Antimicrobial susceptibility among Gram-positive and Gram-negative organisms collected from the Latin American region between 2004 and 2015 as part of the Tigecycline Evaluation and Surveillance Trial

Background The in vitro activity of tigecycline and comparator agents was evaluated against Gram-positive and Gram-negative isolates collected in Latin American centers between 2004 and 2015 as part of the Tigecycline Evaluation and Surveillance Trial (T.E.S.T.) global surveillance study. Methods Minimum inhibitory concentrations (MICs) were determined using the broth microdilution methodology according to the Clinical and Laboratory Standards Institute (CLSI) guidelines. Antimicrobial susceptibility was determined using CLSI breakpoints, except for tigecycline for which the US Food and Drugs Administration breakpoints were used. Results A total of 48.3% (2202/4563) of Staphylococcus aureus isolates were methicillin-resistant S. aureus (MRSA). All MRSA isolates were susceptible to linezolid and vancomycin, and 99.9% (2199/2202) were susceptible to tigecycline. Among Streptococcus pneumoniae isolates, 13.8% (198/1436) were penicillin-resistant; all were susceptible to linezolid and vancomycin, and 98.0% (194/198) were susceptible to tigecycline. Susceptibility was >99.0% for linezolid and tigecycline against Enterococcus faecium and Enterococcus faecalis isolates. A total of 40.8% (235/576) E. faecium and 1.6% (33/2004) E. faecalis isolates were vancomycin-resistant. Among the Enterobacteriaceae, 36.3% (1465/4032) of Klebsiella pneumoniae isolates, 16.4% (67/409) of Klebsiella oxytoca isolates and 25.4% (1246/4912) of Escherichia coli isolates were extended-spectrum β-lactamase (ESBL) producers. Of the ESBL-producing K. pneumoniae and E. coli isolates, susceptibility was highest to tigecycline [93.4% (1369/1465) and 99.8% (1244/1246), respectively] and meropenem [86.9% (1103/1270) and 97.0% (1070/1103), respectively]. A total of 26.7% (966/3613) of Pseudomonas aeruginosa isolates were multidrug-resistant (MDR). Among all P. aeruginosa isolates, susceptibility was highest to amikacin [72.8% (2632/3613)]. A total of 70.3% (1654/2354) of Acinetobacter baumannii isolates were MDR, and susceptibility was highest to minocycline [88.3% (2079/2354) for all isolates, 86.2% (1426/1654) for MDR isolates]. Tigecycline had the lowest MIC90 (2 mg/L) among A. baumannii isolates, including MDR isolates. Conclusions This study of isolates from Latin America shows that linezolid, vancomycin and tigecycline continue to be active in vitro against important Gram-positive organisms such as MRSA, and that susceptibility rates to meropenem and tigecycline against members of the Enterobacteriaceae, including ESBL-producers, were high. However, we report that Latin America has high rates of MRSA, MDR A. baumannii and ESBL-producing Enterobacteriaceae which require continued monitoring. Electronic supplementary material The online version of this article (doi:10.1186/s12941-017-0222-0) contains supplementary material, which is available to authorized users.


Background
Resistance among clinically important organisms to antimicrobial agents is severely threatening the repertoire of treatment options for common infections. The challenge is intensified by the fact that several of these organisms are resistant to multiple antimicrobials. Antimicrobial resistance is a global problem, with some regions noted to have higher rates of resistance than others. For example, Latin America is reported to have high rates of extended-spectrum β-lactamase (ESBL) producing Enterobacteriaceae, methicillin-resistant Staphylococcus aureus (MRSA), and multidrug-resistant (MDR) Acinetobacter spp. [1][2][3][4]. Also of concern are carbapenemaseproducing Klebsiella pneumoniae. There have been many outbreaks in the Latin American region [5], particularly in Panama where there was an outbreak from 2011 to 2013 that was difficult to control [6]. Carbapenemases of the metallo-β-lactamases type, such as NDM-1 and VIM, have also emerged in the region [5,7]. The lack of effective antibiotics against these multi-resistant strains has resulted in an increased use of colistin, and colistinresistant strains of Enterobacteriaceae, Pseudomonas spp. and Acinetobacter spp. are beginning to appear [8].
The Tigecycline Evaluation and Surveillance Trial (T.E.S.T.) is an ongoing global surveillance study that has monitored the in vitro activity of tigecycline and comparator agents since 2004. Tigecycline is a broad-spectrum glycylcycline with activity against Gram-positive and Gram-negative organisms. In this report we examine the activity of tigecycline against Gram-positive and Gram-negative organisms collected from centers across Latin America between 2004 and 2015. Data from isolates collected in Latin America in the earlier years of the T.E.S.T. study have previously been presented. Rossi et al. [9] reported antimicrobial resistance between 2004 and 2007, Fernández-Canigia et al. [10] presented antimicrobial susceptibility between 2004 and 2010 (Gramnegative isolates only), and Garza-González et al. [11] presented susceptibility data for S. aureus isolates collected between 2004 and 2010.

