Antibiotic Strategies in the Era of Multidrug Resistance

George Karam; Jean Chastre; Mark H. Wilcox; Jean-Louis Vincent


Crit Care. 2016;20(136) 

In This Article


There is a general increase in the number of resistant microorganisms worldwide, although specific patterns vary considerably across countries. There has been a significant increase across Europe in the percentages of Klebsiella pneumoniae resistant to fluoroquinolones, third-generation cephalosporins, and aminoglycosides, as well as combined resistance to all three antibiotic groups.[1]Escherichia coli resistance to third-generation cephalosporins has also increased significantly, from 9.6 % to 12.0 % between 2011 and 2014 (population-weighted European Union/European Economic Area (EU/EEA) mean percentage of resistance).[1] For Acinetobacter species, there is considerable variability in resistance rates, but high percentages (>50 %) of isolates with combined resistance to fluoroquinolones, aminoglycosides, and carbapenems have been reported from southern Europe (Fig. 1). Although the percentage of methicillin-resistant Staphylococcus aureus (MRSA) decreased between 2011 and 2014, this decrease was less pronounced compared with the previous 4-year period. In 2014, the EU/EEA population-weighted mean MRSA percentage remained high, with seven out of 29 reporting countries having MRSA percentages >25 %.

Figure 1.

Acinetobacter species: percentage of invasive isolates with combined resistance to fluoroquinolones, aminoglycosides, and carbapenems. European Union/European Economic Area, 2014. From [1]

The increased prevalence of carbapenem-resistant Enterobacteriaceae (CRE), particularly in K. pneumoniae which has seen near untreatable infections occurring in an increasing number of hospitals, is also of concern. Greece, Italy, and Malta in Europe, the USA, South America, and Asia have notably been affected by these bacteria.[2,3] Such is the level of threat that the US Centers for Disease Control and Prevention has named CRE as one of the top three most urgent antimicrobial-resistant challenges.[4]

Mechanisms of Resistance

Resistance can occur in all types of pathogens encountered in the ICU setting, although Gram-negative bacteria are the most likely to exhibit resistance to multiple classes of antibiotics. The three most representative mechanisms of resistance to β-lactam antibiotics in Gram-negative bacteria are: destruction of antibiotics by β-lactamases; impermeability, including closure of porin channels in the bacterial cell wall (most notable as the mechanism of resistance to carbapenems for Pseudomonas aeruginosa); and extrusion of antibiotics by efflux pumps (which can lead to resistance to multiple classes of antibiotics). Analogous mechanisms of resistance occur with classes of antibiotics that are increasingly being used to manage infections due to bacteria resistant to β-lactam antibiotics (Table 1). By contrast to β-lactam antibiotics, which have their mechanisms of action and resistance located within the cell wall of the bacteria, the location of binding sites and modifying enzymes of the other antibiotic classes described in Table 1 are intracellular. Knowledge of this variability of action and resistance mechanisms can contribute to informed decisions in the selection of antimicrobial therapy for resistant organisms.

Extended-spectrum β-lactamases (ESBLs) are broad-spectrum enzymes produced most characteristically by E. coli, Klebsiella, and Proteus species (Table 2). Representative of the ease with which resistance can occur, these enzymes may develop on the basis of a change in only one amino acid in the β-lactamases normally produced.[5] Despite the minimal structural change, ESBLs have the capacity to inactivate many broad-spectrum β-lactam drugs. It is noteworthy that use of third-generation cephalosporins and fluoroquinolones has been identified as a risk factor for selection of ESBLs.[6,7] Of the clinically relevant attributes of these enzymes, ESBLs represent a classic example of a resistance mechanism in which in vitro susceptibility may not be consistently predictive of clinical efficacy.

By contrast to the plasmid-mediated production of ESBLs, AmpC β-lactamases most classically are chromosomally-mediated and occur in such important ICU pathogens as P. aeruginosa and Enterobacter species. In recent years, however, plasmid-mediated AmpC β-lactamases have been identified in pathogens such as E. coli and K. pneumoniae. Influencing the risk that antibiotics pose for the selection of infection caused by resistant pathogens is the fact that one in every 106–107 organisms with the characteristic potential for AmpC β-lactamase production (listed in Table 2) has a spontaneous mutation that allows it to produce this enzyme.[8] These mutant strains are not naturally competitive and do not, therefore, overgrow to destroy the sensitive nonmutant flora. However, when antibiotics are given that destroy the sensitive flora, the resistant mutant strains can proliferate and establish themselves as the predominant pathogens. In such a scenario, injudicious use of broad-spectrum agents may lead to the development of clinical resistance during therapy.

There has been considerable recent publicity about a study from China that has described the first strains of E. coli with plasmid-mediated colistin resistance.[9] Although there are concerns about the use of colistin, including toxicity and dosing uncertainties, it is one of a very few alternatives that can be used in some cases to treat infections caused by CRE. This novel mechanism of colistin resistance, previously only chromosomally mediated, is of grave concern given that strains harboring the plasmids are already widely prevalent in animals in China, and also (albeit lesser so) in some clinical isolates.

Carbapenemase production by Gram-negative bacteria is one of the most concerning patterns of resistance encountered in the ICU because it is associated increasingly with resistance to all presently marketed antibiotics. Using the Ambler classification of β-lactamases—in which there are Classes A, B, C, and D—carbapenemases occur within three of the four classes. Class B β-lactamases have a metallo base, and the initial carbapenemases described clinically were Class B metallo-enzymes. A representative example from clinical practice of a Class B metallo-enzyme is the New Delhi metallo-β-lactamase, which is found in certain Enterobacteriaceae. Ambler Class A β-lactamases are serine based and include the majority of ESBLs. Carbapenemases produced by K. pneumoniae were identified within this Ambler class and subsequently characterized as K. pneumoniae carbapenemases (KPCs). Gaining increasing importance today are the Class D serine-based carbapenemases, with the most classic being the oxa-type enzymes produced by organisms such as Acinetobacter species. Even though the name carbapenemase identifies the ability of these enzymes to inactivate carbapenem antibiotics, carbapenemases are not specific for carbapenem antibiotics but have the ability to hydrolyze β-lactams of all classes.[10]