Understanding images: correlation between computerized tomographic scans of lung structure with impaired function in ARDS.
Explore the source record for details and available documents.
Humans↗
Biomedical subjects
W M Zapol
Publications and source records attributed to W M Zapol.
Platelet function in acute respiratory failure.
To assess the role of platelets in thrombohemorrhagic complications of acute respiratory failure (ARF), we studied platelet function in 13 ARF patients admitted for intensive care, in six acutely ill intensive care patients without evidence of acute lung injury (non-ARF), and in 10 normal subjects. Platelet counts in ARF and non-ARF patients were similar to the normal range. The bleeding time of the ARF patients (8.5 +/- 0.9 min) was significantly longer (p less than 0.01) than the normal (4.8 +/- 0.2 min) but similar to non-ARF patients (5.4 +/- 0.8 min). The bleeding time prolongations in ARF patients were unrelated to platelet concentration. Platelet aggregation induced by ADP and thrombin was normal in both ARF and non-ARF patient groups. The epinephrine response was impaired in one non-ARF patient and in three ARF patients; collagen-induced aggregation was absent in two ARF patients, with a prolonged bleeding time. Levels of VIII:C and vWF in both groups of patients were similar to the normal level, but VIIIR:Ag levels in ARF patients (407 +/- 45% of normal) were higher (p less than 0.01) than in both non-ARF patients (210% +/- 10%) and normal subjects (106% +/- 4). The electrophoretic mobility of VIIIR:Ag was abnormal in ARF patients. The prolonged bleeding time in ARF patients appears to result from the qualitative and quantitative VIIIR:Ag defect. beta-Thromboglobulin levels were greater (p less than 0.01) in ARF patients (87.6 +/- 6.9 ng/ml; p less than 0.001) than in non-ARF patients (46.2 +/- 3.1 ng/ml) or in normal subjects (25.3 +/- 2.5 ng/ml p less than 0.0001). However, platelet factor 4 plasma levels in ARF patients (18 +/- 1.6 ng/ml) did not differ from those in non-ARF patients (15.0 +/- 3.0 ng/ml), but both were significantly different from normal (6.1 +/- 0.8 ng/ml). Plasma thromboxane B2 (T X B2) levels were not different from normal values in either ARF or non-ARF patients, but 6-keto-PGF1 alpha levels were significantly reduced (p less than 0.01) in ARF patients (215 +/- 43 pg/ml) compared to normal values (381 +/- 34 pg/ml). Non-ARF patients had 6-keto-PGF1 alpha levels (285 +/- 111 pg/ml) midway between the normal values and those of ARF patients. Our results suggest that in vivo platelet activation occurs in ARF. ARF patients have quantitative and qualitative platelet defects that may contribute to thrombotic and hemorrhagic complications.
6-Ketoprostaglandin F1 alpha↗
Dose-dependent effects of a pyridoquinazoline thromboxane synthetase inhibitor on arachidonic acid metabolites and hemodynamics during E. coli endotoxemia in anesthetized sheep.
We investigated the effects of a new pyridoquinazoline thromboxane synthetase inhibitor infused before administering Escherichia Coli endotoxin into 18 anesthetized sheep with lung lymph fistulas. In normal sheep increasing plasma Ro 23-3423 concentrations were associated with increased plasma levels of 6-keto-PGF1 alpha, a reduced systemic vascular resistance (SVR, r = -0.80) and systemic arterial pressure (SAP, r = -0.92), the mean SAP falling from 80 to 50 mm Hg at the 20 and 30 mg/kg doses. Endotoxin infused into normal sheep caused transient pulmonary vasoconstriction associated with increased TxB2 and 6-keto-PGF1 alpha levels while vasoconstriction and TxB2 increase were significantly inhibited by pretreatment with Ro 23-3423 in a dose-dependent manner. When compared to controls, plasma and lymph levels of 6-keto-PGF1 alpha, PGF2 alpha and PGE2 after endotoxin infusion were increased several-fold by administering Ro 23-3423 up to plasma levels of 10 micrograms/ml. Doses over 30 mg/kg with blood levels above 10 micrograms/ml reduced plasma and lymph levels of 6-keto-PGF1 alpha, PGF2 alpha and PGE2, suggesting cyclooxygenase blockade at this dose. The peak 6-keto-PGF1 alpha levels at 60 min after endotoxin infusion in sheep with Ro-23-3423 levels below 10 micrograms/ml were associated with the greatest systemic hypotension due to a reduced SVR (r = -0.86). After endotoxin infusion the leukotrienes B4, C4, D4 and E4 in lung lymph were assayed by radioimmunoassay and high pressure liquid chromatography and remained at baseline values.
