The roles of adjuvant inhaled and systemic corticosteroids in the treatment of pneumonia
Scott A. Helgeson1, Joseph E. Levitt2, and Emir Festic1,*
1Mayo Clinic, Department of Medicine, Division of Pulmonary and Critical Care Medicine, Jacksonville, Florida, USA and 2 Stanford University, Department of Medicine, Division of Pulmonary and Critical Care Medicine, Stanford, California, USA
Community and health-care associated pneumonia remain leading causes of morbidity and mortality despite appropriate antibiotic therapy. The pneumonia-associated adverse outcomes are not only related to the infectious organism, but also to a dysfunctional host-immune response resulting in overwhelming inflammation. Use of systemic corticosteroids as adjuvant therapy in pneumonia remains controversial. Multiple randomized clinical trials evaluating corticosteroids in patients with community acquired pneumonia have found discrepant results in terms of benefits and adverse effects. Although meta-analyses suggest potential benefits in a select group of patients with more severe pneumonia, the ideal timing, dose, route of delivery, duration, and patient selection remain to be established. Inhaled delivery of corticosteroids offer the potential advantage of providing therapeutic benefits directly to the lung, with minimal to no adverse systemic effects. A smaller body of literature suggests benefit of inhaled corticosteroids, with or without inhaled beta agonists, but future large scale clinical trials are needed to establish clinical benefit with inhaled delivery.
Pneumonia | Corticosteroids
Community-acquired pneumonia (CAP) is the leading infectious cause of hospitalization and death in the United States, costing more than 10 billion dollars in 2015. [1, 2] CAP and hospital associated pneumonia cause roughly 50% of the cases of acute respiratory distress syndrome (ARDS) and septic shock resulting in a high mortality rate in patients requiring admission to the intensive care unit (ICU). [3-5] The mortality rate from pneumonia has not decreased substantially since initial improvement with widespread adoption of antibiotic therapy as standard of care in the late 1900¡¯s.  The high mortality rate of pneumonia may in part be due to the broad range of infectious etiologies including: viral (23%), bacterial (11%), combination of viral and bacterial infection (3%), fungal or mycobacterial (1%), and clinically unidentified pathogens (62%).  However, residual mortality related to pneumonia may also involve host-specific factors including comorbidities, and either an inadequate or overzealous host response. 
Once a pathogen enters the host causing pneumonia, there is subsequent release of inflammatory cytokines that activate the vascular endothelium and recruit neutrophils to help contain the local infection.  However, this host response also disrupts the normally preserved barrier between the alveolar (and thus external environment) and vascular (internal systemic environment) compartments. [9, 10] The loss of this barrier can cause alveoli to flood with protein-rich edema fluid, leading to the loss of lung compliance, refractory hypoxemia, and ultimately ARDS in the case of diffuse lung injury.  In addition, activation of vascular endothelium can promote a marked systemic inflammatory response syndrome (SIRS) leading to shock and multiorgan dysfunction, while increased vascular permeability can promote dissemination of bacteremia. There has been ample research into oral, intravenous, and inhaled anti-inflammatory and/or immunomodulating therapies to combat this dysfunctional host response in an effort to improve mortality in patients with pneumonia. However, despite the current state of evidence, it remains unclear which patient subgroups are most likely to benefit from adjuvant corticosteroids, and what dose, duration, and route of delivery are most beneficial.
Preclinical Studies Suggesting Benefits of Corticosteroid Use in Pneumonia
Multiple preclinical studies have shown beneficial effects of systemic corticosteroids on the dysfunctional host response caused by pneumonia. An in vitro study that injected monocytic human cells with different bacteria (Staphylococcus aureus, Pseudomonas aeruginosa, and Acinetobacter species) found treatment with varying doses of methylprednisolone was able to suppress bacterial replication in a concentration dependent manner.  Inflammatory markers including TNF-¦Á, IL-1¦Â, and IL-6 were also reduced in these cells after treatment with methylprednisolone. In an animal model, Li et al. intratracheally introduced Escherichia coli into mice models and then either treated with different doses of hydrocortisone or placebo and found a decrease in serum levels of IL-6, INF-¦Ã, and nitric oxide levels in the mice that were treated with hydrocortisone.  In a separate model of piglets bronchoscopically inoculated with Pseudomonas aeruginosa, Sibila et al. compared effects of ciprofloxacin alone versus ciprofloxacin plus methylprednisolone.  Pro-inflammatory cytokines were measured in both the bronchoalveolar lavage (BAL) and serum. There was both a decrease in the concentration of IL-6 and lower bacterial cell counts in the BAL fluid of the piglets receiving antibiotics and corticosteroids compared to the group receiving antibiotics alone.
