Global Assessment of the Activity of Tigecycline against Multidrug-Resistant Gram-Negative Pathogens between 2004 and 2014 as Part of the Tigecycline Evaluation and Surveillance Trial

Multidrug resistance among bacterial pathogens is an ongoing global problem and renders antimicrobial agents ineffective at treating bacterial infections. In the health care setting, infections caused by multidrug-resistant (MDR) Gram-negative bacteria can cause increased mortality, longer hospital stays, and higher treatments costs. The aim of the Tigecycline Evaluation and Surveillance Trial (TEST) is to assess the in vitro antimicrobial activities of tigecycline and other contemporary agents against clinically relevant pathogens. This paper presents antimicrobial activity data from the TEST study between 2004 and 2014 and examines global rates of MDR Gram-negative isolates, including Acinetobacter baumannii, Pseudomonas aeruginosa, and members of the Enterobacteriaceae, during this time. Our results show that tigecycline retained in vitro activity against many MDR Gram-negative pathogens over the study period, while rates of MDR A. baumannii increased globally. Using these findings, we hope to highlight the current status of multidrug resistance in medical facilities worldwide.

IMPORTANCE Multidrug resistance among bacterial pathogens is an ongoing global problem and renders antimicrobial agents ineffective at treating bacterial infections. In the health care setting, infections caused by multidrug-resistant (MDR) Gramnegative bacteria can cause increased mortality, longer hospital stays, and higher treatments costs. The aim of the Tigecycline Evaluation and Surveillance Trial (TEST) is to assess the in vitro antimicrobial activities of tigecycline and other contemporary agents against clinically relevant pathogens. This paper presents antimicrobial activity data from the TEST study between 2004 and 2014 and examines global rates of MDR Gram-negative isolates, including Acinetobacter baumannii, Pseudomonas aeruginosa, and members of the Enterobacteriaceae, during this time. Our results show that tigecycline retained in vitro activity against many MDR Gram-negative pathogens over the study period, while rates of MDR A. baumannii increased globally. Using these findings, we hope to highlight the current status of multidrug resistance in medical facilities worldwide. KEYWORDS Gram-negative bacteria, multidrug resistance, surveillance studies, tigecycline M ultidrug resistance among Gram-negative organisms is a global problem, with rates of infections caused by multidrug-resistant (MDR) Gram-negative bacteria increasing worldwide (1)(2)(3). MDR Gram-negative pathogens, such as Acinetobacter baumannii, Pseudomonas aeruginosa, and the Enterobacteriaceae, are associated with increased lengths of hospitalization, higher health care costs, and greater rates of mortality (2,(4)(5)(6). These organisms have been highlighted as clinically important bacteria, and some are included among the ESKAPE pathogens (an acronym for Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, A. baumanii, Pseudomonas aeruginosa, and Enterobacter spp.) (7).
Tigecycline is a broad-spectrum glycylcycline antimicrobial agent with in vitro activity against both Gram-positive and Gram-negative organisms. Tigecycline has been approved in the United States and Europe for the treatment of complicated skin and intra-abdominal infections and also in the United States for community-acquired bacterial pneumonia (8,9). The in vitro activity of tigecycline is monitored globally, alongside comparator agents, against clinical Gram-positive and Gram-negative isolates as part of the Tigecycline Evaluation and Surveillance Trial (TEST). This study describes the activity of tigecycline against MDR Gram-negative isolates collected globally between 2004 and 2014. Isolates collected during the earlier years of the study period have been included in previous TEST publications, including reports focused on MDR A. baumannii (10) and MDR Enterobacteriaceae (11) isolates that were collected in the United States between 2004 and 2006 and on MDR Gram-negative isolates collected globally between 2004 and 2013 (12). ). This change was statistically significant (P Ͻ 0.01) ( Table 3). Klebsiella pneumoniae. Multidrug resistance was reported in 12% of 32,888 of K. pneumoniae isolates submitted globally (Table 1). During the study period, global rates of MDR K. pneumoniae increased from 4% in 2004 to 18% in 2010, and then the rate decreased to 8% in 2014 ( Fig. 1). By region, the lowest rates of MDR K. pneumoniae were found in North America (7%), with the highest rates in Latin America and the Middle East (19% and 20%, respectively) ( Table 1). High levels of global resistance were reported to ceftriaxone and levofloxacin (89% and 94%, respectively) ( Table 2). Global resistance to amoxicillin-clavulanic acid, cefepime, levofloxacin, and piperacillin-tazobactam increased significantly during the study period: amoxicillin-clavulanic acid, 55%  Table 3). The lowest rate of resistance among MDR K. pneumoniae was to tigecycline (6%).
