Main

Despite its potential clinical use as a selective pulmonary vasodilator in the treatment of persistent pulmonary hypertension of the newborn, little is known about the changes inhaled NO exerts on Raw and Rti in the newborn lung(1–3). In addition to lowering pulmonary arterial resistance, previous studies report that inhaled NO produces bronchodilatation, reduces histamine- or methacholine-induced bronchoconstriction, and improves ventilation/perfusion matching in adult humans and experimental animals(4–10). There are, however, no reports describing the effect of NO on newborn lung mechanics, and the site of action for any decrease in Rl induced by exogenous NO has yet to be identified.

Endogenous NO is produced in cells by the enzyme NOS, which is found within lung epithelium, airway and vascular smooth muscle, neural tissue, and the endothelium of large pulmonary vessels(11). Endogenous NO may contribute to pulmonary vascular dilatation and airway smooth muscle relaxation and has been also proposed as the mediator for iNANC bronchodilation in adult humans(12–14). Maturational studies suggest that NOS expression and activity are greatest in fetal life and around term gestation, and subsequently decrease with age in endothelial and airway epithelial tissues(15–19). Parenchymal viscoelastic properties and airway caliber are both known to contribute to pulmonary resistance during early postnatal life(20, 21). Viscous resistance is the frictional resistance within lung tissue that is being inflated or deflated. In this study, we used alveolar capsules situated over the lung parenchyma for measuring flow-resistive pressure changes across tissue expressed as lung tissue resistance. This allowed us to partition Rl into its airway and tissue components to characterize the sites of changes in resistance induced by exogenous and endogenous NO. We hypothesized that both exogenous and endogenous NO would modulate baseline pulmonary resistance by affecting airway as well as lung tissue resistance in the newborn piglet.

METHODS

Experimental preparation. Experiments were performed in 13 piglets at 2-4 d of life weighing 2.4 ±.3 kg (mean ± SD). The piglets were initially sedated with intramuscular ketamine hydrochloride (14 mg/kg) and xylazine (2.8 mg/kg) and were anesthetized with i.v.α-chloralose (24 mg/kg) and urethane (120 mg/kg), with additional doses of maintenance anesthesia at 10% of loading dose given approximately every 45 min, depending on acute heart rate and blood pressure changes. A femoral artery was cannulated for measurement of systemic blood pressure and blood gas sampling, and an external jugular vein was cannulated for administration of further anesthesia, NOS blocker, and fluids. A catheter was placed in the pulmonary artery to measure pulmonary arterial pressure.

The piglets were placed on a heating pad to maintain body temperature between 37.5 and 38.5 °C. After a high cervical tracheostomy, the piglets were artificially ventilated with 100% O2 through a tightly fitting tracheal cannula with a side tap connected to a volume ventilator (Harvard model 55-0798) that delivered a tidal volume of 10-12 mL/kg. The ventilator rate was set at approximately 25 cycles/min to maintain an arterial Pco2 of 30-35 mm Hg and an arterial Po2 of >300 mm Hg during baseline conditions. Blood gas tensions and pH were determined with a Radiometer automated blood gas analyzer (ABL3, Copenhagen, Denmark).

Lung mechanics were measured as described previously by ourselves and others(20–23). Briefly, a midline sternotomy was performed, and the chest was widely retracted. Light-weight round-base capsules (10 mm diam) were applied to the pleural diaphragmatic surface of both lungs by means of cyanoacrylate glue. The pleura under each capsule was punctured to a depth of 1-2 mm four to five times with a 20-gauge needle to bring the underlying alveoli into communication with the capsule chamber. Pressures in the capsules and in the tracheal side tap were measured with miniature piezoresistive pressure transducers (model 8510 B-2, Endevco, San Juan Capistrano, CA). Dynamic matching of the transducers was tested, and there was no phase or amplitude distortion of signals in the range of applied pressure and frequency.

