Emergence of High Levels of Extended-Spectrum-β-Lactamase-Producing Gram-Negative Bacilli in the Asia-Pacific Region: Data from the Study for Monitoring Antimicrobial Resistance Trends (SMART) Program, 2007

MATERIALS AND METHODS

All isolates were derived from IAI. Only one isolate per patient was accepted into the study. Gram-negative aerobic and facultative bacteria were cultured from specimens from intra-abdominal body sites, e.g., appendix, peritoneum, colon, bile, pelvis, and pancreas. The majority of intra-abdominal specimens were obtained during surgery, though some paracentesis specimens were also accepted. Isolates from blood, urine, and perirectal abscesses were excluded. Overall, 3,004 clinical isolates from 35 laboratories in the following 11 locations in the Asia-Pacific region (numbers of laboratories are in parentheses) were collected and tested in 2007: Australia (1), China (6), Hong Kong (1), India (9), South Korea (2), New Zealand (2), Philippines (1), Singapore (2), Taiwan (7), Thailand (2), and Vietnam (2). Isolates were identified to the species level and tested at each site. Development of a centralized database of study results was managed by International Health Management Associates, Inc., located in Schaumburg, IL. All organisms were deemed clinically significant by local participant criteria. Isolate inclusion was independent of antimicrobial use, age, or gender. All sites identified each study isolate utilizing local laboratory methods.

MICs were determined using dehydrated MicroScan broth microdilution panels manufactured by Siemens Medical Solutions Diagnostics (West Sacramento, CA) by following Clinical and Laboratory Standards Institute (CLSI) guidelines (7). All antimicrobial agents were supplied by the panel manufacturer. The following antimicrobial agents, with their dilution ranges (expressed in μg/ml), were included on the panels: ertapenem, 0.03 to 4; imipenem, 0.06 to 8; cefepime, 0.5 to 32; ceftazidime, 0.5 to 128; ceftazidime-clavulanic acid, 0.12 to 16; cefoxitin, 2 to 16; ciprofloxacin, 0.25 to 2; amikacin, 4 to 32; levofloxacin, 0.5 to 4; cefotaxime, 0.5 to 128; cefotaxime-clavulanic acid, 0.12 to 16; piperacillin-tazobactam, 2 to 64 (piperacillin) and 4 (tazobactam); ampicillin-sulbactam, 2 to 16 (ampicillin) and 2 (sulbactam); and ceftriaxone, 1 to 32. MIC interpretive criteria followed published guidelines established by the CLSI (6).

E. coli, K. pneumoniae, and Klebsiella oxytoca strains were classified as ESBL producers if there was at least an eightfold reduction (i.e., three doubling dilutions) of the MIC for ceftazidime or cefotaxime tested in combination with clavulanic acid versus their MICs when either drug was tested alone (7). Quality control (QC) testing was done by each testing site on each day of testing using the CLSI-recommended QC strains E. coli ATCC 25922, K. pneumoniae ATCC 700603 (positive ESBL control), and Pseudomonas aeruginosa ATCC 27853. All participating labs used the same MIC panels manufactured by Siemens Medical Solutions Diagnostics and performed susceptibility testing and QC testing by the CLSI protocol as described in the MicroScan package insert. QC testing was performed each time a laboratory carried out batch testing of isolates. Only data from test days for which all pertinent CLSI-recommended QC results were in control were included in the database. Results were included in the analysis only when corresponding QC isolates tested within the acceptable range according to CLSI guidelines (6).

RESULTS

Of the 3,004 GNB isolates collected, 82% were represented by seven species: E. coli (n = 1,368), K. pneumoniae (n = 528), P. aeruginosa (n = 245), Enterobacter cloacae (n = 150), Enterobacter aerogenes (n = 65), Proteus mirabilis (n = 61), and Citrobacter freundii (n = 51). Fifty-eight other species comprised the other 18% of the isolates. The most commonly isolated organism was E. coli, for which 578 of the 1,368 isolates (42.2%) were ESBL positive. Also ESBL positive were 35.8% of the K. pneumoniae isolates. Furthermore, although the number of K. oxytoca isolates was relatively small (n = 51), 22 (43.1%) were ESBL positive.

