Sympathetic blockade blunts hypercapnic pulmonary arterial vasocontriction in newborn piglets

Abstract

Objective: Hypercapnia has been implicated in the pathophysiology of pulmonary hypertensive disease in newborns. However, little has been done to determine how its vasoconstrictive actions are mediated. The purpose of this study is to define the role of the sympathetic nervous system in modulating the response of the newborn pulmonary circulation to hypercapnia. Specifically, we studied the effect of sympathetic blockade on mean and pulsatile pulmonary arterial hemodynamics in 48-h-old, intact, open-chest Yorkshire piglets during hypercapnic ventilation. Methods: All animals were anesthetized and then instrumented for high fidelity measurement of pulmonary artery pressure (PAP), flow (PAF), aortic pressure and radius of the main pulmonary artery (R_mn). Baseline data were acquired in all animals. Control animals (n=7) were subjected to 30 s intervals of hypercapnia (inspired CO₂ fraction (FiCO₂)=0.20). Experimental animals (n=7) were pre-treated with an intravenous bolus of the adrenergic blocking agent guanethidine (20 mg/kg) before being subjected to the hypercapnic stress. Characteristic impedance (Z_o) and input mean impedance (Z_m) were determined through application of a Fourier analysis of the PAP and PAF waveforms. The modulus of elasticity (E_y) was calculated from Z_o and R_mn. Pulmonary vascular resistance (PVR) was calculated as (PAP−LAP/PAF). Results: Control animals underwent significant increases in PVR (4860±341 dyne cm s⁻⁵ versus 8090±387 dyne cm s⁻⁵, P≪0.01) and Z_m (7215±495 dyne cm s⁻⁵ versus 10 228±993 dyne cm s⁻⁵, P≪0.01) when exposed to hypercapnia. Pre-treatment with guanethidine attenuated this response (PVR, 5552±368 dyne cm s⁻⁵ versus 7105±611 dyne cm s⁻⁵, P=0.31 and Z_m, 7922±446 dyne cm s⁻⁵ versus 9745±600 dyne cm s⁻⁵, P=0.31). Characteristic impedance, modulus of elasticity and the radius of the main pulmonary artery were unchanged in both groups. Conclusions: These data indicate that vasoconstriction secondary to hypercapnia in the neonatal pulmonary arterial circulation occurs at the level of the distal arteriolar bed, rather than the more proximal pulmonary arteries. In addition, this response is partially modulated by the sympathetic nervous system and may therefore respond clinically to manipulation of sympathetic input to the pulmonary arterial circulation.

Pulmonary hypertension, Sympathetic nervous system, Hypercapnia, Pulmonary impedance

Introduction

Persistent pulmonary hypertension of the newborn and pulmonary hypertensive crisis following cardiac surgery remain daunting clinical challenges. Hypoxia and hypercapnia are known to initiate and potentiate the powerful vasomotor alterations seen in these disease processes [1],[2],[3],[4],[5],[6],[7]. Considerable debate exists as to the role of the sympathetic innervation in these vasoconstrictive responses. Some investigators have proposed the effects of hypoxia to be local; independent of adrenergic regulation [8],[9]. Others have proposed a contributory role for the abundant sympathetic innervation demonstrated in histopathologic studies [10]. Similar attention has not been paid to the mechanisms responsible for hypercapnic vasoconstriction. Data from our laboratory have shown that the effects of hypercapnia may be more potent than hypoxia in the newborn pulmonary circulation [1]. These studies also show that the vasoconstrictive effects of hypercapnia may persist after a return to alveolar normocapnia.

Most studies in the literature focus on the more conventional measurements of pressure and flow, expressing them in mean terms. This analysis ignores the contribution of oscillatory energy contained within the waveforms. As much as 50% of right heart work may be contained within these pulsatile components [11]. Therefore, measurement of pulmonary vascular resistance is an informative, but incomplete examination of the changes occurring in the pulmonary arterial circuit. Assessment of the pulmonary hydraulic impedance terms allows for a more comprehensive quantification of right heart work and regional responses within the pulmonary arterial circulation [11],[12]. Regional responses are determined by subjecting the pulmonary arterial pressure and flow waveforms to a Fourier analysis and calculating input mean impedance (Z_m), which represents the distal arteriolar region and the resistance vessels and characteristic impedance (Z_o) which is dependent upon the geometry and distensibility of the larger, more proximal pulmonary arteries. Dynamic measurement of the radius of the main pulmonary artery throughout the cardiac cycle allows the modulus of elasticity for the vessel to be calculated. Alterations in the proximal vessels, as evidenced by changes in Z_o, can then be classified as primary alterations in the elasticity or compliance of the vessel wall or simply passive alterations in the radius of the vessel.

