Comparison of Activities of Broad-Spectrum β-Lactam Compounds against 1,128 Gram-Positive Cocci Recently Isolated in Cancer Treatment Centers

Abstract

We report the in vitro activities of broad-spectrum β-lactam antimicrobials tested against 1,128 gram-positive pathogens recently isolated from cancer patients. Cefepime and imipenem were more active than ceftazidime and ceftriaxone against these organisms. Only vancomycin demonstrated reliable activity against oxacillin-resistant staphylococci, Enterococcus spp., and Corynebacterium spp. The spectrum of gram-positive organisms against which cefepime and imipenem have activity provides an advantage over ceftazidime as empiric therapy for cancer patients, potentially reducing the need for the empiric addition of glycopeptides.

Often immunosuppressed by cytotoxic therapies, invasive procedures, and indwelling catheters, patients with neoplastic disease are extremely vulnerable to bacterial infection. In these patients, gram-positive organisms have long since replaced gram-negative bacilli as the predominant pathogens (7). Since traditional empiric antimicrobial regimens for cancer patients who are neutropenic and febrile emphasize broad coverage for gram-negative pathogens (11), these patients may remain vulnerable to infections with gram-positive organisms that are resistant to some extended-spectrum cephalosporins or antipseudomonal penicillins.

There is therefore a clear role for new broad-spectrum antimicrobial agents which are well tolerated and which provide adequate activity against gram-negative organisms with improved activity against gram-positive organisms for use in empiric regimens for hematology-oncology patients. Cefepime and imipenem have previously been demonstrated to have greater in vitro activity against both gram-positive and gram-negative bacteria than expanded-spectrum cephalosporins (1, 4, 10), and they have already been applied as effective empiric therapy for cancer patients with fever and neutropenia (13, 15). However, since resistance among the gram-positive pathogens increases with each passing year (6), it is extremely important that surveillance studies be performed to track resistance trends among these patients.

This surveillance trial was designed to provide reliable quantitative information for physicians regarding the in vitro activities of cefepime, imipenem, and other broad-spectrum β-lactams tested against important gram-positive pathogens isolated in cancer treatment hospitals.

Study design.

Ten cancer treatment centers were recruited to participate in the study. Each hospital laboratory was asked to test 150 gram-positive strains isolated from oncology patients, including up to 50 consecutive strains each of Staphylococcus aureus and coagulase-negative staphylococci (CoNS), up to 15 strains of Enterococcus spp., and 10 strains each of Streptococcus pneumoniae, beta-hemolytic streptococci, and viridans group streptococci, as well as approximately 5 strains of other assorted gram-positive pathogens (Bacillus spp. and Corynebacterium spp., etc.). Organism identification was performed according to the protocols in place at each participating institution’s laboratory. All isolates were processed between 1 March and 31 June 1998. When an institution could not attain the goal for the number of isolates of a given pathogen, it was allowed to test some isolates that had recently been archived at its institution. Bloodstream and other sterile-site isolates were encouraged, but organisms could be obtained from other body sites provided that they were judged to be significant pathogens. Duplicate patient strains were not accepted. Recording forms and susceptibility testing reagents were provided to each laboratory by the study coordinator at the University of Iowa, Iowa City. A total of 1,128 strains were tested, and results for 1,119 of them are presented in Table 1.

TABLE 1.

In vitro antimicrobial susceptibility testing results for 1,119 gram-positive organisms isolated from cancer treatment medical centers^a

