REVIEW

CURRENT
OPINION Inhaled antibiotics for the treatment of pneumonia
Matthew P. Schreiber and Andrew F. Shorr

Purpose of review
To describe recent developments in trials exploring inhaled antibiotics for treating severe pneumonia.
Recent findings
Three recent randomized studies investigated the potential role for aerosolized antibiotics for gram-negative
pneumonia in ventilated patients. One single center, nonblinded investigation suggested a benefit with
inhaled amikacin for resistant gram-negative infections. However, two multicenter, blinded trials found no
benefit to adjunctive nebulized amikacin for severe gram-negative pneumonia.
Summary
Well done clinical trials do not support the routine use of inhaled amikacin for pneumonia in ventilated
patients. There may be a potential role for aerosolized antibiotics when other options are limited.
Keywords
antibiotics, inhalational therapies, multidrug resistant organisms, pneumonia

INTRODUCTION of distribution that, in turn, has implications for

Pneumonia remains a leading cause of morbidity and antibiotic dosing [5]. Concurrent use of inotropes
mortality. Evolving patterns of antimicrobial resis- and vasopressors also impacts antibiotic drug clear-
tance have complicated efforts to improve outcomes, ance and underscores that some individuals receiv-
especially in severe pneumonia necessitating mech- ing traditional doses of renally cleared antibiotics
anical ventilation. Whether requiring mechanical may, actually, be receiving inadequate doses of anti-
ventilation for community-acquired pneumonia microbials. This concept of ‘augmented renal clear-
(CAP) or a new pneumonia in a previously ventilated ance’ (ARC) has been shown, in several analyses, to
patient (VAP), patients suffering for respiratory failure have important implications [6,7]. When patients
and pneumonia face crude mortality rates approach- with severe infection and ARC receive standard
ing 30–40% [1,2]. Such severe infections also remain antibiotic regimens they face lower cure rates. To
associated with significant morbidity and cost [1,2]. address this challenge one can employ higher doses
One key determinant of outcomes in these syn- than usually given. This paradigm, though, may
dromes is the appropriateness of initial antibiotic expose the patient to multiple toxicities. Addition-
therapy. Multiple analyses document that failure to ally, the challenge posed by ARC when coupled with
administer, in a timely manner, an antibiotic with in the increasing prevalence of resistant organism has
vitro activity against the culprit pathogen indepen- led to use of select, older antiinfective agents, such
dently increases a patient’s risk of death two-fold to as colistin and polymixins. These agents carry sub-
five-fold [3,4]. The delivery of appropriate antibiotic stantial risks. All of these factors taken together have
therapy in an era of increasing antimicrobial resis- fostered a search for novel means to address the
tance has become more complicated in that physi- predicament clinicians now face in treating pneu-
cians can no longer insure that commonly utilized monia in ventilated patients. One avenue has,
antibiotics will actually treat multidrug-resistant
pathogens. Of particular concern are organisms
Section of Pulmonary, Critical Care, and Respiratory Services, MedStar
such as Pseudomonas aeruginiosa and Acinetobacter Washington Hospital Center, Washington, District of Columbia, USA
bumanni. Correspondence to Andrew F. Shorr, MD, MPH, MedStar Washington
At the same time, accumulating evidence indi- Hospital Center, Suite 2a70, 110 Irving St. NW, Washington, DC 20010,
cates that the prescription of appropriate antibiotic USA. Tel: +1 202 877 7856; fax: +1 202 291 0386;
therapy to critically ill patients is complicated by a e-mail: [email protected]; [email protected]
number of other factors. Specifically, the use of Curr Opin Pulm Med 2019, 25:289–293
mechanical ventilation changes a patient’s volume DOI:10.1097/MCP.0000000000000557

1070-5287 Copyright ß 2019 Wolters Kluwer Health, Inc. All rights reserved. www.co-pulmonarymedicine.com

