Acinetobacter baumannii: A newly emerging nosocomial pathogen
Carl Urban and James J Rahal
Department of Medicine, The New York Hospital Medical Center of Queens, Queens, New York, USA and Cornell University
Medical College, Ithaca, New York, USA
Certain strains of Acinetobacter baumannii (formerly Acinetobacter
calcoaceticus biotype anitratus) have acquired resistance
to all commonly prescribed antibiotics. Only within the last three years did these strains become resistant to
the beta-lactam of last resort, imipenem (7,8,10). Resistance to imipenem in Acinetobacter can be due to enzymes
that inactivate the antibiotic (5), altered penicillin-binding proteins (PBPs)(2) or a combination of altered PBPs
and reduced outer membrane permeability (3). Only polymyxin, and in some instances sulbactam, have demonstrated
bactericidal activity against these multiresistant bacteria. This report reviews events leading to a nosocomial
outbreak of Acinetobacter baumannii
strains susceptible only to polymyxin and sulbactam in a 487 bed teaching hospital. It will present epidemiological
characteristics of the outbreak and discuss activities of key personnel which led to eradication of this strain
from the hospital environment.
During the mid to late 1980's, two populations of Acinetobacter baumannii
with distinct antibiograms were observed in our institution. Multiresistant isolates were susceptible to ceftazidime,
amikacin and imipenem while sensitive strains were also susceptible to tobramycin, trimethoprim/sulfamethoxazole,
gentamicin and ticarcillin/clavulanate. Ceftazidime was selected as the major beta lactam agent of choice for treatment
of patients infected with multiresistant A. baumannii isolates. The necessary use of ceftazidime from 1987 to 1989 led to an epidemic
of ceftazidime-resistant Klebsiella pneumoniae. Detailed in vitro experiments showed that imipenem was the only beta-lactam
agent consistently bactericidal for the organisms tested (4 ). In addition, the genes encoding ceftazidime resistance
were located on plasmids which medicated resistance to many beta-lactam agents as well as other classes of antimicrobials.
Thus, imipenem was chosen as the therapeutic agent for patients infected with ceftazidime resistant Klebsiella pneumoniae.
As imipenem usage increased in 1990, our multiresistant Acinetobacter soon became resistant to imipenem.
In September of 1991 our first isolates of imipenem resistant Acinetobacter baumannii
were identified in the clinical microbiology laboratory. These isolates were resistant to all antibiotics except
polymyxin and the beta-lactamase inhibitor, sulbactam. When we first reported these multiresistant isolates in
1993, imipenem resistance was a rare finding in Acinetobacter (8). We also demonstrated that sulbactam, a beta-lactamase
inhibitor was bactericidal in vitro and was effective therapeutically for patients infected with these Acinetobacter
strains. A subsequent report on the mechanisms of action of sulbactam showed that it bound to penicillin-binding
proteins (PBPs) 2 and 1 at clinically achievable levels in both imipenem-susceptible and resistant strains (9).
While these experiments were being performed in the laboratory, investigations into the clinical and molecular
epidemiology of Acinetobacter infections susceptible only to polymyxin B and sulbactam were initiated (1). Four
major antibiograms were associated with isolates of A. baumannii (ACB). Imipenem resistant (IR-ACR) isolates were resistant by Kirby Bauer
susceptibility testing to imipenem, amikacin, gentamicin, tobramycin, ceftazidime, trimethoprim/sulfamethoxazole,
ciprofloxacin, tetracycline and ticarcillin/clavulanate. Imipenem-sensitive (IS-ACB) isolates were susceptible
only to imipenem. Imipenem and amikacin sensitive (IAS-ACB) isolates were susceptible only to imipenem and amikacin.
