SUPPLEMENT ARTICLE

Impact of Antibiotic Resistance in Gram-Negative Bacilli on Empirical and Definitive Antibiotic Therapy David L. Paterson University of Queensland, Clinical Research Centre, Royal Brisbane and Women’s Hospital, Brisbane, Queensland, Australia; and Division of Infectious Diseases, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania

Serious infections with gram-negative pathogens continue to be associated with considerable mortality. Increasing antibiotic resistance in organisms such as Pseudomonas aeruginosa, Acinetobacter baumannii, and Klebsiella pneumoniae is contributing to difficulties with choosing antibiotics to prescribe for these infections. Optimization of therapy against these organisms starts with the initial empirical antibiotic choice. Surveillance data and hospital or unit antibiograms may inform this decision, although individualization of the initial regimen on the basis of prior antibiotic use and prior isolation of resistant pathogens may be more important. Combinations of antibiotics are often required empirically, and “combination antibiograms” may need to be developed for this purpose. Preliminary data suggest that extending the time over which a dose of antipseudomonal b-lactam antibiotics is infused may improve clinical outcomes; however, this idea remains to be confirmed in randomized trials. The role of direct susceptibility testing in aiding more-rapid initiation of appropriate antibiotic therapy is also being studied. When identification and susceptibility testing is complete, the antibiotic regimen for infections due to gram-negative pathogens can be “fine tuned.” On some occasions, this fine tuning necessitates the introduction of “salvage” antibiotics, such as colistin or tigecycline; on others, it necessitates de-escalation and early termination of therapy. The lack of new antibiotic options against gramnegative pathogens underscores the need for optimization of current therapies and prevention of the spread of these organisms. Available evidence shows that the proportion of gramnegative bacilli resistant to commonly used antibiotics is increasing [1–5]. Prominent examples include fluoroquinolone resistance in Escherichia coli, cephalosporin resistance in a variety of the Enterobacteriaceae, carbapenem resistance in Acinetobacter baumannii and Pseudomonas aeruginosa, and now even resistance to antibiotics used as “salvage” therapy, such as tigecycline and the polymyxins [6]. The mechanisms of this resistance are often complex but include production of multiple b-lactamase types, outer-membrane impermeability, up-regulated efflux pumps, and target-site mutation [3, 5].

Reprints or correspondence: Dr. David L. Paterson, Ste. 3A, Falk Medical Bldg., 3601 Fifth Ave., Pittsburgh PA 15213 ([email protected]). Clinical Infectious Diseases 2008; 47:S14–20  2008 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2008/4706S1-0003$15.00 DOI: 10.1086/590062

S14 • CID 2008:47 (Suppl 1) • Paterson

While the rate of antibiotic resistance is worsening, the antibiotic armamentarium available for use against gram-negative bacilli is barely increasing [7]. The purpose of the present article is to discuss optimization of antibiotic choice for serious infections with gram-negative bacilli. Points of discussion include (1) methods for determining initial empirical antibiotic choice for infections suspected to be caused by gram-negative bacilli; (2) optimizing choice for therapy when the organism is known but susceptibility test results are not yet available; and (3) “fine tuning” therapy when identification is completed and susceptibility test results have been obtained. METHODS FOR DETERMINING INITIAL EMPIRICAL ANTIBIOTIC CHOICE FOR INFECTIONS SUSPECTED TO BE CAUSED BY GRAM-NEGATIVE BACILLI On a daily basis, clinicians are forced to choose an antibiotic for a patient with symptoms and signs of a

