INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: nitric oxide; parameter estimation; airways; inflammation; healthy children

Nitric oxide (NO) is an important endogenous mediator that arises from the airways and alveoli and can be detected in the exhaled breath (1, 2). Several inflammatory disorders, such as asthma, are associated with elevated exhaled NO concentrations (3). Although plateau NO concentration (C_NO,plat) in patients with cystic fibrosis (CF), a congenital lung disease marked by inflammation, has been reported to be decreased (7), it has also been reported to be unchanged (5, 6, 10) when compared with healthy control subjects. These paradoxical results for CF can be explained potentially by both real alterations in the underlying physiology due to differences in disease penetration, duration of disease, types of colonizing organisms, and the coexistence of atopy or asthma, as well as differences in the experimental protocol (e.g., variation in exhalation flow rate) (14, 15).

C_NO,plat may not be a specific or sensitive indicator of NO exchange dynamics in CF, because in and of itself, C_NO,plat provides little or no anatomic information. We have recently described a new, simple, and robust technique that utilizes a preexpiratory 20-s breathhold followed by decreasing flow rate (~ 6-1% vital capacity/s) maneuver to simultaneously determine several key flow-independent parameters that could reveal underlying physiological mechanisms (16): maximum flux of NO from the airways (J_NO,max, pl s¹), diffusing capacity of NO in the airways (D_NO,air, pl s¹ ppb¹), steady-state alveolar concentration (C_alv,ss, ppb), and mean (over radial position) concentration of NO within the tissue phase (C_tiss,air, ppb, equal to the ratio J_NO,max/D_NO,air). These provide potentially useful and reproducible information regarding net production and diffusion of NO in the airways and alveolar regions of the lungs. Thus, by providing a more detailed description of NO exchange dynamics throughout the lung, we hypothesize that this new technique will distinguish NO exchange dynamics in patients with CF from healthy children.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Nine healthy children and nine children with previously diagnosed CF (ages 10-14) participated in this study. Subjects were categorized as healthy on the basis of standard spirometry (FEV₁ > 75% of predicted on the basis of their race, age, and height) and no clinical history of chronic lung disease. Subject characteristics are presented in Table 1 including details of their clinical history. The Institutional Review Board at the University of California, Irvine approved the protocol, and informed consent was obtained from subjects and parents.

Experimental Protocol

Forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV₁) were measured in all subjects (Vmax229; Sensormedics, Yorba Linda, CA) by using the best performance (see Table 1) from three consecutive maneuvers before measuring the indices of NO exchange dynamics.

Each subject performed two types of exhalation maneuvers: (1) vital capacity maneuvers in triplicate at a constant exhalation flow of ~ 50 ml/s and ~ 250 ml/s according to the ATS (American Thoracic Society) and the ERS (European Respiratory Society) (17, 18) to determine C_NO,plat without breathhold; (2) five repetitions of a 20-s preexpiratory breathhold followed by a decreasing flow rate maneuver (16). During the breathhold, a positive pressure of > 5 cm H₂O was maintained to prevent nasal contamination (17). A schematic of the experimental apparatus has been previously presented (16). After the breathhold, the exhalation valve was opened allowing the patient to expire. The expiratory flow rate progressively decreased during the exhalation from 6% to 1% of vital capacity per second using a Starling resistor (Hans Rudolph Inc., Kansas City, MO) with a variable resistance.

Airstream Analysis

A chemiluminesence NO analyzer (NOA280; Sievers, Inc., Boulder, CO) was used to measure the exhaled NO concentration. The instrument was calibrated on a daily basis using a certified NO gas (45 ppm in N₂; Sievers, Inc.). The zero point calibration was performed with an NO filter (Sievers, Inc.) and done immediately before the collection of a profile. The flow rate and pressure signals were measured using a pneumotachometer (RSS100; Hans Rudolph Inc.). The pneumotachometer was calibrated daily and was set to provide the flow in units of STPD and pressure in units of mm Hg.

