Inhaled Nitric Oxide : Basic Biology and Clinical Applications

Nitric oxide is a colorless, almost odorless gas that is slightly soluble in water (2 or 3 mM). 13Environmental NO arises from combustion processes (e.g. , fossil fuel combustion and tobacco smoke) and lightning. 14Atmospheric concentrations of NO usually range between 10 and 500 parts per billion (ppb), but can exceed 1.5 parts per million (ppm) in areas of heavy traffic. 15Concentrations of NO produced in the hot cone of a glowing cigarette can reach 1,000 ppm in a 40-ml puff. 16The Occupational Safety and Health Administration has set 8-h time-weighted average exposure limits in the workplace at 25 ppm for NO breathing and at 5 ppm for nitrogen dioxide (NO²). 17Commercially, NO is manufactured from the reaction of sulfur dioxide with nitric acid, from the reaction of sodium nitrite and sulfuric acid, or by the oxidation of ammonia over a platinum catalyst at a high temperature (> 500°C). 18In an anaerobic environment (i.e. , in highly purified nitrogen), NO can be stored for several years.

In the gaseous phase, NO reacts with molecular oxygen to form NO². The conversion rate of NO to NO²can be described by the relation

where k  is the rate constant for conversion of NO to NO². The rate constant has been reported to be between 0.79 × 10⁻⁹to 2.26 × 10⁻⁹· ppm⁻²· min⁻¹, dependent on experimental conditions. 19Approximately half of a 10,000-ppm NO mixture in air is converted into NO²within 24 s, whereas 50% of a 10-ppm NO mixture in air is converted into NO²within 7 h at 20°C. 21In aqueous solution, NO²decomposes to give equal amounts of nitrite (NO²⁻) and nitrate (NO³⁻). 13 

The pathologic effects of NO²inhalation have been studied in various animal species. High levels of inhaled NO²(> 10 ppm) induce pulmonary edema, alveolar hemorrhage, changes in the surface tension activities of surfactant, hyperplasia of type 2 alveolar epithelial cells, intrapulmonary accumulation of fibrin, neutrophils, and macrophages, and death. 22,23Lower inhaled NO²concentrations (< 2 ppm) can alter surfactant function, produce alveolar cell hyperplasia, and alter the epithelium of the terminal bronchioles. 24Inhalation of 2 ppm NO²in humans increases alveolar permeability 25and airway reactivity. 26–28Inhalation of 0.5–1.5 ppm NO²for 9 weeks caused focal degeneration of pulmonary interstitial cells, with mild emphysematous changes, in rats. 29 

Nitric oxide and superoxide (O²⁻) readily react to form peroxynitrite (⁻OONO) at nearly a diffusion-limited rate. 13During physiologic conditions, O²⁻is scavenged by endogenous O²⁻scavengers (e.g. , superoxide dismutase) and formation of ⁻OONO is minimal. During pathologic conditions, such as in the presence of increased concentrations of O²⁻or after O²⁻scavengers are exhausted, significant concentrations of ⁻OONO may be produced. 30Peroxynitrite directly causes oxidation, peroxidation, and nitration of biologically important molecules (e.g. , lipids, proteins, DNA; for review articles see Szabo et al.  31,32). The cytotoxic effects of ⁻OONO provide protective functions if they are directed by inflammatory cells against invading microorganisms or tumor cells.

An important example of a reaction caused by ⁻OONO is the nitration of tyrosine. Tyrosine nitration inhibits tyrosine phosphorylation, alters the dynamics of assembly and disassembly of cytoskeletal proteins, and inhibits tyrosine hydroxylase, thereby reducing dopamine production by neurons and inhibiting cytoskeletal movements of endothelial cells. 31Nitrotyrosine has been detected in lung tissue sections from patients with lung injury, 33,34in atherosclerotic lesions, 35,36and in lungs after ischemia–reperfusion injury. 37 

Exposure of surfactant to high concentrations of ⁻OONO in vitro  reduced its minimum surface tension. 38Peroxynitrite exposure impaired pulmonary surfactant function, because of peroxidation of surfactant lipids, and decreased the ability of the major hydrophilic surfactant, protein A, to aggregate lipids and act synergistically with other surfactant proteins to reduce the minimum surface tension. 39,40These changes of surfactant protein A were associated with nitrotyrosine formation. 39A mixture of surfactant proteins B and C exposed to ⁻OONO was incapable of reducing phospholipid minimum surface tension during dynamic compression. 41 

Peroxynitrite can cause cell apoptosis by DNA strand breakage, activation of poly-adenosine-diphosphate-ribosyltransferase and by inhibition of mitochondrial respiratory enzymes. 31,32Peroxynitrite rapidly reacts with carbon dioxide to form an adduct that participates in nitration and oxidation reactions. 42Interestingly, in a model of thrombin or hydrogen peroxide (H²O²)–induced vascular injury of the rat mesenteric endothelium and in an ischemia–reperfusion model of the rat heart, infusion of ⁻OONO significantly reduced neutrophil adhesion to the endothelium and expression of adhesion molecules, suggesting that ⁻OONO exerts inhibitory effects on neutrophil adhesion in inflammatory processes. 43 

In summary, ⁻OONO is more cytotoxic than NO in a variety of experimental systems, 32and the balance of NO, O²⁻, and O²⁻–⁻OONO scavenging systems determines whether biologically relevant ⁻OONO concentrations will occur in tissues. 30 

Nitric oxide binds to intracellular iron and heme-containing proteins. Examples of heme proteins that are directly affected by NO are oxyhemoglobin, soluble guanylate cyclase (sGC), cyclooxygenase, and cytochrome p450. Guanylate cyclase is stimulated by NO; cyclooxygenase is stimulated by low NO concentrations 44and inhibited by high NO concentrations. 45The cytochrome p450 system is inhibited by NO. 46,47The ratio of rates of uptake and release of NO for ferrous (Fe²⁺) hemoglobin is 10⁵–10⁶times larger than for oxygen. 48Different from other iron–heme ligands, such as carbon monoxide or oxygen, NO can bind with the ferric (Fe³⁺) and Fe²⁺oxidation state of hemoglobin. The vasodilating effects of NO in vivo  are limited by its rapid reaction with oxyhemoglobin or oxymyoglobin to form nitrosylhemoglobin or nitrosylmyoglobin. Methemoglobin (Fe³⁺hemoglobin) is produced when the heme iron is oxidized from Fe²⁺to Fe³⁺and NO³⁻is released. 49Most of the methemoglobin is reduced back to Fe²⁺hemoglobin by NADH-cytochrome b⁵/cytochrome b⁵methemoglobin reductase within erythrocytes. In addition, reduced glutathione reduces methemoglobin. 50 

Nitric oxide can nitrosate thiol groups to form S-nitrosothiols. Common, naturally occurring S-nitrosothiols include S-nitrosocysteine, S-nitrosohomocysteine, and S-nitrosoglutathione. 51S-nitrosothiols have similar platelet-inhibitory and vasorelaxant activities to NO, which are mediated through guanylate cyclase activation, 52but which differ in other important physiologic characteristics from gaseous NO. 53–55 

In addition to binding to the iron–heme center of hemoglobin, NO participates in transnitrosation reactions with the sulfhydryl group of hemoglobin to form S-nitrosohemoglobin. Such reactions may serve as important steps in the uptake and distribution of NO in the systemic circulation. In vivo  analysis reveals that arterial blood samples from normal rats contained larger concentrations of S-nitrosohemoglobin than did venous samples, 56suggesting that S-nitrosylation is regulated by hemoglobin oxygenation and changes with erythrocyte transit through the lungs. Stamler et al.  57demonstrated that hemoglobin cysteines (Cysβ93) participate in the binding and release of NO. When deoxygenated hemoglobin with a high oxygen affinity enters the pulmonary circulation, the affinity of the hemoglobin thiol groups for NO is high and NO uptake occurs. In the peripheral circulation, where oxygenated hemoglobin with a low oxygen affinity releases oxygen to tissues at a low partial pressure of oxygen (PO²), release of NO is enhanced. Such localized release of NO permits vasodilation and increased oxygen delivery to occur in tissues with reduced PO². 57 

Nitric oxide can alter DNA by the formation of mutagenic nitrosamines, 58,59by direct modification and strand breakage of DNA from the formation of radical nitrogen oxide species (e.g. , ⁻OONO), 60–62and by inhibition of enzyme systems that are necessary to repair DNA lesions. 63–65NO deaminates desoxynucleosides and desoxynucleotides in mammalian cell preparations and in aerobic solutions of nucleic acids 66and causes dose-dependent DNA strand breakage. 67In contrast, NO can abate DNA damage caused by xanthine oxidase and H²O². 68–70 

