INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Prostaglandin E₂ (PGE₂) is considered to be "immunomodulatory" and predominantly "bronchoprotective" (1). PGE₂ is a dominant cyclooxygenase product of airway epithelium and smooth muscle. PGE₂ aerosols inhibit the early and late asthmatic response to allergen inhalation, and the bronchoconstriction to exercise or aspirin inhalation in sensitive asthmatic patients, while having no effect on baseline FEV₁ and no change in methacholine reactivity (2). This action of PGE₂ differs from the effect of short-acting -adrenergic agents which can blunt the early, but not the late asthmatic response. These findings suggest that PGE₂ has a role in inflammatory mediation, and might be of therapeutic benefit. In vitro studies have shown that PGE₂ inhibits many inflammatory events, including mast cell degranulation, leukotriene (LT) B₄ production by alveolar macrophages, and eosinophil activation (9). Additionally, it has been shown in vitro to have inhibitory effects on neurally mediated airway smooth muscle constriction (12). The mechanisms of action of PGE₂ in allergic asthma, however, are not completely understood.

Our hypothesis to account for the physiologic effects reported, is that PGE₂ inhibits the early-phase allergic asthmatic response through inhibition of mast cell mediator release, in particular prostaglandin D₂ (PGD₂). The acute early bronchoconstriction in sensitized asthmatics to allergen is thought to be dependent on mast cell activation, with the release of histamine, tryptase, PGD₂, and the cysteinyl leukotrienes (LTC₄, LTD₄, LTE₄), mediators that can induce bronchospasm (9, 13- 15). We tested this hypothesis in a randomized, double-blind, placebo-controlled crossover study comparing aerosolized PGE₂ to placebo in a group of patients undergoing bronchoscopy in order to measure cells and mediators affected by pretreatment with nebulized PGE₂.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The study comprised 10 nonsmoking, mild atopic asthmatic subjects age 18 to 43 (mean ± SD = 30 ± 7.4 yr) who had a baseline FEV₁ less than 70% predicted and were controlled by intermittent use of inhaled ₂-agonists. All subjects gave written informed consent approved by the Vanderbilt Institutional Review Board. All subjects had reversible airway disease with  15% increase in FEV₁ to -agonists or a positive methacholine challenge. The specific allergen used for bronchoprovocation was an allergen to which the individual gave an allergic history, was skin test positive, and was out-of-season. All subjects were required to have avoided use of inhaled anti-inflammatory agents, systemic corticosteroids, aspirin, antihistamines, and nonsteroidal anti-inflammatory agents for at least 3 wk before their study date. They were to have avoided short-acting antihistamines for 48 h and theophylline and salmeterol for 5 d prior to their study date.

Inhalation of 100 µg of PGE₂ was accomplished by having the volunteers breathe PGE₂ via a PARI-LC Plus nebulizer (PARI Respiratory Equipment, Inc., Richmond, VA) over 5 to 10 min immediately before bronchoscopy. Stock solution of PGE₂ (Pharmacia Upjohn, Puurs, Belgium) was prepared by diluting 5 mg Prostin E₂ (0.5 ml of 10 mg/ml concentration) with 2 ml dehydrated alcohol, resulting in a concentration of 2 mg/ml. The resulting solution was aliquoted into single-use vials of 0.05 ml (100 µg) each, and was stored at 70° C until immediately prior to use, at which time 4.95 ml saline was added, resulting in a concentration of 20 µg/ml for inhalation. The placebo inhalational agent was the same solution as that used to prepare the PGE₂.

Bronchoscopy and Bronchoalveolar Lavage (BAL)

Patients received either inhaled PGE₂ or placebo in a double-blind manner followed by immediate bronchoscopy. BAL of a segment of the lingula was performed as a baseline, followed by lavage of a right middle lobe segment beginning 4 min after endobronchial allergen instillation. BAL was performed with 100 ml of sterile saline in each segment. Patients crossed over to the other arm of the study no sooner than 3 wk after the first study day (16).

A cell count of the BAL fluid (BALF) was immediately performed after bronchoscopy. Wright-Giemsa stains were prepared for differential evaluation of 200 cells and determination of the percentages of macrophages, lymphocytes, neutrophils, mast cells, monocytes, basophils, and eosinophils. The lavage supernatant was stored at 70° C for later determination of eicosanoid concentrations.

Measurements

Prostaglandins and thromboxane B₂ were measured by modified stable isotope dilution assays that used gas chromatography-negative ion chemical ionization-mass spectrometry (GC-NICI-MS) (17).