serious infection before identification of the bacteria and before susceptibility test results are available. Such treatment can be described as initial empirical therapy. When choosing to use initial empirical therapy, clinicians should be aware that some patients with symptoms and signs of a serious infection do not, in fact, have an infection at all. Examples include patients with pulmonary edema or hemorrhage with radiologic appearances similar to those of ventilator-associated pneumonia (VAP) [8], as well as patients with cerebral hemorrhage, advanced cirrhosis, severe burns, malignancy, or autoimmune conditions with fever or hypotension (symptoms that normally would be expected in a severe infection). In these situations, antibiotic therapy creates an increased risk of disturbance of endogenous flora, potentially leading to Clostridium difficile infection or colonization with antibiotic-resistant bacteria [9, 10]. Habitual use of the same antibiotic regimen for all patients with suspected significant bacterial infection may lead to increased resistance [11] and/or increased rates of inadequate coverage. As described by Kollef [12], in this situation, inadequate initial empirical therapy is associated with increases in adverse patient outcomes. Three methods can be used to increase the probability that initial empirical antibiotic choice will be adequate: (1) individualizing therapy on the basis of a formal review of previously isolated organisms and prior antibiotic use, (2) using unit-based antibiograms, and (3) using data from surveillance studies. Use of data from surveillance studies. Masterton [13] outlined the various types of national and international antibiotic resistance surveillance programs in place and their role in informing antibiotic policy. At the individual patient level, these surveillance programs assist clinicians in ensuring that their antibiotic prescribing choices have a substantial probability of effectiveness. For example, antibiotic resistance surveillance programs indicate that, in most parts of the world, E. coli has a substantial likelihood of being resistant to ampicillin or amoxicillin [13]. Therefore, amoxicillin would be a poor empirical choice for treatment of an infection in which E. coli is a probable cause, such as urinary tract infection. National and international trends in antibiotic resistance, derived from sequential examination of surveillance data, can also alert clinicians to a need to reexamine their empirical antibiotic choices. For example, the emergence of substantial degrees of fluoroquinolone resistance in P. aeruginosa described in national surveillance studies [2] necessitates cautious use of fluoroquinolones as the sole therapy for infections such as VAP. Individualizing initial empirical therapy. Every patient differs according to factors such as underlying comorbidities, severity of illness, type of infection, prior antibiotic use, allergies, and colonization with antibiotic-resistant organisms. Thus, it is reasonable to consider that these variables necessitate an

individualization of initial empirical antibiotic therapy for every seriously ill patient. Infection type is a major determinant of the initial empirical regimen. Regimens aimed against gramnegative pathogens can be divided into those that require antipseudomonal therapy and those that do not. Infections in which P. aeruginosa is a pathogen that needs to be covered in initial empirical therapy include VAP, fever in neutropenic patients, and most cases of undifferentiated infection in critically ill patients. In contrast, skin and soft-tissue infections (including foot infection in diabetic patients), community-acquired pneumonia, community-onset intra-abdominal infections, and community-acquired bacterial meningitis do not typically require empirical antipseudomonal coverage [14–17]. There is also a need to consider whether therapy active against methicillin-resistant Staphylococcus aureus is necessary; the present article, however, pertains strictly to considerations regarding therapy against gram-negative bacilli. b-Lactam, aminoglycoside, and fluoroquinolone antibiotics form the core antipseudomonal drugs. Aminoglycosides are rarely used as monotherapy, except in the treatment of urinary tract infection [18]. Studies performed 110 years ago showed that fluoroquinolones were as effective as b-lactam therapy for some infections with gram-negative bacilli, such as VAP [19]. However, most surveillance studies have shown that the resistance of gram-negative bacilli (especially P. aeruginosa) to fluoroquinolones has been increasing during the past decade [2]. Similarly, the resistance of P. aeruginosa and many Enterobacteriaceae to aztreonam is substantial [3]. Thus, b-lactam antibiotics have remained the most widely used antibiotics for initial empirical therapy, especially when P. aeruginosa may be present. Reasonable evidence suggests that 2 factors—recent receipt of antibiotics and colonizing flora—have a substantial effect on the adequacy of initial empirical therapy. Bhat et al. [20] recently showed how such knowledge could be used to improve the adequacy of initial empirical therapy; after reviewing 158 episodes of serious infection with P. aeruginosa occurring in 140 intensive care unit (ICU) patients, they found that initial empirical antibiotic therapy was microbiologically adequate in only 67% of these episodes [20]. The rate of adequacy of initial empirical antibiotic choice could have been improved to 82% by universally recommending an antipseudomonal carbapenem (imipenem or meropenem) as the first-line choice. Routinely adding amikacin to the carbapenem would have improved the adequacy of the initial empirical antibiotic choice to 95%. However, mathematical modeling suggests that homogeneity of antibiotic regimens is inferior to heterogeneity of antibiotic regimens. Thus, algorithms were developed in which piperacillin/tazobactam or cefepime could also be used as initial empirical antibiotic choices. (It is noteworthy that the addition of tazobactam to piperacillin does not add to piperacillin’s anti-

