin+ceftaroline prior to the necessity of salvage therapy (9, 10) indicate that in certain circumstances, starting combination therapy earlier may prove beneficial and that the potential for combination therapy may be underutilized. A better understanding of how combinations perform in vitro, including trends relating single-drug minimum inhibitory concentrations (MICs) to combination performance, will likely provide valuable information for the design of future in vivo and clinical studies (20).
To identify potentially underutilized combinations, we quantified combination performance across MRSA bacteremia isolates with a range of susceptibility profiles. To do this, we adapted a recently developed in vitro methodology for high-throughput antibiotic combination testing known as DiaMOND (Diagonal Measurement of N-way Drug interactions) (21–23). DiaMOND allows for the measurement of combination potency and interaction for large numbers of combinations at a fraction of the time and materials cost of traditional checkerboard assays. Using DiaMOND, we demonstrated the utility of a non-traditional metric for evaluating daptomycin combinations against two series of MRSA bacteremia isolates with decreasing daptomycin susceptibility. We validated these observations against additional MRSA bacteremia isolates with a variety of susceptibility profiles (Table S1) and expanded our combination set to include vancomycin and ceftaroline-based combinations. This allowed us to identify underutilized combinations that warrant additional in vivo investigation and demonstrate the different behavior of specific β-lactams in combination therapy.
We tested two series of MRSA bacteremia isolates from different patients, which both showed decreases in daptomycin susceptibility, as well as 29 additional isolates from different patients at Tufts Medical Center, selected to encompass a range of susceptibilities. Susceptibilities were determined by MIC testing as per Clinical Laboratory Standards Institute (CLSI) guidelines (24, 25) (Table S1). The two series of MRSA bacteremia isolates were tested with daptomycin in pairwise combination with ceftaroline, rifampicin, gentamicin, and cefazolin. The third isolate from the second patient (TR258), along with 29 additional isolates, was tested with an extended set of pairwise combinations. This extended set consisted of daptomycin and vancomycin, each with gentamicin, rifampicin, cefazolin, and oxacillin, and ceftaroline with daptomycin, rifampicin, and gentamicin. Combinations involving gentamicin or rifampicin were only tested against isolates with gentamicin MIC ≤8 µg/mL or rifampicin MIC ≤2 µg/mL, respectively.
For antibiotic testing involving daptomycin alone or in combination, isolates were grown in cation-adjusted Mueller–Hinton broth (CAMHB) supplemented with a total of 50 µg/mL Ca^2+^ (referred to as CAMHB+Ca^2+^), and for testing involving oxacillin alone or in combination, CAMHB was supplemented with 2% NaCl. The single drug IC95 values for each antibiotic correlated across all media conditions (CAMHB alone, CAMHB+Ca^2+^, CAMHB with 2% NaCl, and CAMHB+Ca^2+^ with 2% NaCl) used for testing that antibiotic alone and in combinations (shown in Fig. S1). For daptomycin in all media conditions and ceftaroline in CAMHB alone, the single drug IC95 values correlated (Pearson r ≥ 0.5, P-value ≤0.05) with the MIC values (shown in Fig. S2). This was not the case for all vancomycin MICs and single-drug IC95 values, or for ceftaroline MICs and single-drug IC95 values in CAMHB+Ca^2+^; however, for 77% of the vancomycin isolates tested, the single-drug IC95 values in any medium were within twofold of the MIC values.
We determined growth inhibition (from measured optical density at 600nm (OD600) values) of 10-step dose–response curves for single and pairwise combinations. Dose steps increased by 1.8× of the single drug or each drug in the pair. For single antibiotics except cefazolin or oxacillin, dose #5 of 10 was the antibiotic concentration required to reach 50% growth inhibition (IC50). For all pairwise combinations not involving cefazolin or oxacillin, dose #5 consisted of half of the IC50 of each antibiotic (shown in Fig. 1A). When cefazolin or oxacillin was tested with daptomycin or vancomycin, daptomycin and vancomycin were used at the same doses as when they were used alone, cefazolin and oxacillin were used at a constant 2 µg/mL, and growth inhibition was normalized to growth inhibition with cefazolin or oxacillin alone at 2 µg/mL (shown in Fig. 1B).
