INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary involvement in cystic fibrosis (CF) is characterized by chronic inflammation punctuated by infective exacerbations. There is a massive neutrophil load in the lung with associated tissue destruction and progression to respiratory failure in the majority of patients. Studies examining bronchoalveolar lavage in these patients have shown an increase in inflammatory cytokines even during chronic stable states (1). Studies in infants without evidence of pulmonary bacterial colonization have demonstrated increased levels of inflammatory cytokines (2), suggesting that pulmonary inflammation is an early characteristic of the disease rather than simply a response to infection.

Nitric oxide (NO) is synthesized in a wide variety of mammalian tissues by nitric oxide synthase (NOS) using L-arginine as a substrate. There is evidence of involvement of this reactive gas in the control of vascular tone, neural signaling, and inflammatory processes (3). Three isoforms of NOS are recognized; endothelial (eNOS), neuronal (nNOS), and inducible (iNOS). eNOS and nNOS are constitutively expressed and their activity is tightly regulated through calcium-calmodulin-dependent pathways. In contrast, iNOS is not found in most normal tissues but is induced in the presence of inflammation by a variety of proinflammatory cytokines (4). This enzyme is calcium-independent and, once expressed, continues to produce large quantities of NO, which are rapidly metabolized to other reactive nitrogen intermediates, including NO₂, NO₂, N₂O₃, N₂O₄, S-nitrosothiols, and peryoxynitrite (5). Pharmacologic concentrations of glucocorticoids potently inhibit expression of iNOS but have no effect on the activity of this enzyme after induction.

NO is detectable in exhaled gas from humans (6), and inhalation of nebulized N^G-nitro-L-arginine methyl ester reduces exhaled NO concentrations, suggesting an endogenous pulmonary source (7). Further evidence for intrapulmonary production is provided by the demonstration that NO can be detected in the exhaled gas from isolated ventilated and perfused porcine lungs (8).

Abnormally elevated exhaled NO concentrations are seen in patients with untreated asthma (9), and administration of therapeutic doses of inhaled glucocorticoids to these patients causes exhaled NO concentrations to fall (10). Inhalation of nebulized aminoguanidine, a relatively selective inhibitor of iNOS, also reduced exhaled NO concentrations in asthmatic subjects while leaving exhaled NO unchanged in healthy control subjects (11). An increase in immunostaining for iNOS has been reported in lung biopsy samples from patients with asthma who have not received glucocorticoid therapy (12). Elevated levels of exhaled NO have been found in non-CF bronchiectasis (13) and normal subjects with upper respiratory tract infections (14). Taken together, these data suggest that iNOS is expressed in the presence of pulmonary inflammation causing NO to be released in increased concentrations. In marked contrast, a number of recent reports have shown that exhaled NO is not elevated in patients with stable CF (15).

The evidence indicating that lung inflammation causes activation of iNOS and the demonstration that acute upper respiratory tract infection causes an increase in exhaled NO in normal subjects led us to hypothesize that NOS activity is increased in the lungs of patients with CF with an acute infective pulmonary exacerbation causing elevated concentrations of exhaled NO. Such elevations could potentially provide a useful early clinical index of acute exacerbations. However, in our initial studies (see below) we found no such increase of exhaled NO in CF patients with acute pulmonary infections when compared with stable patients. This suggested either that iNOS activity was not increased in these circumstances or that NO production was increased but did not lead to increased concentrations in the exhaled gas because it was trapped and oxidized to its stable oxidation products nitrite and nitrate (NO₂/ NO₃) in the viscid sputum that lines the airways in CF. To distinguish between these possibilities, we examined sputum NO₂/NO₃ concentrations of patients with CF undergoing an acute infective pulmonary exacerbation, on admission to hospital and at the time of discharge, and compared these with the concentrations observed in patients with stable CF and in healthy control subjects.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Three patient categories were studied: (1) adult patients with CF recruited within 24 h of presentation to hospital with an acute pulmonary exacerbation (Acute Patients), (2) adult patients with CF who had not had an acute pulmonary exacerbation within 1 mo of recruitment and whose medications had not changed over the same period (Stable Patients), (3) healthy nonsmoking volunteers with no history of respiratory disease and no history of recent respiratory tract infection who were of similar age to the patient groups (Control Subjects).