Treatment of Gram-Negative Bacillus Infection • CID 2008:47 (Suppl 1) • S15

pseudomonal spectrum; however, it does add to its activity against class A b-lactamase–producing organisms [3].) Thirtyseven percent of patients who used piperacillin/tazobactam in the month before the current infection were now infected with a piperacillin/tazobactam-resistant strain, and 64% of patients who had isolation of piperacillin/tazobactam-resistant gramnegative bacilli in the month before the current infection were now infected with piperacillin/tazobactam-resistant P. aeruginosa. Therefore, piperacillin/tazobactam was considered to be appropriate initial empirical therapy only if, during the past month, the patient neither had received the antibiotic nor had isolation of a piperacillin/tazobactam-resistant organism. (The use of a 1-month cutoff for prior antibiotic use or prior isolation of resistant organisms was purely arbitrary; it is not known whether the use of other time periods would change the utility of this algorithm.) Comparable findings and recommendations were made with respect to cefepime. By use of this algorithm, antibiotic heterogeneity was maintained, because substantial proportions of patients used piperacillin/tazobactam, cefepime, or an antipseudomonal carbapenem as initial empirical therapy. Use of unit-based antibiograms and combination antibiograms. Unit-based antibiograms are cumulative antibiotic susceptibility reports for patients in a particular ward of the hospital (for example, a specific ICU) in a specified time period. They are potentially more useful than international, national, or hospital-specific antibiograms for a prescriber making decisions about antibiotic therapy in a specific ICU. Unit-based antibiograms have been shown to have substantially different antibiotic susceptibility results, compared with those of hospital-wide antibiograms [21]. Traditional antibiograms provide data on the susceptibility of organisms to a range of antibiotics, but they do not answer the question of what antibiotic combinations may be optimal against any given organism. Combination antibiotic therapy is necessary as initial empirical antibiotic therapy for infections in which P. aeruginosa is prominent, such as VAP [22]. It is important to point out that combination therapy for P. aeruginosa infection does not prevent the emergence of resistance [23]. Its purpose is to maximize the possibility that the infecting organism will be “covered,” because, in most ICUs, !90% of P. aeruginosa strains are susceptible to any particular antipseudomonal b-lactam. A traditional antibiogram does not give data regarding what aminoglycoside or fluoroquinolone should be used in combination with a core antipseudomonal b-lactam (e.g., an antipseudomonal penicillin, cephalosporin, or carbapenem) to ensure that initial empirical coverage is optimized. Mizuta et al. [24] discussed the role of the “combination antibiogram” to address this question. A combination antibiogram provides information on the percentage of isolates susceptible to a particular antibiotic if the isolate is resistant to a S16 • CID 2008:47 (Suppl 1) • Paterson

