|Home | About | Journals | Submit | Contact Us | Français|
From 2002 to 2008, there was a significant increase in extended-spectrum beta-lactamase (ESBL)-positive Escherichia coli isolates in European intra-abdominal infections, from 4.3% in 2002 to 11.8% in 2008 (P < 0.001), but not for ESBL-positive Klebsiella pneumoniae isolates (16.4% to 17.9% [P > 0.05]). Hospital-associated isolates were more common than community-associated isolates, at 14.0% versus 6.5%, respectively, for E. coli (P < 0.001) and 20.9% versus 5.3%, respectively, for K. pneumoniae (P < 0.01). Carbapenems were consistently the most active drugs tested.
Gram-negative bacilli (GNB) continue to play an important role as a cause of hospital-associated (HA) infections (8, 19) and have the leading role in intra-abdominal infections (IAIs) (3, 20). Furthermore, reports from the Asia/Pacific region show strikingly high levels of extended-spectrum beta-lactamase (ESBL)-producing isolates of Escherichia coli and Klebsiella pneumoniae and diminished susceptibility to many antimicrobial agents (11, 16). Antimicrobial resistance in European GNB has been reported (4, 7, 10, 17, 25, 27), but none of these studies focused specifically on IAI pathogens. The Study for Monitoring Antimicrobial Resistance Trends (SMART) monitors the susceptibilities of aerobic GNB from IAIs worldwide. This report describes the frequency of occurrence and antimicrobial susceptibilities of ESBL-producing E. coli and K. pneumoniae isolates in community-associated (CA) and HA IAIs in Europe in 2008 and compares these data with those published previously (1, 11, 22).
Isolates were derived from intra-abdominal body sites (e.g., appendix, peritoneum, colon, bile, pelvis, and pancreas) and were mostly obtained during surgery, though some paracentesis specimens were also accepted. Isolates were considered to be community associated if they were recovered from a specimen taken less than 48 h after the patient was admitted to the hospital and hospital associated if the specimen was taken 48 or more hours after hospital admission, as described previously (11). Isolates were from 37 hospitals in 12 countries (1 in Estonia, 4 in France, 5 in Germany, 2 in Greece, 4 in Italy, 1 in Latvia, 1 in Lithuania, 3 in Portugal, 9 in Spain, 1 in Switzerland, 3 in Turkey, and 3 in the United Kingdom). Isolates were sent to the central laboratory of International Health Management Associates, Inc. (Schaumburg, IL) for confirmation of identification and antimicrobial susceptibility testing. MICs were determined using MicroScan dehydrated broth microdilution panels (Siemens Medical Solutions Diagnostics, West Sacramento, CA), following Clinical and Laboratory Standards Institute (CLSI) and the panel manufacturer's guidelines (5). The MIC interpretive criteria followed guidelines established by the European Committee for Antimicrobial Susceptibility Testing (EUCAST) (http://www.eucast.org ). Using 2008 CLSI guidelines, E. coli and K. pneumoniae isolates were classified as ESBL producers if there was at least an 8-fold reduction (i.e., three doubling dilutions) of the MIC for ceftazidime or cefotaxime tested in combination with clavulanic acid versus the MIC when tested alone (5). Quality control (QC) testing was performed each day of testing using the CLSI QC strains E. coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, and K. pneumoniae ATCC 700603 (positive ESBL control) (6). Fisher's exact test (two tailed) was used to determine significance, and the modified Wald method was used to calculate 95% confidence intervals.
The percentages of isolates from HA or CA infections were 57.9% and 38.5% for E. coli (P < 0.001) and 70.8% and 23.9% for K. pneumoniae (P < 0.001), respectively. Another 3.6% of E. coli isolates and 5.3% of K. pneumoniae isolates were from patients without information regarding length of stay at the time of specimen collection. After declining slightly in 2004, 2005, and 2006, the rates of ESBL-positive E. coli isolates rose during the last 2 years of the study, reaching 12.3% in 2007 and 14.0% in 2008 (Table (Table1).1). The rates of ESBL-positive K. pneumoniae isolates in HA infections were more variable before rising to 20.9% in 2008. ESBL-positive E. coli isolate rates among CA IAIs remained relatively stable, from 4.0% in 2002 to 7.4% in 2007 and 6.5% in 2008. The rate of ESBL-positive K. pneumoniae isolates in CA infections was 5.3% in 2008, lower than the rates observed in 2005, 2006, and 2007 and comparable to that in 2003 (4.4%). However, the generally smaller numbers of ESBL-positive K. pneumoniae isolates led to correspondingly wider 95% confidence intervals. The rates of ESBL-positive E. coli isolates in HA infections ranged from 0% (Lithuania and Switzerland) to 30% (Greece), while those for CA infections ranged from 0% (Latvia and Lithuania) to 24% (Italy) (Table (Table2).2). The rate of CA ESBL-positive E. coli isolates in most countries was lower than the HA rate. ESBL-positive E. coli isolates were detected in all countries with the exceptions of Lithuania and Switzerland. The rates of ESBL-positive K. pneumoniae isolates in HA IAIs also ranged widely,with the highest rate of 56.0% observed in Italy, followed closely by 50% in both Latvia and Lithuania. The rates of ESBL-positive K. pneumoniae isolates in CA IAIs were 0% in all cases except for Germany and Spain, where the rates were 18.2% and 5.4%, respectively. It is worthwhile noting that the numbers of E. coli and K. pneumoniae isolates in several countries were small, leading to very broad confidence intervals. Furthermore, the number of investigator sites varied from country to country, with several countries having one site while Spain had nine.