The purpose of this study was to determine the role of the sympathetic nervous system on the mean and pulsatile pulmonary hemodynamic alterations observed with hypercapnia in the newborn piglet. We hypothesize that the pulmonary hypertension observed with exposure to high inhaled CO₂ concentrations is at least in part mediated by the sympathetic innervation of the pulmonary arterial system. If this hypothesis is correct, a chemical sympathectomy should blunt the increased impedance response seen with hypercapnia.

Materials and methods

All animals received humane care in compliance with the European Convention on Animal Care and `Principles of Laboratory Care' formulated by the National Society for Medical Research and the `Guide for the Care and Use of Laboratory Animals' prepared by the National Institutes of Health (NIH publication No. 85–23, revised 1985) and with the approval of the Georgetown Animal Care and Use Committee (GUACUC).

Surgical preparation

A total of 14, 48-h-old (±4 h, mean weight 1.6–2.4 kg) Yorkshire pigs of either sex were utilized in this study. The ear vein was catheterized and animals were anesthetized with an intravenous administration of thiopental sodium (25 mg/kg). A half dose of this anesthetic agent was administered every twenty minutes to maintain an adequate level of anesthesia.

Following endotracheal intubation the animals were placed supine and mechanically ventilated with a pediatric positive pressure ventilator (Health dyne 105, Marietta, GA). An F_iO₂ of 1.0 was used and an arterial oxygen saturation of 93% or greater was maintained throughout all experiments. Positive inspiratory pressure was pre-set between 20–30 cm H₂O and respiratory rate between 9–10 ventilations/min. These settings achieved a PaCO₂ between 25–30 mmHg. A positive end-expiratory pressure of 3 cm H₂O was maintained to prevent atelectasis. To avoid the effects of respiratory motion on pulmonary artery pressure and flow, ventilation was briefly interrupted during data collection intervals without observable changes in pulmonary or systemic parameters [13].

Pancuronium bromide (0.1 mg/kg IV every hour) was administered to produce complete muscle relaxation. A median sternotomy was performed. Sonomicrometry crystals for measurement of instantaneous diameter changes were placed on the lateral aspects of the main pulmonary artery with 4-0 silk sutures. A 6 or 8 mm ultrasonic flow probe (type 6S or 8S) (Transonic Systems, Ithaca, NY), chosen to avoid constriction of the vessel, was placed around the most proximal portion of the pulmonary artery. A medium sized titanium clip (Ethicon, Rochester, NY) was used to occlude the ductus arteriosus. High fidelity pressure transducers (model MPC-500, Millar, Houston, TX) were placed into the left atrial appendage, main pulmonary artery, right internal carotid artery. Each transducer was secured by a 4-0 silk purse string suture. Pre-measurement of the pulmonary artery catheter length in relation to the main pulmonary artery ensured that the transducer tip was positioned 2 mm beyond the ultrasonic flow probe to avoid any perturbation of waveforms.

Experimental protocol

Guanethidine (Sigma, St. Louis, MO) was dissolved in sterile 0.9% NaCl for intravenous infusion. Pigs of either sex (48-h-old) (n=14) were randomly assigned to two study groups. After obtaining baseline hemodynamic measurements, control animals (n=7) were subjected to 30 s intervals of hypercapnia (inspired CO₂ fraction (FiCO₂)=0.20) via the ventilator. The second group of animals (n=7) were pre-treated with an intravenous bolus of the adrenergic blocking agent, guanethidine (20 mg/kg), via a right internal jugular vein catheter before being subjected to the hypercapnic stress. Ablation of the rise in systemic pressure after temporary bilateral carotid occlusion (loss of the carotid baroreceptor response) was taken as evidence of complete chemical sympathectomy. Arterial blood oxygen tension (PaO₂), partial pressure of CO₂ (PaCO₂) and pH were determined with a CIBA Corning analyzer (Model 278, Medfield, MA). Alveolar gas concentration (O₂ and CO₂ tension) and arterial oxygen saturation were constantly monitored with a pulse oximeter (POET II, Criticare systems, Waukesha, WI).