Organism (no. tested)	Antimicrobial agent	MIC (μg/ml)		% by categoryb
		50%	90%	Susceptible	Resistant
S. aureus
Oxacillin susceptible (240)	Cefepime	4.0	6.0	98.8	0.8
	Ceftazidime	12	16	15.4	5.4
	Ceftriaxone	6.0	24	77.1	10.0
	Imipenem	0.032	0.064	100.0	0.0
	Vancomycin	1.0	1.0	100.0	0.0
Oxacillin resistant (106)	Cefepime	>32	>32	0.0	100.0
	Ceftazidime	>256	>256	0.0	100.0
	Ceftriaxone	>32	>32	0.0	100.0
	Imipenem	>32	>32	0.0	100.0
	Vancomycin	1.0	1.5	100.0	0.0
CoNS
Oxacillin susceptible (101)	Cefepime	1.0	3.0	100.0	0.0
	Ceftazidime	8.0	12.0	75.2	3.0
	Ceftriaxone	2.0	12.0	88.1	5.0
	Imipenem	0.023	0.047	100.0	0.0
	Vancomycin	1.5	2.0	100.0	0.0
Oxacillin resistant (308)	Cefepime	>32	>32	0.0	100.0
	Ceftazidime	>256	>256	0.0	100.0
	Ceftriaxone	>32	>32	0.0	100.0
	Imipenem	16	>32	0.0	100.0
	Vancomycin	1.5	2.0	100.0	0.0
Enterococcus spp.c (150)	Penicillin	4.0	>32	60.7	38.7
	Cefepime	>32	>32	0.7 (≤8.0)	98.7 (≥32)
	Ceftazidime	>256	>256	1.3 (≤8.0)	96.7 (≥32)
	Ceftriaxone	>32	>32	3.3 (≤8.0)	95.3 (≥64)
	Imipenem	2.0	>32	60.0 (≤4.0)	38.7 (≥16)
	Vancomycin	2.0	>256	75.3	22.7
Viridans group streptococci (62)	Penicillin	0.125	3.0	66.1	11.3
	Cefepime	0.25	3.0	75.8 (≤0.5)	11.3 (≥2.0)
	Ceftazidime	2.0	16	12.9 (≤0.5)	51.6 (≥2.0)
	Ceftriaxone	0.19	2.0	83.9	11.3
	Imipenem	0.064	1.0	88.7 (≤0.5)	11.3
	Vancomycin	0.5	1.0	100.0	0.0
S. pneumoniae (59)	Penicillin	0.023	1.5	59.3	6.8
	Cefepime	0.125	2.0	69.5	10.2
	Ceftazidime	0.75	16	52.5 (≤0.5)	37.3 (≥2.0)
	Ceftriaxone	0.047	1.0	79.7	3.4
	Imipenem	0.016	0.25	76.3	3.4
	Vancomycin	0.5	0.75	100.0	0.0
Beta-hemolytic streptococcid (44)	Penicillin	0.064	0.094	100.0	0.0
	Cefepime	0.125	0.19	97.7 (≤0.5)	0.0 (≥2.0)
	Ceftazidime	0.75	1.0	34.1 (≤0.5)	6.8 (≥2.0)
	Ceftriaxone	0.094	0.125	100.0	0.0
	Imipenem	0.032	0.064	100.0 (≤0.12)	0.0
	Vancomycin	0.38	1.0	100.0	0.0
Corynebacterium spp.e (17)	Penicillin	1.0	>32	52.9	47.1
	Cefepime	>32	>32	47.1	52.9
	Ceftazidime	>256	>256	5.9	82.4
	Ceftriaxone	4.0	>32	52.9	47.1
	Imipenem	6.0	>32	47.1	47.1
	Vancomycin	0.38	0.75	100.0	0.0
Bacillus spp. (21)	Penicillin	>32	>32	38.1	61.9
	Cefepime	>32	>32	23.8	71.4
	Ceftazidime	>256	>256	23.8	76.2
	Ceftriaxone	>32	>32	28.6	71.4
	Imipenem	0.125	1.5	100.0	0.0
	Vancomycin	1.0	4.0	90.5	9.5
Micrococcus spp. (11)	Penicillin	0.094	0.25	72.7	27.3
	Cefepime	0.094	0.19	100.0	0.0
	Ceftazidime	16	32	36.4	27.3
	Ceftriaxone	0.25	0.5	100.0	0.0
	Imipenem	0.047	0.064	100.0	0.0
	Vancomycin	0.19	0.25	100.0	0.0

^aNote that only 9 of the 1,128 strains tested are not listed (three Lactobacillus spp., one Leuconostoc sp., three Listeria monocytogenes strains, and two Streptococcus bovis strains). 

