Model simulations of gas mixing and ventilation distribution in the human lung

A new lung model that incorporates intra-acinar diffusion- and convection-dependent inhomogeneities (DCDI) and interregional and intraregional convection-dependent inhomogeneities (CDI) is described. The model is divided into two regions, each containing two subunits. Each of the four subunits in the model consists of a multi-branch-point structure, based on the anatomic data from Haefeli-Bleuer and Weibel (Anat. Record 220: 401-414, 1988). The subunit turnover (TO), i.e., the ratio of subunit tidal to resting volume, and the flow sequences (FS) between the subunits are used as model parameters. The model simulates the normalized alveolar slope (Sn), Fowler and Bohr dead space (VDF and VDB), and alveolar mixing efficiency (AME) as a function of breath number (n) during a multiple-breath N2 washout (MBNW). For the first breath of the MBNW, these indexes are poorly sensitive to the TO distribution or FS between the subunits. However, as the washout proceeds, the n dependence of both Sn and VDB becomes markedly distinct for simulations with different TO and FS. VDF increases only slightly with n during the MBNW for a large range of TO and FS combinations, and AME is independent of FS. Comparison of published experimental observations with model simulations gave a consistent picture of ventilation maldistribution in the human lung. MBNW simulations in conditions of weightlessness, which will be performed shortly in Spacelab, suggest that it will be possible to evaluate quantitatively the intraregional elastic inhomogeneities in the human lung.     

Respiratory mechanics and breathing pattern during and following maximal exercise

We looked for evidence of changes in lung elastic recoil and of inspiratory muscle fatigue at maximal exercise in seven normal subjects. Esophageal pressure, flow, and volume were measured during spontaneous breathing at increasing levels of cycle exercise to maximum. Total lung capacity (TLC) was determined at rest and immediately before exercise termination using a N2-washout technique. Maximal inspiratory pressure and inspiratory capacity were measured at 1-min intervals. The time course of instantaneous dynamic pressure of respiratory muscles (Pmus) was calculated for the spontaneous breaths immediately preceding exercise termination. TLC volume and lung elastic recoil at TLC were the same at the end of exercise as at rest. Maximum static inspiratory pressures at exercise termination were not reduced. However, mean Pmus of spontaneous breaths at end exercise exceeded 15% of maximum inspiratory pressure in five of the subjects. We conclude that lung elastic recoil is unchanged even at maximal exercise and that, while inspiratory muscles operate within a potentially fatiguing range, the high levels of ventilation observed during maximal exercise are not maintained for a sufficient time to result in mechanical fatigue.     

Effect of increased static lung recoil on bronchial dimesions of excised lungs.

We measured bronchial diameters and lengths during static deflation and inflation in eight excised dog lobes before and after static lung recoil (Pst(L)) had been significantly increased by cooling the lobe for 48 h at 4 degrees C and ventilating it for 3 h. In control lobes, bronchial diameters were the same at any volume even though Pst(L) was different during inflation and deflation. These results agree with those of Hughes et al. (J. Appl. Physiol. 32: 25-35, 1972). However, when Pst(L) was increased, diameters at a given volume were significantly increased over control values; diameters at a given pressure were nearly the same as the controls. Therefore, under these conditions, bronchial diameter did not conform to lung volume. The ventilation process appeared to alter the circumferential elastic properties of the bronchi because diameters at all pressures were slightly larger after ventilation. Bronchial length-volume relationships were the same in both control and ventilated lobes. Thus, when Pst(L) was markedly increased, diameter corresponded best to lung recoil and length to lung volume.     

Static lung-lung interactions in unilateral emphysema.

