Pseudomonas aeruginosa is an opportunistic pathogen that can cause severe invasive disease in critically ill and immunocompromised patients.1 This microorganism is an important cause of nosocomial infections, including pneumonia, wound infections, bacteremia, and urinary tract infection,2,3P. aeruginosa is frequent cause of ventilator-associated pneumonia in intensive care units (ICUs), the associated risk factors being mechanical ventilation lasting longer than 7 days, severity of the underlying condition, and previous antibiotic treatment, surgery, or immunosuppression.3,4

P. aeruginosa is a uniquely problematic microorganism because of a combination of inherent resistance to many drug classes and an ability to acquire resistance to all relevant treatments.5 Thus, carbapenems remain one of the best therapy options, although their use is threatened by the emergence of carbapenem-hydrolyzing, enzyme-producing strains and dissemination of multidrug-resistant clones. One reported outbreak of P. aeruginosa was only susceptible to polymyxin.6,7 In Brazilian hospitals, P. aeruginosa is a leading cause of nosocomial infection and the first cause of nosocomial pneumonia, notably in ICUs.8–10

This study describes a nosocomial outbreak of imipenem-resistant, metallo-β-lactamase-producing P. aeruginosa with horizontal transmission in patients hospitalized in a mixed ICU.

Following the index case, 46 additional cases of hospital-acquired infection were documented, including mechanical ventilation-associated pneumonia (41), bloodstream infection (3), surgical wound infection (1) and urinary tract infection (1) caused by Pseudomonas aeruginosa resistant to imipenem and other antimicrobial agents, including cephalosporins, aminoglycosides, and fluoroquinolones (data not shown).

Fig. 1 shows the mortality data and the temporal/spatial relationships (ICU stay) among the 47 patients, which provides evidence of cross-transmission of IRPa.

Fig. 1.

Schematic representation of temporal and spatial relationships among patients infected with imipenem-resistant Pseudomonas aeruginosa during the study period. The arrows represent referral of patients from other hospitals to the ICU in Uberlandia Federal University Hospital, black sections indicate the time the patient was another ward, sections with hatched lines indicate time the patient was in the ICU and infected by imipenem-resistant P. aeruginosa, and crosses indicate death of the patient.

(0.5MB).

During the scrutiny for contamination, P. aeruginosa was not detected on culture of samples from the hands of 23 professionals, although initially this was suspected as the route by which the microorganism had been disseminated. However, P. aeruginosa was found colonizing 15/86 (17.4%) of the patients examined and on surfaces located near the patients, including the mechanical ventilation nebulizer in 10/86 (11.6%) and the headboard of the bed in 3/86 (3.5%), but not the sinks.

The majority of imipenem/ceftazidime-resistant P. aeruginosa strains in infected patients (33/43; 76.7%) had an MBL-positive phenotype, with no differences in detection between mercaptopropionic acid or EDTA (Table 1). In the contamination scrutiny, 17.8% (5/28) of P. aeruginosa imipenem/ceftazidime-resistant strains had an MBL-positive phenotype. Only 26.7% (4/15) of MBL-producing strains were positive for the blaSPM−1 gene. The blaVIM−2, bla IMP−1, and blaSPM genes were not found.

The presence of MBL-producing P. aeruginosa has been described in many hospitals around the world.18 Nonetheless, dissemination of an SPM-1 MBL-producing P. aeruginosa epidemic has been demonstrated only in hospitals of different Brazilian regions.8,14,19 As compared to other studies carried out in Brazil, the presently reported nosocomial outbreak of imipenem-resistant P. aeruginosa infection involved the largest number of cases to date (47 patients) and was associated with 4 clones. This study is the first to include an expressive number (≈60%) of critical IRPa-infected patients who died within 30 days after the infection.

Pseudomonas aeruginosa is a leading pathogen in mechanical ventilation-associated nosocomial pneumonia, with an associated mortality rate of 20% to 80%.8,9,20 During the outbreak in our hospital, pneumonia was the predominant infection, accounting for more than two-thirds (85%) of the cases, followed by bloodstream infection and surgical wounds.

The multivariate logistic regression analysis showed that patients with nosocomial IPRa infection were more likely to have been exposed to mechanical ventilation, and that tracheotomy and use of carbapenems were predisposing factors for the infection. These findings are consistent with those of other studies.4,20–22

Although the advent of carbapenems in the 1980s heralded a new treatment option for serious bacterial infections, carbapenem resistance is now observed in Enterobacteriaceae and Acinetobacter spp., and is becoming commonplace in P. aeruginosa.18P. aeruginosa possessing MBLs has been increasingly reported worldwide, mainly in ICUs and including Brazil,14,18,19 but at rates comparatively lower than those found in our study (≈90%).

These enzymes are clinically relevant because of their ability to hydrolyze most β-lactams, including carbapenems (but excluding aztreonam),18 and their association with mobile genetic elements, increasing the possibility of rapid spread.19,23–25 In the Americas, 6 different MBLs have been described in P aeruginosa, including SPM-1, IMP-1, IMP-16, VIM-2, VIM-8 and VIM-1118 and three MBL types (SPM-1, VIM-2 and IMP-1) have been detected in Brazilian hospitals, with most strains expressing blaSPM−1, which is disseminated among our hospitals.8,14,19,22 We found that MBL-producing strains accounted for a substantial number of cases (76.7%), and that the strains were clinically relevant, because 70.0% of the patients died. In contrast to the reported results of Sader et al,26 neither blaIMP−1 or blaVIM−1 genes were found in this study.

The acquisition of resistant bacteria in hospitals may be a consequence of selective pressure exerted by the use of antibiotics and/or horizontal dissemination.27 Molecular typing of strains presenting the blaSPM−1 gene showed the presence of a single clone (A). In the present study, multidrug-resistant P. aeruginosa strains were clustered in two major genotypes; these showed coresistance to most of the antimicrobial agents tested. Multidrug resistance can be explained by the accumulation of several resistance mechanisms, including gene mutation, over-expression of efflux pumps, loss or modification of porins, and acquired extended-spectrum β-lactamases.28 This suggests that the selective pressure of previous antimicrobial use in the unit contributed to the emergence of resistant clones.

Both the SPM and non-SPM-infected patients were clustered for 18 months in a single unit, with overlapping during their hospitalization. These data suggest that horizontal transmission between these patients may have played a role in the dissemination of the infection. Clonal dissemination was detected within and between intensive care units in São Paulo (4) and Brasilia,10 but with unrelated strains recovered from one of them. These data differ from those of Pellegrino et al29 who reported clonal dissemination between private and public institutions in Rio de Janeiro, in which both presented an apparently endemic background.

Detection of clonal dissemination usually indicates inadequate nosocomial infection control practice.10 This outbreak of P. aeruginosa infection was probably linked to transmission of the microorganism via the hands of healthcare workers. We did not identify hand carriage among the workers examined, but a spatial/temporal relationship was seen among the cases in our unit, and low adherence (23%) to proper hand washing30 practice was detected in a recent study.

In conclusion, to our knowledge, this is the first study to conduct an analysis using both classical and molecular techniques of an outbreak of infection associated with a highly prevalent MBL-producing P. aeruginosa, including mainly blaSPM−1 strains clustered in 2 major genotype, resulting in a significant increase in mortality of elderly patients and in previous use of antibiotics. The clonal dissemination among patients in our unit indicates problems in nosocomial infection control practice, likely associated with low adherence to hand hygiene, and selective pressure that provides resistant strains with an advantage over their susceptible ancestors.