Tumoricidal and tumor-promoting effects of NO have been reported. 65Nitric oxide derived from the inducible NOS (iNOS) of macrophages, Kupffer cells, natural killer cells, and endothelial cells produces tumoricidal effects against many types of tumors, 71–79reduces the viability of several tumor cell lines, 80and inhibits angiogenesis, tumor growth, and metastasis. 81Transfection of iNOS into metastatic melanoma cells reduces their potential for metastasis. 82NO inhibits tumor cell adhesion 83and decreases the metastatic activity of colon cancer cells. 84In other studies, NO mediates tumor growth through NO-mediated control of angiogenesis and of growth factors. 65,85,86Wink et al.  65recently concluded that the role of NO in carcinogenesis is multidimensional. Tissues that are exposed for prolonged durations to high NO concentrations in combination with long-term inflammation and production of reactive oxygen species may accumulate mutations caused by the direct or indirect effects of NO. As a tumor develops, NO produced from iNOS can kill tumor cells through cytostatic and cytotoxic activity. As the tumor progresses, NO may inhibit or support angiogenesis, may limit leukocyte infiltration, and may limit metastasis or kill tumor cells through the induction of apoptosis. 65 

Nitric oxide has contrasting effects on lipids, particularly on the oxidation of low-density lipoproteins in the pathogenesis of atherosclerotic lesions (for review see Rubbo et al.  87). NO inhibits lipid peroxidation by inhibiting radical chain propagation reactions via  radical–radical reaction with lipid peroxyl and alkoxyl groups. 87,88As a ligand to iron (and other transition metals), NO modulates the prooxidant effects of iron and thereby limits the formation of hydroxyl radicals and iron-dependent electron-transfer reactions. 87NO inhibits cell and ⁻OONO-mediated lipoprotein oxidation in macrophage and endothelial cell systems. 89However, NO-induced ⁻OONO formation can oxidize low-density lipoproteins to potentially atherogenic species. 90,91The antioxidant versus  prooxidant outcome of these reactions appears to depend on the relative concentration of the various reactive molecules. 88,92 

Nitric oxide synthase catalyzes a partially tetrahydrobiopterin-dependent five-electron oxidation of the terminal guanidino nitrogen of L-arginine. 93The reaction stoichiometrically consumes oxygen and nicotinamide adenine dinucelotide hydrogen phosphate (NADPH), requires the cofactors flavin adenine dinucleotide, flavin mononucleotide, and calmodulin, and produces L-citrulline and NO. NOS does not produce detectable levels of NO unless superoxide dismutase is present. 94During conditions of L-arginine depletion, NOS generates O²⁻. 95,96NOS is homologous to the cytochrome p450 reductase enzyme containing iron–protoporphyrin IX. 97,98 

Three NOS isoforms have been identified and classified based on the tissue in which they were first identified, the regulation of their activity, and their substrate–inhibitor profile. Constitutive neuronal NOS (nNOS, NOS1) initially was discovered in nerve tissue. Inducible NOS (iNOS, NOS2), a cytokine-inducible isoform, is expressed in a variety of inflammatory cells. Constitutive endothelial NOS (eNOS, NOS3) was originally described in vascular endothelial cells. More recent studies have shown that expression of the constitutive NOS isoforms (nNOS and eNOS) is regulated, 99–102and that the inducible isoform (iNOS2) is constitutively present without previous stimulation. 103NOS isoforms are expressed in many different cell types and intracellular organelles, and most cells are able to synthesize NO. 104,105Altered NOS expression and endogenous NO synthesis have been reported in a large variety of ischemic, traumatic, neoplastic, inflammatory, and infectious diseases. 106–115In addition to enzymatic generation of NO by NOS isoforms, nonenzymatic formation of NO in vivo  during reduced and acidotic conditions (e.g. , organ ischemia) has been reported 116and can contribute to NO production during pathologic conditions.

Neuronal NOS fulfills a myriad of disparate functions in a wide variety of tissues. In the peripheral nervous system, NO acts as a neurotransmitter, regulating smooth muscle relaxation in the gastrointestinal, urogenital, and respiratory tracts via  nonadrenergic noncholinergic nerves containing nNOS. 114Neuronal NOS expression is also present in vasodilator nerves that innervate large cerebral vessels. 117In the central nervous system, NO is essential in neuronal plasticity to modulate information storage in the brain 118and has effects on brain development, memory function, behavior, and pain perception. 114In human skeletal muscle, nNOS modulates contractile force, myocyte development, myofiber differentiation, and myotube innervation. 114Other nNOS expression sites include cardiac nerve terminals that regulate the release of catecholamines in the heart, 119and the retina, where nNOS is involved in NO production in photoreceptors and bipolar cells. 120 

Inflammatory cells (e.g. , macrophages and granulocytes), among many other cell types, express iNOS in response to a variety of infectious and inflammatory stimuli. Inducible NOS produces effects that are beneficial and critical for survival during important bacterial and parasitic infections (e.g. , Mycobacterium tuberculosis, Toxoplasma gondii ) and in the response to inflammation (e.g. , decrease of neutrophil adhesion in endotoxemia, increase of wound closure, and neovascularization of wounds), as shown in murine models of congenital iNOS deficiency. 115In contrast, increased iNOS expression has been associated with the worsening of other infectious diseases (e.g. , influenza pneumonitis) and inflammatory states (e.g. , endotoxin-induced hypotension, autoimmune vasculitis). 115Enhanced expression of iNOS and increased vascular NO synthesis and release have been associated with systemic arterial vasodilation and the “low-tone state” in sepsis. 121 

The 1998 Nobel prize in physiology or medicine was awarded to three researchers who discovered endothelium-derived relaxing factor and demonstrated that NO, generated from eNOS in vascular endothelial cells, is endothelium-derived relaxing factor. 1,2,122Endothelial NOS activity is increased by acetylcholine, bradykinin, and other mediators that increase intracellular calcium concentrations. 122Endothelial NOS activity modulates systemic 122and pulmonary vascular tone 123and plays important roles in lung development and disease. Endothelial NOS expression in the fetal lung changes with lung maturation. 124,125NO production and eNOS expression by endothelial cells is increased by vascular shear stress. 126,127For example, pulmonary vascular eNOS expression is reduced in patients with chronic pulmonary hypertension. 128Congenital absence of eNOS in mice results in pulmonary hypertension and increased right ventricular remodeling if the mouse is stressed by long-term hypoxic breathing. 129Endothelial NOS can be reversibly inhibited by NO 130and eNOS expression can be upregulated by cyclic guanosine monophosphate (cGMP). 102In cardiac endothelial cells, eNOS activity inhibits contractile tone and proliferation of the underlying vascular smooth muscle cells, reduces platelet aggregation and monocyte adhesion, promotes diastolic relaxation, and decreases the oxygen consumption rate of cardiac muscle. 119Endothelial NOS is constitutively expressed in cardiac myocytes, where its activity opposes the inotropic action of catecholamines after muscarinic cholinergic or β-adrenergic receptor stimulation. 119 

Soluble guanylate cyclase (sGC) mediates many of the biologic effects of NO and is responsible for conversion of guanosine-5′-triphosphate to cGMP (fig. 1). Cyclic 3′:5′ GMP is an important second messenger in a variety of cell types and is found in the cytosol of almost all mammalian cells. A variety of nitrovasodilators (e.g. , nitroglycerin, sodium nitroprusside) stimulate cGMP synthesis, which, in turn, is responsible for the smooth muscle relaxation mediated by these drugs. 131The common mechanism of action of these drugs is attributed to the release of NO. 132 

Soluble guanylate cyclase is a heme-containing protein composed of an α and β subunit. The heme moiety in sGC is essential for activation of the enzyme. The presence of heme results in a 100-fold increase of enzyme activity after stimulation with NO, whereas basal enzyme activity is low without heme and does not change with the addition of NO. 133Effects of cGMP on vascular tone, cardiac function, and intestinal water and ion transport by protein kinase-dependent and -independent mechanisms have been reviewed in detail by others. 134,135 

The physiologic action of cGMP is limited by its hydrolysis to GMP by a family of cyclic nucleotide phosphodiesterases. Of the seven known phosphodiesterase isozymes, phosphodiesterases 1 and 5 hydrolyze cGMP. Phosphodiesterase 1 catalyzes cyclic adenosine monophosphate and cGMP hydrolysis and is found in high concentrations in the brain, the heart, the lung, and the testis. Phosphodiesterase 5 is cGMP specific and has been found in lung tissue, platelets, vascular smooth muscle, and the kidney. It has a high affinity for cGMP and can be inhibited by the selective phosphodiesterase 5 inhibitors zaprinast, sildenafil, and dipyridamole. Inhibition of phosphodiesterase 5 enhances endothelium-dependent vasorelaxation, reduces pulmonary vascular tone, and enhances the hypotensive effects of nitrovasodilators. 136,137 

The ability of NO to selectively dilate the pulmonary vasculature was evaluated in an awake lamb model of alveolar hypoxia. Alveolar hypoxia produces reversible pulmonary vasoconstriction mediated by an unknown mechanism. During normoxia, inhalation of 80 ppm NO did not alter the normally low mean pulmonary artery pressure (MPAP) and pulmonary vascular resistance (PVR). With alveolar hypoxia (inspired fraction of oxygen [FI^O2], 0.1), pulmonary vasoconstriction increased MPAP from 17 mmHg to 28 mmHg. With NO inhalation (40 ppm), MPAP decreased to 20 mmHg and further to 18 mmHg (80 ppm NO) within 3 min of NO breathing. 4Cardiac output and systemic arterial pressure were not affected by NO inhalation. These results were confirmed in mechanically ventilated sheep, 138in awake healthy volunteers breathing low oxygen concentrations at ambient pressure 139(fig. 2), and in volunteers at high altitude (at 4,559 m, hypobaric hypoxia). 140 

