Development of High-Level Daptomycin Resistance in Abiotrophia and Granulicatella Species Isolates from Patients with Infective Endocarditis

ABSTRACT

Abiotrophia and Granulicatella species are fastidious organisms, representing the causative agents of ∼1% to 3% of cases of infective endocarditis (IE). Little is known about the optimal antibiotic treatment for these species, and daptomycin has been suggested as a therapeutic option. We describe the antimicrobial profiles of Abiotrophia and Granulicatella IE isolates, investigate high-level daptomycin resistance (HLDR) development, and evaluate daptomycin activity in combination therapy. In vitro studies with 16 IE strains (6 Abiotrophia defectiva strains, 9 Granulicatella adiacens strains, and 1 G. elegans strain) were performed using microdilution to determine MICs and time-kill methodology to evaluate combination therapy. Daptomycin nonsusceptibility (DNS) (MIC ≥ 2 mg/liter) and HLDR (MIC ≥ 256 mg/liter) were based on existing Clinical and Laboratory Standards Institute (CLSI) breakpoints for viridans group streptococci. All isolates were susceptible to vancomycin: G. adiacens was more susceptible to penicillin and ampicillin than A. defectiva (22% versus 0% and 67% versus 33%) but less susceptible to ceftriaxone and daptomycin (56% versus 83% and 11% versus 50%). HLDR developed in both A. defectiva (33%) and G. adiacens (78%) after 24 h of exposure to daptomycin. Combination therapy did not prevent the development of daptomycin resistance with ampicillin (2/3 strains), gentamicin (2/3 strains), ceftriaxone (2/3 strains), or ceftaroline (2/3 strains). Once developed, HLDR was stable for a prolonged time (>3 weeks) in G. adiacens, whereas in A. defectiva, HLDR reversed to the baseline MIC at day 10. This study is the first to demonstrate rapid HLDR development in Abiotrophia and Granulicatella species in vitro. Resistance was stable, and most combination therapies did not prevent it.

INTRODUCTION

Abiotrophia defectiva and Granulicatella species (G. elegans, G. adiacens, and G. balaenopterae) were formerly known as nutritionally variant streptococci (NVS) because of their requirement for l-cysteine or pyridoxal hydrochloride to promote growth in culture media. These microorganisms make up part of the normal human microbiota; however, they are responsible for serious infections such as pneumonia, sepsis, and infective endocarditis (IE) (1, 2). IE caused by Abiotrophia and Granulicatella represents around 1% to 3% of all cases of IE (3). The identification of these microorganisms relies on blood cultures as the gold standard despite the pathogens being frequently involved in negative blood cultures. This is because they are slow-growing and fastidious pathogens that render identification difficult (4). Newer diagnostic modalities, e.g., matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry, 16S rRNA gene PCR, and routine sequencing in microbiological laboratories, could help to identify these microorganisms as being responsible for a higher number of cases of IE (5).

There is poor knowledge about the clinical characterization and optimal management of these patients. Therefore, the impact of antibiotic regimens on outcomes is poorly understood. As a result of the scant available evidence, current international guidelines have major gaps regarding treatment recommendations. Due to the resemblance to streptococcal species, most treatment recommendations are based on streptococcal IE (6, 7).

Based on just two case series, European Society of Cardiology (ESC) guidelines recommend 6 weeks of penicillin (PEN), ceftriaxone, or vancomycin (VAN) in combination with a minimum of 2 weeks of gentamicin (GEN) (7). American Heart Association (AHA) guidelines recommend that Abiotrophia and Granulicatella spp. be treated as streptococci with an elevated (≥0.5 mg/liter) penicillin MIC with penicillin or ampicillin in combination with gentamicin, with the length of treatment tailored by infectious diseases consultation (6). In patients with penicillin intolerance, vancomycin monotherapy is an alternative, and ceftriaxone combined with gentamicin may be another option for susceptible organisms (6, 8). However, the optimal antimicrobial treatment for NVS and especially for Abiotrophia and Granulicatella strains has not yet been defined. Several case series describe high failure rates when treating NVS IE with the recommended combination of penicillin plus gentamicin, and nationwide antimicrobial testing has shown reduced penicillin susceptibility, particularly among Abiotrophia IE strains (9, 10). Moreover, prolonged administration of aminoglycosides such as gentamicin is associated with a notable risk of nephrotoxicity; therefore, safer treatment alternatives are needed (11).

