INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Expired nitric oxide (eNO) is the first noninvasive marker for airway inflammation, and previous studies have demonstrated increased eNO levels in patients with asthma (1). Increased eNO levels appear to result from increased expression of an inducible nitric oxide synthase (iNOS) in the presence of inflammatory cytokines (4). Administration of a nitric oxide synthase inhibitor or corticosteroids has been shown to diminish concentrations of eNO (5). More recent work has also shown strong correlations between eNO and airway hyperresponsiveness (8) and between eNO and an allergen-induced late decrease in FEV₁ (9). However, eNO is also known to be affected by other pulmonary conditions, such as upper respiratory tract infections, smoking, and cystic fibrosis (10).

To date, no studies have been reported of eNO in adults presenting to emergency departments (ED) with acute asthma. The present study was undertaken to validate the utility of mixed VC expirate measurements of NO in assessing the severity of an asthma exacerbation in an acute clinical setting. For the purposes of this study, asthma was defined solely on the clinical grounds of breathlessness, wheezing on examination, and a prior history of episodic wheezing. In contrast, most of the existing data on eNO have been gathered under controlled laboratory conditions, and asthmatic subjects from whom such data were collected were often characterized in advance on the basis of extensive laboratory tests.

In addition to examining the relationship between pulmonary function and eNO, we also wished to assess the relationship between eNO and other laboratory values commonly associated with asthma, such as elevated levels of serum IgE and peripheral blood eosinophilia.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Sixty-six randomly selected patients with acute asthma and 69 age- and sex-matched controls between the ages of 18 and 50 yr were recruited in the emergency department (ED) of the University of Virginia Health Sciences Center in Charlottesville between August 1997 and March 1998. Subjects were enrolled under a protocol approved by the University of Virginia Human Investigations Committee. All subjects provided informed consent before they were enrolled.

The exclusion criterion for the controls was acute breathlessness at the time of the study. Eight of the 69 control subjects identified themselves as asthmatic. These eight subjects with stable, asymptomatic asthma who presented to the ED with nonrespiratory complaints, were analyzed as a subset of the entire group.

Fifty-two patients with acute asthma, eight patients with stable asthma, and 53 controls were included in the final analysis. Ten asthma patients and six controls were excluded because ambient levels of NO in the emergency department were elevated (> 30 ppb) at the time these individuals were seen. Previously published guidelines for measuring eNO recommend that when ambient NO levels are elevated, patients breathe NO-free air from a face mask before providing a breath sample (15). However, the purpose of the present study was to validate mixed expired NO as a clinical tool for assessing acute asthma in the ED, and requiring patients to breathe NO-free air in the midst of an acute exacerbation of asthma is impractical. In addition, three asthmatic subjects were excluded because they were subsequently diagnosed as having other illnesses. One of these three subjects was diagnosed as having vocal cord dysfunction, another as having pneumonia, and a third with congestive heart failure. One asthmatic subject and three control subjects were excluded because NO samples were not collected.

All 52 patients with asthma complained of breathlessness, had wheezing on examination, and reported a history of asthma. All were treated by an ED physician with bronchodilator therapy. Sixteen (31%) of the acutely asthmatic patients reported using inhaled but not oral corticosteroids, seven (13%) reported using both inhaled and oral corticosteroids, and one (2%) reported using only oral corticosteroids. Forty-two of the acutely asthmatic patients (81%) reported using inhaled -agonists at the onset of symptoms (Table 1).

The group with stable asthma consisted of eight subjects who presented to the ED with complaints other than breathlessness. These subjects reported past episodes of wheezing, but did not have asthma symptoms at the time of the study. One of the eight patients with stable asthma reported using inhaled steroids and -agonists regularly. Four others reported using -agonists as needed, but none had used their inhalers in the preceding week. A smoking history was obtained from all participants (Table 1).

Spirometry

Subjects performed three FVC maneuvers in the sitting position while connected to a Renaissance Spirometer (Puritan-Bennett, Wilmington, MA). Spirometry was performed and evaluated with American Thoracic Society (ATS) criteria for acceptability and reproducibility. Subjects were asked to exhale for at least 6 s or until zero flow was achieved. Subjects were asked to perform a maximum of three efforts even if ATS criteria were not met (16). ATS recommendations are that subjects perform up to eight efforts in order to obtain reproducible results. Multiple efforts for spirometry in the midst of an acute asthma attack would be time consuming and might contribute to patient discomfort and fatigue.