The pulmonary selectivity of inhaled NO and its rapid inactivation by hemoglobin was first evaluated in an isolated perfused rabbit lung model. The effluent of the perfused lung was conducted to an isolated, pharmacologically preconstricted segment of aorta. The pulmonary vasculature was then preconstricted with U46619, a thromboxane analog. When the perfusate was a hemoglobin-free aqueous buffer, inhalation of NO first decreased MPAP and subsequently the tone of the sequentially perfused aorta. When erythrocytes were added to the perfusate, inhaled NO still caused pulmonary vasodilation, but its vasodilatory effect on the effluent-perfused aorta was abolished, 141suggesting that inhaled NO was inactivated by contact with hemoglobin. The pulmonary vascular selectivity of NO inhalation (i.e. , pulmonary vasodilation in the absence of systemic arterial vasodilation) has been confirmed in numerous subsequent studies. 142–147 

The longitudinal effects of inhaled NO within the pulmonary vasculature are important because increases of arterial or venous vascular tone differentially influence hydrostatic fluid exchange within the lung. Pulmonary venoconstriction increases pulmonary capillary pressure and promotes edema formation. 148,149If venous and arteriolar constriction both contribute to increased PVR, a vasodilator selectively acting on arterial tone may worsen edema formation by increasing mean pulmonary vascular surface area and pressure. Lindeborg et al.  150reported that inhalation of 5, 20, and 80 ppm NO decreased arterial, microvascular, and venous resistances to the same extent in an isolated rabbit lung model. Shirai et al.  151used an X-ray television system to visualize the in vivo  effects of NO inhalation on the internal diameter of pulmonary arteries and veins in a feline model. Inhaling 5–40 ppm NO caused a dose-dependent increase of the diameter of small arteries and veins during normoxic conditions. After induction of hypoxic vasoconstriction by lobar anoxia, NO inhalation dilated smaller constricted and larger nonconstricted arteries, as well as veins. These results suggest that the pulmonary vasodilator response to inhaled NO is similar in pulmonary arteries and veins.

The intrapulmonary distribution of blood flow and ventilation (ventilation–perfusion [V̇/Q̇] distribution) is a major determinant of transpulmonary oxygenation effectiveness, and the resulting partial pressure of oxygen in arterial blood (Pa^O2). In a normal, healthy lung, ventilated areas are well perfused. The shunt from the right to the left side of the circulation is mainly extrapulmonary (e.g. , bronchial veins) and is less than 5–8% of cardiac output. 152Local alveolar hypoxia constricts the vascular bed adjacent to hypoxic regions and redistributes blood flow to lung regions with better ventilation and a higher intraalveolar PO². It has been proposed that inhaled NO amplifies this mechanism by increasing blood flow through well-ventilated lung areas. Pison et al.  138studied the effects of inhaled NO on distribution in an ovine model of acute hypoxia. Because they studied generalized alveolar hypoxia, no improvement of Pa^O2during NO inhalation was expected. However, they demonstrated increased blood flow to better ventilated (but still hypoxic) lung areas and a stable Pa^O2during NO inhalation.

The effects of NO inhalation on gas exchange have been assessed using lung injury models that induce mismatch. The mismatch induced by oleic acid injury in sheep was significantly improved by inhalation of 40 ppm NO 153and was augmented by the simultaneous use of continuous positive airway pressure to open collapsed alveoli. 154Hopkins et al.  155studied the effects of inhaled NO on gas exchange in dogs by selectively creating areas of shunt or areas with a low ratio. NO (80 ppm) decreased blood flow to shunting regions. In areas with a mismatch, NO produced an inconsistent response. When the PVR of the partially obstructed airway regions was decreased by NO inhalation, inequality was increased because blood flow to the relatively poorly ventilated areas was increased by vasodilation. When NO did not reach the lung regions distal to the partial obstruction, and thus did not reduce local PVR, matching was improved.

Nitric oxide synthase inhibitors suppress the bronchodilator actions of nonadrenergic noncholinergic–mediated bronchodilation, suggesting that endogenous NO synthesis is involved in the control of bronchial tone. 156,157The expression of various NOS isoforms in peripheral nonadrenergic noncholinergic nerve endings 158and in human bronchial epithelium 159supports this finding. Dupuy et al.  160demonstrated that inhaled NO decreased airway resistance after bronchoconstriction with methacholine in guinea pigs, later confirmed in various experimental models using rabbits, 161,162dogs, 163–165and pigs. 166,167In human volunteers, however, inhaled NO only reduced airway resistance minimally after a methacholine challenge. 168Large airways appeared to be preferentially dilated by inhaled NO. 169 

Surfactant synthesized by type 2 alveolar epithelial cells affects lung mechanics by reducing surface tension, modifies pulmonary gas exchange, and has antimicrobial functions. 40Isolated type 2 alveolar epithelial cells exposed to NO (generated by the NO donor drugs S-nitroso-N-penicillamine, spermine NONOate, or 3-morpholino-sydnonimine) in the presence of superoxide dismutase reduced their surfactant synthesis by approximately 60%. 170Exposure of surfactant to NO ex vivo  was not associated with changes of surface activity. 38,In vivo , a combination of high inspired oxygen concentrations and high inspired NO concentrations (100 ppm) inhaled by newborn piglets for 48 h significantly decreased the minimum surface tension of surfactant recovered by bronchoalveolar lavage. 38 

In lambs, high inhaled NO concentrations (80–200 ppm) resulted in abnormal surface activities and inhibition of surfactant protein A lipid aggregation. 171In addition, NO (from NO donor drugs) can decrease surfactant protein A gene expression by distal respiratory epithelial cells. 172In contrast, in an experimental model of acute lung injury, combining exogenous surfactant therapy with inhaled NO improved ventilation–perfusion matching and arterial oxygenation. 173Other interactions of NO and surfactant were recently summarized by Hallman and Bry. 40 

Nitric oxide is inactivated by reaction with its biologic target molecules. The half-life of NO in vivo  is only a few seconds. The main metabolic pathways are the binding of NO to O²⁻and to the heme iron of hemoglobin with the subsequent release of NO³⁻. The binding and release of NO to thiols presents another important metabolic pathway. Approximately 90% of NO is absorbed during a steady state inhalation. Almost 70% of the inhaled gas appears within 48 h as NO³⁻in the urine. 174The remaining 30% of inhaled NO is recovered as NO²⁻in the oral cavity through secretion from salivary glands. NO²⁻is also partly converted to nitrogen gas in the stomach and some NO²⁻in the intestine is reduced to ammonia, reabsorbed, and converted to urea. 175,176 

Hypoxia in the preterm and term newborn is usually characterized by severe PAH, extrapulmonary right-to-left shunting, hypoxemia, and acidosis. Roberts et al.  177studied the effects of inhaled NO in hypoxic and acidotic term newborn lambs delivered by cesarean section. Hypoxia associated with hypercapnia doubled PVR. In this model, inhaling 20 ppm NO during hypoxia completely abolished pulmonary vasoconstriction, despite the presence of a marked respiratory acidosis. Similar results were obtained in hypoxic, mechanically ventilated late-gestation ovine fetuses. 178In an experimental model of persistent pulmonary hypertension of the newborn (PPHN) in lambs (induced by ductal ligation), inhaled NO decreased PVR and markedly increased survival rates. 179,180 

The responsiveness of the premature lung to inhaled NO depends on gestational age and the maturity of the pulmonary vasculature. In the immature lung of the ovine fetus at 0.78 of term, initiation of mechanical ventilation caused maximal pulmonary vasodilation and the addition of NO (20 ppm) or 100% oxygen did not increase vasodilation further. If surfactant was administered and mechanical ventilation at 0.78 term continued with 100% oxygen for 2 h, PVR continued to increase and the initial beneficial effect of mechanical ventilation on PVR and oxygenation decreased. With initiation of NO inhalation 2 h after commencing mechanical ventilation, PVR was again decreased and oxygenation improved. 181At 0.86 of term, initiation of mechanical ventilation caused pulmonary vasodilation, which was further increased by NO inhalation but not by 100% oxygen. Near term (0.96), NO inhalation and 100% oxygen administration both further increased the pulmonary vasodilation caused by mechanical ventilation. 182 

Selective pulmonary vasodilation during NO inhalation has been shown in numerous animal models: after pharmacologic preconstriction of the pulmonary vasculature with U46619, a synthetic thromboxane analog 4; after a heparin–protamine reaction that induces thromboxane-mediated pulmonary vasoconstriction 183; after pulmonary oleic acid instillation, which induces endothelial and alveolar edema, cell necrosis, and PAH 184,185; and after bilateral lung lavage, which depletes surfactant. 186Inhaled NO is also an effective pulmonary vasodilator in endotoxin-induced PAH 187–189and after smoke inhalation injury. 190 