6-Ketoprostaglandin F1 alpha↗
Diving adaptations of the Weddell seal.
Explore the source record for details and available documents.
Adaptation, Physiological↗
C5a and thromboxane generation associated with pulmonary vaso- and broncho-constriction during protamine reversal of heparin.
The authors conducted a study in humans to determine the mediators associated with acute pulmonary vaso- and broncho-constriction occurring episodically with protamine reversal of heparin anticoagulation. Of 48 adult patients investigated prospectively after termination of cardiopulmonary bypass, two presented a sudden increase of airway pressure, acute pulmonary hypertension, and systemic hypotension 1-3 min after right atrial protamine injection. In these two subjects, plasma levels of C5a increased from 0.7 and 2.2 to 9.8 and 9.9 ng/ml, respectively, and thromboxane B2 increased from 0.26 and 0.34 to 7.5 and 16.2 ng/ml 1 minute after drug injection. A third subject not identified prospectively had an identical reaction and mediator profile (C5a, 10.2 ng/ml; TxB2, 18.6 ng/ml at 1 min). The plasma levels of these mediators were unchanged in the remaining patients (C5a, 0.7 +/- 1.1 [x +/- S.D.] to 0.6 +/- 0.9 ng/ml; TxB2, 0.16 +/- 0.12 to 0.15 +/- 0.07 ng/ml). Plasma histamine was not involved in this type of reaction, but increased from 0.7-10.4 ng/ml in a fourth patient who became hypotensive without acute pulmonary hypertension, bronchoconstriction, or elevation of C5a or TxB2. The authors' data indicate that the generation of high plasma levels of C5a anaphylatoxins and thromboxane is associated with pulmonary vaso- and broncho-constriction induced by protamine reversal of heparin in humans.
Adult↗
Bedside measurement of pulmonary capillary pressure in patients with acute respiratory failure.
In this report, the authors present the results of 34 estimates of pulmonary capillary pressure (Pcap) in 15 adult patients receiving intensive care for acute respiratory failure (ARF). Within the pulmonary artery pressure profile during transient balloon occlusion, the authors identified two exponential pressure decay components-the slower one representing the discharge of the pulmonary capillary pressure through the pulmonary venous resistance. By extrapolating this exponential to its origin at the moment of pulmonary artery occlusion, a pressure within the pulmonary vascular bed which approximates pulmonary capillary pressure (Pcap) was identified. Pcap, and not the pulmonary artery occlusion pressure (PAOP), is the major driving pressure forcing fluid from the pulmonary microvasculature. The results indicate that a discrete value for pulmonary capillary pressure can be reproducibly measured in paralyzed ventilated patients. The data report that mean pulmonary artery pressure, pulmonary capillary pressure, and total pulmonary vascular resistance (PVR) are increased in acute respiratory failure, but there is considerable variation in the distribution of pulmonary vascular resistance between the arterial and venous beds. The data suggest that there is unequal and variable partitioning of the increased PVR during acute respiratory failure. Bedside pressure profile Pcap measurements will allow optimum reduction of Pcap during ARF by infusing vasoactive agents to modify the distribution of PVR or reducing the PAOP.
Acute Disease↗
Heart rate and body temperature during free diving of Weddell seals.
We have developed and successfully used the first microprocessor-controlled monitors for collection of data on depth, heart rate, and body temperature of one fetal and five adult male freely swimming Weddell seals. Adult seals almost invariably experienced a prompt bradycardia at the start of each dive, and the mean heart rate during diving was significantly lower for dives greater than 20 min (P greater than 0.999). The heart rate was also significantly greater during the ascent portion of dives when compared with the descent portion (P greater than 0.95). The fetal seal experienced a slow onset of bradycardia when its mother dived; during diving the fetal heart rate decreased by an average of 1.1 beats/min for each minute of the dive. The fetal heart rate generally took approximately 10 min to recover to predive levels after its mother resurfaced to breathe. The body temperature of one adult male Weddell seal showed a decrease of greater than 1.5 degrees C from resting levels before dives of greater than 15 min were initiated and a drop of over 2 degrees C before dives of greater than 30 min duration.
