The enemy within hospital-acquired, antibioticresistant bacteria Peter Hawkey

It is perhaps ironic that the recognition of the aetiology of hospital-acquired (nosocomial) infection was by the surgeon and self-taught microbiologist, Joseph Lister. In 1865, at the Edinburgh Royal Infirmary, the mortality following amputation of a limb was 43 %, the principal cause of death being bacterial sepsis. Mortality was reduced to single figures by the introduction of the antiseptic phenol spray which bathed the operative site and the use of post-operative dressings of phenol. Such was the impact of this medical breakthrough that he was the first member of the medical profession to be elevated to the Peerage as Lord Lister of Lyme. Quite quickly, antiseptic surgery passed out of fashion to be replaced by aseptic surgery, championed in the UK by surgeons such as Lawson Tait. The subsequent introduction of effective antimicrobial therapy further dramatically reduced the incidence and importance of nosocomial infection. q The rise of nosocomial infections The emergence in the late 1950s of strains of Staphylococcus aureus resistant to penicillin, erythromycin, tetracycline and chloramphenicol, which also were good at surviving and spreading in hospitals, such as phage type 80/81, led to a sharp rise in morbidity and mortality.

This was effectively terminated when the isoxyl penicillins, resistant to penicillinase, such as cloxacillin, were introduced. The subsequent development of both intensive care and treatment for cancers in the 1960s and 1970s created a new group of vulnerable patients prone to infection caused by Gram-negative bacteria such as Pseudomonas aeruginosa, Klebsiella pneumoniae and Serratia marcesens (Fig. 1). The heavy use of gentamicin in turn drove the emergence of resistance mediated by aminoglycoside-inactivating enzymes, which by virtue of being located on transposons, integrons and plasmids resulted in epidemics of resistance genes spreading amongst different species and genera. The availability for the first time of molecular techniques for plasmid isolation and visualization, restriction endonuclease fingerprinting/ mapping of plasmids and Southern blotting with DNA probes enabled molecular epidemiology studies to be undertaken to identify antibiotic resistance gene flow. This problem of nosocomial Gram-negative bacterial infections in hospitals was countered by the recognition that the bacteria were surviving as transient flora on the hands of medical and nursing staff. The development and application of handwashing using either aqueous or alcoholic disinfectant preparations (those containing chlorhexidine or triclosan being favoured) were instrumental in controlling the problem. In addition, the early 1980s saw the introduction of extended spectrum cephalosporins, such as cefotaxime and ceftazidime, which had good activity against these multi-resistant nosocomial Gram-negative bacteria. It is perhaps far from coincidental that with rapidly increasing use of these antibiotics in hospitals (they were only available parenterally) the incidence of infection caused by methicillin-resistant S. aureus (MRSA) rose progressively in most countries with a developed or semi-developed hospital healthcare system. q MRSA The rise of MRSA in the last decade and a half has been spectacular and demonstrates the threat from antibioticresistant bacteria. There has also been a similar but smaller increase in infections caused by methicillinsensitive S. aureus (MSSA), primary blood stream nosocomial infections increasing in the USA twofold between 1980 and 1989. Also, in the USA in 1981, 5% of all isolates of S. aureus at a large teaching hospital were MRSA; by 1991 38% were methicillin-resistant. Methicillin resistance is conferred on strains by the possession of a chromosomal copy of modified penicillin-binding protein, PBP2a. This protein has a low affinity for most -lactam antibiotics (including isoxyl penicillins such as flucloxacillin and nafcillin) and therefore mediates cross-resistance to all such compounds. The mecA and associated genes are not frequently found to be transferred in vivo, although mecA

The battle against the rise of hospitalacquired infections caused by antibiotic-resistant micro-organisms is one of the major challenges faced by microbiologists today. Peter Hawkey takes a look at the problem and discusses how it may be tackled in the future.

