INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

RECENT STUDIES have shown that bronchoconstrictor agents induce a substantial increase in tissue resistance (Rti) in several species (15, 29, 23). Three main mechanisms have been invoked to induce changes in Rti after constrictor challenge (11, 21). The first is a pure airway effect in which heterogeneity of airways resistance (Raw) among parallel pathways produces ventilation inhomogeneity and therefore artifactual increases in Rti; the second is a direct constriction of contractile cells and smooth muscle in the parenchyma and in blood vessels; and the third is a secondary response driven by airways smooth muscle contraction in the lung periphery.

According to Fredberg et al. (6), in the absence of disease and gradients in pleural pressure, functional inhomogeneity of the lung arises from intrinsic nonuniformity of lung tissue properties, structural asymmetry of the airway tree, and the mechanical interactions of these features. The combination of inhomogeneities and substantial constriction could produce a nonuniform parallel distribution of ventilation time constants, which would further contribute to the frequency-dependent features of tissue mechanical properties (20). If the influence of these phenomena is ignored, the consequence is an artifactual overestimation of the degree of increase in Rti and hence hysteresivity () (21). Furthermore, several authors recently concluded that there is little change in tissue mechanical properties during bronchoconstriction and that most of the change could be attributed to an artifact of increased airways inhomogeneity (2).

To investigate whether in vivo changes in Rti correspond to an intrinsic mechanism, we have studied the transition from basal conditions to the constricted state in normal and highly heterogeneous rabbit lungs, in which four alveolar capsules in four different lobes were used to detect parallel inhomogeneities during intravenous methacholine (MCh) challenge. Our hypothesis is that, if the changes in Rti are an artifact of parallel inhomogeneities, the time course of the increase in Rti and the development of alveolar heterogeneity should be similar.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animal preparation. Thirteen New Zealand White male rabbits were included in this study. Six of them (weight 2.79 ± 0.53 kg, mean ± SD) were previously exposed to 1-1.5 parts/million ozone (O₃ group) for 2 h in a controlled-ambiance chamber, with a laminar flow of 6 l/min, and were studied 18 h after the beginning of exposure. The other seven rabbits (weight 2.96 ± 0.38 kg) were exposed to air (air group) in the same conditions and were also studied 18 h later.

Experimental procedure. At the time of the study, animals were preanesthetized with diazepam (2 mg/kg) and anesthetized with pentobarbital sodium by slow infusion of 50-60 mg/kg in the marginal vein of the ear. Anesthesia was maintained by an additional hourly dose of 10% of the initial dose. After anesthesia was induced, the upper trachea was dissected and cut. A 3-cm-long Portex no. 4 cannula was inserted into the trachea and tightly bound. Through the anterior part of the tracheal cannula a narrow (90 P) polyethylene catheter was introduced for tracheal pressure (Ptr) measurement. Total length of this catheter was 10 cm, and it protruded 2 mm from the internal opening of the cannula to record Ptr. A jugular venous line was placed for fluid and drug administration. Rabbits were paralyzed with pancuronium (2 mg) and ventilated with a Siemens 900C constant-flow servo ventilator. A warming pad prevented cooling of the animal. Through an upper midline abdominal incision, the diaphragm was cut at the level of the xiphoid, and bilateral pneumothoraxes were introduced. The thorax was widely opened by a midline sternotomy and dissection of costal diaphragm insertions. Basal mechanical ventilation was set at a frequency of 60 breaths/min, a tidal volume (VT) of 5 ml/kg, positive end-expiratory pressure (PEEP) of 5 cmH₂O, and a duty cycle ratio of 0.33.

Protocol. End-inspiratory breath-holding maneuvers were performed to test for leaks. The preparation was allowed to stabilize (~20 min), and afterward total lung inflation maneuvers (peak inspiratory pressure of 20-40 cmH₂O) were performed twice so as to standardize lung volume history. Five minutes later, three baseline recordings (10-s duration each at PEEP of 5 cmH₂O) were sampled 5 min apart to test the stability of the preparation. A continuous recording of 220 s was then performed. At the 30th s, 1 ml of physiological saline, warmed to 37°C, was injected in the jugular vein. Twenty minutes later, a second continuous recording of 220 s was performed. At the 30th s, a bolus of 0.8 mg/kg MCh in 1 ml of physiological saline, warmed to 37°C, was injected in the jugular vein. Five minutes before every recording, two total lung capacity maneuvers were performed to warrant identical volume histories. During the experiment, Ptr was continuously monitored.