Parameter Estimation and Data Analysis

A previously described two-compartment model was used to estimate the four key flow-independent parameters in healthy children and patients with CF (16, 19). Only three of the four parameters are independent, as C_tiss,air is the ratio of two of the other parameters, maximum total volumetric flux per unit airway volume (pl/s) of NO from the airway wall (J_NO,max)/diffusing capacity (pl s¹ ppb¹) of NO in the airways (D_NO,air). Figure 1 is a simple schematic of the two-compartment model and flow-independent parameters. Mathematical estimation of the parameters (C_alv,ss, D_NO,air, J_NO,max) has been previously described in detail (16), and is accomplished by nonlinear least-square minimization utilizing a conjugated direction minimization algorithm.

View larger version (18K): [in this window] [in a new window]   Figure 1.   Schematic of two-compartment model used to describe NO exchange dynamics. Exhaled NO concentration, Cexh, is the sum of two contributionsthe alveolar region and the airway regionwhich depends on three flow-independent parameters: maximum total volumetric flux of NO from the airway wall (JNO,max, pl s1), diffusing capacity of NO in the airways (DNO,air, pl s1 ppb1), and steady-state alveolar concentration (Calv,ss, ppb). JNO is the total flux (pl/s) of NO between the tissue and gas phase in the airway, and inversely depends on the residence time of a gas bolus and thus the exhalation flow rate, E. The expression presented for JNO is valid for exhalation flows larger than approximately 1% of the vital capacity per second.

The alveolar region is characterized by the steady-state alveolar concentration, C_alv,ss (19, 20). Upon exhalation, the air passes through the airway tree and additional NO is transferred from the airway tissue. This additional volume of NO is referred to as the flux of NO from the airways, volumetric flux of NO (J_NO) (pl/s), and is not a constant, but a linear function of the gas phase concentration (19, 21, 22):

(1)

Thus, airway compartment is not well mixed, but the concentration depends on the axial or longitudinal direction and is characterized by two parameters: J_NO,max and D_NO,air. J_NO,max is the maximum flux of NO from the airway tissue, which is equal to the airway compartment flux if the gas phase concentration, concentration (ppb) of NO in the airway compartment (C_air), were zero.

An alternative presentation of the three flow-independent parameters includes the use of the mean (over radial position) tissue concentration in the airways, C_tiss,air, instead of J_NO,max (22). C_tiss,air is simply the ratio J_NO,max/D_NO,air. This is more easily demonstrated by expressing J_NO in an alternate but equivalent form to Equation 1 (16, 21, 22):

(2)

Note that J_NO,max is simply the product D_NO,air · C_tiss,air. In Equation 2, J_NO is expressed as proportional to the appropriate concentration difference between the airway tissue (C_tiss,air) and the gas stream (C_air). The coefficient of proportionality is D_NO,air. Thus, D_NO,air can be interpreted as a conductance for mass transfer of NO between the airway tissue and the gas phase.

Figures 2A (healthy) and 2C (CF) are representative experimental exhalation profiles as well as the model simulation. Figures 2B (healthy) and 2D (CF) are the corresponding exhaled volume, flow, and pressure tracings. The model does not predict phase I and II of the exhalation profile, where the accumulated NO during breathholding in the conducting airways and transition region of the lungs exits the mouth. This discrepancy is attributed to axial diffusion that our model neglects. Although the precise shape of phase I cannot be accurately simulated with the model, the absolute amount of NO in phase I and II can be predicted. Thus, our technique utilizes the information from phase I and II by forcing the model to simulate the total amount of NO eliminated in phase I and II (area under the curve as depicted in Figures 2A and 2C) of the exhalation in addition to simulating the precise exhaled concentration (ppb) (C_exh) over phase III. To ensure complete emptying of the airway compartment after breathhold, we define the transition from phase II and III as the point in the exhalation for which the slope (dC_exh/dV) of the exhalation profile is zero.

View larger version (35K): [in this window] [in a new window]   Figure 2.   Representative exhaled NO concentration profiles (dashed lines) and model fit (dark line) utilizing the 20-s preexpiratory breathhold followed by a decreasing flow rate maneuver are shown as a function of exhaled volume for healthy child (A) and a patient with CF (C ). In addition, the corresponding exhalation flow rate, exhaled volume, and pressure tracings during the exhalation are also presented as a function of time for the healthy child (B) and the patient with CF (D). Flow rates are controlled at ~ 6-1% of the each subject's vital capacity per second, and > 5 cm H2O of pressure is maintained through the breathing maneuver. Detailed description of parameters and regression techniques are described in the text and Appendix .

