In Vitro Activities of 15 Antimicrobial Agents against Clinical Isolates of South African Enterococci

Abstract

The activities of a panel of currently available antibiotics and the investigational agents LY 333328, linezolid, CL 331,002, CL 329,998, moxifloxacin (BAY 12-8039), trovafloxacin, and quinupristin-dalfopristin against 274 clinical isolates of enterococci were determined. No vancomycin resistance or β-lactamase production was observed. Except for 12 isolates (all non-Enterococcus faecalis) showing reduced susceptibility to quinupristin-dalfopristin (MIC, ≥4 μg/ml), the new agents exhibited promising in vitro antienterococcal activity.

The enterococci display a natural capacity to acquire and accumulate resistance determinants, enabling them to survive the activity of virtually every antimicrobial agent currently available for clinical use. Nosocomial enterococcal infections may be complicated by therapeutic difficulties when caused by multiresistant strains (11, 13). With therapeutic options running out progressively, the necessity of investigating new antimicrobial agents in order to formulate potential strategies required to address the challenge of multiresistant enterococci has to be emphasized. Only two studies investigating the extent of resistance to antimicrobial agents presently in therapeutic use have been conducted on South African isolates of enterococci (8, 20). High-level aminoglycoside resistance in up to 77% of Enterococcus faecium isolates (20), concurrent resistance to β-lactam agents and gentamicin in more than 20% of E. faecium isolates, and an increase in ampicillin- and penicillin-resistant strains of E. faecium from 0 to 50% and from 0 to 59.4%, respectively, over a 5-year period (8) have been reported. The frequent use of amikacin as the first-line aminoglycoside, especially in pediatric intensive care units, has been suggested as a factor contributing to selection of resistant E. faecium strains (20).

The objectives of this study were to determine resistance patterns of enterococci isolated from clinically significant infections in Bloemfontein training hospitals and to assess a broad range of investigational agents with proposed antienterococcal activity. The following new agents were included in the study: the semisynthetic glycopeptide LY 333328 (15); the oxazolidinone antibiotic PNU 100,766 (linezolid) (9); the glycylcyclines CL 331,002 (DMG-DMDOT) and CL 329,998 (DMG-MINO), derivatives of tetracycline and minocycline, respectively (22); the fluoroquinolones moxifloxacin (the proposed international nonproprietary name for BAY 12-8039) (3, 4) and trovafloxacin (CP-99,219) (2); and the streptogramin combination compound quinupristin-dalfopristin (RP 59500) (1).

(This work was presented in part at the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada, 28 September to 1 October 1997 [21a].)

A total of 274 enterococcal isolates were collected consecutively during the period from May 1996 through July 1997 from clinical specimens obtained from patients in two local tertiary training hospitals. Duplicate isolates and strains isolated from different specimens from the same patient were excluded. Specimens included in the study were pus swabs (101 specimens), urine (74 specimens), intravenous catheter tips (36 specimens), blood cultures (20 specimens), tissue biopsy material (16 specimens), endotracheal aspirates (8 specimens), pleural fluid (6 specimens), ascites fluid (5 specimens), cerebrospinal fluid (2 specimens), breast milk (2 specimens) and 1 specimen each from bone biopsy tissue, synovial fluid, bile, and endometrial aspirate. Identification of enterococci to species level was based on a series of conventional biochemical tests (6, 7). The API 20 Strep and Rapid ID 32 Strep systems (bioMérieux, Marcy-l’Etoile, France) were employed to identify isolates that did not conform to the biochemical scheme of Facklam and Sahm (7). Interpretation of API profiles was confirmed by Apilab Plus Version 3.2.2 software (bioMérieux).