Animals↗
Vascular components of ARDS. Clinical pulmonary hemodynamics and morphology.
Mild pulmonary artery hypertension (PAP, 29.6 +/- 10.6 mm Hg, mean +/- SD) due to a 3-fold elevation of pulmonary vascular resistance (PVR, 2.2 +/- 1.1 mm Hg X L/min) is a common finding in severe ARDS. A vasodilator such as nitroprusside (151 micrograms/kg X min) can be administered in early ARDS and will lower PAP and PAOP (capillary wedge pressure) while increasing cardiac output (CO) from 6.9 to 8.75 L/min X M2 and venous admixture from 23% to 31.6%. This suggests diffuse vasoconstriction is present in early ARDS. In 19 patients with severe ARDS, 13 had vascular occlusions on balloon occlusion angiography, and these occlusions correlated with increased post mortem counts of PA thrombi. At autopsy, there was a considerable increase of PA medial thickness and a reduction of lumen diameter. This hemodynamic and morphologic evidence suggests both vasoconstrictor and anatomic changes play major roles elevating the PVR in ARDS.
Acute Disease↗
99mTc-fibrinogen scanning in adult respiratory distress syndrome.
Fibrin is often seen occluding the lung vessels of patients dying from ARDS and is surrounded by regions of lung necrosis. To learn if we could observe increased or focal fibrin deposition and assess the kinetics of plasma fibrinogen turnover during severe acute respiratory failure, we injected technetium 99m-labeled human purified fibrinogen (Tc-HF) and used gamma camera scanning for as long as 12 h in 13 sequential patients as soon as possible after ICU admission. The fibrinogen uptake rates were determined by calculating the lung:heart radioactivity ratios at each time point. Slopes of the lung:heart ratio versus time were compared between ARDS and mild acute respiratory failure (ARF). The slope of the lung:heart Tc-HF ratio of the 9 patients with ARDS (2.9 +/- 0.4 units) was markedly higher (p less than 0.02) than the slope of the 4 patients with mild ARF (1.1 +/- 0.4) and the 3 patients studied 5 to 9 months after recovery from respiratory failure (0.7 +/- 0.07). In the 1 patient with ARDS and the 2 patients with mild ARF studied both during acute lung injury and after recovery, the lung:heart Tc-HF ratio had decreased at recovery. To compare the pulmonary uptake of Tc-HF to 99mTc-labeled human serum albumin (Tc-HSA), 5 patients were injected with 10 mCi of Tc-HSA, and scanning of the thorax was performed with a similar sequential imaging protocol 24 h after conclusion of the Tc-HF study.(ABSTRACT TRUNCATED AT 250 WORDS)
Adolescent↗
Pulmonary and systemic contribution to canine tracheobronchial blood flow.
To investigate the quantity and pathophysiological relationship of pulmonary and bronchial artery supply to the tracheobronchial tree in dogs in absence and in presence of increased tissue metabolic requirements, pulmonary and bronchial blood flow to the trachea and bronchi was determined by the radioactive microsphere method in nine adult mongrel dogs. Microspheres were injected into the right atrium of three dogs (pulmonary injection group) and into the left ventricle of four dogs (systemic injection group) just before and three, eight and fourteen days after resection and anastomosis of the right main bronchus. The remaining two dogs served as controls without surgery. After the fourth injection, the tracheobronchial tree was harvested, divided into segments, and the radioactivity counted in each segment. At baseline the systemic flow to the tracheobronchial tree expressed as percentage of cardiac output was 0.188% +/- 0.121 and the pulmonary blood flow was 0.007% +/- 0.004 (mean +/- S.D.). Three, eight and fourteen days after transection and anastomosis of the right main stem bronchus, the systemic blood flow to the bronchial segments proximal and distal to the anastomosis increased significantly whereas pulmonary blood flow did not change. Depending on the anatomical segment considered the systemic contribution was between 5 and 100 times larger than the pulmonary one, both at baseline and after transection and anastomosis of the right main stem bronchus, but the ratio diminished in more distal segments where pulmonary blood flow was larger. It is concluded that the systemic blood contribution to the canine tracheobronchial tree is larger than the pulmonary one and that, in the presence of increased metabolic needs, it responds more promptly with increased blood flow.