We considered that capsules were successfully installed if during slow tidal mechanical ventilation the changes in tracheal and alveolar pressures were eqal at end inspiration and end expiration when there was no flow (Fig. 1). A significant concern of the capsule technique is that pressure within the capsule may not dynamically equal the alveolar pressure beneath the capsule. This discrepancy between alveolar and capsule pressures depends on the time constant of the capsule system, defined by the combined compliance of the transducer and air in the capsule and the resistance to air flow through the pleural openings. For frequencies less than 1 Hz, as used in this study, the capsule pressure should accurately estimate regional alveolar pressure(24). Resistance calculated from the two capsules exhibited minimal differences and was averaged for each experiment. Based on our previous experience, Rti and Raw had comparable results whether derived from an individual capsule or averaged from two capsules(25). The animals were all mechanically ventilated at a positive end-expiratory pressure of 3 cm H2O to prevent closure of the most distal airways. In addition, the lungs were inflated every 10 min by occluding the expiratory line of the ventilator for two to three consecutive volume cycles. Each hyperinflation was performed at an identical time before any experimental maneuver (e.g. NO administration) to avoid confounding effects of different volume histories. The expiratory line was reopened when transpulmonary pressure reached approximately 25 cm H2O.

Figure 1
figure 1

Oscilloscope tracing of simultaneous readings of tracheal (top) and alveolar pressure (bottom) against volume. Direction of the curve is clockwise.

Full size image

Tracheal flow signals were obtained with a Fleisch pneumotachograph and were electrically integrated to derive volume. Pressure and flow signals were filtered electrically with matched 15-Hz low pass filters and were recorded on a Gould (Cleveland, OH) six-channel recorder along with pulmonary and systemic arterial blood pressure. The coefficient of variation for 15 control breaths was minimal (0.04 for both Rl and Rti, and 0.006 for Cdyn) during baseline conditions. To rapidly detect possible capsule malfunction, alveolar pressures were displayed against tidal volume on a Tektronix storage oscilloscope. All data were recorded on an FM magnetic tape for later playback and analysis.

A custom-made computer program was used to calculate resistance on the basis of the method of Von Neergaard and Wirz(26). Total Rl and Rti were calculated from tracheal flow and tracheal and alveolar pressure changes, respectively. Raw was derived as the difference between Rl and Rti (Raw = Rl - Rti). Cdyn was calculated as a ratio between volume and pressure measured at tracheal opening. All studies were approved by the Case Western Reserve University Institutional Review Board.

Experimental protocol. In all 13 piglets, baseline measurements of lung mechanics were made over the last minute of a 5-min period of quiet ventilation during inhalation of 100% O2. NO inhalation was used in 12 animals. In 10 piglets, inhaled NO was sequentially administered at 20, 40, and 80 ppm to the inspired gas consisting of 100% O2. In two additional piglets, the three NO dosages were administered in a random sequence. Intervals of 5 min were used between treatments, and measurements of Rl, Raw, Rti, and Cdyn were repeated over 60-s periods after 5 min of exposure to each concentration of NO. The NO dosages used are the same as intially used in the Neonatal Inhaled Nitric Oxide Study conducted by the NICHD Neonatal Research Network. NO concentration was measured by continuously sampling the inspired gas at the airway opening by means of an NO analyzer (Ecophysics; Switzerland), and NO2 concentrations were measured in two piglets using the same analyzer. NO2 concentrations of 0.3-2.5, 8-9, and 10-15 ppm were recorded during delivery of 20, 40, and 80 ppm NO, respectively. After discontinuing NO administration, the piglets were ventilated with 100% O2, and flow and pressure measurements were again recorded to ensure that baseline conditions were restored. The NOS blocker L-NAME at 30 mg/kg was given i.v. to 12 piglets, and Rl, Rti, Raw, and Cdyn were again measured after 5 min.

To determine whether changes induced by blockade of NOS could be reversed by exogenous NO, inhaled NO at 20 and 80 ppm was then readministered to eight piglets, and the same measurements were made. Throughout the experiment, pulmonary and systemic arterial pressures were recorded. Arterial blood gas tensions were checked periodically and did not change significantly during any experiment, with arterial Po2 consistently in the hyperoxic range.

