Elsevier

The Lancet

Volume 348, Issue 9028, 7 September 1996, Pages 639-642
The Lancet

Articles
Spread of β-lactam-resistant Pseudomonas aeruginosa in a cystic fibrosis clinic

https://doi.org/10.1016/S0140-6736(96)05169-0Get rights and content

Summary

Background

Pseudomonas aeruginosa colonisation of the airways of patients with cystic fibrosis (CF) is associated with considerable respiratory morbidity. Although segregation of colonised patients from non-colonised patients to prevent cross-infection has been recommended, there is little evidence that such cross-infection is widespread. We observed that a high proportion of children attending our CF clinic were colonised with P aeruginosa that was resistant to ceftazidime and other β-lactam antibiotics. We used two genomic fingerprinting techniques to see whether this may have arisen from epidemic spread of a single strain.

Methods

The prevalence of P aeruginosa colonisation and the antibiotic susceptibility of the organisms was determined from review of laboratory reports in the case-notes of 120 children with CF. Isolates were cultured from the sputum of 65 children colonised with ceftazidime-resistant P aeruginosa. Polymorphisms in total bacterial DNA from 92 isolates were analysed with two molecular fingerprinting techniques-pulsed-field gel electrophoresis after restriction enzyme digestion and assessment of flagellin gene polymorphisms by amplification of the whole gene and restriction enzyme digestion.

Results

92 (76–7%) of 120 children were colonised with P aeruginosa, and 65 of the 92 harboured isolates that were resistant to ceftazidime. Only three of the 92 children had never been treated with ceftazidime. The results of the two molecular-fingerprinting techniques were concordant and showed that 55 of 65 children harboured the same epidemic strain. This strain was resistant to ceftazidime, azlocillin, and imipenem, and sensitive to tobramycin and ciprofloxacin.

Interpretation

This study provides the first molecular evidence of a long-term outbreak of P aeruginosa in a CF centre. We suggest that careful surveillance of the prevalence of antibiotic resistance in CF centres should be instituted with measures to prevent cross-infection. We believe that antipseudomonal monotherapy should be considered with caution.

Introduction

Colonisation with Pseudomonas aeruginosa and Burkholderia cepacia is associated with progressive lung disease and mortality in patients with cystic fibrosis (CF).1, 2 There is also compelling evidence that epidemic spread of highly transmissible strains of B cepacia influences the prevalence of pulmonary colonisation in some centres.3, 4 By contrast, epidemic spread of P aeruginosa in unrelated patients, either in CF clinics or through social contacts, is a contentious issue, and there is a need for scientific evidence to support a policy of segregation as practised for patients colonised with B cepacia.

Historically, the use of simple phenotypic typing systems suggested that CF siblings tend to harbour the same P aeruginosa strain, which indicates the possibility of cross-infection if personal contacts are intimate and frequent. More recently, this form of person-to-person spread has been confirmed by reliable DNA-based genomic typing techniques, including pulsed-field gel electrophoresis.5, 6 Epidemic spread of multidrug-resistant P aeruginosa in unrelated patients has also been reported; Pedersen and colleagues7 showed that the incidence of new cases was reduced by strict isolation of colonised patients. However, their hypothesis that a single strain of P aeruginosa was responsible for cross-infection was criticised8 on the grounds that the study was based on the use of phage-typing and serotyping-phenotypic techniques that are unreliable for characterising the mucoid, polyagglutinable P aeruginosa that is characteristic of CF.9 Nevertheless, institution of cross-infection control measures in this unit interrupted spread of the bacterium. Subsequently, the development and use of genomic typing methods have indicated that small groups of patients within a CF clinic may harbour the same strain, which suggests that cross-infection, or acquisition from a single source, can happen.6, 10

In contrast to the above reports, most epidemiological studies of patients with CF have not shown spread of P aeruginosa in association with attendance at summer camps11, 12, 13, or when sharing hospital rooms,14 or during regular attendance at a CF clinic over many years.2 However, there is general agreement15 that patients with CF attending a CF centre usually harbour individual unrelated strains of P aeruginosa over a period of many years, and that, although cross-infection can occur, it is usually restricted to CF siblings or a minority of patients, the precise mode(s) of spread remaining unclear.

In 1995, we became aware that a large number of patients were colonised with a ceftazidime-resistant strain of P aeruginosa among the 120 patients who received their CF care in Liverpool but not among the 150 other patients who were regularly reviewed by our CF team at their local hospital. Here, we report the use of two genomic fingerprinting techniques to provide the first unequivocal evidence of an extensive epidemic in unrelated CF patients involving a β-lactam-resistant strain of P aeruginosa.

Section snippets

Patients

Case-notes of 120 patients attending the CF clinic at Alder Hey Children's Hospital in August, 1995, were retrospectively assessed to identify those colonised with P aeruginosa and to determine the date of first isolation. Diagnosis of CF was based on two positive sweat tests and confirmed by genetic testing. Antibiotic susceptibility for P aeruginosa isolates based on disk-diffusion tests was available from case-notes. According to the treatment policy of the unit from 1986 to 1994, all

Results

Review of the case-notes showed that, in March, 1995, 92 (76·7%) of the 120 children were colonised with P aeruginosa. Disk susceptibility tests had shown that 65 (70·6%) of these colonised patients harboured isolates that were resistant to ceftazidime and other (3-lactam antibiotics but sensitive to tobramycin and ciprofloxacin. Retrospective analysis of the patients' case-notes indicated that a ceftazidime-resistant P aeruginosa had first been detected in 1987. Since then, there has been a

Discussion

The proportion (76·7%) of our clinic population colonised with P aeruginosa is typical of many other CF centres. By contrast, the observation that 70·6% of these patients harboured ceftazidime-resistant and (3-lactam-resistant P aeruginosa is higher than the figure of 14% expected from previous surveillance studies in our centre20 and of 18–6% from national studies.21 A major objective of our study was to see whether acquisition of (3-lactam-resistant P aeruginosa reflected antibiotic-induced

References (23)

  • A multicenter comparison of methods for typing strains of Pseudomonas aeruginosa predominantly from patients with cystic fibrosis

    J Infect Dis

    (1994)
  • Cited by (0)

    View full text