INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: carbon monoxide; nitric oxide; carbon dioxide; oxygen; asthma

High concentrations of nitric oxide (NO) in exhaled gases of asthmatic individuals have been proposed as a noninvasive surrogate marker of airway inflammation and predictor of response to therapy (1). Recent work proposed that CO in exhalate of asthmatic individuals may also be increased owing to airway inflammation and subsequent induction of heme oxygenase in the airway (2). Evaluation of exhaled CO is of increasing interest because of the potential protective role of CO against oxidative stress (5). CO is one of the byproducts of heme catalysis by heme oxygenase (HO), which catalyzes the first step in oxidative degradation of heme to bilirubin. Whereas HO-2 and HO-3 are constitutively expressed, HO-1 is an enzyme induced by a variety of agents causing oxidative stress, including hyperoxia and proinflammatory cytokines (5). In vitro models of hyperoxic exposure of cells and an in vivo rat model of oxygen toxicity demonstrate increased HO-1 expression in various cell types, including bronchoalveolar epithelium, interstitial cells, fibroblasts, smooth muscle cells, and inflammatory cells such as macrophages (3, 10). Exposure to low concentrations of CO imparts tolerance to otherwise lethal hyperoxia in vivo. This may be due to reduction of neutrophil infiltration and apoptosis conferred by the presence of CO (7, 11). However, the relationship of CO to airway inflammation and airflow parameters is not well-established. Spontaneous asthma attacks and experimental allergen-induced asthmatic responses lead to oxidative stress by eliciting an immediate as well as delayed increase in reactive oxygen and reactive nitrogen species, concomitant with a loss of reducing potential (12, 13). If CO is produced in the inflamed airway by HO-1, we hypothesized that inflammation resulting from an experimental allergen challenge would result in an increase of CO in exhaled breath of atopic asthmatic individuals. In comparison, O₂, CO₂, and NO were also measured simultaneously with CO, to determine the relationship of CO to these three well-defined gases.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population and Design

To determine eligibility for the study, individuals underwent history, physical examination, allergy skin prick testing to a standard panel of aeroallergens, spirometry before and after two puffs of inhaled albuterol, or methacholine provocation to produce a 20% decline in FEV₁. Volunteers were classified as having asthma based on the National Asthma Education and Prevention Program Guidelines (14), which include: episodic respiratory symptoms; reversible airway obstruction by documentation of variability of FEV₁ or FVC by 12% and 200 ml either spontaneously or after two puffs inhaled albuterol; or a positive methacholine challenge test. Allergic asthmatics had two or more positive skin tests and met the criteria for asthma. All individuals were studied outside of the specific allergen season; they had not taken oral antihistamines and inhaled anti-inflammatory agents for more than 4 wk and had not taken oral corticosteroids for more than 8 wk. All were nonsmokers, defined as less than 5 pack-years and no smoking for the past 5 yr. Recent upper respiratory infections (within 8 wk), chronic pulmonary disorders other than asthma, significant medical history, including diabetes or coronary artery disease, recent surgery, and pregnancy were exclusion criteria. Spirometry and body plethysmography were performed in compliance with the American Thoracic Society Standards using a Jaeger Master Screen Body (Erich Jaeger, Inc., Wurzburg, Germany) (15). Methacholine provocation testing was performed according to published standards (16). The study was approved by the institutional review board, and written informed consent was obtained from all individuals enrolled in the study. Once the subjects met entry criteria, whole lung allergen challenge was performed between 1 and 4 wk after skin testing. After exhaled gases and spirometry at baseline were collected, whole lung allergen challenge was performed. Exhaled gases were collected before bronchodilator administration when FEV₁ decreased by 20% after allergen (Ag PC₂₀), designated as time zero. Exhaled gases and spirometry were performed subsequently at 1, 3, and 6 h after time zero. Subjects also returned at 24 and 48 h after time zero for measurement of exhaled gases and spirometry.

