Management of pulmonary hypertension: physiological and pharmacological considerations for anesthesiologists.

Abstract

For decades the pulmonary circulation was not considered as important as the systemic (“greater”) circulation. However, pulmonary hypertension can arise because of many diseases of the heart and lung. Therefore, increasing efforts in research have been undertaken leading to a profound increase in understanding pulmonary vascular physiology and pathophysiology. This review discusses basic physiology, clinical concepts, and treatment options for pulmonary hypertension and the related right ventricular heart failure focusing on secondary pulmonary hypertension in patients during anesthetic procedures. Physiology of Pulmonary Circulation The pulmonary vascular bed is a high-flow, low-pressure circulation system. Pulmonary vessels have less resistance in comparison to the systemic circulation under normal conditions because of higher compliance of pulmonary precapillary arterioles with a thinner media and less smooth muscle cells (SMCs) compared with the corresponding systemic arterioles. In addition, the cross-sectional area of the pulmonary vascular bed is large and highly distensible with recruitable vessels available to accommodate increase in flow resulting in low pressure and low resistance. In contrast to the systemic arteries, pulmonary vessels constrict with hypoxia (Euler-Liljestrand reflex) and relax in the presence of hyperoxia. Furthermore, changes in cardiac output (CO), airway pressure, and gravity affect the pulmonary more than the systemic circulation. Increases in CO distend open vessels and recruit previously closed vessels. Therefore, the cross-sectional area of pulmonary circulation enlarges and results in a decrease in pulmonary vascular resistance (PVR). An increase in CO has very little effect on pulmonary arterial pressure (PAP) because of the recruitment and distension of pulmonary vessels. An increased PAP or left atrial pressure (LAP) may also distend and recruit pulmonary vessels. Clinically, this means that enhanced CO caused by the administration of inotropic drugs or enlarged blood volume will passively decrease PVR. The contribution of intra- and extraalveolar vessels accounts for the unique U-shaped relationship between lung volume and PVR, which is minimal at functional residual capacity and increased at large and small lung volumes (Fig. 1). Clinically, this may be observed when hyperinflation of the lungs greatly increases PVR.Figure 1: Relationship between lung volume and pulmonary vascular resistance (PVR) (123). RV = residual volume, FRC = functional residual capacity, TLC = total lung capacity.Gravity influences the distribution of blood flow in the pulmonary circulation. Blood flow as well as ventilation increases in the dependent areas of the lung. The relationship between alveolar and hydrostatic pressure implies important clinical consequences. Patients with unilateral lung disease should be positioned with the diseased side up as was already shown by Remolina et al. (1). Lying with the sick lung dependent resulted in the worst gas exchange and the lowest arterial oxygen pressure (1). Application of high levels of positive end-expiratory pressure (PEEP) will narrow the capillaries in the well-ventilated lung areas and divert flow to less well-ventilated or nonventilated areas. Therefore, a decrease in PaO2 is the consequence. Constantly changing hemodynamics, mechanical forces, and hormonal environment influence the vascular endothelium and the underlying SMCs. Under normal circumstances, the interaction between endothelium and SMCs results in a low vascular resistance in the pulmonary circulation. An increasing number of molecules seem to be involved in these biochemical transductions: L-arginine-nitric oxide (NO)-cyclic-3′-5′guanosine monophosphate (cGMP) pathway seems to have a predominant role (2). Physiological agonists such as bradykinin and acetylcholine as well as mechanical stress derived from pulmonary blood flow can activate endothelial cells. This results in oxidation of the guanidino-nitrogen atom of the amino acid L-arginine by the constitutive NO synthase to form and release NO. NO diffuses from endothelial cells to the SMCs, and produces relaxation by activating guanylate cyclase with increasing intracellular cGMP (2). Pulmonary endothelial cells (surface area of 200 m2 in the adult) have a very strategic location: they are exposed to the entire CO, link the pulmonary and systemic circulation, and also regulate SMC tone by signaling to the vascular wall. Endothelial cells both seem to sense local shear stress and to initiate a response (e.g., production of NO) in order to accommodate rapid changes of blood flow. NO and prostacyclin (PGI2) support oxygenation and lung inflation in dilating the pulmonary vasculature after birth, keeping a key regulatory role in the lung because of its prompt, local powerful action compared with a brief half-life time (3). Pathophysiology of Pulmonary Hypertension It is now understood that the balance between vasoconstrictors and vasodilators and mitogenic and antimitogenic factors derived from the endothelium is disturbed in situations with an increase in PAP (4, 5). Endothelial dysfunction is promoted by hypoxia, acidosis, free radicals (6), inflammatory mediators, shear stress caused by increased pulmonary blood flow from left-to-right intracardiac shunt (7), and fibrin from thromboembolism (8). There is no widely accepted range for normal PAP. However, mean PAP >25 mm Hg (normal 15 mm Hg) at rest or >30 mm Hg with exercise is generally accepted as indicative for pulmonary hypertension (9). The enhanced pressure in the pulmonary circulation is associated with an increase in PVR and results in a progressive inability of the right ventricle (RV) to sustain its output leading to RV hypertrophy and RV failure depending from acute or chronic changes (4). In a study by Vizza et al. (10), prevalence of RV dysfunction (defined as RV ejection fraction [RVEF] <45%) was significantly higher in patients with pulmonary hypertension compared with all other groups with end-stage pulmonary disease. Determination of PVR is difficult in clinical settings even if the respective variables such as LAP, mean PAP, pulmonary capillary wedge pressure (PCWP), and thermodilution CO are measured directly because of their intrinsic inaccuracies. Therefore, indirect estimations use the equation: Normal PVR is approximately 1.1–1. 4 Wood units or about 90–120 dynes · s · cm−5, and a PVR >300 dynes · s · cm−5 is indicative of pulmonary hypertension. Pulmonary blood flow and volume are not always equal to or correlated with CO, because of intracardiac or other shunts. Vascular Remodeling Chronic pulmonary hypertension leads to structural alterations of the pulmonary vasculature and to a progression of histological changes known as “vascular remodeling” (11). Under physiological conditions, pulmonary arterioles are thin-walled vessels with the media occupying only 7% of the vessel thickness. The major finding in remodeled vessels caused by a chronic stimulus like hypoxia is the increase in SMCs in already muscularized arteries and extension of SMCs into vessels that are normally thin and nonmuscular (12) and thickening of the adventitial layer (13). Thickening of the adventitial layer is the result of marked proliferation of the fibroblasts, which has been shown to be modulated by protein kinase C and mitogen-activated protein kinase (14). After the proliferative response, there is an increase in adventitial connective tissue including a switch in SMC phenotype to a more synthetic cell that is responsible for deposition of increased connective tissue (12). This deposition of most notably collagen is probably a protective mechanism strengthening the vascular wall against the increase in intravascular pressure (15). Additionally, damage to intimal endothelium as well as intimal hyperplasia and fibrosis can be detected (5). The central regulatory function of the pulmonary endothelium is underlined by the fact that a dysfunctional endothelium can influence the development of pulmonary hypertension at the levels of coagulation control, vasomotor tone, and pulmonary vascular remodeling (16). The major stimulus for remodeling is hypoxia (17). The increase in PVR is predominantly caused by the hypoxic pulmonary vasoconstriction (HPV), which resides primarily distal to lobar arteries and proximal to the capillaries and occurs in resistance arterioles 30–300 μm in diameter (18). Pulmonary arteriolar SMC oxygen-sensitive voltage-dependent potassium channels seem to have an important role for initiating HPV (19). Inhibition of these channels by decreased PO2 inhibits outward potassium current, causing membrane depolarization and calcium entry through voltage-dependent calcium channels (19). The main determinant of HPV is alveolar PO2, but mixed venous PO2 contributes to approximately 20% of the response (19). HPV is inhibited by substances such as substance P, atrial natriuretic peptides, by mediators such as PGI2 and NO, by increased LAP, by increased alveolar pressure, and by alkalosis. Enhanced HPV, however, could be observed with acidosis, by using epidural anesthesia and by inhibition of cyclooxygenase or NO synthase (19). Subsequent increases in PAP are predominantly caused by vascular remodeling as well as secondary polycythemia. Proliferation and differentiation of pericytes and intermediate cells to SMCs are a consequence of chronic exposure to hypoxia followed by elastin and collagen synthesis and deposition (20). In addition, vasomotor function may be altered in these remodeled vessels (20). Among the candidates of biochemical mediators of hypoxia-induced pulmonary hypertension are voltage-gated potassium channels, mitochondrial oxygen sensing, and an imbalance in vasoactive factors (see below) (21). Remodeling also happens in inflammation secondary to sepsis (22) and in