Annals of Clinical Microbiology and Antimicrobials Open Access Trends in Antibacterial Resistance among Streptococcus Pneumoniae Isolated in the Usa: Update from Protekt Us Years 1–4

Background: The increasing prevalence of resistance to established antibiotics among key bacterial respiratory tract pathogens, such as Streptococcus pneumoniae, is a major healthcare problem in the USA. The PROTEKT US study is a longitudinal surveillance study designed to monitor the susceptibility of key respiratory tract pathogens in the USA to a range of commonly used antimicrobials. Here, we assess the geographic and temporal trends in antibacterial resistance of S. pneumoniae isolates from patients with community-acquired respiratory tract infections collected between Year 1 (2000–2001) and Year 4 (2003–2004) of PROTEKT US.

During the last decade, antimicrobial resistance has increased both in the USA [6][7][8][9][10][11][12][13] and worldwide [14][15][16][17]. In particular, resistance to penicillin and the macrolides has spread rapidly among isolates of S. pneumoniae [18]. Surveillance data have shown that, among S. pneumoniae isolates obtained from pediatric patients, the proportion exhibiting nonsusceptibility to penicillin increased each year after 1994, reaching 45% in 2000 [19]. There has also been a marked shift to high-level resistance to penicillin and cephalosporins among isolates of S. pneumoniae. Furthermore, several published studies and case reports (reviewed by Rzeszutek and colleagues [20]) have suggested a link between pneumococcal macrolide resistance and treatment failure (resulting in hospitalization) in patients with community-acquired RTIs.
Macrolide resistance among S. pneumoniae is mediated by two major mechanisms: methylation of ribosomal macrolide target sites, encoded by the erm(B) gene, and drug efflux, encoded by the mef(A) gene [18]. While erm(B)mediated resistance predominates across much of the world, the dominant genotype in the USA is mef(A) [21]. S. pneumoniae isolates with both erm(B) and mef(A) genes have also been documented in the USA [21,22], and are typically multidrug resistant and clonal in nature [11,23]. These findings have raised concerns over the continued clinical utility of antibacterial agents, such as the βlactams and macrolides, for the empiric treatment of many community-acquired RTIs.
PROTEKT US (Prospective Resistant Organism Tracking and Epidemiology for the Ketolide Telithromycin in the US) -a longitudinal surveillance study -was initiated in 2000 to monitor resistance trends in S. pneumoniae and other common RTI pathogens in the USA [16]. A major aim of the program is to evaluate the activity of telithromycin, the first in a new class of antibacterial agents, and to compare its activity to that of other commonly used antibacterials. Data from the PROTEKT US study for 2000-2001 and 2001-2002 (Years 1 and 2) indicated a national prevalence of pneumococcal macrolide resistance of 31%, with rates approaching 40% in some southern regions of the country [10,24]. Data from 2000-2003 (Years 1-3) of the PROTEKT US study [13] demonstrated that the proportion of S. pneumoniae isolates exhibiting multidrug resistance has stabilized (31%). However, geographic variations remain and there is an increasing prev-alence of isolates with both erm(B) and mef(A) genes, which is associated with high-level macrolide and multidrug resistance. The potential spread of macrolide and multidrug resistance is of serious concern and requires further monitoring. Telithromycin, however, continues to display potent in vitro activity against S. pneumoniae, including against isolates with the combined erm(B)+mef(A) genotype. This paper reports on the phenotypic susceptibility and the distribution of macrolide resistance genotypes for S. pneumoniae isolates collected in Year 4 (2003Year 4 ( -2004) of the PROTEKT US study; in addition, it provides an update on temporal and geographic trends in resistance patterns over the 4 years of the study.
Details of the methods for isolate storage, transportation, and identification have been reported previously [16].

Antibacterial susceptibility testing
Antibacterial minimum inhibitory concentrations (MICs) were determined using the Clinical and Laboratory Standards Institute (CLSI) broth microdilution method [25] at a central laboratory (CMI, Wilsonville, OR, USA). CLSI MIC interpretive criteria were used to determine susceptibility [26]. Susceptibility of S. pneumoniae isolates to telithromycin was determined using breakpoints approved by both the CLSI and the US Food and Drug Administration (susceptible ≤ 1 μg/mL; intermediate 2 μg/mL; resistant ≥ 4 μg/mL) [26,27]. Multidrug-resistant S. pneumoniae were defined as isolates resistant to ≥ 2 of the following antibiotics: penicillin; second-generation cephalosporins, e.g. cefuroxime; macrolides; tetracyclines; and trimethoprim-sulfamethoxazole.

Genotyping
Erythromycin-resistant (MIC ≥ 1 μg/mL) pneumococcal isolates were analyzed for the presence of erm(B), erm(A) subclass erm(TR), and mef(A) macrolide resistance genes. Isolates in Year 1 were analyzed using a multiplex rapidcycle polymerase chain reaction (PCR) with microwell-format probe hybridization, as described previously [28]; a Taqman ® -based PCR assay was used in Years 2-4 [29].  . In all regions, resistance rates for erythromycin were higher than those for penicil- The temporal trends in the prevalence of resistance to between 1 and 6 antibacterial agents are shown in Figure  2. During Years 1-4, there was no increasing trend towards resistance of isolates to a greater number of antibacterial agents. The prevalence of isolates resistant to 3, 4, or 5 classes of antibacterial, however, was high and approximately 8% of isolates demonstrated resistance to 5 classes of antibiotic.

