INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It has been known for many years that the response of asthmatic subjects to a deep inspiration differs from that observed in normal healthy subjects (1). A deep inspiration causes a decrease in airway resistance in normal subjects, whereas asthmatics demonstrate either no change or a slight increase in airway resistance. Skloot and coworkers (8) postulated that the inability to dilate airways during lung inflation is a primary defect in asthma. Their study (8) also showed that in the absence of a deep inspiration during methacholine (MCh) challenge, normal subjects had a greatly exaggerated and sustained response to this agonist. It was suggested that asthmatic airways could be modeled by this condition in normal subjects. Recent experiments by Brusasco and coworkers (9), however, suggest that there are more intrinsic differences between the responses to lung inflation in airways from asthmatic and normal subjects. To better understand the mechanisms underlying this controversy, it is necessary to be able to assess the responses of airways directly, something that conventional pulmonary function tests in human subjects unfortunately cannot do.

The size of the airways during normal tidal ventilation is also controlled by similar factors that determine the response to deep inspiration, i.e., the lung volume, the degree of smooth muscle activation, and the normal tidal stresses exerted by the parenchyma. The effect of such rhythmic cycling in decreasing the smooth muscle tone has been shown in vitro (10, 11), and more recently, similar in vivo effects of tidal stresses have been documented (12). Warner and Gunst (13) originally demonstrated that the rhythmic stretching associated with tidal breathing can decrease not only the baseline lung resistance, but also the response to MCh. This effect of tidal breathing limiting the degree of airway smooth muscle constriction was supported by the more recent studies of Tepper and coworkers and Shen and coworkers (12, 14). This group proposed a mechanistic explanation that involves changes in the plasticity of the smooth muscle cellular cytostructure (15, 16). Another hypothesis that could account for the mechanism underlying these observations in humans was published in recent studies by Fredberg and associates (17, 18). In this work, it was proposed that the steady-state muscle force would be determined by a balance between high and low energy cross bridge dynamics. If the contractile stimulus is increased, then the number of cross bridges increases, and, given enough time at a constant load, the rapidly cycling cross bridges progressively convert to slowly cycling "latch" bridges, and the muscle stiffness increases. This model also predicted that the hysteresis of such latched smooth muscle would be decreased (17). If this state occurred in vivo, then one might predict that there might be little effect of a deep inspiration in dilating the airways. Fredberg and associates also speculated that normal tidal breathing provides a sufficient stress to keep the latch state from occurring in normal lungs (18).

In the present study, we have performed experiments using a direct imaging approach that allow us to obtain measurements of airway and parenchymal dimensions that bear on the applicability of these models in vivo. Our results show that removing tidal stresses at FRC after MCh challenge is sufficient to change the normal dilatory response to deep inspiration into one of contraction. This mechanism thus may be operative in the development of the asthmatic pathology.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our study protocol was approved by The Johns Hopkins Animal Care and Use Committee. Five beagle dogs weighing approximately 10 kg were anesthetized with thiopental (15 mg/kg induction dose followed by 10 mg/kg/h intravenous maintenance dose). After induction of anesthesia, the dogs were paralyzed with 0.5 mg/kg of succinylcholine with occasional supplemental doses as required to ensure no respiratory motion during imaging. After tracheal intubation with a Univent endotracheal tube having an 8.0-mm interior diameter (Vitaid Ltd., Lewiston, NY), the dogs were placed supine and their lungs were ventilated with room air with a volume-cycled ventilator (Harvard Apparatus, Millus, MA) at a tidal volume of 15 ml/kg and a rate of 18 breaths/min. A stable depth of anesthesia was maintained by monitoring heart rate changes and eyelash reflex, and a bilateral vagotomy was performed.

Imaging and Analysis of Airways and Lung Volumes

Airways. High-resolution computed tomography (HRCT) scans were obtained with a Somatom Plus Scanner (Siemens, Iselin, NJ) using a spiral mode to acquire 33 contiguous images in a single 20-s breath hold (apnea) at 137 kVp, and 165 mA. The images were reconstructed as 2-mm slice thickness and a 256 × 256 matrix using a maximum zoom of 4.0 (12 cm field of view) and a high spatial frequency (resolution) algorithm that enhanced edge detection, at a window level of 450 Hounsfield units (HU) and a window width of 1,350 HU. These settings have been shown to provide accurate measurement of airway lumen size in airways as small as 2 mm (19, 20). For repeated airway measurements in a given dog within each experimental protocol, adjacent anatomic landmarks, such as airway or vascular branching points, were defined and the airways were matched by these adjacent landmarks and measured.