Reactive oxygen species (e.g. , H²O², O²⁻) promote lung injury in various clinical settings. 191–194The use of high inspired concentrations of oxygen is sometimes necessary during the treatment of acute lung injury, but these high oxygen concentrations may cause or worsen lung injury. 195,196Thus, the effects of NO inhalation during lung injury caused by neutrophil oxidants or oxygen are of particular interest. Reaction of NO with oxygen or O²⁻results in NO²or ⁻OONO formation, which may damage the lung. 33,34,96In contrast, NO may also protect against the cellular toxicity of H²O², alkyl peroxides, and O²⁻. 197 

Data from isolated lung studies suggest that the net physiologic effects of NO inhalation protects against tissue injury by typical neutrophil oxidants. In buffer-perfused isolated rabbit lungs, NO inhalation decreased H²O²-induced pulmonary edema, pulmonary vascular permeability, 198and edema formation after injury with O²⁻(generated by the reaction of purine with xanthine oxidase). 199In hyperoxic lung injury, low-dose (10 ppm) but not high-dose (100 ppm) inhaled NO completely prevented pulmonary edema caused by inhalation of 100% oxygen for 40 h (fig. 3). 200In rats exposed for 60 h to 95% oxygen, pulmonary endothelial permeability, protein transfer, and type I alveolar epithelial cell injury were attenuated by inhalation of 20 ppm NO. 201,In vitro , however, using isolated microvascular endothelial cells and alveolar epithelial cells, simultaneous exposure to a NO donor drug and hyperoxia was associated with earlier cell death, as compared to hyperoxia alone. These in vitro  findings have been related to an increased production of ⁻OONO. 202 

Migration and adherence of neutrophils to the pulmonary vasculature and the local release of their oxidants are believed to be key events in oxidant lung injury. 203,204The effects of inhaled NO on neutrophil activity and neutrophil–endothelium interactions, in addition to direct effects of their oxidizing products, have been studied extensively. NO inhalation reduced pulmonary neutrophil accumulation after intestinal ischemia–reperfusion injury in rats 205and after dialysis in pigs. 206In an isolated rat lung model perfused with a mixture of human neutrophils and either N-formyl-L-methionyl-L-leucyl-L-phenylalanine 207or interleukin 1, 208inhalation of 50 ppm NO markedly decreased lung edema formation, neutrophil accumulation and neutrophil migration from the vascular into the alveolar space. Similarly, inhaled NO reduced pulmonary leukocyte sequestration in premature lambs with severe respiratory distress. 209In an in vivo  porcine model of Pseudomonas aeroginosa  sepsis, inhalation of NO (20 ppm) for 5 h after bacterial infusion had significant beneficial effects on pulmonary neutrophil sequestration and neutrophil oxidant activity. Inhaled NO reduced protein and neutrophil sequestration into the alveolar space. Neutrophil oxidant activity (stimulated O²⁻production) and alveolar structural damage (assessed by electron microscopy) were also reduced in these septic lungs treated with inhaled NO. 210However, opposing effects also have been reported. Increased oxidant activity (production of O²⁻and ⁻OONO) from intraalveolar neutrophils and increased protein sequestration into the alveolar space after NO inhalation have been observed. 211In a rat model of intratracheally administered endotoxin, inhaled NO (15 ppm) failed to prevent neutrophil sequestration and activation when inhalation was commenced 8 h after endotoxin challenge. 212One important difference in these studies was the time point when NO inhalation was begun; early NO inhalation (with respect to bacterial or endotoxin challenge) appeared to be associated with more effective inhibition of neutrophil activation and sequestration than did NO inhalation begun at a later time.

The effects of inhaled NO on pulmonary neutrophil sequestration may be mediated by modification of adhesion molecule expression and inhibition of the adherence of stimulated neutrophils to the endothelium and their migration through endothelial cell layers. In vitro  studies report that expression of a variety of endothelial and neutrophil adhesion molecules in response to inflammatory stimuli or ischemia–reperfusion injury is modified by molecular NO, NO donor drugs, or the inhibition of endogenous NO synthesis. 213–217NO also scavenges O²⁻released from migrating neutrophils and thereby reduces neutrophil oxidant activity after adherence. 218,219Recent in vitro  data showed a dose-dependent effect of NO on neutrophils: exposure of isolated human neutrophils to an environment of 80% oxygen and 20 ppm NO increased cell death by DNA inhibition, whereas 5 ppm NO did not induce significant DNA fragmentation. 220Whether this apoptotic effect is significant in vivo  is unknown.

Severe (hypoxic) pulmonary vasoconstriction and hypertension characterize high-altitude pulmonary edema (HAPE). Scherrer et al.  140hypothesized that inhalation of NO would reduce MPAP and thus the severity of HAPE. NO (40 ppm), inhaled at high altitude (4,559 m), decreased MPAP both in subjects prone to HAPE and those with HAPE, but not in HAPE-resistant control subjects. In subjects with HAPE, perfusion scintigraphy showed that inhaled NO redistributed pulmonary blood flow from edematous to nonedematous lung regions. This was associated with an improved Pa^O2.

Prolonged exposure to hypoxia induces PAH, pulmonary vascular wall remodeling with neomuscularization, and right ventricular hypertrophy. 221,222Inhaled NO (10–20 ppm), added while breathing at FI^O20.1 for 2–3 weeks, effectively prevented PAH, pulmonary vascular remodeling, and right ventricular hypertrophy in adult rats, 223,224in newborn rats, 225and in wild-type and eNOS-deficient mice. 129These salutary effects of inhaled NO may be mediated by direct vasodilatory mechanisms in the pulmonary vasculature and antiproliferative effects of NO on smooth muscle cells. 226 

Because inhaled NO can reduce reactivity and adhesion of circulating blood cells (e.g. , leukocytes, thrombocytes), it has been hypothesized that thrombus formation may be decreased by inhaled NO. In a rat model of collagen-induced pulmonary thrombosis, inhalation of 80 ppm NO reduced the MPAP increase associated with collagen injection and inhibited ex vivo  collagen-induced platelet aggregation. Rats treated with inhaled NO showed fewer platelet thrombi in small pulmonary vessels and a higher residual circulating platelet count. 11In an in vivo  porcine model of microsphere-induced pulmonary embolism, inhaled NO (5–80 ppm) reduced the increase of MPAP and increased the end-tidal carbon dioxide concentration. Platelet aggregation was increased with pulmonary embolism in control animals. Inhaled NO decreased the initial and maximum platelet aggregation. 227 

A beneficial effect of NO inhalation on ischemia–reperfusion injury, graft function, PAH, and oxygenation after lung transplantation has been reported in various experimental studies. 228–232The shortage of suitable donor lungs allows only a small percentage of potential recipients to receive a lung transplant. 233,234It has been suggested to harvest donor lungs from non–heart-beating donors to increase the number of lungs available for transplantation. 235,236Bacha et al.  studied whether NO inhalation can improve the function of lungs harvested after cardiac arrest in the donor. In pig and rat models, they treated the donor (after cardiac arrest) and recipient with inhaled NO (30 ppm) and demonstrated a significant improvement of oxygenation and short-term graft survival after transplantation, as well as reduction of PAH and decreased pulmonary neutrophil accumulation. 237,238 

Nitric oxide stimulates cGMP formation in platelets and, thus, NO inhalation may inhibit platelet function and augment a bleeding tendency in some species. Högman et al.  239reported that the bleeding time increased after rabbits inhaled NO. They noted that breathing 30 ppm NO for 15 min increased the bleeding time from 51 ± 5 to 72 ± 7 s (mean ± SE), and breathing 300 ppm NO for 15 min increased the bleeding time from 48 ± 12 to 78 ± 17 s. However, bleeding time was not altered in rats breathing 80 ppm NO for 1 h 240or in dogs breathing 20–200 ppm NO for 45 min. 10The reason for these differences among species is unknown.

Inhaled NO may affect the systemic vasculature and cells circulating within the systemic circulation, possibly by reversibly binding to hemoglobin 57or other proteins, with subsequent transport and release of NO at distant sites, or by modification of leukocytes and platelets during their transit through the lung. In a rat model of carotid injury, which is associated with migration and proliferation of smooth muscle cells in the arterial intima, inhalation of 80 ppm for 2 weeks decreased the degree of neointimal formation. 240In a canine model of coronary artery thrombosis and lysis, NO inhalation decreased the cyclic flow variation frequency and increased the duration of periods of coronary artery patency. In the latter model, inhaled NO had no vasodilatory effects on the pharmacologically preconstricted coronary artery segment and did not change the bleeding time. Therefore, coronary patency after thrombolysis was increased by NO inhalation, independent of direct vasodilatory activity or increased bleeding time. 10In contrast to pulmonary vasodilation, these actions of inhaled NO cannot be mediated by any direct effects of gaseous NO on smooth muscle cell relaxation.