We assessed the ability of the model to simulate the profile by comparing the area under the curve in phase I and II in the experimental and theoretical data. Figure 3 presents the percentage of the experimental area under the curve predicted by the model for all 45 breathing maneuvers (nine subjects times five repeated maneuvers). It is evident that there are a few outliers (solid circles, < 80% of the experimental area under the curve) in each population, which represent profiles that the two-compartment model cannot adequately simulate. These profiles were subsequently discarded. There were no more than two discarded profiles from any individual subject. Thus, the population mean values for each of the parameters for the healthy children and patients with CF are based on exhalation profiles in which the model-predicted area under the curve for phase I and II is > 80% of the experimental area under the curve.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Absolute amount of NO eliminated in phase I and II of the exhalation profile as predicted by the model is shown as a percentage of the NO eliminated experimentally. Each circle represents one of the 45 individual data points from the individual breathing maneuvers. Solid circles indicate outliers (model predicted is < 80% of experimentally measured), and means with 95% confidence interval are calculated after elimination of the outliers.

The intramaneuver (superscript m), intrasubject (superscript s), and intrapopulation (superscript p) variability have been described previously (16), and will be characterized by the 95% confidence interval expressed as a percentage of the estimated parameter value, intramaneuver 100(1)% confidence interval (CI^m), intrasubject 100(1)% confidence interval (CI^s), and intrapopulation 100(1)% confidence interval (CI^p), respectively. C_NO,plat and the flow-independent parameters are expressed as a mean and 95% confidence interval (CI). To determine the relation between pulmonary function (FVC and FEV₁/ FVC) and the estimated three flow-independent parameters, linear regression and correlation were performed. A p value of < 0.05 is considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

FVC, FEV₁, FEV₁/FVC, and the clinical history of the CF subjects are presented in Table 1. Seven of the nine subjects with CF had concurrent reactive airways diseases, and all of these subjects were currently taking inhaled corticosteroids. The two subjects who did not have reactive airways disease were taking only nasal corticosteroids. FVC and FEV₁ were significantly lower in CF than in healthy children (p < 0.05). Means for each of the nine subjects in each category and the population mean are presented in Figure 4. In the nine children with CF, only six were able to complete the 20-s breathhold. In the three children who were not able to hold their breath for 20 s, one held their breath for 15 s, and two for 10 s. Estimated J_NO,max is not significantly different between healthy control subjects and patients with CF (p = 0.52); however, D_NO,air is significantly elevated (p = 0.007) and C_alv,ss and C_tiss,air are significantly reduced (p = 0.05 and 0.002, respectively) in CF.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Population means for the four flow-independent parameters (A: JNO,max, B: DNO,air, C: Calv,ss, D: Ctiss,air) of healthy control subjects and patients with CF are presented as a column with individual mean values for each of the nine subjects (solid circle: healthy control subjects, open circle: patients with CF) superimposed. Error bar indicates the intrapopulation 95% confidence interval (CIp). *Statistically different from healthy control subjects (p < 0.05).

Mean values for CI^p for healthy children and CF are 46% and 82%, 49% and 53%, 60% and 46%, and 51% and 51% for J_NO,max, D_NO,air, C_alv,ss, and C_tiss,air, respectively. Thus, the variation across the population of healthy children and patients with CF is approximately the same for each of the parameters.

The population means of CI^m and CI^s for each parameter are presented in Figure 5 for the healthy control subjects and patients with CF. J_NO,max and C_alv,ss for healthy controls are estimated with high accuracy (population mean CI^m = 38% and 97%, respectively) relative to D_NO,air (population mean CI^m = 128%). For CF, the population mean for CI^m for the three flow-independent parameters are 118%, 167%, and 284% for J_NO,max, D_NO,air, and C_alv,ss, respectively, which are much larger than the healthy children. In contrast to CI^m, population mean values for CI^s for each of the parameters are not statistically different between healthy control subjects and patients with CF, and range from 26% to 65% (Figure 5).

