Aerosolized Antibiotics for Treating Hospital-acquired and Ventilator-associated Pneumonia
Abstract and Introduction
Abstract
Hospital-acquired pneumonia is a common complication that
continues to have a poor cure rate in some patients with intravenous therapy
alone. Aerosolized antibiotics are theoretically attractive in an attempt to
optimize lung concentrations of antibiotics. Limited data suggest that
aerosolized aminoglycosides or colistin in addition to intravenous therapy
results in good response rates in patients with multidrug-resistant organisms
or nonresponding pneumonia. Adverse events can occur, especially with colistin.
When used, care should be taken to properly compound and administer aerosolized
antibiotics to ensure tolerability and good drug delivery.
Introduction
Hospital-acquired pneumonia (HAP) continues to be a common
and often devastating complication of modern healthcare. Approximately 90% of
HAP patients are mechanically ventilated and may also be referred to as having
ventilator-associated pneumonia (VAP).[1]
Approximately 9–27% of mechanically ventilated patients in the intensive care
unit (ICU) develop HAP/VAP. This results in an increased attributable mortality
of up to 50%, an increased length of stay by 7–9 days, and increased healthcare
costs of up to US$40,000 per patient.[1] Despite
the availability of modern ICU care and modern antibiotics, the overall
clinical cure rate for HAP in randomized clinical trials is only 63%. The cure
rate for Pseudomonas aeruginosa is even lower.[1,2] It is
unclear why HAP/VAP cure rates are so low. A number of patient and
pathogen-related factors could be involved including patients' underlying
disease states, interpatient differences in the immuno-inflammatory response to
illness and bacterial colonization, and pathogenic differences among bacteria
including the potential to develop resistance during therapy. Compounding the
problem is the lack of novel Gram-negative antibiotics in a time of increasing
resistance from organisms such as P. aeruginosa, Acinetobacter
baumannii and various Enterobacteriacae.
However, problems with antibiotic delivery may also be
important. Many antibiotics used to treat HAP, including β-lactams, aminoglycosides
and vancomycin, have poor penetration into the epithelial lining fluid (ELF) of
the lungs.[3]
The ELF concentrations for these antibiotics are usually less than 50% of the
concentrations achieved in the serum.[3]
Unfortunately, pharmacokinetic studies to determine optimal dosing of
antibiotics are normally done in normal volunteers and not critically ill
patients who can have markedly different pharmacokinetics.[4] Even
fewer formal pharmacokinetic studies are done to assess lung ELF penetration of
antibiotics in actual patients. As such, relatively little is known about the
optimal dosing of most antibiotics for HAP. Many antibiotics have doses
specific to treating HAP or life-threatening infections; however, correct doses
may not be used.[1]
It is clear that antibiotic therapy needs to be optimized in HAP.
One method of improving antibiotic delivery to the lung ELF,
and potentially improving outcomes, is aerosolized delivery into the lungs.
Since the 1960s there have been a number of reports using aminoglycosides,
colistin and β-lactams for treating HAP. However, there has been a sharp
increase in these reports over the last decade largely in an effort to treat
emerging multidrug-resistant (MDR) strains of Gram-negative bacilli such as P.
aeruginosa and A. baumannii. This article will summarize the
available literature on the use of aerosolized antibiotics for treating HAP in
adult patients.
Aminoglycosides
Until the 2000s, aminoglycosides were the most widely
studied aerosolized agents for treating HAP (Table 1). Pines et al.
initially reported two observational case series with aerosolized gentamicin.[5,6] The
first series included patients that had not responded to intravenous therapy for
chronic bronchial infections.[5] The
clinical improvement rate after 4–8 weeks of therapy was rather dismal (18%)
regardless of whether systemic therapy was continued or not. However, in the
second series 12 patients undergoing initial treatment for HAP had aerosolized
therapy added to carbenicillin.[6] A more
typical response rate of 67% was reported after 7–10 days of therapy.
Subsequently, Klastersky et al. increased the rigor
of study design with two randomized trials.[7,8] It is
important to note that antibiotics were endotracheally instilled (ET) as a solution
in these two studies rather than aerosolized. In the first, 15 patients were
randomized to receive either ET or systemic gentamicin.[7] The cure
rate was significantly better in the patients in the ET group (100 vs 25%). In
the second study, 38 patients were randomized to receive either ET sisomicin or
placebo added to systemic β-lactam/aminoglycoside therapy.[8] Again, the
cure rate was significantly better in the ET group (77 vs 45%). Incredibly,
this study from 1979 remains the largest randomized, placebo-controlled trial
of aerosolized antibiotics for treating HAP.[8] In the
1980s, three case reports totaling eight patients showed a cure rate of 100%
with ET or aerosolized therapy for 7–16 days.[9–11] Seven
of the eight cases also received intravenous antibiotics. Importantly, five of
the patients had previously not responded well to intravenous antibiotics
alone.
