Treatment of Bacteremia Due to ESBL-Producing Klebsiella pneumoniae

Abstract & Commentary

Synopsis: A carbapenem (most often imipenem) was found in a prospective observational study to be the optimal choice for therapy of bacteremia due to ESBL-producing Klebsiella pneumoniae.

Source: Paterson DL, et al. Antibiotic Therapy for Klebsiella pneumoniae Bacteremia: Implications of Production of Extended-Spectrum B-Lactamases. Clin Infect Dis. 2004;39:31-37.

Paterson and colleagues performed a prospective, observational study of 440 patients with 455 episodes of bacteremia due to Klebsiella pneumoniae who were managed at 12 centers in Asia, Africa, Europe, and the Americas in 1996-1997. Antibiotic choice was at the discretion of the treating physicians.

Eighty-five (19%) of the episodes were due to an ESBL-producing K. pneumoniae; 20 (24%) of these episodes ended with the patient’s death. The mortality rate was 14% in those who received early antibiotic therapy with in vitro activity against the isolated pathogen and 64% (P = 0.001) in those whose antibiotics lacked such activity. Administration of a carbapenem (mostly imipenem) was associated with significantly lower mortality at 14 days than was administration of other antibiotics with in vitro activity against the infecting pathogen. Multivariate analysis demonstrated that use of a carbapenem during the 5 days after the onset of bacteremia with an ESBL-producing K. pneumoniae was independently associated with lower mortality (OR, 0.09; 95% CI, 0.01-0.65; P = 0.09) and was superior to monotherapy with other agents. There was no difference in the incidence of superinfections.

Comment by Stan Derensinski, MD, FACP

Extended spectrum beta lactamases (ESBL) are enzymes whose evolution from standard plasmid mediated enzymes, such as TEM and SHV, has resulted in their ability to hydrolyze 3rd generation cephalosporins and aztreonam, but not cephamycins or carbapenems. ESBLs, of which more than 200 genetically distinct types have been identified, have been found in most of the Enterobacteriaciae, but are most highly prevalent in K. pneumoniae, Klebsiella oxytoca, and Escherichia coli. It is currently recommended by the NCCLS, that all isolates of these 3 species, with an MIC > 2.0 mg/mL, be tested against cefpodoxime, ceftazidime, aztreonam, cefotaxime, or ceftriaxone, and examined for the presence of an ESBL.1 The sensitivity of screening for ESBLs in enteric bacteria is further increased by testing with more than 1 of these 5 antibiotics. For instance, the production of an ESBL with relative specificity for cefotaxime (CTX-M) would be missed by screening with ceftazidime. The NCCLS recommends that ESBL production be confirmed by testing both cefotaxime and ceftazidime alone and in combination with clavulanic acid, which, like tazobactam and sulbactam, inhibits most ESBLs, resulting in increased susceptibility to the cephalosporin.

Such phenotypic confirmatory testing does not, however, detect all ESBLs. The presence of these enzymes maybe masked by, for example, the presence of an AmpC ß-lactamase, since the hydrolysis of third generation cephalosporins by AmpC is not inhibited by clavulanic acid. The coexistence of inhibitor-resistant TEM or hyperproduction of TEM or SHV b-lactamases in ESBL-producing organisms may also lead to false-negative phenotypic screening tests.

The NCCLS recommends that all isolates confirmed to produce an ESBL be reported as resistant to all penicillins, cephalosporins and to aztreonam. There is controversy, however, regarding the potential efficacy of cefepime against infections caused by ESBL-producers.

In contrast to monobactams and third generation cephalosporins, carbapenems, such as imipenem, are resistant to hydrolysis by these enzymes and thus retain their activity against ESBL-producers. While non-b-lactam antibiotics are not affected by ESBLs, the large plasmids which carry the genes for these enzymes often carry genes encoding resistance to other antibiotics, including fluoroquinolones and aminoglycosides.2

While third generation cephalosporins are ineffective in the treatment of these infections, the role of cefepime, a putative fourth generation cephalosporin, remains incompletely understood. In a recent randomized trial of treatment of nosocomial pneumonia in intensive care patients, cefepime failed in 4 of 13 patients infected with an ESBL-producing organisms while no failures were observed in 10 patients treated with imipenem.3 Nonetheless, successes have been reported elsewhere.

This study has a number of drawbacks, not the least of which is its lack of randomization. Furthermore, as Peterson et al point out, they did not collect information about antibiotic dosing. Systematic underdosing of a particular antibiotic could bias the results against that agent. Nonetheless, these results provide a large database of use to the clinician faced with therapeutic choices in the hospital setting.


1. National Committee for Clinical Laboratory Standards. 1999. Performance standards for antimicrobial susceptibility testing. NCCLS approved standard M100-S9. National Committee for Clinical Laboratory Standards, Wayne, PA.

2. Procop GW, et al. Cross-Class Resistance to Non-Beta-Lactam Antimicrobials in Extended-Spectrum Beta-Lactamase-Producing Klebsiella pneumoniae. Am J Clin Pathol. 2003;120:265-267.

3. Zanetti G, et al. Cefepime Versus Imipenem-Cilastatin for Treatment of Nosocomial Pneumonia In Intensive Care Unit Patients: A Multicenter, Evaluator-Blind, Prospective, Randomized Study. Antimicrob Agents Chemother. 2003;47:3442-3447.

Stan Deresinski, MD, FACP, Clinical Professor of Medicine, Stanford; Associate Chief of Infectious Diseases, Santa Clara Valley Medical Center, and Editor for Infectious Disease Alert.