In vitro activity of ceftazidime–avibactam and comparators against Gram-negative bacterial isolates collected in the Asia–Pacific region as part of the INFORM program (2015–2017)

Background Antimicrobial resistance among nosocomial Gram-negative pathogens is a cause for concern in the Asia–Pacific region. The aims of this study were to measure the rates of resistance among clinical isolates collected in Asia–Pacific countries, and to determine the in vitro antimicrobial activities of ceftazidime–avibactam and comparators against these isolates. Methods CLSI broth microdilution methodology was used to determine antimicrobial activity and EUCAST breakpoints version 9.0 were used to determine rates of susceptibility and resistance. Isolates were also screened for the genes encoding extended-spectrum β-lactamases (ESBLs) or carbapenemases (including metallo-β-lactamases [MBLs]). Results Between 2015 and 2017, this study collected a total of 7051 Enterobacterales isolates and 2032 Pseudomonas aeruginosa isolates from hospitalized patients in Australia, Japan, South Korea, Malaysia, the Philippines, Taiwan, and Thailand. In the Asia–Pacific region, Enterobacterales isolates that were ESBL-positive, carbapenemase-negative (17.9%) were more frequently identified than isolates that were carbapenemase-positive, MBL-negative (0.7%) or carbapenemase-positive, MBL-positive (1.7%). Multidrug-resistant (MDR) isolates of P. aeruginosa were more commonly identified (23.4%) than isolates that were ESBL-positive, carbapenemase-negative (0.4%), or carbapenemase-positive, MBL-negative (0.3%), or carbapenemase-positive, MBL-positive (3.7%). More than 90% of all Enterobacterales isolates, including the ESBL-positive, carbapenemase-negative subset and the carbapenemase-positive, MBL-negative subset, were susceptible to amikacin and ceftazidime–avibactam. Among the carbapenemase-positive, MBL-positive subset of Enterobacterales, susceptibility to the majority of agents was reduced, with the exception of colistin (93.4%). Tigecycline was active against all resistant subsets of the Enterobacterales (MIC90, 1–4 mg/L) and among Escherichia coli isolates, > 90% from each resistant subset were susceptible to tigecycline. More than 99% of all P. aeruginosa isolates, including MDR isolates and the carbapenemase-positive, MBL-positive subset, were susceptible to colistin. Conclusions In this study, amikacin, ceftazidime–avibactam, colistin and tigecycline appear to be potential treatment options for infections caused by Gram-negative pathogens in the Asia–Pacific region.


Background
Antimicrobial resistance is recognized to be a healthcare issue in the Asian continent [1], where there are high levels of resistant Gram-negative organisms, such as extended-spectrum β-lactamase (ESBL)-positive Enterobacterales and multidrug-resistant (MDR) Pseudomonas aeruginosa [2]. Of particular concern is the presence of metallo-β-lactamase (MBL)-positive Enterobacterales and newly-emerging resistance to the polymyxins (colistin or polymyxin B) [3][4][5]. Recent publications have shown that further afield in the Asia-Pacific region, Australia has relatively low levels of resistant Gram-negative organisms, although carbapenem-resistant Enterobacterales and P. aeruginosa are emerging as health risks to the Australian public [6,7].
Ceftazidime-avibactam, a combination of a thirdgeneration cephalosporin and a non-β-lactam β-lactamase inhibitor, has a spectrum of activity that covers MDR Enterobacterales and P. aeruginosa and is able to inhibit ESBLs and Klebsiella pneumoniae carbapenemases (KPCs) [8]. In Australia, Thailand and Taiwan, as well as in the United States and in Europe, ceftazidime-avibactam has been approved for the treatment of complicated intra-abdominal infection (in combination with metronidazole), complicated urinary tract infection (including pyelonephritis), and hospital-acquired and ventilatorassociated bacterial pneumonia [9,10]. In Europe, ceftazidime-avibactam is also indicated for the treatment of infections due to aerobic Gram-negative organisms in adult patients with limited treatment options [10].
The aims of this study were to determine the rates of resistance among clinical isolates of Enterobacterales and P. aeruginosa collected in Asia-Pacific countries between 2015 and 2017, and to show the in vitro antimicrobial activities of ceftazidime-avibactam and comparators against these isolates. The countries included are Australia, Japan, South Korea, Malaysia, the Philippines, Taiwan, and Thailand. Isolates collected in 2015 are also included in an INFORM publication on the Asia-Pacific region between 2012 and 2015 [11], and isolates collected in 2015 and 2016 are also included in INFORM global publications [12,13].