After years without new data, since 2007 there have been
four new publications. All four reports used increased rigor in that they all
used aerosolization, included only patients with VAP, and required
bacteriologic confirmation of infection. A small trial randomized ten patients
to receive either aersosolized or intravenous tobramycin for P. aeruginosa
or A. baumannii.[12] The two
treatment failures were in the intravenous group. A retrospective observational
study of 22 patients receiving either aerosolized tobramycin or amikacin for a
mean of 7 days added to intravenous therapy reported a 41% recurrence rate
(i.e., 59% cure rate).[13] This
cure rate is not impressive; however, many of the treatment failures were in
patients with difficult-to-treat situations such as previous episodes of VAP or
MDR organisms. In the largest observational study to date, Czosnowski et al.
reported the outcomes of 53 trauma ICU patients treated with aerosolized
tobramycin or amikacin added to intravenous therapy for approximately 10 days.[14] The
clinical cure rate was a reasonable 73% despite a high rate of patients with
previous treatment failure with intravenous therapy alone or the presence of
MDR organisms. This study is also the only one with microbiologic confirmation
of outcomes in a large percentage of patients. Lastly, there was an observational
study in 16 cancer patients treated with aerosolized aminoglycosides and/or
aerosolized colistin. The overall clinical response rate was 100% compared with
55% in a matched group of patients who did not receive aerosolized therapy.
However, the outcomes for colistin and aminoglycosides were not reported
separately.[15]
Colistin
Similar to aminoglycosides, reports with aerosolized
colistin date from the 1960s (Table 2). Pines et al.
again reported poor outcomes (clinical improvement 24%) when adding aerosolized
colistin to intravenous colistin in 17 patients treated for 7–10 days.[5] However,
the use of systemic and aerosolized colistin has reemerged over the past decade
owing to the emergence of MDR P. aeruginosa and A. baumannii,
which are resistant to all other antibiotic choices such as carbapenems,
aminoglycosides and floroquinolones. Hamer reported a 100% cure rate in three
patients with MDR P. aeruginosa HAP treated for 11–14 days with
aerosolized colistin added to intravenous therapy.[16] Similarly,
16 patients with MDR A. baumannii VAP in another report also had a 100%
cure rate after 15 days of therapy.[17]
Interestingly, rifampicin was the systemic therapy used rather than intravenous
colistin.
Four other relatively small observational reports since 2004
involved a total of 46 patients with either MDR P. aeruginosa or A.
baumannii HAP.[14,18–20]
Various systemic antibiotic regimens were used in all but one patient and the
mean treatment duration was 10–19 days. The clinical cure rates in these
reports were rather consistent (73–88%) and seem reasonable for treating MDR
organisms.
Fortunately, the most recent group of four studies had
better study designs and/or increased numbers of patients. All of the patients
in these studies were mechanically ventilated and higher quality culture
techniques were used. Michalopoulos et al. reported a prospective
observational study of 60 patients with P. aeruginosa, A. baumannii
or Klebsiella pneumoniae VAP treated for a mean of 16 days.[21] The
vast majority also received intravenous therapy (57 out of 60) and half of the
isolates were MDR. The clinical cure rate was 83% and the mortality rate
attributable to VAP was 17%. Similarly, Lin et al. reported a
retrospective observational study of 45 patients with MDR A. baumannii
VAP treated for a mean of 10 days.[22] All
patients also received intravenous therapy; however, the clinical cure rate was
lower than most other reports (58%). The overall mortality rate was high at
42%, but the VAP-related mortality was not reported. Pereira et al. also
reported an observational case series of 14 patients; mostly with P.
aeruginosa. The combined cure or improvement rate was 93%.[23]
The largest and most well-designed study was a retrospective
matched case–control study of 43 VAP patients who received aerosolized and
intravenous colistin compared with 43 patients treated with intravenous
colistin monotherapy.[24]
Patients were matched based on age and APACHE II score and the mean treatment
duration was 10–13 days. Approximately 75% of patients had A. baumanii
and the remaining patients had K. pneumoniae or P. aeruginosa.