Animals↗
Cyclooxygenase blockade elevates leukotriene E4 production during acute anaphylaxis in sheep.
We examined changes in the levels of eicosanoids in blood and pulmonary lymph of anesthetized sheep undergoing acute anaphylaxis. Within 1-3 min of intravenous antigenic challenge of previously sensitized sheep, there were approximately 7-30-fold elevations in mean arterial plasma levels of thromboxane B2 and 6-ketoprostaglandin F1 alpha, respectively, as measured by RIA. Negligible changes in levels of these cyclooxygenase products were found in both nonsensitized sheep and in sensitized sheep treated with indomethacin before antigenic challenge. In contrast, no changes in levels of sulfidopeptide leukotrienes (SPLT) in pulmonary lymph were detectable by RIA during anaphylaxis in sensitized or nonsensitized sheep, but levels of SPLT in indomethacin-treated sensitized sheep increased more than fivefold above levels in lymph from both other groups of animals. The immunoreactive SPLT in lymph from indomethacin-treated sheep was accounted for as LTE4, as demonstrated by mobility on HPLC and absorbance at 280 nm. These results support the possibility that certain undesirable effects of nonsteroidal antiinflammatory drugs, such as cardiopulmonary reactions in aspirin-sensitive individuals, and impaired renal and cardiac function during therapy with these drugs, may be related in part to augmented synthesis of the 5-lipoxygenase pathway products, especially those of the sulfidopeptide class. Increased LT production could also limit the antiinflammatory effectiveness of these drugs in many disease states.
6-Ketoprostaglandin F1 alpha↗
Microcomputer-assisted metabolic studies of voluntary diving of Weddell seals.
Utilizing a microprocessor-controlled peristaltic withdrawal pump, arterial blood samples were obtained from Weddell seals (Leptonychotes weddelli) while diving voluntarily under the sea ice at McMurdo Sound, Antarctica. Plasma concentrations of glucose, lactate, free fatty acids, urea, and amino acids were determined in seals at various times during rest, diving, and recovery. In addition, radiolabeled palmitate, glucose, lactate, p-aminohippurate, inulin, galactose, and cholate were injected into the descending aorta of seals in the resting state or during voluntary diving at sea. Sequential plasma samples were collected, and the kinetics of wash-in and clearance of each component were determined. Under resting conditions, the change in specific activity of palmitate, glucose, and lactate after bolus injection followed smooth multiexponential decay kinetics as in terrestrial animals. Similar decay curves for the clearance of organ-specific compounds were also obtained in seals at rest. If, during voluntary diving, Weddell seals were analogous to exercising animals, the wash-in and clearance kinetics of all metabolites and organ-specific compounds would be qualitatively similar to those observed at rest. In contrast, wash-in and clearance appeared to be qualitatively and quantitatively modified in a way consistent with utilization of the classical Scholander diving response in both short and long dives.
Amino Acids↗
Hemoglobin concentrations and blood gas tensions of free-diving Weddell seals.
Arterial blood gas tensions, pH, and hemoglobin concentrations were measured in four free-diving Weddell seals Leptonychotes weddelli. A microprocessor-controlled sampling system enabled us to obtain 24 single and 31 serial aortic blood samples. The arterial O2 tension (PaO2) at rest [78 +/- 13 (SD) Torr] increased with diving compression to a maximum measured value of 232 Torr and then rapidly decreased to 25-35 Torr. The lowest diving PaO2 we measured was 18 Torr just before the seal surfaced from a 27-min dive. A consistent increase of arterial hemoglobin concentrations from 15.1 +/- 1.10 to 22.4 +/- 1.41 g/100 ml (dives less than 17 min) and to 25.4 +/- 0.79 g/100 ml (dives greater than 17 min) occurred during each dive. We suggest that an extension of the sympathetic outflow of the diving reflex possibly caused profound contraction of the Weddell seal's very large spleen (0.89% of body wt at autopsy), although we have no direct evidence. This contraction may have injected large quantities of red blood cells (2/3 of the total) into the seal's central circulation during diving and allowed arterial O2 content to remain constant for the first 15-18 min of long dives. The increase of arterial CO2 tensions during the dive and the compression increase of arterial N2 tensions were also moderated by injecting red blood cells sequestered at ambient pressure. After each dive circulating red blood cells are oxygenated and rapidly sequestered, possibly in the spleen during the first 15 min of recovery.