Table 1. Summary of randomized controlled trials on systemic corticosteroids in pneumonia
Abbreviations: LOS, length of stay; CAP, community acquired pneumonia; CRP, C-reactive protein; IV, intravenous; HR, hazard ratio
Clinical Studies of Systemic Corticosteroids in Pneumonia
Multicenter Randomized Controlled trials
Several randomized clinical trials (RCTs) beginning in 2011 trialing different corticosteroid preparations in different degrees of pneumonia severity have found relative benefits from adjuvant-treatment with corticosteroids (Table 1). [15-17] In the initial single-center study from Meijvis et al. in 2011, 304 patients were randomized to receive 5 mg of intravenous dexamethasone daily versus placebo.  The patients had a wide range of pneumonia severity based on a Pneumonia Severity Index (PSI) risk class but no difference in severity between the treatment groups.  Length of stay (LOS) was reduced in the steroid group by 1 day (6.5 versus 7.5; p=0.048) but there was no difference in mortality, and rates of hyperglycemia were significantly increased (44% versus 23%; p<0.001). 
In the largest study to date, Blum et al. performed a multicenter RCT in adults with CAP. The study randomized nearly 800 patients to receive 50 mg of oral prednisone or placebo for 7 days. While not a specific enrollment criterion, the majority of patients were moderate risk (PSI II-IV) and did not require ICU admission with no difference in baseline severity between treatment groups. In steroid treated patients, there was a significant reduction in time to clinical stability, defined as stable vital signs for at least 24 hours (3.0 vs. 4.4 days; hazard ratio 1.33; 95% CI, 1.15-1.50; p<0.001).  Similar to Meijvis et al., there was also a 1 day reduction in hospital LOS in the treated group (6 versus 7 days, p=0.01) without differences in pneumonia related complications, but hyperglycemic events were again increased (OR 1.77; 95% CI, 1.24-2.52; p=0.002). 
In a more recent RCT, Torres et al. performed a multicenter trial of patients with severe CAP and a C-reactive protein level > 150 mg/L, suggesting a strong inflammatory host response.  Severe CAP was defined based on the modified American Thoracic Society criteria or a PSI of V.  These additional enrollment criteria likely contributed to the study requiring over 8 years (compared to 3 and 4.5 years in the prior studies, respectively) to enroll 120 patients. Patients were randomized to receive 0.5 mg/kg of intravenous methylprednisolone every 12 hours or placebo for 5 days.  Treatment failure, defined as development of shock, the need for mechanical ventilation, death, or radiographic progression (¡Ý 50% increase in size of infiltrate), was significantly reduced in the steroid group (OR 0.34; 95% CI, 0.14-0.87; p=0.020).  When treatment failure was separated into early (0-72 hours) versus late (72-120 hours) in a post-hoc analysis, there was no difference in early treatment failure between the groups (6 out of 61 in the steroid group versus 6 out of 59 patient in placebo group; 95% CI, -10-11; p=0.95). However, late treatment failure was significantly reduced in the steroid group (2 out of 61 patients in the steroid group versus 15 out of 59 patients in the placebo group; 22% difference; 95% CI, 10-34; p=0.001).  While, the overall improvement in the combined endpoint was driven mainly by the lack of late radiographic progression (15% versus 2%, p=0.07) in the steroid group, late incidences of other composite outcomes were low, possibly resulting in non-significant trends in reduced rates of late respiratory failure (2% versus 5%, p=0.11), mechanical ventilation (2% versus 7%, p=0.20), and septic shock (0% versus 7%, p=0.06). However, time to clinical stability, ICU and hospital LOS, and in-hospital mortality were not different between the groups limiting definitive interpretation of trial results in regard to more clinically-relevant endpoints.  In addition, despite the stated inclusion criterion of severe CAP, nearly one third of the enrolled patients were in PSI class I-III at the baseline (30% in treatment and 24% in placebo groups) and may not have routinely required hospitalization by current guidelines, potentially diluting treatment effects. 
Not all RCTs have shown benefit from adjuvant corticosteroids in pneumonia. Snijders et al. randomized 213 patients with CAP over a 3 year period to receive either 40 mg prednisolone daily or placebo for 7 days.  The severity of CAP was not an inclusion criterion but the majority of the patients were in the mild to moderate risk class by PSI scores. There was no difference between the groups in the primary endpoint of clinical cure, defined by resolution or improvement of signs and symptoms of CAP, and there was a trend towards more ICU admissions in the treatment group (14.4% versus 6.4%; p=0.06). However, C-reactive protein levels were higher in the treatment group (258.5 mg/L vs. 214.5 mg/L; p=0.03) suggesting that these patients may have been more severely ill at baseline and patients in the treatment group did have a more rapid decline in C-reactive protein. In contrast to Torres et al., more steroid-treated patients had worsening of symptoms after 7 days (OR 2.36; 95% CI, 1.05-5.31; p=0.04), suggesting a possible delay in negative sequelae due to blunting of the initial SIRS response in this overall lower-risk population. 
