Global in vitro activity of tigecycline and comparator agents: Tigecycline Evaluation and Surveillance Trial 2004–2013

The Tigecycline Evaluation and Surveillance Trial (TEST) is a global antimicrobial susceptibility surveillance study which has been ongoing since 2004. This report examines the in vitro activity of tigecycline and comparators against clinically important pathogens collected globally between 2004 and 2013. Antimicrobial susceptibility was determined using guidelines published by the Clinical and Laboratory Standards Institute. The Cochran Armitage Trend Test was used to identify statistically significant changes in susceptibility between 2004 and 2013. Among the Enterobacteriaceae susceptibility was highest to the carbapenems [imipenem 97.1% (24,655/25,381), meropenem 97.0% (90,714/93,518)], tigecycline (97.0%, 115,361/118,899) and amikacin (96.9%, 115,200/118,899). Against Acinetobacter baumannii the highest rates of susceptibility were for minocycline (84.5%, 14,178/16,778) and imipenem (80.0%, 3,037/3,795). The MIC90 for tigecycline was 2 mg/L. 40% (6,743/16,778) of A. baumannii isolates were multidrug-resistant. Enterococci were highly susceptible to tigecycline and linezolid (>99%); vancomycin resistance was observed among 2% of Enterococcus faecalis (325/14,615) and 35% of Enterococcus faecium (2,136/6,167) globally. 40% (14,647/36,448) of Staphylococcus aureus were methicillin-resistant while 15% (2,152/14,562) of Streptococcus pneumoniae were penicillin-resistant. Against S. aureus and S. pneumoniae susceptibility to linezolid, vancomycin, and tigecycline was ≥99.9%. Globally, 81% (331/410) of statistically significant susceptibility changes during the study period were decreases in susceptibility. Amikacin, the carbapenems, and tigecycline were active against most gram-negative pathogens while linezolid, tigecycline, and vancomycin retained activity against most gram-positive pathogens collected in TEST during 2004–2013.


Introduction
While the initial development of antimicrobial resistance mechanisms may be a local event, antimicrobial resistance has become a problem of global concern, usually resulting in prolonged and expensive therapy [1]. Global travel and migration as well as international trade have blurred the traditional geographical boundaries between countries and continents, enabling the rapid and global spread of resistant organisms [2]. Numerous important resistance mechanisms have shown alarming increases in distribution in recent years, such as extended-spectrum β-lactamases (ESBLs) and carbapenemases [3,4]. This situation is further complicated by the current shortage of new antimicrobial development, increasing the probability that today's resistant organisms may become tomorrow's pan-resistant pathogens [5].
The Tigecycline Evaluation and Surveillance Trial (TEST) is a global surveillance study which has been ongoing since 2004. It has been designed specifically to monitor the in vitro activity of the broad-spectrum antimicrobial tigecycline plus comparator antimicrobial agents against numerous clinically important gram-negative and gram-positive organisms. In this report, we examine the in vitro activity of tigecycline and comparators against a collection of gram-negative and -positive organisms collected from medical centres globally between 2004 and 2013. This report updates that of Garrison et al. [6], who examined global antimicrobial susceptibility and resistance rates between 2004 and 2007.

Materials and methods
Materials and methods for the TEST study have been published previously e.g. [7] with minimum inhibitory concentrations (MICs) determined according to the broth microdilution method of the Clinical and Laboratory Standards Institute (CLSI) [8].
After receipt by the central laboratory, International Health Management Associates, Inc. (IHMA, Schaumberg, IL, USA), organism identification confirmation was carried out on all isolates using matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (Bruker Daltronics, Bremen, Germany).
Antimicrobial susceptibility was determined using breakpoints approved by the CLSI [9] with the (US) Food and Drug Administration (FDA) breakpoints used for tigecycline [10]. FDA tigecycline breakpoints for Enterococcus faecalis (vancomycin-susceptible) were used for all Enterococcus isolates in this study and penicillin oral breakpoints (susceptible ≤0.06 mg/L, resistant ≥2 mg/L) were used for Streptococcus pneumoniae.