Collection of Exhaled Gases

Based on previous methods for exhaled CO measure (2, 17), exhaled gases were collected in Mylar bags using an inspiratory breath hold maneuver. Individuals inhaled to total lung capacity, and after a 15-s breath hold exhaled against 10 cm H₂O pressure into a Mylar collection bag. All individuals were seated at rest for at least 15 min before gases were collected. Exhaled CO and CO₂ were measured with a Siemens Ultramat 6 infrared analyzer (Karlsruhe, Germany) that was adapted for use in this study. The analyzer was calibrated daily using CO free gas and a gas with a known CO and CO₂ concentration. The analyzer was sensitive to a concentration of 100 ppb for CO and 0.1% for CO₂. The response time of the CO analyzer at a sample flow rate of 750 ml/ min was 7 s. Absorbed wavelengths for CO and CO₂ are characteristic and separable to the individual gases, so that CO₂ interference with CO does not occur until CO₂ levels reach > 8%. Carboxyhemoglobin (COHb) was measured using a combined arterial blood gas analyzer and hemoximeter (Model 1715; Instrumentation Laboratories, Inc., Lexington, MA), which measures COHb with multiple-wavelength spectrophotometry. Exhaled NO concentrations were determined using a chemiluminescence analyzer (Sievers Instruments, Boulder, CO) as previously described (18). A Teledyne UFO-130 micro-fuel oxygen sensor (City of Industry, CA) was used for exhaled O₂ sampling, with a response time of 0.150 ms. The oxygen analyzer was calibrated using zero air, followed by high gain calibration with 100% O₂ (Praxair, Cleveland, OH).

Zero air was prepared by passing ultrapure nitrogen (99.999% pure nitrogen; PraxAir, Cleveland, OH) through a NuPure II Eliminator room temperature purifier for inert gases (Manotick, Ontario, Canada). The gas purifier reduces gaseous impurities to concentrations of less than 1 ppb for O₂, CO₂, CO, hydrogen dioxide, hydrogen, and methane. Purified gas was then collected and used as a zero calibration gas for the analyzers.

Allergy Skin Testing

Allergy skin prick testing was performed with the following Ag: cat allergen; dog hair; Dermatophagoides farinae; cockroach; tree mix; grass mix; ragweed mix; molds including Alternaria, Aspergillus, and Cladosporium; normal saline as negative control; and histamine as positive control. Allergens were obtained from Hollstier Stier, Spokane, WA. Skin tests were read after 15 min. A positive reaction was a 3-mm-diameter wheal with a 10-mm erythema flare. Allergy or atopy was defined as two or more positive skin tests in the presence of positive histamine reaction.

Whole Lung Allergen Challenge

Subjects with more than two positive skin tests were exposed to aerosolized allergen (Ag) provocation (12). A stock solution of allergen (10,000 protein nitrogen units [PNU]/ml) was diluted with nonbuffered saline to produce eight concentrations (1, 3.16, 10, 31.6, 100, 316, 1,000, and 3,162 PNU/ml). The concentration of allergen in PNU/ml producing a 20% reduction in FEV₁ (Ag PC₂₀) was determined.

Statistical Analysis

Quantitative data are summarized as mean ± SE; categorical data are summarized by frequencies. Absolute and percent changes from baseline were computed for the quantitative variables. One-sample t tests comparing the mean changes to zero were used to identify trends. In the case of absolute changes, this is equivalent to using the paired Student's t test to compare baseline values with subsequent time points. Associations between pairs of variables or their changes from baseline are described by Pearson's correlation coefficient and a test for nonzero correlation. All tests were performed at individual significance levels of alpha 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

Eight atopic asthmatic individuals participated in the study. All had asthma and atopy as defined in the selection criteria. Three individuals underwent methacholine challenge to document airway reactivity. The group consisted of two men and six women between the ages of 25 and 53 yr. Clinical characteristics of study subjects are listed in Table 1. Six healthy nonatopic control individuals also participated in collection of exhaled gases at baseline (5 men, 36 ± 5 yr [ages 27-43 yr]).

Lung Function

FEV₁ decreased by 23 ± 3% with Ag (p = 0.002) (Figure 1). The average Ag concentration in PNU that decreased the FEV₁ by 20% (PC₂₀) was 1,337 ± 544, i.e., log dose 3.13 ± 1.12. Spirometric peak expiratory flow rates (PEFR), specific conductance of the airways (SGaw), and FVC also decreased with Ag (PEFR [L/s]: baseline 5.8 ± 0.6, Ag 3.8 ± 0.6, p < 0.001; SGaw [%]: baseline 55 ± 12, Ag 44 ± 14, p = 0.037; FVC [%]: baseline 85 ± 6, Ag 73 ± 4, p = 0.006) (Figure 1). After -agonist administration, lung functions returned to baseline. The presence of a late-phase response to allergen was confirmed by a decrease of PEFR 6 h after Ag challenge (L/s: 5.1 ± 0.7, versus baseline, p = 0.011). FEV₁ also decreased at 6 h after Ag consistent with a late-phase response in five of the eight individuals (FEV₁ [%]: baseline 67 ± 4, 6 h after Ag 62 ± 10, p = 0.026). PEFR, FEV₁, and SGaw were not significantly different from baseline at 24 h and 48 h after Ag (Figure 1).

View larger version (13K): [in this window] [in a new window]   Figure 1.   Airflow with whole lung Ag challenge. FEV1 %, SGaw %, and PEFR significantly decreased from baseline after atopic asthmatic individuals were exposed to Ag (arrow) (all p < 0.05). PEFR decreased at 6 h after Ag challenge, consistent with a late-phase response (p = 0.01).