Lung volume. Standard resolution computed tomography scans were obtained with a Somatom Plus Scanner (Siemens, Iselin, NJ) using a 1-s scan time, 137 kVp, and 210 mA. The images were reconstructed as a 256 × 256 matrix using a maximum zoom of 2.0. Twenty-three contiguous sections were obtained, starting at the apex of the lungs and moving caudally to the bases using a 8-mm table feed and an 8-mm slice thickness. Images were reconstructed with the use of a standard lung algorithm at the aforementioned window settings. The area of the lungs on each computed tomographic (CT) scan was defined as the area within the pleural border excluding the heart and diaphragm. The total lung volume was calculated as the sum of the lung areas of each CT slice times the slice thickness.

The HRCT images were analyzed using the airway analysis module of the Volumetric Image and Display Analysis (VIDA) image analysis software package (Department of Radiology, Division of Physiologic Imaging, University of Iowa, Iowa City, IA) as previously described and validated (21, 22). The HRCT images were transferred to a UNIX-based Sun workstation. An initial isocontour was drawn within each airway lumen, and the software program then automatically located the perimeter of the airway lumen by sending out rays in a spoke-wheel fashion to a predesignated pixel intensity level that defines the lumenal edge of the airway wall. Intraobserver and interobserver accuracy and variability of the software program using this HRCT technique in phantoms, consisting of rigid tubes to measure known areas, have been previously documented by us (20) and others (21) to be highly resistant to operator bias.

Protocol

Dogs were anesthetized and ventilated as previously described. To standardized lung volume history, the dogs were initially given a deep inspiration of both lungs to 30 cm H₂O. Under direct fiberoptic visualization, the deflated blocker cuff to the Univent endotracheal tube was advanced into either the right or the left mainstem bronchus. The dogs then received a continuous intravenous infusion of 67 µg/min methacholine (Sigma Chemical, St. Louis, MO) while both lungs were ventilated. After 10 min, a stable constriction plateau was reached. Ventilation was stopped, 1 min later the HRCT scans were acquired (over a 20-s period), and ventilation was resumed. One lung was randomized for subsequent ventilation. At end-expiration, ventilation was stopped and the cuff of the bronchial blocker was inflated and the selected lung was blocked from ventilation at FRC. Ventilation was then resumed for the other lung. After 10 min, this ventilation was stopped, the blocker cuff was deflated and both lungs were inflated to 30 cm H₂O for 10 s; they were then allowed to passively expire to FRC. One minute later, the HRCT scans were repeated as previously described. After the scans, normal two-lung ventilation was resumed.

Analysis

The completely relaxed airway after atropine (0.2 mg/kg intravenously) at 30 cm H₂O was defined as 100% (relaxed state, maximum size), and airway lumenal areas were expressed as a percent of this maximally relaxed area. The maximal lung volume used for normalization was also measured after atropine at 30 cm H₂O. Atropine was administered at the end of all experimental protocols; previous work has demonstrated that this dose of atropine abolishes baseline smooth muscle tone in dogs (22). All measurements of airway and lung size were measured at FRC, and each airway in each dog served as its own control. Mean airway areas as a percent of relaxed areas were compared before and after a deep inspiration in the ventilated and the nonventilated lungs by paired t test. To compare changes in airway area with corresponding dimensional changes in lung volume, we calculated lung volume to the two-thirds power (V^2/3). Mean lung volumes to the 2/3 power as a percent of the maximal lung volume were compared before and after a deep inspiration in the ventilated and the nonventilated lungs by paired t test. Significance was considered to be p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eighty-four airways (2.3 to 13.9 mm relaxed diameter) in the five dogs were matched and measured under all challenge conditions. Figure 1 shows that a similar distribution of airways was studied in the ventilated and nonventilated lungs. Figure 2 shows the changes in lung volume and mean airway area after the MCh challenge and deep inspiration. During MCh infusion, the airways constricted to an average of 40 ± 2% (mean ± SEM) of maximal area (35 ± 3% and 44 ± 8% for the ventilated and the nonventilated airways, respectively, p = 0.33). (Lung volume)^2/3 averaged 58 ± 1% of maximal volume (59 ± 1% and 57 ± 2% for the ventilated and the nonventilated lungs, respectively, p = 0.23). However, after the 10-min period of ventilatory stasis in one lung, the subsequent deep inspiration led to an increase in airway area in the ventilated lung and a decrease in area of the unventilated lung. Figure 2 shows that, on the ventilated side, lung volumes increased after a deep inspiration (DI), mean 65 ± 3% (p = 0.007 compared with pre-DI), concomitant with an increase in the mean airway area to 39 ± 2% (p = 0.0001, compared with pre-DI). On the nonventilated side, (lung volume)^2/3 decreased after a deep inspiration, mean 48 ± 2% (p = 0.01 compared with pre-DI), concomitant with a decrease in mean airway area after a deep inspiration to 34 ± 2% (p < 0.0001 compared with pre-DI). Figure 3 shows the correlation between the mean changes in airway area and the changes in (lung volume)^2/3 for both ventilated and nonventilated lungs.