Ertapenem and imipenem were the most active agents against ESBL-positive and -negative isolates from both HA and CA origins (Table (Table3).3). The susceptibilities of HA and CA ESBL-positive isolates were similar. However, for ESBL-negative E. coli isolates, several drugs had significantly lower (P < 0.05) levels of activity against HA than CA isolates. This result potentially denotes selection processes of patients' endogenous gut microorganisms during hospitalization or the acquisition of more resistant isolates despite the absence of ESBL production.
The low number of ESBL-positive K. pneumoniae isolates collected from HA and CA infections in 2008 (47 and 4, respectively) makes comparison of susceptibility data difficult, and in fact, differences as high as 20.2% failed to achieve statistical significance (Table (Table3).3). Comparisons of the activities of the study agents against HA and CA ESBL-negative K. pneumoniae isolates revealed significant differences (P < 0.05) for all agents, with the exceptions of imipenem (P = 0.063) and amikacin (P = 0.0738), which were within the group of most active agents (Table (Table33).
Various reports have described increases in the incidence of ESBL-positive E. coli and, to a lesser extent, ESBL-positive K. pneumoniae in Europe, particularly in the community compared to the incidence in health care facilities (4, 7, 9, 18). Many of these reports were single-year “snapshots” or presented multiyear analyses ending in 2006 or earlier and comprised isolates from different types of infections. In contrast, SMART has been on-going since 2002, collecting data exclusively on hospital-associated and community-associated IAIs.
Some investigators have reported dramatic increases in the incidence of E. coli isolates possessing CTX-M ESBLs in the community (4, 7, 18, 21). In addition, Ben-Ami et al. (2) and Pitout et al. (21) reported that isolates producing CTX-M beta-lactamases were significantly more resistant to fluoroquinolones than were isolates producing other types of ESBLs. Although the specific enzyme types of ESBL producers were not identified as part of the SMART program in 2008, a high percentage of fluoroquinolone resistance was observed, denoting the potential presence of these enzymes.
In the study reported by Ben-Ami et al. (2), the investigators evaluated risk factors for nonhospitalized patients to be infected with an ESBL-positive isolate. These included recent use of antimicrobials, residence in a long-term care facility, recent hospitalization, age of ≥65 years, and male gender. However, 34% of the ESBL isolates in their analysis came from patients with no recent health care contact. Nevertheless, increases of fecal carriage of ESBL-producing isolates in both hospitalized and nonhospitalized patients have been reported (24), which may increase the risk for IAI due to these organisms. It is essential for physicians to be aware of the extent of increasing ESBL rates in both species and of the subsequent reduced susceptibilities that these species exhibit to most agents. Based on the results of the present study, the number of options for empirical therapy of IAI in various parts of Europe exhibiting high ESBL-positive rates may therefore be decreasing, particularly for cephalosporins, fluoroquinolones, and ampicillin-sulbactam. However, a limitation of the current study is that the number of investigator sites in some countries was not balanced to correlate with population differences; therefore, more extensive investigations in some of the countries studied are warranted in order to solidify general conclusions and trends.
Although the carbapenems were consistently the most active agents in this study, the emergence of carbapenemases in isolates with or without the coproduction of an ESBL (7, 13, 15, 23, 24, 26) reinforces the need for continuous local and global monitoring of carbapenem susceptibility.
We thank Joseph Chow (Merck & Co., Inc.) for his review of the manuscript. We thank all the SMART investigators in Europe for their participation in this program.
The authors are responsible for the work described in this paper. All authors were involved in at least one of the following: conception, design, acquisition, analysis, statistical analysis, interpretation of data, and drafting of the manuscript and/or revision of the manuscript for important intellectual content. All authors provided final approval of the version to be published. The authors did not receive honoraria for writing this article.
The SMART surveillance program is funded by Merck & Co., Inc.
Stephen Hawser, Samuel Bouchillon, Daryl Hoban, and Robert Badal served as scientific advisors or consultants to Merck, and received research support from Merck to conduct this study. Rafael Cantón received honoraria for lectures from MSD and AstraZeneca. Fernando Baquero and Rafael Cantón have served as scientific advisors on the SMART steering committee.
Published ahead of print on 26 April 2010.