View larger version (12K): [in this window] [in a new window]   Figure 5.   Mean values for each subject for intramaneuver (CIm, closed symbols) and intrasubject (CIs, open symbols) confidence intervals are shown for each parameter (JNO,max, DNO,air, and Calv,ss) for healthy control subjects and patients with CF. Solid bar indicates population mean. HC; CF.

None of the flow-independent parameters had any significant correlation with standard indices of lung volume (FVC) or airway obstruction (FEV₁/FVC) for both healthy children and patients with CF.

Mean C_NO,plat (± CI) from the experimental maneuvers were 20.2 ± 10.1 ppb and 10.2 ± 3.97 ppb for healthy children and 8.91 ± 5.38 ppb and 4.79 ± 2.48 ppb for patients with CF at the target flow rates of ~ 50 ml/s and ~ 250 ml/s, respectively. The experimental values for each of the individuals are presented in Table 2 with model-predicted C_NO,plat in parenthesis. C_NO,plat at the targeted flow rate of 50 ml/s and 250 ml/s are lower for patients with CF when compared with healthy control subjects; however, the mean exhalation flow rates for patients with CF were larger in both cases (Table 2). Because a larger flow rate will decrease C_NO,plat, one must attempt to control for this confounding variable. We used the estimated flow-independent parameters for each of the subjects to estimate C_NO,plat using the two-compartment model (see governing equation in the Appendix ) at precisely an exhalation flow rate of 50 and 250 ml/s (Table 2). Mean C_NO,plat (± CI) from the model were 17.5 ± 8.97 ppb and 7.11 ± 4.03 ppb for healthy children and 11.5 ± 6.89 ppb and 4.28 ± 2.64 ppb for patients with CF at 50 ml/s and 250 ml/s, respectively. The model-predicted C_NO,plat are still lower for children with CF when compared with healthy children, but the difference is no longer significant (p > 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study has estimated several flow-independent parameters characteristic of NO exchange dynamics in the lungs, which have not been measured in CF before, and compared these to the plateau concentration at a constant exhalation flow rate. When compared with healthy control subjects, we found differences in C_NO,plat that were not significant if one controlled for differences in the exhalation flow rate. In contrast, the flow-independent parameters differed substantially between patients with CF and healthy control subjects. Thus, our new method provides greater sensitivity in distinguishing NO exchange dynamics between healthy control subjects and patients with CF.

The difference between C_NO,plat and the flow-independent parameters is highlighted by examining the composite experimental and model-predicted exhalation profiles from the population of patients with CF and healthy control subjects (Figure 6). Figure 6A depicts the composite experimental profile for each population by taking the mean exhaled concentration at equivalent exhaled volumes. Figure 6B presents the composite model predicted profiles by using the average values for the flow-independent parameters in the model equation (see Appendix ). It is evident that there are small changes in the exhaled NO concentration during phase III due to small changes in C_alv,ss and J_NO. However, there are very large differences in the concentration of NO in phase I and II in which the differences in NO exchange dynamics between healthy control subjects and patients with CF are magnified during a breathhold. The magnification of the airway compartment allows one to estimate D_NO,air, J_NO,max, and C_tiss,air, achieve more detailed information regarding NO exchange dynamics, and distinguish healthy children from patients with CF.

View larger version (19K): [in this window] [in a new window]   Figure 6.   Composite experimental (A) and model-predicted (B) NO exhalation profiles are presented for the 20-s breathhold followed by a decreasing flow rate maneuver for healthy control subjects and patients with CF. Population mean values for the three parameters, JNO,max, DNO,air, and Calv,ss, for healthy control subjects and children with CF were utilized in the model prediction and are equal to 784 and 607 pl s1, 4.82 and 17.6 pl s1 ppb1, and 4.63 and 1.96 ppb, respectively. FVC was set equal to the mean value for each population, and the flow rate decreased from 6% to 1% of the vital capacity per second.

Interpreting the significant increase in D_NO,air observed in patients with CF requires understanding the factors that affect its value. Thicker and more viscous mucus present in patients with CF would tend to increase the diffusion distance for NO as well as the "ease" at which NO can diffuse. Both of these observations would decrease D_NO,air and contrast with our experimental observation of an elevated D_NO,air.