Motivated by the introduction of single-lung transplantation into clinical practice, we compared the static mechanical properties of the respiratory system in six supine dogs before (at baseline) with those after the induction of unilateral emphysema. Relaxation volume (Vrel), total lung capacity (TLC), and static compliance of the emphysematous lung increased to 214 +/- 68, 186 +/- 39, and 253 +/- 95% (SD) of baseline, respectively. Vrel of the nonemphysematous lung fell to 81 +/- 28% of baseline, with no significant change in TLC of the nonemphysematous lung or its pressure-volume relationship, indicating that unilateral hyperinflation does not cause dropout of contralateral lung units. After unilateral emphysema, the chest wall shifted to a higher unstressed or neutral volume (when pleural pressure equals atmospheric pressure) in three of six animals, minimizing the anticipated decrease in lung recoil pressure at the higher respiratory system Vrel. The pattern of relative lung emptying in the intact dog and in the excised lungs was similar during stepwise deflations from TLC, suggesting that mean pleural pressure of the hemithoraces is equal. We conclude that in the dog the static volume distribution between emphysematous and nonemphysematous lungs is determined only by differences in lung recoil and compliance.     

Contribution of tree structures in the lung to lung elastic recoil

The direct contribution of forces in tree structures in the lung to lung recoil pressure and changes in recoil pressure induced by alterations of the forces are analyzed. The analysis distinguishes the contributions of axial and circumferential tensions in the trees and indicates that only axial tensions directly contribute to static recoil. This contribution is derived from analysis of the axial forces transmitted across a random plane transecting the lung. The change in recoil pressure induced by changes in axial tension is similarly derived. Alterations of circumferential tensions in the trees indirectly change recoil by causing nonuniform deformations of the surrounding lung parenchyma, and a continuum elasticity solution for the stress induced by the deformations is derived. Sample calculations are presented for the airway tree based on available data on airway morphometric and mechanical properties. The increase in recoil pressure accompanying increases in axial and circumferential tensions with contraction of airway smooth muscle is also analyzed. The calculations indicate that axial stresses in the airway tree out to bronchioles directly contribute only a small fraction of the static recoil pressure. However, it is found that contraction of smooth muscle in these airways can increase recoil pressure appreciably (10-20%), mainly by the deformation of the parenchyma with increases in circumferential tension in smaller airways. The results indicate that the geometric and mechanical properties of the airway tree are such that only peripheral elements of the tree can substantially affect the elastic properties of the lung. The possible contributions of vascular trees for which data on mechanical and morphometric properties are more limited are also discussed.     

Relative hysteresis of the airways and lung parenchyma in normal subjects

We evaluated the mechanical properties of the airways sequentially from the glottis toward the main bronchi in 10 normal subjects. Plots of airway cross-sectional area vs. lung volume, measured during inspiration and expiration, were used to determine the relative magnitude of the airways vs. parenchymal hysteresis. Airway cross-sectional area was measured by means of the acoustic reflection technique. We found that the hysteresis of the proximal part of the trachea was greater than that of the lung parenchyma, whereas the hysteresis of the distal trachea and subcarinal segments of the airways was smaller than that of the lung parenchyma. The transition zone between the proximal and the more distal airway properties occurred 8-26 cm distal to the glottis. This transition zone was reproducible in its location on repeated testing in each subject but varied among subjects. To the extent that relative hysteresis of the airways depends on bronchomotor tone, our findings suggest that the bronchomotor tone is inhomogeneous, being maximal at the proximal part of the trachea and gradually decreasing toward the more distal trachea and subcarinal airway segments.     

Effect of bronchomotor tone on work rate of breathing

We studied the optimal airway caliber for minimizing the work rate of breathing in the lung (W) with different bronchomotor tones in six normal subjects. The inhalation of methacholine contracted airway smooth muscle, and the inhalation of salbutamol relaxed it. To calculate W at a given alveolar ventilation (VA), anatomical dead space (VDanat), pulmonary resistance (RL), and dynamic compliance were measured simultaneously, breath by breath, during various breathing maneuvers. VDanat increased and RL decreased with both increased breathing frequency and tidal volume, even at a given airway tone. This suggests that the airway caliber varied even at a given bronchomotor tone. The minimum W at a given VA increased in constricted airways, but there was no significant difference between control airways after saline inhalation and relaxed airways. It has been suggested that airway smooth muscle tones at both control and relaxed conditions bring W to a minimum and that the airway smooth muscle tone existing in the control state acts to keep the airway caliber optimal in order to minimize the W and stabilize the airway mechanics.     

CO2 relaxes parenchyma in the liquid-filled rat lung.