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

Infectious diseases

could obviate any potential benefit associated with

KEY POINTS inhaled drug dispensation [9].
 Current suggest a possible role for inhaled antibiotic in Reflecting the possibilities associated with
the treatment of severe pneumonia. inhaled antibiotics for treating pneumonia in ven-
tilated patients, a recent survey of over 400 ICUs
 Recent trials of inhaled antibiotics have failed to revealed that many have already adopted aerosol-
demonstrate a benefit.
ized antibiotics. In this international investigation,
 There may remain a role for inhaled therapies as part more than 25% of responding units reported regular
of rescue regimens when other options are limited. &
use of the inhaled approach [10 ]. Inhaled therapies
were used predominantly for pneumonias because
of multidrug-resistant organisms and as an adjunct
to standard intravenous therapies. In other words,
necessarily, focused on the development and testing aerosolized antibiotics were given only as an adjunct
of novel new agents whereas other approaches have when a pathogen was identified rather than as part
emphasized the optimization of pharmacodynamics of an empiric therapy regimen. Emphasizing the
and pharmacokinetics at the bedside [8]. burden of resistant organisms, colistin and amino-
An alternative paradigm, though, has stressed glycosides were the most commonly prescribed
innovative means for antibiotic delivery. Specifically, &
inhaled agents [10 ]. Notably, the use of inhaled
interest has emerged in inhaled antibiotics. Potential agents for treating pneumonia had increased when
advantages of inhaled administration include the &
compared to a prior survey 3 years earlier [10 ]. It is
delivery of high doses of drug directly to the target also important to note that another contemporary
tissue. Theoretically, one optimizes efficacy by survey of practices surrounding the use of nebulized
guaranteeing that drug levels at the site of infection antibiotics reported that only 28% of ICUs relied
are many fold higher than the minimum inhibitory upon optimal technique when administering
concentration (MIC) of the responsible pathogen. inhaled antibiotic therapy [11].
Beyond eradicating the pathogen, such high drug This change in behavior was likely in part moti-
levels in the lung might also help to mitigate the vated by the increasing prevalence of resistant
emergence of drug resistance. Inhaled administra- pathogens encountered in severe pneumonia. Addi-
tion would also possibly allow the clinician to avoid tionally, guidelines from several professional socie-
select systemic toxicities of some specific antimicro- ties suggested that there might be a potential role for
bials. For example, inhaled administration of amino- aerosolized antibiotics when treating VAP [12,13].
glycosides would overcome not only the fact that More importantly, though, was the publication of
these drugs have poor lung penetration when given a meta-analysis by Zampieri et al. [14] that suggested
intravenously but would also potentially protect the that adjunctive inhaled antimicrobials increased
patient from concerns about nephrotoxicity. Inhaled rates of clinical cure significantly. Readers should
use of antibiotics might also help to prevent Clostrid- note, however, that a subsequent meta-analysis
ium difficile diarrhea because only the lungs, and not focusing only on randomized studies (as opposed
the gastrointestinal tract, would see drug deposition. to including observational analyses) and which
The success of inhaled antibiotics in cystic fibro- included six studies of varying quality concluded that
sis has further helped to increase interest in this there was insufficient evidence to recommend the
approach for treating ventilated patients with severe utilization of aerosolized antibiotics as either primary
pneumonia. One key lesson learned in the develop- or adjuvant therapy in severe pneumonia [15 ].
&

ment of inhaled therapies for cystic fibrosis has been

the importance of particle size. Particles of the wrong
size will never successfully reach the lower airways REVIEW
and will likely only deposit in the medium and larger Because these earlier meta-analyses, three clinical
airways – and thus will not effectively treat pneumo- trials employing inhaled antibiotics as adjunctive
nia. This is a particular concern in ventilated patients therapy in pneumonia have either been published
because inhaled antibiotic administration essentially or have reported preliminary findings. Although
requires reliance on nebulization. Traditional nebu- conflicting in their general conclusions, these stud-
lization, though, does not insure that one achieves ies indicate that there is little evidence to recom-
the correct particle size. Standard nebulizers addi- mend the broad application of this approach.
tionally are not necessarily designed so as to make The first trial compared not adjunctive inhaled
certain that the drug is administered during the antibiotics but rather inhaled versus intravenous
proper part of the respiratory cycle. Failure to prop- formulations of the same aminoglycoside. In a study
erly time drug delivery with breath administration of 133 postcardiac surgery patients, Hassan et al.