Statistics. Statistical analyses of changes in pulmonary mechanics in response to NO inhalation used one factor repeated measure analysis of variance. Comparisons of measured parameters before and after L-NAME used two-tailed paired t test. Data are expressed as mean± SD. Differences were considered significant at a p < 0.05.

RESULTS

Addition of NO to the inspired gas decreased both the airway and tissue components of Rl. Inhalation of NO at 20, 40, and 80 ppm caused a significant decrease in Rl (p < 0.001), Raw(p < 0.05), and Rti (p < 0.001) in comparison to initial control measurements (Fig. 2). At each concentration of NO, resistance values were significantly lower than during the initial control period. These results did not differ in the subset of animals who received randomized dosages. Increasing NO concentration from 20 to 80 ppm was not followed by further decreases in Rl, Raw, or Rti. Control values before and after termination of NO inhalation also did not differ. Cdyn tended to increase at all dosages, but only at 40 ppm was the increase significant (p < 0.05). Systemic arterial pressure did not change significantly in response to NO inhalation and fluctuated between 60 and 85 mm Hg. Pulmonary arterial pressure, however, demonstrated a significant reduction in response to NO, falling from 17.4± 4.1 to 14.6 ± 3.0 mm Hg after NO was added (p < 0.001). An example of the responses of tracheal and alveolar pressures to NO inhalation is shown in Figure 3.

Figure 2
figure 2

Responses of Rl (circles), Raw(triangles), and Rti (boxes) during inhalation of NO at concentrations of 20, 40, and 80 ppm. Control data are shown before NO inhalation. Data are presented as mean ± SD.

Full size image
Figure 3
figure 3

An example of the responses of tracheal pressure(Ptr), alveolar pressures (Palv), and air flow (˙V) to NO administration at 20 and 80 ppm before and after L-NAME administration. The slight decrease in ˙V as measured by the tracheal pneumotachometer during constant tidal volume ventilation at 80 ppm(but not 20) is probably due to the change in gas density as NO (mixed with N2) is added to the inspired gas.

Full size image

Blockade of NOS increased both the total and tissue component of Rl. After L-NAME administration, Rti and Rl increased significantly(p = 0.005 and 0.02, respectively), whereas Raw, derived from Rl - Rti, demonstrated a variable response when compared with control values before L-NAME (Table 1). Dynamic lung compliance decreased significantly (p < 0.0001) after L-NAME administration, falling 13% from baseline. L-NAME administration increased both systemic arterial pressure from 66.4 ± 8.1 to 98.9 ± 10.2 mm Hg (p < 0.001) and pulmonary artery pressure from 17.3± 4.4 to 26.3 ± 9.1 mm Hg (p < 0.001).

Table 1 Measurements of R1, Rraw, and Rti as well as Cdynbefore and after L-NAME administration
Full size table

Readministration of NO at 20 and 80 ppm to eight piglets after L-NAME caused Rl to fall from 21.6 ± 5.8 to 16.5 ± 4.4(p < 0.001) and 14.9 ± 3.5 (p < 0.001) cm H2O/L/s, and Rti to fall from 13.6 ± 5.3 to 11.3 ± 5.5 (p < 0.01) and 9.6 ± 4.4 cm H2O/L/s(p < 0.05) at 20 and 80 ppm, respectively. Raw changed from 9.3 ± 4.3 to 6.8 ± 3.3 cm H2O/L/s (NS) at 20 ppm of NO and significantly fell to 6.2 ± 2.6 cm H2O/L/s (p< 0.05) at 80 ppm of NO. Cdyn was significantly increased from 5.5± 0.9 to 6.2 ± 1.2 mL/cm H2O at 20 ppm (p < 0.01) and 6.3 ± 1.2 mL/cm H2O at 80 ppm (p < 0.001). The responses of tracheal and alveolar pressures to administration of L-NAME followed by inhaled NO in one animal are demonstrated (Fig. 3).

