In Vitro Activity of Ceftazidime-NXL104 against 396 Strains of β-Lactamase-Producing Anaerobes

Abstract

NXL104, a novel β-lactamase inhibitor, was tested at a constant concentration of 4 μg/ml in combination with ceftazidime (CAZ), with and without added metronidazole, against 396 β-lactamase-producing strains of anaerobic bacteria. MIC₅₀/MIC₉₀ values for Bacteroides fragilis and the B. fragilis group were 8/16 and 64/>128 μg/ml, respectively. Although CAZ-NXL104 had limited activity against most anaerobic strains, in combination with metronidazole it shows potential for treating mixed infections involving resistant Enterobacteriaceae and anaerobes.

TEXT

The worldwide emergence of antimicrobial resistance in Enterobacteriaceae containing extended-spectrum β-lactamases (ESBLs), metallo-carbapenemases, and KPC enzymes has radically altered clinicians' empirical selection of antimicrobial agents (3, 6, 10), and new alternative therapeutic agents are urgently needed to treat these serious infections. These multidrug-resistant organisms can also be part of mixed infections, such as intraabdominal and diabetic foot infections, where anaerobic bacteria are important pathogens with their own unique resistance issues. NXL104 is a novel β-lactamase inhibitor with demonstrated activity against AmpC, TEM, SHV, and CTX-M β-lactamases as well as non-metallo-carbapenemases (6, 10); combined with ceftazidime, it is currently in development for treatment of serious intraabdominal infections. While NXL104 has been shown to be effective in vitro in combination with various cephalosporins against these resistant aerobic organisms, there is scant data on its activity against the anaerobic bacteria (6, 10). Consequently, we tested ceftazidime (CAZ) and NXL104 alone and in combination against 396 clinical strains of anaerobic bacteria that are often encountered in mixed infections, selected on the basis of various MIC levels (determined previously) for β-lactamase inhibitor combinations, with an emphasis on the less susceptible strains. CAZ-NXL104 was further tested in combination with various concentrations of metronidazole to look for potential synergy.

The test strains included Bacteroides spp., Parabacteroides spp., Prevotella spp., Porphyromonas spp., Bilophila wadsworthia, Desulfovibrio spp., Campylobacter spp., Fusobacterium spp., Clostridium clostridioforme group spp., and Eggerthella lenta, recovered from patients with primarily intraabdominal infections, identified by standard methods (5, 7). Test strains were stored at −70°C in 20% skim milk. The individual species are listed in footnotes to Tables 1 and 2. For strains that could not be identified by conventional techniques, partial sequencing of 16S rRNA genes was performed as described elsewhere (12).

Table 1.

MIC distributions of Gram-negative anaerobes selected for decreased susceptibility to β-lactams for ceftazidime-NXL104, other β-lactamase inhibitor combinations, metronidazole, imipenem, and clindamycin

