Main

Increasing evidence suggests that the traditional concept of changes in airway caliber and/or changes in the lung parenchyma as the determinants for alterations in lung function needs to be reconsidered. In recent years, the potential roles of the bronchial and pulmonary circulations in altering lung function and in modifying the response of the lung to agonists have gained increasing attention. Experiments have been carried out both in animals and humans to evaluate the relationship between the airway vasculature and pulmonary mechanics. In dogs, it has been found that both bronchial(1) and pulmonary blood flow(2) affected recovery time constants from i.v. challenge with histamine to a substantial and similar degree. Ishii et al.(3) reported that an increase in left atrial pressure caused increased resistance of both the central and peripheral airways measured by a retrograde catheter technique. Also Hogg et al.(4) observed an increase in peripheral Raw immediately after elevating left atrial pressure.

It has been suggested recently that the resulting changes in lung function are due to reflex bronchoconstriction of the small airways(5) and can at least in part be prevented by vagotomy(3). However, the functional integrity of cholinergic efferent innervation is controversial in the newborn, and there is evidence that efferent vagal control differs between juvenile and adult animals(6).

We have described a technique for partitioning the pulmonary response to different agents into airway and parenchymal components(7, 8). This has been shown to be useful in the evaluation of the sites of action of different agents such as histamine, methacholine, and hypertonic saline in altering lung function.

The present study was undertaken to determine whether vagal influences contribute to changes in lung function due to vascular engorgement in a juvenile animal model. Furthermore, we aimed to determine the pulmonary sites of action of different mechanisms elevating PAP. Two different techniques for producing acute vascular engorgement, namely acute fluid loading and increasing left atrial pressure by a inflatable balloon were used. These two techniques could conceivably have differential effects on the pulmonary and bronchial circulations and thus have differential effects on airway and parenchymal mechanics. Physiologic data were obtained from 11 piglets during vascular engorgement.

METHODS

Animal preparation. Our study was approved by the Institutional Animal Ethics Committee. Eleven piglets (weight mean, 7.1 ± 0.4 kg; age, 3.9 ± 0.2 wk) were studied. Anesthesia was induced with halothane and maintained with fentanyl (1 μg/kg/h i.v.) and pentobarbitone sodium (10 mg/kg i.v.). The piglets underwent a tracheotomy, were intubated, and mechanically ventilated (Harvard Apparatus model 708) with a tidal volume of 10 mL/kg body weight. Paralysis was obtained with pancuronium bromide (0.2 mg/kg i.v.). After insertion of central venous and femoral arterial catheters the chest was widely opened by midline sternotomy. Blood pressure and heart rate were monitored continuously. Once set, tidal volume and ventilator frequency were kept constant throughout the study.

Catheters were inserted into the pulmonary artery and the left atrium of the heart, the former to measure pulmonary artery blood pressure continuously and the latter to elevate the left atrial pressure when inflated. The atrial catheter consisted of a self-made latex balloon, capable of being inflated with up to 10 mL of fluid or air, connected to a modified 20-gauge venous cannula (Insyte, Becton Dickinson, Sandy, UT).

Measuring at the airway opening. Pao was measured with a piezoresistive pressure transducer (Endevco, San Juan Capistrano, CA, 8510 B-2). Flow was measured with a pneumotachograph (Hans Rudolph Inc., Kansas City, MO).

Alveolar capsule technique. The alveolar capsule technique(9) was used to measure PA in the open-chest piglets. Three small plastic capsules were glued to the pleural surface with cyanoacrylate glue (Loctite Corp., Ontario, Canada, 41621). The visceral pleura was punctured four times to a depth of approximately 0.5 mm with a 19-gauge needle to bring the underlying alveoli into communication with the capsule chamber. A piezoresistive pressure transducer (Endevco 8507C-2) was lodged into each capsule to measure PA. The capsules were glued to the right middle and left upper lobe.

Measurement of respiratory mechanics. Mechanical ventilation. Respiratory mechanics were calculated from the Pao,V' (flow), and V (volume) signals using a multilinear regression implementation of the equation of motion of a single compartment model of the respiratory system, as follows: equation 1 where EL represents the elastance of the lung, RL represents the resistance of the lung and PAee represents the Pa at end-expiration(10). We have found that EL frequently varies with volume during mechanical ventilation(11). We therefore allow EL to vary during ventilation by including a volume-dependent term i.e.: equation 2 where E1L represents the volume-independent component of elastance and E2L represents the volume-dependent component.

