PhillyEBM.comESBL-Producing Bacteria and Third Generation Cephalosporins: Implications for Empiric Therapy and Formulary Changes
1. Does the restricted usage of ceftazidime result in reduced third-generation cephalosporin resistance?
2. Does the replacement of third generation cephalosporins with cefepime for empiric treatment of bacterial infection result in improved clinical outcomes?
Extended spectrum beta-lactamases are plasmid-mediated enzymes that confer resistance to penicillins, third-generation cephalosporins and aztreonam, and are inhibited by beta-lactamase inhibitors. These are distinct from other beta-lactamases capable of hydrolyzing extended-spectrum cephalosporins (e.g. carbapenemases, AmpC beta-lactamases). During the past two decades, the prevalence of ESBL-producing Escherichia coli and Klebsiella spp. (ESBL-EK) increased markedly, worldwide. This change in resistance has been found to be associated with the use of extended spectrum (third generation) cephalosporins. In particular, ceftazidime use has been implicated as a significant factor in the induction of ESBLs and selection of highly-resistant Gram negative bacteria. Cefepime, a fourth-generation cephalosporin, has been found to be more stable against ESBLs as well as other beta-lactamases, and has not been associated with ESBL-induction.
In efforts to control the rising incidence of ESBL-producing bacteria and to prevent inadequate empiric therapy for Gram negative blood stream infections, hospitals increasingly have adopted various antimicrobial formulary interventions. The CHOP antibiotic stewardship program initiated its own formulary intervention July of 2007 to use cefepime, a fourth generation cephalosporin, in place of ceftazidime for empiric therapy of suspected gram negative bacterial infections, in certain patient populations.
1. We found no studies examining the effects of ceftazidime restriction alone and subsequent change in antimicrobial resistance patterns, in the non-outbreak setting. Although difficult to isolate the effect of ceftazidime restriction vs. effect of ceftazidime substitute (cefepime, piperacillin/tazobactam, imipenem) on rates of ESBL-producing isolates and subsequent reduced third generation cephalosporin resistance, studies of antimicrobial resistance patterns at several institutions after formulary changes demonstrate that restriction of third generation cephalosporin use results in improved susceptibility of E. coli and K. pneumonia to this same class of cephalosporins. These studies also suggest that restriction in third generation cephalosporin use may result in a decrease in prevalence of ESBL-producing isolates. Substitution of third generation cephalosporins with other antibiotics may result in the subsequent emergence of other resistant organisms.
2. In vitro studies show that cefepime is a better choice for treatment of ESBL-producing gram negative bacteria, and retrospective studies show that cefepime is more successful in treating ESBL-producing gram negative bacteria. However, current data show no difference in mortality for patients empirically treated with cefepime versus ceftazidime. There are slightly better outcomes in length of hospital stay and clinical course for those treated with cefepime versus ceftazidime, but these differences in outcome have not been statistically significant. In children with cancer presenting with fever and neutropenia, cefepime is equally effective as ceftazidime as monotherapy. From an economic outcome standpoint, there are data to show that cefepime is associated with lower costs, which include outright cost of the antibiotic, and concomitant antibiotic therapy and drugs used to treat adverse event and clinical failures.