LTC₄, LTD₄, and LTE₄ were analyzed by enzyme immunoassay using a peptidoleukotriene polyclonal antiserum (Cayman Chemical, Ann Arbor, MI) after prior purification on C18 silica columns. The recovery was determined using 4,000 counts per minute (cpm) of tritiated LTC₄ standard (DuPont NEN Research Products, Boston, MA).

Tryptase and eosinophilic cationic protein (ECP) were measured in 1 ml of BALF by radioimmunoassay (RIA) kits (Pharmacia Diagnostics AB, Uppsala, Sweden), after 10-fold concentration on Centricon microconcentrators (Amicon Inc., Beverly, MA) with a molecular weight cutoff of 10 kD. Histamine was assayed using an enzyme immunoassay kit (Immunotech, Inc., Westbrook, ME).

Data Analysis

The primary purpose of the analysis of study outcome measures was to test within-individual responses. The observed responses were calculated as before-to-after changes (response to endobronchial allergen challenge), and as differences in these changes between placebo and PGE₂ treatments. Responses were tested for normality, and power transformations (including logarithmic and square-root transformations) were applied to improve normal approximations where needed. Square-root transformations were adequate for most outcome measures; in these cases, parametric methods were used for hypothesis testing. For cases in which power transformations were inadequate, nonparametric rather than parametric methods were used.

The primary a priori hypothesis for this study is that the increase in PGD₂ and its metabolite PGF₂ from baseline values to after allergen would be less in the PGE₂ arm than in the placebo arm. A secondary a priori hypothesis is that the increase in preformed mast cell mediators, histamine and tryptase, from baseline values to postallergen in the PGE₂ arm of the study is less than in the placebo arm. Additionally, other mediators of the early asthmatic response were measured for exploratory analyses. The parametric paired t test and the nonparametric exact Wilcoxon signed-rank test were used for hypothesis testing. One-sided p values were calculated for one-sided hypotheses. For hypotheses involving multiple tests, Bonferroni's correction was applied to the otherwise designated significance level: alpha = 0.05. All analyses were performed using Stata statistical software version 6 (Stata Corporation, College Station, TX), S-PLUS software (MathSoft, Seattle, WA), and MathSoft (MathSoft, Inc., Seattle, WA). Means are reported ± SD; a value of p less than 0.05 was considered statistically significant.

One patient receiving placebo had baseline LTC₄, LTD₄, and LTE₄ concentrations equal to 1,285 pg/ml, which was both 25 times the next highest value and that subject's other baseline value. As this was a crossover study, this individual had baseline values in both arms of the study, and although there were no significant differences between baseline values in the other nine study subjects, this individual had baseline leukotriene concentrations significantly different at baseline between the two study days, representing a 25-fold difference. Additionally, the baseline level in this individual is on the order of magnitude of 25 times higher than reported baseline BALF leukotriene concentrations in other human asthma studies (15, 18). This subject was excluded from the analyses.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Subject Description

PGE₂ was well tolerated by the 10 subjects enrolled, with no serious adverse events. Cough was present in 60% of subjects receiving inhaled PGE₂ and 40% of subjects receiving placebo. The subjects ranged in age from 18 to 43 yr (mean age, 30 ± 7.4 yr); 40% were female, 20% African American, 80% white. The mean baseline FEV₁ was 3.15 ± 0.82 L. There were no significant differences between the baseline FEV₁ on the PGE₂ day compared with the control day (mean difference = 0.094 ± 0.45; p = 0.5). Fifty percent of subjects had an increase in FEV₁ after inhaled PGE₂ (mean increase, 0.38 ± 0.30 L). Seventy percent had a decrease in FEV₁ after placebo inhalation. The median number of days between study days was 21 (range, 19 to 82 d).

BALF Cell Counts and Differentials

There were no significant differences in cell counts at baseline values between the placebo and PGE₂ arms. There was a nonsignificant decrease in total cell counts (p = 0.07) and a significant decrease in eosinophil counts (p = 0.03) in those subjects receiving PGE₂ after endobronchial allergen challenge (Figure 1). Total baseline cell counts were 10.97 ± 4.59 × 10⁴ and after allergen challenge 12.56 ± 4.61 × 10⁴ in the placebo group, and 8.03 ± 4.54 × 10⁴ at baseline values and 6.94 ± 3.89 × 10⁴ after allergen in the group treated with PGE₂. There were no significant differences in neutrophil or lymphocyte counts between those receiving placebo versus PGE₂ after endobronchial challenge.