Combined, these trials suggest a possible benefit from treatment with corticosteroids in patients with more severe CAP due mainly to a decrease in hospital length of stay while also slightly increasing rates of hyperglycemia. Importantly, these trials have not shown any reduction in mortality or respiratory failure with corticosteroids as adjuvant therapy. The heterogeneity of patients in these trials and the inability to properly define high-risk patients (prognostic enrichment) or identify the most inflamed patients that would benefit most from corticosteroid therapy (predictive enrichment) likely led to these studies being under-powered to find benefits in mortality and other clinically-meaningful benefit. 
Secondary Analyses of Clinical Trial Data
Recently, Blum et al. performed a secondary analysis of their original study to better discern which patients might benefit the most from corticosteroid treatment defined by their primary endpoint of time to clinical stability, defined by stabilization of vital signs at two consecutive measurements ¡Ý 12 hours apart.  They assessed interactions between treatment effects and pneumonia etiology (any pathogen identified, bacterial, viral, streptococcus pneumonia, and influenza), antibiotic regimen, and baseline markers of inflammation (fever and procalcitonin). Overall, interpretation was limited by small sample sizes and differences in comorbidities between subgroups added to the limitations of performing multiple comparisons. No significant interactions were found for any subgroup. In the unadjusted analysis, patients with pneumococcal pneumonia did not appear to benefit from steroids, however, after adjustment for pneumonia severity and other comorbidities, differences in time to clinical stability was similar to other subgroups.
A meta-analysis by Siemieniuk et al. of twelve trials and 1,974 patients provided a more nuanced interpretation of these discordant study results. Overall, there was a small but nonsignificant reduction in all-cause mortality (RR 0.67; 95% CI, 0.45-1.01).  Subgroup analysis showed a significant improvement in mortality (RR 0.39; 95% CI, 0.2-0.77; p=0.01) in patients treated with steroids meeting the ¡°severe¡± CAP criteria - defined by a PSI class of IV or V; a confusion, urea nitrogen, respiratory rate, blood pressure, and age 65 years or older (CURB-65) score of ¡Ý 2 ; 1 major or 3 minor criteria per the 2007 Infectious Disease Society of America and American Thoracic Society consensus guidelines; or a British Thoracic Society score ¡Ý 3. [24-26] There was also a significant reduction in the need for mechanical ventilation and development of ARDS among patients treated with steroids. However, the overall methodological strength of the included studies was considered to be low.  One of the included trials, by Confalonieri et al., demonstrated remarkable benefit of corticosteroids towards mortality and was stopped early.  The magnitude of this benefit was considered to be an outlier to the other RCTs. However, even after the exclusion of this trial¡¯s data, there was a significant benefit of systemic corticosteroids towards all-cause mortality in patients with severe CAP in this meta-analysis (RR changed from 0.39; 95% CI, 0.20¨C0.77, to RR 0.51; 95% CI, 0.27¨C0.98, with and without the Confalonieri trial, respectively).
A more recent meta-analysis attempted to address the limits of heterogeneity among trials in the prior meta-analysis of aggregate data by performing an individual patient meta-analysis. Nine completed trials (all included in Siemieniuk et al.) were identified, but individual patient data was only available for 6 trials and 1,509 patients. Overall, there was no difference in the primary outcome of mortality (5.0% versus 5.9%, p=0.24 and 5.0% versus 5.2%, p=0.61, with and without the Confalonieri trial, respectively. Time to both clinical stability (-1.03 days; 95% CI, -1.62 - -0.43; p<0.001) and hospital length of stay (-1.15 days; 95% CI, -1.75 - -0.55; p<0.001) were reduced by approximately 1 day in patients receiving corticosteroids.  However, there was also an increased risk of hyperglycemia (adjusted OR of 2.15; 95% CI, 1.60-2.90; p<0.001) and an increased risk of CAP-related re-hospitalizations (adjusted OR of 1.85; 95% CI, 1.03-3.32; p=0.04). No significant effect modifications were found for subgroups although trends existed for greater effects in more severe pneumonia. In addition to a meta-analysis of patient level data, this review differed from Siemieniuk et al. by omission of 5 small trials totaling 255 patients and 2 larger trials (239 patients) conducted in the 1970¡¯s.