Question 1:
Lipworth et al conducted a prospective, pre/post-intervention comparative study investigating the results of a formulary intervention at two university hospitals: Hospital A, an academic tertiary care center; Hospital B, an urban community hospital. Two sequential formulary restrictions were made: the first, replacing ceftriaxone with ampicillin/sulbactam +/- gentamicin with few exceptions (e.g. empiric treatment of bacterial meningitis); the next, replacing ceftazidime with cefepime. Five years after implementation of formulary restrictions, the use of ceftriaxone and ceftazidime decreased significantly, at both hospitals. At Hospital A the prevalence of ESBL-EK decreased significantly. There was no statistically significant decrease seen at Hospital B. Patients with ESBL-EK at hospital B were more likely to have been transferred from a long-term care facility (LTCF), to have advanced age and to have decubitus ulcers. This may have been reflective of the fact that these patients from LTCF may have already been colonized with EBSL-EK at time of admission. Usage of cefepime, fluoroquinolone and gentamicin increased at both hospitals, and usage of ampicillin-sulbactam increased at Hospital A. There was a subsequent increase in fluoroquinolone resistance among several pathogens, as well as increase in ampicillin/sulbactam-resistant K. pneumoniae at Hospital A. [1]
Empey et al performed a retrospective, pre/post-intervention comparative study investigating effects of a hospital formulary intervention at a university-affiliated, tertiary care, level one trauma center for adults and children. Hospital-wide use of third generation cephalosporins was replaced with cefepime. Penicillin-based therapy (e.g. ampicillin-sulbactam for community-acquired infections, pipercillin-tazobactam for nosocomial infections) was encouraged; cefepime was added to the formulary, intended for penicillin-allergic patients with nosocomial infections. Six months after formulary change there was a statistically significant decrease in infection rate from ceftazidime-resistent K. pnuemoniae, with additional benefit of decrease in piperacillin-resistant and ceftazidime-resistant P. aeruginosa. Actual prevalence of ESBL-producing isolates was not reported, but presumed based upon extended-spectrum cephalosporin resistance. [2]
A prospective, pre/post-intervention comparative study was conducted by Du et al. A hospital formulary at a university-affiliated tertiary care hospital in China replaced usage of third generation cephalosporins with cefepime or carbapenems in the general intensive care unit, with exception of perioperative prophylaxis. One year after intervention, use of cefepime increased significantly and there was no statistically significant change in imipenem use. There was a statistically significant decrease in resistance of E. coli and Klebsiella spp to cefotaxime, ceftriaxone and ceftazidime. Prevalence of ESBL-producing E. coli and Klebsiella species decreased non-significantly. [3]
A prospective, pre/post-intervention comparative study by Bantar et al investigated changes in resistance patterns after a hospital formulary at a public teaching hospital in Argentina replaced usage of ceftriazone and ceftazidime with piperacillin-tazobactam, with exception of ceftriaxone use for central nervous sytem infections. Six months after intervention, there was no statistically significant decrease in combined third generation cephalosporin resistance for E. coli. However, rates of third generation cephalosporin resistance in K. pneumoniae decreased significantly. In addition, rates of third generation cephalosporin resistance in P. mirabilis also decreased. Rates of ESBL producing-isolates were not reported. [4]
Question 2
A 1999 study by Ambrose et al evaluated adult patients with nosocomial pneumonia who were treated empirically with either ceftazidime or cefepime. At the end of therapy, 80% of patients treated with cefepime had a therapeutic response in comparison to 68% of those treated with ceftazidime, but this difference was not statistically significant. Konstantinou et al [6] performed the same comparison in adult patients with pneumonia and found that the treatment groups had similar clinical outcomes but cefepime was associated with shorter post-therapy hospitalization, lower incidence of vancomycin co-administration and longer time to vancomycin initiation.
There are very few studies evaluating empiric use of cefepime in children. One such study on a pediatric population in Taiwan focused on the question of cefepime versus ceftazidime monotherapy in pediatric cancer patients with fever and neutropenia. [7] This study found that after 72 hours of treatment, 82% of patients in the cefepime group had not received additional antibiotics compared to 87% in the ceftazidime group. The duration of therapy with unmodified cefepime versus ceftazidime was comparable in both groups. Three (6.4%) patients in the cefepime group and 2 (4.3%) patients in the ceftazidime group died.Their study showed no statistical difference between cefepime and ceftazidime, therefore their conclusion was that they were equivalent in efficacy and safety for empiric treatment of the febrile neutropenic pediatric cancer patient. A similar study was performed at University of Texas Southwestern Medical Center at Dallas [8]. 74% (26 of 35) of cefepime-treated patients and 70% (23 of 33) of ceftazidime-treated patients responded to treatment. Patients in the cefepime group developed fewer new infections than those in the ceftazidime group (9% vs. 21%, respectively). They also showed that the use of concomitant systemic antimicrobial therapy (mostly vancomycin) occurred less often in the cefepime-treated patients, as compared with the ceftazidime group [35% [17 of 49] vs. 44% (24 of 55), respectively]. They reported no deaths or serious adverse events related to study therapy.
The cost of treating patients empirically with cefepime versus ceftazidime is an additional criterion for comparing these treatment options. The Ambrose et al 1999 study performed in a critical care setting found that cefepime was associated with significantly lower costs, which included cost of study antibiotic, and concomitant antibiotic therapy and drugs used to treat adverse event and clinical failures.Other studies in adults also reiterate this conclusion. [9, 10]. No similar studies have been performed in pediatric populations.