core antipseudomonal antibiotic. For example, Bhat et al. [20] showed that, of the P. aeruginosa isolates resistant to cefepime in their institution, 19.5% were resistant to amikacin, 100% to aztreonam, 82.9% to ciprofloxacin, 43.9% to gentamicin, 87.8% to levofloxacin, and 39.0% to tobramycin. Therefore, combinations of cefepime plus amikacin would be much more likely to improve the adequacy of initial empirical antibiotic therapy than would a combination of cefepime plus a fluoroquinolone. Unit-based, combination antibiograms can be constructed to complement individualized antibiotic choices. The following strategies should be used to optimize empirical antibiotic administration for seriously ill patients: • Use unit antibiograms to determine potentially useful antibiotics, especially those active against P. aeruginosa (e.g., cefepime, ceftazidime, piperacillin/tazobactam, imipenem, doripenem, and meropenem, all of which have been found to be active in vitro against 180% of P. aeruginosa isolates from an ICU in the past 12 months). • Determine antibiotics used by the patient in the past month and avoid these antibiotics as empirical therapy for the current infection (e.g., if piperacillin/tazobactam was used 2 weeks ago, avoid it for treating the current infection). • Scrutinize past microbiological records to determine whether organisms have been isolated that are resistant to potentially useful antibiotics; avoid antibiotics that have not demonstrated in vitro activity against gram-negative bacilli isolated from the patient in the past month (e.g., if Enterobacter cloacae resistant to ceftazidime was isolated 2 weeks ago, avoid ceftazidime for the current infection). • Use a local combination antibiogram to determine an optimal secondary antibiotic for the current infection. A combination antibiogram shows that, for example, if P. aeruginosa from the ICU is cefepime resistant, there is a 90% chance that it is susceptible to tobramycin but only a 30% chance that it is susceptible to ciprofloxacin; if P. aeruginosa from the ICU is resistant to an antipseudomonal carbapenem, then there is an 80% chance that it is susceptible to tobramycin but only a 40% chance that it is susceptible to ciprofloxacin, which leaves tobramycin as a sensible antibiotic choice for empirical combination therapy with cefepime or an antipseudomonal carbapenem. • Optimize the dosing of the preferred antibiotic regimen on the basis of pharmacodynamic principles (e.g., if the preferred antibiotic regimen is tobramycin plus cefepime, the aminoglycoside should be dosed aiming for a high peak: MIC ratio and the b-lactam dosed as an extended infusion). THE IMPORTANCE OF DOSE, FREQUENCY, AND DURATION IN OPTIMIZING THE REGIMEN Once the decision regarding the choice of antibiotic regimen has been made, it is necessary to ensure that the dose, frequency

of administration, and duration over which the antibiotic is infused are optimized. The product information of each antibiotic mentioned thus far gives some guidance for appropriate dosage regimens; however, some exceptions exist. For example, guidance regarding dosage regimens for patients receiving continuous renal replacement therapy typically is absent from approved product information. Furthermore, a few lines of evidence suggest that some adjustment of dosage regimens may optimize pharmacodynamic parameters and could potentially result in superior clinical outcomes [25]. At present, most of these data do not come from randomized controlled trials. Lodise et al. [25] examined the impact of a novel dosing regimen of piperacillin/tazobactam on outcome. They compared the clinical outcomes of patients who received the antibiotic in doses of 3.375 g infused over 30 min every 4–6 h (conventional dosing) with those who received the antibiotic in doses of 3.375 g infused over 4 h every 8 h (extended-infusion dosing). Among patients with Acute Physiology and Chronic Health Evaluation (APACHE) II scores of ⭓17 who were infected with P. aeruginosa, 14-day mortality was significantly lower among those treated with the extended-infusion strategy than those given conventional dosing (12.2% vs. 31.6%; P p .04). These data come from a cohort study rather than a randomized controlled trial, but they do provide preliminary evidence to suggest that the dosing of piperacillin/tazobactam should be altered for seriously ill patients with suspected P. aeruginosa infection. The product information for cefepime, approved by the US Food and Drug Administration, suggests a range of possible doses, from 0.5 g every 12 h to 2 g every 8 h, for patients with normal renal function. Dosing at the lower end of this range may be appropriate for urinary tract infection but is most likely inappropriate for patients with serious infections with organisms that may have cefepime MICs in the higher portion of the susceptible range. Specifically, P. aeruginosa and extendedspectrum b-lactamase–producing Enterobacteriaceae may have cefepime MICs of 8 mg/mL and yet may be reported by laboratories using Clinical and Laboratory Standards Institute breakpoints as cefepime susceptible. A recent analysis of 204 patients with bloodstream infection with gram-negative bacteria primarily treated with cefepime showed that patients infected with an organism with a cefepime MIC of 8 mg/mL had a 28-day mortality rate of 56.3%, compared with 24.1% for those with a cefepime MIC !8 mg/mL [26]. Pharmacodynamic assessments using Monte Carlo simulation suggest that dosing regimens of 1–2 g administered over 30 min every 12 h have a low probability of “target attainment” when the cefepime MIC is 8 mg/mL [26]. Again, these data do not come from a randomized trial but suggest that cefepime should be administered empirically at a dose of 2 g every 8 h or as an extended infusion, to adequately treat infection with P. aeruginosa and