The study protocols were approved by the ethics committee of Saint Vincent's Hospital, and all subjects and patients gave fully informed written consent. The diagnosis of CF was made on the basis of a compatible history and a positive sweat test.

Pulmonary Function Tests

FEV₁ and FVC were measured in all subjects at the time of recruitment (Vitalograph, Buckingham, UK), and values were expressed both in liters and as a percentage of the predicted value (18). In the acute patients, spirometry was repeated 14 d later after intensive inpatient therapy.

For the acute group baseline pulmonary function was determined as the values of FEV₁ and FVC recorded at routine outpatient clinic visits during the six mo prior to admission, at a time when there was no clinical evidence of an acute infective pulmonary exacerbation.

Sputum Collection and Analysis

Sputum was induced in control subjects by inhalation of nebulized hypertonic saline, as previously described (19), and collected into sterile plastic containers. For acute and chronic patients sputum was spontaneously expectorated or induced if necessary. Once collected, samples were processed within 1 h. Samples were diluted fivefold with 0.07 M sodium phosphate buffer (pH, 6.0) containing 0.5 M NaCl and 50 mg/L dithiothreitol, agitated gently to uniform consistency, and then centrifuged at 13,000 g for 30 min. The resulting supernatant was stored at 70° C for later analysis. NO₂/NO₃ was measured as previously described using the Griess reaction (20). Briefly, nitrate in the supernatant was converted to nitrite using nitrate reductase. Griess reagent was then added and total nitrite was calculated by measuring absorbance of the sample at 540 nm and expressed as µmol/L sputum.

Measurement of Exhaled NO

Exhaled NO was measured using a chemiluminescence analyzer (LR 2000; Logan Research Ltd, Rochester, UK). Subjects exhaled at a constant flow rate against a fixed resistance. Patients wore noseclips applied immediately before the maneuver and did not breathhold before exhalation. Readings were taken only if ambient NO was below 4 ppb. Gas was sampled from a sidearm at a constant rate of 250 ml/ min. Carbon dioxide was simultaneously measured using an infrared analyzer. Lower airway NO was measured by sampling gas at the end of exhalation, as indicated by the exhaled CO₂ concentration. The procedure was repeated three times and the mean value taken. The analyzer was calibrated regularly using standard calibration gases (BOC Gases Ireland Ltd, Dublin, Ireland).

Statistical Analysis

Data are expressed as means ± SD. To test for statistical differences between mean values paired or unpaired t tests were used when appropriate. For multiple comparisons across three groups, analysis of variance was performed and when a significant F-value was found, the Student-Newman-Keuls post hoc test for multiple comparisons was used to assess the significance of the differences between the means. A p value < 0.05 was accepted as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exhaled NO was measured in 13 acute patients, 25 stable patients, and nine control subjects. The characteristics of these patient are shown in Table 1. Mean FEV₁ and FVC were significantly higher in the stable group than in the acute group. Mean FEV₁ and FVC were significantly higher in the control group when compared with the acute and stable groups. There was no difference in age or height between the three groups. Stable patients were receiving routine medications, including inhaled bronchodilators, nebulized recombinant human dornase alpha, nebulized and oral prophylactic antibiotics, oral and inhaled glucocorticoid, vitamin E, and multivitamin and calorie supplements. After admission acute patients were treated with supplementary oxygen, intravenous antibiotics, glucocorticoids, and aminophylline, in addition to their routine medications. The results of measurement of exhaled NO in the three groups are shown in Figure 1. Mean exhaled NO was significantly higher in the control group (8.8 ± 4.9 ppb) than in both the acute (3.8 ± 3.9 ppb) and stable (5.0 ± 2.5 ppb) groups. There was no significant difference in exhaled NO between acute and stable groups. Four patients in the stable group were receiving neither inhaled nor oral steroids. Mean exhaled NO for these four patients (5.3 ± 2.6 ppb) did not differ significantly from the remainder of the stable group (4.9 ± 2.6 ppb) (p = 0.82, unpaired t test).

