Eur J Clin Microbiol Infect Dis (2007) 26:229–237 DOI 10.1007/s10096-007-0279-3

REVIEW

Antimicrobial therapy for Stenotrophomonas maltophilia infections A. C. Nicodemo & J. I. Garcia Paez

Published online: 3 March 2007 # Springer-Verlag 2007

Abstract Stenotrophomonas maltophilia has emerged as an important nosocomial pathogen capable of causing respiratory, bloodstream, and urinary infections. The treatment of nosocomial infections by S. maltophilia is difficult, as this pathogen shows high levels of intrinsic or acquired resistance to different antimicrobial agents, drastically reducing the antibiotic options available for treatment. Intrinsic resistance may be due to reduced outer membrane permeability or to the multidrug efflux pumps. However, specific mechanisms of resistance such as aminoglycosidemodifying enzymes or the heterogeneous production of metallo-β-lactamase have contributed to the multidrugresistant phenotype displayed by this pathogen. Moreover, the lack of standardized susceptibility tests and their interpretative criteria hinder the choice of an adequate antibiotic treatment. Recommendations for the treatment of infections by S. maltophilia are based on in vitro studies, certain nonrandomized clinical trials, and anecdotal experience. Trimethoprim-sulfamethoxazole remains the drug of choice, although in vitro studies indicate that ticarcillin-clavulanic acid, minocycline, some of the new fluoroquinolones, and tigecycline may be useful agents. This review describes the main resistance mechanisms, the in vitro susceptibility profile, and treatment options for S. maltophilia infections.

A. C. Nicodemo (*) : J. I. G. Paez Department of Infectious Diseases, University of São Paulo Medical School, São Paulo, SP, Brazil e-mail: [email protected] A. C. Nicodemo Rua Barata Ribeiro 414, Conjunto 104, CEP 01308-000 São Paulo, SP, Brazil

Introduction Stenotrophomonas maltophilia is a nonfermentative gramnegative bacillus, previously known as Pseudomonas maltophilia and later as Xanthomonas maltophilia [1–4]. This bacterium is found in various environments such as water, soil, plants, food, and hospital settings, among others [5, 6]. The pathogenic factors and virulence associated with S. maltophilia include the production of proteases and elastases and the ability to adhere to synthetic materials. S. maltophilia adheres avidly to medical implants and catheters, forming a biofilm that renders natural protection against host immune defenses and different antimicrobial agents [7–10]. The incidence of S. maltophilia isolates provided by different hospitals ranges from 7.1 to 37.7 cases per 10,000 discharges [6, 11, 12]. Nosocomial S. maltophilia pneumonia is associated with high mortality, particularly when associated with bacteremia or obstruction. In uncontrolled clinical trials, mortality rates associated with S. maltophilia bacteremia range from 21 to 69% [13, 14]. Senol et al. [13] reported an attributable mortality rate of 26.7% in S. maltophilia bacteremia. The risk factors for infection by S. maltophilia include prolonged hospitalization requiring invasive procedures, previous exposure to broad-spectrum antibiotics, mechanical ventilation, and severe mucositis [12, 15–22]. Stenotrophomonas maltophilia is associated with a broad spectrum of clinical syndromes, including pneumonia, bloodstream infection, skin infections and surgical-site-related infections, urinary tract infections, endocarditis, meningitis, intra-abdominal infections, and endophthalmitis [6, 23–35]. Patients with cystic fibrosis, a hereditary metabolic disorder of the exocrine glands that mainly affects the pancreas, respiratory system, and sweat glands, are commonly colonized by S. maltophilia.