Organism (no. of strains) and β-lactamase inhibitora	Cumulative % of strains with MICs (μg/ml) of:
	≤0.125	0.25	0.5	1	2	4	8	16	32	64	128	>128
Bacteroides caccae (15)
Ceftazidime							13	20	27			100
NXL104												100
Ceftazidime-NXL104						7	27	40	60	73	80	100
Metronidazole		7	27	87	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 11)	29	36		43	57	64	79	93	100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 12)	85			100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 4)	100
Amoxicillin-clavulanate 2:1			7	40	47	53	87	93		100
Ampicillin-sulbactam 2:1				13	20	53	67	87	93	100
Piperacillin-tazobactam	7	13	33	53	80	93						100
Imipenem	7	33	53	73	93					100
Clindamycin	14	21		28	42		49					100
B. fragilis (68)
Ceftazidime							7	22	40	56	66	100
NXL104												100
Ceftazidime-NXL104					15	44	57	90	94	97	99	100
Metronidazole	1	7	50	99	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 67)		2	5	30	43	56	92	97	98			100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 68)	80	86	91	94	95	97	100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 34)	100
Amoxicillin-clavulanate 2:1			9	37	69	82	94	99		100
Ampicillin-sulbactam 2:1			2	10	28	63	88	94	99	100
Piperacillin-tazobactam	9	21	47	79	87		94			99		100
Imipenem	10	21	44	78	90	96	99			100
Clindamycin	13	19	37	63	72	73						100
B. ovatus (38)
Ceftazidime							13	20	27			100
NXL104												100
Ceftazidime-NXL104									16	37	68	100
Metronidazole		8	71	89	97	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 37)						6	29	47	85	97	100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 34)	68	71		76		79	91	100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 26)	100
Amoxicillin-clavulanate 2:1			16	34	40	53	71	97	100
Ampicillin-sulbactam 2:1				8	34	42	76	90	100
Piperacillin-tazobactam		5	16	32	61	76	48	97	100
Imipenem	5	47	79	90	100
Clindamycin	5	8	19	35	43				46			100
Bacteroides stercoris/B. uniformis/Bacteroides salyersiae (28)b
Ceftazidime									11			100
NXL104											11	100
Ceftazidime-NXL104						4	7	18	29	64	93	100
Metronidazole	7	18	82	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 25)						4	13	39	74	96	100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 26)	65			69		73	81	88	96	100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 21)	100
Amoxicillin-clavulanate 2:1			11	25	39	71	96	100
Ampicillin-sulbactam 2:1				18	21	36	86	93	100
Piperacillin-tazobactam	11	14	29	68	86	96				100
Imipenem		32	75	93		100
Clindamycin	18	22	26	40	54	58					62	100
B. thetaiotaomicron (52)
Ceftazidime												100
NXL104												100
Ceftazidime-NXL104								2		33	81	100
Metronidazole		6	48	87	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 50)	2		4	6	8	10	27	47	76	92	100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 45)	95					98		100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 25)	100
Amoxicillin-clavulanate 2:1			8	27	33	60	89	94	100
Ampicillin-sulbactam 2:1				8	23	31	56	96	100
Piperacillin-tazobactam		2		4		23	48	73	89	98	100
Imipenem	14	39	77	90	98		100
Clindamycin	2	4	14	22	24	44	54	56				100
Bacteroides vulgatus (23)
Ceftazidime									4		8	100
NXL104												100
Ceftazidime-NXL104								26	61	83	100
Metronidazole		22	61	91	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 18)	28				39	44	61	83	94	100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 21)	63	74	79				100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 14)	100
Amoxicillin-clavulanate 2:1					9	26	83	100
Ampicillin-sulbactam 2:1						4	57	91	100
Piperacillin-tazobactam				13	39	74	91	100
Imipenem		4	39	83	96			100
Clindamycin	30	47		51	60							100
Parabacteroides spp. (50)c
Ceftazidime							2	4		14	22	100
NXL104												100
Ceftazidime-NXL104						4	22	64	80	90	94	100
Metronidazole	2	18	68	96	98	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 43)	10	13	15		20	45	73	83	98			100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 48)	87			89		96	98	100
Ceftazidime-NXL104 plus 2× MTZ MIC (n = 34)	100
Amoxicillin-clavulanate 2:1				6	16	40	68	96	100
Ampicillin-sulbactam 2:1				2	6	10	42	86	94	98	100
Piperacillin-tazobactam				4	26	70	86	96				100
Imipenem	2	12	46	70	90	100
Clindamycin	8		12	30	44	70	74	76			78	100
Prevotella/Porphyromonas spp. (49)d
Ceftazidime			4	10	16	37	43	47	65	78	88	100
NXL104									12	26	43	100
Ceftazidime-NXL104	16	29	31	49	76	92	100
Metronidazole	2	16	61	92	100
Ceftazidime-NXL104 plus 0.5× MTZ MIC (n = 44)	38	56	64	74	97	100
Ceftazidime-NXL104 plus 1× MTZ MIC (n = 32)	88	97	100
Amoxicillin-clavulanate 2:1	16	39	49	65	90	100
Ampicillin-sulbactam 2:1	4	27	45	65	82	100
Piperacillin-tazobactam	100
Imipenem	93	100
Clindamycin	71		73	75							77	100

^aCeftazidime-NXL104, NXL constant at 4 μg/ml; piperacillin-tazobactam, tazobactam constant at 4 μg/ml; n, the number of strains which had metronidazole MICs in range for the calculation.

^bB. stercoris (4 strains), B. uniformis (23 strains), B. salyersiae (1 strain).

^cP. distasonis (35 strains), P. goldsteinii (3 strains), P. merdae (12 strains).

^dPrevotella bivia (11 strains), Prevotella buccae (5 strains), Prevotella corporis (1 strain), Prevotella denticola (2 strains), Prevotella disiens (7 strains), Prevotella intermedia (7 strains), Prevotella melaninogenica (7 strains), Prevotella loeschii (1 strain), Prevotella oris (2 strains), Porphyromonas levii (1 strain), Porphyromonas somerae (5 strains).

Table 2.