Klebsiella oxytoca. Over the study period, 8,000 isolates of K. oxytoca were submitted globally, of which 2.5% were MDR ( Table 1). The global rates of MDR K. oxytoca increased from 0.4% in 2004 to 5% in the 2008-2009 period and then decreased to 0.6% in 2014 (Fig. 1). The highest percentages of MDR K. oxytoca isolates were reported in the Middle East (13%), while rates in all other regions were Յ7% (Table 1). More than 80% of MDR K. oxytoca isolates collected globally were resistant to minocycline, ceftriaxone, and levofloxacin (86% to 91%) ( Table 2). The lowest rate of resistance was to tigecycline (6%).

DISCUSSION
This study describes the global rates of multidrug resistance among a selection of clinically important Gram-negative organisms collected between 2004 and 2014, and it shows the in vitro antimicrobial activity of tigecycline and a panel of other contemporary antimicrobial agents against these resistant isolates.
Tigecycline retained in vitro activity against the majority of MDR organisms collected between 2004 and 2014, with the exception of P. aeruginosa (MIC 90 , Ն32 mg/liter), against which tigecycline is known to have limited activity (13). Furthermore, none of the antimicrobial agents tested in the present study demonstrated potent in vitro activity against MDR P. aeruginosa, which was highlighted among the ESKAPE organisms as a cause for global concern (7). Against MDR A. baumannii, another of the ESKAPE pathogens, tigecycline had the lowest MIC 90 (2 mg/liter) of the agents on the TEST panel. This MIC 90 was comparable with that reported from a global study of A. baumannii isolates collected between 2005 and 2009, which showed that tigecycline inhibited 95% of MDR Acinetobacter spp. isolates at Յ2 mg/liter (14). Also, a study by        Among the MDR Enterobacteriaceae collected in this study, rates of resistance were lowest to tigecycline. The overall global rates of tigecycline resistance among MDR Enterobacteriaceae were 15% (357/2,402) for Enterobacter spp., 6% (235/4,098) for Klebsiella spp., and 0.2% (8/3,222) for E. coli. Although the highest rates of tigecycline resistance among MDR Enterobacteriaceae were recorded for Enterobacter spp., the yearly rates of resistance among E. aerogenes and E. cloacae isolates fluctuated between 1% and 42% during the study period, and this could be explained by the low numbers of isolates submitted in some years. By year, the number of MDR E. aerogenes isolates collected ranged from 6 in 2014 to 83 in 2008. Yearly totals of MDR E. cloacae isolates were higher than those for MDR E. aerogenes isolates, but these totals only exceeded 200 isolates in four out of seven study years (2007 to 2010).
The overall global rate of tigecycline resistance among isolates of MDR Klebsiella spp. in the current study was 6% (K. oxytoca, 12/203; K. pneumoniae, 223/3,895). For MDR K. oxytoca isolates, the overall rate of tigecycline resistance may be difficult to interpret due to low isolate numbers (Յ50 isolates collected per study year).  [3/13]). Two of the three tigecycline-resistant K. oxytoca isolates collected in North America and Europe in 2006 and 2008, respectively, were submitted by the same center from each region. This suggests a localized incidence of tigecycline resistance during these two study years.
Higher numbers of MDR K. pneumoniae isolates were collected than numbers of MDR K. oxytoca isolates, and global rates of tigecycline-resistant MDR K. pneumoniae isolates ranged from 3% (4/147) to 11% (9/83). Despite this, the rates of tigecyclineresistant MDR K. pneumoniae isolates decreased from 9% (27/307) in 2012 to 3% (4/147) in 2014, which may signify the start of a decline in global resistance. Further surveillance will be needed to follow this trend. In their study of bla KPC -carrying K. pneumoniae in Palermo, Italy, Bonura et al. (16) reported a shift toward a polyclonal epidemic, which highlights that the evolution of resistance is complex and that the importance of changing patterns of resistance should not be underestimated.