NO Collection

Subjects provided a mixed VC expirate for NO sampling, by performing an FVC maneuver while breathing into an NO-impermeable Mylar balloon (Amscan, Harrison, NY) through a 5-mm-diameter mouthpiece. The subjects were seated and asked to inspire fully. At end inspiration they immediately exhaled forcefully into the balloon, if possible for a period of 6 s. An ambient air sample from the ED was collected by allowing air to flow into the open end of an evacuated 8.1-L Tedlar gas sampling reservoir (Cole Parmer, Niles, IL).

Mixed samples of VC expirate were analyzed within 1 h after collection with a Sievers NOA 280 chemiluminescence device (Sievers Instruments, Boulder, CO) with a lower limit of detection of 2 ppb. NO gas in a Tedlar reservoir bag appears to remain stable for at least 2 h after collection (17). The device was calibrated with 0 ppm (ambient air passed through a permanganate filter) and 9.96 ppm NO (BOC Gases, Murray Hill, NJ). At least 2 ppb NO was detected in all samples.

Determination of Flow Rate and Mouth Pressure

In order to determine that expiratory flow rates were sufficient to remain above a level that affects NO, we performed a second study. A group of 21 volunteers were recruited from the pulmonary function laboratory of our institution. Flow-volume curves were recorded for each subject with the SensorMedics 6200 Spirometer (Yorba Linda, CA) while the subject exhaled through the same resistive mouthpiece used in the ED during eNO sample collection. An in-line pressure manometer was placed proximal to the Mylar balloon to measure mouth pressures (CP100; Bicore, Irvine, CA). Expiratory flow rates were recorded at three time points in each FVC maneuver. Mouth pressure tracings were analyzed for a plateau pressure. Exhaled NO levels were not measured during this study. Simultaneous measurement of expiratory flow rates and mouth pressures required the addition of a significant dead space volume in the collection apparatus, which could have altered eNO values.

Blood Sampling

Subjects had blood drawn for a complete blood count, differential count, and total serum IgE assay. Serum was aliquoted and stored at -30° C. IgE was measured in duplicate, using the CAP system (Pharmacia Diagnostics, Uppsala, Sweden).

Statistics

The True Epistat (Epistat Services, Richardson, TX) statistical program was used for data analysis. Values are presented as mean ± SEM. Differences between mean control NO levels and those in mixed VC expirate of asthma patients were analyzed with the Mann- Whitney U test for nonparametric data, since values for NO, IgE, and eosinophils were not normally distributed. All correlations were generated with Spearman's rank sum test. A value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The sex, race, smoking history, and medication use of the study subjects are shown in Table 1. Forty-three of 52 (83%) patients with acute asthma, five of eight (63%) with stable asthma, and 39 of 52 (75%) controls had reproducible ( 5% variation between the two best efforts) FEV₁ measurements. Mean FVC, FEV₁, FEV₁%, FEF_25-75, and FEV₁/FVC ratios are presented in Table 2. The mean FEV₁ and FEF_25-75 differed significantly among the three groups. Of note was that the eight patients with stable asthma were initially enrolled as part of the control group; however, their pulmonary function was significantly different from that of the controls (p < 0.01), and they were therefore treated as a separate group in the present analysis.

Mean eNO was 15.0 ± 1.0 ppb (range: 4.6 to 35 ppb) in the patients with acute asthma (n = 52) versus 8.2 ± 0.5 ppb (range: 2.5 to 22 ppb) in the controls (n = 53) (Table 2, Figure 1). The eNO differed significantly in the acute asthma group and controls (p < 0.001). However, the eNO was 8.8 ± 1.5 ppb (range: 5.7 to 19 ppb) in the group with stable asthma (n = 8), which was not different from that of the controls (p = 0.76). Twenty-three of the 52 patients with acute asthma had mixed eNO levels  15 ppb, as compared with only two of 53 controls (p < 0.001, odds ratio [OR] = 20.6 [95% confidence interval (CI): 4.2 to 69.4]). A subset of 11 subjects in the acute asthma group who had low eNO values ( 10 ppb) and peripheral blood eosinophil counts in the normal range (< 200 cells/µl) was identified. Four of the subjects in this group had an FEV₁/FVC ratio indicating airway obstruction (FEV₁/FVC < 0.8). In comparison, 37 of 41 patients with acute asthma who had either an eNO > 10 ppb or eosinophilia had an FEV₁/FVC ratio indicating airway obstruction (FEV₁/FVC < 0.8) (p < 0.001, OR = 16.2 [95% CI: 2.6 to 117]).