CO(2) regulation of lung compliance is currently explained by pH- and CO(2)-dependent changes in alveolar surface forces and bronchomotor tone. We hypothesized that in addition to, but independently of, those mechanisms, the parenchyma tissue responds to hypercapnia and hypocapnia by relaxing and contracting, respectively, thereby improving local matching of ventilation (Va) to perfusion (Q). Twenty adult rats were slowly ventilated with modified Krebs solution (rate = 3 min(-1), 37 degrees C, open chest) to produce unperfused living lung preparations free of intra-airway surface forces. The solution was gassed with 21% O(2), balance N(2), and CO(2) varied to produce alveolar hypocapnia (Pco(2) = 26.1 +/- 2.4 mmHg, pH = 7.56 +/- 0.04) or hypercapnia (Pco(2) = 55.0 +/- 2.3 mmHg, pH = 7.23 +/- 0.02). The results show that lung recoil, as indicated from airway pressure measured during a breathhold following a large volume inspiration, is reduced approximately 30% when exposed to hypercapnia vs. hypocapnia (P < 0.0001, paired t-test), but stress relaxation and flow-dependent airway resistance were unaltered. Increasing CO(2) from hypo- to hypercapnic levels caused a substantial, significant decrease in the quasi-static pressure-volume relationship, as measured after inspiration and expiration of several tidal volumes, but hysteresis was unaltered. Furthermore, addition of the glycolytic inhibitor NaF abolished CO(2) effects on lung recoil. The results suggest that lung parenchyma tissue relaxation, arising from active elements in response to increasing alveolar CO(2), is independent of (and apparently in parallel with) passive tissue elements and may actively contribute to Va/Q matching.     

Effect of chest strapping on regional lung function.

We studied lung mechanics and regional lung function in five young men during restrictive chest strapping. The effects on lung mechanics were similar to those noted by others in that lung elastic recoil increased as did maximum expiratory flow at low lung volumes. Chest strapping reduced the maximum expiratory flow observed at a given elastic recoil pressure. Breathing helium increased maximum expiratory flow less when subjects were strapped than when they were not. These findings indicated that strapping decreased the caliber of airways upstream from the equal pressure point. Regional lung volumes from apex to base were measured with xenon 133 while subjects were seated. The distribution of regional volumes was measured at RV, and at volumes equal to strapped FRC and strapped TLC; no change due to chest strapping was observed. Similarly, the regional distribution of 133Xe boluses inhaled at RV and strapped TLC was unaffected by chest strapping. Closing capacity decreased with chest strapping. We concluded that airway closure decreased during chest strapping and that airway closure was not the cause of the observed increase in elastic recoil of the lung. The combination of decreased slope of the static pressure-volume curve and unchanged regional volumes suggested that strapping increased the apex-to-base pleural pressure gradient.     

Inter- and intraregional ventilation inhomogeneity in hypergravity and after pressurization of an anti-G suit.

This study assessed the effects of increased gravity in the head-to-foot direction (+G(z)) and anti-G suit (AGS) pressurization on functional residual capacity (FRC), the volume of trapped gas (V(TG)), and ventilation distribution by using inert- gas washout. Normalized phase III slope (Sn(III)) analysis was used to determine the effects on inter- and intraregional ventilation inhomogeneity. Twelve men performed multiple-breath washouts of SF(6) and He in a human centrifuge at +1 to +3 G(z) wearing an AGS pressurized to 0, 6, or 12 kPa. Hypergravity produced moderately increased FRC, V(TG), and overall and inter- and intraregional inhomogeneities. In normogravity, AGS pressurization resulted in reduced FRC and increased V(TG), overall, and inter- and intraregional inhomogeneities. Inflation of the AGS to 12 kPa at +3 G(z) reduced FRC markedly and caused marked gas trapping and intraregional inhomogeneity, whereas interregional inhomogeneity decreased. In conclusion, increased +G(z) impairs ventilation distribution not only between widely separated lung regions, but also within small lung units. Pressurizing an AGS in hypergravity causes extensive gas trapping accompanied by reduced interregional inhomogeneity and, apparently, results in greater intraregional inhomogeneity.     