In the most recent systematic review and meta-analysis for the Cochrane Database, Stern et al. included 17 RCTs totaling 2264 patients (including 4 RCT¡¯s with 310 children) and found that only three trials had low reporting bias while the remaining 14 trials had high bias.  Corticosteroid therapy reduced overall mortality (RR 0.58; 95% CI, 0.40-0.84; moderate quality evidence) in adults with severe pneumonia but not in non-severe pneumonia (RR 0.95; 95% CI, 0.45-2.00). Corticosteroids also improved early clinical failure (defined as death from any cause, radiographic progression, or clinical instability at days 5 to 8) in all patients regardless of pneumonia severity (RR 0.32; 95% CI, 0.15-0.7; high quality evidence). Similar benefits were also seen for children with bacterial pneumonia. Corticosteroid therapy was associated with hyperglycemia (RR 1.72; 95% CI, 1.38-2.14), but there were no differences for other adverse effects or secondary infections (RR 1.19; 95% CI, 0.73-1.93). While no patient level data was available, there was no association between treatment effect and proportion of patients with shock, COPD, pneumococcal pneumonia, or pneumonia caused by an atypical bacteria or virus in individual trials. However, there was a significantly smaller effect with increasing mean age of the study population. This review with inclusion of additional studies and patients, likely benefited from greater power to detect significant differences in the severe pneumonia subgroup. Despite the authors rating the overall quality of evidence as moderate to high, 11 of the 17 trials were single center, 9 were open-label, and 14 were assessed as high-risk for selective reporting bias due to inadequate blinding of participants and personnel. Whether inclusion of smaller and lower quality studies provides more precision about the true treatment effect of adjuvant corticosteroids is unclear.
Corticosteroids in Viral Pneumonia
The Blum et al. secondary analysis was the first study to suggest a potential benefit of corticosteroids in viral pneumonia. To date, there are no RCTs specifically investigating corticosteroids in viral pneumonia and a recent Cochrane review could only identify low quality observational studies with 1917 individuals addressing adjuvant corticosteroids as treatment in influenza pneumonia.  This systematic review suggested an increased odds of mortality (259 out of 744 patients in the corticosteroid group vs 165 out of 1173 patients in the control group; OR 3.06; 95% CI, 1.58-5.92) in patients with influenza pneumonia treated with corticosteroid.  An increase in hospital-acquired infections was also seen in the steroid treated patients, but differences in severity of illness could not be adequately addressed and the data was judged to be of very low methodological strength.  Recently, Moreno et al. performed a secondary analysis of an earlier multicenter observational study to determine the effect of corticosteroids in severe influenza pneumonia.  This study included 1846 patients who were admitted to the ICU with influenza pneumonia and treated with relatively higher doses of corticosteroids (median daily dose equivalent to 80 mg of methylprednisolone) for a mean duration of 7 days.  Based on a propensity score matching analysis, corticosteroid use was associated with an increase in ICU mortality (27.5% versus 18.8%; HR 1.32; 95% CI, 1.08-1.60; p<0.006). 
While it is not possible to draw definitive conclusions from even large observational studies, there is rational biologic plausibility to explain these potential discrepant results for adjuvant corticosteroids in viral versus bacterial pneumonia. Effective antimicrobial therapies are lacking for viral infections compared to our armamentarium of antibiotics for bacterial pneumonia. Corticosteroids may provide benefit for bacterial infection by preventing negative systemic sequelae of a robust SIRS response while still allowing for effective bacterial clearance with potent antibiotics. On the other hand, a blunted host response may allow greater viral replication and dissemination resulting in greater systemic organ involvement and late-onset complications when treating viral pneumonia, and more specifically, influenza. This theoretical difference also exists, likely to an even larger extent, for fungal infections, although the lower incidence of fungal pneumonias in trials and observational studies precludes rigorous comparisons.
Studies with Inhaled Delivery of Corticosteroids
There is a theoretical concern that systemic corticosteroids may have variable anti-inflammatory effects in the alveolar compartment, and that inhaled delivery of corticosteroids may provide more effective anti-inflammatory benefits directly to the target organ while limiting negative systemic effects, such as, hyperglycemia and immunosuppression. (Figure 1) Both systemic and inhaled corticosteroids affect the same inflammatory mediators, but there may be an earlier and more potent effect in the lung compartment with inhaled delivery. A recent study showed that after instillation of endotoxin into the lungs of healthy patients, 40mg of intravenous dexamethasone markedly reduced anti-inflammatory markers in the peripheral blood, while only slightly affecting IL-6, IL-8, and TNF-¦Á in bronchoalveolar lavage fluid.  Therefore, it is plausible that inhaled administration of corticosteroids would result in more timely and potent anti-inflammatory effect in the lungs than after the systemic administration of corticosteroids. This, however, needs to be contrasted to the long-term administration of inhaled steroids for chronic obstructive pulmonary diseases, where a higher incidence of pneumonia requiring hospitalization has been observed, but without an increase in mortality. [33, 34].
Figure 1. Key pathways in progression of lung injury. Infection causes neutrophils to release inflammatory cytokines IL-6, IL-8, and TNF-¦Á which in turn causes epithelial injury (marked by increased levels of RAGE and SP-D) and vascular permeability (marked by increased ANG-2). There is potential for inhaled corticosteroids (ICS) to reduce this inflammatory response and limit epithelial permeability while inhaled beta agonists (IBA) could increase alveolar fluid clearance (AFC) and preserve vascular barrier function.