Data analysis and calculations. Volume (V) was calculated by digital integration of the flow signal. From , V, and Ptr signals, lung elastance (EL) and total lung resistance (RL) were calculated cycle by cycle by adjusting the equation of motion (26) where K is a constant term reflecting PEEP and the error linked to the residuals of the least-squares adjustment method (12), t is time, and dV(t)/dt is flow as a function of time. Anadat software (RHT-Infodat, Montreal, Quebec, Canada) was used for this purpose.

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Schematic diagram showing electrical analog of linear monocompartmental model used to determine mechanical parameters. dV/dt, change in volume over time; Ptr, tracheal pressure; Raw, airways resistance; Palv, alveolar pressure; Rti, tissue resistance; Eti, tissue elastance.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Typical tracings of flow, volume (Vol), Ptr, and Palv at baseline (A) and after intravenous methacholine (MCh) administration (B) in an air-exposed rabbit. Broken lines show fit of linear model to both pressures.

Stress decomposition analysis. Fredberg and Stamenovic (7) and more recently Ludwig et al. (16) argued that dissipative and elastic processes are coupled at a fundamental level. This is the structural damping hypothesis. The presumption of such coupling leads to the consideration of a new attribute of organ-level dissipative behavior, called tissue , and also referred to as structural damping coefficient (7). According to this theory, the fractional change in Rti in response to neurohumoral or biophysical stimulation can be decomposed into the product of two distinct contributions, the change in  and the change in EL where the subscript 0 denotes values in the control or basal state. The value for  was calculated according to the equation  =  · Rti/EL, where  is the angular frequency (7).

Statistics. To account for individual differences at the beginning of the pulmonary reaction, individual trends have been aligned taking RL as reference. Accordingly, MCh challenge trends were aligned to coincide at the point at which total RL increased. The value of RL was considered to increase significantly (z = 0.05) at a time (T₀) at which it exceeded RL_basal + 1.64_basal, where RL_basal is the average of the values previous to MCh administration, and _basal is the SD for the same values. For graphical representations and group averaging, signals have been cut from 20 s before T₀ to 180 s after T₀. Saline trends have been aligned only according to the recorded time of the beginning of intravenous injection.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

In 11 of 13 cases, the four Pcap values met the criteria of validity at the basal periods previous to saline and to MCh injections. In the other two cases (both in O₃ group), only three valid Pcap could be obtained. We rejected the two "bad" Pcap, since we could not identify the cause of the failure. In all cases, at least one capsule permanently met the criteria of validity; so we were able to determine Rti for every animal and every trend.

Table 1 shows the average values of pulmonary mechanics and heterogeneity parameters in both groups at basal state. As expected, alveolar heterogeneity was significantly greater in the O₃ group, and differences were statistically significant for all mechanical parameters except Raw, probably in relation to the high variability of this parameter in O₃-exposed animals.

As shown in Fig. 3, saline injection did not induce any significant change in the air group. However, it significantly increased CV_i and CV_e in the O₃ group 42 and 25 s after saline injection, respectively, without noticeable changes in mechanical parameters. Increases of both CV_i and CV_e were significant with respect to the average basal value. Basal values of both heterogeneity parameters recovered before MCh injection. CV_mean was lower than both and showed a similar time course (see Fig. 6).

View larger version (35K): [in this window] [in a new window]   Fig. 3.   Time course of ventilatory inhomogeneity and tissue mechanical parameters after saline intrajugular injection in air-exposed (n = 7; A), and O3-exposed (n = 6; B) rabbits. Vertical line indicates time at which saline (Sal) infusion was initiated. CV, coefficients of variation of capsule pressures; solid line, end-inspiratory CV (CVi); dotted line, end-expiratory CV (CVe). EL, dynamic lung elastance. Bars represent SE. Only mean values are represented for both CV.