extended-spectrum b-lactamase–producing Enterobacteriaceae, which may have cefepime MICs as high as 8 mg/mL. The antipseudomonal carbapenems (imipenem, meropenem, and doripenem) share with antipseudomonal penicillins the pharmacodynamic characteristic that best predicts antimicrobial activity—that is, the percentage of time during the dosing interval during which the serum concentration exceeds the MIC. As Nicolau and colleagues [27, 28] point out, Monte Carlo simulation predicts that extending the infusion duration of meropenem from 30 min to 3 h increases the probability of bactericidal target attainment. In a nonrandomized assessment, Lorente et al. [29] assessed outcomes of patients with VAP who were given meropenem by continuous infusion (1 g over 360 min every 6 h) or by 30-min infusions (1 g over 30 min every 6 h). The group receiving meropenem by continuous infusion had a clinical cure rate superior to that of the group receiving the 30-min infusions (cure rate, 90.5% vs. 59.6%; P ! .001). Extended infusions of meropenem or imipenem are not approved by any regulatory agency. An extended-infusion dosing regimen for doripenem is currently under review by the US Food and Drug Administration. OPTIMIZING CHOICE FOR THERAPY WHEN THE ORGANISM IS KNOWN BUT SUSCEPTIBILITY DATA ARE NOT YET AVAILABLE When blood samples are collected for culture, they are placed in specialized equipment, which regularly assesses for changes in the bottle that are consistent with bacterial growth. If growth occurs (typically 8–24 h after the blood sample is collected for culture), microbiology technicians perform a Gram stain on the broth within the bottle and indicate the Gram stain appearance to the clinician. The broth is then plated onto solid media, incubated, and examined for growth after another 12– 16 h. At this point, the microbiologist is usually able to differentiate Pseudomonas-like organisms from coliforms. Simple biochemical tests may be able to provide species identification (e.g., E. coli) at this point. However, most laboratories use semiautomated identification and susceptibility testing methodologies, which take another 8–16 h to provide results. Thereafter, final identification and susceptibility test results are typically available. Close liaison with clinical microbiologists may allow clinical use of preliminary identification data. Antipseudomonal therapy must be ensured if P. aeruginosa is provisionally identified. An important role of the microbiology laboratory is to alert prescribers to the possibility of the presence of an organism not routinely covered by standard antipseudomonal therapy. For example, the provisional identification of Stenotrophomonas maltophilia should allow coverage of this organism with trimethoprim/sulfamethoxazole if a clinically significant infection is apparent. Earlier appropriate antibiotic therapy may be ini-

Treatment of Gram-Negative Bacillus Infection • CID 2008:47 (Suppl 1) • S17

tiated for infection with multidrug-resistant epidemic organisms, such as carbapenem-resistant A. baumannii or KPC- or extended-spectrum b-lactamase–producing K. pneumoniae, if the characteristic phenotype of these organisms is observed on incubation of specimens on solid media [3]. Despite these not being final identifications or susceptibilities, a 24-h “head start” on the initiation of therapy with antibiotics not typically used as empirical therapy, such as colistin or trimethoprim/sulfamethoxazole, may provide some clinical benefit. Direct susceptibility tests, although clearly not standardized, may also provide a similar benefit. Bouza et al. [30] performed a prospective trial in which patients with VAP were randomized into 2 groups: one group had conventional respiratory specimen culture results reported using standard procedures, whereas the intervention group had a rapid antibiotic susceptibility result obtained by the immediate placement of E-test strips (AB Biodisk) on respiratory tract samples. Patients in the group with rapid reporting of susceptibility testing had fewer days of fever, decreased antibiotic consumption, decreased rates of C. difficile infection, and fewer days of receiving mechanical ventilation. Rapid PCR-based identification methods for gramnegative bacilli are on the horizon [31], although the diversity and complexity of mechanisms of resistance are likely to make rapid determination of antibiotic resistance much more difficult. FINE TUNING THERAPY WHEN IDENTIFICATION AND SUSCEPTIBILITY DATA ARE COMPLETE When identification and susceptibility testing results are available to the clinician, antibiotic regimens can be fine tuned. Recommendations for antibiotic therapy for infections with a range of gram-negative organisms are listed in table 1. Even with the best efforts at optimizing initial empirical regimens, surprises sometimes occur, necessitating intensification of antibiotic therapy. The need for intensification of antibiotic therapy may imply a necessity for the use of salvage options, such as polymyxins or tigecycline. The polymyxins were developed