Multiple studies involving experimental models of lung injury have demonstrated amelioration of inflammation, histologic injury, and derangement in oxygenation, hemodynamics, and lung mechanics in animals treated with inhaled corticosteroids compared to placebo. These beneficial effects were demonstrated in multiple studies despite being heterogeneous in several important aspects. Walther et al. studied effects of nebulized beclomethasone starting 30 minutes after infusion of live Staphylococcus aureus in mechanically ventilated pigs.  Oxygenation, hemodynamics, and lung mechanics were better maintained in pigs receiving nebulized beclomethasone after onset of sepsis relative to the animals receiving placebo. Forsgreen et al. studied effects of prophylactic treatment with an aerosolized corticosteroid in a porcine model of early ARDS induced by Escherichia coli sepsis.  Pretreatment with aerosolized corticosteroid, either 15 minutes or 2 hours before the start of an endotoxin infusion, improved lung mechanics and hemodynamics relative to placebo. Moreover, the administration of aerosolized corticosteroid did not reduce endogenous cortisol production. Jansson et al. also investigated the effect of pretreatment with aerosolized corticosteroid on lung injury and inflammation following varying doses of intratracheally instilled lipopolysaccharide (LPS).  Pretreatment with aerosolized budesonide decreased the production of inflammatory cytokines TNF-¦Á, IL-1¦Â, and monocyte chemoattractive protein-1 after challenge with both low and high concentrations of LPS. Greater difference in pulmonary edema and development of acute lung injury (ALI) were seen at higher LPS concentrations suggesting that therapeutic effects of aerosolized budesonide were greater with more severe disease.
Timing of administration of therapy has varied across multiple studies. Wang et al. examined the effect of varying the time window between injury and treatment with a nebulized corticosteroid on lung injury induced by chlorine gas inhalation.  Nebulized budesonide was given immediately, 30, or 60 minutes following chlorine gas exposure. The mean arterial oxygen tension in animals with treated immediately and 30 minutes post-delivery of budesonide was significantly higher than in controls or animals treated 60 minutes post-delivery. However, the recovery of lung compliance was enhanced and the pulmonary wet-to-dry weight ratio was lower in all three budesonide groups relative to controls suggesting that while earlier is better, benefit may still occur with delayed treatment. Despite inherent heterogeneity, the above studies demonstrate relatively consistent beneficial effects of aerosolized corticosteroids in animal models of lung injury caused by either inhaled or systemic insults with different mechanism of the inciting injury and timing of treatment.
Not all preclinical studies of inhaled corticosteroids in experimentally induced lung injury and pneumonia resulted in improvement relative to the sham inhalation. For example, Sjoblom et al. randomized 16 rabbits into groups receiving inhalation of ammonia, followed by 30 minutes and 2 hours delayed treatments with 0.5 mg inhaled budesonide or placebo.  The ammonia inhalation resulted in severe ALI after only 15 minutes. In comparison with placebo, budesonide did not result in improved gas exchange or reduced airway pressure levels during the observation period. This result may exemplify the limitations of delayed treatment with a relatively low dose of inhaled corticosteroid in the setting of a potent inciting injury. While difficult to extrapolate differences in animal models to clinical relevance, the preclinical data suggests that inhaled corticosteroids are likely to be most beneficial when given early before onset of more severe lung injury.
Role of Beta Agonists
Beta agonists in experimental studies have demonstrated the ability to preserve pulmonary vascular stability and up-regulate alveolar fluid clearance. Spindler and Waschke tested whether ¦Â adrenergic receptor signaling contributes to the maintenance of baseline endothelial barrier properties.  By using postcapillary venules of rats in vivo and cultured microvascular endothelial cells, they compared effects of propranolol and epinephrine and concluded that activation of ¦Â adrenergic receptor signaling contributes to the stability of endothelial barrier properties under baseline conditions. An extensive physiological review by Matthay et al. provides a detailed discussion of the in vivo and in vitro experiments that have enhanced our understanding of the complex regulation of lung fluid balance by active transport mechanisms across both the alveolar and distal airway epithelium.  Pertinent to our topic, the authors commented on the potential beneficial effects of both beta agonist and corticosteroids on the reabsorption of edema from the distal airspaces of the lung, during physiological and pathological conditions.