Method
Each participating study center was required to collect a predefined number of Enterobacterales and P. aeruginosa isolates every year from adult and pediatric patients with specific bacterial infections. Only patient-derived, nonduplicate isolates suspected as the cause of each infection were included in the study. Isolates from external sources, such as drainage bottles or environmental samples, were excluded from the study. Demographic information was recorded for each isolate, including source of isolate and patient information, for example, age, sex and referring ward.
Isolates were identified at the local center of collection and shipped to a central laboratory (International Health Management Associates [IHMA] Inc., Schaumburg, IL, USA) for further testing. The central laboratory confirmed species identity and determined antimicrobial minimum inhibitory concentrations (MICs) using the Clinical Laboratory Standards Institute (CLSI) broth microdilution method [14]. MICs were interpreted using version 9.0 of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) clinical breakpoint tables [15]. Enterobacterales and P. aeruginosa isolates were screened for the presence of the genes encoding ESBLs or carbapenemases (including MBLs), if they met the MIC criteria described below.
Original-spectrum β-lactamases (TEM-type β-lactamases without substitutions at amino acid positions 104, 164 or 238, or SHV-type β-lactamases without substitutions at amino acid positions 146, 238 or 240) were not included. All detected genes were amplified using flanking primers and sequenced. Sequences were compared with publically available databases.

Resistance in the Asia-Pacific region and by country among E. coli and K. pneumoniae
For all E. coli and K. pneumoniae isolates collected in the Asia-Pacific region, the rate of ESBL-positive, carbapenemase-negative E. coli was greater than for K. pneumoniae (30.1% and 25.6%, respectively; Table 1). Rates of carbapenemase-positive, MBL-negative and carbapenemase-positive, MBL-positive K. pneumoniae in the Asia-Pacific region (1.9% and 2.5%, respectively) were greater than for E. coli (0.2% and 0.6%, respectively). By country, the highest rate of ESBL-positive, carbapenemase-negative E. coli was collected in Thailand (51.8%) and the highest rates of ESBL-positive, carbapenemasenegative K. pneumoniae were collected in South Korea, Malaysia, the Philippines and Thailand (33.5-38.8%). The rates of ESBL-positive, carbapenemase-negative E. coli and K. pneumoniae were both lowest in Australia (12.2% and 7.6%, respectively). Country rates of carbapenemasepositive, MBL-negative and carbapenemase-positive, MBL-positive E. coli were ≤ 0.3%, with the exception of the Philippines (carbapenemase-positive, MBL-positive E. coli, 0.9%) and Thailand (0.8% and 2.2%, respectively). The rates of carbapenemase-positive, MBL-negative and carbapenemase-positive, MBL-positive K. pneumoniae were highest in the Philippines (5.2% and 4.3%, respectively) and Thailand (2.5% and 7.3%, respectively). Figure 2 shows the rates of resistant subsets among P. aeruginosa isolates from the Asia-Pacific region and by country. Among all P. aeruginosa isolates collected in the Asia-Pacific region, the rate of carbapenemase-positive, MBL-positive isolates (3.7%) was greater than for ESBL-positive, carbapenemase-negative or carbapenemase-positive, MBL-negative isolates (0.4% and 0.3%, respectively). No ESBL-positive, carbapenemase-negative P. aeruginosa isolates were collected in Australia, Japan, South Korea, the Philippines, or Taiwan and no carbapenemase-positive, MBL-negative P. aeruginosa isolates were collected in Australia, Japan, Malaysia, or the Philippines. The rates of carbapenemase-positive, MBLpositive P. aeruginosa were highest in the Philippines (9.3%) and lowest in Taiwan and Australia (0.0% and 0.3%, respectively). The rate of MDR phenotypes among all P. aeruginosa isolates in the Asia-Pacific region was 23.4% and the country rates were: Australia, 14.9%; Malaysia, 18.3%; Taiwan, 19.1%; Japan, 21.1%; the Philippines, 24.1%; South Korea, 30.9%; and Thailand, 31.0%.