View larger version (14K): [in this window] [in a new window]   Figure 1.   NO measurements of eNO in mixed VC expirate in Mylar balloons for patients with acute asthma (closed circles) and controls (open circles) in the ED. Bars represent the mean ± SEM. There was a significant difference in eNO in the patients with acute asthma as compared with the controls (p < 0.001).

In order to validate the method of eNO collection used in the ED, we studied 21 volunteers in the pulmonary function laboratory. The mean expiratory flow rate (average of three time points) as measured with the Mylar bags was 1.14 ± 0.08 L/s. The mean mouth pressure recorded during the FVC maneuver was 42.5 ± 5.5 cm H₂O. The mean FVC% in this group of 21 subjects was 80.8 ± 5.5%, the mean FEV₁% was 76.5 ± 7.9%, and the FEF_25-75% was 64.6 ± 9.7%. Eleven of the 21 subjects had an obstructive pattern on spirometry, with an FEV₁%  76% (range: 20 to 76%). In this subgroup, the mean flow was 902 ± 163 ml/s and the mean mouth pressure was 31.2 ± 8.3 cm H₂O as measured with the Mylar bag apparatus. The subject from this group of volunteers with the lowest FEV₁% (20%) exhaled at a recorded flow rate 250 ml/s or more. This flow rate is well above levels believed to influence eNO measurement (18).

The peripheral blood eosinophil count was 372 ± 64 cells/ µl in the patients with acute asthma, 151 ± 39 cells/µl in those with stable asthma, and 104 ± 8 cells/µl in the controls. The level of eNO correlated with peripheral blood eosinophilia in the study group as a whole (n = 111, r = 0.33, p < 0.001). The level of eNO also correlated with peripheral blood eosinophilia in the subgroup consisting of patients with acute and stable asthma (n = 60, r = 0.34, p < 0.001) and in the subgroup with acute asthma (n = 51, r = 0.31, p = 0.03). Total serum IgE was measured in all subjects. Mean total IgE was 409 ± 77 IU in patients with acute asthma, 197 ± 62 IU in those with stable asthma, and 137 ± 39 IU in the controls. No significant correlation was observed between total IgE and eNO in any group.

The level of eNO also showed a significant inverse correlation with FEV₁% in the study group as a whole (n = 113, r = 0.46, p < 0.001), in the group consisting both of patients with stable and acute asthma (n = 60, r = 0.42, p < 0.001), and in the group consisting only of patients with acute asthma (n = 52, r = 0.29, p = 0.04) (Figure 2). Expired NO also correlated significantly with FEF_25-75% in the entire study group (n = 113, r = 0.47, p < 0.001), in the group with acute and stable asthma (n = 60, r = 0.40, p < 0.01), and in the group with acute asthma (n = 52, r = 0.31, p < 0.03).

View larger version (13K): [in this window] [in a new window]   Figure 2.   Correlation between eNO in mixed expired air sampled from Mylar balloons and FEV1% predicted in a group of 52 asthma patients presenting to the ED with an acute exacerbation.

The reported use of glucocorticoids did not influence eNO levels. The eNO concentration in mixed VC expirate collected in the Mylar balloon was 14.6 ± 1.3 ppb in the group using inhaled steroids (n = 23) and 14.9 ± 3.1 ppb in the group using oral glucocorticoids (n = 9), versus 14.7 ± 1.3 ppb in the group using no glucocorticoids (n = 29). Self-reported smoking habits also showed no relationship to eNO, since eNO was 8.3 ± 0.8 ppb in controls who reported smoking as compared with 8.2 ± 0.7 ppb in nonsmoking controls.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study was designed to evaluate the measurement of eNO as a noninvasive test for acute asthma in the ED. Previous studies have shown levels of mixed eNO to be significantly increased in asthmatic individuals. However, most other studies of this examined highly selected populations whose asthma was well characterized on the basis of extensive laboratory tests in advance of the study. The present study examined a random population of acutely ill asthma patients presenting to the ED. There was no screening done in the selection of subjects, which reduced the chance of introducing a selection bias into the study. Consequently, the significant finding of increased eNO levels in the acute asthmatic group was increased because they were enrolled randomly.