LEFT: Fig. 1. Typical intensive care unit patient. A portal of entry for nosocomial bacteria is the endotracheal tube providing the artificial ventilation (tracheal secretions are being removed by suction, another potential route for infection). Intra-vascular lines providing hydration, nutrition and a route for drug administration are the next most important portal of entry. The temperature (Bennett) probe is visible at the Y piece of the ventilator tubing.

MICROBIOLOGYTODA YVOL28/FEB01

is carried on a transposon. It seems likely that the gene has transferred into several clonal lines of S. aureus, although some hold the view that all of the clones seen today have descended from a common ancestor. Different countries and geographical areas have different clones (recognizable by the digestion of genomic DNA using the restriction endonuclease SmaI and display of the large fragments by pulse field gel electrophoresis). At the moment the most successful clones seen in the UK are EMRSA (Epidemic MRSA) 15 and 16. In contrast, the clones seen in France are quite different, although some other clones have spread across continents (e.g. South America and Europe). The reasons for this are not altogether clear, but undoubtedly the ability of different strains to survive both at carriage sites and in the environment, as well as their ability to colonize new patients is important. There is no evidence that MRSA strains are per se more pathogenic than MSSA. Most surveys show that >50 % of MRSA are also resistant to aminoglycosides, fluoroquinolones, macrolides and lincosamides. This multiple resistance undoubtedly allows cross-selection by antimicrobial agents other than isoxyl penicillins. Delays in therapy also expand the number of patients acting as reservoirs. Therapy invariably involves vancomycin, heavy usage of which has probably led to the pre-eminence of vancomycin-resistant enterococci (VRE). S. aureus is probably the most common cause of nosocomial infections and represents a huge challenge to the medical and scientific community. A recent study has shown that primary nosocomial blood stream infections result in an approximate threefold increase in direct cost (mean cost US$27,000) compared with those due to MSSA (US$9,600 ). q Problems with Gram-negative bacteria As if these problems with Gram-positive pathogens were not enough, the heavy usage of extended spectrum cephalosporins has resulted in the selection of an interesting and growing problem caused by Gramnegative bacteria carrying the so-called extended spectrum -lactamases (ESBLs). These arise by mutation and alteration of one or two amino acids in TEM and SHV plasmid-mediated -lactamases. The extended spectrum cephalosporins had been specifically designed to be stable to degradation by the unmutated forms of these enzymes. The amino acid substitutions allow the extended spectrum cephalosporins to gain access to the active site of the enzyme as well as enhancing the rate of hydrolysis of the substrate. These enzymes are generally recognized by the fact that they are inhibited by conventional plasmid-mediated -lactamase inhibitors, such as clavulanic acid. The comparative ease with which mutation can lead to resistance and the evolution of ESBLs has been shown in a number of studies with

examples of identical genotypes arising by convergent evolution from different wild type genes. There are currently some 80 or so types of TEM gene and nearly 30 SHV genes researched (see http://www.lahey.org/ studies/webt.stm). In an effort to treat hospitalized patients with these infections caused by multi-resistant Gram-negative bacilli, many physicians have increased usage of carbapenem antibiotics. Inevitably, the bacterial population demonstrates a resilient response and resistance has developed via two unpredicted routes. By overexpressing certain types of -lactamase normally encoded on the chromosome, but now mobilized onto plasmids (e.g. CMY), low levels of resistance to drugs such as imipenem are seen. Although these levels of -lactamase are too low to allow clinically significant resistance, when combined with the loss of an outer-membrane protein (porin), clinically significant levels of resistance are seen, resulting in treatment failure. More sinister has been the appearance in widely dispersed locations around the globe in the last year or so of zinc-dependent -lactamases (IMP24, VIM1 and 2) related (by nucleic acid sequence) to the highly transmissible enzyme IMP-1 first seen in Japan more than 10 years ago, but which has not spread outside of that country. These genes are often plasmid-associated and capable of moving from one replica onto another because of carriage as integrons. q Intensive care units (ICUs) Modern medical techniques have produced patients highly vulnerable to nosocomial infection who are managed in areas of the hospital where antibiotic usage is particularly high. Intensive care unit patients are particularly susceptible to infection with antibioticresistant bacteria by virtue of invasive procedures such as artificial ventilation and intravascular lines (Fig. 1). Some bacteria, such as Pseudomonas spp. and Acinetobacter baumannii, have become highly significant pathogens in this group. Treatment can be extremely difficult because they often carry a huge array of resistance mechanisms. Studies such as the EPIC study (a one day in 1992 point prevalue study of infections in ICUs in different European countries, including nearly 1500 units) showed that ventilator-associated pneumonia (VAP) accounted for nearly half of the ICU infections. q Combating the problem The microbiologist can help reduce this burden of infection in several ways. The development of rapid molecular typing methods based on PCR has brought the prospect of real-time investigation of environmental and patient sources of infection, and of routes of transmission. An example of such a successful investigation is shown in Fig. 2 and Table 1. Improvements in both the accuracy and speed of anti-