Daptomycin (DAP) is a cyclic lipopeptide antibiotic that offers some advantages over vancomycin due to its bactericidal activity against clinical strains of Gram-positive bacteria, including viridans group streptococci (VGS), and various types of unusual Gram-positive bacteria (12, 13). Nevertheless, the appearance of daptomycin nonsusceptibility (DNS) or resistance has been well documented in Staphylococcus aureus and enterococcal strains (14–16), as has high-level daptomycin resistance (HLDR) in viridans group streptococci (17). This resistance is characterized by being fast, stable, and elevated (MIC of ≥256 mg/liter), which appears to be preventable in vitro in a notable proportion of strains by adding either beta-lactams or gentamicin (17, 18). Recently, a clinical case has been reported where the addition of daptomycin to amoxicillin was proposed as a safe and effective alternative for the treatment of Abiotrophia defectiva IE (11). Nevertheless, few studies have investigated Abiotrophia defectiva and Granulicatella species antimicrobial susceptibilities, and to date, experimental data are unavailable for both the development of daptomycin resistance and the activity of daptomycin combined with ampicillin or cephalosporins in these species.

We undertook the current study with the following objectives: (i) to describe the antimicrobial profile of Abiotrophia and Granulicatella isolates from 16 IE patients; (ii) to investigate if Abiotrophia and Granulicatella isolates develop HLDR; (iii) to evaluate daptomycin activity in combination with gentamicin, ampicillin, ceftaroline, or ceftriaxone and determine if the development of HLDR is prevented by combination therapy; and (iv) to compare the in vitro activities of daptomycin in combination with penicillin or vancomycin plus gentamicin.

(This work was presented at the 23rd National Congress of the Spanish Society of Infectious Diseases and Clinical Microbiology held in 2019 in Madrid, Spain [poster 0080] [39], and the 8th National Congress of the Spanish Society of Cardiovascular Infections held in 2019 in Madrid, Spain [poster PF-001] [40].)

RESULTS

Microorganisms.

The strains from 16 consecutive IE episodes included in the study were 6 A. defectiva strains (37.5%), 9 G. adiacens strains (56.3%), and 1 G. elegans strain (6.2%).

In vitro susceptibility testing.

The MIC results from in vitro susceptibility testing of ampicillin, penicillin, gentamicin, ceftaroline, ceftriaxone, and vancomycin are shown in Table S1 in the supplemental material. The three strains selected for the in vitro studies are in boldface type. Overall, all the isolates were susceptible to vancomycin, but all the selected strains were tolerant to vancomycin (minimal bactericidal concentration [MBC]/MIC ≥ 128 mg/liter). G. elegans was uniformly susceptible to all the antibiotics tested. On the other hand, both A. defectiva and G. adiacens showed reduced susceptibility to penicillin, with zero of six (0%) and two of nine (22%) strains being fully susceptible, respectively, and most strains being in the intermediate MIC group: five of six and six of nine (83% and 67%) strains, respectively. The ampicillin MICs of A. defectiva were distributed evenly among susceptible, intermediate, and resistant groups, i.e., two of six strains (33%) in each, whereas the G. adiacens strains were ampicillin susceptible for six of nine strains (67%), intermediate for two of nine strains (22%), and resistant for one of nine (11%) cases. Regarding ceftriaxone, most A. defectiva strains, five of six strains (83%), were susceptible, whereas almost half of the G. adiacens strains were intermediate or resistant: four of nine (44%) cases.