In this study, the identification of tigecycline-resistant MDR E. coli from 2009 onwards indicates that these organisms have recently acquired mechanisms of resistance to the glycylcyclines. In the literature, occurrences of emerging tigecycline resistance among patients with E. coli infections have been reported in the United Kingdom (17) and in Italy (18). In both cases, E. coli isolates that were initially susceptible to tigecycline in vivo became tigecycline resistant after prolonged antimicrobial administration and, furthermore, in vitro the resistant isolates were found to produce carbapenemases: New Delhi metallo-␤-lactamase 1 (NDM-1) (17) and K. pneumoniae carbapenemase 3 (KPC-3) (18). These carbapenemases are active against thirdgeneration cephalosporins and carbapenems; therefore, the acquisition of tigecycline resistance likely confers an MDR phenotype. Stone et al. (17) reported the development of resistance in vivo after 53 days of tigecycline treatment, and Spanu et al. (18) reported resistance after 21 days of treatment. Despite these reports of development of tigecycline resistance among E. coli isolates, the current TEST study shows that tigecycline remains active against the majority of MDR E. coli isolates, and the TEST publication by Hoban et al. (12) reported that tigecycline was active against carbapenemresistant E. coli isolates collected between 2004 and 2013.
By organism, the highest overall rates of multidrug resistance reported in the present study were among A. baumannii isolates, for which 44% of isolates collected globally were MDR. By year, the results presented in Fig. 1 show an increase in the rates of MDR A. baumannii during the study period, from 23% (309/1,323) in 2004 to 63% (447/712) in 2014. The previous TEST publication by Garrison et al. (19) reported increasing global rates of MDR A. baumannii isolates between 2004 and 2007, and Mendes et al. (14) described a global increase in the rates of MDR Acinetobacter spp. between 2005 and 2009. Our report shows that multidrug resistance among A. baumannii isolates contin-ues to increase; given the limited treatment options for infections caused by such organisms, this is a cause for concern.
The majority of MDR isolates collected globally were resistant to levofloxacin, with rates of resistance ranging from 83% of E. aerogenes isolates to 98% of E. coli. The World Health Organization (WHO) recently published a report on global antimicrobial resistance that included national data on rates of resistance from Africa, the Americas, Eastern Mediterranean, Europe, Southeast Asia, and the Western Pacific (20). In their report, resistance rates among E. coli isolates of greater than 50% were reported to fluoroquinolones in all regions except Europe. They also showed that infections caused by fluoroquinolone-resistant E. coli isolates were associated with increased mortality. Fluoroquinolone resistance has been linked with extended-spectrum ␤-lactamase production among the Enterobacteriaceae (4, 6); therefore, extended-spectrum ␤-lactamase production may be an indicator of multidrug resistance.
Although global surveillance studies, such as TEST, have reported important information on changes in antimicrobial activity and resistance, there are certain limitations to the data presented. One such limitation is the yearly variation in the numbers of participating centers. For example, the Asia-Pacific region stopped submitting isolates between 2010 and 2014. In this region, the rates of MDR A. baumannii isolates increased from 29% (14/49) in 2005 to 60% (97/161) during the final year of participation. The lack of isolates from this region after 2010 will have impacted the global results. Furthermore, the majority of centers participating in TEST were located in Europe and North America; therefore, changes in these regions could have had a greater impact on the global data.
Despite these limitations, the data presented in this study show that tigecycline has remained active against this global collection of Gram-negative pathogens. The collection of small numbers of tigecycline-resistant MDR E. coli isolates is of concern, however, and highlights the importance for the continued surveillance of tigecycline activity globally. The increasing rates of MDR A. baumannii isolates must also be monitored, and this information should be used to aid health care facilities in reducing MDR infections worldwide. Overall, more global studies of MDR pathogens are needed if the ongoing problem of antimicrobial resistance is to be addressed.

MATERIALS AND METHODS
A total of 611 TEST centers submitted MDR Gram-negative isolates between 2004 and 2014. The numbers of centers located in each study region were as follows: Africa, 20; Asia-Pacific Rim, 52; Europe, 219; Latin America, 69; the Middle East, 23; North America, 228. Not all study centers submitted isolates during all study years. Centers from the Asia-Pacific Rim did not participate in the study after 2010. Isolates were collected from all body sites from patients with known hospital-or community-acquired infections.
MICs were determined in local laboratories using Clinical and Laboratory Standards Institute (CLSI) guidelines for the broth microdilution methodology (21). Antimicrobial susceptibility was assessed using breakpoints approved by the CLSI (22), except for tigecycline, for which the U.S. Food and Drug Administration (FDA) breakpoints were used (8). Breakpoints were not available for tigecycline against A. baumannii or P. aeruginosa isolates. Full methodology details for the TEST study have been published previously (23).
The Cochran-Armitage trend test was used to identify statistically significant changes in susceptibility between 2004 and 2014, with a cutoff value of P Ͻ 0.01 to indicate significance, due to the large number of trend tests performed.