In vitro antimicrobial activity of 12 agents in the Asia-Pacific region against all Enterobacterales
The in vitro antimicrobial activities of a panel of agents against the pooled collection of Enterobacterales are presented in Table 2. Among all Enterobacterales isolates, the highest susceptibility rates of 96.8% to 98.1% were observed for amikacin, ceftazidime-avibactam, and meropenem. The activity of ceftazidime alone (MIC 90 , 64 mg/L) was lower than that of ceftazidime-avibactam (MIC 90 , 0.5 mg/L). Susceptibility to colistin was 83.9% and a colistin MIC 90 value of ≥ 16 mg/L was observed against the Enterobacterales overall; however, colistin was active against E. coli and K. pneumoniae isolates (MIC 90 , 0.5 and 1 mg/L, respectively). Tigecycline was active against the collection of Enterobacterales (MIC 90 , 1 mg/L).

In vitro antimicrobial activity of 12 agents in the Asia-Pacific region against E. coli and K. pneumoniae
More than 95% of all E. coli and K. pneumoniae isolates were susceptible to amikacin, ceftazidime-avibactam, colistin, imipenem and meropenem (   Table 2). Data for the carbapenemase-positive, MBLnegative subset of E. coli are not presented in Table 2 because of the small isolate numbers. Against this subset, the lowest MIC ranges were observed for colistin and tigecycline (both agents, 0.12-0.5 mg/L), followed by ceftazidime-avibactam (0.25-2 mg/L), whereas amoxicillin-clavulanic acid, aztreonam, ceftazidime, and piperacillin-tazobactam were not active. For the carbapenemase-positive, MBL-positive subset of E. coli and the carbapenemase-positive, MBL-negative and carbapenemase-positive, MBL-positive subsets of K. pneumoniae, the susceptibility rates to the majority of agents were < 24%, with the exception of amikacin (67.3-94.9%) and colistin (87.2-100%; Table 2). Susceptibility to tigecycline among carbapenemase-positive, MBL-positive E. coli isolates was 92.9%. All carbapenemase-positive, MBL-negative K. pneumoniae isolates were susceptible to ceftazidime-avibactam; however, no carbapenemasepositive, MBL-positive E. coli or K. pneumoniae isolates were susceptible to ceftazidime-avibactam.

In vitro antimicrobial activity of 12 agents in the Asia-Pacific region against P. aeruginosa
The highest rate of susceptibility among P. aeruginosa isolates was to colistin (99.7%), followed by ceftazidime-avibactam and amikacin (92.7% and 92.3%, respectively; Table 3). The activity of ceftazidime alone (MIC 90 , 64 mg/L) was less potent than that of ceftazidime-avibactam (MIC 90 , 8 mg/L). Among the resistant subsets of P. aeruginosa, eight isolates were identified as ESBLpositive, carbapenemase-negative and seven as carbapenemase-positive, MBL-negative. These data are not presented in Table 3 because of the small isolate numbers; however, the lowest MIC range was observed for colistin (0.5-2 mg/L against both subsets). Susceptibility to the majority of agents was reduced among the carbapenemase-positive, MBL-positive subset, when compared with the susceptibility among all P. aeruginosa isolates, with the exception of colistin (100% susceptible; Table 3). Amoxicillin-clavulanic acid (MIC 90 , ≥ 64 mg/L) and tigecycline (MIC 90 , ≥ 16 mg/L) were inactive against all MDR, or carbapenemase-positive, MBL-positive P. aeruginosa isolates.

