Exhaled NO: first, hold your breath

the discovery of nitric oxide (NO) in expired air a decade ago (1) raised hope that measurements of exhaled concentrations of this important mediator would provide new insights into physiological and pathophysiological processes of the lung. Unfortunately, things are never as simple as they seem, and this hope is only now beginning to be realized. A major obstacle has been our inability to answer a classic question in pulmonary physiology; namely, what determines the concentration of an endogenous gas in exhaled air?

With respect to exhaled NO, the answer to this question must explain four observations in normal human subjects (4, 5). First, when NO produced in the sinuses was excluded from exhaled air by keeping mouth pressure at 20 mmHg, exhaled NO concentration achieved a stable level during constant expiratory flow. Second, this stable concentration decreased as expiratory flow was increased. Third, despite these flow-dependent decreases in exhaled NO concentration, the output of exhaled NO increased with increasing expiratory flow. Fourth, as air traveled toward the mouth during constant expiratory flow, its NO concentration progressively increased; for example, at an expiratory flow of 45 ml/s, one-half of the total increase occurred in airways larger than lobar bronchi.

Despite differences in approach and assumptions, several recently proposed models of NO exhalation provide fundamentally similar explanations for these observations (2, 3, 6, 7). NO in exhaled air derives from 1) convection of NO-containing alveolar gas and 2) diffusion of NO from airway walls. Because the reaction rate of NO with hemoglobin is extremely rapid and the NO-binding capacity of alveolar capillary blood is virtually unlimited, alveolar NO concentration is thought to rapidly achieve a low equilibrium value at the end of inspiration and remain constant at this value during the ensuing expiration. Consequently, the contribution of alveolar NO convection (the product of alveolar NO concentration and expiratory flow) to exhaled NO output varies linearly and directly with expiratory flow. The amount of NO diffusing into exhaled air from airway walls is determined by the NO concentration difference between airway walls and lumen and the airway NO diffusing capacity. The airway NO diffusing capacity, in turn, depends on the luminal surface area of NO-producing airways and the diffusion distances between sites of NO production and the airway luminal surface. Because airway NO diffusion increases with decreasing luminal NO concentration, its contribution to exhaled NO output also increases with increasing expiratory flow; however, because luminal NO concentration cannot fall below alveolar NO concentration, the contribution of airway NO diffusion to exhaled NO output achieves nearly maximal levels at relatively low flow rates. Thus exhaled NO output is dominated by airway diffusion at low expiratory flows and by alveolar convection at high expiratory flows.

The equations describing these processes make it possible to estimate parameters of exhaled NO output. For example, by measuring exhaled NO concentrations over a wide range of constant expiratory flows and using nonlinear regression to fit the data to model equations, it is possible to estimate NO concentrations in alveolar gas and airway wall, airway NO diffusing capacity, and maximum possible diffusional flux of NO from airway walls (the product of NO concentration in airway walls and airway NO diffusing capacity). Although this approach is just beginning to be applied, early results suggest that it may indeed lead to new insights about the lung. For example, it now appears that the increased exhaled NO concentrations characteristic of asthma are not due to increased airway wall NO concentration, as expected, but rather to increased airway NO diffusing capacity (6). Treatment with inhaled corticosteroids, which reduce activity of inducible NO synthase, decreased airway wall NO concentration but did not alter the increased airway NO diffusing capacity. Furthermore, the degree of airways obstruction and hyperreactivity to methacholine after steroids was greatest in patients with the smallest diffusional flux of NO from airways. On this basis, it was suggested that a major feature of asthma is decreased ability of endogenous NO to relax airway smooth muscle, leading to compensatory upregulation of constitutive NO synthase in nonadrenergic-noncholinergic airway nerves, reflected by increased airway NO diffusing capacity.