290 www.co-pulmonarymedicine.com Volume 25  Number 3  May 2019

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

 Inhaled antibiotics Schreiber and Shorr

&&
[16 ] concluded that inhaled amikacin (as com- to Hassan et al., there were no differences in any
pared to intravenous amikacin) results in higher secondary endpoints such as earlier liberation from
cure rates when used along with empiric piperacil- the ventilator or ICU length of stay. Finally, there
lin-tazobactam. The authors also state that inhaled was no difference in either mortality or clinical cure
&&
amikacin results in less nephrotoxicity and faster at Day 14 [17 ]. Despite the overall general lack of
liberation from the ventilator. There was no differ- efficacy or impact on any meaningful clinical mea-
ence in mortality between the two cohorts. The sure, the investigators did observe that among 13
study enrolled both patients with hospital-acquired patients with pan-drug-resistant Acinetobacter
pneumonia and VAP and relied on a randomized greater microbiological eradication compared to
&&
design. Importantly, the investigators utilized the placebo [17 ]. This last finding implies that when
inhaled approach early, as soon as an organism was the patient fails to receive initially appropriate ther-
identified, but still reserved aerosolization for drug apy because no such alternatives exist, there may be
resistant organisms. Despite these strengths the trial a role for inhaled antibiotics as a rescue option.
was not blinded and comes from only a single center The largest study exploring inhaled antibiotics
&&
[16 ]. Moreover, they employed a simple pneu- for severe pneumonia finished in mid-2017. The
matic nebulizer for ventilated patients, and more INHALE study compared aerosolized amikacin to
patients in the intravenous treatment arm received placebo in severe Gram-negative pneumonia [18].
initially inappropriate therapy (but this difference This phase III trial was designed based on earlier
represented a trend rather than a statistical differ- results from a smaller phase II study. In the phase II,
ence). Not surprisingly, given the sample size, there placebo controlled trial by Niederman et al. [19],
was no difference in mortality. These factors taken researchers utilized a proprietary formulation of
together limit the generalizability of the trial find- amikacin (400 mg daily or 400 mg twice daily) along
ings, especially since several of their endpoints are with a novel delivery system optimized for the
prone to ascertainment bias and therefore may be creation of appropriately sized particles. The unique
affected by the lack of blinding. More importantly, delivery device was synchronized with the respira-
although the authors claim that inhaled amikacin tory cycle to assure the proper timing for lung
results in fewer days of mechanical ventilation, delivery during appropriate phases of the breath
there was no actual difference in ventilator-free days delivery. This preliminary study enrolled 69 patients
&&
in the cohort [16 ]. This dichotomy suggests there and confirmed that the innovative drug-delivery
was some imbalance in post-ICU mortality. combination system resulted in high levels of ami-
Conversely, in a more rigorously designed, mul- kacin in tracheal aspirates [19]. Intriguingly,
&&
ticenter trial, Kollef et al. [17 ] assessed the effec- although there was no difference in cure rates, indi-
tiveness and safety of inhaled amikacin combined viduals receiving amikacin required fewer intrave-
with fosfomycin. More specifically, Kollef et al. stud- nous antibiotics [19].
ied a dose 300 mg amikacin/120 mg fosfomycin Unfortunately, though, the larger INHALE trial
twice daily for 10 days. Unlike Hassan et al. failed to demonstrate any value to aerosolized ami-
&& &&
[16 ,17 ] this trial utilized a specially designed kacin in severe pneumonia in mechanical ventila-
formulation of amikacin and fosfomycin in order tion patients [18]. This analysis represents the
to insure proper particle size for lung tissue deposi- largest trial on the question of inhaled therapy
tion. They also employed a placebo to facilitate and enrolled over 700 patients from across the globe
comparisons, and the trial was one of the few in [18]. As with the Kollef et al. report, this study was
this area that relied upon double blinding. Unlike both double blinded and placebo controlled. Both
earlier reports, including that by Hassan et al., they trials also only included mechanical ventilation
used a high-efficiency vibrating mesh nebulizer that patients. In contrast to the report by Hassan et al.,
also was devised to help insure target tissue drug INHALE explored the role for nebulized amikacin as
delivery. As noted earlier, utilization of simple jet part of an essentially empiric therapy regimen.
nebulizers often fails to guarantee that sufficient INHALE was intended to demonstrate superiority
doses of antibiotic reach areas of infection in the in survival rates with adjunctive aerosolized treat-
lung. The trial cohort included 142 patients with ment. This may have represented a very high bar for
VAP (as opposed to including HAP) who were documenting the benefit of any innovative inter-
severely ill with many in concurrent shock. Despite vention in disease states such as VAP. VAP, for
multiple efforts to optimize the engineering and example, is associated with high baseline mortality
drug delivery, the investigators found that the rates and many persons die ‘with VAP’ rather than
inhaled therapy had no effect on the primary end- ‘of VAP’. Thus, altering mortality may be nearly
point – a change in the clinical pulmonary infection impossible to show without an exceedingly large
&&
score from baseline [17 ]. Furthermore, in contrast sample size. However, amikacin also failed to show

1070-5287 Copyright ß 2019 Wolters Kluwer Health, Inc. All rights reserved. www.co-pulmonarymedicine.com 291