Data collection and analysis

The signal generation of high fidelity pressure waveforms was electronically processed through a calibration control unit (Millar, Houston, TX) and a Gould transducer (Model 13, Cleveland, OH). The analog signals were amplified, transmitted and temporarily stored in an IBM 486 PS model 30 computer for analog to digital conversion at a sampling rate of 200 hz/s. Ultrasonic flow probe signals were processed and amplified through a Transonic flowmeter system (Model T-201, Ithaca, NY) and transmitted to the same computer for analog to digital conversion. Simultaneous analog pressure and flow waveforms were displayed on an oscilloscope (Tektronic, Model 5B10 Beaverton, OR) for signal verification during experimental manipulation. Data were subsequently downloaded onto a Dell 25 MHz 486x PC with a math co-processor for Fourier analysis and hemodynamic computations. Between 8–16 heart beats were analyzed per data collection period per animal resulting in 112–224 heart beats which underwent analysis per group. All animals were in normal sinus rhythm during the entirety of the experiment, including data collection periods.

Statistical analysis

All statistical analyses were performed using a computer-based software package, Instat (GWU, Washington, DC). A paired two-tailed non-parametric test (Wilcoxon signed rank) was employed to determine statistical significance within groups. An unpaired two-tailed Mann-Whitney U was used to determine significance between groups. A P value equal to or less than 0.05 was considered statistically significant.

Hemodynamic calculations

Pulmonary vascular resistance, input mean impedance and characteristic impedance

Pulmonary vascular resistance was calculated in the usual fashion:  

where P_PA is the mean pulmonary artery pressure, P_LA is the mean left atrial pressure, Q_PA is the mean pulmonary artery flow.

Pulmonary arterial impedance calculations were based on a Fourier analysis of pressure and flow waves as previously described [12]. Data collection periods were 30 s long and 8–16 random heart beats were analyzed for each data interval for each pig. Ten harmonics were calculated for each heart beat. Total pulmonary flow is expressed as  

where Q_m is the mean flow, Q_n is the amplitude of the nth harmonic, ω is the fundamental angular frequency (2πf where f is the frequency in Hz), I is the length of the sequence and Θn is the phase angle of the nth harmonic. Pressure waveforms are expressed as  

where P_m is the mean pressure, P_n is the amplitude of the nth harmonic and β_n is the phase angle of the nth harmonic. Dividing mean pressure by mean flow produces the input impedance to (Z_m) at the zeroth harmonic. Similarly, the division of each of the sinusoidal terms gives the input impedance for the nth harmonic. The corresponding phase angle (Φ_n) was calculated from subtraction of the flow phase angle from the pressure phase angle. Characteristic impedance (Z_o) is defined as the impedance in the absence of wave reflections and was calculated between 3–10 Hz.

E_y calculations

Wave velocity (C_o) was calculated for the main pulmonary artery assuming the relationship Womersley derived between characteristic impedance, wave velocity and radius (R) of a strongly tethered elastic tube [14] 

where ρ=1.055 g/ml, which is the density of blood; σ=0.5, which is Poisson's ratio and j=√−1. M'₁₀ and ε are functions of Womersley's non-dimensional parameter α.  

where μ=0.04 poise which is fluid viscosity.Using the calculated Womersley's wave velocity, a value for the elastic modulus (E_y) is determined using the Moens–Korteweg equation  

where h=wall thickness.