patients with chronic lung disease (23) and adult respiratory distress syndrome (ARDS) (24). The inflammation process and increased blood flow lead to vascular remodeling by damaging endothelial cells and disturbs the sensitive balance of pulmonary tone. Under physiological conditions, these cells eliminate factors that initiate SMC proliferation such as angiotensin II, endothelin, thromboxane A2, prostaglandin H2 and O2−(25). Alternatively, damaged endothelial cells may fail to produce inhibitory factors, possible heparin-like substances, which normally decrease SMC proliferation. Removal of the inciting stimulus can lead to reversal of structural changes. Etiology of Pulmonary Hypertension Pulmonary hypertension is caused by a variety of acute and chronic pulmonary diseases with an increase in PAP (Table 1). In contrast to secondary pulmonary hypertension, primary pulmonary hypertension (PPH) is not related to a known underlying disease.Table 1: World Health Organization Diagnostic Classification of Pulmonary Hypertension (1998) (21)PPH PPH is rare (1–2 per million in the general population), affects more women than men (1.7:1), sometimes has a familial link (6% of all cases), and has a poor prognosis (median survival 2–3 yr after diagnosis). PAP usually is >60 mm Hg in these patients (9, 26). PPH is a diagnosis of exclusion, because it is idiopathic. Hypoxemia and increased PVR resulting in an increased PAP can lead to RV failure and death. Altered vasoreactivity alone is not responsible for PPH. Vascular remodeling contributes to the vasculopathic process. However, regression of extensive changes caused by hypertension can occur. Secondary Pulmonary Hypertension Secondary pulmonary hypertension is more common, the increases in PAP are generally less severe (mean PAP <40 mm Hg) (26), and it is also more frequently seen in the perioperative period. Therefore, in this review, we have focused on secondary forms of pulmonary hypertension. Most cases of pulmonary hypertension are secondary to cardiac or pulmonary disease (Table 1) and may be reversible in some cases. Clinical manifestations might be overshadowed by the symptoms of the underlying disease (Table 2). Blood vessel changes are not confined to PPH but are also found in many forms of secondary pulmonary hypertension. Congenital cardiac disease with an atrial or ventricular septal defect causes an increase in pulmonary blood flow. Over time, in response to the increased blood flow, the PVR is increased and eventually will exceed systemic vascular resistance (SVR). Under these conditions, blood flow is shunted from right to left, often referred to as Eisenmenger syndrome. In contrast, high-pressure shunts such as those associated with truncus arteriosus, ventricular septal defects, or patent ductus arteriosus (28) cause pulmonary hypertension much earlier and it is often more severe.Table 2: Common Symptoms and Signs of Pulmonary Hypertension (27)The relevance of pulmonary hypertension in the perioperative period was shown in a study by Reich et al. (29). In patients undergoing coronary artery bypass grafting with cardiopulmonary bypass (CPB), the development of pulmonary hypertension was a significant predictor of increased mortality and perioperative myocardial infarction (29). The severity of the observed vasoconstriction correlates with the extent of CPB-induced endothelial injury (30). Levels of thromboxane A2(31) and endothelin (32) were increased after CPB, whereas PGI2 and NO levels were reduced (33). For a long time, it was thought that CPB did not cause any pulmonary endothelial injury due to ischemia-reperfusion because of the protection via the vasa vasorum of the bronchial circulation. It has been shown, however, that total CPB may cause complete cessation and reestablishment of pulmonary artery flow, with inadequate pulmonary endothelial blood supply by the vasa vasorum, resulting in an ischemia-reperfusion injury (34). The intraoperative injury and postoperative endothelial dysfunction of the pulmonary endothelium is promoted by the following factors (2): Preoperative status of the pulmonary vascular bed, i.e., pulmonary hypertension, valvular pathology; chronic obstructive pulmonary disease, pulmonary thromboembolism, and shear stress caused by increases in pulmonary blood flow and pressure of left-right intracardiac shunt. Bando et al. (35) demonstrated in patients with congenital heart disease that preoperative pulmonary hypertension, absence of mixed venous saturation monitoring, and absence of prophylactic α-blockade significantly increased postoperative pulmonary hypertension. Intraoperative vasospastic stimuli, such as hypoxia, hypercarbia, acidosis, duration of total CPB, ischemia-reperfusion injury, free radical formation, inflammatory mediators, pulmonary leukosequestration, excess thromboxane or endothelin production, and microemboli. Postoperative factors such as adrenergic tone, atelectasis, and HPV. Other examples of pulmonary hypertension