The BALTI trial was the first trial to suggest that beta agonist therapy improves alveolar edema resorption in patients with ALI/ARDS.  This trial used intravenous salbutamol versus placebo in patients with ALI/ARDS and showed that treated patients had lower lung water (9.2 mL kg-1 versus 13.2 mL kg-1; 95% CI difference, 0.2-8.3 mL kg-1; p=0.038) and plateau airway pressure (23.9 cm H2O versus 29.5 cm H2O; p=0.049) after 7 days of treatment. However, five years later, the BALTI-2 trial showed that the same treatment regimen of intravenous salbutamol did not decrease mortality in ARDS patients and was actually stopped early because of a safety concerns.  Concurrently, the ALTA trial, which randomized patients with ALI to either 5 mg of aerosolized albuterol or placebo for up to 10 days, was also stopped early for futility after finding no difference in ventilator free days and hospital mortality between the treatment groups on an interim analysis.  The ALTA trial did show that systemic absorption of aerosolized albuterol in patients with ALI/ARDS was sufficient based on measurements of plasma albuterol levels. A common factor of these negative trials was that the enrolled patients were already mechanically ventilated for fully established ALI/ARDS, so the therapeutic window for potential effects of beta agonist therapy might have been missed.
The only RCT to test whether pretreatment with beta agonists might prevent ALI was performed in patients undergoing elective esophagectomy in 12 centers in the United Kingdom. Although the treatment with salmeterol 100 µg twice daily did not prevent early lung injury (32 [19.2%] of 168 vs. 27 [16.0%] of 170; OR 1.25; 95% CI, 0.71¨C2.22), the patients receiving salmeterol experienced a lower incidence of postoperative pneumonia (7 vs. 17; OR 0.39; 95% CI, 0.16¨C0.96). 
Combined Use of Inhaled Corticosteroids and Beta Agonists
In a secondary analysis of the Lung Injury Prediction Score (LIPS) cohort, patients receiving inhaled corticosteroids prior to hospitalization (two thirds of whom were receiving concomitant inhaled beta agonists) had a reduced risk of developing ARDS based on the multivariate logistic regression model (OR 0.56; 95% CI, 0.30-0.98; p=0.045). The only other variables from this model significantly associated with the development of ARDS were the LIPS and APACHE II scores. The protective benefit seemed to have been more pronounced in patients with risk factors for direct lung injury (mainly pneumonia). However, following adjustment in a propensity-matched analysis, pre-admission use of inhaled corticosteroids in a hospitalized population of patients at-risk for ALI was no longer significantly associated with a lower incidence of ALI. The overall direction, magnitude, and stability of the association suggested a non-statistically significant benefit. 
In a population-based study from Olmsted county, Minnesota, United States, the authors showed that inhaled beta agonists taken prior to hospitalization provided a similar protective effect against development of ARDS (OR 0.48; 95% CI, 0.31-0.72; p<0.001).  Sixty-three percent of patients receiving inhaled beta agonists were also receiving inhaled corticosteroids. This protective effect was more pronounced in patients with pneumonia previously receiving inhaled beta agonists (OR 0.50; 95% CI, 0.34-0.74; p<0.001) compared to patients without pneumonia (OR 0.72; 95% CI, 0.46-1.13; p=0.15).
There is also substantial data that suggests a potential synergistic benefit of combined inhaled corticosteroid and beta agonists in patients with obstructive pulmonary disease. The bronchodilating effect of inhaled beta agonists may improve distal delivery of inhaled corticosteroids to the alveoli, while inhaled beta agonists have also been shown to increase the efficacy of inhaled corticosteroids by increasing the nuclear localization of the corticosteroid receptors.  Moreover, the timely delivery of inhaled beta agonists was shown to enhance alveolar epithelial barrier stability, thus improving its function. Both inhaled corticosteroids and beta agonists are medications broadly available, inexpensive, and with established safety profiles when used in combination.
In a recent multicenter RCT, the authors studied the effect of combined inhaled corticosteroid and beta agonist therapy in patients at high risk of developing ARDS, defined by a LIPS ¡Ý 4.  Sixty patients were randomized to receive inhaled budesonide/formoterol versus placebo for up to 5 days. The primary outcome was longitudinal change in oxygen saturation divided by the fraction of inspired oxygen (S/F) through day 5. The categorical change in S/F by a greater than 20% from baseline was also analyzed, which was assumed to be clinically significant. This study showed that patients at risk of ARDS randomized to receive inhaled budesonide and formoterol, compared to those receiving placebo, had improved oxygenation based on more rapid improvement in the S/F (p = 0.02), with significant separation occurring at day 2. No patient in the treatment group had a decrease in S/F compared with eight patients (27%) in the placebo group. However, the benefit appeared to be restricted to patients with pneumonia (p=0.03) versus those without pneumonia (p=0.51).  Overall, patients in the treatment arm had lower rates of mechanical ventilation (53% vs. 21%, p = 0.01) and ARDS (23% vs. 0%, p = 0.01), as well as a shorter hospital and ICU length of stay. There was an imbalance in baseline severity of illness with more patients having baseline shock in the placebo group. While shock was not associated with a change in S/F and did not modify the effect of treatment on S/F, it cannot be ruled out that some differences in outcomes may have been due to differences in baseline covariates in this small trial. It is important to stress that the median time from the Emergency Department presentation to the first dose of study drug was less than 9 hours. Therefore, it is likely that early treatment with inhaled medications is crucial in order to provide protective effects prior to development of more severe lung injury (as in larger trials of mechanically ventilated patients with ARDS). Beta agonists are likely to be most effective early in the course of developing lung injury when the alveolar epithelium and epithelial and endothelial barriers are still relatively maintained and thus more responsive to up-regulation of edema resorption. Similarly, smaller reductions in pulmonary edema may have enhanced benefits for the prevention of respiratory failure relative to improving outcomes once intubation and mechanical ventilation have occurred.