The multiple constant flow approach to exhaled NO analysis has some practical limitations, particularly in patients with lung disease. In a typical maneuver, a subject inhales to total lung capacity, holds his breath for a few seconds, and then exhales through a selected fixed resistance, maintaining mouth pressure (and therefore flow) constant by means of a visual-feedback device. Multiple exhalations over a wide range of flow rates are necessary. At very low flow rates, it may be difficult for dyspneic patients to sustain exhalation long enough to achieve stable exhaled NO concentrations, thereby compromising estimation of airway NO diffusing capacity and airway wall NO concentration, which depend critically on data obtained at lower flows. At very high flow rates, duration of exhalation is short. As a result, stability of exhaled NO concentration may be difficult to confirm and lead to inaccurate estimation of alveolar NO concentration, which is sensitive to data obtained at high flow rates.

In this issue of the Journal of Applied Physiology, Tsoukias and colleagues (8) propose an interesting and creative solution to these problems, using transient analysis of variable exhaled flow rates following a breath-hold maneuver to provide additional information about airway NO diffusion. In this approach, the subject inspires NO-free air to total lung capacity, holds his breath for a known period of time, and then exhales through a variable resistance, keeping mouth pressure constant while the resistance is adjusted to produce an exponentially decreasing flow. The exhaled gas can be divided into three sequential components defined by location during the breath hold: 1) dead space (tubing and proximal airways that do not produce NO), where NO concentrations remain at zero during the breath hold; 2) NO-producing airways, where NO concentrations rise progressively toward airway wall concentrations during the breath hold; and 3) alveoli, where NO concentrations quickly achieve a stable low level during the breath hold. With the assumption that there is no axial mixing of gas, this model predicts a time course of exhaled NO concentration characterized by a brief period of no change from inspired concentration (dead-space gas), a rapid rise followed by a slow rise (gas in NO-producing airways), and a rapid fall followed by a slow rise (alveolar gas traversing NO-producing airways at progressively lower flows). By selecting points on the exhaled NO concentration signal and integrating the simultaneously recorded flow signal backward from these points until integrated volume equaled the assumed volume of NO-producing airways to which each exhaled bolus of gas was exposed, the authors were able to obtain a large number of data points relating the NO concentration in a bolus of exhaled gas to the time that the bolus spent in NO-producing airways. Thus data from a single maneuver were sufficient to obtain fits to the model equations [which turn out to be identical to those previously derived (see Refs. 2,3, 6, 7)], allowing parameters of exhaled NO output to be determined. Results in two normal subjects suggest that the technique has acceptable within-subject reproducibility, yielding values within ranges obtained with the more demanding and difficult multiple constant flow approach. Further evaluation in normal subjects and patients with lung disease is clearly warranted.

The potential weaknesses of this approach derive from its assumptions. The assumption that axial mixing of gas does not occur during breath hold and exhalation is clearly invalid in a qualitative sense, and the quantitative effects of this assumption on parameter estimation remain unknown. As admitted by Tsoukias et al. (8), axial mixing may explain why their model does not fit the actual data very well early in expiration (see Fig. 5 of Tsoukias et al.). Perhaps more important is the need to assume the volume of airways that produce NO (V_air). This parameter is required not only to include in the model equations as a known value but also to determine when to stop the backward integration procedure for exhaled boluses of gas that dwelled in alveoli during the breath hold. Tsoukias et al. recommend using predicted values of anatomic dead space as V_air; however, as they demonstrate in Fig. 8 of their manuscript, the value chosen for V_air can have profound quantitative consequences. For example, assuming that V_air was 200 rather than 100 ml caused a ninefold increase in estimated airway NO diffusing capacity and a sevenfold decrease in estimated airway wall NO concentration. These results suggest that assuming any value for V_air may be unwise, particularly in patients with disease states that may alter airway NO production.

Such considerations need not be cause for great concern. It is highly likely that the innovative approach of Tsoukias et al. (8) will stimulate additional work, which will clarify the impact of their assumptions, assess the utility of their methods, and lead to development of alternative approaches that require fewer assumptions while retaining the power of the breath hold and convenience of transient analysis that their paper so clearly demonstrates.

FOOTNOTES

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.