in the 1950s, when knowledge of pharmacodynamics was undeveloped [32]. As a result, few assessments have been made of optimal dosing strategies for polymyxin B or colistin. Muchneeded work is now proceeding in this area. Tigecycline has activity against some b-lactamase–producing organisms that are resistant to all b-lactam therapies [33]. This antibiotic has unusual pharmacokinetic properties, resulting in low blood concentrations but somewhat higher tissue levels. The utility of the drug must be considered suspect in therapy for bloodstream infections with organisms that have relatively high tigecycline MICs [34]. Evolution of resistance during therapy has been observed, raising further concerns about the drug [35]. The utility of tigecycline in therapy for serious infections, such as VAP, is not yet known. In some circumstances, empirical antibiotic regimens may be broader than is required on the basis of antibiotic susceptibility testing. In this circumstance, de-escalation is required. Penicillins are appropriate antibiotics for de-escalation (e.g., from piperacillin/tazobactam to piperacillin for P. aeruginosa infections and from antipseudomonal b-lactams to ampicillin for some infections with Enterobacteriaceae). It is debatable whether it is appropriate to de-escalate to cephalosporins or fluoroquinolones, because these antibiotics are prone to cause considerable “collateral damage” [36]—that is, promoting resistance not only to that antibiotic but to other classes of antibiotics as well. An important consideration in de-escalation is assessment of the need to discontinue antibiotic therapy. Increasingly, clinical data suggest that many infections with gram-negative bacilli require a shorter duration of therapy than has historically been thought necessary [37]. In summary, clinicians should apply logic in their selection and use of antibiotics in treating severely ill patients, rather than using a “tradition-based” method of selecting treatment. Tradition is hard to break, but it results in patients receiving inappropriate antibiotics and contributes to the development of resistance in bacteria through increased selection pressure. Information is available with which to make a more logical choice of antibiotic, and closer liaison with a microbiologist

Table 1. Recommendations for therapy for serious infections due to specific multidrugresistant gram-negative organisms. Organism Pseudomonas aeruginosa ESBL-producing Enterobacteriaceae

Recommended therapy Antipseudomonal b-lactam Carbapenem

Ceftazidime resistant Enterobacter cloacae Carbapenem resistant Acinetobacter baumannii Stenotrophomonas maltophilia

Fluoroquinolone, carbapenem, or cefepime Colistin/polymyxin B, tigecycline, or ampicillin/sulbactam, depending on susceptibility and location of infection Trimethoprim/sulfamethoxazole plus ticarcillin/clavulanate

NOTE. ESBL, extended-spectrum b-lactamase.

S18 • CID 2008:47 (Suppl 1) • Paterson

will greatly facilitate the decision-making process. Clinicians must be prepared to alter the dose and or infusion rates for certain patients, relying on evidence from small clinical trials to support such actions until they are confirmed in larger clinical trials.

16.

Acknowledgments

18.

Scientific Connexions provided assistance in editing the manuscript (funded by AstraZeneca). Supplement sponsorship. This article was published as part of a supplement entitled “Update on the Appropriate Use of Meropenem for the Treatment of Serious Bacterial Infections,” sponsored by AstraZeneca LP. Potential conflicts of interest. D.L.P. receives research support from Elan, AstraZeneca, and Pfizer and serves on advisory boards for Novartis, AstraZeneca, Merck, Acureon, Protez, and Johnson & Johnson.

17.

19.

20.