Fig 1 DiaMOND dosing schematics and combination interaction does not always indicate the greatest decrease in primary antibiotic required to reach the IC
95in combination. (A) DiaMOND dosing schematic for all combinations not involving oxacillin or cefazolin and (B) DiaMOND dosing schematic for all combinations involving oxacillin or cefazolin. See Materials and Methods for further description. (C) Dose–response curves showing % growth inhibition versus microgram per milliliter antibiotic for daptomycin alone (black circles with dashed line), ceftaroline alone (light aqua triangles with dashed line), daptomycin when used in combination with ceftaroline (dark aqua circles with solid line), and ceftaroline when used in combination with daptomycin (dark aqua triangles with solid line) against isolate TR399 (grown in CAMHB+Ca^2+^). (D) Dose–response curve fits showing % growth inhibition versus microgram per milliliter antibiotic for daptomycin alone (black circles with dashed line), gentamicin alone (red diamonds with dashed line), daptomycin when used in combination with gentamicin (dark red circles with solid line), and gentamicin when used in combination with daptomycin (dark red diamonds with solid line) against isolate TR399 (grown in CAMHB+Ca^2+^). Antibiotic DAP, daptomycin; VAN, vancomycin; CRF, ceftaroline; GEN, gentamicin; RIF, rifampicin; CZN, cefazolin; OXA, oxacillin.
Cultures were grown (at 37°C with shaking) overnight to saturation, diluted back to 500, and allowed to grow to mid-log (OD600 = 0.2–0.5), then used to make an OD600 = 0.001 dilution. Then, this dilution was added to 384-well plates (50 µL/well) to which drug stocks had been pre-added using an HP D300E digital drug dispenser. Plates were then incubated at 37°C with shaking for 24 hours, and OD600 values were measured using a Biotek plate reader.
Using our MATLAB analysis pipeline (22), we fitted Hill curves to the growth inhibition values for each dose–response curve and calculated antibiotic concentration and combination interaction at pre-chosen points below and up to the IC95. All drug interaction and potency results shown are an average of at least three biological replicates. All drug interaction and potency results shown in this paper are provided in Data S1.
DiaMOND (21–23) has been used to identify pairwise and higher-order synergistic antibiotic combinations against Mycobacterium tuberculosis and Mycobacterium abscessus. DiaMOND metrics have also been used to predict in vivo outcomes of drug combinations in mouse tuberculosis models. DiaMOND is based on measurements along the information-rich diagonal of a traditional checkerboard for combinations of antibiotics, which all possess sufficient activity against the bacterium in question (shown in Fig. 1A), or measurements where one antibiotic mainly acts as a sensitizer (shown in Fig. 1B). This allows for combination measurements to be done with a fraction of the number of doses (and time and materials) as a traditional checkerboard assay. For testing MRSA bacteremia isolates, we adapted the previously used DiaMOND protocol to use growth conditions and treatment times specified by CLSI for antibiotics used against MRSA (24) and modified our computational pipeline to report both concentration and antibiotic interaction metrics for all combinations at pre-specified points along the combination dose–response curve. We used a drug interaction metric known as the fractional inhibitory concentration (FIC) score at the 95% growth inhibition level, usually reported as a log2-transformed value (log2FIC95). The more negative the log2FIC95 score is for a combination tested against an isolate, the more synergistic the combination is against that isolate, and the more positive the log2FIC95 score is, the more antagonistic, with log2FIC95 scores close to zero indicating additivity. In some initial testing, we found that combination synergy was not always the best indicator of what combinations resulted in the greatest decrease in daptomycin, vancomycin, or ceftaroline required to reach the IC95 (example shown for daptomycin+ceftaroline and daptomycin+gentamicin in Fig. 1C and D). Thus, along with log2FIC95 scores, we always considered the fold change in daptomycin, vancomycin, or ceftaroline required to reach the IC95, when different secondary antibiotics were added, as this may be a more important metric for antibiotics that are concentration-dependent (such as daptomycin).