Results

Non-pulsatile data

No significant difference existed between baseline PAP, PAF or PVR in the two groups of animals. Control animals (n=7) were not pre-treated with guanethidine and exhibited an increase in PAP (20.1±1.9 mm Hg versus 24.5±2.4 mm Hg, P≪0.05) and no alteration in pulmonary artery flow (4.6±0.9 ml/s versus 4.4±0.6 ml/s, P=1.0, NS). PVR increased by 66% (4860±341 dyne s cm⁻⁵ versus 8090±387 dyne s cm⁻⁵, P≪0.01) with hypercapnia. In contrast, animals pre-treated with guanethidine (n=7) still underwent an increase in PAP (21.8±2.4 mm Hg versus 28.3±3.7 mmHg, P≪0.05), but this was accompanied by an equally large increase in pulmonary artery flow (4.0±0.4 ml/s versus 4.8±0.4 ml/s, P≪0.05). The net result was a markedly attenuated change in PVR (5552±368 dyne s cm⁻⁵ versus 7105±611 dyne s cm⁻⁵, P=0.31, NS), a rise of only 28%. Left atrial pressure and aortic pressure were not altered by CO₂ inhalation in either group. Arterial partial pressure of CO₂ was increased in the control group and in the guanethidine group; resulting in a significant decline in pH in both groups during hypercapnia (Table 1 ).

Table 1

Non-pulsatile hemodynamic response to hypercapnia

Pulsatile data

Input mean impedance, seen at 0 Hz in Fig. 1 , underwent an increase of 44% upon administration of the hypercapnic challenge (7215±495 dyne s cm⁻⁵ versus 10228±993 dyne s cm⁻⁵, P≪0.01) in control animals. Characteristic impedance (Z_o) calculated between 3–10 Hz was not altered (Table 2 ). In the animals pre-treated with guanethidine, input mean impedance, seen at 0 Hz in Fig. 2 , rose by only 23% an increase that did not reach significance (7922±446 dyne cm s⁻⁵ versus 9745±580 dyne cm s⁻⁵). Characteristic impedance was again unaltered with hypercapnia. No change was seen in the R_mn or E_y of the proximal pulmonary arterial vessels in the control or the guanethidine treated group during the stress periods (Table 2).

Fig. 1

Pulmonary arterial impedance moduli and corresponding phase angles in control animals. Input mean impedance is at zero Hz and increases with hypercapnia. Characteristic impedance is between 3–10Hz and is not altered by hypercapnia.

Table 2

Pulsatile hemodynamic response of the proximal pulmonary arterial circulation to hypercapnia

Fig. 2

Pulmonary arterial impedance moduli and corresponding phase angles in guanethidine pre-treated animals. Input mean impedance is at 0 Hz. Note that the large increase seen in the control animals is no longer present after sympathetic blockade. Characteristic impedance is between 3–10 Hz and is not altered, similar to the response observed in the control animals.

Discussion

Hypercapnia, even very brief exposures, has been shown to have a potent vasoconstrictor effect on the neonatal pig pulmonary arterial circulation [1]. Clinically, adult patients undergoing coronary artery bypass grafting experience large increases in pulmonary artery pressure and pulmonary vascular resistance after only mild elevations in PaCO₂[7],[15],[16]. Similar studies in infants following cardiopulmonary bypass showed increased pulmonary artery pressures and resistances with increased end-tidal CO₂ levels [4],[17]. Despite the obvious clinical relevance of carbon dioxide on the newborn and adult pulmonary arterial circulation, almost no work has been done to determine the mechanisms through which its effects are mediated. Specifically, little has been done to determine the precise role of the sympathetic nervous system in this response. While stimulation of the sympathetic nervous system has consistently been shown to alter pulmonary arterial hemodynamics, multiple studies utilizing receptor blockade have demonstrated little effect of the autonomic nervous system on pulmonary arterial tone at rest [10],[18],[19],[20]. Sympathetic stimulation produces both vasoconstriction and vasodilation in feline pulmonary arteries secondary to selective stimulation of α and β receptors respectively [18]. In addition to this receptor specific action, there appears to be a regional specificity in the response to the sympathetic nervous system. Stellate ganglion stimulation in dogs produces a vasoconstrictor response in the larger, more proximal pulmonary arteries with the smaller resistance vessels unaffected [20]. These studies provide the rationale for studying the role of the sympathetic nervous system in modulating the pulmonary arterial hemodynamic response to hypercapnia.