It has been previously hypothesized that inflammation may increase the surface area of the airways producing NO by stimulating the inducible form of nitric oxide synthase (iNOS) (34, 35). This would increase D_NO,air (proportional to surface area, see Appendix ), which is consistent with our observation; however, iNOS expression is reduced in epithelial cells from patients with CF even though levels of inflammatory mediators such as tumor necrosis factor (TNF)- and interleukin (IL)-1 are increased in bronchoalveolar lavage (36). Furthermore, neutrophils, which increase iNOS expression in normal cells, do not enhance the iNOS expression in CF epithelial cells (37).

The impact of chemical consumption of NO, as characterized by k (a first-order rate constant), is also a possibility. In vivo, NO reacts with several substrates including oxygen, protein thiols (e.g., glutathione), and superoxide. In CF, there are increased numbers of neutrophils, which can produce toxic radical species including superoxide. Normally, glutathione (GSH) provides an effective antioxidant role in the lung; however, in CF, the concentration of GSH has been reported to be decreased (23) as a result of abnormal GSH transport related to the defective CFTR (24). The presence of neutrophils and lower levels of GSH may provide an environment in which excess superoxide can react quickly with NO to form peroxynitrite and other more stable end products such as nitrite and nitrate (25, 26).

It has been postulated that the half life, t_1/2, in vivo in healthy control subjects is between the range of 0.1 to 10 s, which is equivalent to between 7 and 0.07 s¹ based on the simple relationship k = ln(2)/t_1/2 (the shaded region of Figure 7). Figure 7 plots a theoretical estimate of D_NO,air (see Appendix for details) as a function of k. D_NO,air increases as k increases (t_1/2 decreases). The relationship is highly nonlinear: D_NO,air is independent of k for k < 1, but becomes a strong positive function of k for k > 10. The positive relationship between D_NO,air and k is due to an increase in the radial gradient of NO concentration from chemical consumption and is a well-known phenomenon in chemical reaction engineering (33).

View larger version (17K): [in this window] [in a new window]   Figure 7.   Predicted DNO,air (by Equation A2) is plotted as a function of reaction rate constant k. Values for Aair, tiss:air, NO,tiss, and Ltiss are 9,100 cm2, 0.0412, 3.3 × 105 cm2/s, and 0.01 cm, respectively, and are obtained from a previous study (28). The shaded region depicts that predicted for a healthy subject. Patients with CF may have a larger value for k as indicated.

The increase in chemical consumption of NO acts to increase D_NO,air, or the conductance for mass transfer, but would decrease the mean tissue concentration, C_tiss,air. In addition, reduced iNOS expression would also tend to reduce C_tiss,air. Thus, the effect on D_NO,air and C_tiss,air would tend to negate each other in terms of the impact on the actual flux of NO (J_NO or J_NO,max, see Equations 1 and 2) from the airways. C_alv,ss is reduced in patients with CF, but this represents only 17-24% of C_NO,plat at an exhalation flow rate of 50 ml/s. This provides a potential explanation for why some studies on CF have reported no statistical change, or a small decrease in exhaled NO concentration, and why the flow-independent parameters are potentially more sensitive. A patient with CF could have highly elevated D_NO,air but greatly reduced C_tiss,air; thus, J_NO and J_NO,max are near normal, and a decrease in C_alv,ss results in a small decrease in C_NO,plat, which may not be detectable when compared with healthy control subjects (compare healthy subject #8 with CF subject #5). Figure 8 describes schematically the probable cascade of events in CF that leads to changes in the flow-independent parameters and little or no changes in the exhaled NO concentration as previously described.

View larger version (18K): [in this window] [in a new window]   Figure 8.   Schematic of the possible mechanisms underlying nitric oxide exchange dynamics in CF. A cascade of events may be initiated by abnormal GSH transport and influx of inflammatory cells, which leads to increased consumption of NO, decreased tissue phase concentration of NO, slightly decreased alveolar gas phase concentration, and increased airway diffusing capacity. This combination of changes leads to relatively normal airway flux of NO and normal to low exhaled NO concentration.