View larger version (22K): [in this window] [in a new window]   Figure 1.   Change in total cell counts (×104) and eosinophil cell counts (×104) in BALF from baseline to after endobronchial allergen challenge in the placebo and PGE2 arms of the study ± SEM. There was a nonsignificant reduction in total cell count change (p = 0.07), and a significant reduction in eosinophil count change (p = 0.028) in those subjects receiving PGE2 after endobronchial allergen challenge. Total baseline cell counts were not significantly different at baseline values between the groups.

Arachidonic Acid Metabolites and Mast Cell Mediators

There were no significant differences in baseline values of the mediators measured: tryptase, histamine, cysteinyl leukotrienes, PGD₂, PGF₂, PGF₂, and thromboxane B₂, between the placebo and PGE₂ arms of the study. As expected, there were significantly greater quantities of PGE₂ at baseline values in those subjects receiving nebulized PGE₂ compared with those receiving nebulized placebo (mean difference = 24.2 ± 27.6; p = 0.025). After endobronchial allergen challenge in those treated with placebo there were significant increases in PGD₂, PGF₂, cysteinyl leukotrienes, tryptase, and ECP.

After endobronchial allergen challenge, those treated with PGE₂ had significantly lower PGD₂ concentrations (p = 0.014). Nebulized PGE₂ also decreased cysteinyl leukotrienes by 64% after allergen challenge, which was not significant (Figure 2). There was an 80% reduction in the PGD₂ metabolite, PGF₂, which did not reach statistical significance (p = 0.027 with Bonferroni correction using alpha = 0.025 for statistical significance) (Table 1). There were no significant differences between concentrations of tryptase, histamine, ECP, thromboxane B₂, and PGF₂ after endobronchial allergen challenge in those receiving PGE₂ compared with those receiving placebo.

View larger version (22K): [in this window] [in a new window]   Figure 2.   The dot plot shows data points for each subject during their control period at baseline values and after allergen instillation (left), and for each subject during their crossover period after administration of inhaled PGE2 before bronchoscopy at baseline and after allergen administration (right). A line connects each subject's baseline and post- allergen instillation values for each arm of the study. (A) Nebulized PGE2 significantly reduced PGD2 in BALF after allergen challenge (p = 0.014). (B) Nebulized PGE2 reduced the cysteinyl leukotrienes in BALF after allergen challenge by 64% (p = 0.049).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study demonstrates that PGE₂ can be safely delivered to the airways by aerosol, with higher BALF concentrations of PGE₂ after inhalation, than after placebo. This study also demonstrates for the first time that PGE₂ inhalation suppresses mast cell release of prostaglandin metabolites, resulting in significantly lower concentrations of PGD₂ after antigen stimulation, and a nonsignificant 80% reduction in PGD₂ metabolite, PGF₂. The degree of suppression of PGD₂ significantly correlates with BALF concentration of PGE₂ post- allergen challenge, supporting the hypothesis that PGE₂ acts to suppress PGD₂ release. In addition, there was a nonsignificant 64% reduction in the rise in cysteinyl leukotrienes in BALF after allergen challenge in those receiving PGE₂.

There was a significant increase in PGD₂ (p = 0.01), cysteinyl leukotrienes (p = 0.04), and tryptase (p = 0.02) after endobronchial allergen challenge on the placebo day, indicating mast cell activation. It appears that newly generated mast cell mediator release (e.g., PGD₂) may be controlled by PGE₂. Others have shown that preformed mast cell mediator release is also inhibited by PGE₂, supporting its role in mediation of mast cell function (19). Although the post-allergen challenge rise in preformed mast cell mediators (e.g., tryptase and histamine) was not significantly inhibited in those receiving PGE₂, small numbers and large variations in BALF concentrations might account for this. Alternatively, PGE₂ may have differential activity on newly generated mast cell products, compared with release of preformed mast cell products. Differential activity on the mast cell is reasonable to hypothesize, as ₂-agonists are potent in inhibiting mast cell degranulation, but are much less active in preventing the release of newly generated mediators such as PGD₂ (20). Additionally, there was a 64% decrease in leukotriene concentrations, suggesting that prostanoids may impact leukotriene production or appearance. Data on canine kidney cells suggest that phospholipase A₂ can be inhibited by increased adenosine 3',5'-cyclic monophosphate, and that this can be increased by PGE₂ (21). This could be hypothesized to be a mechanism for reduction in PGD₂ production.