As shown in Fig. 4, RL and Rti increased after MCh injection in both groups. Changes were greater in O₃-exposed than in air-exposed animals. The % peaked immediately after MCh injection, reflecting a faster increase of RL and Rti in the O₃ group. It is interesting to note that %(Rti) rose before %(RL).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Time course of total lung resistance (RL; A) and Rti (B) after MCh intrajugular injection in air-exposed (n = 7) and O3-exposed (n = 6) rabbits. Vertical dotted line indicates cycle before first significant increase in RL was noticed. %, Significant between-means difference at  = 0.05 as a percentage of average value in air-exposed group. Bars represent SE.

Figure 5 shows the time course of alveolar heterogeneity parameters in both groups. In the air group, a significant increase in CV_i was observed 10 s after the increase in RL, and CV_e increased significantly as late as 54 s after the increase in RL. Initial changes were small, and a "substantial" increase was not observed until 1 min after the first change in RL. In the O₃ group, alveolar heterogeneity developed earlier, as shown by the initial peak change in % of both variables. Comparison with Figs. 4 and 6 shows that, in the air group, there was a noticeable delay of Rti with respect to CV_i, CV_e, and CV_mean. In our results, Raw changes also preceded changes in CV.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Time course of ventilatory inhomogeneity parameters, in air-exposed (n = 7) and O3-exposed (n = 6) rabbits. A: CVe; B: CVi. Dotted line indicates cycle before first significant increase in RL was noticed. Bars represent SE.

Figure 6 shows the time course of the components of stress decomposition in air-exposed rabbits. The dissociation between hysteretic and elastic changes is evident, with hysteretic changes predominating during the early transition to the constricted state. For ease of reference, average trend values of CV_i, CV_e, and CV_mean are represented, and a clear out-of-phase behavior between lung tissue hysteresis and both heterogeneity parameters can be observed.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Time course of dynamic elastance (EL/EL0) and hysteresivity (/0) changes over basal values after MCh intrajugular injection in air-exposed rabbits (n = 7). Bottom: time course of ventilatory heterogeneity parameters (CV); dotted line, CVe; continuous line, CVi; dots, average CV by cycle (CVmean). Only mean values of CVi, CVe, and CVmean are represented. Vertical dotted line indicates cycle before first significant increase in RL was noticed.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The main finding of this study has been the delayed time course of parallel inhomogeneity changes in relation to changes in total or tissue resistances and hysteresivity. Alveolar capsules have served both for alveolar heterogeneity evaluation and the measurement of tissue mechanical properties. Alveolar heterogeneity has been estimated from all capsules, accepting that they reflect underlying subpleural pressure. By contrast, measurement of tissue mechanics has been performed only with those capsules that actually reflect alveolar pressure in phase with pulmonary expansion as measured at the trachea.

An intrinsic drawback of the alveolar-capsule method is that it samples only those units close to the pleura. The atypical distribution of path lengths from alveolus to airway opening of such units makes them especially susceptible to parallel inhomogeneities (9). Previous authors (1, 6, 32) have taken advantage of this anatomical arrangement to quantify alveolar heterogeneity from capsule pressures. Another problem is that the capsules are placed at the top of the lung, so that they do not see what happens in the dependent regions underneath. In a complementary set of experiments performed in six rabbits, capsules were placed in the top of 11 lower lobes and in the lower third of their corresponding diaphragmatic side. Tissue resistance showed no significant difference (paired t-test) when measured from top and bottom capsule pressure (mean difference ± SD: 0.339 ± 1.407 cmH₂O · l¹ · s). Furthermore, in a single experiment after intravenous MCh administration, no lag between mechanical parameters, calculated from either top or bottom capsules, was observed; thus it seems that there was no preferential delivery of drug to the basal parts of the lung. An increase in alveolar closure in the dependent parts of the lung after MCh cannot be excluded, because no capsule was placed in the bottom part of the lung. Such dependent alveolar closure would be expected to influence lung tissue mechanical parameters either by increasing dynamic intrinsic PEEP or by increasing tidal ventilation to the nondependent parts of lung. Dynamic intrinsic PEEP (minimal Ptr  0-flow Ptr) increased by only 0.98 ± 0.12 cmH₂O (mean ± SD) in the air-exposed rabbits. Its time course lagged the increase in RL, Rti, and , showing a behavior similar to that of the heterogeneity parameters. Increase of VT does increase tissue resistance but has no effect on hysteresivity, in agreement with previous studies (7, 24). Therefore, although alveolar closure can influence the changes observed in constricted state, it is unlikely that they can explain the transitional changes in our experimental conditions.