Further insights into the mechanisms of the systemic effects of NO inhalation were recently provided by Fox-Robichaub et al.  9In a cat model of intestinal ischemia and reperfusion, inhalation of 80 ppm attenuated the reduction of perfusion, increase of leukocyte rolling, adhesion and migration, and endothelial dysfunction. Changes in leukocyte activity and vessel size were directly visualized by in vivo  microscopy and were induced by inhalation of 80 ppm but not 20 ppm NO. These effects were independent of intrapulmonary modification of leukocyte adhesion molecules, suggesting that inhaled NO was bound to transport molecules and was released in the peripheral circulation. 9 

Homozygous sickle cell anemia is a genetic disease characterized by severe hemolytic anemia, frequent vasoocclusive events, and a reduced life expectancy. A single amino acid substitution from valine to glutamic acid of the hemoglobin β chain results in hemoglobin S (HbS) formation. At deoxygenation, an erythrocyte containing HbS changes its shape from a biconcave disk to a crescent sickle cell because of intracellular hemoglobin polymerization. Sickle cells can occlude the microcirculation. 241–243 

Hemoglobin S has a markedly decreased oxygen affinity, measured as a markedly increased P⁵⁰(partial pressure of oxygen at half saturation of hemoglobin), compared with adult hemoglobin. In studies by Head et al. , 12inhalation of 80 ppm NO for 45 min by patients with homozygous sickle cell disease shifted the oxygen dissociation curve of their erythrocytes 4.6 ± 2.0 mmHg to the left, significantly decreasing the P⁵⁰(fig. 4). Methemoglobin concentrations did not increase substantially. In five of seven volunteers with sickle cell disease, the effect persisted for at least 60 min after discontinuing NO. 12In normal volunteers, the P⁵⁰was not affected by breathing NO. 12Precisely how inhaled NO alters sickle hemoglobin is unknown. One hypothesis is that the Cysβ93 residue of HbS is modified by NO, increasing HbS solubility and decreasing the tendency to polymerize during deoxygenation. 244,245 

The first clinical studies of inhaled NO focused on whether the physiologic effects measured in animal models were reproducible in patients. Acute respiratory distress syndrome (ARDS) and persistent PPHN have been the most commonly studied clinical syndromes. The results of large multicenter studies of NO inhalation in the treatment of critically ill newborns and adults recently have been reported. Inhaled NO also has been tested clinically in other conditions, including chronic PAH, chronic obstructive pulmonary disease (COPD), and lung transplantation and heart surgery.

Persistent pulmonary hypertension of the newborn is a clinical syndrome characterized by sustained pulmonary hypertension and severe hypoxemia, resulting in cyanosis unresponsive to oxygen therapy. Persistent pulmonary hypertension of the newborn may be caused by a variety of etiologies (e.g. , aspiration) or can be idiopathic. 246Diagnostic confirmation of PPHN includes echocardiographic observation of a right-to-left shunt through the ductus arteriosus or foramen ovale, caused by increased PVR, in the absence of congenital heart disease. Conventional treatment strategies include breathing high inspired concentrations of oxygen, hyperventilation, and infusion of bicarbonate to produce alkalosis, inhalation treatments with bovine surfactant, and intravenous vasodilator therapy. ECMO may be used to treat hypoxemia. However, the anticoagulation and cannulation of large vessels required for ECMO is associated with hemorrhagic complications.

In 1992, Roberts et al.  247and Kinsella et al.  248reported that 80 ppm 247or 6–20 ppm 248inhaled NO improved oxygenation in patients with PPHN. Several large controlled, randomized multicenter trials of the effects of inhaled NO in near-term and term hypoxic newborn patients were reported in 1997 (tables 1 and 2). 6–8,249,250In the majority of patients with PPHN and hypoxic respiratory failure (in whom the decision to initiate ECMO were made by the clinical team on the basis of center-specific ECMO entry criteria and without knowledge of assignment of the patient to the treatment group or placebo), NO improved oxygenation and decreased the requirement for ECMO. Kinsella et al.  reported that NO inhalation and high-frequency oscillatory ventilation were an effective combination that may increase the rate of responsiveness to inhaled NO. 8In a 1- to 2-yr follow-up study of children who received inhaled NO treatment for PPHN, neurodevelopment scores, growth rates (growth percentiles for weight, length, and occipitofrontal circumference), the frequency of airway disease, and the need for supplemental oxygen were comparable to conventionally ventilated or ECMO-treated patients. 251In summary, NO improved oxygenation in many newborns and, although it did not change overall survival, it reduced the need for ECMO (P < 0.05). 6,7Inhaled NO therapy did not appear to impart any benefits, however, to newborns with congenital diaphragmatic hernia. 249 

Respiratory distress syndrome (RDS), or hyaline membrane disease, of the premature newborn is characterized by deficiency or dysfunction of surfactant and is often associated with acute PAH. 181After promising preliminary studies of inhaled NO in the premature newborn with RDS, 252,253Skimming et al.  254studied the effect of inhaled NO at 5 and 20 ppm in preterm neonates (without systemic hypotension or congenital malformations and mechanically ventilated at FI^O2> 0.5). They demonstrated that arterial oxygenation improved and systemic arterial blood pressure was unaffected during a 15 min NO inhalation trial. The conclusions of this study were limited, however, because MPAP was not measured, and only 7% of the initially evaluated premature infants were included in the study.

In severe ARDS, PAH augments pulmonary edema and may impede right ventricular function and decrease cardiac output. Rossaint et al.  144demonstrated in patients with severe ARDS that inhaled NO produced selective pulmonary vasodilation. This was later confirmed by larger studies. 142,147Occasionally, the NO-induced pulmonary vasodilation has been associated with improved right ventricular performance, as indicated by improvements in right ventricular ejection fraction and decreased right ventricular end-diastolic and end-systolic volumes. 255In children with ARDS, inhaled NO (20 ppm) decreased MPAP by 25% and increased cardiac index by 14%. 256Inhaled NO also effectively decreased MPAP associated with the use of permissive hypercapnia in patients with ARDS. 257 

Inhaled NO (40 ppm) has been reported to decrease pulmonary capillary pressure 258and pulmonary transvascular albumin flux, 259partly caused by its effect on venous PVR in patients with acute lung injury. 258Such reductions of pulmonary venous and pulmonary capillary pressure should promote resolution of pulmonary edema, an important component of ARDS.

Severe hypoxemia caused by extensive intrapulmonary right-to-left shunting is characteristic of ARDS. Common current strategies of management include lung recruitment by high levels of positive end-expiratory pressure, prone positioning, and ventilation with a high FI^O2. Therapies that permit lower airway pressures and FI^O2might reduce the risk of barotrauma and oxidant injury to the lung. Inhaling 18 ppm NO for 40 min reduced the shunt fraction by 5% and increased the Pa^O2/FI^O2ratio by 30% in patients with ARDS. 144In a dose-ranging study of ARDS patients breathing NO, the ED⁵⁰(the dose producing 50% of maximal effect) for increasing Pa^O2(10–100 ppb) was markedly less than the ED⁵⁰producing pulmonary vasodilation (1–10 ppm, fig. 5). 260 

In a phase 2 multicenter trial, the effects of NO inhalation on oxygenation were studied in 177 patients who met the criteria of ARDS (Pa^O2/FI^O2< 200 mmHg within the last 72 h, bilateral chest infiltrates, pulmonary capillary wedge pressure (PCWP) < 18 mmHg, positive end-expiratory pressure requirement > 8 cm H²O and FI^O2requirement > 0.5). Sixty-five percent of the patients who received inhaled NO (pooled results of patients receiving 1.25, 5, 20, 40, 40, or 80 ppm NO) had a significant (P = 0.0002 vs.  placebo) improvement in Pa^O2(defined as a 20% increase of Pa^O2after 4 h of therapy). Only 24% of the patients receiving placebo (nitrogen) responded similarly (fig. 6). 142The improved oxygenation induced by NO allowed physicians to reduce FI^O2and positive end-expiratory pressure and thereby decreased the oxygenation index (FI^O2× mean airway pressure × 100)/Pa^O2) for the first 4 days of therapy. The MPAP was slightly lower in the inhaled NO group, compared with placebo, for 2 days. Similar transient improvements of Pa^O2/FI^O2during inhaled NO therapy in patients with ARDS have been shown in prospective studies conducted by Michael et al.  261and Troncy et al.  262The reasons for such transient effects remain unclear.

The phase 2 U.S. multicenter study of the effects of inhaled NO on ARDS patients (discussed previously) reported a mortality rate of 30% in NO-(all doses pooled) and placebo-treated patients both. 142The patients were assigned randomly to receive 1.25, 5, 20, 40, or 80 ppm NO or nitrogen placebo. Subgroups receiving the same dose of NO for the treatment period consisted of 8–34 patients. Mortality rates in the subgroups were 32%(7 of 22 patients) in the 1.25 ppm group, 24%(8 of 24 patients) in the 5 ppm group, 31%(9 of 29 patients) in the 20 ppm group, 30%(8 of 27 patients) in the 40 ppm group, and 38%(3 of 8 patients) in the 80 ppm group. A prospective randomized study reported by Troncy et al.  262similarly found no significant difference between ARDS patients receiving inhaled NO and controls with regard to 30-day mortality or days of mechanical ventilation. Several hypotheses should be addressed in the interpretation of these data and should be kept in mind for the design of future studies:

1. The beneficial effects of NO inhalation (e.g. , improvement of gas exchange) may not alter the overall outcome because the survival of patients with ARDS may not be primarily dependent on gas exchange. The majority of patients dying with ARDS also have severe sepsis or multiple organ failure. The incidence of death primarily because of respiratory failure varies among studies. In a study by Montgomery et al.  published in 1985, the percentage of ARDS deaths specifically because of hypoxemia and respiratory failure was reported to be 16%. 263A study of patients treated at the LDS Hospital in Salt Lake City reported that 40% of deaths in ARDS patients were caused by respiratory failure. 264It is unknown whether this subgroup of patients would benefit from inhaled NO because such severely hypoxemic patients have been excluded from prospective studies.