Exhaled Gases

Ambient CO levels were measured daily and were never greater than 1.2 ppm. Serum COHb measured in all subjects on the day of gas collection was 0.4 ± 0.1% COHb. In determination of reproducibility of exhaled CO in healthy control subjects, 93% of replicate measures had an absolute difference within 10% of their mean value, and within individuals across two different days, 61% of measures had an absolute difference within 20% of their mean. Comparison of the infrared determination to an electrochemical analyzer (Bedfont Scientific, Rochester, UK) revealed good concordance for values over a range of CO concentrations. Measurements were within one unit for 70% of cases and within 50% for 90% of cases.

Baseline exhaled CO of asthmatics was similar to healthy control individuals (CO [ppm]: controls 1.8 ± 0.2 [n = 6], asthma 1.9 ± 0.4 [n = 8], p = 0.82). Exhaled CO decreased in all asthmatics immediately after Ag (CO [ppm]: baseline 1.9 ± 0.4, Ag 1.4 ± 0.4, p = 0.003). CO values at all time points after 1 h were similar to baseline (all p > 0.05) (Figure 2). Overall, CO was not related to lung function (all p > 0.1).

View larger version (13K): [in this window] [in a new window]   Figure 2.   Exhaled gases after whole lung allergen challenge. Exhaled NO increased after Ag (arrow), whereas exhaled CO significantly decreased during the immediate asthmatic response to Ag (p < 0.01) and subsequently returned to baseline values at later times. Exhaled CO2 during the immediate asthmatic response was lower than baseline (p  0.01), but at later times returned to baseline levels. O2 did not change significantly with Ag challenge (all p > 0.05).

Baseline exhaled NO was significantly increased in the asthmatics as compared with healthy control subjects (NO [ppb]: controls 7.3 ± 0.6, asthma 15 ± 2, p = 0.015). NO increased in all but one asthmatic individual at 3 h after Ag, and in six of eight individuals at 24 and 48 h (Figure 2). Baseline NO in asthmatics was inversely correlated with SGaw [r = 0.72; p = 0.044) (Figure 3) and tended to correlate with FEV₁ (r = 0.67; p = 0.07). On the other hand, baseline NO concentrations were positively correlated with SGaw 24 h after Ag (r = 0.89, p = 0.003), i.e., higher concentrations of NO at baseline predicted better SGaw 24 h after Ag. CO and NO concentrations correlated in individuals at baseline (r = 0.71, p < 0.01); however, this was lost after Ag challenge. There was no significant correlation of CO and NO with exhaled O₂ and CO₂ at any time, suggesting that the derivation of exhaled O₂ and CO₂ was different than CO or NO.

View larger version (9K): [in this window] [in a new window]   Figure 3.   Exhaled NO was correlated to SGaw at base- line (r = 0.72, p = 0.04).

Exhaled CO₂ decreased immediately after Ag in all but one individual, and was persistently lower than baseline at 1 h after Ag challenge (CO₂ [%]: baseline 5.0 ± 0.1, 1 h after Ag 4.8 ± 0.1, p = 0.045). At later times after Ag, exhaled CO₂ levels were similar to baseline (all p > 0.1) (Figure 2). Notably, exhaled CO₂ was related to FVC at baseline (r = 0.78, p = 0.024) and immediately after Ag (r = 0.74, p = 0.035) (Figure 4). Thus, decrease in vital capacity with Ag-induced asthmatic response was associated with decrease in CO₂ clearance from the lung.

View larger version (11K): [in this window] [in a new window]   Figure 4.   Exhaled CO2 was correlated to FVC at baseline (r = 0.78, p = 0.02).

In general, the exhaled gases of the late asthmatic responders were similar to the entire group. However, in the subset of patients with the late-phase response, a small but significant increase of CO₂ occurred at 3 h and 6 h after Ag challenge (CO₂ [%]: baseline 5.0 ± 0.1%, 3 h after Ag 5.3 ± 0.1%, 6 h after Ag 5.4 ± 0.4; all p < 0.02 in comparison to baseline values).