Unfortunately, there were no significant differences between the aerosolized
group and the control group in the combined clinical cure or improvement rate
(74 vs 60%) or VAP-related mortality (16 vs 26%). However, there was a
statistical trend towards decreased mortality in the aerosolized group (23 vs
42%; p = 0.066). Three important limitations of the study are the retrospective
design, the possibility of a type II error because of the small study size, and
a lack of reporting of antibiotic therapy prior to initiating colistin that
could have affected patient outcomes. Nonetheless, this is an important study
because of the case-matched design.
Three other reports deserve mention. A case series of five
patients with chronic P. aeruginosa pulmonary colonization were treated
with 4–8 days of aerosolized colistin alone.[25] All
patients had eradication of the organism. A larger observational study reported
on a mixed group of 71 patients with either HAP or pulmonary colonization with P.
aeruginosa or A. baumannii.[26] Most
patients received intravenous therapy, and the mean duration was 11 days. The
eradication rate was 92%. These two reports are less important because it
wasn't clear which patients actually had HAP versus colonization. Last, a case
report showed successful treatment of Stenotrophomonas maltophilia VAP
with aerosolized colistin added to intravenous doxycycline in one patient who
failed trimethoprim/sulfamethoxazole therapy.[27]
β-lactams
Pines et al. again provided an early report, this
time with aerosolized carbenicillin added to intravenous in 15 patients with P.
aeruginosa HAP (Box 1).[6] Only 47%
of patients responded well to 7–14 days of therapy. Other researchers in the
1980s added aerosolized cefotaxime or ceftazidime to the same antibiotic given
intravenous for HAP caused by various Gram-negative organisms.[28] The
cure rate was 96% in 25 patients.
Administration Considerations in
Mechanically Ventilated Patients
Because 90% of HAP patients have VAP, this article will
focus on administration issues only in mechanically ventilated patients. Also,
there are a number of unique issues in delivery of aerosolized medications in
mechanically ventilated patients. Outpatient administration of aerosolized
tobramycin results in delivery of approximately 10–20% of the dose to the lung.[29]
However, initial data examining the lung delivery of aerosolized medications in
mechanically ventilated patients showed that less than 3% of a dose reached the
lung.[30]
Not only was drug delivery poor, but it was highly variable with over a tenfold
difference in delivery depending on the administration techniques used.[31] It is
unclear why delivery is so poor during mechanical ventilation, but some
potential problems include deposition of drug in the tubing, the distance from
the site of aerosolization to the lung, the inability of patients to control
their breathing and the nonphysiologic nature of mechanical ventilation. Other
patient-related factors that may impair drug delivery include atelectasis,
mucus production, or possibly preexisting lung diseases such as asthma or
chronic obstructive pulmonary disease.
Thus, a series of laboratory and clinical studies were
performed to determine how to improve aerosol drug delivery during mechanical
ventilation.[29,32]
It has subsequently been shown that using optimal technique can increase lung
delivery by as much as 650%.[31] Factors
related to optimal aerosol delivery in mechanically ventilated patients are
listed in Box 1.
Ideally, hospitals would only use ventilators for which good aerosolized
delivery has been confirmed in studies. One study showed a high degree of
variability between four ventilators in aerosol delivery.[32]
However, very few comparative data are available. A reasonable alternative is
to ensure that ventilators used for antibiotic delivery should have a flow rate
greater than 6 l/min and that they only nebulize during inspiration rather than
continuously.[32]
The choice of nebulizer can also be important. One of the
primary determinants of lung deposition is the median mass aerodynamic diameter
(MMAD) of the particles generated. Lung deposition is optimized at a MMAD of
1–5 µm. Larger particles tend to deposit in the upper airways and smaller
particles tend to be exhaled.[29,32] Jet
nebulizers are the most widely used in ICUs. They are inexpensive, disposable
and generally produce acceptable particle size. The aforementioned studies
almost universally used jet nebulizers. Ultrasonic nebulizers can perform
better than jet nebulizers in producing better particle size and shorter
administration times. However, they require a specialized power supply at the
bedside, must be cleaned, and are more expensive.[32] Also,
some large drug molecules may be broken down during ultrasonic nebulization.[32] As
such, jet nebulizers are preferred for aerosolized antibiotic use in HAP. There
is no preferred model of jet nebulizer for HAP. Clinicians should ensure that
the nebulizer model being used creates a MMAD of 1–5 µm based on the
manufacturer's information.
Clinicians should also know the maximum fill volume for the
nebulizer being used. Starting administration with a full nebulizer results in
improved drug delivery.[32]
Aerosolized doses should be compounded to a total volume equivalent to the fill
volume.[29,32]
The placement of the nebulizer in the ventilator circuit also affects delivery.