2. The beneficial effects of NO inhalation are offset by its toxic effects. As described previously, NO has several negative effects on biologic molecules and tissues, especially if ⁻OONO is formed. This hypothesis needs further testing in patients, e.g. , by analyzing lung specimens for evidence of NO or ⁻OONO toxicity after NO inhalation.

3. NO inhalation has a narrow therapeutic range. Small doses, e.g. , 1–5 ppm, could be effective and improve survival, but smaller doses might be not effective, and larger doses could be toxic. The effective dose may change over time and vary among different patients and different disease states. 265It is necessary to establish improved dosing criteria. Larger patient groups receiving similar doses are necessary to discover statistically robust differences. Currently, a blinded, multicenter phase 3 study is being completed that investigated the effects of 5 ppm inhaled NO, compared with placebo, in patients with ARDS. Such a protocol assumes that the dose–response relation of inhaled NO is similar among ARDS patients and over time. This may not be true. 265 

Performing such trials is difficult and expensive. The incidence of ARDS is relatively low, and the precipitating events are often multifactorial. Usually a large number of centers must participate to recruit sufficient numbers of patients. There may be significant differences in response rates and outcomes among different centers, as reported for PPHN patients by Kinsella et al.  8Different treatment strategies and the experiences of individual institutions and caregivers may provide confounding variables. As a result, conclusive studies evaluating the effects of a drug such as inhaled NO, which at best has a modest effect on the survival of a diverse population of patients with ARDS, may not be economically viable.

The pathophysiology of chronic PAH includes a partially reversible increase of MPAP in the early stages of the disease, leading to a nonreactive and irreversibly remodeled pulmonary vasculature after long-standing PAH. 266Vascular remodeling is characterized by muscularization of previously nonmuscular small resistance arteries, medial hypertrophy of proximal pulmonary arteries, and a reduced number of arteries within the lung. Diagnosis and medical treatment of chronic PAH relies on vasodilator therapy. When medical treatment is no longer possible (e.g. , no vasodilator response) or ineffective (e.g. , tachyphylaxis or tolerance), lung transplantation remains as the last option to prolong life.

The determination of pulmonary vascular responsiveness is essential for prognosis and long-term treatment. Drugs commonly used to assess pulmonary vasodilatory responses include intravenous prostacyclin (PGI²), adenosine, and calcium channel blockers. 267After studying adult patients with inhaled NO and infused prostacyclin, Sitbon et al.  suggested using inhaled NO as the “gold-standard” to assess pulmonary vasoreactivity 268because inhaled NO was reported to selectively reduce MPAP in patients with pulmonary hypertension. 5A recent survey of long-term vasodilator treatment in approximately 800 patients with primary pulmonary hypertension reported that inhaled NO was used as the primary vasodilator to test pulmonary vascular responsiveness by 32% of the participating U.S. tertiary hospitals. 269 

As opposed to commonly used intravenous vasodilators, which can produce systemic hypotension, inhaled NO does not significantly affect systemic vascular resistance. This permits the rapid and safe evaluation of changes of biventricular function during a brief trial of pulmonary vasodilation and, therefore, provides an important diagnostic tool for the decision to begin medical treatment or to plan lung transplantation or combined heart and lung transplantation. 270 

In chronic PAH, a positive response to a short-term vasodilator trial usually results in long-term drug treatment that may include a wide spectrum of systemic vasodilator drugs (e.g. , acetylcholine, α-adrenergic agonists, direct-acting vasodilators, angiotensin-converting enzyme inhibitors, calcium channel blockers, prostaglandins) and permanent anticoagulation. 267The effectiveness of current vasodilator treatment often is limited by systemic hypotension. Channick et al.  271tested an ambulatory NO delivery system consisting of an 80-ppm NO tank, a gas–pulsing device, and a nasal cannula in eight PAH patients with a pulmonary artery catheter. They reported that this technique produced effective pulmonary vasodilation without evidence of significant nitrogen dioxide formation. One of the eight patients was discharged from the hospital and treated for 9 months with inhaled NO without any apparent adverse events. Long-term domiciliary NO inhalation as an alternative or a bridge to lung transplantation 272requires investigation in larger patient groups, after consideration of the beneficial and toxic effects caused by long-term NO inhalation.

Inhaled NO has been tested for use as a pulmonary vasodilator in COPD and as a bronchodilator in COPD and asthma. COPD is characterized by irreversible airway obstruction and is associated with irregular enlargement of alveoli and destruction of alveolar walls after chronic inflammation. Hypoxia produces pulmonary vasoconstriction, resulting in chronic PAH and right ventricular hypertrophy. Important characteristics of asthma include inflamed, hyperreactive airways with reversible bronchoconstriction.

Inhaled NO has been reported to be a bronchodilator in many experimental animal models. Högman et al.  169evaluated inhalation of 80 ppm NO in healthy volunteers, in patients with hyperreactive airways, bronchial asthma, and COPD. Inhaled NO caused mild bronchodilation in patients with asthma but not in patients with COPD. In other studies, however, the bronchodilator action of NO has been reported to be much weaker than commonly used inhaled β²-adrengeric agonists. 168,273,274 

Hypoxemia in COPD is primarily caused by a mismatch and not by intrapulmonary right-to-left shunting (as in ARDS). Hypoxic vasoconstriction augments blood flow to better ventilated regions and improves oxygenation. Inhaled NO may oppose this physiologically useful mechanism by vasodilating poorly ventilated areas in the obstructed lung, and thus increasing blood flow to these areas, as reported by Hopkins et al.  155Indeed, transcutaneous arterial oxygen tension, 275Pa^O2, and V̇/Q̇ distribution 276were worsened by NO inhalation in air-breathing COPD patients. However, when NO was used in combination with modest oxygen enrichment, 277Pa^O2was improved to a greater extent than with oxygen therapy alone. The combination also more effectively decreased MPAP. Thus, the combined use of supplemental oxygen and inhaled NO (e.g. , via  an ambulatory inhalation device) may offer a valuable therapeutic strategy for improving oxygenation and providing pulmonary vasodilation in selected COPD patients.

Pulmonary artery hypertension frequently occurs in the immediate postoperative period after lung transplantation and has been effectively treated with inhaled NO. 278Inhaled NO has been reported to be effective in the treatment of post–lung transplant pulmonary dysfunction. In a retrospective study by Date et al. , 279243 patients undergoing lung transplantation over 6 yr were analyzed. Thirty-two patients had immediate severe graft dysfunction, as indicated by a Pa^O2/FI^O2ratio < 150 mmHg. Comparing patients in whom NO treatment was not available with patients in whom NO treatment was begun after graft dysfunction was diagnosed, inhaled NO reduced MPAP and increased the Pa^O2/FI^O2ratio within the first hour of treatment. The requirement for ECMO was similar in both patient groups. The rate of airway complications and hospital mortality (7% in NO group vs.  24% in control group) was markedly reduced in patients receiving NO therapy. As in all retrospective studies using historical controls, changes of treatment strategies and increased clinical experience over time must be considered. Although this study suggests that inhaled NO may decrease post–lung transplant organ dysfunction, it should be confirmed in prospective controlled studies.

The degree and reversibility of the increased PVR determine the various treatment options and outcome in children with congenital heart disease and PAH. 280Roberts et al.  281demonstrated that inhaled NO (80 ppm for 10 min) decreased MPAP without causing systemic vasodilation in children between 3 months and 7 yr of age with congenital cardiac lesions (e.g. , atrioseptal defect, ventricular septal defect, atrioventricular canal). The ability of inhaled NO to decrease the PVR of children with congenital cardiac defects has been confirmed by others. 282,283In preoperative patients with severe right-to-left shunting, inhaled NO increased pulmonary blood flow, decreased extrapulmonary shunt flow and improved oxygenation. 284Inhaled NO, therefore, might provide a therapeutically useful tool for the acute nonsurgical treatment of these patients.

Transient PAH is common after repair of congenital cardiac lesions and has been related to damage to the pulmonary vascular endothelium, probably induced by the use of cardiopulmonary bypass. 285Inhaled NO has been reported to ameliorate the postoperative PAH of congenital heart disease 286–291and decrease the need for postoperative ECMO. 292 

Patients with left ventricular valvular disease may have preoperative PAH caused by an increased left atrial pressure with retrograde transmission of pressure into the pulmonary arterial circulation. Pulmonary vascular remodeling occurs as a result of chronic pulmonary venous hypertension and PAH. Pulmonary vascular remodeling and vasoconstriction may persist or slowly decrease with time after valve replacement. Treatment with NO after repair of the valvular disease might relieve the vasoconstrictor component of PAH in these patients.