Given the challenge posed by decreasing daptomycin susceptibility during treatment (26), we wanted to understand the extent of differences in daptomycin-based combinations for restoring or circumventing decreased daptomycin activity. We tested daptomycin in pairwise combination with ceftaroline, cefazolin, gentamicin, or rifampicin against a series of MRSA bacteremia isolates from two different patients (A and B) (Fig. 2). For both patients, there was a decrease in daptomycin susceptibility for isolates taken over the course of treatment. Figure 2 shows the fold decrease in the amount of daptomycin required to reach the IC95 when tested with ceftaroline, cefazolin, gentamicin, or rifampicin against serial isolates from patient A (four isolates) and patient B (three isolates). Of the secondary antibiotics tested, only the addition of cefazolin lowers the amount of daptomycin required to reach the IC95 below 1 µg/mL for all seven isolates, including the isolates (third and fourth isolates from patient A, third isolate from patient B) with daptomycin MIC values of ≥1 µg/mL.
Fig 2 Adding cefazolin lowers the amount of daptomycin required to reach the IC
95, particularly as daptomycin susceptibility decreases over the course of treatment. This figure shows two series of MRSA bacteremia isolates (isolates are shown in the order that they were taken) from two different Tufts Medical Center patients during their courses of treatment. For patient A, daptomycin alone was started at 6 mg/kg between the second positive culture and the third positive culture. The third positive culture was obtained approximately 5 days after daptomycin was started. The fourth positive culture was obtained after approximately 12 days of daptomycin and 5 days of gentamicin. No antibiotic administration data were available for patient B. Both series of isolates showed a decrease in daptomycin susceptibility for the third isolate in the series, as seen in the daptomycin alone IC95(µg/mL) presented on the x-axis of the graph for each isolate. The graphs for each isolate show the fold decrease in daptomycin required to reach the IC95when cefazolin (blue), gentamicin (red), ceftaroline (light aqua), or rifampicin (green) are added to daptomycin. The third isolate from patient B is resistant to rifampicin, so it was not tested with daptomycin+rifampicin. See Table S1 for isolate and MIC info.
To understand to what extent the benefit of adding cefazolin to daptomycin (including relative to adding other secondary antibiotics to daptomycin) extends beyond these two series of isolates, we expanded our testing to additional MRSA bacteremia non-serial isolates from 29 different patients, with a range of susceptibility profiles (Fig. 3A). We also added daptomycin plus oxacillin in our set of combinations. Adding cefazolin to daptomycin always lowered the amount of daptomycin required to reach the IC95. Furthermore, compared to the other four antibiotics tested with daptomycin, adding cefazolin to daptomycin resulted in the greatest fold decrease in daptomycin required to reach the IC95 for 39% of the isolates tested (14 out of 36 total isolates). This was a higher percentage than was achieved by any of the other daptomycin combinations. The addition of cefazolin resulted in the largest decrease in daptomycin required to reach the IC95 for 5 of the 10 isolates with vancomycin MIC = 2 µg/mL and daptomycin MIC <1 µg/mL, 2 of the 4 isolates with daptomycin MIC ≥1 µg/mL and vancomycin MIC <2 µg/mL, and 1 of the 2 isolates with vancomycin MIC = 2 µg/mL and daptomycin MIC ≥1 µg/mL. The average decrease in daptomycin required to reach the IC95 when cefazolin was added was 0.29 µg/mL for all 36 isolates tested (0.28 µg/mL when considering only TR258 and the additional 29 non-serial isolates, see Table 1), and the average fold decrease in daptomycin was 1.92. Only the addition of gentamicin to daptomycin resulted in a greater average fold decrease (Table 1).
Fig 3 Cefazolin performs relatively well compared to other antibiotics in terms of decreasing but not increasing the amount of daptomycin or vancomycin required to reach the IC
95. (A) For the 30 MRSA bacteremia isolates tested with all DAP combinations (TR258 and the 29 non-serial isolates), this graph shows the percentage of isolates, specified on the y-axis, that achieved at least an x-Fold Change in Primary antibiotic (FCP), specified on the x-axis. For this graph, daptomycin is always the primary antibiotic, and percentages are indicated for the combination of daptomycin with ceftaroline (light aqua line), gentamicin (red line), cefazolin (blue line), rifampicin (green line), or oxacillin (brown line). Negative values on the x-axis indicate a fold decrease in daptomycin required to reach the IC95when the secondary antibiotic is added, and positive values on the x-axis indicate a fold increase in daptomycin required to reach the IC95when the secondary antibiotic is added. (B) This graph is the same style as the graph in (A), except this graph shows the results for the vancomycin-containing combinations.