Antimicrobial agents investigated were gentamicin and kanamycin (Sigma Chemical Co., St. Louis, Mo.); ampicillin (SmithKline Beecham Pharmaceuticals, Worthing, United Kingdom); erythromycin, vancomycin, and LY 333328 (Eli Lilly & Co., Indianapolis, Ind.); teicoplanin (Hoechst Marion Roussel, Gerenzano, Italy); ciprofloxacin and moxifloxacin (Bayer AG, Wuppertal, Germany); trovafloxacin (Pfizer Inc., Groton, Conn.); tetracycline (Bristol-Myers Squibb, Princeton, N.J.); linezolid (Pharmacia & Upjohn Inc., Kalamazoo, Mich.); CL 331,002 and CL 329,998 (Lederle Laboratories, Pearl River, N.Y.); and quinupristin-dalfopristin (Rhône-Poulenc Rorer, Collegeville, Pa.). Susceptibility testing was performed on Mueller-Hinton agar (Difco Laboratories, Detroit, Mich.), according to the approved National Committee for Clinical Laboratory Standards (NCCLS) agar dilution method (14). NCCLS resistance breakpoints were applied for ampicillin, vancomycin, teicoplanin, tetracycline, ciprofloxacin, and erythromycin (14). Since NCCLS breakpoints for aminoglycosides are not relevant in determining high-level resistance to this group of agents, MIC breakpoints proposed by Ounissi et al. (16) were used. Because high-level resistance to amikacin cannot be detected reliably by in vitro methods (5, 13, 18), kanamycin MICs were determined in order to assess amikacin resistance. Pharmaceutical company breakpoint recommendations were (i) a quinupristin-dalfopristin MIC of ≥4 μg/ml that is indicative of reduced susceptibility and (ii) a trovafloxacin resistance breakpoint MIC of ≥8 μg/ml. Resistance breakpoints for the other new agents have not yet been proposed by the manufacturers. All isolates exhibiting resistance to ampicillin (MIC, ≥16 μg/ml) were sonicated and examined for β-lactamase production with the rapid chromogenic nitrocefin disk test (19).

Of the 274 clinical isolates of enterococci investigated, 213 (77.7%) were identified as E. faecalis, 47 (17.2%) were E. faecium, and 14 (5.1%) were non-E. faecalis, non-E. faecium species, the latter group being comprised of E. avium (four isolates), E. hirae (four isolates), E. gallinarum (two isolates), E. durans (two isolates), and one isolate each of E. casseliflavus, E. flavescens, and an asaccharolytic variant of E. faecalis. The distribution of enterococcal species isolated from clinically significant infections was in agreement with results of other epidemiological studies (10, 17). The comparative in vitro antienterococcal activities of the 15 agents investigated are shown in Table 1. β-Lactamase activity was not detected in 35 E. faecium and three E. faecalis isolates for which ampicillin MICs were 32 to 64 μg/ml and therefore did not contribute to ampicillin resistance. Only 5 of 47 (10.6%) E. faecium isolates were fully susceptible to all the agents investigated.

TABLE 1.