Two principal methods of eNO sampling have been reported in the literature: on-line and mixed VC sampling. In on-line sampling, patients expire into the analyzer, which rapidly analyzes eNO; in mixed VC testing, patients inspire fully and then exhale a full VC volume into a collection bag, with this volume being analyzed later. On-line measurement of eNO has some significant limitations as compared with the mixed VC technique. Patients must breathe directly into the collection device with sufficient flow to produce enough force to reach a plateau pressure of 20 cm H₂O. Sustained flows of this magnitude may be difficult to achieve for asthma patients with severely obstruction and acute dyspnea. With the mixed VC method, there is no requirement that subjects exhale at a target flow rate. Moreover, the on-line technique requires using bulky equipment, including cylinders of oxygen and calibration gas, which is not possible in the ED; mixed expired air samples can be collected in the ED and transported to a central site for subsequent analysis.

Mixed eNO sampling also has some potential drawbacks. The expired gas may be contaminated with eNO from the nasopharynx, which is known to be present at a higher concentration than oral eNO (17). However, the resistive mouthpiece used in the present study, when tested in a group of similarly obstructed subjects, resulted in substantial increases in mouth pressure (mean = 31.5 cm H₂O). These levels of back pressure are adequate to close off the velum and isolate the nasopharynx from the rest of the upper respiratory tract (19). Furthermore, other investigators have compared on-line eNO collection with balloon collection and found no significant differences between the two methods. One group, in a study of 72 children, found that mean eNO was 5.1 ± 0.3 ppb with mixed expired balloon collection and 5.3 ± 0.2 ppb with on-line sampling (20). Canady and colleagues have reported a strong, linear relationship (r² = 0.83) between eNO measured with a Tedlar bag and the on-line technique (21).

Another drawback to the use of the mixed expired air technique in the present study was that the flow rate was not monitored during sample collection, and flow rate has been shown to influence eNO measurements in at least two separate studies. Hogman and colleagues have shown that eNO rises when the flow rate falls below 150 ml/s, and Silkoff and coworkers showed a similar effect at flow rates < 100 ml/s (18, 22). There was no effect on eNO when flow rates were above these thresholds. This variation with flow introduces the possibility that the correlations observed between FEV₁ and eNO (Figure 1) in the present study may be an artifact related to lower flow rates in the asthmatic subjects. However, it is our conclusion that the asthmatic subjects in the present study performed forced exhalations into Mylar balloons at flow rates well above rates known to affect eNO.

Both the spirometric measurements and the eNO maneuvers in the present study involved forced exhalations. Table 2 shows the spirometric results obtained in the ED. During almost all of the FVC maneuver, flow rates were well above 150 ml/s, since the average FEF_25-75 was 1,550 ± 130 ml/s, the average PEF was 3,940 ± 280 ml/s, and the average FEV₁ was 1,790 ± 100 ml/s. Two subjects, however, did have measured FEF_25-75 values below 350 ml/s; such low flows may have affected their eNO readings (13.5 ppb in one subject and 24.0 ppb in the other). In addition, when further stable subjects with a variety of degrees of obstruction were tested in the laboratory, the measured rates of flow into the Mylar balloon apparatus remained above the critical value of 150 ml/s reported by Silkoff and coworkers (18).

Another problem in interpreting mixed eNO measurements in the acute setting is the influence of ambient NO levels on these measurements. Recently published work confirms our observations that a high ambient level of NO increases eNO (9, 23). On the basis of these investigations and published guidelines for measuring eNO, 10 asthmatic and six control subjects were excluded from the present study because of high ambient NO levels (> 30 ppb). Ambient NO levels in the ED ranged from 2 ppb to 171 ppb over the course of the study. If days with ambient NO > 30 ppb are included in the analysis, a highly significant correlation emerges between ambient NO and mixed eNO (r = 0.58, p < 0.0001). After discarding measurements made on those days with ambient NO > 30 ppb, a modest correlation remains between ambient NO and mixed eNO (r = 0.30, p < 0.04). These observations underscore the importance of measuring ambient NO whenever collecting breath samples for eNO.