Table 1 and Fig. 1 show the distribution of ESBL isolates by location. While ESBL rates for the Asia-Pacific region as a whole were relatively high, several countries exhibited ESBL rates that were strikingly higher than 50%. For example, rates for India for ESBL-positive E. coli, K. pneumoniae, and K. oxytoca were 79.0%, 69.4%, and 100%, respectively. Furthermore, of the nine sites in India, none had an ESBL-positive rate for E. coli of less than 54.5% and the highest rate was 94.1% (data not shown). ESBL-positive E. coli rates were also high in China (55.0%) and Thailand (50.8%), making ESBL-positive isolates more common than ESBL-negative ones in all three countries. Furthermore, Thailand also exhibited relatively high rates of ESBL-positive K. pneumoniae isolates, whereby nearly one-half (45.5%) of isolates were ESBL positive (Fig. 1). The lowest rates of ESBL-positive isolates appeared in Australia and New Zealand; however, the numbers of K. pneumoniae and K. oxytoca isolates in these two countries were relatively small.

• Open in new tab
• Download powerpoint

FIG. 1.

Frequency distribution of ESBL-positive isolates by country in the Asia-Pacific (AP) region during 2007. Korea refers to South Korea.

Table 2 describes the antimicrobial susceptibilities of the eight most commonly isolated organisms (those with n > 50), including ESBL-positive isolates of E. coli, K. pneumoniae, and K. oxytoca. The carbapenems, ertapenem and imipenem, were consistently the most active agents against all strains, with the exception of P. aeruginosa strains, for which 71% of isolates were susceptible to imipenem. The most active antipseudomonal antibiotic was amikacin, to which 85% of isolates were susceptible, followed by piperacillin-tazobactam (81% susceptible). The most active of the noncarbapenem antibiotics was generally amikacin: susceptibility of all organisms ranged from 73 to 100%. For all other agents, susceptibilities were typically considerably lower. Ampicillin-sulbactam exhibited notably poor in vitro activity against all species in this study (Table 1).

The susceptibilities to ertapenem and imipenem of non-ESBL-producing isolates ranged from 94 to 100% and 97 to 100%, respectively. ESBL-negative E. coli isolates exhibited 100% susceptibility to ertapenem and imipenem, and these drugs maintained most of their activity against ESBL-positive E. coli, with susceptibilities of 96% and 98%, respectively (Table 2). Furthermore, against ESBL-negative K. pneumoniae, ertapenem and imipenem exhibited susceptibilities of 99 and 100%, respectively. It is noteworthy that ertapenem exhibited marginally weaker activity against ESBL-positive K. pneumoniae, with 86% of isolates being susceptible. With imipenem, the susceptibility of ESBL-positive K. pneumoniae was 3% lower than that of ESBL-negative isolates, representing a smaller drop in activity than that observed with ertapenem.

Table 3 summarizes rates of ESBL-positive E. coli and K. pneumoniae isolates from community-acquired and hospital-acquired infections in the Asia-Pacific region overall, China, India, and the Asia-Pacific region excluding those two countries. In the Asia-Pacific region excluding China and India, ESBL-positive E. coli rates for community-acquired and hospital-acquired infections were 13.7% and 31.5%, respectively. China's rates were roughly double those, at 36% and 64.7%, respectively, but showed approximately the same overall ratio of hospital-acquired infections to community-acquired infections: roughly twice as many hospital-acquired infections as community-acquired infections. In India, however, the community-acquired-infection rate of 79.0% was virtually identical to that of hospital-acquired infections (78.9%). ESBL-positive K. pneumoniae rates for hospital-acquired infections were three to four times higher than rates for community-acquired infections, the exception again being India, with rates of 72.7% and 61.8%, respectively.