View larger version (34K): [in this window] [in a new window]   Figure 1.   Distribution of airway size in airways analyzed from the ventilated and nonventilated lungs.

View larger version (29K): [in this window] [in a new window]   Figure 2.   Changes in airway area (upper graph) and (lung volume)2/3 (lower graph) in the ventilated and nonventilated lungs before and after deep inspiration (DI). p Values (paired t test) refer to significance levels between adjacent bars.

View larger version (20K): [in this window] [in a new window]   Figure 3.   Correlation between the changes in airway area and (lung volume)2/3 in the ventilated and nonventilated lungs before and after DI.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results clearly demonstrate that the airway response to a deep inspiration can be qualitatively changed by altering the pattern of normal tidal stresses. By eliminating tidal stresses in one lung, we were able to change the normal airway dilation that would occur in response to a deep inspiration to one of airway constriction. This change occurred after the airways were moderately constricted with intravenous MCh. These experimental observations are striking, because they provide a potential mechanism which allows conversion of the normal dilation after a deep inspiration to that of constriction. All that is needed to cause this qualitative change is to minimize the tidal stresses for a period of time, and with a constant level of agonist stimulation, changes occur in the smooth muscle that result in a constriction after the deep inspiration. We believe that this qualitative change in the response to deep inspiration that we observe in the dog is related to the qualitative difference seen between asthmatics and normals. However, as discussed subsequently, the mechanisms underlying the observation still are not completely clear.

The model of Fredberg and coworkers argues that the stresses associated with normal tidal breathing are sufficient to keep the airway smooth muscle from attaining a low energy latch state (23). According to predictions based on this model, if one were to eliminate tidal stresses, the muscle cross bridges must soon attain a latch state, consisting of increased stiffness and little hysteresis. Although this model is not directly concerned with the response to deep inspiration, it can be used to predict specific responses. That is, if decreasing or eliminating tidal stresses can cause airway smooth muscle to obtain a very stiff latch state with low hysteresis, then the airway response to a deep inspiration in situ should be one of further narrowing. The logic behind this interpretation is based on the interaction of the airway with surrounding parenchyma. If the airway were isolated, then the passive recovery from any large stretch would leave the airway at a larger size. However, in situ the final airway size depends on the resultant recoil pressure exerted by the surrounding lung (24). If the deep inspiration did cause a normal reduction in lung elastic recoil, but had little direct effect on the stiff airway, then the airway would end up smaller.

Indeed, our results appear just as extrapolations from this model would predict. However, the situation in the intact living lung is more complex than just the response of smooth muscle in the airway wall. In vivo the lung parenchyma behaves as if it were comprised of smooth muscle, showing an increase in both lung tissue resistance and parenchymal stiffness in response to contractile agonists (25). The impact of such contractile responses of the tissue surrounding the airways has not been analyzed, and the effect of a transient deep inspiration on this interaction is even less well modeled. Our study provides data on which to test models as they develop. The CT methods used allowed us to also measure lung volumes in addition to airway size. Results shown in Figure 2 demonstrate that the lung parenchymal dimensions behave similarly to the changes observed in the airways to a deep inspiration.

This interaction is emphasized in Figure 3. In the ventilated lung, a deep inspiration leads to proportional increases in (lung volume)^2/3 and airway area. In the nonventilated lung, there is a proportional decrease in both (lung volume)^2/3 and airway area. These changes suggest that whatever changes occur in the smooth muscle of the airway walls whose lumenal size we measure by HRCT (2 to 12 mm in diameter) also occur in the smooth muscle responsible for changes in parenchymal stiffness. There is negligible smooth muscle in the alveolar walls, though often there is some in the mouths of the alveolar ducts (26). Because this smooth muscle is a continuation of the bronchial smooth muscle (26), parenchymal contraction with a cholinergic agonist thus reflects contraction of airway smooth muscle. Although increased stiffness can occur from contraction of the conducting airways (27), it is likely dominated by contraction in the smallest membranous airways. Thus, normal ventilatory stresses and their absence likely have a major effect on contracted smooth muscle throughout the airway tree. Although this mechanism is related to what has been called the relative hysteresis hypothesis (2, 28), the actual situation in vivo depends on more than just the relative hysteresis between airways and parenchyma. What matters is how the smooth muscle in the airway responds to stress and stretch. Both the lung volume and lung recoil pressure following the deep inspiration can have independent effects on the resultant size of the contracted airways, and the airway tone itself may be altered by the deep inspiration.