Comparative activities of 15 antimicrobial agents against enterococci^a

Species (no. of isolates) and antibiotic	Resistance breakpoint (μg/ml)	% Resistant at breakpoint	MIC (μg/ml)
			50%	90%	Range
E. faecalis (213)
Gentamicin	≥128	39.9	8	>1,024	≤0.5–>1,024
Kanamycinb	≥512	49.3	128	>1,024	≤0.5–>1,024
Ampicillin	≥16	1.4	1	1	0.5–64
Vancomycin	≥32	0	0.5	2	0.25–4
Teicoplanin	≥32	0	0.25	0.25	≤0.06–1
LY 333328	—	—	1	2	0.5–2
Linezolid	—	—	1	1	0.25–2
Tetracycline	≥16	80.3	32	128	0.5–128
CL 329,998	—	—	0.5	1	0.125–2
CL 331,002	—	—	0.25	0.5	≤0.06–0.5
Ciprofloxacin	≥4	14.1	0.5	8	0.125–>32
Trovafloxacin	—	—	0.125	2	≤0.06–16
Moxifloxacin	—	—	0.125	4	≤0.06–16
Erythromycinc	≥8	61.5	>128	>128	0.125–>128
E. faecium (47)
Gentamicin	≥128	46.8	16	512	≤0.5–>1,024
Kanamycinb	≥512	74.5	>1,024	>1,024	2–>1,024
Ampicillin	≥16	74.5	32	64	0.125–64
Vancomycin	≥32	0	0.5	1	0.25–2
Teicoplanin	≥32	0	0.5	0.5	≤0.06–1
LY 333328	—	—	1	1	0.25–2
Linezolid	—	—	1	1	0.25–1
Tetracycline	≥16	36.2	0.5	64	0.125–128
CL 329,998	—	—	0.5	1	0.125–1
CL 331,002	—	—	0.25	0.25	≤0.06–0.5
Ciprofloxacin	≥4	19.1	2	4	0.125–4
Trovafloxacin	—	—	0.5	1	≤0.06–2
Moxifloxacin	—	—	1	1	≤0.06–2
Erythromycin	≥8	66.0	>128	>128	≤0.06–>128
Q-Dd	—	—	1	4	0.25–8
Non-E. faecalis, non-E. faecium (14)
Gentamicin	≥128	0	4	8	1–8
Kanamycinb	≥512	7.1	32	128	4–>1,024
Ampicillin	≥16	7.1	0.5	8	0.25–32
Vancomycin	≥32	0	0.5	1	0.25–4
Teicoplanin	≥32	0	0.125	0.5	≤0.06–0.5
LY 333328	—	—	0.5	1	0.25–1
Linezolid	—	—	1	1	0.25–1
Tetracycline	≥16	28.6	0.5	32	0.25–32
CL 329,998	—	—	0.5	1	0.25–2
CL 331,002	—	—	0.25	0.5	0.125–0.5
Ciprofloxacin	≥4	7.1	0.5	1	≤0.06–4
Trovafloxacin	—	—	0.125	0.25	≤0.06–2
Moxifloxacin	—	—	0.125	0.25	≤0.06–2
Erythromycin	≥8	14.3	0.5	64	≤0.06–>128
Q-Dd	—	—	2	4	0.5–4

^aResistance of E. faecalis to quinupristin-dalfopristin was not assessed. —, resistance breakpoints not available. 

^bKanamycin MICs were determined in order to assess amikacin resistance. 

^cOnly 96 strains of E. faecalis were tested against erythromycin. 

^dQ-D, quinupristin-dalfopristin. 

The different patterns of resistance to multiple antimicrobial agents observed in clinical isolates of E. faecalis and E. faecium are compared in Fig. 1. It is noteworthy that 28 of 30 (93.3%) E. faecalis isolates represented by pattern D in Fig. 1 (i.e., resistant to ciprofloxacin and an aminoglycoside) were resistant to both gentamicin and kanamycin-amikacin, while 6 of 7 (85.7%) E. faecium strains displaying pattern D multiresistance were also resistant to ampicillin. Concomitant resistance to gentamicin (MIC range, 128 to >1,024 μg/ml) and to kanamycin-amikacin (MICs, >1,024 μg/ml) was observed in 22 of 35 (62.9%) ampicillin-resistant isolates of E. faecium. The three E. faecium strains isolated from blood cultures and three of four strains isolated from intravenous catheter tips were resistant to ampicillin, and, with the exception of one gentamicin-susceptible strain (MIC, 8 μg/ml), these isolates associated with potentially life-threatening infections were also resistant to gentamicin (MICs, 256 to 1,024 μg/ml) and kanamycin-amikacin (MICs, >1,024 μg/ml). Because 39.9% of E. faecalis strains and 46.8% of E. faecium strains exhibit resistance to both gentamicin and kanamycin-amikacin and 70.2% of E. faecium strains are resistant to both ampicillin and an aminoglycoside, affecting loss of susceptibility to synergy (13), new strategies must be explored.

FIG. 1.

Multiresistance patterns of clinical isolates of E. faecalis and E. faecium. A, resistant to gentamicin and kanamycin-amikacin; B, resistant to kanamycin-amikacin and susceptible to gentamicin; C, resistant to ampicillin and an aminoglycoside; D, resistant to ciprofloxacin and an aminoglycoside.