Results of screening for DNS and HLDR strains.

Baseline daptomycin MICs were between 1 and 4 mg/liter for A. defectiva, of which three of six strains were susceptible to daptomycin with an MIC of ≤1 mg/liter; for G. adiacens, baseline daptomycin MICs were between 1 and 16 mg/liter, of which one of nine strains was susceptible to daptomycin with an MIC of ≤1 mg/liter; and for G. elegans, the daptomycin MIC was 0.5 mg/liter. After being incubated with inhibitory concentrations of daptomycin, all strains increased the baseline MIC of daptomycin and showed resistance to daptomycin with either DNS or HLDR (Table 1).

TABLE 1.

Selection of resistance and high-level resistance after exposure to daptomycin^a

Strain	Daptomycin MIC (mg/liter)
	Basal	Postexposure
		DNS	HLDR
Abiotrophia spp.
A. defectiva 288	1		≥256
A. defectiva 569	4	8–64
A. defectiva 580	2	12–24
A. defectiva 1051	1	6–8
A. defectiva 1093	4	8–64
A. defectiva 1120	1		≥256
Granulicatella spp.
G. adiacens 295	2		≥256
G. adiacens 318	4		≥256
G. adiacens 413	16		≥256
G. adiacens 659	2		≥256
G. adiacens 761	1		≥256
G. adiacens 814	4		≥256
G. adiacens 991	4		≥256
G. adiacens 1083	4	12–16
G. adiacens 1097	8	8–16
G. elegans 289	0.5	8–64

^aDNS, daptomycin nonsusceptible (MIC ≥ 2 mg/liter); HLDR, high-level daptomycin resistance (MIC ≥ 256 mg/liter). According to CLSI breakpoints for viridans group streptococci, isolates were defined as daptomycin susceptible when the MIC was ≤1 mg/liter.

The distribution of DNS and HLDR after the subculturing of strains with inhibitory concentrations of daptomycin showed that A. defectiva has higher DNS rates than G. adiacens (67% [4 strains] versus 22% [2 strains] [P = 0.14]), whereas G. adiacens showed higher rates of HLDR development than A. defectiva (78% [7 strains] versus 33% [2 strains] [P = 0.14]). Only the G. elegans strain of the collection developed DNS.

In vitro time-kill studies.

Figure 1 shows the results of the synergy studies for penicillin and vancomycin plus gentamicin (Tables S2 and S3). Penicillin combined with gentamicin (Fig. 1A) showed synergistic and bactericidal activity for the A. defectiva 288 strain at all concentrations. For G. adiacens 295, synergistic and bactericidal activity was observed under the combination therapy with the highest concentration of 1× MIC PEN and 1× MIC GEN, and for G. adiacens 761, synergistic activity was observed only for the combination therapy with the highest concentration of 1× MIC PEN plus 1× MIC GEN. In vitro, vancomycin alone at up to 64 μg/ml was not bactericidal due to the three strains that were tolerant. However, for the A. defectiva 288 strain, the combination of vancomycin plus gentamicin (Fig. 1B) showed synergistic and bactericidal activity only at the highest concentration of 1× MIC VAN plus 1× MIC GEN, and for the G. adiacens 295 strain, the combination showed synergistic and bactericidal activity at the highest concentration. However, for G. adiacens 761, no activity was observed. It is worth noting that at different vancomycin concentrations (1× MIC, 2× MIC, 3× MIC, and 4× MIC), no bactericidal activity against A. defectiva 288, G. adiacens 295, and G. adiacens 761 was observed in any case.

FIG 1.

Time-kill curves for A. defectiva 288, G. adiacens 295, and G. adiacens 761. The strains were incubated with penicillin (PEN) (A) or vancomycin (B) plus gentamicin (GEN) at concentrations of 0.5× MIC and 1× MIC for all antibiotics. Values are means ± standard deviations from two independent experiments. The dashed line is indicative of bactericidal activity.