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

Infectious diseases

any benefit in multiple secondary endpoints to bacteria that alter the target pathogens MIC in vivo.
include: pneumonia-related mortality, early clinical Further work is clearly required to better understand
cure rates, days on mechanical ventilation, and days why the inhaled approach has failed to prove effective
in the ICU [18]. Subgroup analyses, furthermore, did in broader populations given the theoretical appeal of
not identify any relevant population that benefited this paradigm. Additionally, for clinicians wishing to
from nebulized amikacin. Readers should note that rely upon aerosolized rescue therapy, they must rec-
the results of INHALE are currently only available as ognize that effective nebulized antibiotic therapy is
a press release and formal results have not been actually a complex process that requires attention to
presented in a peer reviewed venue (as of yet). detail and a focus on multiple complex steps includ-
One finding consistent across all three trials was ing nebulizer selection, dose selection, and adjust-
the general tolerability of inhaled antibiotics. This ments to ventilator settings. Pending other large
agent seemed associated with minimal toxicity. The trials, it seems prudent to continue to look to inhaled
most commonly reported adverse events were cough antibiotics for use in select cases where other options
&& &&
and bronchospasm [16 ,17 ,18,19]. In instances are very limited.
where it was measured, serum drug levels of the
agents delivered via nebulization were de minims. Acknowledgements
None.
CONCLUSION
Financial support and sponsorship
The most recent guidelines for nosocomial pneumo-
nia from the American Thoracic Society and the None.
Infectious Disease Society of America suggest ‘for
patients with VAP due to gram-negative bacilli that Conflicts of interest
are susceptible only to aminoglycosides or polymyx- M.P.S. has no competing interests. A.F.S. has served a
ins, we suggest both inhaled and systemic antibi- consultant to, speaker for, or received research support
otics, rather than systemic antibiotics alone.’ [13]. from: Achaogen, Astellas, Alios, Aridis, Bayer, Cidara,
This represents a weak recommendation [13]. The Entasys, Melinta, Merck, Nabriva, Paratek, Pfizer, Spero,
most recent clinical trials on this question indicate and Tetraphase
the appropriateness of this recommendation. Broad,
early use of inhaled amikacin when a gram-negative
pathogen is suspected seems inappropriate based on REFERENCES AND RECOMMENDED
&&
the findings of Kollef et al. [17 ,19] and the INHALE READING
Papers of particular interest, published within the annual period of review, have
trial. These two reports also underscore that many of been highlighted as:
the potential benefits of targeted pulmonary antibi- & of special interest
&& of outstanding interest
otic administration are ephemeral at best. Even with
modern nebulizers and devices created to optimize 1. Wunderink RG, Waterer G. Advances in the causes and management of
community acquired pneumonia in adults. BMJ 2017; 358:j2471.
drug delivery and timing of drug release during the 2. Schreiber MP, Shorr AF. Challenges and opportunities in the treatment of
respiratory cycle, no investigators have shown a ventilator-associated pneumonia. Expert Rev Anti Infect Ther 2017;
15:23–32.
benefit when the patient is already receiving ini- 3. Zilberberg MD, Shorr AF, Micek ST, et al. Multidrug resistance, inappropriate
tially appropriate systemic therapy. Rather, as the initial antibiotic therapy and mortality in Gram-negative severe sepsis and
septic shock: a retrospective cohort study. Crit Care 2014; 18:596.
findings of Hassan et al. suggest, particularly when 4. Micek ST, Lang A, Fuller BM, et al. Clinical implications for patients treated
coupled with the results of the subgroup of patients inappropriately for community-acquired pneumonia in the emergency depart-
&& ment. BMC Infect Dis 2014; 14:61.
given inappropriate therapy in the Kollef et al. [17 ] 5. Felton TW, Hope WW, Roberts JA. How severe is antibiotic pharmacokine-
analysis, antibiotic nebulization represents one ticvariability in critically ill patients and what can be done about it? Diagn
Microbiol Infect Dis 2014; 79:441–447.
option for rescue therapy. 6. Tsai D, Lipman J, Roberts JA. Pharmacokinetic/pharmacodynamic considera-
Why have inhaled antibiotics employed early not tions for the optimization of antimicrobial delivery in the critically ill. Curr Opin
Crit Care 2015; 21:412–420.
proven beneficial? It seems that the issue is not related 7. Blot SI, Pea F, Lipman J. The effect of pathophysiology on pharmacokinetics in
to achieving adequate drug delivery to the lower air- the critically ill patient–concepts appraised by the example of antimicrobial
agents. Adv Drug Deliv Rev 2014; 77:3–11.
ways and to the alveoli. On one hand, it may be that 8. Roberts JA, Abdul-Aziz MH, Davis JS, et al. Continuous versus intermittent b-
inhaled drug is incapable of penetrating the mucus lactam infusion in severe sepsis. A meta-analysis of individual patient data
from randomized trials. Am Respir Crit Care Med 2016; 194:681–691.
associated with pneumonia. On the other hand, one 9. Ari A, Atalay OT, Harwood R, et al. Influence of nebulizer type, position, and
can speculate that the MICs measured in a test tube bias flow on aerosol drug delivery in simulated pediatric and adult lung models
during mechanical ventilation. Respir Care 2010; 55:845–851.
may not actually represent the MIC encountered in 10. Alves J, Alp E, Koulenti D, et al. Nebulization of antimicrobial agents in
the pulmonary milieu when the alveoli are filled with & mechanically ventilated adults in 2017: an international cross-sectional sur-
vey. Eur J Clin Microbiol Infect Dis 2018; 37:785–794.
secretions. Put another way, there may be important A large international survey describing current practices regarding the use of
interactions between respiratory secretions and nebulized antibiotics for pneumonia in the ICU.