These studies highlight the need for future research in adjunctive treatment with inhaled therapies in pneumonia. Some trials using systemic corticosteroids have shown an increase in rates of hyperglycemia and a possible increase in delayed complications such as secondary infections. The use of short-term, adjuvant inhaled corticosteroid therapy would not be expected to yield these complications. Further studies are needed to determine whether the adjuvant treatment with combination of inhaled corticosteroids and beta agonists may provide additional benefit. The potential synergy between the anti-inflammatory effects of corticosteroids and increased capillary stability and edema resorption from beta agonists combined with timely inhaled delivery directly to the target organ prior to onset of respiratory failure has potential to maximize clinical benefit while limiting negative systemic side-effects.
The consensus guidelines created by the Society of Critical Care Medicine and European Society of Intensive Care Medicine in 2017 recommended using corticosteroids for 5-7 days at a daily dose less than 400 mg of hydrocortisone IV for all patients hospitalized with CAP. In this recommendation, they mention that patients with severe disease derived more benefit, but that all hospitalized CAP patients should be given corticosteroids.  The consensus guidelines for CAP by the Infectious Disease Society of America and American Thoracic Society (IDSA/ATS) published in 2007 did not recommend adjunctive corticosteroids to be used in patients with CAP due to lack of sufficient evidence.  However, these guidelines are currently being updated and it is expected that based on the more recent evidence, use of adjunctive corticosteroids in CAP will be addressed. It is our opinion that the level of recommendation will be weak and level of confidence low. Moreover, data on the use of corticosteroids in hospital acquired pneumonia is lacking and the current supporting evidence is extrapolated from the trials on patients with CAP.
Future studies should address what patients are likely to derive the most benefit from adjuvant corticosteroid therapy. CRP, as used in the Torres et al. trial is widely available and adjusting the dose and duration of corticosteroids based on initial CRP level and subsequent CRP level in response to therapy is a possibility. There is also a necessity to better understand the host response and to identify other biologic markers of inflammation that may or may not already be clinically available. There is much patient heterogeneity in all the trials referenced above, so the inconsistent results may be a result of the one size-fits-all approach. There is also a need to identify factors that increase the host¡¯s inflammatory response, such as specific bacterial and viral pathogens, comorbidities, and genetic factors. The ability to differentiate patients that have more inflammation and who may benefit from corticosteroids from those that have minimal inflammation and may be harmed could provide both predictive and prognostic enrichment for future trials.
In summary, there may be a role for the use of corticosteroids as adjuvant therapy in patients with severe pneumonia, but their role still needs to be better defined. Despite several RCT¡¯s suggesting benefit in CAP, at the present time, adjuvant corticosteroids are not considered standard of care in patients either with CAP or hospital acquired pneumonia and hence they are not routinely administered. We suspect that even updated societal guidelines recommending using of adjuvant corticosteroids are unlikely to have significant uptake in clinical practice, especially in the United States where multicenter RCTs are lacking. Stronger RCT evidence from appropriately powered trials to confirm benefit in harder endpoints like mortality will likely be needed to change practice. Given the remaining equipoise for this important question despite multiple multicenter RCTs, future trials should likely target more novel methodology such as cluster randomization with waiver of consent to allow timely randomization of several thousand patients stratified by disease severity to more definitively address overall mortality and heterogeneity of treatment effects in subgroups of pneumonia patients. In addition to ensuring timely (early) delivery, future studies need to discern which patients would be the most amenable to the beneficial effects relative to their host characteristics, severity, and etiology of pneumonia, along with best route of delivery, dose, and duration of therapy. Currently, the strongest data supports use of corticosteroids for severe bacterial pneumonia. Therefore, it is a reasonable approach to use systemic corticosteroids early in the hospital course in patients with suspected severe bacterial pneumonia and/or those admitted to the ICU and found to have strong inflammatory response. Whether some clinical measures of the disease severity or biologic measures of the host inflammatory response, such as CRP, provide predictive enrichment relative to clinical measures of pneumonia severity also remains to be determined.