Figure 2 shows the results of the synergy studies for daptomycin combined with ampicillin, gentamicin, ceftaroline, or ceftriaxone (Tables S4 to S7). After 24 h of incubation, all those isolates recovered from the combinations with daptomycin were tested for changes in the MIC of daptomycin. HLDR isolates were recovered from daptomycin monotherapies at all concentrations. Daptomycin plus ampicillin (Fig. 2A) showed synergistic and bactericidal activity for the A. defectiva 288 strain at all concentrations. For the G. adiacens 295 and 761 strains, synergistic activity was observed only with the combination therapy with the highest concentration of 1× MIC DAP. Indifference was observed at lower concentrations of 1/2× MIC DAP. Additionally, among the Granulicatella strains, independently of the concentration tested and the activity, HLDR or DNS (MIC of DAP of between 16 and 128 mg/liter) isolates were recovered in all cases. Daptomycin plus gentamicin (Fig. 2B) showed synergistic and bactericidal activity at 24 h against A. defectiva 288 at all concentrations. For G. adiacens 295, a synergistic and bactericidal effect was observed only at a concentration equal to 1× MIC gentamicin. At gentamicin concentrations of 1/2× MIC, activity was lost, and HLDR or DNS (MIC of DAP of 64 mg/liter) isolates were recovered. For G. adiacens 761, the combination was indifferent, and HLDR or DNS (MIC of DAP of between 32 and 96 mg/liter) isolates were recovered in all cases. Daptomycin plus ceftaroline (Fig. 2C) showed synergistic and bactericidal activity against A. defectiva 288 at all concentrations, but despite the bactericidal activity, HLDR or DNS (MIC of DAP of 4 mg/liter) isolates were recovered. Synergistic and bactericidal activity was observed in G. adiacens 295 at a concentration of 1× MIC, and synergistic activity was observed in the remaining combination therapies with daptomycin plus ceftaroline (at least one of the antibiotics at 1/2× MIC). In G. adiacens 761, daptomycin plus ceftaroline showed synergistic activity, and HLDR isolates developed with all combinations. At the highest concentration (1× MIC), the combination of daptomycin plus ceftriaxone (Fig. 2D) was synergistic and bactericidal against A. defectiva 288 and G. adiacens 295, but HLDR or DNS (MIC of DAP of between 4 and 48 mg/liter) isolates were still recovered for A. defectiva 288. For G. adiacens 761, HDLR isolates were recovered from all combinations and concentrations, with daptomycin plus ceftriaxone showing synergistic activity (Fig. 2D).

FIG 2.

Time-kill curves for A. defectiva 288, G. adiacens 295, and G. adiacens 761. The strains were incubated with daptomycin (DAP) plus ampicillin (AMP) (A), gentamicin (GEN) (B), ceftaroline (CTL) (C), or ceftriaxone (CTR) (D) at concentrations of 0.5× MIC and 1× MIC for all antibiotics. Values are means ± standard deviations from two independent experiments. The dashed line is indicative of bactericidal activity. *, recovered isolates with daptomycin nonsusceptibility (DNS) (MIC of 8 to 96 mg/liter); #, recovered isolates with high-level daptomycin resistance (HLDR) (MIC ≥ 256 mg/liter).

The results after 24 h of incubation with a high concentration of daptomycin (10 mg/liter) plus ampicillin or gentamicin are shown in Fig. 3. Synergistic and bactericidal activity was observed for the two G. adiacens strains when daptomycin was combined with ampicillin (Fig. 3A), and no HLDR strains were recovered except for G. adiacens 295 at a concentration of 1/2× MIC ampicillin. Daptomycin plus gentamicin (Fig. 3B) showed synergistic and bactericidal activity at 24 h for the two G. adiacens strains with the exception of daptomycin plus gentamicin at 1/2× MIC for G. adiacens 761, where only synergistic activity was achieved and HLDR isolates developed (see also Tables S8 and S9 in the supplemental material). In A. defectiva 288, a bactericidal effect was found for daptomycin monotherapy at a concentration of 10 mg/liter (Table S8).