DISCUSSION

Various country-based reports describing the incidence and susceptibility of ESBL-positive isolates of GNB are available in the literature, and each one reports quite different ESBL rates. For example, a one-center study in Vietnam reported that, of 350 isolates from clinical specimens, 87.4% were GNB. Of these GNB isolates, 88.9% were Enterobacteriaceae, of which 14.7% were ESBL positive (13). A study of 493 isolates from a single center in South Korea of Enterobacter, Serratia marcescens, C. freundii, and Morganella morganii revealed rates of ESBL-positive isolates of 12.8%, 12.4%, 4.9%, and 0%, respectively (5). Ko et al., from South Korea, described 22.4% of K. pneumoniae isolates and 10.2% of E. coli isolates as ESBL producers (14). In a recent study from Taiwan, 28.4% of K. pneumoniae isolates collected from various body sites were ESBL positive (4). Furthermore, in a study from India the percentage of ESBL-positive isolates was similarly elevated, with 23.1% of isolates being ESBL positive (18). Of the isolates from India, 48.4% of isolates were E. coli and 51.6% were K. pneumoniae. Notably, all of the ESBL-producing isolates were consistently susceptible only to carbapenems.

Asia is almost certainly a part of the world in which ESBLs have emerged de novo, with early antimicrobial resistance studies showing elevated levels of ESBL phenotypes, particularly among Klebsiella isolates and particularly in China, South Korea, Japan, and India (10). In most countries, there are mixtures of CTX-M types, with VEB appearing significantly in Vietnam and Thailand and ESBL isolates from India being completely dominated by the presence of bla_CTX-M-15 alone, with no other CTX-M types reported (10). The data from the SMART 2007 Asia-Pacific study show that E. coli and Klebsiella sp. ESBL frequencies are on the whole elevated in Asia compared both to previous years (2) and other regions of the world (2). Overall, ESBL frequencies in the Asia-Pacific region from 2007 were 40% compared with 30%, 17%, 10%, and 8% for Latin America, the Middle East and Africa, the European Union, and North America, respectively (2). Moreover, ESBL rates have markedly increased during the last few years in the Asia-Pacific region, with ESBL rates rising steadily from 2003 (15%) to 40% in 2007 (2, 3). However, there are two major features of the data from SMART 2007 (3) that are highly noteworthy. The first is that this is the first study demonstrating that, on the basis of the ESBL frequencies observed in India, China, Thailand, and Vietnam, the ESBL-positive phenotype has now become the dominant phenotype, alarmingly so in India, with E. coli and K. pneumoniae ESBL rates of 79.0% and 69.4%, respectively. Furthermore, the ESBL-positive E. coli rate in China was also notably high, running at 54.7%. India's near equivalence of rates of ESBL-positive isolates in community- and hospital-acquired IAI of both E. coli and K. pneumoniae is particularly significant. Additionally, these high rates were observed in all nine study centers in India, each representing distinct cities in various regions of India: New Delhi, Lucknow, Indore, Mumbai, Hyderabad, Bangalore, Chennai, Tamil Nadu, and Kolkata. The very high rates of ESBL-positive strains are seen all over the country and are not restricted to any single city or region. ESBL-positive K. oxytoca rates were also high in several countries. However, as molecular characterization of the strains was not performed, we are not able to distinguish K1-hyperproducing K. oxytoca from ESBL-producing K. oxytoca.

With the combined populations of India and China being circa 2.5 billion and with relatively high fecal carriage rates, these countries represent major ESBL gene reservoirs. Second, resistance to the majority of the antimicrobials used in the SMART 2007 study was high, with the exception of the carbapenems ertapenem and imipenem. It is important to mention, therefore, that, despite the high and increasing ESBL rates within Asia, and in particular in India, China, Thailand, and Vietnam, most ESBL-producing isolates remain susceptible to this important antibiotic class. Nevertheless, rates of resistance to ertapenem and imipenem in the current SMART study, albeit relatively low, may be creeping somewhat higher. Since almost all diminished ertapenem activity occurred in K. pneumoniae, we suspect it was due to the increasing incidence of KPC-producing strains; however, since study strains prior to 2008 were not saved or sent to a central laboratory for further testing, this supposition cannot be confirmed. Hence, further in-depth studies such as SMART and others are highly warranted in order to keep pace with the evolving clinical importance of ESBL producers and their susceptibility to currently available agents.