MICROBIOLOGYTODA VOL28/FEB01 Y

microbial susceptibility testing may in the future bring the prospect of being able to use antimicrobials more precisely. The clinical input from microbiologists is also crucial in maintaining appropriate levels of antimicrobial usage in all areas of the hospital. Microbiology laboratories, by processing routine specimens and performing antimicrobial sensitivity testing, form the front line of surveillance for antibiotic resistance and therefore are the intelligence service in the battle against nosocomial infection.

10

11

12

q What does the future hold? q Increasing mobility of the world population and the opportunity for the faster spread of resistance genes q An ageing population with increased expectations of complex health care q Better environments for antibiotic-resistant nosocomial bacteria to flourish in q Increasing addiction to antibiotics by both patients and physicians particularly in some developing industrial societies q An apparent increase in the rate of emergence of novel antibiotic resistance mechanisms
Lane Specimen*

LEFT: Fig. 2. Repetitive element primer (REP)-PCR profiles of Acinetobacter spp. strains isolated from an outbreak of cross-infection in an intensive care unit. Lanes 16 are patient infections and 712 are environmental isolates. Table 1 shows the source of the isolates. An inadequate disinfection regime for the Bennett temperature probes was shown to be the cause. Improving the disinfection procedure eliminated the outbreak.

Perhaps it does not look hopeful, but with the necessary political and professional will the battle against nosocomial infection cannot just be joined but even turned in the favour of mankind. If we lose it, the consequences for health care will be serious indeed. q Peter Hawkey BSc MB BS MD FRCPath is Professor of Medical Microbiology, Division of Microbiology, School of Biochemistry and Molecular Biology, University of Leeds, Leeds LS2 9JT. Tel. 0113 233 5597; Fax 0113 233 5649; email p.m.hawkey@ leeds.ac.uk
Date of isolation Site REP-PCR profile

Table 1. Sources and properties of the Acinetobacter spp. causing an ICU outbreak, shown in Fig. 2

1 2 3 4 5 6 7 8 9 10 11 12

Acinetobacter baumannii Acinetobacter baumannii Acinetobacter baumannii Acinetobacter baumannii Acinetobacter baumannii Acinetobacter haemolyticus Acinetobacter baumannii Acinetobacter baumannii Acinetobacter baumannii Acinetobacter baumannii Acinetobacter lwoffii Acinetobacter baumannii

1/5/93 1/7/93 1/8/93 1/9/93 12/31/92 12/20/92 1/12/93 1/12/93 12/6/92 12/8/92 12/20/92 1/22/93

Sputum Tracheal secretion Sputum Sputum Leg wound Blood culture Bennett temp probe Sluice sink Dust Dust Blood culture Bennett temp. probe

1 1 2 1 1 3 1 4 5 6 7 1

*Detected by randomly amplified polymorphic DNA (RAPD) analysis.

MICROBIOLOGYTODA YVOL28/FEB01