DISCUSSION

Despite the explosion of information concerning the physiologic function and therapeutic uses of NO, to our knowledge there are no published reports dealing with the effects of NO on pulmonary function in the newborn. In this study we have demonstrated that inhaled NO reduces Rl in the newborn by affecting both Raw and Rti. In contrast, inhibition of endogenous NOS caused a significant increase in total Rl by increasing Rti. No consistent effect on Raw was observed.

Previous studies in adults report a bronchodilatory role for inhaled NO. After NO inhalation, Dupuy et al.(5) demonstrated a reduction in Rl in adult guinea pigs whereas Hogman et al.(4) and Sanna et al.(27) reported a small increase in specific airway conductance (reciprocal of resistance) in adult human volunteers with asthma and methacholine-induced bronchospasm. The physiologic effects of inhaled NO on pulmonary resistance in normal adult humans and pediatric asthmatic patients appear to be minimal(27, 28). Based on their finding that NO inhalation did not change resistance in adult rabbits after methacholine nebulization, Hogman et al.(29) suggested that the action of NO was predominantly in the central airways. However, Gwyn et al.(7) recently documented that inhaled NO acted as a bronchodilator of peripheral airways in mature dogs. In our study, we have demonstrated for the first time a decrease in both Raw and Rti using capsules placed on the lungs of open-chested newborn piglets during NO inhalation. Our protocol did not allow us to demonstrate the minimal dose of NO needed to decrease Rl. Whether the mechanisms responsible for bronchodilation are the same as the mechanisms responsible for lowering pulmonary arterial pressure remains to be determined. The presence of NO2 (an upper respiratory tract irritant and bronchoconstrictor) in the inhaled gas may have lessened the magnitude of NO-induced bronchodilation in the lung, and may have attenuated the ability of NO to lower Rl and/or increase Cdyn significantly at high doses of NO(30). Furthermore, low resting bronchomotor tone, partly due to mechanical ventilation and low Paco2, may have limited the ability of NO to induce bronchodilatation(21, 31).

If inhaled NO is able to act as a bronchodilator as well as a vasodilator, this may have important implications for matching ventilation and perfusion throughout the lung. Putensen et al.(8) described improved ventilation/perfusion distribution with NO inhalation in pigs during methacholine-induced bronchoconstriction. A recent report describing improved oxygenation after NO inhalation in newborn piglets with meconium aspiration suggested that selective bronchodilation and the resultant vasodilation in ventilated alveoli improved ventilation/perfusion ratios(32). Further studies are, however, clearly needed to assess the role of exogenous NO in modulating ventilation/perfusion matching under various pathophysiologic conditions.

Our experiment also demonstrates that endogenously produced NO is a determinant of baseline Rl. Because L-NAME infusion increased Rl by increasing Rti, the effect of endogenous NO is physiologically limited to the distal contractile elements of the newborn piglet and actively reduces baseline resistance at this level but not at the larger airways. We speculate that NOS blockade, by decreasing endogenous NO production in the newborn lung, increases Rti by promoting constriction of distal smooth muscle and/or interstitial contractile elements. The fall in dynamic lung compliance after L-NAME infusion is consistent with previous data that demonstrate a simultaneous increase in Rti(20, 25).

Using adult subjects, several studies investigating the role of endogenous NO on pulmonary function have demonstrated conflicting findings. In contrast to our results, Kips et al.(33) infused L-NAME into adult rats and demonstrated no effect on basal airway tone or basal airway responsiveness. It must be noted, however, that the varying forms of NOS are expressed within the lung at different levels throughout maturation. Levels of NOS have been shown to increase during gestation and subsequently decrease with age(15–19). Our results suggest that the significant increase in Rti after NOS blockade is due to differences in expression of NOS in the newbron compared with the adult. The absence of a significant effect of endogenously released NO on baseline Raw may be explained by a low airway smooth muscle tone in the newborn piglet(34). In this study we did not have an independent assessment of the contractile state of the airways under control conditions. The modest increase in Rti observed after NOS blockade is probably not of great physiologic significance, however, under pathologic conditions associated with increased airway tone, an opposing airway relaxation induced by release of endogenous NO may assume much greater prominence. Recently we have shown that NOS blockade significantly potentiates airway smooth muscle responses to bronchoconstrictive agents in early postnatal life(35). The ability of inhaled NO at 20 and 80 ppm to reverse the effects of NOS blockade on Rti further supports a role for endogenous NO in modulating baseline pulmonary function.