Methods
The Latin American countries that participated in T.E.S.T. were Argentina, Brazil, Chile, Colombia, El Salvador, Guatemala, Honduras, Jamaica, Mexico, Nicaragua, Panama, Puerto Rico and Venezuela. Not all study centers submitted isolates during all study years. All body sites were acceptable sources for isolate collection and a maximum of 25% of isolates could be from urine. Isolates were collected from both inpatients and outpatients with documented hospital-or community-acquired infections, and one isolate was permitted per patient.
Detailed materials and methods for the T.E.S.T. study have been described in previous publications (e.g. [12]). Isolate identification and susceptibility testing were performed at the individual centers. Minimum inhibitory concentrations (MICs) were determined using the broth microdilution methodology according to the Clinical and Laboratory Standards Institute (CLSI) guidelines [13]. Antimicrobial susceptibility was determined using breakpoints approved by the CLSI [14], except for tigecycline for which the US Food and Drugs Administration (FDA) breakpoints were used [15]. When determining Streptococcus pneumoniae susceptibility to penicillin, oral penicillin V breakpoints were used. In 2006, four antimicrobials (azithromycin, clarithromycin, erythromycin and clindamycin) were added to the S. pneumoniae T.E.S.T. panel and, where available, isolates were tested retrospectively.
ESBL production among Klebsiella spp. and Escherichia coli were determined by IHMA according to CLSI guidelines using cefotaxime, cefotaxime-clavulanic acid, ceftazidime and ceftazidime-clavulanic acid disks. Haemophilus influenzae isolates were tested for β-lactamase production using center specific methodology.
Year by year data shows susceptibility rates were stable for levofloxacin (≥97.0% in all years) and ceftriaxone (≥89.0% in all years), however susceptibility rates were more variable for the other agents on the panel (Additional file 1: Table S1). All S. pneumoniae isolates were susceptible to tigecycline between 2010 and 2015, prior to that susceptibility increased from 78.0% ( Table S1). Susceptibility to levofloxacin and tigecycline among penicillin-resistant isolates was also high (98.0% for each for the 2004-2015 pooled time period); however, susceptibility to ceftriaxone and amoxicillin-clavulanic acid among penicillin-resistant isolates was reduced (69.2 and 63.6%, respectively) ( Table 3).
A total of 1339 isolates of S. agalactiae were submitted to T.E.S.T. between 2004 and 2015 in Latin America (Table 3). Susceptibility to the majority of agents was unchanged over the course of the study (Additional file 1: Table S1) and all isolates were susceptible to ampicillin, linezolid, meropenem, penicillin and vancomycin ( Table 3). More than 97.0% were susceptible to tigecycline, ceftriaxone and levofloxacin; however, susceptibility to minocycline was lower (27.6%) and variable over the course of the study (Table 3; Additional file 1: Table S1).
A total of 576 isolates of E. faecium were collected in Latin America between 2004 and 2015, and vancomycin   resistance was seen in 40.8% ( Table 2). Rates of vancomycin resistance among E. faecium isolates were highest in Brazil (77.3%), Chile (55.9%) and Argentina (52.2%).
Rates of vancomycin resistance among E. faecium isolates were lower in countries in Central America (Guatemala, Honduras and Panama) than in the rest of the Latin America. Vancomycin resistance rates were variable over the course of the study (Fig. 1). High percentages (>99.0%) of E. faecium isolates were susceptible to linezolid and tigecycline (Table 3) and rates were unchanged over the course of the study (Additional file 1: Table S1).
A single E. faecium isolate was non-susceptible to linezolid. Among the vancomycin-resistant isolates, all were susceptible to linezolid and 98.7% were susceptible to tigecycline (Table 3).
Of the 2004 E. faecalis isolates submitted between 2004 and 2015, 1.6% were vancomycin-resistant ( Table 2). Rates of vancomycin resistance were ≤2.5% in all countries except Brazil, which had a resistance rate of 14.3%. None of the E. faecalis isolates submitted by Central American countries were resistant to vancomycin. No vancomycin-resistant E. faecalis isolates were collected in 2004, 2010, 2011, 2013 or 2014, and less than 10 isolates were collected for any other year (Additional file 1: Table S1). Susceptibility rates for all E. faecalis isolates were >98.0% for linezolid, tigecycline, ampicillin, penicillin and vancomycin (Table 3) and were unchanged over time (Additional file 1: Table S1). This high level of susceptibility to linezolid and tigecycline was maintained among vancomycin-resistant isolates, whereas susceptibility to ampicillin and penicillin decreased to 78.8 and 75.8%, respectively (Table 3). Susceptibility among all E. faecalis isolates to minocycline decreased from 56.0% (14/25) in 2004 to 20.0% (8/40) in 2011; rates after 2011 were variable but did show a trend towards increasing susceptibility (Additional file 1: Table S1).