Our study used a pulsatile engineering analysis to determine regional differences within the pulmonary arterial circulation of neonatal piglets. Piglets (2-day-old) were chosen because of the close similarity to the neonatal human pulmonary arterial circulation with regard to both baseline hemodynamics and maturation during the first several weeks of life [1]. Rapid alterations in pulmonary arterial hemodynamics occur in the early newborn period and by 7–10 days of age more closely resemble the adult than they do the neonate. In addition to the much higher values for PAP and PVR in the immediate post-partum circulation, newborn animals have been shown to exhibit greater vasoreactivity than older animals [3]. The studies described herein use carefully selected animals to accurately represent the unique neonatal period, and may therefore be more clinically applicable to newborn human physiology than studies using adult animals. Other possible explanations are discussed below.

Response of the neonatal piglet pulmonary circulation to hypercapnia

Control animals exposed to elevated levels of inspired carbon dioxide showed an increase in pulmonary artery pressure and pulmonary vascular resistance, consistent with the findings of other animal and human studies [1],[2],[4],[7],[15],[16],[21],[22]. Left atrial pressure remained constant during all protocols and therefore, did not account for the alterations we observed in vascular resistance. Input mean impedance (Z_m) is calculated from the mean terms for pulmonary artery pressure and pulmonary artery flow and is therefore highly analogous to pulmonary vascular resistance. It is represented in Fig. 1 by the impedance modulus at 0 Hz. The rise in Zm and PVR with hypercapnia represents vasoconstriction in the region of the distal pulmonary arterioles. In contrast, characteristic impedance or Z_o (impedance modulus between 3–10 Hz in Fig. 1) is a complex term with multiple variables. It occurs in that part of the circulation that is minimally affected by reflected waves and therefore more closely describes the geometry and compliance of the larger, more proximal pulmonary arteries. A careful examination of the two most critical components of characteristic impedance, modulus of elasticity and radius, allows a more complete analysis of the larger pulmonary arterial vessels. By showing that the elastic modulus was unchanged, we demonstrate that no active alteration of the visco-elastic properties of the main pulmonary artery occurred. Similarly, no passive dilatation or constriction occurred, as evidenced by the lack of alteration in the radius measurements. Failure to alter the characteristic impedance term therefore represents a true lack of response in the larger proximal pulmonary arteries to alterations in PaCO₂ and pH. These results agree with earlier work using an X-ray TV system to study alterations in small diameter pulmonary arteries and veins in which pulmonary arteries with a diameter between 200–600 microns were seen to decrease in diameter under the influence of hypercapnia, with no change seen in the larger arteries [23]. Thus, in our animals with an intact sympathetic nervous system, only the distal or arteriolar `resistance' region was affected by hypercapnia, while the larger pulmonary arteries were unaffected.

Alteration of the hypercapnic response after chemical sympathectomy

In this investigation, the pulmonary arteriolar vasoconstrictor response to hypercapnia was markedly attenuated in chemically sympathectomized animals. The increase in PVR and Z_m observed in control animals was markedly blunted in the treated animals; resistance rose only 28%, versus 66% in untreated animals (P≪0.05), and Z_m rose only 23%, versus 44% in untreated animals (P≪0.05). These results demonstrate a requirement for intact sympathetic innervation for full expression of the pulmonary vascular response to hypercapnia. However, guanethidine did not completely ablate the hemodynamic effects of hypercapnia. While the increases in PVR and Z_m did not reach significance in the treated animals, both pulmonary artery pressure and flow did increase. Other studies support the observation of increased cardiac output with hypercapnia and have proposed a direct vasodilator effect for increased PaCO₂ when not accompanied by a fall in pH [18],[24]. Our studies show that hypercapnic acidosis in an intact neonatal piglet model is a pulmonary arterial vasoconstrictor in the presence of an intact sympathetic nervous system, but that response is sharply attenuated and PAF is augmented in the presence of sympathetic blockade. Interestingly, the alteration in the response did not occur in the proximal circulation as suggested by the work of Pace, in which stellate ganglion stimulation in adult dogs produced vasoconstriction of the larger, more proximal vessels [20]. The contribution of the sympathetic innervation in our studies appears to be at the level of the resistance vessels. The differences between these earlier studies and our current studies may be a result of either species or age-dependent differences.