The optimal site for placement is 30 cm from the endotracheal tube in the
inspiratory loop.[29,32] There
are also data showing that drug delivery is improved when the ventilator
humidification is turned off during nebulization.[29,32]
Obviously, it is important that humidification be resumed after the dose is
administered. Implementing this into practice will require more training and
vigilance than the other methods described here.
Safety & Tolerability
A number of pulmonary adverse events such as coughing and
bronchospasm can occur with outpatient use of aerosolized antibiotics. The only
adverse events reported in the aminoglycoside HAP studies were coughing and
dizziness in four patients.[5,10] No
other adverse events were reported. The vast majority of aminoglycoside HAP
reports used the intravenous formulation. A preservative-free formulation of
tobramycin for inhalation used in cystic fibrosis may theoretically provide a
better safety profile, but this has not been studied widely in HAP patients.
There are fewer data on the safety of aerosolized β-lactams; however, no
adverse events were reported in the HAP treatment studies or when ceftazidime
was used in trials of VAP prevention.[6,28,33,34]
Compared to aminoglycosides, aerosolized colistin may have a
worse adverse event profile. It was 'badly tolerated' in an earlier report but
a number of more recent studies reported that there were no adverse events.[14,17,19,24]
Hypotension was reported in one patient treated and cough or bronchospasm was
reported in four others.[23,33,35] This
did not occur with intravenous colistin in this patient. Another case report
showed hypersensitivity pneumonitis and fibrosis from aerosolized colistin used
to treat HAP.[36]
Most worrying is the report of death in a patient treated as an outpatient with
aerosolized colistin.[101] It is
thought that this patient suffered from a reaction to a toxic breakdown product
of colistin that occurred because the doses were prepared more than 24 h prior
to administration. Subsequently, the US FDA issued a public health advisory
that colistin be administered promptly after preparation.
Clinicians should be aware of several factors that can
improve tolerability of aerosolized antibiotics.[29,32]
Doses should be diluted in normal saline when dilution is needed to QS to the
maximum fill volume of the nebulizer being used. A sodium concentration of
77–154 mEq/l is associated with better tolerability. The pH of the dose should
also be roughly physiologic (4.0–8.0). There is a wide range of tolerable
osmolarity (150–1200 mOsm/l), but hypotonic solutions should be avoided. For patients
with chronic lung disease or who have had previous adverse events from an
aerosolized dose, pretreatment with albuterol should be considered (Box 1). Another option may
be to switch to the preservative-free tobramycin if the organism is sensitive
to it.
One of the potential benefits of aerosolized administration
is the potential for lower systemic concentrations and thus fewer adverse
events such as nephrotoxicity. It appears that aerosolized aminoglycosides do
not generate detectable serum concentrations in patients with normal renal
function.[29,32]
A recent study of nine critically ill HAP patients showed that the only patient
with a detectable tobramycin serum concentration (0.80 µg/ml) during
aerosolized therapy also had moderate renal dysfunction.[37]
Similarly, a lack of detectable serum concentrations was seen with aerosolized
ceftazidime.[33]
However, accumulation has been rarely reported, especially in patients with
renal dysfunction.[10,37]
Thus, assessing systemic accumulation in patients with renal dysfunction may be
prudent.
Expert
Commentary
Despite the fact that adjunctive aerosolized
antibiotics are commonly used in HAP, there is a disappointing amount of data
in the medical literature. Almost all of the studies were observational and did
not have control groups. Nonetheless, the relatively poor response rates seen
with intravenous therapy of HAP and the emergence of MDR organisms makes new
treatment options desirable. The ATS/IDSA HAP guidelines recommend that
"adjunctive therapy with an inhaled aminoglycoside or polymyxin (colistin)
for MDR Gram-negative pneumonia should be considered, especially in patients
who are not improving.[1] "
The recommendations by the Society of Infectious Diseases Pharmacists are
similar.[38]
However, is it still unclear from the literature if
aerosolized antibiotics provide an additional benefit in HAP. If one discounts
the universally poor results seen by the early studies by Pines et al., aminoglycoside cure rates
were 67–100% in a total of over 165 patients[6–14] and
colistin cure rates were 58–100% in a total of over 200 patients.[14,16–24]
Given than many of these patients had MDR organisms or had failed intravenous
therapy alone, these cure rates seem to compare favorably to the baseline cure
rate of 63% for VAP seen in clinical trials. The only randomized study did show
a benefit; but that study is from the 1970s.[8]
Unfortunately, the most recent and best designed study with colistin failed to
show a benefit.[24]
There are also questions regarding the quality of the
data. There could have been publication bias over time such that clinicians
only reported positive experiences. Obviously, there is a lack of controlled
studies and the numbers are low. Perhaps more concerning are the vast
differences in ICU care over the decades between the early reports and today in
any number of areas. Many of the early reports included nonmechanically
ventilated patients or did not report whether patients were mechanically
ventilated. It is also difficult to determine the severity of illness of those
patients. The culture techniques were also not well described in earlier
studies. Thus, it is not clear if those patients truly had HAP. Some patients
were treated with aminoglycoside monotherapy for HAP which is not acceptable in
the modern context based on poor cure rates.[1]
Fortunately, reports from the 2000s have generally become larger and more
rigorous in important aspects of HAP care. For instance, the recent reports
focused only on VAP patients, used far better microbiologic techniques for
diagnosis and were likely to have used better administration techniques.