during anesthesia are shown in Table 3.Table 3: Occurrence of Pulmonary Hypertension During Anesthesia (36)Symptoms and Diagnosis of Pulmonary Hypertension The most common clinical signs of pulmonary hypertension are dyspnea and fatigue (9) (see also Table 2). These major symptoms might be explained through the associated decrease in CO; however, an exact etiology is still missing. On physical examination, a prominent P2 heart sound, a tricuspid regurgitation murmur, an atrial (S3), or ventricular (S4) heart sound might be heard. The chest radiograph can display an enlarged main pulmonary artery and enlarged hilar vessels, whereas the electrocardiogram shows a right axis deviation suggesting right ventricular hypertrophy. However, the validity of acute changes in the electrocardiogram attributed to right heart insufficiency in the intraoperative period is limited. Transesophageal echocardiography (TEE) as a more advanced technique diagnoses pulmonary hypertension indirectly attributed to RV enlargement, paradoxical interatrial and interventricular movement, partial systolic closure of pulmonary valve, tricuspid regurgitation, and an increased RV systolic pressure (Fig. 2). There are two approaches to obtaining hemodynamic measurements using TEE. By using spatial imaging methods, cardiac chamber volumes can be estimated to obtain both preload and stroke volume. In addition, Doppler-based methods can be used to estimate both right ventricular filling and CO (37). Because of the inherent difficulty in developing a simple geometric model of the RV, regression equations may inaccurately predict ventricular volumes. In addition, TEE measurements are further altered by changing loading conditions, which alter both the size and geometry of the RV. Although TEE evaluation of RV volume can be difficult to determine in adults, high-quality images are routinely obtained in pediatric patients. One of the main limitations on further use of TEE is the availability of equipment as well as costs of echocardiography machines and probes, and expertise (37). The definitive diagnosis is obtained by right heart catheterization with direct measurement of PAP, right atrial pressure (RAP), PCWP, and CO.Figure 2: Transesophageal echocardiography for diagnosis of pulmonary hypertension. Dilated right ventricle (RV) in the midesophageal four chamber view. The figure shows the classical septal (S) interventricular shift to the left side. RA = right atrium, LA = left atrium, LV = left ventricle.RV and Pulmonary Hypertension Like the pulmonary circulation, the RV was not considered as important as the left ventricle (LV) in maintaining normal hemodynamics and counted for a long time as a “quantite negliable” (38). Today it is recognized that RV and LV are interdependent and both have vitally important functions. The RV is a thin-walled, highly compliant, but poorly contractile chamber. Under normal loading conditions and when function is not compromised, RV ejects blood against 25% of the afterload of the LV, resulting in a smaller RV wall thickness (39). The RV is bound by the RV free wall and the interventricular septum. Failure of the septum (e.g., because of ischemia) to contract normally will decrease RV systolic function. Blood supply for the RV and the septum depends mostly on whether a right or left dominant, or a “balanced” coronary circulation is present. Usually the right and the left anterior descending coronary artery supply the septum and parts of the free wall of the RV. The continuous pressure gradient between the aorta and the RV (coronary perfusion pressure) is responsible for the coronary blood flow to the RV free wall throughout systole and diastole (40). Therefore, the RV blood/oxygen supply is proportional to the systemic pressure but inversely proportional to the RV pressure. Systemic hypotension or increased RV pressure results in a decreased RV coronary perfusion pressure (41). Vlahakes et al. (42) previously showed that right heart performance is directly related to systemic pressure during pulmonary hypertension. RV function is very sensitive to increased RV preload or afterload, septic shock, coexistent LV dysfunction, or right coronary artery occlusive disease (43, 44). In a fundamental study, Urabe et al. (45) demonstrated that an increase in the afterload of the RV resulted in a rightward shift of the relationship between perfusion pressure and regional shortening. Also, end-diastolic segment lengths increased significantly after banding of the pulmonary artery. In contrast to LV performance, RV function is relatively sensitive to increases in afterload. Acute increase in mean PAP above approximately 40 mm Hg results in a significant decrease in RVEF even in the presence of a normal RV contractility. However, gradual increases in afterload are well tolerated, because the RV has time to assemble new sarcomeres in parallel to increase wall thickness (39). In the presence of decreased RV contractility, the RV is even more susceptible