Fullerton et al.  293reported that 20–40 ppm inhaled NO produced pulmonary vasodilation in patients after aortocoronary bypass. In patients after heart transplantation, inhalation of 20 ppm NO caused significant pulmonary vasodilation but also decreased systemic vascular resistance. 294The decreased SVR was most likely secondary to an improved cardiac output because systemic arterial pressure and PCWP remained unchanged. Intravenous administration of the phosphodiesterase inhibitor dipyridamole markedly augmented the pulmonary vasodilatory response to inhaled NO in some patients after aortic or mitral valve replacement. 295 

A considerable number of patients who receive inhaled NO therapy do not respond by either pulmonary vasodilation or improvement of systemic oxygenation. The reported rate of hyporesponders ranges from 30% to 45%, depending on the threshold value chosen to define hyporesponsiveness. 142,147Several hypotheses have been raised to explain the mechanisms of hyporesponsiveness. Manktelow et al.  147and Krafft et al.  296reported that ARDS patients with sepsis were less likely to respond to inhaled NO (60–70% of septic patients were hyporesponders). The presence of high levels of endogenously produced NO and the opposing pulmonary vasoconstrictor action of catecholamines used for the treatment of septic vasodilation have been suggested as possible reasons for the decreased response to inhaled NO. Holzmann et al.  297evaluated the effects of sepsis on NO responsiveness in an isolated rat lung model. They reported that hyporesponsiveness was associated with decreased pulmonary cGMP release (table 3), suggesting that signal transduction in the NO response pathway is downregulated in sepsis. This was attributed to increased phosphodiesterase activity and therefore increased cGMP breakdown. Bigatello et al.  reported that hyporesponsive patients with ARDS have a reduced accumulation of plasma cGMP during NO breathing. 298 

Increased vascular production of O²⁻, as observed in systemic nitrate tolerance, also may contribute to hyporesponsiveness to inhaled NO. Munzel et al.  reported that long-term nitrate treatment of rabbits resulted in increased O²⁻production by the aorta and hyporesponsiveness (tolerance) to acute nitroglycerin administration, related to an activated membrane-associated oxidase. 299,300Increased O²⁻production in the pulmonary vasculature may have similar effects.

Lastly, Weimann et al.  301recently reported a linkage between ABO blood group distribution and hyporesponsiveness to inhaled NO, demonstrating that ARDS patients with the major blood groups A or O had a larger increase of Pa^O2/FI^O2in response to NO inhalation than patients with blood groups B or AB. The underlying mechanism of these results is unknown, but they indicate that the pulmonary vascular response to inhaled NO may be determined or modified by genetic factors.

Because cGMP is hydrolyzed by phosphodiesterase, inhibition of phosphodiesterase may increase the effectiveness and duration of the action of inhaled NO. In awake lambs with U46619-induced PAH, intravenous infusion of zaprinast, a phosphodiesterase 5–specific inhibitor, increased the duration of action of inhaled NO and accentuated the NO-induced reduction of PVR. With administration of zaprinast, the half-time of pulmonary vasodilation after discontinuing a 4-min 40-ppm NO inhalation trial was increased from 1 or 2 min to 10–12 min (fig. 7). 302Potentiation of the effects of NO by another phosphodiesterase inhibitor, dipyridamole, has been reported in the ovine fetal pulmonary circulation. 303,304 

No experimental or clinical studies have yet determined whether inhibition of pulmonary vascular O²⁻production may increase responsiveness to inhaled NO. Inhibition of the O²⁻-generating membrane-associated oxidase in rabbit aorta by hydralazine normalized vascular O²⁻production and restored the vasodilatory activity of nitroglycerin. 299,300 

Almitrine bismesylate acts as an agonist at peripheral arterial chemoreceptors and increases discharge of the carotid sinus nerve. 305Intravenous administration of low concentrations of almitrine has been reported to restore or enhance hypoxic vasoconstriction in the acutely injured lung. Inequalities and the Pa^O2in ARDS patients usually are improved by almitrine administration, which augments hypoxic vasoconstriction and redistributes pulmonary arterial blood flow toward better ventilated areas with a higher PA^O2. The combination of almitrine infusion and NO inhalation, each having different mechanism, might synergistically reduce mismatch. Almitrine administration enhanced the oxygenation beneficial effects of inhaled NO on oxygenation 306and allowed a further reduction of FI^O2in patients with ARDS. 307The relatively long in vivo  half-life (mean tissue half-life, 12 h) and possible toxic side effects of almitrine (including a slowly reversible peripheral sensory neuropathy) require careful investigation before this regimen can be recommended for routine clinical use.

Perfluorocarbons are inert liquids that lower surface tension in surfactant-depleted lungs and dissolve large concentrations of respiratory gases. Zobel et al.  308demonstrated that inhaled NO enhanced the effects of perfluorocarbons on pulmonary gas exchange in a piglet model of acute lung injury induced by repeated bilateral lung lavage. An additive effect of inhaled NO and perfluorocarbons on pulmonary gas exchange in acute lung injury has been confirmed by others. 309,310 

The risks of inhaled NO should be carefully considered in patients with markedly impaired left ventricular function (e.g. , heart transplant candidates). Inhalation of NO may vasodilate the pulmonary circulation and increase blood flow to the left ventricle, thereby acutely increasing left atrial pressure and PCWP 311–313and promoting pulmonary edema formation. 314Cardiac output, left atrial pressure, or PCWP should be monitored if NO is administered to patients with severe left ventricular dysfunction.

Rebound PAH, an increase of intrapulmonary right-to-left shunting and a decreased Pa^O2after acute NO discontinuation is well-described. 144Lavoie et al.  315reported four patients with severe acute respiratory failure in whom NO therapy was discontinued abruptly after a decision to discontinue life-extending measures. A sustained decrease of arterial oxygen saturation occurred immediately after NO discontinuation, which was reversed by restarting NO therapy. All four patients died within 24 h after NO discontinuation.

It has been suggested that downregulation of endogenous NO synthesis by NO inhalation is responsible for rebound PAH. 130,316–318However, recent data obtained in rats with hypoxic pulmonary hypertension suggest that inhibition of endogenous NO synthesis plays a minor role in rebound PAH: no changes of lung eNOS protein levels, NOS activity, endothelium-dependent and -independent vasodilation were reported after 3 weeks inhaling 20 ppm NO. Lung GC activity was transiently decreased after 1 week of NO inhalation, but GC activity was normal after 3 weeks of NO inhalation. 319 

To avoid rebound PAH, a slow stepwise reduction of the inhaled NO concentration with immediate control of any adverse effects (e.g. , reduced oxygen saturation or blood gas tensions, increased MPAP 144) are important to safely wean the patient from inhaled NO. In addition, administration of the phosphodiesterase 5 inhibitor dipyridamole has been reported to prevent rebound PAH in children after cardiac surgery. 320 

Inhaled NO can inhibit platelet function and increased the bleeding time in rabbits. In ARDS patients receiving inhaled NO (3–100 ppm), NO decreased platelet aggregation and agglutination in vitro . However, the in vivo  bleeding time (Ivy bleeding time) was not altered. 321In neonates, a recent study reported that bleeding time doubled after 30 min of 40-ppm NO inhalation. 322However, in the multicenter studies in newborns, no difference in the frequency of bleeding events was observed in NO-treated compared with placebo-treated patients. 6–8,249 

Nitric oxide inhalation therapy should be instituted after careful consideration of potential acute and long-term toxicity. The major concerns are (1) methemoglobinemia, (2) NO²formation, and (3) cellular toxicity. Acute inhaled NO overdose (> 500–1,000 ppm) leads to rapid NO²formation, severe methemoglobinemia, pulmonary alveolar edema and hemorrhage, hypoxemia, and death within minutes to hours. 323 

Blood methemoglobin concentrations and inspired NO²concentrations have been regularly monitored in clinical trials of inhaled NO in adults and neonates. 6,8,142,249In the large number of patients studied in these trials (n = 471) receiving inhaled NO therapy at doses ranging from 1.25 to 80 ppm, significant methemoglobinemia or NO²formation was uncommon (table 4). If methemoglobin or NO²levels increased above predetermined limits, the inhaled NO concentration was decreased. Discontinuation of NO administration because of NO²or methemoglobin formation was only necessary in 3 of 471 patients (0.6%).

The most important requirements for safe NO inhalation therapy are (1) continuous analysis of NO and NO²concentrations (using chemiluminescence or electrochemical analyzers 324,325);(2) frequent calibration of the monitoring equipment;(3) frequent analysis of blood methemoglobin levels;(4) the use of certified tanks; and (5) administration of the lowest NO concentration required.