To be confident that capsule pressure really does correspond to alveolar pressure, we have stated a number of criteria based on the following assumptions: first, we have assumed a standard linear model for lung mechanics; second, lung elastance is attributed to lung parenchyma; and third, air pressure is considered to equilibrate at end expiration in well-communicated air spaces. We have no doubt that some arguments can be raised against a rigid application of our assumptions. The margins of tolerance serve to smooth these criteria.

We have used two ways to determine the influence of parallel inhomogeneities on lung tissue mechanical properties: first, by analysis of the time changes of parallel inhomogeneities and mechanical parameters in a group of normal rabbits; and second, by comparison of this behavior with that observed in a group of rabbits having inhomogeneous lungs. Exposure to O₃ has been used to induce lung inhomogeneities, because it is well tolerated and the preparation is stable. In agreement with previous pathological studies, mechanical changes in O₃-exposed animals suggest peripheral damage and a significant degree of alveolar inhomogeneity (3).

A significant change in alveolar heterogeneity was observed from the 1st min after saline administration in O₃-exposed but not in air-exposed rabbits. This change was noticed in both CV_i and CV_e and had no effect on lung tissue mechanics. The origin of the increase in alveolar heterogeneity could be related either to an inflammatory increase in vascular permeability, since peripheral inflammatory changes are usually present 18 h after acute O₃ inhalation (24), or to a more rapid dissipation of the effect of volume history (previous inflations to total lung capacity) in the damaged parenchyma. Both CV_i and CV_e returned to basal values before MCh administration. It is generally accepted that the response of the lungs to the administration of a bronchoconstrictive agent is as follows: the smooth muscle in the airways constricts, reducing the airway diameters and increasing Raw. It has been reported in recent studies (15, 18, 23, 29, 30) that bronchoconstrictor agents induce a substantial increase in tissue resistance in several species. Some investigators (13, 17, 22, 28) believe that this tissue response is a secondary response driven by smooth muscle contraction in the lung periphery and a subsequent response manifesting airway-tissue interdependence. Other authors (10, 18, 29, 31) have postulated that smooth muscle in the parenchyma and in blood vessels may also constrict, additionally increasing Rti and tissue elastance (Eti). Constrictive processes occur in a markedly inhomogeneous manner throughout the lungs, often making it difficult to accurately infer their precise nature from individual capsule signals. This has led some authors to believe that there is no substantive tissue response and what actually occurs is that increased inhomogeneities in peripheral airway resistance produce an artifactual increase in Rti due to the nature of the measurement and subsequent analysis (2, 14, 21).

To a large extent, disagreement arises from the way alveolar capsules are used to identify tissue mechanical changes in the presence of substantial heterogeneities. Indeed, if we accept capsule pressure readings without prior conditions, smooth muscle constriction often leads to negative Rti, which is meaningless according to the laws of physics (4). This observation is induced by the violation of the assumptions behind the application of the equation of motion to lung mechanics.

A linear monocompartmental model has been used to calculate mechanical parameters by assuming that the global mechanical behavior of lung can be described by three integrative magnitudes (Fig. 1): EL, Raw, and Rti. This is a usual way of analyzing lung mechanical behavior, even in the presence of limited nonlinearities (4, 14, 17). It could be argued, however, that changes in the nonlinear behavior of lung tissue after MCh would increase Rti as an artifact of mathematical analysis. To test this possibility, we have used an independent approach to calculate Rti independently from linear behavior. According to previous authors (5, 9, 10), the in presence of nonlinearities, it is possible to define lung tissue resistance based on three unambiguous quantities, VT, , and hysteresis area (A): Rti' = (4 · A)/[ ·  · (VT)²]. Rti' is therefore defined as a dimensional parameter proportional to the amount of energy dissipation in the tissue. In Fig. 7, tissue resistance values obtained by both methods are plotted together during MCh challenge. Both measurements of Rti had a similar behavior regardless of the measurement method. The ratio Rti/Rti', an index of the amount of the error related to nonlinearities, did not vary after MCh administration. It can therefore be concluded that nonlinearities do not induce any overestimation of the increase in Rti.