We presume that PGD₂ is coming predominantly from the mast cell, although it is certainly possible that PGE₂ may mediate the asthmatic response via other cell types, such as the alveolar macrophage. Prior studies have demonstrated that thromboxane is a good marker of alveolar macrophage stimulation, and thromboxane concentrations were not significantly changed after endobronchial allergen challenge in those receiving PGE₂ (18). This provides indirect evidence that activation of another cell accounts for PGD₂ appearance, likely the mast cell.

PGE₂ inhalation results in significantly lower appearance of eosinophils in the airways of atopic asthmatics after allergen challenge compared with placebo. Whether PGE₂ affects eosinophil recruitment or accumulation in the BALF related to cellular detachment from airway surfaces cannot be answered in this study. Leukotrienes that are involved in eosinophil recruitment were reduced in subjects receiving PGE₂, although not significantly. The time between PGE₂ delivery and subsequent bronchoscopy was short, and it does not seem plausible to hypothesize that PGE₂ had an impact on eosinophil recruitment. However, it could be hypothesized that PGE₂ has an impact on eosinophil accumulation or adhesion and therefore on the appearance of eosinophils in BALF. We hypothesize that PGE₂ inhibits the late allergic bronchoconstriction via inhibiting cellular mechanisms for eosinophil appearance. One additional basis for this hypothesis is that PGE₂ will inhibit the migration of eosinophils into an area of passive cutaneous anaphylaxis, and has some direct inhibitory effect on the eosinophil (22, 23). Measurement of cytokines that might have been impacted by PGE₂, such as interleukin-5 (IL-5), or assessment of mechanisms that interfere with transendothelial migration might shed light on the mechanism of action of PGE₂ on eosinophil recruitment, but were not assessed in this study. Also, this study only measured airway presence in BALF, which does not necessarily reflect what is happening in the airway wall.

There are obvious problems with such a study model. PGE₂ has a very short half-life, and the time between delivery and endobronchial allergen challenge likely decreases the efficacy of the drug. In our study, there was a nearly 40% decrease in PGE₂ concentrations during the 4 min between baseline lavage and post-allergen lavage in those receiving PGE₂. Given the short half-life of PGE₂, it seems intuitive that the shorter the time from PGE₂ administration to allergen challenge, the greater the effect of the drug. An improved study model might therefore use direct instillation of PGE₂ into the airways, followed by allergen challenge. Additionally, the activation or release of certain mediators is likely time-dependent, with various mediators peaking at particular time points. Wenzel and colleagues, in a similar study model, performed endobronchial allergen challenge followed by BAL at 5 min, 1 min longer than in our study model. After endobronchial allergen challenge they demonstrated significantly higher BALF concentrations of the cysteinyl leukotrienes (15). Additionally, cysteinyl leukotrienes appear to peak in the urine at approximately 2 h (24). In our study, cysteinyl leukotrienes and preformed mast cell mediators were not significantly altered by PGE₂ administration, although cysteinyl leukotrienes were reduced by 64%. It may be, therefore, that small numbers and time from administration to endobronchial allergen challenge may impact the ability to show significant differences and the efficacy of the drug, respectively. Additionally, the study is underpowered to address mechanisms of action of PGE₂ on inflammatory mediators other than the a priori hypothesis that PGE₂ acts through suppression of PGD₂ release.

PGE₂ has been shown to affect airway physiology, inhibiting both the early and late asthmatic response, without resulting in significant bronchodilation (6). The early asthmatic response begins almost immediately after exposure to allergen, reaches its peak at 15 to 20 min, and is followed by resolution over the next 1 to 2 h. This process, involving the actions of histamine, PGD₂, and cysteinyl leukotrienes (LTC₄, LTD₄, LTE₄) is thought to be largely mast cell dependent (25). Because PGD₂ is thought to be derived principally from the mast cell, our finding of a decrease in PGD₂ and its metabolite PGF₂, supports the idea that PGE₂ is acting by inhibiting mast cell activation. Also, given the physiologic effect on attenuation of the late asthmatic response, it seems more plausible that the mechanisms of action of PGE₂ are anti-inflammatory, rather than bronchodilatory.

The new findings of this study are the demonstration of inhibition of PGD₂ in the airways of atopic asthmatics after endobronchial allergen challenge. In conclusion, it seems likely that the inhibition of the early asthmatic response by PGE₂ is via its action on downregulation of prostaglandins and other mediator release from the mast cell. Given that PGE₂ has been shown to result in attenuation of the physiologic early and late asthmatic response, and our data supporting the anti-inflammatory mechanism of PGE₂, its exogenous administration or increase in endogenous production may well have roles in the protective anti-inflammatory management of allergic asthma.