The tissue contribution to energy dissipation (Vti) during regular mechanical ventilation was calculated using theequation with equation 4 where E1ti is the volume-independent and E2ti is the volume-dependent component of tissue elastance. RL, E1L, E2L, Vti, E1ti, and E2ti were calculated by multiple linear regression fitted to 30-s data epochs containing periods of regular mechanical ventilation. Data were acceptable if the multiple linear regression correlation coefficient was >0.97. Raw was then calculated by subtracting Vti from RL, i.e.equation 5

Signal conditioning. All signals were amplified (PR signal conditioner, Endevco 4423), low pass filtered (Frequency Devices, Haverhill, MA, 902 LPF, 8 pole Bessel, Fc = 10 Hz) and recorded by a 12-bit analog-to-digital converter at a rate of 100 Hz. The data were stored on a computer for analysis.

Experimental protocol. Baseline measurements. After surgical preparation, a stable ventilation pattern was established with a tidal volume of 10 mL/kg. Three discrete 30-s data epochs were then recorded for calculation of baseline RL, Vti, and EL, as described above.

Acute fluid overload. Five piglets received a volume of 10 mL/kg polygeline (Haemaccel®) via the central venous line within 5 min. After the end of the infusion three separate 30-s data epochs were collected. Subsequently, 10 mL/kg blood was removed within 5 min using the femoral artery catheter, and data were collected after this procedure. Next, an amount of 20 mL/kg polygeline was infused and subsequently 10 mL/kg blood removed. This was repeated once. These procedures resulted in the following fluid balance: 0 mL/kg, +10 mL/kg, 0 mL/kg, +20 mL/kg, +10 mL/kg, +30 mL/kg, +20 mL/kg.

Inflation of the left atrial balloon. The left atrial balloon was inflated with isotonic saline in 2-mL steps to establish a dose-response curve. Between the different steps the balloon was deflated and measurements were recollected. This procedure was repeated until a maximal elevation in mean PAP was seen without significant depression of systemic blood pressure. Acute fluid load or inflation of the left atrial balloon were carried out in random order.

Effects of bilateral vagotomy. Six additional piglets were investigated to study the influence of vagotomy on changes in airway and tissue mechanics. Three pigs underwent an acute fluid load with 20 mL/kg polygeline and subsequent removal of the same amount of blood. After bilateral vagotomy, ensuing the baseline measurements, the fluid load was repeated. In another three animals a balloon catheter was inflated and deflated before and after the vagi were cut as above.

Statistical analysis. Multivariate ANOVA with repeated measures using Excel 5.0 was performed to test for a significant change of respiratory mechanics during the three different experimental procedures in comparison with baseline measurements. Two-way ANOVA with repeated measures (each test comparing only two treatment conditions) was used to locate significant differences between treatment conditions. Using a Bonferroni correction for multiple comparisons these tests were performed at the 0.002 and 0.005 level of significance for the fluid load and balloon technique, respectively, to preserve an overall type I error rate of 5%. ANOVA without repeated measures was used for comparison of mean PAP.

For the effects of vagotomy, power tests were performed to determine the sensitivity to detect a type II error. All data are given as mean ± SD.

RESULTS

Acute fluid load. Initial total RL was 1.59 ± 0.29 kPa·L-1 ·s. The changes of RL and concomitant mean PAP during acute fluid administration are shown in Fig. 1. After rapid administration of 10 mL/kg polygeline and subsequent withdrawal of an equal amount of blood, resistance measurements returned almost to baseline in all five animals without statistically significant increases in RL, Raw, or Vti (Figs. 1 and 2). Mean PAP rose from 3.03 ± 0.60 kPa (22.7 ± 4.5 mm Hg) to maximal 5.65 ± 1.12 kPa (42.4 ± 8.4 mm Hg). During administration of 30 mL/kg an attenuation of the pulmonary pressure occurred. Resulting changes in Raw and Vti in comparison with baseline are given in Fig. 2. Raw and Vti increased also significantly from 10 to 20 mL/kg and from 20 to 30 mL/kg (p < 0.002).

Figure 1
figure 1

Effect of acute fluid load on total RL (upper curve) and mean PAP (lower curve). Different volumes of fluid were administered and extracted in a stepwise fashion. The labels on the x axis represent the resulting fluid balance. All values are shown as mean ± SD. The asterisks indicate significant differences of resistance in comparison with baseline (bs).

Full size image
Figure 2
figure 2

Effect of acute fluid load on Raw and Vti.

Full size image

The baseline ratio of Vti/RL was 33.2% and did not change significantly during fluid overload. The relative changes of Vti during fluid administration were 2-3-fold larger than changes in Raw (p = 0.03 for 10 mL/kg and 30 mL/kg,p = 0.08 for 20 mL/kg).