References 1. National Nosocomial Infections Surveillance System. National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004; 32:470–85. 2. Neuhauser MM, Weinstein RA, Rydman R, Danziger LH, Karam G, Quinn JP. Antibiotic resistance among gram-negative bacilli in US intensive care units: implications for fluoroquinolone use. JAMA 2003; 289:885–8. 3. Paterson DL, Bonomo RA. Extended-spectrum beta-lactamases: a clinical update. Clin Microbiol Rev 2005; 18:657–86. 4. Giske CG, Monnet DL, Cars O, Carmeli Y, for ReAct-Action on Antibiotic Resistance. Clinical and economic impact of common multidrug-resistant gram-negative bacilli. Antimicrob Agents Chemother 2008; 52:813–21. 5. Perez F, Hujer AM, Hujer KM, Decker BK, Rather PN, Bonomo RA. Global challenge of multidrug-resistant Acinetobacter baumannii. Antimicrob Agents Chemother 2007; 51:3471–84. 6. Paterson DL, Doi Y. A step closer to extreme drug resistance (XDR) in gram-negative bacilli. Clin Infect Dis 2007; 45:1179–81. 7. Spellberg B, Powers JH, Brass EP, Miller LG, Edwards JE Jr. Trends in antimicrobial drug development: implications for the future. Clin Infect Dis 2004; 38:1279–86. 8. Chastre J, Luyt CE, Combes A, Trouillet JL. Use of quantitative cultures and reduced duration of antibiotic regimens for patients with ventilator-associated pneumonia to decrease resistance in the intensive care unit. Clin Infect Dis 2006; 43(Suppl 2):S75–81. 9. Muto CA, Pokrywka M, Shutt K, et al. A large outbreak of Clostridium difficile-associated disease with an unexpected proportion of deaths and colectomies at a teaching hospital following increased fluoroquinolone use. Infect Control Hosp Epidemiol 2005; 26:273–80. 10. Donskey CJ. Antibiotic regimens and intestinal colonization with antibiotic-resistant gram-negative bacilli. Clin Infect Dis 2006; 43(Suppl 2):S62–9. 11. Bonhoeffer S, Lipsitch M, Levin BR. Evaluating treatment protocols to prevent antibiotic resistance. Proc Natl Acad Sci USA 1997; 94: 12106–11. 12. Kollef MH. Inadequate antimicrobial treatment: an important determinant of outcome for hospitalized patients. Clin Infect Dis 2000; 31(Suppl 4):S131–8. 13. Masterton RG. Surveillance studies: how can they help the management of infection? J Antimicrob Chemother 2000; 46(Suppl T2):53–8. 14. Lipsky BA, Armstrong DG, Citron DM, Tice AD, Morgenstern DE, Abramson MA. Ertapenem versus piperacillin/tazobactam for diabetic foot infections (SIDESTEP): prospective, randomised, controlled, double-blinded, multicentre trial. Lancet 2005; 366:1695–703. 15. Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007; 44(Suppl 2):S27–72. Solomkin JS, Mazuski JE, Baron EJ, et al. Guidelines for the selection of anti-infective agents for complicated intra-abdominal infections. Clin Infect Dis 2003; 37:997–1005. Tunkel AR, Hartman BJ, Kaplan SL, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis 2004; 39:1267–84. Vidal L, Gafter-Gvili A, Borok S, Fraser A, Leibovici L, Paul M. Efficacy and safety of aminoglycoside monotherapy: systematic review and meta-analysis of randomized controlled trials. J Antimicrob Chemother 2007; 60:247–57. Fink MP, Snydman DR, Niederman MS, et al. Treatment of severe pneumonia in hospitalized patients: results of a multicenter, randomized, double-blind trial comparing intravenous ciprofloxacin with imipenem-cilastatin. The Severe Pneumonia Study Group. Antimicrob Agents Chemother 1994; 38:547–57. Bhat S, Fujitani S, Potoski BA, et al. Pseudomonas aeruginosa infections in the intensive care unit: can the adequacy of empirical beta-lactam antibiotic therapy be improved? Int J Antimicrob Agents 2007; 30: 458–62. Binkley S, Fishman NO, LaRosa LA, et