Discussion
This study reports the in vitro antimicrobial susceptibilities and the rates of resistant subsets for clinical isolates of Enterobacterales and P. aeruginosa from the Asia-Pacific region. For the pooled collection of Enterobacterales, susceptibility was highest to amikacin, ceftazidime-avibactam and meropenem, and among P. aeruginosa was highest to amikacin, ceftazidime-avibactam and colistin. Enterobacterales isolates that were ESBL-positive, carbapenemase-negative were more common than carbapenemase-positive, MBL-negative or carbapenemase-positive, MBL-positive isolates. MDR isolates of P. aeruginosa were more frequently observed than the other resistant subsets.
This study is an update to the 2012-2015 INFORM study, which reported similar susceptibility rates to amikacin, ceftazidime-avibactam, colistin and meropenem among a pooled collection of Enterobacteriaceae [11]. Any comparisons should be treated with caution, however, as the isolates collected in 2015 were included in both studies and the 2012-2015 INFORM study also included isolates from China and Hong Kong. Yet, the findings of these studies together would indicate a level of stability in antimicrobial susceptibility among the Enterobacterales in the Asia-Pacific region.
In the current study, susceptibility to colistin and its antimicrobial activity was lower for the pooled collection of Enterobacterales, when compared with E. coli and K. pneumoniae isolates individually. This reduced susceptibility to colistin among the pooled group of Enterobacterales is likely to be due to the presence of intrinsically colistin-resistant species, such as P. mirabilis and Serratia marcescens [19].
Among the resistant subsets presented in this study, > 98% of ESBL-positive, carbapenemase-negative Enterobacterales isolates remained susceptible to amikacin, ceftazidime-avibactam, colistin, imipenem and meropenem. The range of antimicrobial susceptibilities to these agents was similar to the 2012-2015 INFORM study for the isolates of Enterobacteriaceae molecularly-characterized as ESBL-positive (95.1-99.6%) [11]. In the present study, susceptibility to colistin, imipenem and meropenem was reduced among the carbapenemase-positive, MBL-negative subset of Enterobacterales, compared with the ESBL-positive, carbapenemase-negative subset; however, susceptibility to amikacin and ceftazidimeavibactam among this subset remained > 93%. Similar trends were observed among K. pneumoniae isolates, although there were relatively low isolate numbers in the resistant subsets for this species. Studies of antimicrobial resistance in South Korea and Taiwan, on isolates collected between 2012 and 2017, have reported the rates of amikacin resistance to be < 10% among E. coli and K. pneumoniae isolates [20][21][22], and among carbapenemnonsusceptible E. coli isolates (Taiwan only) [23].
In the current study, the group of carbapenemasepositive, MBL-positive Enterobacterales remained susceptible to colistin (93.4%) and similar results were seen for these resistant subsets of K. pneumoniae (94.2%) and E. coli (100%) isolates. Furthermore, nearly 100% of P. aeruginosa isolates, including all MDR or carbapenemase-positive, MBL-positive isolates, were susceptible to colistin. Colistin has been described in the literature as a 'last-resort' antimicrobial agent due to its clinical efficacy against antimicrobial-resistant isolates of Enterobacterales and P. aeruginosa [19]. However, resistance to colistin, along with the carbapenems, has been emerging in Southeast Asia, leading to increased hospital mortality rates in the case of the carbapenems [5]. A clinical trial involving patients with infections predominantly caused by carbapenem-resistant K. pneumoniae found that ceftazidime-avibactam was superior to colistin and may be an alternative treatment option [24]. Avibactam is inactive against MBL-positive isolates, therefore it is presumed that carbapenem resistance among K. pneumoniae from these clinical studies was mediated by the serine carbapenemases, such as KPC and OXA-48 [8,25], which are susceptible to ceftazidime-avibactam. For P. aeruginosa isolates in the current study, the MDR subset accounted for 23.4%. Lower rates of MDR P. aeruginosa were recently reported in the literature for the Asia-Pacific region (15.0% [2013-2016] and 14.8% [2012][2013][2014][2015]) [11,26]. Each study used the MDR definition proposed by Magiorakos et al. [18]; however, variations in the participating countries and time periods for each study may account for the differences in the overall regional rates.
One of the limitations of this antimicrobial surveillance study is that a predefined number of isolates were collected at each center, so the findings cannot describe the epidemiology of resistance. Additionally, the lower numbers of isolates collected from Japan and Malaysia than elsewhere may have been insufficient to allow the detection of resistant subsets, resulting in relatively low resistant rates being reported for these countries. Furthermore, this study did not include data from China and India where antimicrobial resistance is prevalent, exemplified by the discovery of the first colistin-resistant Enterobacteriaceae isolate [27] and the first bla NDM-1 gene in K. pneumoniae [28] in these two countries, respectively. Nevertheless, reporting the local antimicrobial activities of contemporary agents in this region, through programs such as INFORM, is a key to guide physicians toward appropriate use of antimicrobials.
Among all P. aeruginosa isolates, including the resistant subsets, the rate of colistin susceptibility (> 99%) was higher than rates for ceftazidime-avibactam and amikacin, particularly among carbapenemase-positive, MBL-positive isolates (3.9% and 10.5%, respectively). Data from this study have shown the variability in antimicrobial activity depending on the β-lactamase profile of isolates, and multicenter surveillance of antimicrobial resistance remains essential for public health and clinical management of bacterial infections.