Multidrug‐resistant Acinetobacter: a threat to the antibiotic era

Hospital-acquired infections caused by Acinetobacter have now been reported worldwide.1 The organism is a highly resilient, Gram-negative coccobacillus and has emerged as one of the most difficult bacteria to control and treat.2 The range of infections associated with Acinetobacter is wide and includes pneumonia, bacteraemia, urinary tract infection, skin and wound infection, meningitis and endocarditis, with the distribution of these dependent on the reporting institution.2 The overall success of Acinetobacter can be attributed to two ‘secret weapons’– adaptation to the hospital environment and antibiotic resistance. Acinetobacter is ubiquitous in the environment and has therefore acquired the necessary skills for survival under harsh conditions. Such skills are used in the hospital setting, where the organism can survive for long periods in both wet and dry surroundings.2–4 The described reservoirs of the organism are extensive.5 As an example, Woods et al. in Brisbane found substantial rates of colonization of intensive care unit (ICU) computer keyboards.6 Also, survival on hospital bed rails for up to 9 days after the discharge of an infected patient has been reported.7 The ability for such persistence provides the organism with an excellent opportunity for cross-transmission within an institution, either through the hands of health-care workers or through hospital equipment.8–10 Also, as a result of such widespread habitation, the inability to identify a point source of infection may make control extremely difficult.5 In addition to its adaptation to the hospital environment, the organism has a renowned ability to rapidly acquire resistance to antimicrobials. This provides a selection advantage in environments where large volumes of antibiotics are being prescribed, such as the ICU. This can also create a challenging therapeutic experience for the clinician, as antimicrobial options become greatly limited. In a recent review of Acinetobacter outbreaks from 1977–2000, a stepwise trend in the reporting of outbreaks with increasingly resistant organisms was noted.5 Initial reports, mostly during the 1980s, focused on aminoglycoside resistance. This was soon followed by a larger number of outbreaks reporting resistance to multiple antibiotic classes (multidrug resistance), including third- and fourth-generation cephalosporins, quinolones and aminoglycosides.5 As a result, carbapenems became the agents of choice for serious Acinetobacter infections, inevitably leading to the most recent outbreaks involving carbapenem-resistant organisms.5,8,11 This path of antibiotic resistance accumulation is characteristic of Acinetobacter and is commonly observed within institutions over time.12 One of the first descriptions of highly resistant Acinetobacter (ceftazidime and imipenem resistant) from New York City is a prime example.8 Initially, unrestricted use of ceftazidime for Acinetobacter infections led to both ceftazidime-resistant Acinetobacter and Klebsiella. In an attempt to treat such organisms, the use of imipenem increased and subsequently led to an outbreak of Acinetobacter that was resistant to all routinely tested antibiotics, including imipenem.8 Further testing showed susceptibility to polymyxin B and sulbactam only. The evolution of this outbreak not only illustrates the ‘genetic agility’ of Acinetobacter, but also the ‘balloon’ theory of antibiotic resistance, whereby limiting or ‘squeezing’ the use of one antibiotic may be counteracted by the equivalent increase or ‘bulging’ in another, resulting in unforeseen consequences in the development of resistance.13 The New York outbreak also highlights the importance of infection control for Acinetobacter. All isolates were genetically related, indicating cross-transmission or a point source. Fortunately, samples taken from laryngoscopes and the hands of health-care workers were of the same strain, allowing a focused, multifaceted infection control effort to terminate the outbreak.12 Despite this, multidrug-resistant Acinetobacter had already disseminated widely throughout other hospitals in New York, with a single clone making up the majority of isolates.14 City-wide dissemination of a multidrug- or a pan-resistant clone is of great concern and stresses the importance of carrying out molecular epidemiological analysis in the assessment of Acinetobacter outbreaks, thereby allowing knowledge of the local epidemiology of this problematic organism.15,16 It also illustrates the issue of hospital-to-hospital transfer of resistant bacteria. Already, cases of inter hospital transfer of multiply resistant Acinetobacter have been observed in Australia following transfer of patients from overseas,17 as was exemplified following the Bali terrorist bombing.18 Initial reports of Acinetobacter from Australia came from the Northern Territory, where bacteraemic, community-acquired pneumonia is well described.19 In this scenario, isolates were susceptible to aminoglycosides, although resistance to cefotaxime was common. Inappropriate initial antibiotic therapy was strongly associated with a fatal outcome. Risk factors, such as male sex, age >45 years, Aboriginal ethnic background, cigarette smoking, alcoholism, diabetes mellitus and chronic obstructive airway disease were identified and were thought to contribute to the high mortality (64%).19 The first-described Australian outbreak of hospital-acquired Acinetobacter was from Western Australia.20 Over a 2-year period, 45 cases of gentamicin-resistant Acinetobacter from an ICU were identified. All isolates were also resistant to cephalosporins and ticarcillin, with five isolates being resistant to ciprofloxacin as well. All were sensitive to imipenem. Pneumonia was the most common infection type and no deaths were attributable to Acinetobacter. Of particular interest was the fact that 11% of staff hand samples were positive for the same strain of Acinetobacter, as determined by molecular epidemiological analysis. Despite the best infection control intentions, a state of