230

Sometimes it is difficult to distinguish between colonization and infection. The differential diagnosis should be based on the association of factors such as physical examination, radiograph results, other clinical or image findings, and laboratory test results, including the microbiological assays. The treatment of infection caused by S. maltophilia is controversial and difficult due to genotypic and phenotypic variability amongst members of S. maltophilia species; intrinsic resistance mechanisms expressed by S. maltophilia against most antimicrobial agents; the ability of S. maltophilia to develop resistance during treatment; poorly standardized susceptibility tests and their interpretative criteria; and the difficulty of transferring in vitro findings to clinical practice, given the lack of randomized clinical trials comparing the efficacy of antimicrobial agents [6, 36, 37].

Resistance mechanisms Resistance due to production of beta-lactamases Beta-lactam resistance is due to the expression of two inducible β-lactamases, L1 and L2, although not all clinical S. maltophilia isolates express β-lactamases, even after induction with a β-lactam agent. L1 metallo-β-lactamase is a homotetramer of 118 kDa. It is a Zn2+-dependent metalloenzyme that hydrolyzes virtually all classes of β-lactam agents, including penicillins, cephalosporins, and carbapenems, but not monobactams. Furthermore, the L1 enzyme is not inhibited by clavulanic acid. L2 serine-β-lactamase is a cephalosporinase that hydrolyzes aztreonam and is completely inhibited by clavulanic acid and partially inhibited by other β-lactamase inhibitors [38–41]. The expression of such β-lactamases is determined by chromosomal genes, which are highly polymorphic within the species [42]. In 2000, Avison et al. [43] demonstrated a constitutively expressed β-lactamase gene from a clinical isolate of S. maltophilia. Its DNA sequence is almost identical to that of blaTEM2, and the expressed enzyme is a Bush type 2a penicillinase with an amino acid sequence identical to that of TEM-2. This gene was present within a transposon in the genome of this strain. These findings suggest that this pathogen can act as a reservoir for mobile β-lactamase genes. Resistance due to efflux systems Multidrug resistance efflux pumps have been identified as an important resistance mechanism in S. maltophilia. The efflux pump is composed of a membrane fusion protein, an energydependent transporter, and outer membrane proteins (OMPs). Alonso and Martinez [44] described the cloning and the characterization of a multidrug efflux pump from S.

Eur J Clin Microbiol Infect Dis (2007) 26:229–237

maltophilia for the first time and named the new system SmeDEF. In 2001, the same authors [45] showed SmeDEF expression in 33% of the S. maltophilia strains studied and a resultant increase in the MICs of tetracyclines, choramphenicol, erythromycin, norfloxacin, and ofloxacin. Gould and Avison [46] examined a collection of 30 phylogenetically grouped clinical S. maltophilia isolates from Europe and North, South, and Central America and compared their resistance profiles to SmeDEF expression levels. Of 20 spontaneous S. maltophilia drug-resistant mutants tested, four overexpressed SmeDEF, but only two carried mutations within the smeT gene, which is the repressor of the S. maltophilia multidrug SmeDEF efflux pump. Therefore, mutation in smeT might be responsible for SmeDEF overproduction in multidrug-resistant strains of S. maltophilia [47, 48]. In the above-mentioned study of 30 clinical isolates, 6 significantly overexpressed SmeDEF. However, smeT is not the only gene product that affects SmeDEF expression, and no general SmeDEF-mediated phenotype can be defined. Li et al. [49] later described the SmeABC system, identifying the SmeC as an outer membrane multidrug efflux protein of S. maltophilia. However, resistance is dependent only upon the SmeC OMP component of this multidrug efflux system. The fact that SmeC but not SmeAB contributes to antimicrobial resistance and can be expressed independently of these genes suggests that SmeC also functions as part of an additional as-yet-unidentified efflux system. Chang et al. [50] have shown that strains expressing the SmeABC and SmeDEF efflux systems are resistant to ciprofloxacin and meropenem, respectively. Aminoglycoside resistance Current literature suggests that multiple mechanisms may be involved in aminoglycoside resistance, such as aminoglycoside-modifying enzymes, temperature-dependent resistance due to outer membrane changes, the efflux-mediated mechanism, and target modification. The enzymatic modification of the aminoglycosides is due to a family of enzymes that includes O-nucleotidyltransferases, O-phosphotransferases, and N-acetyltransferase. In 1999, Lambert et al. [51] identified the chromosomal aac(6′)-Iz gene of S. maltophilia and established that aac(6′)-Iz enzymeproducing strains show higher resistance to gentamicin. Li et al. [52] have demonstrated that aac(6′)-Iz acetyltransferase enzyme-expressing strains exhibit reduced susceptibility, particularly to tobramycin. Recently, Okazaki and Avison [53] have demonstrated the aph(3′)-IIa determinant of S. maltophilia, which encodes resistance to the aminoglycosides class, except for gentamicin. Changes in the lipopolysaccharide (LPS) structure have been correlated with changes in resistance to a variety of