FIG 3.

Time-kill curves for G. adiacens 295 and G. adiacens 761. The strains were incubated with daptomycin (DAP) at a concentration of 10 mg/liter plus ampicillin (AMP) (A) or gentamicin (GEN) (B) at concentrations of 0.5× MIC and 1× MIC for ampicillin and gentamicin. Values are means ± standard deviations from two independent experiments. The dashed line is indicative of bactericidal activity. #, recovered isolates with high-level daptomycin resistance (HLDR) (MIC ≥ 256 mg/liter).

Stability of daptomycin resistance.

We conducted studies on the stability of HLDR by subculturing the three selected HLDR isolates (Fig. 4). Lower-grade stability was observed for the A. defectiva strain, so that between days 10 and 11, the microorganism recovered the baseline daptomycin MIC of 2 mg/liter in both culture media (Fig. 4A). We observed a higher grade of stability of resistance (from 3 weeks and up to 4 weeks, depending on the culture medium) for the two G. adiacens isolates. HLDR was maintained throughout testing in G. adiacens 761 when subcultures were made in the enriched solid medium. When G. adiacens 761 was cultivated in liquid medium, the MIC decreased at day 21 and stabilized at an MIC of 12 mg/liter (Fig. 4B).

FIG 4.

Stability of daptomycin resistance. The strains A. defectiva 288 (A), G. adiacens 295 (B), and G. adiacens 761 (C) with high-level resistance to DAP at ≥256 mg/liter were placed in series on a daily basis into antibiotic-free medium to determine changes in the MIC of daptomycin for 30 days by an Etest.

In both culture media, G. adiacens 295 showed a change in the resistance profile at around day 20, with a continuous decrease in the daptomycin MIC although not returning to the baseline MIC (Fig. 4C).

DISCUSSION

A. defectiva and G. adiacens are fastidious organisms with unique nutritional requirements that are essential for growth. IE caused by A. defectiva and G. adiacens remains rare, but it is an understudied condition that carries significant morbidity and mortality (2, 19, 20). International guidelines recommend combination therapy of penicillin, ceftriaxone, or vancomycin plus gentamicin (6, 7); however, there is a lack of evidence concerning the optimal treatment of IE caused by A. defectiva and G. adiacens, and therefore, the potential role of daptomycin should be considered.

Our daptomycin investigations have several principal findings. First, we found rather high baseline daptomycin MICs for 50% of A. defectiva isolates and 89% of G. adiacens isolates that were nonsusceptible to daptomycin. Second, penicillin was the least effective antibiotic, with full susceptibility in none of the A. defectiva isolates and only 22% of the G. adiacens isolates. Third, all strains showed the potential to develop daptomycin nonsusceptibility or high-level daptomycin resistance when exposed to daptomycin. Fourth, in A. defectiva, daptomycin plus ampicillin or gentamicin showed bactericidal and synergistic activity without the appearance of daptomycin-resistant strains. Finally, in G. adiacens, daptomycin in combination with cephalosporins showed synergy, but despite this, one of the strains developed daptomycin resistance.

Previous studies of A. defectiva and G. adiacens isolates from a mixture of different clinical infections have shown various results for daptomycin MIC values. A small study by Piper et al. including 10 strains (4 associated with IE) found no strains with DNS (13). On the other hand, larger studies by Prasidthrathsint and Fisher, Alberti et al., and Mushtaq et al. described higher rates of DNS (between 50% and 96%), findings more in keeping with ours (10, 21, 22).