Resistance mechanisms
Data from Years 1-4 show that mef(A)-encoded resistance was consistently the most commonly expressed genotype (65.7%) ( Figure 3). However, across the 4-year study period, the proportion of macrolide-resistant isolates carrying both the mef(A) and erm(B) genes increased dramatically. In Year 1, 9.7% of macrolide-resistant isolates

Discussion and Conclusion
As part of the PROTEKT US study we have analyzed a large dataset for S. pneumoniae isolates sampled from across the USA. This has resulted in the generation of valuable information regarding temporal and geographic changes in antibacterial resistance patterns over the 4-year period from 2000 to 2004.
The latest data collected between 2003 and 2004 (Year 4) confirm the results of previous reports indicating that the prevalence of pneumococcal penicillin resistance across the USA appears to be stable or is decreasing, whilst intermediate penicillin resistance is increasing slightly [13,30].
There are a number of possible reasons why pneumococcal resistance to penicillin and some other antibacterials (e.g., trimethoprim-sulfamethoxazole) may have stabilized or begun to decrease in the USA [30]. First, recent local [31] and national [32] campaigns to promote appropriate antibacterial prescribing may have exerted a downward pressure on resistance rates as this factor is one of the most important drivers of resistance in communityacquired infections [33]. Another factor may be the introduction in 2000 of the pneumococcal conjugate vaccine (PCV7) for routine immunization of infants; resistance rates have traditionally been highest among the pediatric population and the use of PCV7 has been shown to decrease pneumococcal resistance not only among children but also in the population as a whole, via a herding effect [34]. A third factor may be the introduction of fluoroquinolones as a treatment for respiratory tract infections in adults. As the use of these agents has increased, it is possible that use of other more traditional agents may have declined, thus reducing associated rates of resistance.
Although the overall levels of in vitro penicillin-nonsusceptibility in S. pneumoniae may be a cause for concern, the clinical importance of this phenomenon in the management of pneumococcal pneumonia has been questioned [35][36][37]. There is no evidence of widespread clinical failures among respiratory infections caused by S. pneumoniae strains classified as penicillin-resistant in vitro. Moreover, in respiratory infections documented as being due to resistant pneumococci, the infecting S. pneumoniae strain generally exhibits low-level in vitro resistance (penicillin MIC 1-2 μg/mL); in these cases, the infection can usually be successfully treated using high doses of βlactam antibiotics [38]. Nevertheless, careful monitoring of penicillin resistance rates should continue, especially since reports of S. pneumoniae strains with high-level penicillin resistance (MIC ≥ 8 μg/mL) have appeared recently [39,40]. Furthermore, there is clear evidence that infections of the central nervous system caused by penicillinresistant S. pneumoniae strains can be associated with the failure of β-lactam therapy [38,41]. Consideration of the prevalent rates of penicillin resistance among S. pneumoniae is therefore important in the management of such infections.
In common with recent reports from the USA and other countries, macrolide resistance over the 4 years of PRO-TEKT US exceeded penicillin resistance in all US regions [13,16,17,30,42]. The macrolide resistance rate reported here for Year 4 was similar to those found for Years 1-3 (approximately 30%), suggesting that levels may have plateaued [12,13].
In this study, the proportion of isolates in Year 4 exhibiting resistance to both penicillin and erythromycin decreased compared with previous years; this downward trend was also observed over Years 1-3 of the PROTEKT US study [13]. As noted in other recent surveillance studies [7,[10][11][12][13]43], data for Years 1-4 of PROTEKT US showed considerable regional variation in the rates of resistance to penicillin and erythromycin across the USA.
Macrolide resistance mediated by erm(B) has typically been associated with high-level resistance (MIC 90 values Distribution of resistance genes among macrolide-resistant Streptococcus pneumoniae isolates (n = 11 578) collected during the PROTEKT US study Years 1-4 (2000-2004) Figure 3 Distribution of resistance genes among macrolide-resistant Streptococcus pneumoniae isolates (n = 11 578) collected during the PROTEKT US study Years 1-4 (2000Years 1-4 ( -2004. of ≥ 64 μg/mL), while mef(A)-mediated resistance has historically been characterized by lower-level resistance (MIC 90 values of 4-8 μg/mL) [15,44]. The predominant mechanism of pneumococcal macrolide resistance in the USA is mediated by mef(A) [21]. However, the latest PRO-TEKT US data presented here confirm that the prevalence of the mef (A) genotype is decreasing and that clones expressing both erm(B) and mef(A) genes are increasing in prevalence. Of additional importance, the mef(A)-positive isolates were found to exhibit levels of macrolide resistance that were higher (MIC mode 16 μg/mL) than those reported in previous surveillance studies (4-8 μg/mL) [15,43]. This may impact on the ability of the macrolides to eradicate such strains from the sites of infection in patients with community-acquired RTIs.
Molecular epidemiology studies undertaken as part of PROTEKT US have shown that, of the erm(B)+mef(A) isolates analyzed, > 90 % are clonally related to the multidrug-resistant international Taiwan 19F-14 clonal complex 271 [12]. Since these strains show high-level macrolide and multidrug resistance, their spread across the USA represents a serious public health threat.