eNO is thought to be a marker for inflammation within the human respiratory tract. In the present study, eNO correlated with peripheral blood eosinophilia, which itself is indicative of an inflammatory process in the lungs of asthmatic individuals. This observation corroborates earlier ones relating eNO to airway inflammation as measured by eosinophil counts in induced sputum (24). Previous studies have shown a reduction in eNO in individual asthma patients who received oral glucocorticoids (7, 25). Surprisingly, we found no difference in eNO between the asthmatic patients in our study who reported using inhaled or oral steroids and those who did not. Several possible explanations exist for this discrepancy. Use and compliance with antiinflammatory therapy was not verified beyond the questioning done as part of the initial enrollment in the ED. However, all of the subjects in the present study were in the midst of an acute attack, and it therefore appears that their antiinflammatory therapy was insufficient to control lower airway inflammation.

In contrast to other studies, the present study showed no difference in eNO between patients with stable asthma and controls, despite the lack of use of inhaled steroids by seven of eight patients with stable asthma. Byrnes and colleagues, Massaro and associates, and Kharitonov and coworkers all found elevated eNO levels in patients with stable asthma (6, 26, 27). Several methodologic differences between their studies and the present study may explain this discrepancy. All three of the other studies recruited subjects from selected referral populations in whom the diagnosis of asthma had been confirmed by an outside examiner or by laboratory tests such as methacholine challenge. The patients with stable asthma in the present study were characterized only on the basis of self-reported asthma, either in the recent or remote past. Little was known about the accuracy of that diagnosis or the activity of the disease. Also, two of the previous studies used peak eNO instead of mixed eNO measurements (6, 27). None of the three previous studies included asthmatic subjects who smoked, whereas five of eight patients with stable asthma in the present study smoked, which could have decreased eNO. Moreover, the patients with stable asthma in the present study constituted a relatively small group of eight subjects, whose mean FEV₁ was not as low as in other studies (26, 27).

In combination with other commonly used laboratory tests, eNO may have a role as a tool for assessing asthma. For example, a subset of 11 subjects of the 52 patients with acute asthma in the present study had both a "normal" eNO ( 10 ppb) and normal peripheral blood eosinophil counts (< 200 cells/µl). Only four of the 11 subjects in this subset had an FEV₁/FVC ratio suggestive of airway obstruction (FEV₁/FVC < 0.8). On the other hand, 37 of 41 patients with acute asthma and either an increased eNO or peripheral blood eosinophilia had an FEV₁/FVC ratio consistent with obstruction (FEV₁/FVC < 0.8). The sensitivity of the eNO/eosinophilia combination in predicting airway obstruction in this model is 90%, and its specificity is 64%. When eosinophilia alone is used to predict airway obstruction, sensitivity falls to 54%, whereas specificity is unchanged. These data suggest that patients who present with symptoms consistent with acute asthma and laboratory evidence of inflammation, such as an increased eNO concentration or eosinophilia, are likely to have the spirometric abnormalities consistent with acute obstruction. Although normal eNO levels and eosinophil counts do not rule out a diagnosis of asthma, it would appear that a low eNO concentration and normal eosinophil count is uncharacteristic of patients with acute asthma, and other diagnoses, such as vocal cord dysfunction or pulmonary embolism, should be considered in these patients (28).

This study demonstrated the potential clinical utility of mixed VC eNO measurement in assessing pulmonary inflammation in subjects presenting to the ED with symptoms suggestive of asthma. Asthma is a heterogeneous disorder with numerous subclassifications, such as allergic, nonallergic, or exercise-induced asthma. Laboratory parameters such as serum IgE, peripheral blood eosinophil counts, and pulmonary function tests are commonly used in clinical practice to further define an individual patient's phenotype. However, patients are not always able to provide reproducible spirometric measurements in an acute setting, as we have found. Nor are they always willing to have blood drawn. As a simple, noninvasive test for airway inflammation, measurements of eNO may be of great value in diagnosing or classifying asthma. Further investigations, relating eNO to other putative markers of airway inflammation such as chemokines or cytokines in sera or secretions, may be warranted.