^bCriteria are those published by the NCCLS (9). The MIC breakpoint is listed in parentheses if no criteria are found in the NCCLS document. Breakpoints used for staphylococci were applied to the Corynebacterium spp., Bacillus spp., and Micrococcus spp. For oxacillin-resistant staphylococci, resistance to all other β-lactams is assumed. 

^cIncludes Enterococcus casseliflavus (1 strain), E. faecalis (8 strains), Enterococcus faecium (18 strains), and 123 strains not identified to species level. 

^dIncludes Streptococcus serogroups B (4 strains), C (1 strain), and G (1 strain), as well as 38 untyped strains. 

^eIncludes Corynebacterium jeikeium (four strains), Corynebacterium spp. (eight strains), and diphtheroids (five strains). 

Susceptibility testing methods.

Susceptibility testing was performed at each participating center by using the E test (AB Biodisk, Solna, Sweden) according to an explicit test protocol (14). The following antimicrobials were evaluated against strains of staphylococci: cefepime, ceftazidime, ceftriaxone, vancomycin, imipenem, and oxacillin (by disk diffusion test). For nonstaphylococcal gram-positive isolates the same antimicrobials were tested, except that a penicillin E-test strip was substituted for the 1-μg oxacillin disk (8). Interpretive criteria for each antimicrobial tested were published by the National Committee for Clinical Laboratory Standards (NCCLS) (9), except where noted in Table 1.

Isolates with susceptibility patterns that were unusual or of potential epidemiologic importance (e.g., non-vancomycin-susceptible organisms, beta-hemolytic streptococci for which the penicillin MIC was >0.125 μg/ml, and other streptococci for which the penicillin MICs were >2 μg/ml) were referred to the monitor laboratory for confirmation and additional testing. The monitor’s results were used in the final analysis.

Quality assurance.

Each participating laboratory performed quality control by testing Enterococcus faecalis ATCC 29212 and S. aureus ATCC 29213 at least weekly or a total of five times during the study period. The overall rate of E-test-result agreement with expected MIC quality control ranges was 99.6%, and the range for the 10 hospitals was 97.3 to 100.0%.

Overall susceptibility test data are listed in Table 1, and these data are discussed by genus groups.

Staphylococci.

Of 346 S. aureus isolates tested, 69% were susceptible to oxacillin (8, 9). All antimicrobials tested, with the exception of ceftazidime (MIC at which 90% of the isolates were inhibited [MIC₉₀], 16 μg/ml; 15.4% susceptible) and ceftriaxone (MIC₉₀, 24 μg/ml; 77.1% susceptible), demonstrated activity against the oxacillin-susceptible strains. Nearly all isolates tested were susceptible to vancomycin (100.0%), imipenem (100.0%), and cefepime (98.8%). Vancomycin (MIC₉₀, 1.5 μg/ml; 100.0% susceptible) was the only tested agent that demonstrated potent activity against oxacillin-resistant S. aureus.

Of 409 CoNS isolates tested only 25% were oxacillin susceptible. Vancomycin (MIC₉₀, 2.0 μg/ml; 100.0% susceptible), imipenem (MIC₉₀, 0.047 μg/ml; 100.0% susceptible), and cefepime (MIC₉₀, 3.0 μg/ml; 100.0% susceptible) all demonstrated complete activity against oxacillin-susceptible CoNS. Both ceftazidime (MIC₉₀, 12.0 μg/ml; 75.2% susceptible) and ceftriaxone (MIC₉₀, 12.0 μg/ml; 88.1% susceptible) were fourfold less active than cefepime. As with oxacillin-resistant S. aureus, vancomycin was the only agent tested that demonstrated activity against oxacillin-resistant CoNS (MIC₉₀, 2.0 μg/ml; 100.0% susceptible). Notably, there were no glycopeptide-intermediate or -resistant strains of staphylococci detected in these 10 cancer treatment centers (3, 9).

Enterococcus spp.

Table 1 includes all Enterococcus spp. processed in this surveillance study, since nearly all participating laboratories do not routinely identify isolates to the species level. Vancomycin was the most active agent, although 22.7% of strains were resistant.