Total EL increased in response to fluid administration by maximal 30.9± 11.4% compared with baseline (p < 0.00001). Increases between 10 and 20 mL/kg and 20 to 30 mL/kg were also statistically significant(p < 0.00002). The magnitude of the response to fluid overload differed between animals but exhibited a similar pattern in each piglet with only very small intraindividual variations. Both the response of mean PAP as well as resistance and elastance parameters was attenuated at higher levels of fluid administration.

Atrial balloon technique. Fig. 3 summarizes the increase of RL and the elevation of mean PAP after inflation of the atrial balloon with increasing volumes. The changes of Vti and Raw with different levels of PAP are shown in Fig. 4. The relative contribution of both compartments to the increase in total RL was not significantly different from each other(p = 0.69 for ΔPAP = 8 mm Hg, p = 0.31 forΔPAP = 17 mm Hg). The baseline ratio of Vti/RL was 27.7% and did not change significantly with different degrees of elevation of left atrial pressure. EL increased by maximal 23.0 ± 16.3% (p< 0.00001) during inflation of the balloon.

Figure 3
figure 3

Effect of atrial balloon inflation on total RL(upper curve) and mean PAP (lower curve). The labels on the x axis represent the volume with which the left atrial balloon was inflated. The asterisks indicate significant differences of resistance in comparison with baseline (bs).

Full size image
Figure 4
figure 4

Effect of atrial inflation on Raw and Vti.

Full size image

The correlation between changes in mean PAP and total RL,Vti, and Raw was very close and similar with both methods used to cause vascular engorgement (r > 0.90, p< 0.002).

During acute fluid load, each increase of the mean PAP by 10 mm Hg was associated with an increase in total RL by 4.7 ± 0.5%. When PAP was increased by elevating left atrial pressure, by means of the balloon, the increase in RL per 10 mm Hg increase in PAP was 3.7 ± 0.8%, respectively.

Effects of vagotomy. Baseline total RL and Vti resistance before and after vagotomy are shown in Table 1. Vagotomy did not change significantly these parameters in any of the studied animals. Furthermore, no significant effect of vagotomy was found in the effects of acute fluid load or elevation of left atrial pressure on pulmonary mechanics (Table 2). Based on these data, the risk of a type II error was calculated. There was a 80% power of detecting at least a 7% change in baseline RL and at least a 12% change in response to vascular engorgement before and after vagotomy.

Table 1 Baseline values for total RL, tissue resistance Vti, and EL before and after vagotomy
Full size table
Table 2 Changes in total RL, Vti,Raw, and EL expressed in percent of baseline with inflation of the left atrial balloon and acute fluid load
Full size table

DISCUSSION

The results of the present study have shown that: 1) vascular engorgement produces largely reversible increases in resistance of both the airway and the parenchymal compartments of the lung with predominance of the tissue resistive properties as well as increases in EL; 2) vagotomy does not influence basal bronchomotor tone nor changes in airway or lung Vti caused by vascular engorgement; and 3) the primary site of action within the lung does not differ whether vascular engorgement is induced by acute fluid loading or increasing left atrial pressure.

Increasing evidence exists that the vasculature within the lung has a larger impact on pulmonary mechanics than assumed previously. However, the underlying mechanisms are still not clearly defined. Several investigators have reported that elevated left atrial pressure in patients with heart disease can produce obstructive symptoms(12, 13), and airway wall edema has been described in congestive heart failure(14). In patients with acute postoperative pulmonary hypertension, respiratory system resistance is increased and compliance decreased(15). Gilbert et al.(16) administered 2 L of isotonic saline i.v. in normal and asthmatic subjects and reported profound decrements in forced expiratory volume in 1 s.

Few studies have been performed to investigate the role of influences of the nervous system on bronchial tone during vascular engorgement. Ishii et al.(3) elevated left atrial pressure in vagotomized dogs to exclude the effects of vagal reflex on the increase in resistance. They observed an almost complete reversibility of the changes in resistance of the central and peripheral airways within the first 30 min after elevation of left atrial pressure, whereas after this period increases in peripheral Raw were not completely reversible. This supports the assumption that the initial changes in pulmonary mechanics are not influenced by vagal responses. Other authors, however, have found, that vagal stimulation has an effect on central airways and an even more pronounced effect on peripheral airways(17, 18). Wagner and Mitzner(5) recently found that vascular engorgement decreased bronchial luminal area only in the presence of airway smooth muscle tone, suggesting a reflex bronchoconstriction in response to engorgement. However, most of these studies were performed in adult animals. It has been shown that cholinergic efferent innervation develops after birth and, unlike adult pigs, vagotomy in newborn piglets does not decrease repiratory system compliance and resistance(6). Pérez Fontán demonstrated that vagal bronchomotor tone is absent in developing piglets at least to the age of 6 wk(19). The data of our present study support and extend these findings by demonstrating that vascular engorgement in newborn piglets does not act on lung function via vagally mediated bronchoconstriction.