Gram-negative organisms
Data on rates of Gram-negative resistant phenotypes of K. pneumoniae, Klebsiella oxytoca, E. coli, P. aeruginosa, A. baumannii and H. influenzae are presented by country in Table 4, and by year in Fig. 2 (with the exception of K. oxytoca and H. influenzae). Antimicrobial susceptibility data for these organisms, as well as Enterobacter spp. and Serratia marcescens, are presented in Table 5, and year by year susceptibility data are presented in Additional file 1: Table S2.
Among the 4032 K. pneumoniae isolates submitted between 2004 and 2015, 36.3% were ESBL-producers ( Table 4) and rates of ESBL production ranged from 18.3% in Venezuela to 73.7% in Honduras. Figure 2a shows K. pneumoniae ESBL production rate was relatively stable for the 2004-2015 time period. Among K. pneumoniae isolates, susceptibility was highest to tigecycline, meropenem and amikacin (95.7, 90.9 and 86.9%, respectively); susceptibility among ESBL-producers was also highest to these agents (93.4, 86.9 and 75.3%, respectively) ( Table 5). Susceptibility rates to tigecycline and meropenem were stable across the years of the study, whereas rates to amikacin were more variable (Additional file 1:   Table  S2). The susceptibility rate to levofloxacin among all K. pneumoniae isolates was 65.3%, and among ESBLproducing isolates was 39.4% (resistance rates 31.6 and 55.6%, respectively) ( Table 5) and although there was some variability no trend was seen over time (Additional file 1: Table S2). A total of 409 K. oxytoca isolates were collected, of which 16.4% were ESBL-producers (Table 4). Among all K. oxytoca isolates, susceptibility rates were highest to tigecycline, meropenem and amikacin (98.0, 97.6 and 95.6%, respectively) ( Table 5) and little variability was seen over time (Additional file 1: Table S2). Numbers of ESBL-producing K. oxytoca were low in each year (≤14 isolates); in years with ≥10 isolates rates of susceptibility were highest to tigecycline, meropenem and amikacin (Additional file 1: Table S2).
Of the E. coli isolates collected, 25.4% were ESBLproducers and the percentage of isolates that produced ESBLs was highest in Honduras, Guatemala and Mexico (Table 4). Among all E. coli isolates, susceptibility was highest to tigecycline and meropenem (99.7 and 98.1%, respectively), and these rates were similar among the ESBL-producers (99.8 and 97.0%, respectively) ( Table 5).
Rates of susceptibility to tigecycline and meropenem were stable across the 2004-2015 time period (Additional file 1: Table S2) Table S2). For the other agents on the panel, susceptibility rates were lower among ESBL-producing E. coli compared with E. coli isolates overall. The rate of levofloxacin susceptibility among all E. coli isolates was 47.8%, and among ESBL-producing E. coli isolates was 11.8% (resistance rates were 48.8 and 84.4%, respectively). Susceptibility to meropenem was lower among K. pneumoniae isolates (90.9%) than E. coli isolates (98.1%).
Of the 3613 P. aeruginosa isolates submitted by Latin American centers between 2004 and 2015, 26.7% were MDR ( Table 4). The countries that submitted the highest percentages of MDR P. aeruginosa isolates were Guatemala, Venezuela, Honduras and Brazil (43.8, 32.6, 31.9 and 31.5%, respectively). The year on year rates of MDR were variable across the 2004-2015 time period, however <20% of P aeruginosa were MDR between 2012 and 2015 (Fig. 2b). Breakpoints were available for six of the agents on the panel. Of these, the agents with the highest rate of susceptibility against P. aeruginosa was amikacin (72.8%) (  Table S2). Among all P. aeruginosa isolates, 56.8% were susceptible to ceftazidime. Among     Table S2).
Of the 1300 H. influenzae isolates submitted between 2004 and 2015, 20.8% were β-lactamase positive ( Table 4). The country with the highest rate of β-lactamase positive isolates was Mexico (28.0%), whilst the lowest rate was in Colombia (5.4%). All H. influenzae isolates were susceptible to levofloxacin and meropenem ( Table 5) and rates of susceptibility were consistent across the years of the study (Additional file 1: Table S2). Among all H. influenzae isolates and among β-lactamase positive isolates, susceptibility was ≥97.0% to ceftriaxone, amoxicillin-clavulanic acid, cefepime, piperacillin-tazobactam, minocycline and tigecycline.