We then tested isolate TR258 (the third isolate from patient B) and the 29 non-serial isolates with vancomycin in pairwise combination with gentamicin, rifampicin, cefazolin, and oxacillin (Fig. 3B). Adding cefazolin to vancomycin decreased the amount of vancomycin required to reach the IC95 for 26/30 (86.7%) isolates tested, while adding gentamicin to vancomycin decreased the amount of vancomycin required to reach the IC95 for 24/28 (85.7%) isolates tested (two isolates with gentamicin MIC >8 µg/mL were not tested). When cefazolin was added to vancomycin, the average decrease in vancomycin required to reach IC95 was 0.45 µg/mL, and the fold decrease was 1.61. Only the addition of oxacillin to vancomycin resulted in a larger average fold decrease (1.67-fold) in the amount of vancomycin required to reach the IC95, for the 22/30 (73.3%) isolates that showed a decrease in vancomycin required to reach the IC95 when oxacillin was added. However, for 8/30 (26.7%) isolates, adding oxacillin to vancomycin resulted in an increase in vancomycin required to reach the IC95 (Table 1). Compared to the other antibiotics tested with vancomycin, adding cefazolin to vancomycin resulted in the greatest fold decrease in vancomycin required to reach the IC95 for 11/30 (37%) of isolates tested, which was a higher percentage than any of the other vancomycin combinations. This 37% includes 3 of the 10 isolates with vancomycin MIC = 2 µg/mL and daptomycin MIC <1 µg/mL, 1 of the 4 isolates with daptomycin MIC ≥1 µg/mL and vancomycin MIC <2 µg/mL, and 1 of the 2 isolates with vancomycin MIC = 2 µg/mL and daptomycin MIC ≥1 µg/mL. For the 4/30 (13.3%) isolates tested for which adding cefazolin increased the amount of vancomycin required to reach the IC95, the increase was relatively small (average 0.12 µg/mL increase, or 1.15-fold) (Table 1).
The fold change in vancomycin required to reach the IC95 did not correlate for vancomycin+cefazolin versus vancomycin+oxacillin, and the fold change in daptomycin required to reach the IC95 did not correlate for daptomycin+ceftaroline, daptomycin+cefazolin, or daptomycin+oxacillin. However, the primary antibiotic fold change results between some pairs of combinations (for example, daptomycin+rifampicin and daptomycin+gentamicin) did correlate (Table S2). Thus, two different secondary agents (for example, rifampicin and gentamicin) can have similar effects on a primary agent (for example, daptomycin). However, in our data set, this does not hold for comparing two different β-lactams as secondary agents for the same primary antibiotic. Considering the secondary agents beyond the β-lactams, adding the secondary agent lowered the IC95 of daptomycin more than the IC95 of vancomycin in 72% of the comparisons (shown in Fig. S3). It is also worth noting that the isolates for which adding oxacillin resulted in the greatest decrease in vancomycin (or daptomycin) required to reach the IC95 also usually had an increase in the IC95 of vancomycin (or daptomycin) alone when 2% NaCl was added (shown in Fig. S4). This makes it harder to interpret the actual value of adding oxacillin, since the decrease in daptomycin or vancomycin required to reach the IC95 when oxacillin is added is calculated relative to the IC95 of daptomycin or vancomycin alone in CAMHB+2% NaCl, the same condition used for testing daptomycin+oxacillin and vancomycin+oxacillin.