The introduction of the 7-valent pneumococcal vaccine (PCV7) in 2000 was intended to reduce the incidence of pneumococcal disease in children. Recent evidence suggests that this reduction has indeed occurred [19,45], with decreases of 58% in 2001 and 66% in 2002 in the number of invasive pneumococcal infections in children. However, the vaccine does not provide coverage against all S. pneumoniae serotypes. Most dual erm(B)+mef(A) isolates have been characterized as either serotype 19A or 19F and, although serotype 19F is represented in the PCV7 vaccine, 19A is not. As a result, incidence of the nonvaccine serotype 19A multidrug-resistant clone is proportionally higher in the pediatric population than in the past. According to recent surveillance data covering pneumococcal isolates collected in the USA, the prevalence of vaccine serotype 19F has decreased since introduction of PCV7, while that of nonvaccine serotype 19A has concomitantly increased [34,46] [46].
Since the introduction of newer macrolide antibiotics, there has been a steady increase in macrolide resistance from year to year among pneumococci, which has been correlated with their consumption [47]. Macrolide antibacterials are commonly used for the empiric treatment of community-acquired RTIs; therefore, pneumococcal macrolide resistance is of increasing concern in the clinical set-ting [18]. In recent years, a number of reports have been published linking occurrences of macrolide treatment failure (often resulting in hospitalization with breakthrough bacteremia) to infection by macrolide-resistant strains of S. pneumoniae in patients with community-acquired RTIs. Clinical failures in patients treated with azithromycin and clarithromycin have been documented [48][49][50][51], and the number of reports appears to be increasing. It is notable that clinical failures have been reported in patients infected with pneumococcal strains expressing mef(A)encoded macrolide resistance as well as in those with erm(B)-mediated resistance [52,53]. Expression of both erm(B) and mef(A) among S. pneumoniae isolates is strongly associated with the emergence of multidrug resistance. Almost all of the isolates (99.8%) expressing this dual mechanism of macrolide resistance in Year 4 of the study exhibited such resistance. Multidrug resistance has also been linked to an increased risk of clinical failure [54].
Telithromycin represents the first in a new class of antimicrobials -the ketolides. Telithromycin demonstrated potent in vitro activity against S. pneumoniae isolates, including erm(B)+mef(A) macrolide-resistant strains. The in vitro susceptibility of S. pneumoniae to telithromycin was very high in each of the study years, irrespective of the macrolide resistance mechanism. Overall, > 99% of macrolide-resistant S. pneumoniae isolates were susceptible to this agent. These data are in agreement with corresponding longitudinal data from the international PROTEKT Global study (1999)(2000)(2001)(2002)(2003), which indicate that no significant change in telithromycin susceptibility has been observed since the launch of the drug in some European countries in 2000-2001 [55,56]. Currently, telithromycin is licensed in the USA for treating community-acquired pneumonia in adults; however, the most recent Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of communityacquired pneumonia in adults [57] do not carry any recommendations regarding the clinical use of telithromycin. Recommendations will be finalized when further evaluation of the safety of telithromycin by the US Food and Drug Administration has been completed.
The findings in this study highlight the need for the judicious use of antimicrobials and the continued monitoring of pneumococcal resistance patterns -in particular, the spread of multiresistant clones. Physicians should take local or regional resistance patterns into consideration when choosing empiric antibacterial treatment for community-acquired RTIs.
In summary, antimicrobial resistance in S. pneumoniae appears to have stabilized in the USA. However, geographic variations remain, and the prevalence of isolates with the combined erm(B) and mef(A) genotype, associated with high-level macrolide resistance (MIC 50 > 256 μg/mL) and multidrug resistance, continues to increase.
Telithromycin retains potent in vitro activity against S. pneumoniae, including isolates with the combined erm(B)+mef(A) macrolide resistance genotype.

Competing interests
It may appear that SGJ has a competing interest as it relates to this manuscript as he was an employee of Aventis Pharmaceutical, the sponsor of the PROTEKT project, from May of 2001 until August of 2002 and has served as a consultant as well as a member of the company's Speakers Bureau at various times since leaving the company.
SB has received contract funding, expense reimbursements, and honoraria from sanofi-aventis.
DJF has received research grants and consultancy fees from sanofi-aventis related to telithromycin research, publications, and presentations. Sanofi-aventis are financing the production of this manuscript.

Authors' contributions
SGJ contributed to the analysis and interpretation of data related to this manuscript and provided clinical isolates to the project for testing. In addition, SGJ was involved in revising the manuscript critically for important intellectual content.
SB actively participated in the overall design and coordination of the study, collection of data, review of the manuscript, and has granted final approval for the publication of the manuscript.
DJF and colleagues at GR Micro Limited undertook the laboratory testing, data collection and analysis, and drafted the paper.