Endogenous NO production and its contribution to iNANC bronchodilation have been demonstrated in adult subjects, but its presence and activity in newborns has been uncertain. Studies in rabbits suggest that the iNANC system in rabbits is not functional at birth and undergoes considerable postnatal development, whereas the iNANC system does exist in pigs and newborn cats, suggesting significant interspecies differences(36–38). Our findings that NOS inhibition produced an elevation of Rti in newborn pigs suggest that baseline iNANC activity is measurable in more distal airway and lung parenchyma in that species. Recently we demonstrated that piglet lung parenchymal tissue is modulated by central cholinergic output. Hence an opposing relaxant system may counteract constrictor influences at this location(21). In adults, Ward et al.(39) showed that NO-mediated iNANC bronchodilation is present throughout the tracheobronchial tree, but with decreased response in the distal airways secondary to decreased NOS-containing nerves. Future experiments will need to correlate location of physiologic responses with localization of NOS-containing structures throughout the developing lung.

The use of alveolar capsules allowed the partitioning of Rl into tissue and airway components. We believe that the current study accurately demonstrated airway and lung partitioning, because Rti and Raw contributed more equally to Rl under baseline conditions as we previously described for the newborn piglet(25). The actual changes observed at the tissue level may be due to diverse distal contractile elements, including bronchiolar smooth muscle, interstitial cells, and the distal pulmonary vasculature(21, 22). We cannot identify which contractile elements are responsible for the change induced by NO and changes in the state of pulmonary vascular constriction could possibly account for the changes in tissue viscance induced by NO or NOS blockade. As L-NAME increased Rti without a significant increase in Raw, tissue distortion secondary to changes in resistance of larger airways probably cannot explain our findings. It is possible that ventilation at relatively high tidal volumes may have limited bronchoconstriction induced by L-NAME, by inhibiting airway smooth muscle contraction as has been described in immature rabbits when alveolar capsules were used(22). High concentrations of bronchoconstrictive agents may induce the development of inhomogenous ventilation, leading to inaccurate assessment of lung mechanics by alveolar capsules(23). By maintaining constant ventilator settings and observing identical responses from two capsules placed on different lung lobes, we have minimized the likelihood of such inhomogeneity. Furthermore, the mild constrictive effects induced by L-NAME would be unlikely to induce detectable inhomogeneity. Throughout our experiment, responses of systemic and pulmonary artery pressures to NO were consistent with previous findings and arterial blood gas tensions did not physiologically vary significantly during the experiment. Inhaled NO produced a selective decrease in pulmonary artery pressure, whereas L-NAME increased both pulmonary and systemic artery pressure(1, 2, 40, 41). We therefore cannot exclude NO mediated pulmonary arteriolar and venular vasodilation with a resultant increase in pulmonary blood flow (or vasoconstriction after NOS blockade) as possible contributors to our results. There is the additional possibility that NO-mediated mechanisms may alter the properties of surface-active molecules, resulting in changes in Rti(42). Future in vitro studies using distal airway (or vascular) segments or lung parenchymal strips might enable us to identify whether airway smooth muscle contraction is the primary mechanism underlying the observed responses of Rti.

Our study demonstrates a contribution of NO to airway smooth muscle relaxation and change in tissue viscance in the newborn lung with resultant implications for pulmonary function. Inhaled NO in the newborn piglet caused a significant decrease in resistance throughout the lung. Blockade of endogenous NO produced a significant increase in only Rti, implicating a relaxant effect of endogenous NO on lung parenchyma rather than larger airways under baseline conditions. Greater understanding of the location of relaxant responses induced by No may clarify our understanding of ventilation/perfusion balance in both health and disease.