Abstract |
ENGLISH TITLE
Comparison of the effectiveness of inhaled colistin in combination with
intravenous colistin versus intravenous colistin alone, in patients with ventilatorassociated
pneumonia.
ABSTRACT
Introduction
Ventilator-associated pneumonia is a common infectious complication in the
intensive care unit, leading to increase in mortality, prolongation of hospital stay and
increased costs. Ventilator-associated pneumonia is usually caused by multidrugresistant
bacteria. In Greek hospitals, the most common pathogens are
Acinetobacter baumannii, Klebsiella pneumoniae and Pseudomonas aeruginosa.
Intravenous colistin is one of the main therapeutic options. It is administered as a
prodrug (colistimethate sodium) with complex pharmacokinetic properties, and has
debatable efficacy and relatively high toxicity. Colistin can also be administered as an
inhaled antibiotic, according to the experience in patients with cystic fibrosis. The
use of inhaled colistin in patients with ventilator-associated pneumonia can
theoretically lead to higher drug levels at the site of infection, without considerable
risk for systemic toxicity
Aim
We aimed to evaluate the efficacy of treatment with inhaled colistin in
combination with intravenous colistin, compared with intravenous colistin alone, for
ventilator-associated pneumonia.
Methods
This was a retrospective, comparative cohort study, done at the intensive
care unit of a tertiary-care hospital, in Athens, Greece. We included all patients with
microbiologically-documented ventilator-associated pneumonia who received
intravenous colistin for at least 3 days, with or without nebulized colistin, between
May 2005 and August 2007. The usual dose of nebulized colistin was 1 MIU thrice
daily and that of intravenous colistin was 3 MIU thrice daily for normal renal
function.
Patient data regarding demographics, comorbidities, laboratory, imaging and
microbiological tests, antimicrobial treatments, invasive procedures, presence of
155
foreign bodies or catheters, special treatments, as well as the duration of mechanical
ventilation and hospitalization were extracted and analyzed. The primary endpoint
was clinical cure (resolution of pneumonia) at the end of colistin treatment. The
secondary endpoint was all-cause mortality. The variables that were associated with
the outcome were analyzed with multiple logistic regression.
Results
We included 121 patients, of whom 78 (64.5%) received inhaled colistin in
combination with intravenous colistin, whereas 43 (35.5%) received intravenous
colistin alone. Twenty-three (19.0%) of the 121 patients received concomitantly
other active antibiotics. The causative pathogens were A. baumannii in 92 (76.0%)
patients, P. aeruginosa in 22 (18. 2%), and K. pneumoniae in 7 (5,8%) patients. The
patients who received inhaled colistin had fewer isolates that were susceptible only
to colistin, had received fewer blood transfusions, while they received longer
treatment with intravenous colistin.
Clinical cure was more common in the inhaled colistin group [62/78 (79.5%)
patients versus 26/43 (60.5%) patients in the intravenous colistin alone group
(p=0.025)]. In multivariate analysis, the administration of inhaled colistin was the
only independent predictor variable of clinical cure (OR 2.53, 95% CI 1.11-5.76). In
the subgroup of 98 patients who received colistin as the only active antibiotic,
clinical cure was also more common in those who received inhaled colistin [46/60
(76.7%) versus 22/38 (57.9%) in those who received intravenous colistin alone,
p=0.049].
All-cause, in-hospital mortality was 31/78 (39.7%) in the inhaled colistin
group, compared with 19/43 (44.2%) in the intravenous colistin alone group
(p=0.63). All-cause mortality during intensive-care unit stay was 28/78 (35.9%) vs.
17/43 (39.5%), respectively (p=0.69). After adjustment for risk factors, treatment
with inhaled colistin was not associated with in-hospital mortality. The variables that
were independently associated with increased mortality were higher APACHE II
score (OR 1.12, 95% CI 1.04-1.20), underlying malignancy (OR 4.11, 95% CI 1.18-
14.23) and lower daily dose of intravenous colistin [OR (per MIU) 0.81, 95% CI 0.68-
0.96].
Discussion
The main finding of this study is that, among patients with ventilatorassociated
pneumonia caused by multidrug-resistant Gram-negative bacteria, clinical
cure was more common in those treated with inhaled colistin combined with
intravenous colistin, than in those treated with intravenous colistin alone.
The benefit of adjunctive inhaled colistin can be attributed to the limited and
delayed penetration of intravenous colistin in lung parenchyma. On the contrary,
inhaled colistin appears to achieve high lung concentrations and is associated with
156
low risk for nephrotoxicity. However, the mode of administration of nebulized
antibiotics and the penetration in foci of serious pneumonia warrant further study.
The survival of the patients with ventilator-associated pneumonia depends
mainly on their underlying clinical condition, while the infection-attributable
mortality appears to be relatively low. For the demonstration of potential
differences in survival with the use of adjunctive inhaled colistin, a large sample size
would be required.
Most of the relevant in vivo and clinical studies support the idea that
adjunctive inhaled colistin could be beneficial in ventilator-associated pneumonia, in
terms of clinical improvement and microbiological eradication. There is however
substantial heterogeneity between these studies as regards to the mode of
administration of inhaled colistin, the dosage used, other concomitant antibiotics,
and the susceptibility profile of the causative pathogens.
Conclusions
The use of inhaled colistin, as adjunctive to intravenous colistin, in ventilatorassociated
pneumonia due to multidrug-resistant Gram-negative bacteria appears to
be associated with a higher likelihood of clinical cure, without evidence for improved
survival. These findings merit futher investigation in randomized clinical trials on the
safety and efficacy of inhaled colistin.
|