The rate of penicillin-susceptible strains in our results is lower than what was previously reported for A. defectiva strains susceptible to penicillin (11 to 24%) and G. adiacens strains susceptible to penicillin (34 to 39%) (10, 21, 22). Contrary to these findings, another study focusing primarily on oral NVS isolates from healthy volunteers found substantially higher rates of penicillin susceptibility of 57% in A. defectiva strains and 82% in G. adiacens strains (23). Combining our novel findings with the knowledge from these previous studies, it is reasonable to believe that there might be a gradient of penicillin resistance. It would go from low-level resistance in noninfectious oral isolates (23) through increased resistance in mixed infections (10, 21, 22) to a very high degree of resistance in IE isolates. However, our in vitro studies show that the combination of penicillin plus gentamicin was synergistic and bactericidal against A. defectiva, had synergistic activity against both G. adiacens strains, and should remain the therapy of choice. Other treatment options for A. defectiva and G. adiacens IE include vancomycin or ceftriaxone. Exactly as described in previous studies, all NVS strains were fully susceptible to vancomycin (21, 23, 24). The combination of vancomycin plus gentamicin was synergistic and bactericidal against A. defectiva and G. adiacens 295, but no activity was observed against G. adiacens 761. Regarding ceftriaxone, our results showed 83% and 56% susceptibilities for A. defectiva and G. adiacens strains, respectively. This pattern is similar to those in previous studies finding higher rates of ceftriaxone susceptibility in A. defectiva (92 to 100%) and lower rates of ceftriaxone susceptibility in G. adiacens (22 to 76%) (10, 21–23). As with penicillin, the combination of ceftriaxone and gentamicin is highly likely to be synergistic against these strains (6, 7).

Daptomycin may represent an alternative treatment option for NVS IE. However, so far, it has been unknown to which extent A. defectiva and G. adiacens develop resistance to daptomycin. Daptomycin nonsusceptibility and resistance in Gram-positive bacteria are intrinsically related to the types of mechanisms usually seen in acquired resistance, such as drug target modification, drug inactivation, and drug efflux (25–27). The term nonsusceptibility is used in daptomycin resistance since the MIC cutoff for resistance has not yet been established (28). So far, the development of daptomycin nonsusceptibility has been shown in clinical practice during daptomycin therapy for Staphylococcus species and Enterococcus species infections (29–31). In viridans group streptococci, a more worrisome picture has recently been discovered with the emergence of HLDR upon daptomycin exposure both in vitro and in vivo (17, 18, 32). Different studies have demonstrated the rapid development of high-level resistance (≥256 mg/liter), within 24 h, despite antibiotic concentrations as high as 8× the baseline MIC (17, 18). These HLDR isolates were associated with the complete disappearance of phosphatidylglycerol and cardiolipin from cell membranes through substitutions in two enzymes involved in the cardiolipin biosynthesis pathway (33, 34). The present study is the first to describe a similar development of HLDR in Abiotrophia and Granulicatella strains after incubation with inhibitory concentrations of daptomycin. We found that 78% of G. adiacens strains and 33% of A. defectiva strains developed HLDR. In addition, in those strains that were daptomycin susceptible at baseline and that did not develop HLDR, we observed an increase in the basal MIC, rendering them all DNS. The drastic increases in daptomycin MICs, including the development of HLDR, were very fast (24 to 48 h postexposure), and the pattern appears similar to that of the development of resistance in Streptococcus mitis (17, 27, 35).

With the newly established rapid development of HLDR in Abiotrophia and Granulicatella strains, we evaluated whether the combination therapy and high doses of daptomycin treatment improve the antibacterial response and reduce the development of daptomycin resistance. Several interesting observations emerged when we evaluated the in vitro activity of daptomycin combined with a broad range of antibiotics to inhibit bacterial growth and prevent the emergence of HLDR in A. defectiva and G. adiacens IE strains. First, our data demonstrate that the addition of ceftaroline, ceftriaxone, or gentamicin to daptomycin in inhibitory concentrations improves the synergistic and bactericidal activity of daptomycin, but this finding was consistent for only two of the three strains studied. For one strain of G. adiacens, none of the combinations of antibiotics with daptomycin, regardless of the dose, rendered both a synergistic and bactericidal impact. Furthermore, DNS and HLDR isolates were recovered in all assays with this strain.