Finally, we tested isolate TR258 (the third isolate from patient B) and the 29 non-serial isolates with ceftaroline in pairwise combination with gentamicin and rifampicin and compared these results with the ceftaroline+daptomycin results already obtained (Fig. 4). Ceftaroline+rifampicin was synergistic (or at least additive) against 67.9% of the isolates (log2FIC95 < 0), and adding rifampicin to ceftaroline always decreased the amount of ceftaroline required to reach the IC95 (and vice versa). The average decrease in ceftaroline required to reach the IC95 was 0.87 µg/mL, and the average fold decrease was 2.81 (Table 1). This contrasts with the results of adding rifampicin to daptomycin or vancomycin. Daptomycin+rifampicin and vancomycin+rifampicin were not synergistic against any of the isolates tested and resulted in relatively high percentages of isolates for which the amount of daptomycin or vancomycin required to reach the IC95 increased (Table 1). Furthermore, for isolates where adding rifampicin to daptomycin or vancomycin decreased the amount of daptomycin or vancomycin required to reach the IC95, the decrease was relatively low compared to that achieved with the other secondary antibiotics (Table 1). Adding rifampicin resulted in a larger decrease in ceftaroline required to reach the IC95 compared to adding daptomycin for 8/28 (28.6%) isolates tested with both combinations and compared to adding gentamicin for 21/26 (80.8%) isolates tested with both combinations (two rifampicin-resistant isolates were not tested with rifampicin combinations, and two gentamicin-resistant isolates were not tested with gentamicin combinations). Adding gentamicin never resulted in the largest decrease in ceftaroline required to reach the IC95, compared to adding daptomycin or rifampicin.
Fig 4 Adding rifampicin to ceftaroline always lowers the amount of ceftaroline required to reach the IC
95and vice versa, and the combination is often synergistic or at least additive, unlike when rifampicin is added to daptomycin or vancomycin. (A) This graph is the same style as the graphs in Fig. 3, except this graph shows the results for ceftaroline+gentamicin, ceftaroline+rifampicin, and daptomycin+ceftaroline, where ceftaroline is treated as the primary antibiotic in the fold change calculations. (B) This graph shows the fold change in ceftaroline required to reach the IC95when rifampicin is added (y-axis) vs the fold change in rifampicin required to reach the IC95when ceftaroline is added (x-axis). (C) This graph shows the fold change in daptomycin required to reach the IC95when rifampicin is added (y-axis) vs the fold change in rifampicin required to reach the IC95when daptomycin is added (x-axis). (D) This graph shows the fold change in vancomycin required to reach the IC95when rifampicin is added (y-axis) vs the fold change in rifampicin required to reach the IC95when vancomycin is added (x-axis). For the graphs in (B) through (D), the points representing each isolate are colored by the log2FIC95score of ceftaroline+rifampicin (B), daptomycin+rifampicin (C), or vancomycin+rifampicin (D) for that isolate.
Our results agree with a growing body of work supporting the use of β-lactams in combination therapy with other agents against MRSA and highlight some potentially underutilized β-lactam combinations. Even with the “seesaw effect” in which a decrease in vancomycin and particularly daptomycin susceptibility is accompanied by an increase in β-lactam susceptibility (27–30), results from clinical trials and retrospective cohort studies have been mixed about the benefits of adding a β-lactam to daptomycin (31–33) or vancomycin (32, 34) for treating MRSA bacteremia. Along with other confounding factors, these studies have usually grouped the results of different β-lactams used in combination with daptomycin, vancomycin, or both. Our data and those from others (35–37) indicate that not all β-lactams perform the same in combination and that strain-specific differences occur. Investigator-initiated clinical trials, such as the Staphylococcus aureus Network Adaptive Platform (SNAP) trial to examine cefazolin in combination with daptomycin or vancomycin (38), will be essential for understanding the contribution of these confounding factors (39). Randomized control trials such as SNAP and potential follow-up trials are needed to determine under what patient or infection circumstances (including strain susceptibility) the addition of cefazolin to daptomycin or vancomycin provides a demonstrative benefit compared to monotherapy and what the optimum timing of combination therapy initiation is (depending on patient and infection circumstances). Head-to-head comparison via randomized control trials of daptomycin or vancomycin+cefazolin to daptomycin+ceftaroline (the predominant combination therapy in use for MRSA bacteremia) could further indicate in what circumstances cefazolin can be part of a combination regimen that outperforms the currently available options. The potential advantages of cefazolin are clear—it has similar efficacy but is less likely to be nephrotoxic than other anti-staphylococcal β-lactams (40). Daptomycin+cefazolin, in particular, is also an attractive daptomycin+β-lactam combination in circumstances where avoiding ceftaroline might be desirable due to its expense (36) or the need to prevent an adverse reaction (41), such as neutropenia, particularly in at-risk patients (42).