Streptococci.

Only 66.1% of viridans group streptococcal isolates were susceptible to penicillin, and for 11.3% of the strains the MICs were >2 μg/ml (Table 1). Cefepime, ceftriaxone, vancomycin, and imipenem were the most active agents tested (75.8 to 100.0% susceptible). Ceftazidime was least effective against viridans group streptococci and had the narrowest spectrum of activity (MIC₉₀, 16 μg/ml; 12.9% susceptible, with an MIC breakpoint of ≤0.5).

Over 40% of the 59 pneumococcal isolates were not susceptible (MIC, ≥0.12) to penicillin, and high-level resistance (MIC, ≥2 μg/ml) was observed in 6.8% of isolates. Among the cephalosporins, ceftriaxone (MIC₉₀, 1.0 μg/ml; 79.7% susceptible) and cefepime (MIC₉₀, 2.0 μg/ml; 69.5% susceptible) were more active than penicillin against S. pneumoniae, while ceftazidime (MIC₉₀, 16 μg/ml; 52.5% susceptible) was less active against this organism.

All of the antimicrobials tested had some activity against beta-hemolytic streptococci (MIC₅₀s, ≤0.75 μg/ml), although ceftazidime (MIC₉₀, 1.0 μg/ml) was four- to eightfold less active than cefepime (MIC₉₀, 0.19 μg/ml) or ceftriaxone (MIC₉₀, 0.125 μg/ml).

Other species.

Imipenem was the most active agent against Bacillus strains (MIC₉₀, 1.5 μg/ml). Interestingly, 2 strains (MICs, 12 and >256 μg/ml) of the 21 tested demonstrated resistance to vancomycin (MIC₉₀, 4.0 μg/ml; 90.5% susceptible). Vancomycin was active against Corynebacterium spp. (MIC₉₀, 0.75 μg/ml; 100% susceptible). The most active agents against the Micrococcus strains included vancomycin, imipenem, ceftriaxone, and cefepime (100.0% of isolates were susceptible). Ceftazidime was the least active agent against these strains (MIC₉₀, 32 μg/ml; 36.4% susceptible).

In the majority of cases when a bacterial pathogen is isolated from a cancer patient, that pathogen is gram positive (7). In addition to staphylococci and enterococci, the viridans group streptococci have become increasingly prevalent, accounting for 10 to 39% of bacteremias among neutropenic cancer patients and demonstrating increasing antimicrobial resistance (2, 12). Not surprisingly, vancomycin is the drug most frequently added to (non-vancomycin-containing) empiric regimens in febrile neutropenic patients.

In vitro studies have demonstrated cefepime and imipenem to be more active against gram-positive organisms than ceftazidime (1, 4, 10). Our results confirm that cefepime and imipenem have a distinct advantage over ceftazidime in the spectrum of gram-positive organisms against which they have activity and in potency, particularly when the pathogen is an oxacillin-susceptible Staphylococcus sp., a Micrococcus sp., or any one of a variety of streptococci. For some gram-positive organisms tested, cefepime and ceftriaxone demonstrated comparable spectrums of activity. However, ceftriaxone is not often recommended for empiric therapy in cancer patients, particularly those with neutropenia, due to its limited activity against several nosocomial gram-negative pathogens, especially Pseudomonas aeruginosa (5).

In conclusion, this surveillance study of more than 1,100 gram-positive organisms recently isolated from 10 cancer treatment centers in the United States reveals that cefepime and imipenem offer wider empiric coverage than widely used expanded-spectrum cephalosporins. Utilization of one of these agents rather than ceftazidime as part of initial empiric regimens should improve the likelihood that prevalent gram-positive pathogens will be inhibited. This in turn may reduce the need for additional therapy with glycopeptides. However, for oxacillin-resistant Staphylococcus spp. and Corynebacterium spp., vancomycin and teicoplanin remain the only effective agents, and for many glycopeptide-resistant Enterococcus spp., adequate therapeutic options are not readily available. There remains an urgent need for the development of novel agents that are highly active against the increasingly resistant gram-positive pathogens found in cancer treatment centers and other hospitals.