to acute increases in afterload. The decrease in RVEF results from a disproportionate increase in end-systolic volume compared with end-diastolic volume (41). The RV is less preload-responsive than the LV such that a given increase in preload results in a smaller increase in stroke work. Therefore, attempts at volume loading may be less effective in increasing RV output compared with volume loading of the LV. In addition, volume loading may even worsen the ratio of RV oxygen demand/supply (41). Vlahakes (46) concluded that two important principles emerged in the management of right heart failure: First, RV afterload must be reduced and second, systemic pressure must be maintained or increased. Normal atrioventricular conduction and contraction are essential for maintaining normal RV function (47). The maintenance of sinoatrial activity is as important for RV efficiency as it is for LV efficiency (47). RV pump performance is determined by both intrinsic factors (contractile state of the RV myocardium) as well as extrinsic factors (preload, afterload, constraining effects of the pericardium, intrapericardiac pressure, right coronary artery perfusion pressure, LV performance, and the contractile state of the interventricular septum) (48). All abnormalities of the LV function like coronary artery disease, congestive heart failure, valvular heart disease, or systemic hypertension influence RV function by ventricular interdependence. Also, a dilated RV and right atrium can shift the interatrial and interventricular septum and compress the left atrium and reduce LV end-diastolic volume. Central venous pressure does not provide accurate information on RV transmural filling pressure and volume because it is dependent on the compliance of the RV and/or abnormalities in the tricuspid valve apparatus. Angiography, echocardiography, radionuclide methods, thermodilution, and magnetic resonance imaging are better and more reliable modalities in measuring the RV performance. RV dysfunction caused by increased PAP was demonstrated in patients with chronic obstructed pulmonary disease (COPD) (49), ARDS (50), and in patients receiving protamine infusion (51, 52). Mechanical ventilation can also impair RV performance. PEEP (30 cm H2O) shifts the interventricular septum producing paradoxical motion, resulting in right heart dilation and decreased LV chamber size (53). In postoperative ventilated patients, a decrease in RVEF is observed at PEEP levels >15 cm H2O because of compression of alveolar vessels and accompanying increase in RV afterload (54). A significant inverse correlation was noted between RVEF and increasing amounts of airway pressure delivered by high-frequency jet ventilation (55). Treatment Treatment of the underlying disease has priority. Because symptoms often arise late in pulmonary hypertension, the average life expectancy after occurrence of signs of manifest RV insufficiency is <1 yr (56). Basic Considerations Antiobstructive therapy in COPD, corticoids for interstitial lung disease, as well as systemic anticoagulation for chronic lung embolism are fundamentals of therapy of pulmonary hypertension. Elimination of ventilation/perfusion mismatch, which causes dystelectasis and atelectasis and antibiotic therapy for pneumonia can also improve symptoms. Early definitive repair of congenital heart diseases can reduce morbidity and mortality from postoperative pulmonary hypertension (35). A common clinical problem is the decompensation of the right heart because of chronic left heart insufficiency. In this situation, therapy of the LV has priority. Infarction of the right heart needs immediate revascularization by means of thrombolysis, percutaneous transluminal coronary angioplasty, or coronary artery bypass grafting. Symptomatic Therapy Symptomatic therapy supports causal treatment and should reduce PVR, which can be achieved with the following active and passive factors influencing the pulmonary circulation: Improving oxygenation with 100% oxygen Avoidance of respiratory acidosis, moderate hyperventilation (PaCO2 30–35 mm Hg) Correction of a metabolic acidosis (aim: pH >7.4) Recruitment maneuvers, to avoid ventilation/perfusion mismatch Adaptation of respiratory therapy, avoiding overinflation of the lung alveolae Avoidance of catecholamine release caused by stress situations: adequate analgesia and sedation Avoidance of shivering: body temperature 37°C Specific Treatment Treatment options for RV dysfunction with increased PVR include vasodilative drugs, whereas positive inotropic drugs are the treatment of choice in RV dysfunction with normal PVR. Optimization of RV Preload. Optimization of RV preload should be considered if central venous pressure is <10 mm Hg (57). It is recommended, however, that preload of the RV should be estimated by a simple “endogenous volume loading” test (i.e., raising patients’ legs) in every single patient. “Responders” show an increase in mean arterial blood pressure (MAP) and can increase RVEF