It is important to note that asthma, or reactive airways disease, can also substantially impact NO exchange dynamics. Only two of the CF subjects (subjects #3 and #7) did not have concurrent reactive airways disease. However, these two subjects did not display remarkably different NO exchange parameters. One subject with CF had a very high J_NO,max and D_NO,air (subject #2). There was nothing remarkable in this subject's clinical history; however, this subject had by far the best lung function of the subjects with CF (see Table 1). The clinical significance of this relationship is not known.

The larger intramaneuver confidence interval in patients with CF (Figure 5) is due to the smaller lung volume (Table 1 and Figure 6), relative inhomogeneity of gas distribution, and magnitude of the exhaled NO signal. Even though the range of exhalation flow rates is scaled to the lung volume, a smaller lung volume results in a smaller absolute range of sampled flow rates. Thus, the smaller exhaled volume results in a smaller change in the exhaled NO signal with changing flow rate thereby decreasing the accuracy of the parameter estimate (based on the nonlinear least-square regression technique). In addition, the absolute exhaled concentration of NO tends to be smaller in CF, which decreases the signal-to-noise ratio of the analytical instrument.

As previously described, only six of the nine children with CF were able to complete the 20-s breathhold. Although a 20-s breathhold is desirable, it is not required. We have previously reported (16) theoretically and experimentally that the uncertainty of D_NO,air decreases as the breathhold time increases up to 20 s. The estimation of D_NO,air depends almost exclusively on the magnitude of the peak in phase I (see Figure 6), which increases with increasing breathhold time (16).

In spite of the differences observed with CI^m, the intrasubject (CI^s) variability is not significantly different between healthy control subjects and patients with CF (see Figure 4). This finding suggests that the reproducibility of the breathing maneuver does not depend on the presence of CF despite the fact that three of the children with CF were not able to hold their breath for 20 s. Each of these three children held their breath for approximately the same length of time for each of the breathing maneuvers.

The intrapopulation (CI^p) variability did not differ between the healthy children and patients with CF, and is similar for all of the flow-independent parameters (range 46-82%). This variation within the population is much larger than other endogenous gases (e.g., CO₂). The mechanisms underlying the intrapopulation variability both in health and disease are not known, but might be related to the presence of subclinical inflammation in healthy control subjects or genetically determined differences in NO metabolism as characterized by the length of a repeated polymorphism in the neuronal nitric oxide synthase gene in patients with CF (39, 40).

There are permanent structural changes that occur in the CF lung, particularly those related to anatomic and physiological deadspace. The physiological deadspace in CF is normal to slightly increased relative to healthy control subjects (41). Thus, our estimate of the airway compartment volume (volume of airway compartment [V_air, ml] Table 1), which is based on the subjects ideal body weight and age, may underestimate V_air. We have previously shown that D_NO,air is the most sensitive of the three flow-independent parameters to the choice of V_air (16). The dependence is positive; that is, as V_air increases, the estimate of D_NO,air also increases to compensate for the dilutional effect of the larger volume. Thus, a slightly larger V_air for children with CF would actually serve to increase the D_NO,air further from the healthy control population. Nitrogen washout is a common method to estimate anatomical deadspace; however, the accuracy of this method is compromised by the presence of diseases that impact emptying patterns. The dependence of D_NO,air on V_air may explain some of the intersubject variability, and suggests that intrasubject longitudinal changes in the flow-independent parameters may have the greatest clinical utility.

Another possible source of error is performing the spirometric breathing maneuvers before the NO breathing maneuver. According to Silkoff and coworkers (42) and Deykin and coworkers (43, 44), spirometry can depress exhaled NO levels by 10-36% from the baseline in healthy subjects and subjects with asthma by an unknown mechanism. This may perturb one or more of the flow-independent parameters, and should be considered in any future studies.

In summary, we have quantified flow-independent parameters characteristic of NO exchange dynamics in the lungs of healthy children and age-matched patients with CF. The airway diffusing capacity and tissue concentration are significantly different between the two populations and may be a more effective technique than the plateau concentration in distinguishing NO exchange in patients with CF from healthy control subjects. In addition, the flow-independent parameters are not correlated with standard spirometry, which suggests that they may provide additional information to the clinician regarding the inflammatory status of the airways and alveolar region in patients with CF. Their precise interpretation will not be known until future studies establish the correlation of these parameters with more established markers of inflammation obtained by more invasive means.