The parallel changes in parenchymal volume and airway area, however, do suggest a uniformity of constriction throughout the airway tree. Airway size is known to be strongly influenced by lung volume (22), so if lung volume is smaller after deep inspiration, then this by itself could cause the airway to be smaller. Two questions arise from this correlation, one being whether the reduction in lung volume itself could cause the magnitude of changes in airway area observed, and the other is, why is the lung volume smaller? Regarding the first, we don't think that the decrease in lung volume can account for a major part of the reduction in airway area, simply because the change in lung volume is too small (approximately 15%). In a previous report (22) we showed considerable variability in the response of airways to lung volume changes, but on average the mean area of contracted canine airways changed approximately 2% for each cm H₂O change in transpulmonary pressure. We did not measure transpulmonary pressure in the current study, but to obtain the observed 25% change in airway area in Figure 3, the lung volume at FRC would have thus had to be sufficient to cause a greater than 12 cm H₂O change in transpulmonary pressure. This seems highly unreasonable. Regarding the reasons why lung volume decreases at all, there must be some increased contraction of the smooth muscle triggered by the deep inspiration. Because there is little a priori reason to expect there to be any qualitatively different response between the airway smooth muscle that causes conducting airway contraction and the airway smooth muscle that causes parenchymal stiffening, we would thus predict that even in the absence of parenchyma (that is, in an excised airway with as stable MCh concentration), following a period of time without tidal stresses, an airway might show a further contraction after a simulated deep inspiration. Under the conditions of our experiment in vivo, the fact that the surrounding lung also gets smaller would be expected to further augment the reduction in airway size. However, the situation becomes very complex, with changing lung volume, airway size, and airway contraction. Furthermore, the interaction between airways and parenchyma is dependent on the parenchymal shear modulus, and this has been shown to increase with MCh or carbachol (29, 30). An increase in shear modulus per se would lead to tighter coupling between the airways and parenchyma, perhaps even partially reversing the degree of lumenal constriction. However, the recent results of Okazawa and coworkers (30) in rabbit lungs suggest that the increase in parenchymal shear modulus with cholinergic contraction may not be great enough to cause much alteration in the normal coupling. With regard to our experiments, the influence of shear modulus changes must remain speculative, because there are no reported experimental measurements of shear modulus in the constricted canine lung.

Our experiments did not attempt to determine the specific level of reduction of tidal stresses needed to change the response to deep inspiration from one of dilation to one of constriction. Our experimental design of totally eliminating the tidal stresses in one lung was chosen both for experimental simplicity and to maximize any potential effects. There would of course be an expected dose-time effect on the magnitude of the response, but we presently have no information on how this would sort out. That is, if we reduced the tidal volume to say 20% of normal, but lengthened the duration to 20 min, could we achieve the same effect? This information may be very relevant to the difference between the responses to deep inspiration in normals and asthmatics. There also may well be other chemical factors associated with inflammatory processes that could contribute to the contractile response to deep inspiration. One possible factor to consider is that of local hypoxia. Our protocol of sealing one lung was chosen so that the other ventilated lung would cause negligible lung volume (and hence airway) changes in the static lung. However, with no ventilation, the gases in the sealed lung must equilibrate at the mixed venous level. The PCO₂ will thus be only marginally elevated, but there will be a moderate amount of hypoxia in the sealed lung. At the present time we do not know the effect of this hypoxia, but there is little experimental evidence that would suggest that either the MCh contraction or the response to stretch would be greatly affected. Although changes in asthmatic lungs may not mimic the experimental protocol we used, air trapping in the lungs of asthmatics associated with bronchospasm can lead to increased lung volume and altered tidal stresses on the airways and parenchyma. Understanding the dose-time effect of these insidious changes may ultimately affect the management and treatment of asthma.

In summary, we have shown that a constrictor response to deep inspiration can be generated in normal airways by minimizing tidal stresses. The absence of these normal rhythmic stresses alters the smooth muscle throughout the airway tree, such that subsequent large stresses lead to a further constriction. This response may result from increased active smooth muscle tone after the deep inspiration. These results also offer a possible mechanism by which the response to deep inspiration is altered in asthmatic subjects.