15. Blum, C.A., et al., Adjunct prednisone therapy for patients with community-acquired pneumonia: a multicentre, double-blind, randomised, placebo-controlled trial. Lancet, 2015. 385(9977): p. 1511-8.
16. Meijvis, S.C.A., et al., Dexamethasone and length of hospital stay in patients with community-acquired pneumonia: a randomised, double-blind, placebo-controlled trial. The Lancet, 2011. 377(9782): p. 2023-2030.
17. Torres, A., et al., Effect of corticosteroids on treatment failure among hospitalized patients with severe community-acquired pneumonia and high inflammatory response: a randomized clinical trial. JAMA, 2015. 313(7): p. 677-86.
19. Salih, W., S. Schembri, and J.D. Chalmers, Simplification of the IDSA/ATS criteria for severe CAP using meta-analysis and observational data. European Respiratory Journal, 2014. 43(3): p. 842-851.
21. Iwashyna, T.J., et al., Implications of Heterogeneity of Treatment Effect for Reporting and Analysis of Randomized Trials in Critical Care. American Journal of Respiratory and Critical Care Medicine, 2015. 192(9): p. 1045-1051.
23. Siemieniuk, R.C., et al., Corticosteroid therapy for patients hospitalized with community-acquired pneumonia: A systematic review and meta-analysis. Annals of Internal Medicine, 2015. 163(7): p. 519-528.
25. Mandell, L.A., et al., Infectious Diseases Society of America/American Thoracic Society Consensus Guidelines on the Management of Community-Acquired Pneumonia in Adults. Clinical Infectious Diseases, 2007. 44(Supplement_2): p. S27-S72.
26. Myint, P.K., et al., Severity assessment criteria recommended by the British Thoracic Society (BTS) for community-acquired pneumonia (CAP) and older patients. Should SOAR (systolic blood pressure, oxygenation, age and respiratory rate) criteria be used in older people? A compilation study of two prospective cohorts. Age and Ageing, 2006. 35(3): p. 286-291.
28. Briel, M., et al., Corticosteroids in Patients Hospitalized With Community-Acquired Pneumonia: Systematic Review and Individual Patient Data Metaanalysis. Clinical Infectious Diseases, 2017. 66(3): p. 346-354.
33. Festic, E. and P.D. Scanlon, Incident pneumonia and mortality in patients with chronic obstructive pulmonary disease. A double effect of inhaled corticosteroids? American journal of respiratory and critical care medicine, 2015. 191(2): p. 141-148.
36. Frosgren PE, M.J., Dahlback CM, Axelsson BI, Prophylactic treatment with an aerosolized corticosteroid liposome in a porcine model of early ARDS induced by endotoxaemia. Acta Chir Scand, 1990. 156(6-7): p. 423-31.
38. Wang, J., L. Zhang, and S.M. Walther, Inhaled budesonide in experimental chlorine gas lung injury: influence of time interval between injury and treatment. Intensive Care Med, 2002. 28(3): p. 352-7.
43. Gates S, P.G., Lamb S, Kelly C, Thickett D, Young D, McAuley D, Snaith C, McCabe C, Hulme C, Gao-Smith F, Beta-Agonist Lung Injury Trial-2 (BALTI-2): a multicentre, randomised, double-blind, placebo-controlled trial and economic evaluation of intravenous infusion of salbutamol versus placebo in patients with acute respiratory distress syndrome. Health Technology Assessment, 2013. 17(38).
44. The National Heart, L. and B.I.A.R.D.S.C.T. Network, Randomized, Placebo-controlled Clinical Trial of an Aerosolized ¦Â2-Agonist for Treatment of Acute Lung Injury. Am J Respir Crit Care Med, 2011. 184(5): p. 561-568.
46. Festic, E., et al., SpO2/FiO2 ratio on hospital admission is an indicator of early acute respiratory distress syndrome development among patients at risk. J Intensive Care Med, 2015. 30(4): p. 209-16.
49. Festic, E., et al., Randomized Clinical Trial of a Combination of an Inhaled Corticosteroid and Beta Agonist in Patients at Risk of Developing the Acute Respiratory Distress Syndrome. Crit Care Med, 2017. 45(5): p. 798-805.
50. Pastores, S.M., et al., Guidelines for the diagnosis and management of critical illness-related corticosteroid insufficiency (CIRCI) in critically ill patients (Part II): Society of Critical Care Medicine (SCCM) and European Society of Intensive Care Medicine (ESICM) 2017. Intensive Care Medicine, 2018. 44(4): p. 474-477.
Conflict of Interest: No conflicts declared.
* Corresponding Author. Emir Festic, M.D., M.S.
Mayo Clinic, 4500 San Pablo Road S, Jacksonville, Florida 32224, USA
Phone: 904-953-6728; Fax: 904-953-2082
© 2019 by the Journal of Nature and Science (JNSCI).