Early aggressive eradication therapy for intermittent
Pseudomonas aeruginosa airway colonization in cystic fibrosis patients:
15 years experience☆
C.R. Hansena,⁎, T. Presslera, N. Høibyb,1
aDepartment of Pediatrics, 5003, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark
bDepartment of Clinical Microbiology, 9301, Rigshospitalet, University Hospital of Copenhagen, Juliane Maries Vej 22, 2100 Copenhagen, Denmark
Received 21 December 2007; received in revised form 3 March 2008; accepted 16 June 2008
Available online 9 August 2008
Background: Since 1989, CF-patients intermittently colonized with Pseudomonas aeruginosa have been treated with inhaled colistin and oral
ciprofloxacin in the Copenhagen CF-centre. The study evaluates 15 years results of this treatment.
Methods: All isolates of P. aeruginosa from CF-patients intermittently colonized with P. aeruginosa from 1989 to 2003 were identified All anti-
P. aeruginosa treatments were evaluated for antibiotics used, treatment duration, pseudomonas-free interval and development of chronic infection.
All P. aeruginosa isolates were assessed for resistance and for non-mucoid or mucoid phenotype.
Results: 146 CF-patients were included in the study (1106 patient-years). 99 patients had first ever isolate during the study period. Median
observation time 7 years (0.1–14.9). 12 patients developed chronic infection. A Kaplan Meyer plot showed protection from chronic infection in up
to 80% of patients for up to 15 years. 613 colistin/ciprofloxacin treatments were given. There was no difference in pseudomonas-free interval
comparing 3 weeks (5 months) and 3 months (10.4 months) of colistin and ciprofloxacin, but a significant difference compared to no treatment
(1.9 months). Patients developing chronic infection had significantly shorter pseudomonas-free interval after treatment of first ever isolate
compared to patients remaining intermittently colonized (pb0.003). Treatment failure (P. aeruginosa-positive culture immediately after ended
treatment of first ever isolate) was a strong risk factor for development of chronic infection after 3–4 years, OR 5.8. 1093 pseudomonas-isolates
were evaluated (86.6% non-mucoid). No colistin-resistance was found. Ciprofloxacin-resistance was found in 4% of isolates.
Conclusion: Treatment of intermittent P. aeruginosa colonization in CF-patients using colistin and ciprofloxacin can protect up to 80% of patients
from development of chronic infection for up to 15 years. A positive culture immediately after treatment of first ever isolate is a strong risk factor
for development of chronic infection. We found no colistin-resistance and minimal ciprofloxacin-resistance.
© 2008 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.
Keywords: Cystic fibrosis; Pseudomonas aeruginosa; Colomycin; Ciprofloxacin; Pulmonary infection
Cystic fibrosis (CF) is characterised by recurrent coloniza-
tion and ultimately chronic infection of the lower airways.
Pseudomonas aeruginosa is the most common pathogen; up to
80% of the adult CF population is chronically infected with
P. aeruginosa. Chronic airway infection is a major cause of the
increased morbidity and mortality observed in patients with CF
by a period of recurrent, intermittent colonization of the airways
Journal of Cystic Fibrosis 7 (2008) 523–530
☆This work was presented at European Cystic Fibrosis Conference (ECFC) in
Copenhagen, 2006. Hansen CR, Pressler T, Høiby N. Treatment of intermittent
Pseudomonas aeruginosa colonization with colistin and ciprofloxacin result in
very low antibiotic resistance — 15 years follow up. J.Cyst.Fibros. 5
[supplement 1], S22. 2006. Ref Type: Abstract.
⁎Corresponding author. Tel.: +45 3545 4760; fax: +45 3545 6717.
E-mail address: firstname.lastname@example.org (C.R. Hansen).
1Fax: +45 3545 6412.
1569-1993/$ - see front matter © 2008 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.
[1,4]. Chronic infection is characterised by continuous growth of
P. aeruginosa in airway secretions, development of anti-
P. aeruginosa antibodies and also by a change in phenotype of
the microorganism from non-mucoid to mucoid. Once chronic
infection is established, durable eradication is impossible.