Mechanisms involved in hypercapnic vasoconstriction and the response to sympathectomy

Control of cardiovascular hemodynamics is generally considered to be under the influence of both the sympathetic nervous system and the adrenal glands. In this study adrenal function was left intact while the sympathetic nervous system was selectively blocked. Specifically, guanethidine blocks pre-synaptic release of norepinephrine, but does not affect function of the α and β receptors that are known to innervate the pulmonary arterial circulation. Epinephrine, the predominant product of the adrenal medulla has been shown to be reflexly released during hypoxia and cause a β-mediated relaxation of small diameter pulmonary vessels [25]. It has also been shown that this β-relaxation occurs without the requirement for α-blockade. If the response to hypercapnia is similar to that seen with hypoxia, much of what we observed in our guanethidine treated animals can be explained by the action of epinephrine derived from the adrenal medulla, unopposed by norepinephrine release from sympathetic nerve endings. The net result would be an α-mediated vasoconstriction in the intact animals and a β-mediated relaxation in guanethidine treated animals.

Since Liljestrand first proposed in 1958 that hypoxic pulmonary vasoconstriction resulted in increased hydrogen ion concentration in the blood secondary to lactic acid production and carbon dioxide retention, it has been generally accepted that alterations in the local environment are sufficient to produce a vasomotor response [26]. While the concept of lactic acid as the primary mediator quickly fell out of favor, it became apparent that alterations in pH markedly affect the ability of hypoxia to vasoconstrict the pulmonary circulation [19],[21]. Further evidence for the importance of the local cellular environment comes from observations that a superimposed metabolic alkalosis ablates the vasoconstrictor effect of hypoxia, and that acidosis potentiates the effect [5]. Other studies with hypoxic vasoconstriction have demonstrated that not only is the local response critical, but may be all that is required; with no role for the sympathetic nervous system [27]. Since it is known that decreased pH and possibly increased PaCO₂ alone can cause increases in PVR, it is reasonable that our model of respiratory acidosis would cause some alterations in PVR and impedance regardless of sympathetic innervation. This study intentionally chose not to correct for acidosis and therefore does not distinguish between effects secondary to changes in hydrogen ion concentration and carbon dioxide. This more accurately represents the respiratory acidosis commonly seen in the sick post-operative neonate. Past studies that corrected shifts in pH with sodium bicarbonate indicate the hydrogen ion concentration and not PaCO₂ is more important in the hypertensive response [5],[19]. An increase in PaCO₂ without a fall in pH does not seem to be sufficient to cause the hypertensive response and is important in the understanding of clinical scenarios such as the use of permissive hypercapnia [22]. This strategy, coupled with metabolic alkalosis, has been shown to be effective in infants with increased pulmonary vascular resistance after cardiac surgery [17]. Our studies demonstrate the potential efficacy of sympathetic blockade to further blunt the deleterious effects of hypercapnic acidosis.

In addition to direct effects on the pulmonary vasculature, carbon dioxide has been shown to have significant effects on the heart itself. Rose noted that the combination of hypercapnic acidosis and increased afterload caused impaired right ventricular function [28]. However, in the absence of an elevated pulmonary vascular resistance, hypercapnic acidosis had little effect on right ventricular function. In our study, the large increase in pulmonary vascular resistance and input mean impedance seen in control animals did not occur after pre-treatment with guanethidine. The attenuated increase in afterload seen with sympathetic blockade may serve to preserve right ventricular performance and maintain cardiac output as we observed. Therefore, while hypercapnic acidosis results in pulmonary hypertension clinically, the detrimental effects on the neonatal myocardium may be prevented by sympathetic blockade.

Conclusion

In conclusion, an intact sympathetic nervous system is required for full expression of the hypercapnic vasoconstrictor response in the neonatal pulmonary arterial circulation. Sympathetic blockade with guanethidine blunts this response. In neonatal animals the vasoconstrictor response to hypercapnia occurs in the distal arteriolar region, with no alterations in either the distensibility or geometry of the proximal pulmonary arteries. These data provide insight into the mechanisms responsible for hypercapnic vasoconstriction in the newborn population and indicate a potential role for sympathetic blocking agents in the treatment of pulmonary hypertensive disease in the neonate and as an adjunct in the use of permissive hypercapnia.