Both trovafloxacin and moxifloxacin displayed decreased activity against 30 ciprofloxacin-resistant (MIC, ≥4 μg/ml) E. faecalis isolates. Moxifloxacin MICs of ≥4 μg/ml were observed for 26 of 30 (86.7%) strains, while 6 of 30 (20.0%) isolates could be regarded as trovafloxacin-resistant (MICs, ≥8 μg/ml). No trovafloxacin or moxifloxacin MICs for ciprofloxacin-resistant strains in the E. faecium group exceeded the resistance breakpoint of the representative fluoroquinolone, ciprofloxacin. The observation that ciprofloxacin-resistant strains of E. faecalis displayed reduced susceptibility to trovafloxacin and moxifloxacin may reflect low-level fluoroquinolone cross-resistance (23).

As requested by the manufacturer, the assessment of quinupristin-dalfopristin was limited to E. faecium and non-E. faecalis, non-E. faecium species. At 4 μg/ml, the MIC currently recommended as indicative of reduced susceptibility to this agent, 12 of 61 (19.7%) enterococcal strains could be regarded as resistant to quinupristin-dalfopristin. Concomitant ampicillin and aminoglycoside resistance and reduced susceptibility to quinupristin-dalfopristin were displayed by 3 of 47 (6.4%) strains of E. faecium.

When the results of the study presented here are taken into consideration, it is evident that conventional therapeutic options in local tertiary hospitals have reached serious limitations. Without exception, kanamycin MICs for all kanamycin-resistant isolates were ≥1,024 μg/ml, indicating high-level amikacin resistance probably attributable to the liberal use of amikacin in local teaching hospitals (20). The high incidence of ampicillin-resistant E. faecium isolates (35 of 47) might in turn be linked to the frequent prescription of expanded-spectrum cephalosporins for hospitalized patients (12). These results contribute to existing evidence that older antimicrobial agents are progressively losing their efficacy against the enterococci. In the present study no vancomycin-resistant strains were found, but a vancomycin-resistant strain of E. faecalis (MIC, 32 μg/ml) of the vanB genotype was isolated in Bloemfontein in 1995 (21). Fortunately, the threat of vancomycin and teicoplanin resistance has not yet materialized in the two local teaching hospitals covered by this study.

In summary, the novel glycopeptide LY 333328 compared favorably with vancomycin, although it was less active than teicoplanin against all species of enterococci. The glycylcyclines CL 329,998 and CL 331,002 were overall superior to the class representative tetracycline, to which 70.1% of the isolates were resistant. The activities of moxifloxacin and trovafloxacin against E. faecium and the non-E. faecalis, non-E. faecium species were similar and, against both groups, were fourfold greater than those of ciprofloxacin. However, moxifloxacin displayed only twofold-greater MICs at which 90% of the isolates were inhibited (MIC₉₀) than ciprofloxacin (4 and 8 μg/ml, respectively) against the E. faecalis isolates, while trovafloxacin (MIC₉₀, 2 μg/ml) maintained greater antienterococcal activity against this species. Although 12 isolates (seven E. faecium strains and five non-E. faecalis, non-E. faecium strains) exhibited reduced susceptibility to quinupristin-dalfopristin (MICs, 4 to 8 μg/ml), it was more active than the comparator agent erythromycin. The activity of linezolid, an oxazolidinone antibiotic, could not be directly compared to that of a related agent. Nevertheless, its MIC₉₀ of 1 μg/ml against all species of enterococci is reflective of its activity. With the exception of quinupristin-dalfopristin, all the investigational agents were active against the ampicillin-resistant strains of E. faecium.

Infections caused by multiresistant enterococci are increasingly difficult to treat. However, from the results presented here and from assessments conducted worldwide, it does appear that new agents with reliable antienterococcal activity have the potential to become clinically valuable in the near future.