During the 1970s and 1980s, treatment of chronic pseudo-
monas infection became possible, thanks to an increasing
number of i.v. antibiotics and to the possibility of treatment with
inhaled and oral drugs . In 1989, treatment of intermittently
colonized patients using inhaled colistin and oral ciprofloxacin
was tried with encouraging results  and after a few years,
treatment of these patients became standard in the Copenhagen
CF centre. At first patients were treated for 3 weeks, but after a
few years the “Copenhagen Model” was established (Table 1)
and oral ciprofloxacin used with increasing dosage and for
increased duration at re-infections. Being one of the first anti-
P. aeruginosa antibiotics available for inhalation in addition to
inhaled aminoglycosides, inhaled colistin has been used for
about 20 years in Europe [5,9–12]. Oral ciprofloxacin has also
since the mid-1980s [9,13,14].
In 1996 an intern-analysis of colistin and ciprofloxacin
treatments efficacy for 3 weeks and 3 months was performed.
the group of patients receiving 3 months' treatment , and
standard treatment was changed to step 3 for all isolates of
P. aeruginosa in intermittently colonized patients: 3 months of
day divided in two doses (Table 1) [7,8]. During the late 1990′s
treatment with DNA-se (Pulmozyme®) was initiated; first in
patients chronically infected with P. aeruginosa and then as
standard treatment in the entire CF population after 1996.
The aim of the present study was to evaluate 15 years
experience with treatment of intermittent P. aeruginosa-coloniza-
tion in CF patients using inhaled colistin and oral ciprofloxacin in
the Copenhagen CF centre. We assessed the efficacy of the
treatment by means of time to recurrence of P. aeruginosa in
airway secretion and evaluated development of chronic infection,
number and type of treatments and possible development of
2. Materials and methods
The studyisadescriptive,retrospective cohortstudy.A cohort
of 296 CF patients is followed at the Copenhagen CF centre: all
patientsare seen every month inthe out patient clinic, and ateach
visit a sputum sample or lower respiratory tract secretions,
obtained by endo-laryngeal suctioning, is examined for the
presence of microorganisms as described previously . All
P. aeruginosa between 1 January 1989 and 31 December 2003
of P. aeruginosa during the study period. Patients were followed
chronic infection or time of death or loss to follow-up. All anti-
P. aeruginosa treatments were counted and evaluated for
antibiotics used and treatment duration.
Patients withfirsteverisolateofP.aeruginosaduringthe study
period were identified. These patients had not had any un-treated
treatment policy of the centre.Effectwas measuredastime to next
isolate ofP.aeruginosa, and for development of chronic infection.
Treatment failure was defined as a P. aeruginosa-positive culture
during treatment or immediately after ended treatment.
For patients developing chronic infection, time-period from
first ever isolate of P. aeruginosa to establishment of chronic
is defined as 6 consecutive monthly sputum samples with growth
of P. aeruginosa or — in case of shorter colonization — rise in
specific, precipitatingantibodies againstP. aeruginosa to at least 2
[1,15]. Intermittently colonized patients are defined as patients
with at least one isolate of P. aeruginosa, but normal levels of
precipitating antibodies against P. aeruginosa.
Patients were also assessed for development of chronic
infection according to Leeds criteria ; when more than 50%
of months, when samples had been taken, were P. aeruginosa
All sputum samples obtained in the study period were
evaluated for inclusion. Bacteriological investigation comprises
isolation and identification of P. aeruginosa, assessment for
mucoidity, primary sensitivity testing (using agar-diffusion) to
anti-P. aeruginosa antibiotics including ciprofloxacin (suscep-
tible strains MIC b0.2 μg/ml, resistant strainsN1 μg/ml) and
colistin (susceptible strains MIC b4 μg/ml, resistant strains
≥4 μg/ml)  (Neosensitabs®, Rosco, Copenhagen, Den-
mark). All P. aeruginosa positive cultures during the study
period were included.
Pseudomonas-free interval after treatment and time to
development of chronic infection were expressed as a Kaplan–
The “Copenhagen Model”: treatment of intermittent infection with P.
3rd and subsequent
1 mio. unitsab.i.d.
2 mio. units t.i.d.
2 mio. units t.i.d.
10–20 mg/kg b.i.d.
10–20 mg/kg b.i.d.
10–20 mg/kg b.i.d.
a1 mg colistin base=30,000 international units.