Although most studies did a very poor job of describing drug administration in
detail.
Ultimately, the risk versus benefit analysis seems to
be somewhat in favor of using aerosolized antibiotics in selected patients.
Such patients might include those:
- Not responding to
intravenous therapy alone; - Who are being treated for
a recurrence of the same organism; - Who have MDR organisms.
This recommendation is based on what seems to be good
cure rates in difficult-to-treat cases. This is balanced by a rather benign
adverse event profile based on limited data – at least for aminoglycosides.
Aminoglycosides should be the drug of choice when they can be used based on in vitro sensitivity. Colistin
should only be used for aminoglycoside-resistant isolates based on what seems
to be a higher risk of adverse events. Clinicians should also take care that
doses are optimally compounded and administered to ensure the best drug
delivery possible.
Regarding dosing, a wide range of doses were used
clinically in the reports described. The Society of Infectious Diseases
Pharmacists statement also provides dosing recommendations.[38] A
reasonable approach when using gentamicin or tobramycin is to use 300 mg every
12 h. This is the FDA-approved dose in cystic fibrosis. There are very few data
to guide amikacin dosing, but the drug is typically dosed approximately three
times higher than gentamicin and tobramycin when administered systemically, so
a dose of 500–1000 mg every 12 h seems reasonable. A colistin dose of 150 mg every
12 h is reasonable based on recent reports. Although the prescribing
information recommends 25–50 mg two to three times per day. Note that these
doses are based on the Coly-Mycin M product (Monarch Pharmaceuticals, TN, USA)
in the USA, which is dosed in mg of active colistin. The Colomycin product
(Forest Laboratories UK, Dartford, UK) has 30–33 mg of active colistin per 80
of colistimethate sodium (equivalent to 1 million units).
Five-year
View
Current
Research
In 5 years from now, the first large randomized,
placebo-controlled clinical trial of aerosolized antibiotics in HAP might have
been completed.[102]
This will be a clear landmark in this area for two reasons. First, this will be
by far the largest and most rigorous study to date. Second, the study will use
a new generation of nebulizer technology: vibrating mesh plate nebulizers.
These nebulizers seem to provide much better drug delivery than jet nebulizers.[38] A
pharmacokinetic trial of a new aerosolized amikacin product in VAP patients
showed excellent drug delivery to the lung.[39] If this
study is successful at showing that adding the aerosolized plus intravenous
therapy is superior to intravenous therapy alone, then it could potentially
create a dramatic shift in the treatment of HAP. One could also envision a
middle ground where aerosolized therapy only adds a benefit in certain
situations based on patient- or organism-related factors. Although such an
outcome would likely not result in FDA approval. If the study shows no benefit
for aerosolized therapy, then it will cast a very negative light on this
therapy.
Future
Research
With a lack of any other currently registered clinical
trials under development, it appears that clinician-investigators will have to
continue to report on their experiences with aerosolized antibiotics.
Hopefully, some will use the higher-level techniques, as seen in the recent
case-matched study, to better determine if there is a beneficial effect.
Ideally, more prospective, randomized, double-blinded trials will be performed.
However, the funding challenges for such a trial are clear. Funding would have
to come from the federal government or from a manufacturer developing an
aerosolized drug formulation. A welcome advance would be more data on the use
of β-lactams, which appear to be well tolerated. Data on aerosolized quinolones
may also be available in 5 years, although more likely in other indications
than HAP. Even more on the cutting edge are potential advances in drug
formulations using lipid-based products, or new-generation dry powder delivery
that could improve lung deposition.
沒有留言:
張貼留言