View larger version (13K): [in this window] [in a new window]   Figure 1.   Exhaled nitric oxide measured as parts per billion (ppb) for each of three subject categories: patients with acute exacerbation of pulmonary cystic fibrosis (CF), patients with stable CF, and healthy control subjects. Horizontal bars indicate mean values for each group. Asterisk indicates significant difference from acute and stable groups (p < 0.05, ANOVA).

Because mean exhaled NO was not elevated in patients with CF and was in fact lower than in the control subjects, we hypothesized that NO was trapped by excessive sputum production in the patient groups. In order to examine this possibility we measured sputum NO₂/NO₃ in 12 patients with acute exacerbation, 18 patients with stable disease, and the nine control subjects described above. Patient characteristics are shown in Table 2. Mean FEV₁ was significantly higher in the stable group than in the acute group. Mean FEV₁ and FVC were significantly higher in the control group when compared with the acute and stable groups. There was no difference in age or height between the three groups. The results of sputum NO₂/NO₃ measurement in the acute, stable, and control groups are shown in Figure 2. Mean sputum NO₂/NO₃ was significantly higher in the acute group (774 ± 307 µmol/L) when compared with both the stable group (387 ± 203 µmol/L) and the control group (421 ± 261 µmol/L).

View larger version (13K): [in this window] [in a new window]   Figure 2.   Sputum nitrate/nitrite (µM) for each of three subject categories: patients with acute exacerbation of pulmonary cystic fibrosis (CF), patients with stable CF, and healthy control subjects. Horizontal bars indicate mean values for each group. Single and double asterisks indicate significant difference from acute group (p < 0.05 and p < 0.01, respectively, ANOVA).

The differences in mean FEV₁ and FVC between the acute and stable groups suggest that the observed differences in sputum NO₂/NO₃ may have been caused, not by acute infection, but rather by impaired lung function. Therefore, we examined sputum NO₂/NO₃ in a subgroup of stable patients (matched stable) whose FEV₁ lay within the range of values of baseline FEV₁ observed in the acute patients. Mean FEV₁ in the matched stable group was 1.34 ± 0.44 L, 37 ± 12% of predicted, and mean FVC was 2.22 ± 0.50 L, 53 ± 12 % of predicted. Mean sputum NO₂/NO₃ was significantly (p < 0.05, unpaired t test) lower in the subgroup of matched stable patients (386 ± 205 µmol/L) when compared with the acute group (see above).

Paired admission and discharge sputum samples were available for seven of the acute patients. Spirometric data for these patients are shown in Table 3. There was a significant improvement in FEV₁ between admission and discharge. However, as shown in Figure 3, there was no significance difference between mean admission and mean discharge sputum NO₂/NO₃ for these patients: 707 ± 330 and 1,065 ± 526 µmol/L, respectively (p = 0.10, paired t test).

View larger version (15K): [in this window] [in a new window]   Figure 3.   Sputum nitrate/nitrite (µM) for patients with cystic fibrosis with acute infective exacerbation sampled on admission, and at time of discharge after 2 wk of intensive inpatient therapy.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have shown that exhaled NO is not elevated in patients with acute pulmonary exacerbation of CF but that the concentration of the stable NO oxidation products NO₂/NO₃ in sputum are elevated in this patient group when compared with stable patients and normal control subjects. We have also demonstrated that sputum NO₂/NO₃ remains elevated in patients with acute pulmonary exacerbation of CF after 2 wk of intensive therapy with intravenous antibiotics and high dose glucocorticoids. Our data confirm the observation that in patients with stable CF exhaled NO is not elevated when compared with that in normal control subjects.