All (N=146)First ever isolate in
Age at diagnosis
Age at 1st ever isolate
Number of isolates in the study
0.5 years (0–13)
7.6 years (0.1–15)
4.2 years (0–28.1)
0.5 years (0–13)
7 years (0.1–14.9)
4.8 years (0.3–28.1)
524C.R. Hansen et al. / Journal of Cystic Fibrosis 7 (2008) 523–530
Meyer estimate. Cox regression was used to compare P.
aeruginosa free interval. Ciprofloxacin-resistance over time was
compared using Pearson correlation. Success or failure in
different treatments was compared using Chi-square test. Odds
Ratio (OR) was calculated using SPSS version 15. Level of
significance P-value ≤0.05.
146 patients (72 males) were intermittently colonized with P.
aeruginosa during the study period, of these, 99 patients (46
males)had first everisolate ofP. aeruginosaafter 1989 (Table 2).
1patient was losttofollow-upand 1patient diedduringthe study
period. The study includes 1106 patient years.
613 treatments were a combination of inhaled colistin and oral
ciprofloxacin, 164 treatments were combinations of other
antibiotics. 130 colonization episodes were not treated, mainly
in the beginning of the period, when treatment had yet to become
standard of the centre (Table 3). The remaining 186 isolates were
either found during treatment or in patients treated for chronic
infections caused by other Gram-negative microorganisms. Only
treatments with colistin and ciprofloxacin were further evaluated
in the study.
Duration of treatment with colistin and ciprofloxacin was
evaluated (Fig. 1). During the study, an increasing number of
treatments lasted 3 months, as standard treatment of all isolates
was changed to 3 months (step 3 in the “Copenhagen model”,
Table 1). 69 treatments lasted more than 3 months.
3.3. Treatment effect
In the group of 99 patients with first ever isolate during the
colonization episodes were not treated, 87 (24.2%) colonization
episodes were treated for 3 weeks and 239 (66.6%) colonization
episodes were treated for 3months (Fig. 2). In the non-treatment
group, median time to recurrence of P. aeruginosa was
1.9 months (range 0–58.3 months). In the group treated for
3 weeks, median time to recurrence of P. aeruginosa was
5 months (range 0.2–112.8 months), and in the group treated
for 3 months, median time to recurrence of P. aeruginosa was
10.4 months (range 0–98.2). The difference between no
treatment and any treatment was significant at all times
(pb0.02). The difference between 3 weeks and 3 months of
treatment was not significant.
When looking at time to recurrence of P. aeruginosa after
treatment offirst ever isolate with3 weeks or3 monthsofcolistin
and ciprofloxacin in the 99 patients with first ever isolate during
the study, the group of patients eventually developing chronic
interval between first ever and second isolate when comparing to
Fig. 1. Change in treatment-duration of colistin and ciprofloxacin; number of treatments of different lengths during the study period.
Treatment of isolates: number (percentage of all isolates)
Interval Colistin and ciprofloxacin Other treatmentsa
aOther treatments: inhaled antibiotics (colistin, tobramycin, ceftazidim) and/
or oral antibiotics (ciprofloxacin, azithromycin). In many cases, patients were
treated with i.v. antibiotics (beta-lactam in combination with tobramycin),
followed by treatment as described above.
525 C.R. Hansen et al. / Journal of Cystic Fibrosis 7 (2008) 523–530
the group of patients not developing chronic infection during the
study period (Fig. 3).
Treatment failure after treatment with colistin and cipro-
floxacin for first ever isolate was a risk factor for later
development of chronic infection, OR 5.8 (95% confidence
3.4. Development of chronic infection
Of the 99 patients with first ever isolate in the study, 12
patients developed chronic infection according to the Copenha-
gen definition after a median of 3.7 years after first ever isolate
(0.5–10.3 years), and a Kaplan Meyer estimate showed that up
to 80% of patients were protected against development of
chronic infection for up to 15 years (Fig. 4).
All 99 patients were assessed for chronic infection using
Leeds criteria  at any time based on previous 12 months,
more than 50% of months, when samples had been taken, were
P. aeruginosa culture positive. Of the 12 patients developing
chronic infection according to the Copenhagen criteria, all
patients became chronically infected according to the Leeds
criteria. 2 patients had a chronic infection by the Leeds criteria
5 months and 2 years, respectively, prior to the onset of chronic
infection by the Copenhagen criteria. 3 patients became
chronically infected according to both criteria at the same
time, and in 7 patients, chronic infection according to the Leeds
criteria developed a median of 1 year (range 0.3–12.7 years)
after development of chronic infection by the Copenhagen
criteria. Of the 87 patients not developing chronic infection by
the Copenhagen criteria, 3 patients at one time fulfilled the
Leeds criteria of chronic infection, but went back to intermittent
colonization according to the Leeds criteria.
