Production of KPC-2 Alone Does Not Always Result in β-Lactam MICs Representing Resistance in Gram-Negative Pathogens

LETTER

In 2010, the CLSI lowered the susceptibility breakpoints for many cephalosporins and the carbapenems to streamline the screening of potential carbapenemase-producing Enterobacteriaceae. Despite this effort, cases of carbapenemase-producing Enterobacteriaceae that demonstrate susceptibility to the cephalosporins and the carbapenems have been reported (4). The existence of such isolates suggests that lowering the β-lactam breakpoints is insufficient to detect potential carbapenemase producers, a finding which is important for selection of appropriate therapy. Clinically encountered KPC-producing isolates often produce multiple β-lactamases and may have other resistance mechanisms, such as porin modifications that impact susceptibility to β-lactam antibiotics. This makes it difficult to assess the specific contribution of bla_KPC to β-lactam susceptibilities (5, 7, 10). The effect of inoculum concentration on KPC producers has been addressed for the drugs imipenem and meropenem but only using clinical isolates that typically possess multiple β-lactam resistance mechanisms (1, 9). Therefore, the purpose of this study was to examine the impact of KPC-2 alone on β-lactam MICs at both standard and higher inocula.

bla_KPC transformants were constructed by cloning the bla_KPC structural gene and its upstream promoter region into a low-copy-number, broad-host-range vector before transformation into β-lactam-susceptible strains of Escherichia coli, Klebsiella pneumoniae, Enterobacter cloacae, and Pseudomonas aeruginosa as described previously (8). To examine the influence of bla_KPC on antimicrobial susceptibilities, the MICs of ceftazidime, ceftriaxone, cefepime, imipenem, and ertapenem were determined using the CLSI agar dilution method and results were interpreted according to 2010 CLSI breakpoints (2, 3). The observed β-lactam MICs for the bla_KPC-2 transformants at the recommended inoculum of 10⁴ CFU per spot and a high inoculum of 10⁶ CFU per spot are shown in Table 1. Although in tests with the standard inoculum, the MICs of cefepime increased for the Enterobacteriaceae transformants and the MICs of ceftazidime increased for the K. pneumoniae transformant, the increases were not sufficient for the transformants to be categorized as resistant. For all transformants, the MICs of the five β-lactams tested increased above the resistance breakpoint when the higher inoculum was used. An inoculum effect (fold changes ranging from 8- to >64-fold) was observed for the Enterobacteriaceae transformants when tested against ceftriaxone, cefepime, imipenem, and ertapenem. An inoculum effect for ceftazidime was observed with E. coli, K. pneumoniae, and P. aeruginosa but not with E. cloacae, which demonstrated only a 2-fold increase in MIC. Ceftazidime and imipenem MICs increased 8-fold for the P. aeruginosa transformant, but the MICs of ceftriaxone, cefepime, and ertapenem exceeded the upper limit of this assay (512 μg/ml) and the fold increase was not able to be determined.

TABLE 1.

Antimicrobial susceptibility with recommended and increased inocula, determined by agar dilution of bla_KPC transformants

Speciesa	Inoculum amt	β-Lactam MIC (μg/ml)b
		CAZ	CRO	FEP	IPM	ERT
E. coli	104	0.06	0.007	0.06	0.12	0.007
	106	0.25	0.015	2	0.25	0.015
E. coli + KPC-2	104	16	64	8	8	4
	106	256	512	>512	128	256
K. pneumoniae	104	0.12	0.015	0.03	0.12	0.007
	106	0.25	0.03	0.06	0.5	0.03
K. pneumoniae + KPC-2	104	4	32	4	8	4
	106	64	>512	512	512	128
E. cloacae	104	1	4	0.06	0.12	0.06
	106	16	16	0.12	2	0.5
E. cloacae + KPC-2	104	16	32	4	2	2
	106	32	>512	256	32	64
P. aeruginosa	104	1	32	1	1	8
	106	2	128	4	2	16
P. aeruginosa + KPC-2	104	64	>512	256	4	512
	106	512	>512	>512	32	>512

^aE. coli K-12 259, K. pneumoniae KP23, E. cloacae ATCC 13047, P. aeruginosa PAO1.

^bCAZ, ceftazidime; CRO, ceftriaxone; FEP, cefepime; IPM, imipenem; ERT, ertapenem.

This study indicates that the activity of the five β-lactam drugs studied, including ertapenem, which is the most frequently used drug for KPC screening, can be affected by variations in inoculum density when testing KPC-producing E. coli, K. pneumoniae, E. cloacae, and P. aeruginosa. These data suggest that the MIC of these antibiotics may not be a reliable indicator for the presence of bla_KPC. The most recent CLSI guidelines remove the need for clinical laboratories to screen isolates for carbapenemases for therapeutic purposes, recommending such testing as optional for only epidemiological and infection control inquiries. For laboratories that continue to follow the 2009 CLSI guidelines, the marginal increases in ceftazidime and cefepime MICs observed in this study are of particular concern, as resistance to an expanded-spectrum cephalosporin is used as a criterion for further carbapenemase screening. Thus, clinical isolates with similar susceptibility patterns would not meet the criteria for additional screening if ceftazidime or even cefepime were used for initial susceptibility testing. The data presented in this report indicate that clinical isolates that produce only a KPC enzyme in the absence of other mechanisms conferring resistance to oxyimino-cephalosporins and/or carbapenems may not be recognized as possible KPC producers during susceptibility testing. Undetected KPC-only-producing isolates increase the risk of spread for these isolates and/or the transposon carrying bla_KPC. These data support the argument that lowering the breakpoints of certain β-lactam antibiotics may not be sufficient for preventing the spread of KPC-mediated carbapenem resistance (6). Therefore, it is necessary for laboratories to screen for the presence of KPC rather than simply lowering susceptible breakpoints.