Second, high-dose daptomycin used concomitantly with ampicillin or gentamicin achieved bactericidal activity in both Granulicatella strains. It is noteworthy that high-dose daptomycin in monotherapy does not achieve a bactericidal effect and generates the development of HLDR. Similar observations were reported for S. mitis (17, 18).

Third, despite displaying synergistic or bactericidal activity, none of the antibiotic combinations had the ability to suppress completely the evolution of daptomycin resistance in vitro. This strays from reports in the literature for VGS, where several studies have shown that daptomycin in combination with beta-lactams or cephalosporins reduces/suppresses the development of daptomycin resistance (18, 32, 36, 37).

Finally, we investigated the stability of HLDR over time by measuring the MICs daily for 30 days. The current literature has described the generation of HLDR in streptococci in as early as 24 h and stable over time at 4 weeks (32, 38). In our study, the Granulicatella strains showed stable HLDR (3 to 4 weeks), without any of the strains returning to the baseline MIC values. A less stable pattern was observed in A. defectiva, where the effect of HLDR was reversed at around day 10 and MIC values returned to the baseline MIC values. Although the HLDR mechanism has not yet been discovered for A. defectiva and G. adiacens, data are available regarding resistance mechanisms in the S. mitis-oralis subgroup of VGS (33).

Our study has several limitations. First, all strains were selected from one referral center. Second, the percentages of resistance are dependent on the numbers of strains analyzed, and therefore, the accuracy of the estimates should be interpreted in this context. Third, for the synergy studies, we selected three specific strains and a standard inoculum for the study; it is possible that other strains or inocula would have shown different results. However, we chose these representative strains with beta-lactam resistance and with the ability to develop HLDR.

In conclusion, we have demonstrated extensive penicillin resistance in IE strains of Abiotrophia defectiva and Granulicatella adiacens as well as a considerable presence of baseline daptomycin nonsusceptibility at a higher degree than in previous studies of mixed infections. As a novel finding, we demonstrate the ability of Abiotrophia defectiva and Granulicatella adiacens to rapidly develop high-level daptomycin resistance. Combination therapy with ampicillin, gentamicin, or cephalosporins enhanced the activity of daptomycin but did not prevent the development of resistance in many cases. Further studies are necessary to elucidate the molecular basis of daptomycin resistance in these microorganisms and the clinical significance of our findings. In the meantime, patients with endocarditis caused by these strains should be treated with a combination of beta-lactams or vancomycin (in allergic patients) plus gentamicin.

MATERIALS AND METHODS

Bacterial strains.

We studied 16 consecutive strains (6 A. defectiva strains, 9 G. adiacens strains, and 1 G. elegans strain) isolated from blood cultures of patients with IE diagnosed at our center between 2000 and 2018. None of the patients had previously received daptomycin. The three strains (A. defectiva 288, G. adiacens 295, and G. adiacens 761) that developed HLDR in vitro were selected for in vitro synergy testing and studies of the stability of resistance. We chose strains that were intermediately susceptible to penicillin; in addition, the A. defectiva strain was resistant to ampicillin, and one of the G. adiacens strains was resistant to ceftriaxone. The isolates were identified by MALDI-TOF mass spectrometry and stored at −80°C in skim milk.

Antimicrobial agents.

Daptomycin powder was supplied by MSD. Gentamicin, penicillin, ampicillin, ceftaroline, ceftriaxone, and vancomycin were purchased commercially from Sigma-Aldrich. The drugs were prepared according to the manufacturers’ recommendations.

Susceptibility testing.