Discussion
This study reports on the rates of resistant phenotypes and in vitro antimicrobial susceptibility among important Gram-positive and Gram-negative isolates collected in Latin America between 2004 and 2015. It provides an update to previous publications which reported T.E.S.T. data from Latin America [9][10][11]. Tigecycline maintained its in vitro activity against the isolates collected in this study (susceptibility >93.0%, MIC 90 2 mg/L for A. baumannii). As previously reported, tigecycline was not active against P. aeruginosa [16].
Historically, the prevalence of MRSA has been reported to be increasing in the Latin American region.
For example, the SENTRY study reported a significant increase in MRSA rates in Latin America between 1997 and 2006 (from 33.8 to 40.2%; p = 0.007) [17]. Previous T.E.S.T. reports have suggested a stabilization of rates [11] and this T.E.S.T. study of data for isolates collected between 2004 and 2015 continues to suggest that rates are stable in the region, although with country variations. The overall rate of MRSA in this study was 48.3%, which is similar to a SENTRY report from Latin America for the 2011-2014 time period (44.7%) [18]. Recent studies from Europe (between 2012 and 2015) and the USA (between 2005 and 2011) have reported decreasing rates of MRSA [19,20]. Such reports suggest that global efforts regarding infection control and antimicrobial stewardships are having an impact.
Linezolid and vancomycin are key tools in the treatment of MRSA as infections are often caused by organisms resistant to other antimicrobials. As reported by other studies in Latin America [1,3,18,21], all S. aureus isolates (including MRSA) collected as part of T.E.S.T between 2004 and 2015 were susceptible to linezolid and vancomycin. Small numbers of tigecycline non-susceptible isolates were collected in the early years of the T.E.S.T. program, as previously reported by Garza-González et al. [11]. However, from 2010 onwards all S. aureus isolates (including MRSA) were susceptible to tigecycline. This was also the case in the Latin American SENTRY study in which all S. aureus isolates (including MRSA) collected over a similar time (2011-2014) were susceptible to tigecycline [18].  [10], and lower than the global rate for the 2004-2013 T.E.S.T. study period (54.8%) [4]. The year on year data from this study between 2006 and 2015 shows a trend of decreasing A. baumannii susceptibility to meropenem. Acinetobacter spp. strains resistant to carbapenems have increased in prevalence and present a serious treatment challenge to clinicians [27]. As a result older agents, such as colistin, have seen a resurgence in use; however, colistin-resistant and pan-drug-resistant strains have been reported [8,27,29] highlighting the importance of judicious antimicrobial use and stewardship. It is notable, particularly in the case of the Enterobacteriaceae, that from the start of this study until 2009/2010 susceptibility to minocycline decreased and then from 2010/2011 onwards began to increase again so that rates in 2015 are similar to rates from 2004. This has also been reported in both a global analysis of the T.E.S.T. data and also among isolates from skin and soft tissue infections [4,30]. The reasons for this are unclear although there was variability in center involvement throughout the study and the total number of isolates submitted peaked in 2009 with lower numbers of isolates submitted in subsequent years. To our knowledge this has not be reported by other surveillance studies and warrants further analysis.
Surveillance studies such as T.E.S.T are an invaluable tool for monitoring the rate of resistant pathogen phenotypes and antimicrobial susceptibility among clinical pathogens. However, there are a number of limitations to this study. For example, there was a yearly variation in the number of participating centers with a larger number of centers participating in the earlier years of the study than the latter. The center count was at its highest in 2008 (44 centers) and at its lowest in 2012 (4 centers). Furthermore, the number of isolates submitted varied widely from country to country, with almost half of isolates (48.9%) being submitted by Mexico and Argentina combined.

Conclusions
Antimicrobial resistance continues to be a problem in Latin America with high rates of MRSA, ESBL-producing Enterobacteriaceae and MDR A. baumannii. There are limited treatment choices for infections caused by such organisms; however, this study shows that linezolid, vancomycin and tigecycline continue to be active in vitro against Gram-positive organisms such as MRSA. Against resistant Gram-negative organisms, both in Latin America and globally, the rise in antimicrobial resistance is more troubling especially in the context of carbapenem resistance. In vitro, this study reported high percentages of susceptibility to meropenem and tigecycline among Gram-negative organisms (with the exception of P. aeruginosa). However, resistant isolates were identified and warrant continued monitoring. Authors' contributions SV participated in data collection and interpretation as well as drafting and review of the manuscript. MJD was involved in the study design and data interpretation, and drafting and review of the manuscript. Both authors read and approved the final manuscript.