All patients included in the study had specific, precipitating
anti-P. aeruginosa antibodies b2 at inclusion, and 10 of the 12
patients developing chronic infection had increasing precipitating
Fig. 2. Proportion of patients still free of pseudomonas after treatment with colistin and ciprofloxacin.
Fig. 3. P. aeruginosa-free interval after treatment of first ever isolate in patients who eventually did develop chronic infection and patients without development of
526C.R. Hansen et al. / Journal of Cystic Fibrosis 7 (2008) 523–530
From the 146 patients included in the study, 1093 isolates of
P. aeruginosa (946 non-mucoid, 147 mucoid isolates) were
assessed for resistance to colistin and ciprofloxacin. We found
no isolates resistant to colistin in intermittently colonized
patients in the study period.
44 ciprofloxacin-resistant isolates were found in 30 inter-
mittently colonized patients. 8 patients harboured ciprofloxacin-
resistant strains more than once. First ciprofloxacin-resistant
isolate was found a median of 1.6 years after first dose of
ciprofloxacin (range 0–13.9 years), and after a median of
24 weeks of total ciprofloxacin-treatment (range 0–99 weeks).
In all patients, the occurrence of ciprofloxacin-resistant isolates
was intermittent, since the resistant strain subsequently
4% of all isolates included in the study were ciprofloxacin-
resistant; 39 of 946 non-mucoid isolates (4.1%) and 5 of 147
mucoid isolates (3.4%) were ciprofloxacin-resistant (n.s.).
There was no increase in the occurrence of ciprofloxacin-
resistant isolates over time.
Treatment policy of CF patients colonized with P. aerugi-
nosa in the airways using inhaled colistin and oral ciprofloxacin
has evolved over the last 15 years in the Copenhagen CF centre
[6–8,18]. A 3 weeks treatment developed into the “Copenhagen
model” (Table 1) and has now become a standard treatment of
3 months of colistin and ciprofloxacin for all intermittent
colonisation episodes. The implementation of this treatment has
resulted in an increasing proportion of the patient population
free from chronic P. aeruginosa infection of the airways [19,20]
even into adulthood, resulting in improvement of lung function
compared to historic controls . Preservation of normal lung
function in CF patients free of chronic P. aeruginosa infection
 could ultimately lead to increased survival in these patients,
since respiratory failure following chronic P. aeruginosa
infection is still a very common cause of increased mortality
in CF, although introduction of antibiotic treatment has
modified the outcome of chronic P. aeruginosa infection .
Most CF centres agree on treating young CF patients
intermittently colonized with P. aeruginosa [8,9,22–34]. Most
centres use inhaled antibiotics alone or in combination with oral
ciprofloxacin either as primary treatment or after a course of i.v.
antibiotics. Treatment duration varies among centres. The
combination of antibiotics for this treatment is not yet defined
As can be seen from the present study, an official treatment
strategy of a CF centre does not automatically lead to the use of
this strategy in all patients. Individual treatments depend on
many things; implementation of a treatment strategy in the
centre, compliance, exacerbations, early signs of evolving
chronic infection and adverse effects of the treatment. Similar
results have been reported by others [23,32].
The median time to recurrence of P. aeruginosa was
5 months for 3 weeks treatment and 10.4 months for 3 months
treatment, this difference was not significant. The difference
between 5 and 10.4 months could still be clinically relevant,
however, leading to the difference between one or two
treatments for P. aeruginosa in a year in individual patients,
but also to the difference between 3 months or 6 weeks of total
treatment time in a year in individual patients. We found 1093
isolates of P. aeruginosa in 1106 patient years in this study, and
on average, every patient should have received one treatment a
year, if all isolates had been treated according to the treatment
protocol of our centre.
The present study shows a median time to recurrence of P.
aeruginosa similar to results reported by other groups on
treatment of intermittently colonized CF patients: Taccetti et al.
 found a median time to recurrence of P. aeruginosa of
18 months (range 4–80). Munck et al.  found a median time
to recurrence of P. aeruginosa of 8 months (range 3–25).
Treatment was very intensive: 2–3 weeks of i.v. antibiotics
Fig. 4. Proportion of patients without chronic infection in patients with first ever isolate in the study period.
527C.R. Hansen et al. / Journal of Cystic Fibrosis 7 (2008) 523–530