Exhaled O₂ did not change significantly with Ag (Figure 2). However, exhaled O₂ at baseline was the only exhaled gas that correlated with the Ag PC₂₀ of all individuals (r = 0.82; p = 0.013) (Figure 5). Exhaled CO₂ was inversely correlated to O₂ at all times. The linear fit of a composite of all the data from all times reveals the following relationship between CO₂ and O₂:

(1)

View larger version (8K): [in this window] [in a new window]   Figure 5.   Exhaled O2 was directly related to airway reactivity, i.e., the allergen PC20 (r = 0.82, p = 0.013).

where [O₂] is exhaled O₂%, [CO₂] is exhaled CO₂%, 1/0.8 is the slope, and the y-intercept 21 [r = 0.75, p < 0.001] (Figure 6). The linear fit reflects the alveolar gas equation, such that the slope predicts a normal resting respiratory quotient of 0.8 in these individuals. Derivation of the alveolar gas equation and a normal respiratory quotient in the asthmatic volunteers from measures of exhaled O₂ and CO₂ validated the accuracy of the method by which exhaled gases were collected.

View larger version (12K): [in this window] [in a new window]   Figure 6.   Inverse correlation of exhaled O2 and CO2 in asthmatics before and after Ag challenge. Exhaled CO2 from all individuals at each time of study was plotted against the corresponding O2. The linear fit of the data leads to derivation of the alveolar gas equation (Equation 1; r = 0.755, p < 0.001).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In contrast to higher than normal NO concentrations in the asthmatic individuals, CO concentrations were not different than that of control subjects in this study. Furthermore, Ag challenge did not increase exhaled CO, although airflow uniformly decreased immediately after Ag, i.e., FEV₁, FVC, PEFR, and SGaw (19). After administration of short-acting -agonist, airflow improved, but the majority of our volunteers had a late asthmatic response, manifest by significantly decreased FEV₁ or PEFR at 6 h after Ag.

Prior studies have shown higher than normal NO in exhaled breath of asthmatics at baseline, which increases after whole lung Ag challenge (19). Similarly, asthmatic individuals in this study had high concentrations of NO at baseline. The pattern of changes in NO was somewhat different than previously found, in that increases in exhaled NO in some patients were present by 6 h and in others persisted to 48 h. Baseline NO predicted the magnitude of decrease in FEV₁ during the late asthmatic response (4 to 8 h after Ag) in some studies (20). Here, baseline NO concentrations were inversely correlated with SGaw at baseline, i.e., the lower the SGaw the higher the NO levels. Increases in exhaled NO in asthmatic lungs are the result of multifactorial activation of NO synthetic machinery. This includes accelerated degradation of S-nitrosothiols in the airways with acidification of lung water, which increases NO production from NO₂ (22). In addition, NO synthesis is increased in airway epithelium of asthmatic lungs as compared with healthy control subjects, through transcriptional and posttranscriptional mechanisms (20, 25, 26). Inflammatory cytokines activate signal transduction proteins, which increase transcription of NO synthase (NOS) II messenger ribonucleic acid (mRNA). Furthermore, arginine, substrate for NOS II, is increased more than 3-fold in asthmatic epithelial cells (25).

In contrast to NO, CO concentrations after Ag did not increase over time. Rather, CO concentrations decreased immediately after Ag, with levels tending to be lower even 1 h after Ag. The cause of decreased CO is not known. We speculate that if CO is produced primarily within the lung, the decrease of CO immediately after Ag may be related to increased consumption by lung tissue or hemoglobin. Alternatively, edema during the immediate asthmatic response may decrease CO diffusion into gas space from the lung tissues. This mechanism has been implicated in decreased NO immediately after allergen challenge (27). Inflammation of distal lung parenchyma contributes to asthma pathogenesis (28, 29). Specifically, perturbations in diffusion capacity after asthma exacerbation may occur with decreased diffusion for up to 4 h after Ag challenge (27). In support of a diffusion defect leading to decreased CO, CO concentrations returned to baseline values by 3 to 6 h after Ag in our study. Perhaps most supportive of a diffusion mechanism for decreased CO is the similarity between the alterations in CO and CO₂ immediately after Ag challenge, and the lack of significant change in either gas during the late-phase response to Ag.

CO₂ concentrations in exhaled gases are derived from diffusion into the lung. In this study, exhaled CO₂ levels in asthmatics were dependent upon the vital capacity, a relationship likely detected because of the 15-s breath hold maneuver at total lung capacity. Thus, CO₂ exhaled is dependent upon the alveolar volume in the absence of tidal breathing. Close inverse correlation between O₂ and CO₂ provided a linear regression model in agreement with the known alveolar gas equation, confirmed a normal respiratory quotient in the asthmatic volunteers, and validated the accuracy of our methodology. Unexpectedly, baseline concentrations of exhaled O₂ correlated with airway hyperreactivity, as measured by Ag PC₂₀. This suggests that the more reactive asthmatic airways consume oxygen more avidly, perhaps because of increased oxidative and nitrosative chemical events in the lower airway of these individuals at baseline (25, 26). In conclusion, although this study corroborates prior findings of increased NO during the asthmatic response to an Ag challenge, no such dynamic increase in CO was noted after Ag challenge.