MIC distributions of selected anaerobes with decreased susceptibility to β-lactams for ceftazidime, NXL104, and ceftazidime-NXL104

Organism (no. of strains) and β-lactamase inhibitor	Cumulative % of strains with MICs (μg/ml) of:
	≤0.125	0.25	0.5	1	2	4	8	16	32	64	128	>128
Bilophila wadsworthia (20)
Ceftazidime								5	20	85	95	100
NXL104	25					45	70	75				100
Ceftazidime-NXL104	45	50		65			70			80	100
Campylobacter spp. (5)a
Ceftazidime			20		40							100
NXL104							20			60		100
Ceftazidime-NXL104	20	40									80	100
Clostridium spp. (15)b
Ceftazidime							7	14	20	60	67	100
NXL104												100
Ceftazidime-NXL104							7	14	27	60	87	100
Desulfovibrio spp. (9)c
Ceftazidime											22	100
NXL104										11	33	100
Ceftazidime-NXL104										33	89	100
Eggerthella lenta (10)
Ceftazidime												100
NXL104												100
Ceftazidime-NXL104												100
Fusobacterium spp. (10)d
Ceftazidime			20	40	70	80			90			100
NXL104								10	20			100
Ceftazidime-NXL104		20	40	60	80				90			100
Veillonella species
Ceftazidime							25					100
NXL104												100
Ceftazidime-NXL104							25				100

^aC. gracilis (4 strains), C. rectus (1 strain).

^bC. bolteae (8 strains), C. butyricum (2 strains), C. clostridioforme (2 strains), C. hathewayi (2 strains), Clostridium species (1 strain).

^cD. fairfieldensis (8 strains), D. desulfuricans (1 strain).

^dF. gonidiaformans (1 strain), F. mortiferum (3 strains), F. nucleatum (3 strains), F. varium (3 strains).

Comparator drugs included amoxicillin-clavulanate, ampicillin-sulbactam, piperacillin-tazobactam, metronidazole, clindamycin, and imipenem. The antimicrobial agents were reconstituted according to their manufacturers' instructions and stored at −70°C (1). NXL104 was added at a constant concentration of 4 μg/ml to serial dilutions of ceftazidime. MICs are reported for ceftazidime.

The agar dilution test was conducted according to the procedures in CLSI approved standard M11-A7 (1). Serial dilutions of each compound were prepared and added to molten agar to prepare the plates. Three extra dilution series of CAZ-NXL104 with the addition of 1, 0.5, 0.25, or 0.125 μg/ml of metronidazole to each tube of molten agar were prepared to test for possible interactions. Prior to testing, the strains were taken from the freezer and transferred at least twice on supplemented brucella blood agar to ensure purity and good growth. On the day of testing, inocula were prepared in the anaerobic chamber from 48-h plates by suspending organisms into brucella broth to the turbidity of the 0.5 McFarland standard. The inocula were pipetted into the wells of a Steers replicator and applied to the test plates for a final inoculum of ∼10⁵ CFU/spot. Control strains Bacteroides fragilis ATCC 25285, Klebsiella pneumoniae ATCC 700603, and Escherichia coli ATCC 25922 were included each day of testing. (The aerobic organisms were used because they have CLSI-established quality control ranges for ceftazidime and NXL104.) All plates were incubated in the anaerobic chamber at 37°C for 44 h and then examined for growth. The MIC was the drug concentration that completely inhibited growth or resulted in a major reduction of growth compared to the drug-free growth control (1).

Table 1 shows the MIC distribution and cumulative percent at each value for the Bacteroides, Parabacteroides, and Prevotella species against the β-lactamase inhibitor combinations, imipenem, clindamycin, CAZ-NXL104, metronidazole, and CAZ-NXL104 in combination with metronidazole at various concentrations relative to each strain's metronidazole MIC. The number of strains with metronidazole MICs that fell within the calculated range is indicated. Table 2 shows the activity of CAZ and NXL104 alone and in combination against the other organisms tested.