Airway hysteresis in normal subjects and individuals with chronic airflow obstruction

Specific conductance (sGaw) was measured without prior pharmacological induction of bronchoconstriction before and 5-10 s after a total lung capacity (TLC) volume history in normal subjects and in individuals with chronic airflow obstruction (CAO); increased sGaw after inspiration to TLC was considered evidence of airway hysteresis. Lung elastic recoil [Pst(L)] was also measured before and after inspiration to TLC. In the normal subjects 1) prebronchodilator sGaw increased significantly, whereas Pst(L) decreased significantly after inspiration to TLC; 2) modulators of cyclooxygenase activity had no significant effects on sGaw responses to deep inspiration; and 3) airway hysteresis diminished after inhalation of atropine or metaproterenol. In the CAO group 1) prebronchodilator sGaw and Pst(L) decreased significantly after inspiration to TLC, and 2) bronchoconstriction after deep inspiration diminished after inhalation of atropine or metaproterenol. This study demonstrates that normal airways exhibit hysteresis even without alteration of resting airway tone and that airway hysteresis is impaired in CAO.     

Vagal control of central and peripheral pulmonary resistance in developing piglets

To study the postnatal maturation of vagal control of airway muscle tone, we determined the effects of vagotomy and supramaximal vagal stimulation on the resistance of the respiratory system in eight newborn and seven 6-wk-old piglets. Because the lung periphery has distinctive responses to cholinergic agonists and a lower density of vagal fibers and cholinergic receptors than the central airways, we partitioned the respiratory resistance of the piglets between central airways (Rc) and peripheral airways and lung tissue (Rp) with bronchial catheters inserted in a retrograde manner. The piglets were anesthetized with alpha-chloralose and ventilated with positive airway pressure. Vagotomy did not change Rc or Rp in either the newborn or the 6-wk-old piglets. Vagal stimulation, on the other hand, increased both Rc (median increase 53% in the newborn and 72% in the 6-wk-old piglets) and Rp (54 and 42%, respectively). At all states of vagal tone, Rp increased as the lungs were inflated, suggesting a large contribution of tissue viscoelasticity to this resistance. Our results demonstrate that vagal bronchomotor tone is absent during mechanical ventilation with positive pressure in the developing piglet. However, vagal innervation of both central airways and tissue contractile elements is functionally competent at the time of birth in this species.     

Respiratory mechanics and lung histology in normal rats anesthetized with sevoflurane

Respiratory system, lung, and chest wallmechanical properties were subdivided into their resistive, elastic,and viscoelastic/inhomogeneous components in normal rats, to define thesites of action of sevoflurane. In addition, we aimed to determine theextent to which pretreatment with atropine modified theseparameters. Twenty-four rats were divided into four groups ofsix animals each: in the P group, rats were sedated (diazepam) andanesthetized with pentobarbital sodium; in the S group, sevoflurane wasadministered; in the AP and AS groups, atropine was injected 20 minbefore sedation/anesthesia with pentobarbital and sevoflurane,respectively. Sevoflurane increased lung viscoelastic/inhomogeneouspressures and static elastance compared with rats belonging to the Pgroup. In AS rats, lung static elastance increased in relation to theAP group. In conclusion, sevoflurane anesthesia acted not at the airwaylevel but at the lung periphery, stiffening lung tissues and increasing mechanical inhomogeneities. These findings were supported by the histological demonstration of increased areas of alveolar collapse andhyperinflation. The pretreatment with atropine reduced central andperipheral airway secretion, thus lessening lung inhomogeneities.