al. Comparison of unit-specific and hospital-wide antibiograms: potential implications for selection of empirical antimicrobial therapy. Infect Control Hosp Epidemiol 2006; 27:682–7. American Thoracic Society, Infectious Diseases Society of America. Guidelines for the management of adults with hospital-acquired, ventilator-associated, and healthcare-associated pneumonia. Am J Respir Crit Care Med 2005; 171:388–416. Bliziotis IA, Samonis G, Vardakas KZ, Chrysanthopoulou S, Falagas ME. Effect of aminoglycoside and b-lactam combination therapy versus b-lactam monotherapy on the emergence of antimicrobial resistance: a meta-analysis of randomized, controlled trials. Clin Infect Dis 2005; 41:149–58. Mizuta M, Linkin DR, Nachamkin I, et al. Identification of optimal combinations for empirical dual antimicrobial therapy of Pseudomonas aeruginosa infection: potential role of a combination antibiogram. Infect Control Hosp Epidemiol 2006; 27:413–5. Lodise TP Jr, Lomaestro B, Drusano GL. Piperacillin-tazobactam for Pseudomonas aeruginosa infection: clinical implications of an extendedinfusion dosing strategy. Clin Infect Dis 2007; 44:357–63. Bhat SV, Peleg AY, Lodise TP Jr, et al. Failure of current cefepime breakpoints to predict clinical outcomes of bacteremia caused by gramnegative organisms. Antimicrob Agents Chemother 2007; 51:4390–5. Capitano B, Nicolau DP, Potoski BA, et al. Meropenem administered as a prolonged infusion to treat serious gram-negative central nervous system infections. Pharmacotherapy 2004; 24:803–7. Li C, Du X, Kuti JL, Nicolau DP. Clinical pharmacodynamics of meropenem in patients with lower respiratory tract infections. Antimicrob Agents Chemother 2007; 51:1725–30. Lorente L, Lorenzo L, Martin MM, Jimenez A, Mora ML. Meropenem by continuous versus intermittent infusion in ventilator-associated pneumonia due to gram-negative bacilli. Ann Pharmacother 2006; 40: 219–23. Bouza E, Torres MV, Radice C, et al. Direct E-test (AB Biodisk) of respiratory samples improves antimicrobial use in ventilator-associated pneumonia. Clin Infect Dis 2007; 44:382–7. Mussap M, Molinari MP, Senno E, et al. New diagnostic tools for neonatal sepsis: the role of a real-time polymerase chain reaction for the early detection and identification of bacterial and fungal species in blood samples. J Chemother 2007; 19(Suppl 2):31–4. Li J, Nation RL, Turnidge JD, et al. Colistin: the re-emerging antibiotic for multidrug-resistant Gram-negative bacterial infections. Lancet Infect Dis 2006; 6:589–601. Bradford PA, Weaver-Sands DT, Petersen PJ. In vitro activity of tigecycline against isolates from patients enrolled in phase 3 clinical trials of treatment for complicated skin and skin-structure infections and

Treatment of Gram-Negative Bacillus Infection • CID 2008:47 (Suppl 1) • S19

complicated intra-abdominal infections. Clin Infect Dis 2005; 41(Suppl 5):S315–32. 34. Peleg AY, Potoski BA, Rea R, et al. Acinetobacter baumannii bloodstream infection while receiving tigecycline: a cautionary report. J Antimicrob Chemother 2007; 59:128–31. 35. Anthony KB, Fishman NO, Linkin DR, Gasink LB, Edelstein PH, Lautenbach E. Clinical and microbiological outcomes of serious infections

S20 • CID 2008:47 (Suppl 1) • Paterson

with multidrug-resistant gram-negative organisms treated with tigecycline. Clin Infect Dis 2008; 46:567–70. 36. Paterson DL. “Collateral damage” from cephalosporin or quinolone antibiotic therapy. Clin Infect Dis 2004; 38(Suppl 4):S341–5. 37. Chastre J, Wolff M, Fagon JY, et al. Comparison of 8 vs 15 days of antibiotic therapy for ventilator-associated pneumonia in adults: a randomized trial. JAMA 2003; 290:2588–98.