The antibiofilm activity of rifampicin makes it an attractive adjunctive agent for MRSA infections involving biofilms (14, 26). Rifampicin is not recommended for MRSA bacteremia (14, 43), except for some cases involving prosthetic device infections (26), largely due to a lack of demonstrated benefit and rifampicin resistance development in a large, randomized control trial of rifampicin paired with vancomycin or daptomycin for S. aureus bacteremia (44). These recommendations are also supported by a retrospective cohort analysis for infective endocarditis (45) due to S. aureus. Ceftaroline+rifampicin has not been directly assessed in retrospective studies or clinical trials, but it was successful for all five cases of MRSA bacteremia where it was used in a retrospective, multicenter, observational study of adult patients with MRSA bloodstream infection (11). In their assessment of a panel of biofilm-producing MRSA bloodstream isolates, Barber and colleagues found that adding subinhibitory concentrations of ceftaroline lowered rifampicin biofilm MIC in only 11 of 20 isolates (46), and the addition of rifampicin to ceftaroline failed to show a benefit against three of these MRSA strains in an in vitro biofilm pharmacokinetic/pharmacodynamic model (47). However, these three strains were all rifampicin-resistant (MIC >4 µg/mL), so it seems plausible that ceftaroline+rifampicin synergy may be limited to rifampicin-sensitive MRSA strains. Thus, ceftaroline+rifampicin may be preferable in some cases as a daptomycin and vancomycin-sparing regimen (26). In our demonstration of the efficiency of DiaMOND and its potential for development into a point-of-care assay, we identify three combinations—vancomycin+cefazolin, daptomycin+cefazolin, and ceftaroline+rifampicin—that deserve more in vivo study and demonstrate how different β-lactams perform differently in combination therapy. Furthermore, strain-related differences may affect outcomes in prospective studies assessing the benefit of combination therapy, for example, the SNAP trial (39). Currently, a limitation of DiaMOND is that it is based only on growth inhibition measurements in rich medium (CAMHB), which may not fully recapitulate or reflect complexities of infection such as the development of resistance or persister populations, or adaptations to growing in different infection microenvironments, including growth in biofilms. Adaptations to the method are being developed to address these limitations. The ability to assess strain-specific differences with a rapid method such as DiaMOND may inform antibiotic choice at the point of care, resulting in patient benefit.
We thank Cheleste Thorpe, Husain Poonawala, Joan Mecsas, Talia Greenstein, and members of the Aldridge laboratory for insightful discussions and helpful comments.
This work was supported in part by the Stuart B. Levy Center for Integrated Management of Antimicrobial Resistance at Tufts (Levy CIMAR), a collaboration of Tufts Medical Center and the Tufts University Office of the Vice Provost for Research and Scholarship Strategic Plan. B.B.A. and K.P.D. also were supported, in part, by NIH U19142780 (to Deborah T. Hung). D.R.S. and L.M. were also supported in part by a financial grant from the Tupper Family Foundation at Tufts Medical Center.
K.P.D. performed the conceptualization, data curation, formal analysis, investigation, methodology, validation, visualization, writing (original draft), and writing (review and editing); L.M. performed the data curation, investigation, methodology, and validation; D.R.S. performed the conceptualization, funding acquisition, project administration, resources, supervision, visualization, writing (original draft), and writing (review and editing); B.B.A. performed the conceptualization, funding acquisition, project administration, resources, software, supervision, visualization, writing (original draft), and writing (review and editing). All authors have approved this manuscript.
Bree B. Aldridge, Email: Bree.Aldridge@tufts.edu.
Florence Claude Doucet-Populaire, University Paris-Saclay, AP-HP Hôpital Antoine Béclère, Service de Microbiologie, Institute for Integrative Biology of the Cell (I2BC), CEA, CNRS, Clamart, France.
All data generated or analyzed during this study are included in this article and its supplemental figures and tables and in tabular form in Data S1. Further inquiries can be directed to the corresponding author.
All MRSA bacteremia isolates used in this study were collected for prior work and were deidentified and, therefore, exempt from requiring IRB approval.
The following material is available online at https://doi.org/10.1128/spectrum.00976-24.
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.
All data generated or analyzed during this study are included in this article and its supplemental figures and tables and in tabular form in Data S1. Further inquiries can be directed to the corresponding author.