Little is known about the long-term sequalae of NO inhalation in humans. In 12 newborns receiving NO inhalation treatment (< 20 ppm) for up to 4 days, there were no signs of increased lipid peroxidation product, impaired surfactant activity, or changed cytokine profile. 326However, in two infants requiring prolonged ventilation with NO, nitrotyrosine residues were detected in airway specimens. 326The relative contribution of NO inhalation and endogenous NO formation to nitrotyrosine formation in the lung is unclear because nitrotyrosine formation has been demonstrated in acutely injured lungs without the exogenous administration of NO. 33,34Follow-up studies of adult patients 8 months after NO treatment for ARDS showed no obvious differences in pulmonary function compared to ARDS patients not treated with NO. 327 

In summary, reported data of clinical NO toxicity are sparse. Studies appropriately designed to detect long-lasting or irreversible pathologic effects of NO breathing will be necessary to predict the long-term effects of inhaled NO and establish time and dose limits.

Prostacyclin is a natural product of the cyclooxygenase pathway and a potent vasodilator. The initial clinical studies of NO inhalation in ARDS patients compared the pulmonary vasoactive properties of intravenously administered prostacyclin with inhaled NO. 144Although prostacyclin infusion decreased MPAP to the same degree as inhaled NO, prostacyclin decreased Pa^O2, presumably by reducing hypoxic pulmonary vasoconstriction and caused systemic hypotension. Such adverse effects are commonly observered during intravenous infusion of commonly used vasodilator drugs.

It was hypothesized that the administration of prostacyclin via  aerosol would limit its hemodynamic effects to the lung. This hypothesis was tested in ARDS patients, and the effectiveness of prostacyclin when used as a short-term inhaled aerosol was compared with inhaled NO. 328Similar effects, namely decreased MPAP, decreased intrapulmonary right-to-left shunting, and increased Pa^O2/FI^O2were observed. Systemic vasodilation was not reported with either drug. These effects of inhaled prostacyclin have been confirmed in other studies. 329,330A disadvantage of aerosolized prostacyclin therapy is that systemic absorption can occur, which makes it difficult to maintain pulmonary vasodilation without producing systemic vasodilation for periods lasting more than a few hours. Larger studies, including randomized trials studying the effectiveness and responsiveness to inhaled prostacyclin for longer time periods, are necessary to support a useful therapeutic role.

The use of inhaled NO donor drugs has been proposed as an alternative to NO inhalation. Administration of such drugs, which release a defined amount of NO over a prolonged time period, might permit intermittent NO dosing. Adrie et al.  compared the pulmonary vascular effects of inhaled sodium 1-(N,N-diethylamino) diazen-1-ium-1,2-diolate (DEA/NO), which spontaneously generates NO, with the inhalation of sodium nitroprusside and NO gas in awake sheep with pharmacologically induced pulmonary hypertension. 331DEA/NO caused nonselective vasodilation, and sodium nitroprusside was only selective for the pulmonary circulation at low inhaled concentrations, compared with the highly selective effect of NO gas. 331In a pig model of acute pulmonary hypertension, Brilli et al.  compared the effects of the aerosolized NO donors ethylputreanine NONOate (EP/NO) and 2-(dimethylamino) ethylputreanine NONOate (DMAEP/NO). 332They reported that 3-min DMAEP/NO aerosolization caused selective pulmonary vasodilation, which lasted for approximately 30–50 min without effects on systemic arterial pressure or cardiac output. 332Similarly, tracheal instillation of DMAEP/NO resulted in prolonged and pulmonary selective vasodilation. 333EP/NO, aerosolized or instilled, was less effective and the effects were inconsistent. 332,333 

The intravenous infusion of ultra–short-acting NO donor agents might also be an alternative to inhaled NO for producing selective pulmonary vasodilation. In awake, healthy sheep with pulmonary hypertension, the intravenous (systemic) infusion of PROLI/NO (C⁵H⁷N³O⁴Na²· CH³OH), an ultra–short-acting nucleophile/NO adduct that generates NO, produced selective pulmonary vasodilation without affecting the systemic circulation. The selective effect of this intravenous drug was caused by its short half-life in vivo , which resulted in complete NO release during transit of the pulmonary circulation and before reaching the systemic arterial circulation. 334Because such drugs are intravenously administered, oxygenation might be adversely affected in the injured lung through indiscriminant release of hypoxic pulmonary vasoconstriction. Nevertheless, such newly designed NO donor drugs appear promising as a selective pulmonary vasodilator and provide an alternative to NO inhalation.

The use of inhaled phosphodiesterase inhibitors has been investigated as an alternative or adjunct to NO inhalation. Inhalation of nebulized zaprinast induced selective pulmonary vasodilation and enhanced the effects of inhaled NO in awake lambs. 335However, at a zaprinast concentration (50 mg/ml) producing a similar decrease of MPAP to 20 ppm NO, significant systemic vasodilation was observed (fig. 8).

Nitric oxide is produced by almost every healthy mammalian tissue. In health, NO has a myriad of functions that are essential for life. In disease, NO has many effects that can be both helpful and deleterious.

Inhaled NO has made a rapid journey from the laboratory bench to the bedside because of its unique selective pulmonary vasodilator activity and its ready availability. It is the first vasodilator described to provide truly selective pulmonary vasodilation. A large number of basic and clinical research studies have made great steps in delineating its physiology, side effects, and clinical efficacy. Nearly simultaneously, the clinical use of inhaled NO has become widespread. During the past 8 yr, inhaled NO has been used to treat pulmonary hypertension and hypoxemia in thousands of patients worldwide. Inhaled NO is an effective pulmonary vasodilator in many disease states characterized by pulmonary hypertension. In addition, inhaled NO decreases pulmonary venous admixture in diffuse lung injury, and therefore increases systemic oxygenation in many patients. Most importantly, randomized multicenter studies of NO inhalation have shown that this new therapy significantly reduces the necessity for ECMO in newborns with PPHN or hypoxic respiratory failure. NO inhalation has many additional effects that may be clinically beneficial. Inhaled NO has been reported to decrease pulmonary edema formation and lung injury. In the systemic circulation, NO inhalation inhibits peripheral vascular restenosis in animal models of arterial injury, augments the oxygen binding of sickle erythrocytes, and reduces cyclic coronary occlusion in a model of coronary injury and thrombolysis. However, important questions remain:

1. Does the reduction of pulmonary artery pressure and increased Pa^O2caused by NO inhalation improve clinical outcome for patients with acute lung injury? In a relatively uniform and well-defined population of patients, newborns with hypoxic respiratory failure, NO inhalation effectively improves oxygenation and significantly reduces the use of ECMO. Avoiding ECMO, a complicated and expensive invasive procedure with limited availability, is an important clinical endpoint and undoubtedly would justify the use of inhaled NO. 6,7The clinical usefulness of inhaled NO in adults remains unclear. Clinical studies of ARDS in adults are complicated by the diverse nature of the patient population, the precipitating causes of lung injury, and the common occurrence of sepsis and multiple organ system failure. Current data from multicenter trials suggest that the mortality rate in moderate lung injury is not significantly changed by NO inhalation. Whether this is because of inappropriate study design, the complex nature and spectrum of ARDS, inefficacy of NO, inappropriate dosing, or counterbalancing toxic effects of NO is unknown.

2. If available data suggest that survival is unchanged, should clinicians continue to study NO inhalation? Researchers concentrating on the cellular and subcellular effects of NO properly express concerns that NO inhalation may worsen lung injury and damage important structures. Initial clinical studies, however, suggest that toxicity, if present, is extremely low. The doses of NO now commonly used are less than those received with cigarette exposure and are nearly within the atmospheric background range of many urban areas. Many clinical scientists continue to evaluate inhaled NO and find it useful for short-term symptomatic treatment of hypoxemic respiratory failure and pulmonary vasoconstriction.

The pharmacologic and toxicologic profiles of NO inhalation are incomplete. It is necessary to delineate further (1) proper indications, (2) contraindications, (3) sound dosing criteria, (4) organ disease from cellular and subcellular toxicity, and (5) the causes of NO hyporesponsiveness. Randomized clinical studies of patients with carefully defined specific disease states characterized by pulmonary hypertension or hypoxemia (e.g. , pulmonary embolism, severe PAH, postpneumonectomy pulmonary edema, acute rejection after lung transplantation) and in premature newborns with respiratory failure remain to be completed. If such trials are carefully designed and conducted, we may define additional groups of patients that may benefit from, or may be harmed by, inhaled NO. The use of inhaled NO continues to be a unique and fascinating approach to studying and treating diseases as diverse as acute rejection of the transplanted lung and sickle cell crisis. In evaluating this complex field, it is critical that our view does not become colored by a single study or effectiveness in a particular disease state. As with most medical advances, it is the evolution of a wide-ranging body of research that will properly determine the place for NO inhalation therapy in our armamentarium.

The authors thank Dr. Joerg Weimann (Department of Anesthesia and Critical Care, University of Heidelberg, Germany), Dr. Kenneth D. Bloch (Cardiovascular Research Center, Massachusetts General Hospital, Boston), and Dr. Steven Tannenbaum (Massachusetts Institute of Technology, Cambridge) for helpful suggestions and critical discussions.