     

Inhaled nitric oxide reverses PAF-dependent bronchoconstriction in the pig

In six anesthetized, paralyzed, mechanically ventilated pigs we evaluated the respiratory effects of inhaled nitric oxide (NO) (80 ppm in O₂) under control conditions and after platelet-activating factor (PAF) administration (50 ng/kg, i.v.). PAF was also administered to the same pigs after pretreatment with indomethacin (3 mg/kg, i.v.). The mechanical properties of the respiratory system were evaluated by the rapid airway occlusion technique. With this technique the overall respiratory resistances, the airway resistances, and the additional resistances of respiratory system and lung can be evaluated. The results show that NO inhaled by the pig at 80 ppm for 6 min under control conditions reduced static and dynamic elastances of the respiratory system and lung and pulmonary arterial pressure, without modifying bronchomotor tone. Therefore, NO reduced the PAF-dependent changes in resistances and in static and dynamic elastances of the respiratory system and lung. The modest change in elastances caused by PAF in pigs pretreated with indomethacin was reduced by NO inhalation, which also has a mild bronchodilatory effect. The changes in elastances appear to be correlated with the pulmonary vasodilator activity of inhaled NO.     

Effects of ventilation inhomogeneity on DLcoSB-3EQ in normal subjects.

In patients with airflow obstruction, we found that ventilation inhomogeneity during vital capacity single-breath maneuvers was associated with decreases in the three-equation single-breath CO diffusing capacity of the lung (DLcoSB-3EQ) when breath-hold time (tBH) decreased. We postulated that this was due to a significant resistance to diffusive gas mixing within the gas phase of the lung. In this study, we hypothesized that this phenomenon might also occur in normal subjects if the breathing cycle were altered from traditional vital capacity maneuvers to those that increase ventilation inhomogeneity. In 10 normal subjects, we examined the tBH dependence of both DLcoSB-3EQ and the distribution of ventilation, measured by the mixing efficiency and the normalized phase III slope for helium. Preinspiratory lung volume (V0) was increased by keeping the maximum end-inspiratory lung volume (Vmax) constant or by increasing V0 and Vmax. When V0 increased while Vmax was kept constant, we found that the tBH-independent and the tBH-dependent components of ventilation inhomogeneity increased, but DLcoSB-3EQ was independent of V0 and tBH. Increasing V0 and Vmax did not change ventilation inhomogeneity at a tBH of 0 s, but the tBH-dependent component decreased. DLcoSB-3EQ, although independent of tBH, increased slightly with increases in Vmax. We conclude that in normal subjects increases in ventilation inhomogeneity with increases in V0 do not result in DLcoSB-3EQ becoming tBH dependent.     

Elastic characteristics of the lung perivascular interstitial space

An analysis of the elastic behavior of the lung perivascular interstitial space during interstitial fluid accumulation is presented. Fluid accumulation must deform the lung parenchyma and vascular walls that form the interstitial space boundaries. The deformations of these boundaries are predicted from previously published data on the elastic properties of the boundary materials. The analysis gives the relationships among the elastic properties of the boundaries, the compliance of the interstitium, the lung volume, and the lung elastic recoil pressure. Values of the interstitial compliance are predicted to decrease with increasing lung recoil pressure and are dependent on the lung pressure-volume history. At low recoil pressures over 70% of the interstitial compliance results from deformation of the parenchyma. As the recoil pressure increases, either with increasing lung volume or due to the lung pressure-volume history, the contributions of the parenchymal and vascular wall deformations become similar. The predictions are generally consistent with published data on interstitial compliance obtained from measurements of isolated lung weight gain during vascular fluid transudation. This correlation suggests that the elastic behavior of the interstitial space can be accounted for by the known elastic properties of the boundary materials.     

Convection- and diffusion-dependent ventilation maldistribution in normal subjects

We performed multiple-breath N2 washouts (MBNW) with tidal volumes of 1 liter at 8-16 breaths/min and constant flow rates in six normal subjects. For each breath we computed the slope of the alveolar plateau, normalized by the mean expired N2 concentration (Sn), the Bohr dead space (VDB), an index analogous to the Fowler dead space (V50), and the normalized slope of phase II (S2). In four subjects helium (He) and sulfur hexafluoride (SF6) were washed out after equilibration with a 5% gas mixture of each tracer. The Sn for He and SF6 increased in consecutive breaths, but the difference (delta Sn) increased only over the first five breaths, remaining constant thereafter. In all six subjects Sn, VDB, and V50 increased progressively in consecutive breaths of the MBNW, the increase in Sn being the greatest, approximately 290% from the first to the 23-25th breath. In contrast, S2 was unchanged initially and decreased after the sixth breath. The results indicate that after the fifth breath the increase in Sn during a MBNW is diffusion independent and may constitute a sensitive index of convection-dependent inhomogeneity (CDI). Subtraction of this component from the first breath suggests that Sn in a single-breath washout is largely due to a diffusion-dependent mechanism. The latter may reflect an interaction of convection and diffusion within the lung periphery, whereas CDI may comprise ventilation inequality among larger units, subtended by more centrally located branch points.     