Very sensitive (VS-ACB) isolates were susceptible to imipenem, amikacin, gentamicin, tobramycin, ciprofloxacin,
trimethoprim/sulfamethoxazole and ticarcillin/clavulanate. ACB isolates were monitored from the clinical microbiology
laboratory and patient records were reviewed over a twelve month period. Patients infected as defined by the CDC
with IR-ACB were also recorded. Isolation of IR-ACB from different sites in one patient was identified as one case
and isolation of more than one ACB with identical susceptibility patterns from a single site was recorded as once
per site. Environmental surveillance was conducted to determine inanimate and personnel reservoirs from December
1991 to January 1992 which corresponded to the peak of the outbreak. Sterile swabs from bed rails, counter tops,
faucets, sinks, vents, monitor keyboards, door knobs, charts, laryngoscopes, blades and handles and intravenous
drip supports were placed into 5 milliliters of sterile brain heat infusion broth. Hand cultures were randomly
performed on hospital personnel. Individuals were instructed to wash their hands in 100 milliliters of trypticase
soy broth in a large plastic bag. Media from environmental and hospital personnel cultures were incubated at 37°C
for 24 hours and subcultured onto MacConkey agar.
Organisms that grew on the plates were identified using standard
microbiological media and the Vitek System (Hazelwood, MO). Bacteria identified as ACB were subjected to Kirby-Bauer
disk diffusion methods and resulting antibiograms were recorded. ACB strains isolated were categorized according
to antibiogram, and representative strains were subjected to both Hind III and Xba restriction endonuclease digestion
to determine specific genetic patterns. Results from these investigations revealed that 72 isolates of IR-ACB were
obtained from 59 patients during an outbreak period from September 1991 to September 1992. Restriction patterns
demonstrated that these isolates were derivatives of a single strain. Thirty-six patients were male and 23 female
with ages ranging from 16-93. Thirty-nine or 54% of IR-ACB were isolated from the respiratory tract and another
12 (16.6%) isolated from skin or surgical wound infection. The remaining isolates were recovered from intra-operative
biliary drainage, blood, peritoneal fluid, pleural fluid and urinary tract specimens. All patient had received
prior intravenous antibiotics in the hospital and had been on mechanical ventilatory support. Eighteen of 59 patients
were considered infected with IR-ACB. The majority of respiratory isolates were recovered from patients in the
surgical intensive care unit (SICU). A multitude of inanimate objects including tables, beds, monitors, charts,
ventilators and intravenous drip supports in the SICU, were positive for IR-ACB, IS-ACB or IAS-ACB. Hand cultures
of personnel were positive in the SICU, while those of personnel in the medical intensive care unit (MICU) or surgical
progressive care unit were negative. Laryngoscopes used in the SICU, and hands of respiratory therapists also yielded
multiresistant Acinetobacter (IR-ACB, IS-ACB or IAS-ACB). Restriction endonucleases digestion of ACB suggested
that isolates with these three antibiotic susceptibility patterns had the same genetic pattern, while VS-ACB showed
a completely different genetic profile.
Increased selection pressure imposed by imipenem for the treatment
of patients infected with ceftazidime-resistant Klebsiella
pneumoniae resulted in evolution to imipenem resistance
in a previously multiresistant bacterium. The identical digestion patterns of IS-ACB and IAS-ACB suggest that isolates
with these two antibiotype patterns were derived from a single clone which evolved to the totally resistant IR-ACB,
also possessing the same genetic pattern. It is likely that our restriction endonuclease digestion experiments
were not sufficiently sensitive to detect mutation(s) leading to imipenem resistance. We also suspect that imipenem
resistance occurred by a stepwise chromosomally mediated mechanism. The absence of any beta-lactamases which hydrolyze
imipenem in our strains (K. Bush and P. Bradford, personal communication), and relatively low MIC of imipenem (16
uglm) in resistant isolates support this hypothesis.
The outbreak of IR-ACB probably originated in the SICU as demonstrated
by the recovery of this strain from both personnel and environmental cultures. Contamination of mechanical ventilatory
equipment was not implicated as an intermediate vector for spreading organisms in this investigation. However,
laryngoscopes used only in the SICU were found to harbor resistant IR-ACB. Infection control efforts were principally
directed towards the SICU including monitoring and enforcing handwashing techniques, glove changing and use of
polymyxin B for wound irrigation. The SICU was also freshly painted and all inanimate objects within the unit were
cleaned with a disinfectant solution. Since the nosocomial outbreak was recognized early in its development and
concentrated in the SICU, the coordinated effects of the Infectious Disease Section, its Research Laboratory, Infection
Control, and Pharmacy were able to eradicate the IR-ACB from the hospital. Although sensitive strains of Acinetobacter
remain, IR-ACB has not been isolated from patients or from the hospital environment since October, 1992.