View larger version (41K): [in this window] [in a new window]   Fig. 7.   Time course of Rti calculated by 2 different methods: Rti, from fit of equation of motion to Palv; Rti', from formula proposed by Hildebrandt. Means ± SE are shown. Open circles are average Rti/Rti', taken as an indicator of development of nonlinearities.

Some criticisms of the lung tissue constriction theory are based on the failure of input impedance data to show a substantial increase in extrapolated quasistatic Eti in rat lungs after MCh challenge (21). However, dose-dependent increases in lung static elastance during induced constriction have previously been documented in both dogs (19) and rabbits (27). It may therefore well be possible that, in agreement with the same critical authors (20), our ability to discover tissue changes is inversely related to the amount of alveolar heterogeneity.

Absence of changes in tissue mechanics after saline administration in O₃-exposed rabbits is relevant due to the fact that parallel inhomogeneity is seen to increase significantly. Because the extent of this increase is relatively small, one can argue that the effect of ventilatory heterogeneity on lung tissue parameter estimation is a matter of intensity, as previously suggested by Lauzon et al. (14). If we assume that small changes in heterogeneity have no effect on Rti, instead of the immediate increase observed, we would not expect Rti to increase until ~50-60 s after MCh injection in air-exposed animals. Furthermore, a clear increase in alveolar heterogeneity is only observed after Rti has reached a maximum. According to the general assumption that causes must precede effects, parallel airways inhomogeneity is unlikely to be the cause of the increase of Rti, and even less so the increase of , at the initial phase of the constriction. Therefore the immediate increase in Rti in air-exposed animals should be due to physiological events other than parallel inhomogeneity.

In O₃-exposed rabbits, the increase in total and tissue resistances, as well as ventilatory heterogeneity, was faster than in air-exposed animals. The different time pattern of Rti between O₃- and air-exposed animals could be attributed to an earlier increase of parallel inhomogeneity in the most heterogeneous lungs, or possibly the faster increase in CV_i and CV_e in O₃-exposed rabbits actually reflects the early failure of parenchymal mechanisms preventing alveolar inhomogeneity in damaged lungs.

After these considerations, a question still remains: What is the nature of this early tissue response? In Fig. 6, the dissociation between hysteretic and elastic changes is evident: hysteretic changes predominate during the early transition to the constricted state. A transitional uncoupling between Eti and  was previously observed by Salerno et al. (30), who concluded that tissue changes in which contractile elements are involved result in marked alterations in the coupling of conservative and dissipative processes in the lung, predominantly increasing . In contrast, smooth muscle contractions in lung periphery and the subsequent changes manifesting airway-tissue interaction would predominantly increase Eti, just as increasing lung volume causes a decrease in  despite an increase in Rti (29). In agreement with this statement, Fig. 8 shows the linear relationship between Eti and Raw changes during the transition to the constrictor state in air-exposed rabbits, evidence of the mechanical linking between pulmonary elasticity and bronchoconstriction. Our results agree with a previous study by Mitzner et al. (22), who found that, by perfusing the bronchial vessels with MCh, the increase in Raw was associated with a fall in dynamic compliance.

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Relationship between EL/EL0 and airways resistance (Raw/Raw0) during transition to constricted state in air-exposed rabbits. Values are expressed in relation to basal (prechallenge) values.

Alveolar heterogeneity is unlikely to affect hysteresivity during the transitional period to the constrictor state. Time course of changes in  and in CV_i and CV_e are too different, and  increases too early to invoke a causal effect of parallel inhomogeneities.

In conclusion, in homogeneous lungs, the lag between the increase in tissue resistance and hysteresivity on the one side and the increase in parallel airways inhomogeneity on the other suggests that there is a real, not artifactual, tissue reaction immediately after MCh intravenous infusion. In heterogeneous lungs, MCh induces a faster increase in parallel inhomogeneities, which are probably responsible for earlier mechanical changes. In homogeneous lungs, hysteresivity reflects intrinsic tissue changes, probably related to the activation of the contractile machinery in lung parenchyma, whereas dynamic elastance is related to airways reaction and reflects airways-tissue interdependence.