CAZ and NXL104 alone were mostly inactive (overall MIC₉₀ for both was >128 μg/ml). B. fragilis was the most susceptible of the Bacteroides group to ceftazidime alone and in combination with NXL104, with CAZ-NXL104 MIC₅₀ and MIC₉₀ values of 8 and 16 μg/ml, respectively. Most of the Parabacteroides distasonis group showed elevated MICs to all the β-lactamase inhibitor combinations and to imipenem, with the CAZ-NXL104 MIC₅₀ and MIC₉₀ values at 16 and 64 μg/ml, respectively. The strains that were susceptible to other β-lactamase inhibitor combinations were also more susceptible to CAZ-NXL104 (data not shown). The indole-positive strains, including Bacteroides thetaiotaomicron, Bacteroides ovatus, and Bacteroides uniformis, were the most resistant to ceftazidime and CAZ-NXL104, regardless of susceptibility to the other β-lactamase inhibitor combinations, with most MICs at ≥128 μg/ml. For all members of the B. fragilis group, the MIC₅₀ and MIC₉₀ values were 64 and >128 μg/ml, respectively. For Prevotella and Porphyromonas spp., the MIC₅₀ and MIC₉₀ values were 2 and 4 μg/ml, respectively. For many of the Bacteroides fragilis and Prevotella strains, the addition of metronidazole at 0.5× or 1× their MIC resulted in a further reduction in the MIC of CAZ-NXL104 by one or more dilutions for some strains. But even at its original MIC, the metronidazole MICs for several organisms (B. fragilis, 3 strains; B. ovatus, 7 strains; B. thetaiotaomicron, 2 strains; B. uniformis, 9 strains; P. distasonis, 2 strains) appeared higher by 1 to 4 dilutions, suggesting possible antagonism. However, upon further examination, the MICs at the next-higher metronidazole concentration for those strains were ≤0.125 μg/ml, indicating that there were likely minor differences in MIC test interpretations and not antagonism between the drugs. This is shown in Table 1. For most strains, the effect was additive or indifferent.

NXL104 alone showed activity of ≤4 μg/ml against 9 of 20 strains of Bilophila wadsworthia, and CAZ-NXL104 was more active than either drug alone, with MICs of ≤1 μg/ml for 13 of the 20 strains. Two of the Campylobacter species were susceptible to ≤0.25 μg/ml CAZ-NXL104. Among the other organisms, 8 of 10 strains of Fusobacterium (the exception was 2 strains of F. mortiferum) were inhibited by ≤2 μg/ml of CAZ-NXL104. NXL104 did not increase the activity of ceftazidime against β-lactamase-producing strains of the Clostridium clostridioforme, Clostridium bolteae, Clostridium butyricum, Desulfovibrio spp., penicillin-resistant strains of Veillonella, or the cephalosporin-resistant strains of Eggerthella lenta.

Overall, piperacillin-tazobactam was the most active of all the β-lactamase inhibitor combinations tested, with only 1 strain each of B. fragilis and B. thetaiotaomicron and two strains of P. distasonis resistant, with MICs of ≥128 μg/ml, and 7 other strains with intermediate MICs of 64 μg/ml. Sixty-eight percent of all Bacteroides strains were fully susceptible to ampicillin-sulbactam at a breakpoint of 8 μg/ml compared to 59% for amoxicillin-clavulanate at a breakpoint of 4 μg/ml. Predictably, metronidazole showed the greatest activity against the anaerobes tested, with overall MIC₅₀ and MIC₉₀ values of 0.5 and 1 μg/ml, respectively. Clindamycin resistance was present in 52% of the Bacteroides fragilis group species, 30% of the Parabacteroides species, and 28% of Bacteroides fragilis strains. Imipenem resistance was present in only five isolates of Bacteroides species, but many showed elevated MICs of 2 to 4 μg/ml (Table 1). These included 12 B. fragilis strains, 16 other B. fragilis group species, and 13 Parabacteroides species.

Resistance to beta-lactam antibiotics can occur through a variety of mechanisms, both for aerobic Gram-negative rods, especially the Enterobacteriaceae, and for anaerobic bacteria. For anaerobes, the most common resistance mechanism is β-lactamase production, as detected by the presence of genes such as cepA or cfiA, which also codes for carbapenemase (2, 4, 8, 9, 11). Other mechanisms include reduced penetration of the antibiotic into the periplasmic space by loss of porins or alterations in the penicillin binding proteins; efflux pumps may also play an important role in certain strains (13). Some organisms contain several resistance mechanisms and can be multiresistant to a wide range of beta-lactam antibiotics. Thus, the MIC range for β-lactam antibiotics against Bacteroides spp. is quite variable and cannot be explained by any one mechanism.

Strains selected for this study included those with reduced susceptibility to the standard β-lactamase inhibitor combinations to see if NXL104 would show enhanced activity in counteracting their β-lactamases. While there are no anaerobic breakpoints established for ceftazidime, the breakpoint for full susceptibility for other third-generation cephalosporins is ≤16 μg/ml (1). CAZ-NXL104 was remarkably active against B. fragilis, with 61 of 68 (90%) strains inhibited by ≤16 μg/ml. Although CAZ-NXL104 showed limited activity against most of the strains, in combination with metronidazole, CAZ-NXL104 is potentially effective for treating serious polymicrobial infections containing multidrug-resistant Enterobacteriaceae and anaerobic Gram-negative organisms.