Following the description of our outbreak, two other reports of imipenem-resistant episodes have been reported
(7,10). The first occurred in 1990 with five patients having imipenem and sulbactam resistant isolates identified
(10). Three of these patients were considered colonized; and two were treated with colistimethate (polymyxin E).
We were fortunate in that sulbactam still retained bactericidal activity against our isolates and was used therapeutically
(ampicillin/sulbactam). We have subsequently identified sulbactam resistant Acinetobacter isolates that remain
susceptible to other antibiotics. These possess a different restriction pattern from that of our previously described
imipenem-resistant isolates (unpublished results). A more recent report describes an outbreak of imipenem resistant
Acinetobacter which occurred in France from February 1991 to March 1992(7). The epidemiological characteristics
were similar to ours since they demonstrated both environmental and hospital personnel as contributory factors
in the dissemination of these organisms. The use of antibiotyping, biochemical biotyping and pulse-field gel electrophoresis
resolved their imipenem resistant isolates into two genotypes (7).
An abstract presented at the 1994 IDSA meetings (Orlando, Fla.)
has demonstrated the successful treatment of a patient with meningitis and ventriculitis with polymyxin B caused
by an initially reported imipenem susceptible isolate of Acinetobacter that failed therapy (6).
In addition, we have had personal communications with two institutions
in New York City which have experienced outbreaks of imipenem resistant Acinetobacter that were susceptible to
polymyxin B and sulbactam. Acinetobacter is truly an emerging pathogen in our geographical area. Although these
organisms remain susceptible to polymyxins, given time and selective pressure, resistance to polymyxins, a rare
phenomenon, may yet occur. Cautious and selective use of late generation beta- lactam agents is critical to the
prevention of multiresistance, and "total" resistance.
- Go ES, Urban C, Burns J, Kreiswirth B, Eisner W, Mariano
N, Mosinka-Snipas K, and Rahal JJ (1994). Acinetobacter infections resistant to all antibiotics except polymyxin
B and sulbactam: clinical and molecular epidemiology. Lancet, 344:132a-1332.
- Gehrlein M, Leying H, Cullmann W, Wendt S. and Opferkuch
W. (1991) Imipenem resistance in Acinetobacter baumannii is due to altered penicillin-binding proteins. Chemotherapy
- Obara M, and Nakae T. (1991) Mechanisms of resistance
antibiotics in Acinetobacter calcoaceticus. J Antimicrob Chemother, 28, 791-800.
- Meyer, KS, Urban, C, Eagon, JA, Berger, BA, and Rahal,
JJ (1993) Nosocomial-outbreak of Klebsiella infection resistant to late generation cephalosporins. Ann.Intern,
- Paton R, Miles RS and Amyes SGB. (1993) ARI1:B-lactamase-mediated
imipenem resistance in Acinetobacter baumannii. International. J. of Antimicrob. Agents. 2: 81-88.
- Simon B.C., Lang F., and Holzman R.S.(1994) Successful
treatment of Acinetobacter calcoaceticus ventriculitis and development of subsequent chemical meningitis following
intraventricular polymyxin B. abstract #207 In: Program and Abstracts of 32nd Infectious Disease Society of America,
- Tankovic J, Legrand P, DeGatines G, Chemineau V, Brun-Buisson
C. and Duval J. (1994) Characterization of a hospital outbreak of imipenem resistant Acinetobacter baumannii by
phenotypic and genotypic typing methods. Antimicrob. Agents Chemother. 32:2677-2681.
- Urban C, Go E, Mariano N, Berger BJ, Avraham I, Rubin
D and Rahal JJ. (1993). Effect of sulbactam on infections caused by imipenem-resistant Acinetobacter calcoaceticus
biotype anitratus. J. Infect Dis. 167, 448-451.
- Urban C, Go E, Mariano N, and Rahal JJ. (1995). Interaction
of sulbactam, clavulanic acid and tazobactam with penicillin binding proteins of imipenem-resistant and susceptible
Acinetobacter baumannii. Fems. Microb. Letters. 125: 193-198.
- Wood CA, and Reboli (1993). Infections caused by imipenem-resistant
Acinetobacter calcoaceticus biotype anitratus. J. Infect Dis.168:1602-1603.