Eur J Clin Microbiol Infect Dis (2007) 26:229–237

antimicrobial agents [54]. S. maltophilia exhibits a temperature-dependent variation in susceptibility to several antibiotics, including aminoglycosides and polymyxin B [55]. Temperature-dependent changes in outer membrane fluidity [56], LPS side-chain length [57], and, possibly, core phosphate content [58] seem to explain the temperaturedependent variation in aminoglycoside susceptibility, implicating LPS as determinant in the aminoglycoside resistance in this organism. The ability of S. maltophilia to alter the size of O-polysaccharide and the phosphate content of LPS at different temperatures, increasing resistance to aminoglycosides at 30°C compared to 37°C, has been shown. McKay et al. [59] cloned a spgM gene from S. maltophilia that was shown to encode a bifunctional enzyme with both phosphoglucomutase and phosphomanomutase activities. Mutants lacking spgM produced less LPS than the spgM+ parent strain and tended to have shorter O-polysaccharide chains. However, spgM mutants displayed a modest increase in susceptibility to several antimicrobial agents and were completely avirulent in an animal infection model. The latter may be related to the resultant serum susceptibility of spgM mutants, which, unlike the wild-type parent strain, were rapidly killed by human serum. This data highlights the contribution made by LPS to the antimicrobial resistance of S. maltophilia. Proteins of the small multidrug resistance (SMR) family have been characterized in some gram-negative bacteria in which resistance is attributed specifically to aminoglycosides. Chang et al. [50] detected the smr gene in six S. maltophilia strains analyzed, although the role of the smr gene in drug resistance by S. maltophilia requires further study. The resistance to aminoglycosides can also be due to target modification (16S rRNA methylation or ribosomal mutations), which has been documented in some gramnegative pathogens and Mycobacterium spp. [60]. Trimethoprim-sulfamethoxazole resistance Stenotrophomonas maltophilia resistance mechanisms to trimethoprim-sulfamethoxazole (SXT) have not been studied thoroughly. Barbolla et al. [61] mentioned the presence of the sul I gene (plasmid-mediated resistance) in three clones for which the MICs of SXT were increased. According to the authors, these findings not only support the increased spread of class one integrons compared to other mechanisms, but also reveal the potential limitations of using SXT therapy in severe infections.