292 www.co-pulmonarymedicine.com Volume 25  Number 3  May 2019

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

 Inhaled antibiotics Schreiber and Shorr

11. Solé-Lleonart C, Rouby JJ, Chastre J, et al. Intratracheal administration of 16. Hassan NA, Awdallah FF, Abbassi MM, et al. Nebulized versus IV amikacin as
antimicrobial agents in mechanically ventilated adults: an international survey && adjunctive antibiotic for hospital and ventilator-acquired pneumonia post-
on delivery practices and safety. Respir Care 2016; 61:1008–1014. cardiac surgeries: a randomized controlled trial. Crit Care Med 2018;
12. Torres A, Niederman MS, Chastre J, et al. International ERS/ESICM/ESCMID/ 46:45–52.
ALAT guidelines for the management of hospital-acquired pneumonia and A moderate-sized randomized trial of inhaled versus intravenous amikacin along
ventilator-associated pneumonia: Guidelines for the management of hospital- with piperacillin/tazobactam showing a benefit with the nebulized paradigm.
acquired pneumonia (HAP)/ventilator associated pneumonia (VAP) of the Although only conducted in a single center and in a select population, this study
European Respiratory Society (ERS), European Society of Intensive Care showed several clinical advantages of relying on inhaled as opposed to systemic
Medicine (ESICM), European Society of Clinical Microbiology and Infectious aminoglycoside use for pneumonia.
Diseases (ESCMID) and Asociación Latinoamericana del Tórax (ALAT). Eur 17. Kollef MH, Ricard JD, Roux D, et al. A randomized trial of the Amikacin
Respir J 2017; 50:; pii: 1700582. && Fosfomycin inhalation system for the adjunctive therapy of gram-negative
13. Kalil AC, Metersky ML, Klompas M, et al. Management of adults with hospital- ventilator-associated pneumonia: IASIS trial. Chest 2017; 151:
acquired and ventilator-associated pneumonia: 2016 clinical practice guide- 1239–1246.
lines by the Infectious Diseases Society of America and the American One of the few large, multicenter, blinded trials on aerosolized adjunctive therapy
Thoracic Society. Clin Infect Dis 2016; 63:e61–e111. for severe pneumonia in ventilated patients. This analysis revealed no benefit to
14. Zampieri FG, Nassar AP Jr, Gusmao-Flores D, et al. Nebulized antibiotics for this approach except for in a subgroup of patients with pan-drug resistant
ventilator-associated pneumonia: a systematic review and meta-analysis. Crit pathogens.
Care 2015; 19:150. 18. https://www.bayer.us/en/newsroom/press-releases/article/?id=123159.
15. Russell CJ, Shiroishi MS, Siantz E, et al. The use of inhaled antibiotic therapy [Accessed 12 November 2018]
& in the treatment of ventilator-associated pneumonia and tracheobronchitis: a 19. Niederman MS, Chastre J, Corkery K, et al. BAY41-6551 achieves bacter-
systematic review. BMC Pulm Med 2016; 16:40. icidal tracheal aspirate amikacin concentrations in mechanically ventilated
A more rigorous meta-analysis of clinical trials of inhaled antibiotics that suggest that patients with Gram-negative pneumonia. Intensive Care Med 2012;
the current level of evidence is insufficient to support reliance on this approach. 38:263–271.

1070-5287 Copyright ß 2019 Wolters Kluwer Health, Inc. All rights reserved. www.co-pulmonarymedicine.com 293

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.