with volume therapy, if RV contractility is normal and PAP is increased only slightly (57). However, administration of volume also can have deleterious effects in patients with hemodynamically relevant right heart infarction. Routine treatment in our institution includes diuretics for “nonresponders” (defined as patients showing no increase in MAP to the above-mentioned “volume loading” test) who present with central venous pressure <20 mm Hg. Reduction of RV Afterload. Reduction of the cross-section of the pulmonary circulation of >60%–70% decreases CO and induces RV failure (58). As previously mentioned, baseline support, IV or inhaled vasodilators should be used, keeping in mind to avoid a significant decrease in SVR. If the decrease in SVR exceeds the improvement in RV stroke volume, greater systemic hypotension can be the consequence. Magnesium is thought to produce vasodilation by blocking calcium channels (59). Magnesium has other beneficial properties such as the ability to enhance NO synthase activity, activate adenylate cyclase, and release PGI2(59). Adenosine stimulates purine receptors in both endothelial and SMCs. Relaxation of SMCs is mediated by both release of NO and direct stimulation of SMCs (60). Adenosine has a short half-life (9 s), therefore it is relatively selective for the pulmonary circulation. It has been shown to be a potent vasodilator and predictive of other drugs to produce pulmonary vasodilation in studies of patients with PPH (61, 62). Ventilation/perfusion mismatching remains a problem, but there are no arrhythmias observed because of the small dosing schedule (50–200 μg/kg body weight/min). Angiotensin-converting enzyme inhibitors: In contrast to long-term treatment (3–6 mo) with oral captopril that led to reduced PAP and PVR in patients with secondary pulmonary hypertension (63), short-term treatment did not show any improvement (64). This difference might be explained by the fact that angiotensin-converting enzyme inhibitors depend on their ability to reduce vascular remodeling (65). The angiotensin receptor antagonist losartan successfully reduced PAP and PVR in patients with secondary pulmonary hypertension 4 h after application (66). Calcium antagonists have been shown to be effective in the treatment of pulmonary hypertension secondary to connective tissue vascular disease (67), but not in patients with COPD (68), where treatment was limited because of the deleterious effects on venous admixture (69). The effectiveness of therapy with calcium antagonists in secondary pulmonary hypertension seems to depend on the initial level of PAP, i.e., the higher the initial level of PAP, the less effective the drug (67, 68). A metaanalysis of eight long-term trials (70) to investigate the effectiveness of nifedipine on reducing PAP in patients with pulmonary hypertensive disorders showed that a reduction in PAP occurred in seven of eight trials. The largest reduction in PAP occurred in patients treated with the largest dosages and the reduction in PAP corresponded with subjective clinical improvement. Although these drugs have clearly shown their efficacy, they have severe adverse outcomes if used improperly; therefore, it is recommended that initial calcium antagonist treatment should be limited to specialized centers to avoid complications (71). Patients with pulmonary hypertension and severe clinical right heart failure (mean RAP >20 mm Hg, CO <2 L/min) should be excluded from treatment with calcium antagonists because of the negative inotropic effects of these drugs. Milrinone/amrinone (phosphodiesterase [PDE] inhibitors) act by inhibiting one or more enzymes responsible for the breakdown of cyclic 3′-5′-adenosine monophosphate (cAMP)/cGMP leading to an increased amount of these cyclic nucleotides with increased LV contractility and pulmonary vasodilation. Both PDE inhibitors, given IV, have been used successfully in patients with pulmonary hypertension after cardiac surgery (72). Inhaled aerosolized milrinone was shown for cardiac surgical patients with increased PVR to induce selective pulmonary vasodilation without systemic effects. It also seems to have an additive pulmonary vasodilatory effect to inhaled PGI2(73). PGI2: The potent vasodilator PGI2 was first reported in 1980 to reduce PAP in PPH (74). It is produced mainly by the vascular endothelium and acts via specific prostaglandin receptors linked to adenylate cyclase with following increase of cAMP. PGI2 production is impaired in patients with pulmonary hypertension (75), so therapeutical addition of PGI2 may act in part to replace these deficiencies. Additional beneficial effects are an inhibition of both platelet aggregation and SMC proliferation (76). Apart from treatment of PPH, PGI2 or its stable analog iloprost also decreases PAP in persistent pulmonary hypertension of the newborn (77), pulmonary hypertension after heart surgery in infants (78), ARDS (79), or pulmonary hypertension secondary to connective tissue diseases in a