Exhaled NO in patients with stable CF was not elevated and, in fact, was lower than in control subjects. Concentrations seen in this study are similar to those reported previously (15- 17). It has been suggested that low exhaled NO in stable CF may be due to trapping of NO in pulmonary secretions (16, 17). Our results do not support this hypothesis since we have shown normal sputum NO₂/NO₃ in association with low exhaled NO in stable patients. Four patients in the stable group were not receiving any form of glucocorticoid, yet these patients had concentrations of exhaled NO similar to those of the rest of the stable group, suggesting that glucocorticoid suppression of iNOS was not responsible for the low concentrations of NO seen. This lack of effect of glucocorticoids on exhaled NO in stable CF has been noted previously (17). Exhaled NO was also lower in acute patients when compared with control patients. We are unaware of any previous report of this observation. It is surprising that exhaled NO was not elevated in patients with CF given that iNOS is thought to be activated by pulmonary inflammation in other disease states and that even in asymptomatic infants with CF increased markers of pulmonary inflammation have been demonstrated (2).

In contrast, our results show that sputum NO₂/NO₃ was elevated in acute patients compared with that in stable patients and control subjects (Figure 2). Although we show higher absolute values than those reported previously for acute patients (21), trends from stable to acute groups are similar. This suggests that NOS is activated during an acute exacerbation, producing NO that does not diffuse into the airway lumen but rather is trapped by airway secretions, resulting in high levels of NO₂/NO₃ in sputum.

Baseline lung function as determined by routine outpatient spirometry, was poorer in the acute group than in the stable group. This is probably because those patients with more advanced lung disease had a greater likelihood of being in the hospital at any one time. There are little published data on the relationship between exhaled NO and the severity of pulmonary impairment in obstructive airway diseases. Kanazawa and colleagues have reported that, in asthma, the concentration of exhaled NO is correlated with the severity of disease as assessed by FEV₁/FVC (22). To avoid the possibility that differences in lung function between the groups influenced the comparisons of sputum NO₂/NO₃ in the acute and stable groups, we examined sputum NO₂/NO₃ in a subgroup of stable patients whose FEV₁ measurements fell within the range of baseline FEV₁ for the acute group. Our observation that NO₂/NO₃ concentrations in the acute group were significantly elevated above those found in stable patients with similarly impaired lung function indicates that the elevated concentrations were a consequence of the acute infection and not caused by more severely impaired baseline lung function.

Sputum NO₂/NO₃ was not elevated in the stable patients. This suggests that iNOS is not activated in stable CF despite the documented chronic pulmonary inflammation in these patients. Partial or complete inhibition of iNOS may explain this finding. Neutrophil-derived taurine chloramine has been reported to inhibit synthesis of iNOS in murine macrophages (23). Neutrophils are the predominant inflammatory cell in the lung in CF and may contribute to low iNOS activity in CF. Indeed, sputum levels of taurine and chloramines are elevated in patients with CF when compared with those in asthmatic patients (24). Thus, if iNOS production is attenuated in the stable CF lung then this may represent an immune defect in the host response contributing to the pathophysiology of CF.

Excess NO may be metabolized locally at an increased rate before it has an opportunity to diffuse into the airway lumen. It is known that pyocyanin found in Pseudomonas species, a common respiratory pathogen in adult CF, is nitrosylated in the presence of NO causing rapid inactivation of NO (25).

