Approximately 444 subjects who are greater than or equal to 18 to 95 years of age, are non-pregnant, and are in the inpatient setting of one of the study sites will be evaluated to treatment efficacy. Analysis will include subjects with bloodstream infection (BSI) or pneumonia due to at least one of the following gram-negative bacilli organisms: Acinetobacter baumannii, Klebsiella spp, Escherichia coli, Enterobacter spp. and/or Pseudomonas aeruginosa that demonstrates in vitro non-susceptibility defined as extensively drug-resistant Gram-negative bacilli (XDR-GNB) which includes XDR-AB, XDR-PA and CRE. If a subject has both BSI and pneumonia at the time of study enrollment, they will be included as a subject with pneumonia.

Objectives:

Primary:

•Determine whether the treatment regimen of Colistimethate sodium (colistin) combined with a carbapenem (imipenem or meropenem) is associated with a decreased risk for mortality compared to colistin alone for subjects with bloodstream infection (BSI) and/or pneumonia due to XDR-GNB.

Secondary:

•Determine what treatment regimen (colistin monotherapy or colistin combined with a carbapenem (imipenem or meropenem) is more likely to reduce the emergence of colistin resistance among XDR-GNB isolates during therapy.

The Gram-negative bacilli organisms Acinetobacter baumannii, Klebsiella spp., Escherichia coli, Enterobacter spp. and Pseudomonas aeruginosa have become a frequent cause of bloodstream infection and pneumonia in the hospital and other healthcare settings. Among these pathogens, antimicrobial resistance has emerged to many classes of antimicrobial agents. Most concerning, has been the emergence of resistance to group 2 carbapenems (such as imipenem). In several regions of the world, including Southeastern Michigan, strains of extensively-drug resistant Gram-negative bacilli (XDR-GNB) that exhibit resistance to most, and in some cases all types of available antimicrobial agents, including group 2 carbapenems, have emerged and disseminated. Treatment options for XDR-GNB typically include Colistimethate sodium (referred to as colistin in this study), used alone (monotherapy) or in combination with other agents. Unfortunately, resistance to colistin has begun to emerge in some strains of XDR-GNB, which is a truly concerning development, since colistin is one of the last remaining treatment options for XDR-GNB. No prospective, randomized controlled trials have been conducted to evaluate the clinical efficacy of colistin monotherapy versus colistin-containing combination therapy or the impact of these therapeutic modalities on the emergence of colistin resistance among XDR-GNB. We plan to conduct a double-blind randomized controlled trial including patients with pneumonia and bloodstream infection due to XDR-GNB. After enrollment, subjects will be randomized to receive 14 days of either colistin monotherapy or colistin plus meropenem.

In the Detroit metro area, infections due to XDR-GNB have developed into a regional challenge and common problem. We have assembled a multi-disciplinary team that includes Infectious Diseases researchers, clinicians, infectious diseases pharmacists, microbiologists, epidemiologists and statistical experts to address critically important questions and challenges regarding the management of bloodstream infection and pneumonia due to XDR-GNB. Specifically, we hypothesize that the combination of colistin and imipenem will provide superior efficacy in the treatment of XDR-GNB pneumonia and bloodstream infection and will prevent the emergence of decreased susceptibility to colistin among XDR-GNB strains. We also aim to analyze tools that could be used in "real time" to aid clinicians treating patients with infection due to XDR-GNB. For example, we aim to analyze the association between the presence of in vitro synergy of the colistin and carbapenem (imipenem or meropenem) combination (as determined by E-test) and clinical outcomes; and the association between colistin plasma levels and clinical outcomes and the development of nephrotoxicity.

Due to the growing threat of XDR-GNB around the world, several international sites were subsequently added including sites in Asia, Israel, and Europe.

Inclusion Criteria:

• Hospitalized Adults (> 18 years to 95 years of age), at one of the study sites.

• Diagnosis of BSI and/or pneumonia due to a preliminary result of gram-negative non-lactose fermenter that is oxidase negative; or a final results of XDR-A. baumannii; carbapenem-resistant Enterobacteriaceae; or XDR- P. aeruginosa and/or patients with suspected BSI and/or HAP (hospital acquired pneumonia) and who have had a prior history (within last 6 months) of XDR-GNB that was susceptible to colistin.

  o If final results do not indicate that the pathogen is an XDR-GNB, and identifies alternative treatment options, the patient would be eligible for the study if the subject is allergic to all the alternative treatment options.

• Patients with polymicrobial respiratory or blood infections, including XDR-GNB and one or more pathogens, will be included in the study, as long as the XDR-GNB is determined to be a true pathogen (AB, CRE or PA). Other pathogens will be treated with antimicrobial agents as determined by the treating physician.

• If more than one XDR-GNB study pathogens is identified as a study pathogen causing BSI and/or pneumonia, then the first study pathogen recovered will be considered as the primary study pathogen. If more than one study pathogen is recovered from the same culture, then the infection will be categorized as being caused by multiple study pathogens.

• Patients with a life expectancy of > 24 hours

• Signed written informed consent and HIPAA Authorization form (US sites)

Exclusion Criteria:

• Female patients who are pregnant
• Female patients who are nursing
• Patients who are prisoners
• Patients who are less than 18 years of age or greater than or equal to 96 years of age
• Patients with neutropenia (WBC < 500 cells/mm3)
• The presence of any of the following known clinical syndromes involving XDR-GNB as a pathogen which necessitate durations of antimicrobial therapies greater than 14 days: endocarditis, osteomyelitis, prosthetic joint infections, meningitis and/or other central nervous system infections.
• Patients receiving valproic acid (with or without a known seizure disorder).
• Patients who received 72 hours or more of polymyxin treatment (intravenous or inhaled [pneumonia]) within 96 hours of enrollment.
• Patients who have end-stage renal disease requiring hemodialysis, will be excluded from evaluation pertaining to nephrotoxicity in the per protocol population.
• Patients with known Type 1 or other severe drug allergy to either of the study drugs or to β-lactams.