endemicity ensued. Unfortunately, such a scenario is more often the norm rather than the exception. A more positive outcome occurred after a smaller outbreak in Geelong, Victoria, Australia.21 This involved 10 patients, 9 of whom were carrying isolates resistant to all antibiotics except meropenem, tobramycin and amikacin, and one isolate that had a different susceptibility profile, including resistance to meropenem. A possible breach in infection control precautions with respiratory equipment in the ICU was found and after a thorough, multifaceted intervention period, including the closure of a surgical ward, the outbreak was successfully terminated. Such prompt action and appropriate use of both clinical and molecular epidemiological techniques should be commended. Unfortunately, outbreaks of carbapenem-resistant Acinetobacter have been described in tertiary referral centres throughout Australia. In such an outbreak, from Westmead Hospital, Sydney, Australia, infection with carbapenem-resistant Acinetobacter was independently associated with increased hospital mortality and prolonged hospital and length of ICU stay.22 A recent outbreak in Melbourne is also illustrative of this emerging problem.11 Sixty-nine patients developed Acinetobacter bloodstream infections in a single institution from September 2002 to April 2004. The significant trends in the emergence of resistance over time were particularly worrying. Initial rates of meropenem resistance were less than 20%, but rose substantially to 95% during the period from January to April 2004. This increase in resistance was temporally related to a rapid increase in meropenem use at this institution. Overall, 64% of blood culture isolates were resistant to meropenem and 10% were non-susceptible to all tested antimicrobials. Of these, three isolates were tested against colistin and were found to be susceptible. With these data in mind, how should Acinetobacter resistant to carbapenems and other first-line antibiotics be treated? Colistin, a member of the polymyxin class of antibiotics, has been available for over 50 years. Its use was abandoned in the early 1980s because of unacceptable rates of nephrotoxicity compared with newer antimicrobial options. However, use of colistin has undergone a necessary revival as a result of the dearth of agents available for treating highly resistant Gram-negative bacilli, including Acinetobacter.23 Unfortunately, the optimal dosing regimen of colistin is unknown, because its pharmacokinetics were never studied in detail when the drug was developed. This is of particular concern in seriously ill patients, such as those with acute renal failure on renal replacement therapy.24 Nevertheless, reasonable outcomes have been reported when colistin is used for multiply resistant Acinetobacter infections.23 Furthermore, recent studies suggest that toxicity may not be as profound as was observed in the early decades of its use.25 Combination therapy of colistin plus rifampicin has shown favourable results in preliminary studies26,27 and may provide a more effective option for management of these challenging infections. Nebulized colistin has also been investigated, but studies of this treatment method are limited28 and we would only recommend such therapy in combination with an active systemic agent. The numerous uncertainties of colistin dosing and potential usefulness in combination therapy should act as a stimulus for further research. Of concern, and in keeping with the behaviour of this organism, rates of resistance to colistin have recently been reported as high as 58% in multidrug-resistant Acinetobacter in Israel.29 Although, in a larger study of 2621 Acinetobacter isolates from four major geographic regions, including Australia, rates of resistance of less than 5% were reported.30 Li and colleagues from Melbourne have recently shown heteroresistance to colistin in an Acinetobacter isolate.31 Other, non-traditional antimicrobials have shown promising results in treating multidrug-resistant Acinetobacter. First, sulbactam, a β-lactamase inhibitor, has intrinsic activity against such isolates, with clinical studies showing efficacy if the isolate is susceptible.32 Unfortunately, sulbactam resistance is common in certain areas and therefore susceptibility testing is required before use. A second potential agent is trimethoprim-sulphamethoxazole, but as with sulbactam, rates of resistance in multidrug-resistant isolates can be high and therefore, susceptibility results are also required before use.16 Finally, tetracycline derivative antibiotics, most importantly doxycycline, minocycline and the new agent, tigecycline, have all shown activity against Acinetobacter. Tigecycline has recently been licensed by the US Food and Drug Administration for the treatment of complicated skin, soft tissue and intra-abdominal infection. The agent has activity against Acinetobacter, but as a result of its pharmacokinetic characteristics (rapid movement from the bloodstream to the tissues) its use for Acinetobacter bloodstream infection is questionable. As with other tetracyclines, tigecycline is bacteriostatic for Acinetobacter. In conclusion, multidrug-resistant Acinetobacter has now emerged as an impressive organism, with unique skills of adaptation that will allow it to persist as a problem well into the future. The epidemiology of this organism is complex and emphasizes the importance of use of molecular tools in outbreak investigation (Table 1). The challenges of antimicrobial treatment are mounting and have led clinicians to seek older or non-traditional agents in an attempt to maximize patient outcomes. Our ability to differentiate colonization from infection is of critical importance, especially for respiratory samples, as inappropriate antibiotic use will only add fuel to an already raging fire. Future research should be targeted towards the most effective strategies to prevent the spread of multidrug-resistant Acinetobacter, as well as, new or novel therapeutics for the treatment of highly resistant Gram-negative bacilli, including Acinetobacter.