Question 1: The following search strategy was used in the MeSH Database: "Cephalosporins"[Majr] AND "beta-Lactam Resistance"[Mesh]AND "Formularies, Hospital"[Mesh]. Two articles resulted from this search, and related links were explored. Studies describing outcomes of extended spectrum cephalosporin restriction and E. coli or K. pnuemoniae resistance were then selected for review. References from these studies were searched, and additional studies not found by the above search strategy were then reviewed for inclusion. Studies pertaining to ESBL outbreaks were excluded, as these may have reflected proliferation of specific ESBL types, as well as other infection-control factors that could confound results and limit ability to generalize findings.
Question 2: Pubmed, on-line journals and the reference section of previously identified papers were used to obtain articles. MeSH terms used were cefepime, ceftazidime, ESBL, ß lactamase inhibitors, mortality and outcomes.
Note: The Ambrose et al 1999 study quoted from Question 2 was found from a journal reference, and obtained through the on-line journal website for Infectious Diseases in Clinical Practice. The citation is as follows: Ambrose PG, Richerson MA, Stanton ME, Cost-effectiveness analysis of cefepime compared with ceftazidime in intensive care unit patients with hospital-acquired pneumonia. Infect Dis Clin Pract 1999;8:245-51.
1) Lipworth AD, Hyle EP, Fishman NO, Nachamkin I, Bilker WB, Marr AM, Larosa LA, Kasbekar N, Lautenbach E. Limiting the emergence of extended-spectrum Beta-lactamase-producing enterobacteriaceae: influence of patient population characteristics on the response to antimicrobial formulary interventions. Infect Control Hosp Epidemiol. 2006 Mar;27(3):279-86. [Penn Proxy]
2) Empey KM, Rapp RP, Evans ME. The effect of an antimicrobial formulary change on hospital resistance patterns. Pharmacotherapy. 2002 Jan;22(1):81-7. [Penn Proxy]
3) Du B, Chen D, Liu D, Long Y, Shi Y, Wang H, Rui X, Cui N. Restriction of third-generation cephalosporin use decreases infection-related mortality. Crit Care Med. 2003 Apr;31(4):1088-93. [Penn Proxy]
4) Bantar C, Vesco E, Heft C, Salamone F, Krayeski M, Gomez H, Coassolo MA, Fiorillo A, Franco D, Arango C, Duret F, Oliva ME. Replacement of broad-spectrum cephalosporins by piperacillin-tazobactam: impact on sustained high rates of bacterial resistance. Antimicrob Agents Chemother. 2004 Feb;48(2):392-5. [Penn Proxy]
5) Kang CI, Kim SH, Park WB, Lee KD, Kim HB, Kim EC, Oh MD, Choe KW. Bloodstream infections due to extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae: risk factors for mortality and treatment outcome, with special emphasis on antimicrobial therapy. Antimicrob Agents Chemother. 2004 Dec;48(12):4574-81. [Penn Proxy]
6) Konstantinou K, Baddam K, Lanka A, Reddy K, Zervos M. Cefepime versus ceftazidime for treatment of pneumonia. J Int Med Res. 2004 Jan-Feb;32(1):84-93. [Penn Proxy]
7) Chuang YY, Hung IJ, Yang CP, Jaing TH, Lin TY, Huang YC. Cefepime versus ceftazidime as empiric monotherapy for fever and neutropenia in children with cancer. Pediatr Infect Dis J. 2002 Mar;21(3):203-9. [Penn Proxy]
8) Mustafa MM, Carlson L, Tkaczewski I, McCracken GH Jr, Buchanan GR. Comparative study of cefepime versus ceftazidime in the empiric treatment of pediatric cancer patients with fever and neutropenia. Pediatr Infect Dis J. 2001 Mar;20(3):362-9. [Penn Proxy]
9) Paladino JA. Cost-effectiveness comparison of cefepime and ceftazidime using decision analysis. Pharmacoeconomics. 1994 Jun;5(6):505-12. [Penn Proxy]
10) Bonfitto P, Lamorgese V, De Vietro T, Malerba M, Rizzello L, Scoditti S, Zuin R. A randomized trial of cefepime and ceftazidime for the treatment of community-acquired pneumonia. J Chemother. 1999 Aug;11(4):273-7. [Penn Proxy]