Statistical analysis
The Cochran Armitage Trend Test was used to identify statistically significant changes in susceptibility between 2004 and 2013. A positive change designated a statistically significant decrease in susceptibility; conversely, a negative change indicated that susceptibility had increased significantly. A statistical significance cut-off value of p < 0.01 was used in this analysis. Imipenem and meropenem data were excluded from the statistical analysis. As previously reported imipenem was replaced by meropenem in 2006 so data for the full period of study were not available for these two antimicrobials. Comparative data from 2004 and 2013 are sometimes presented here to support statistically significant changes in susceptibility; where 2004 and/or 2013 numbers were lower than intervening study years, data from other study years (most often 2005 or 2012) have been presented.
Among MDR Enterobacteriaceae (n = 9,372; Table 1), statistically significant changes in susceptibility were observed in all regions excluding the Middle East; most of these changes (20/28) represented decreases in susceptibility (Additional file 1: Table S1).
Significant decreases in susceptibility were observed to all antimicrobials with available breakpoints globally (all p < 0.0001) (Additional file 1: Table S1). Among MDR A. baumannii, significant reductions in global susceptibility were noted to minocycline, and piperacillin-tazobactam (p < 0.0001). Significant increases in MDR A. baumannii susceptibility were observed to amikacin (p < 0.001) in Latin America, although susceptibility was only 8.8% during the complete study interval, and to minocycline (p < 0.001) in Africa.

Haemophilus influenzae
A total of 15,925 isolates of H. influenzae were collected globally. All isolates were highly susceptible (>98.5%) to the antimicrobial agents on the panel with the exception of ampicillin (78.3% susceptibility) (Table 3).
Globally, increased ampicillin susceptibility was recorded (p < 0.001; 76 β-lactamase production was reported in 20.1% (3,207/ 15,925) of H. influenzae isolates globally; these isolates retained high susceptibility to most agents with the exception of ampicillin (0.2% susceptible; Table 3). No statistically significant changes in susceptibility were seen among β-lactamase-positive H. influenzae.
A. baumannii susceptibility decreased in this report, both globally and regionally, to most antimicrobials; 47.1% of A. baumannii isolates from Asia/Pacific Rim were MDR. Molton et al. [12] reported similar results, with 55% of A. baumannii isolates from Singapore being MDR. A. baumannii are adept at acquiring resistance mechanisms and in expanding their global distribution [18]; for example, meropenem susceptibility among A. baumannii from a tertiary care teaching hospital in Mexico decreased from 91.7% in 1999 to 11.8% in 2011 while imipenem susceptibility decreased from 88.2% to 13.9% [19]. These examples highlight the importance of monitoring highly resistant pathogens such as A. baumannii, which have the potential to become panresistant [12].
The activity of tigecycline against resistant clinical bacteria collected globally in 2011 was described by Sader et al. [20] as a part of the SENTRY Antimicrobial Surveillance Program. Tigecycline susceptibility levels of 100% were reported among MRSA, 99.9% among ESBLpositive E. coli, 99.5% among vancomycin-resistant Enterococcus spp., 99.4% among PRSP, and 97.7% among ESBL-positive Klebsiella spp.; also, a tigecycline MIC 90 of 2 mg/L was reported by Sader et al. [20] for Acinetobacter. In a separate study, Sader et al. [21] reported good tigecycline activity against clinical isolates collected in the USA between 2006 and 2012, including MDR phenotypes, with no upwards trends in tigecycline resistance reported. These results accentuate the good in vitro activity of tigecycline against resistance phenotypes in the current study, including ESBL-positive and carbapenem-resistant E. coli, vancomycin-resistant enterococci, MRSA, and PRSP. Tigecycline was also active against carbapenem-resistant Enterobacter spp., with 95.4% of all isolates and 83.0% of carbapenem-resistant isolates susceptible to tigecycline; the next most active antimicrobial against carbapenem-resistant isolates was amikacin, with 68.3% of isolates susceptible. The good in vitro activity of tigecycline against resistant pathogens suggests it may have an important role in the treatment of infections caused by these difficult-to-treat pathogens.
Global susceptibility of Enterobacter spp., K. pneumoniae, S. marcescens, and S. pneumoniae to minocycline increased in this study by ≥20% between 2011 and 2012; this increase has not been reported in previously published studies. The numbers of centres participating in the TEST study increased globally from 197 in 2011 to 398 in 2012; this large influx of isolates from new centres in 2012 may be responsible for the observed susceptibility changes that year. This highlights one of the limitations of longitudinal surveillance studies: inconsistent centre involvement over time, with some centres contributing isolates in several years but others in only a single year. Another bias in TEST is the regional distribution of centres: Europe and the US account for more than two thirds of the centres participating in this study (72.9%, 443/608), thus global results reported are heavily influenced by trends in these regions.