Activation of neurons in the rostral ventrolateral medulla increases bronchomotor tone in dogs.

Previous work from this laboratory has demonstrated that the chemical activation of cell bodies in the caudal ventrolateral medulla of chloralose-anesthetized dogs decreased bronchomotor tone by withdrawing cholinergic input to airway smooth muscle. In the present study we determined the bronchomotor responses to microinjection of DL-homocysteic acid (100 mM; 25-50 nl) into the rostral ventrolateral (RVL) medulla of chloralose-anesthetized dogs. Total lung resistance was used as a functional index of bronchomotor tone. Microinjection of DL-homocysteic acid into the 20 sites located in the lateral aspect of the RVL medulla increased both total lung resistance [from 6.5 +/- 0.4 to 9.1 +/- 0.8 (SE) cmH2O.l-1.s; P less than 0.05] and mean arterial pressure (from 125 +/- 5 to 148 +/- 8 mmHg; P less than 0.05). Microinjection of this amino acid into nine sites located in the medial aspect of the RVL medulla increased mean arterial pressure (from 130 +/- 6 to 153 +/- 6 mmHg; P less than 0.05) but had no effect on total lung resistance. We confirmed in three sites that the increase in total lung resistance evoked by microinjection of DL-homocysteic acid was accompanied by an increase in tracheal smooth muscle tension. The increase in total lung resistance evoked by DL-homocysteic acid was not affected by beta-adrenergic blockade but was abolished by muscarinic blockade.     

Rapidly adapting receptors monitor lung compliance in spontaneously breathing dogs

We examined the ability of rapidly adapting receptors (RARs) to monitor changes in dynamic lung compliance (Cdyn) in anesthetized spontaneously breathing dogs by recording RAR impulses from the vagus nerves. We decreased Cdyn in steps through the physiological range by briefly restricting lung expansion with an inflatable cuff around the chest and recording the response after deflating the cuff; we restored Cdyn to control by hyperinflating the lungs. Of 45 RARs, 34 were stimulated by a 40 +/- 2% reduction in Cdyn, their inspiratory discharge increasing on average more than threefold. Two-thirds of responsive RARs were stimulated by less than or equal to 20% reductions in Cdyn; in most, firing increased proportionately with lung stiffness (1/Cdyn) as Cdyn was decreased further. Stimulation by reduced Cdyn was not simply a function of the concomitant increase in transpulmonary pressure, because similar increases in pressure produced by increasing tidal volume produced smaller increases in firing. RAR stimulation was unaffected by atropine and, hence, was not dependent on neurally mediated changes in bronchomotor tone. Our results indicate that during spontaneous breathing RARs provide a signal inversely proportional to Cdyn.     

Lung elastic recoil during breathing at increased lung volume.

During dynamic hyperinflation with induced bronchoconstriction, there is a reduction in lung elastic recoil at constant lung volume (R. Pellegrino, O. Wilson, G. Jenouri, and J. R. Rodarte. J. Appl. Physiol. 81: 964-975, 1996). In the present study, lung elastic recoil at control end inspiration was measured in normal subjects in a volume displacement plethysmograph before and after voluntary increases in mean lung volume, which were achieved by one tidal volume increase in functional residual capacity (FRC) with constant tidal volume and by doubling tidal volume with constant FRC. Lung elastic recoil at control end inspiration was significantly decreased by approximately 10% within four breaths of increasing FRC. When tidal volume was doubled, the decrease in computed lung recoil at control end inspiration was not significant. Because voluntary increases of lung volume should not produce airway closure, we conclude that stress relaxation was responsible for the decrease in lung recoil.