231

biofilms. Stenotrophomonas maltophilia has the ability to adhere to abiotic surfaces. The positive charge of the cell surface of the bacterium seems to be an important element that favors its adhesion to negatively charged surfaces [8]. The biofilm formation on prosthetic materials such as central venous catheters, urinary tract catheters, and heart valves, amongst others, is a biological property of this bacterium. Biofilms are structured communities of bacterial cells enclosed in a self-produced expolysaccharide matrix and adherent to an inert surface. Di Bonaventura et al. [10], in an in vitro study, characterized the kinetics of S. maltophilia biofilm formation: bacteria attach rapidly to polystyrene after 2 h of incubation, and then the biofilm formation increases over time, reaching maximum intensity at 24 h of culture. The production of extracellular slime or glycocalyx is a crucial factor in bacterial adherence and in bacterial protection against host defense mechanisms and antimicrobial agents, which commonly fail to eradicate the biofilms and consequently, the infection [62]. This highlights the need to remove these prosthetic devices in order to eradicate the infection. Susceptibility tests There are several uncertainties surrounding the in vitro susceptibility testing of S. maltophilia, which range from the selection of the antimicrobial agents to be tested, to the best in vitro methodology to be used, to the accuracy of the in vitro methods used, to the correlation between the different methods available [37]. The recommendations established by different professional societies for susceptibility testing of S. maltophilia vary with regard to the selection of antimicrobial agents to be tested, the disk content, the zone diameter interpretative criteria, and the equivalent MIC breakpoints. The Clinical and Laboratory Standards Institute (CLSI) recommends the disk diffusion technique in order to establish the susceptibility of S. maltophilia, but only to SXT, minocycline, and levofloxacin. Other agents may be approved for therapy, but according to the CLSI, their performance has not been sufficiently studied to establish disk diffusion breakpoints. The MIC interpretative breakpoints are available only for ticarcillin-clavulanic acid, ceftazidime, minocycline, levofloxacin, SXT, and chloramphenicol [63]. Therefore, further studies are necessary in order to enhance the in vitro susceptibility testing of S. maltophilia to different antimicrobial agents. Treatment

Biofilm formation Trimethoprim-sulfamethoxazole Although the biofilm formation is not precisely a “resistance mechanism,” it can increase the resistance to antimicrobial agents, which typically fail to eradicate

Trimethoprim-sulfamethoxazole should be considered the empirical choice for clinically suspected S. maltophilia

232

infections and as the treatment of choice for culture-proven infections by this agent. Susceptibility to this combination is above 80%, according to the results of studies using several in vitro methods [12, 14, 37, 64–77]. Sader and Jones [78], studying 2,076 strains as part of the worldwide Sentry Antimicrobial Surveillance Program, reported a resistance rate of 4.7%. Nevertheless, resistance to this combination is increasing in certain centers. Ticarcillin-clavulanic acid and aztreonam-clavulanic acid In general, the β-lactam antibiotics show low activity against S. maltophilia, owing to the previously mentioned resistance mechanisms. Rates of resistance of S. maltophilia to β-lactam agents such as ampicillin, amoxicillin, piperacillin, and aztreonam are invariably high [12, 18, 70–84]. Beta-lactamase inhibitors such as clavulanic acid can sometimes increase the susceptibility of S. maltophilia to such agents [82]. The ticarcillin-clavulanic acid combination has been recommended as a second therapeutic option, mainly in the treatment of patients who experience adverse effects with SXT therapy [6]. Several studies have demonstrated susceptibility above 70% to this in vitro drug combination [66–68, 70, 81, 83]. However, Sader and Jones [78], studying 2,076 strains as part of the worldwide Sentry Antimicrobial Surveillance Program, reported a resistance rate of 54.7%. Nicodemo et al. [37] reported an in vitro resistance rate of 41%, similar to the rates shown in other studies [79, 85]. Garrison et al. [86], using the pharmacodynamic model to evaluate the ticarcillin-clavulanic acid combination, have shown that S. maltophilia strains exhibit partial growth suppression followed by regrowth, suggesting the need for controlled studies to establish the true efficacy of this combination in the treatment of S. maltophilia infections. The aztreonam-clavulanic acid combination (2:1 and 1:1) has good in vitro activity, although difficulties with the interpretation of the diffusion tests in the component ratios and differences in the pharmacokinetics of these drugs restrict their use in the treatment of S. maltophilia infections [72, 82, 84, 85, 87]. Other combinations such as ticarcillinsulbactam, piperacillin-tazobactam, and ampicillin-sulbactam do not show good activity against this bacterium [12, 19, 75, 80, 82, 83, 85]. Cephalosporins and carbapenems Cephalosporins in general show low activity against S. maltophilia, while cefoperazone, ceftazidime, and cefepime exert some in vitro activity. However, resistance rates are undesirably high, as reported in various trials [12, 64, 66, 67, 73, 75, 80, 84, 85, 88–90]. The risk of resistance induction due to β-lactamase production and low β-lactam