Daptomycin, gentamicin, penicillin, ampicillin, ceftaroline, ceftriaxone, and vancomycin MICs were determined using the microdilution method in cation-adjusted Mueller-Hinton broth (MHB; Oxoid Ltd., Hampshire, England) liquid medium. For testing daptomycin, MHB was supplemented with calcium chloride at 50 mg/liter, 5% lysed horse blood, and 0.001% pyridoxal hydrochloride (Sigma-Aldrich). All solid cultures were performed by using PolyViteX chocolate agar (bioMérieux SA, Marcy l’Etoile, France). The experiments were carried out according to the recommended procedures of the CLSI (28), using Streptococcus pneumoniae ATCC 49619 as the reference control strain for assay validation.

Resistance screening methodology.

Resistance screening was performed as described previously (17). Briefly, all strains were subcultured in the presence of daptomycin at 0.5 mg/liter, 1 mg/liter, and 2 mg/liter on Mueller-Hinton agar plates (Oxoid Ltd., Hampshire, England) supplemented with 50 mg/liter of calcium, 5% (vol/vol) defibrinated sheep blood (Thermo Fisher Scientific), and 0.001% (vol/vol) pyridoxal hydrochloride. The final inoculum was adjusted to 5 × 10⁵ CFU/ml. Plates were incubated for up to 48 h in a 5% CO₂ atmosphere. S. mitis 351 was used as a positive quality control strain, and S. pneumoniae ATCC 49619 was used as a negative quality control strain. In plates with positive growth in the presence of daptomycin, the same CFU were harvested and retested by the Etest method to determine changes or increases in the MIC of daptomycin.

Since daptomycin resistance MIC breakpoints for NVS have not been established, we used the CLSI breakpoints for VGS as the closest comparator. Strains with an MIC of ≥2 mg/liter were considered not susceptible to daptomycin (daptomycin nonsusceptible [DNS]). High-level daptomycin resistance (HLDR) was defined as an MIC of ≥256 mg/liter.

Daptomycin resistance stability.

To assess the stability of daptomycin resistance, the A. defectiva 288, G. adiacens 295, and G. adiacens 761 strains that had developed high-level resistance to DAP at ≥256 mg/liter were subjected to serial daily passages in antibiotic-free medium. Every 24 h, the growth in antibiotic-free plates or liquid medium was analyzed by the Etest method to determine changes in the MIC of daptomycin for 30 days. At the end of the experiment, all recovered isolates were identified by MALDI-TOF mass spectrometry.

Synergy studies.

The time-kill methodology was used to test the activity of penicillin or vancomycin plus gentamicin and of daptomycin plus gentamicin, ampicillin, ceftaroline, or ceftriaxone against the selected strains. A final inoculum of between 5 × 10⁵ and 1 × 10⁶ CFU/ml was used. Prior to inoculation, each tube of fresh MHB, adjusted to 50 mg/liter of calcium, was supplemented with 5% lysed horse blood and 0.001% pyridoxal hydrochloride, and the different antibiotics alone or in combination were evaluated. Concentrations of 0.5× MIC and 1× MIC were chosen for testing. Daptomycin was also evaluated at a concentration of 10 mg/liter (high concentration relative to the MIC) plus ampicillin or gentamicin at 0.5× MIC and 1× MIC. A tube without antibiotic was used as a growth control. Viability counts were performed at 0, 4, and 24 h. Drug carryover was prevented by dilution. Bactericidal activity was defined as a 3-log₁₀ reduction in CFU per milliliter at 24 h in comparison to the initial inoculum. Synergistic activity was defined as a 2-log₁₀ reduction in CFU per milliliter at 24 h in comparison to the reduction by the most active single-therapy antibiotic. Time-kill studies were performed in duplicate. It should be noted that in combinations with daptomycin where the values were below the limit of detection (2 log₁₀ CFU/ml), the MIC change for daptomycin was not evaluated.

Statistical analysis.

Statistical analysis was performed using the SPSS software package (version 25.0; SPSS, Chicago, IL, USA). All assays were repeated independently at least two times. Differences in susceptibility prevalences across species were assessed using the chi-square test. A P value of less than 0.05 was considered statistically significant.