Although we have shown, in acute patients, an increase in sputum NO₂/NO₃ we have not determined its source or mode of synthesis. It seems clear that NO can be produced in the lung. Inducible NOS activity has been demonstrated in human neutrophils (26), macrophages (27), and epithelial cells (28). The contribution to exhaled NO of other NOS isoforms more remote from the airway lumen, such as nNOS in pulmonary neurons and eNOS in pulmonary endothelium are unknown and these isoforms may have contributed to the sputum NO₂/ NO₃ measured. It is possible, therefore, that the sputum NO₂/ NO₃ was not produced by the inducible isoform of NOS in the lung but rather by constitutive isoforms. CF lung tissue homogenates have been shown to have increased NOS activity that is calcium-dependent (29), a feature of the constitutive isoforms of NOS rather than of the inducible isoform (30).

It is also possible that extrapulmonary synthesis can contribute to exhaled NO. Nitrite in solution can form gaseous NO, and it has been shown that porcine lungs perfused with nitrite solution have an elevated exhaled NO when compared with those perfused with nitrite-free fluid (31). Therefore, extrapulmonary NO₂/NO₃ production, brought about by a systemic inflammatory response to acute pulmonary exacerbation of CF, could lead to the elevated levels of NO₂/NO₃ seen in sputum.

All of the acute group were treated during hospital admission with high dose glucocorticoid. Despite this there was no difference in sputum NO₂/NO₃ levels seen before and after 2 wk of intensive treatment. This suggests that NOS remained activated despite the high doses of systemic glucocorticoid used. It may be that the acute pulmonary inflammation had not sufficiently resolved after 14 d of intensive therapy. Our data should not be interpreted to mean that high dose glucocorticoid therapy has no effect on iNOS expression during acute pulmonary infections. The stimulus to express iNOS provided by the increased pulmonary inflammation may have been of sufficient magnitude to partially overcome the inhibitory effect of high dose glucocorticoid therapy. In the absence of this therapy NO₂/NO₃ may have continued to rise throughout the 2-wk observation period. Alternatively, it may be that the elevated NO₂/NO₃ seen during an acute exacerbation is not a product of iNOS but of a steroid-insensitive isoform analogous to the high output constitutive NOS reported in paranasal sinus mucosa (32).

There is considerable evidence suggesting that induction of iNOS and the resultant production of reactive nitrogen intermediates may have important pathophysiologic consequences in patients with CF. NO is a potent vasodilator and contributes to the formation of airway edema in inflamed lung by increasing blood flow to leaky post-capillary venules (33). Its vasodilator activity in the lung may also contribute to ventilation-perfusion mismatch. In addition to its proinflammatory functions NO also has lymphocytostatic effects (34). A further important activity of NO in the infected lung is its ability to kill pathogens (35). It has been shown that host defense response to several infectious agents is markedly impaired in iNOS-deficient mice. However, in other inflammatory conditions in mice, including influenza virus pneumonitis, LPS- induced hypotension, and ovalbumin-induced allergic airway disease, absence of iNOS worsens inflammatory tissue damage (5). Interestingly, it has recently been suggested that iNOS deficiency reduces lung damage in response to LPS (5). Thus, it is clear that induction of iNOS may lead to both beneficial and harmful effects depending on the exact circumstances. Whether or not the balance of the effects of increased NO production during acute infective pulmonary exacerbations of CF is beneficial or harmful remains to be determined.

In conclusion, exhaled NO is not elevated in patients undergoing an acute pulmonary exacerbation of CF and is not therefore a useful clinical indicator of such exacerbations. Sputum NO₂/NO₃, however, is elevated in patients with acute CF when compared with patients with stable CF and healthy control subjects. This supports the hypothesis that NOS is activated in patients with acute pulmonary exacerbation of CF and that the excess NO produced is trapped in viscid airway secretions. Concentrations of NO₂/NO₃ seen in acute patients did not change during intensive antibiotic and high-dose glucocorticoid therapy.

Further research is needed to determine the origin of elevated NO₂/NO₃ seen in acute pulmonary exacerbation of CF, to characterize the factors influencing its production, to investigate a possible role in the pathophysiology of CF, and to determine whether or not iNOS may be a useful therapeutic target.