Eur J Clin Microbiol Infect Dis (2007) 26:229–237

activity, particularly of the cephalosporins, limits their empirical use in the treatment of S. maltophilia infections [12]. Combinations of cephalosporins with β-lactamase inhibitors, such as ceftazidime-clavulanic acid, cefoperazone-sulbactam, and cefepime-clavulanic acid, are often mentioned anecdotally, but demonstration of in vitro effectiveness is scarce [81, 84, 85]. Stenotrophomonas maltophilia is intrinsically resistant to carbapenems. Howe et al. [91] have shown that both imipenem and meropenem are L1 β-lactamase inducers and, thus, are not effective against in vitro S. maltophilia. Fluoroquinolones New fluoroquinolones such as clinafloxacin, levofloxacin, gatifloxacin, moxifloxacin, and sitafloxacin show superior in vitro activity compared to earlier quinolones [37, 66, 68, 70, 73, 85, 88, 90, 92–95]. The MIC90 of ciprofloxacin has increased over the last few years, which can be explained by ciprofloxacin’s poor Cmax MIC90 ratio [85]. Several studies have shown the low in vitro activity of this agent against S. maltophilia strains [19, 64, 71–75, 79, 83, 85, 88–90, 92– 94]. Gesu et al. [92], in an in vitro study comparing the activities of levofloxacin and ciprofloxacin against clinical bacterial isolates, evaluated 124 S. maltophilia strains and verified susceptibility rates of 85.5 and 58.9%, respectively, to levofloxacin and ciprofloxacin. Valdezate et al. [96] showed that more than 95% of the S. maltophilia strains tested were susceptible to the new fluoroquinolones. Clinafloxacin seems to be the most active fluoroquinolone, as shown by Pankuch et al. [97] and confirmed in further studies that showed clinafloxacin to be two- to fourfold superior to levofloxacin, moxifloxacin, trovafloxacin, and sparfloxacin [70, 93]. Weiss et al. [93], in a comparison of seven fluoroquinolones, showed that clinafloxacin was the most active, inhibiting 95% of the 326 strains analyzed, followed by trovafloxacin (84.3%), moxifloxacin (83.1%), and sparfloxacin (81.5%). Gales et al. [68], as part of the worldwide Sentry Antimicrobial Surveillance Program, demonstrated resistance rates for gatifloxacin of around 2% in Europe and 15% in Canada. Sader and Jones [78] showed low resistance rates for gatifloxacin (14.1%) and levofloxacin (6.5%). Cohn and Waites [98], using a time-kill assay, showed that gatifloxacin had a bactericidal effect against S. maltophilia isolates, suggesting that gatifloxacin might be used to treat strains that show in vitro susceptibility. Biedenbach et al. [94] suggested that gatifloxacin may be used as a monotherapy or together with a second drug in the treatment of refractory infections due to S. maltophilia strains. Giamarellos-Bourboulis et al. [95] demonstrated the in vitro bactericidal effect of moxifloxacin against genetically

Eur J Clin Microbiol Infect Dis (2007) 26:229–237

distinct isolates resistant to SXT; nevertheless, monotherapy with moxifloxacin against respiratory tract infections due to strains for which the MIC is greater than 2 g/l may select resistant mutants. Ba et al. [99] have shown that moxifloxacin exhibits greater bactericidal activity than ciprofloxacin; however, they also verified the selection of mutants exhibiting resistance to both quinolones. Garrison et al. [86] detected the appearance of mutants resistant to both quinolones in a pharmacodynamic model that evaluated ciprofloxacin and levofloxacin. Di Bonaventura et al. [10] showed that rufloxacin, ofloxacin, and grepafloxacin exert significant static activity ( p