In our experiment, the animals were anesthetized initially with halothane, and anesthesia was maintained with pentobarbitone and fentanyl. Pancuronium bromide was used for muscle relaxation. These agents are commonly used in animal studies of respiratory function but may themselves change pulmonary mechanics(20–22). Pancuronium is known to enhance the increase in pulmonary resistance due to vagal nerve stimulation, probably via block of prejunctional muscarinic receptors that physiologically inhibit vagally mediated increases in pulmonary resistance(22). We cannot completely rule out that effects of the above agents may have influenced the results of our study, especially in enhancing the changes in pulmonary mechanics. However, the response of pulmonary resistance to vascular engorgement was not altered by vagotomy, which suggests that vagal stimulation is not responsible for the effects of vascular engorgement, and pancuronium should be expected to enhance any differences if present.

In our fluid experiment we used polygeline rather than saline because we wanted to decrease the possibility of extravasation of fluid into the interstitium and alveoli. For hemodynamic reasons we could not extract the same amount of fluid as administered within a short period of time. However, we found that each single step of fluid administration was largely reversible, indicating that interstitial edema was unlikely to be responsible for the alterations in resistance and elastance. In addition, one may have expected changes due to edema to be progressive and increasing throughout the study. This thesis is supported by the fact that both the changes in elastance and resistance induced by elevation of left atrial pressure with the inflation of the balloon were completely reversible. These findings are compatible with observations of Wetzel et al.(23). The authors investigated the change in airway cross-sectional area in piglets after intravascular volume expansion with 30 mL/kg Ringer's lactate by high resolution computed tomography and found a decreased internal cross-sectional area in both large (2-5 mm) and small (0.75-2 mm) airways. These changes were rapidly reversed by intravascular volume reduction suggesting a reversible bronchial mucosal vascular engorgement rather than edema as the potential cause of increased Raw.

On a morphologic level, the bronchial circulation would be expected to be the principal site of vasodilation when hyperemia occurs in the conducting peripheral airways, whereas the pulmonary circulation would be expected to be involved in hyperemia of the alveolar tissue. An important aim in our study was to partition RL into its airway and tissue components using the alveolar capsule technique(9). The relative alteration in Vti exceeded Raw 2-3-fold with both methods used to cause vascular engorgement, but the relationship between Vti/RL did not change throughout the experiments. This indicates that both tissue and airway compartments (and thus, bronchial and pulmonary circulations) are affected. This is consistent with work of Kelly et al.(1, 2) who concluded that both vascular beds affect the recovery time from i.v. histamine challenge in the lung periphery.

There is still controversy in the literature regarding whether pulmonary artery blood flow or mean PAP is the major determining factor for changes in respiratory mechanics. Some authors observed an effect on pulmonary mechanics only when both PAP and pulmonary artery blood flow were elevated(24). In animal models some investigators even failed to detect any changes in respiratory function after experimental changes in PAP(25, 26). In children undergoing cardiopulmonary bypass we have previously reported that respiratory mechanics were correlated with pulmonary blood flow but not PAP(27). Griffen et al.(28), however, reported a decrease in dynamic compliance with increasing PAP alone. Bancalari et al.(29), studying infants with increased and decreased pulmonary artery blood flow, suggested that PAP rather than flow is the primary factor affecting compliance in humans with left-to-right shunts. The results of the present study are consistent with this conclusion. The two different techniques we used to cause vascular engorgement have opposing effects on the pulmonary blood flow. Elevation of left atrial pressure diminishes the blood flow, whereas acute fluid administration increases it. Despite different blood flows we have found that the responses of both the airway and the tissue compartments to vascular engorgement were closely correlated to mean PAP.

We observed an attenuation of the response of both PAP and pulmonary mechanics to increasing fluid volumes. The former is very likely due to the autoregulatory response of the heart, the Frank-Starling mechanism. With increasing venous return and elevated end-diastolic volume the cardiac output will also increase, thus compensating a further elevation of pulmonary venous and PAP. If we assume the PAP to be the driving force for changes in lung mechanics, the attenuation of RL and elastance in response to larger fluid volumes may be attributed to the above mechanism.

In summary, we found that in 4-wk-old piglets cholinergic innervation does not play a role in changes in airway and lung tissue mechanics caused by vascular engorgement. Both techniques we used (acute fluid administration and increasing left atrial pressure by a balloon catheter) cause similar increases in both Raw and Vti with and without vagi intact. The alterations in pulmonary mechanics were closely associated with changes in pulmonary artery blood pressure. This model provides the opportunity to gain further valuable insight into the role of the vasculature altering lung mechanics.