Animals↗
Rat pulmonary artery wall injury by chronic intermittent infusions of Escherichia coli endotoxin. Obliterative vasculitis and vascular occlusion.
To examine the effect of intermittent endotoxemia on rat pulmonary artery structure and hemodynamic function we infused purified Escherichia coli endotoxin on four occasions over 3 weeks (at 7-day intervals), through an indwelling catheter placed in the external jugular vein. The fourth infusion of endotoxin was associated with widespread but focal alveolar consolidation, reduced perfusion of small pulmonary arteries by lumen occlusion and obliteration, pulmonary vascular wall injury, and a peripheral leukocytosis (mean +/- SEM total leukocytes, endotoxin 133.5 +/- 23 X 10 mm3, control 12.2 +/- 1.2 X 10 mm3, p less than 0.001) in which polymorphonuclear (PMN) leukocytes predominated at the expense of lymphocytes (p2x less than 0.01). The alveolar wall was thickened and the alveolar space was consolidated by degenerating polymorphonuclear leukocytes, mononuclear cells, lipid laden alveolar macrophages, erythrocytes, fibrin, and cell debris. In regions of alveolar consolidation vessel lumens were either narrowed by subendothelial cell collections that consisted either of mononuclear cells or degenerating mural and inflammatory cells, or they were occluded by degenerating inflammatory cells and cell debris. The walls of occluded vessels were evident only by their residual elastic laminae: remnants of lysed endothelial cells lined the intima and the media consisted of degenerating mural and inflammatory cells. Capillary endothelial cells showed extensive hydropic degeneration and lysis of cell contents. Intimal precursor smooth muscle cells were hypertrophied but were not associated with the appearance of mature smooth muscle cells in the walls of small pulmonary arteries. In regions of less severe alveolar consolidation by inflammatory cells, vessel wall injury was still evident but less marked; precursor smooth muscle cells were hypertrophied; subepithelial and subendothelial collections of fluid and fibrin were present; and plasma membranes of endothelial cells were disrupted. Despite extensive pulmonary vascular injury, chronic intermittent endotoxemia did not produce the structural changes associated with pulmonary hypertension (medial thickening and appearance of medial muscle in previously nonmuscular arteries) nor a significant change in pulmonary artery pressure.
Animals↗
Seal lungs collapse during free diving: evidence from arterial nitrogen tensions.
Arterial blood nitrogen tensions of free-diving Weddell seals (Leptonychotes weddelli) were measured by attaching a microprocessor-controlled blood pump and drawing samples at depth to determine how these marine mammals dive to great depths and ascend rapidly without developing decompression sickness. Forty-seven samples of arterial blood were obtained from four Weddell seals during free dives lasting up to 23 minutes to depths of 230 meters beneath the sea ice of McMurdo Sound, Antarctica. Peak arterial blood nitrogen tensions of between 2000 and 2500 millimeters of mercury were recorded at depths of 40 to 80 meters during descent, indicating that the seal's lung collapses by 25 to 50 meters. Then arterial blood nitrogen tensions slowly decreased to about 1500 millimeters of mercury at the surface. In a single dive, alveolar collapse and redistribution of blood nitrogen allow the seal to avoid nitrogen narcosis and decompression sickness.
Animals↗
Oxygen toxicity and restructuring of pulmonary arteries--a morphometric study. The response to 4 weeks' exposure to hyperoxia and return to breathing air.
This study describes the pulmonary vascular lesions in rat pulmonary arteries and altered right ventricular weight after 1) prolonged exposure to hyperoxia (87% O2 for 4 weeks) at ambient pressure, 2) weaning from hyperoxia to air over 7 days, and 3) return to breathing air for 2, 4, or 8 weeks. Hyperoxia for 28 days narrows the lumen of intraacinar and preacinar arteries, increasing the percent medial thickness (%MT) by reducing the external diameter and thickening medial muscle. The ratio of patent intraacinar arteries to alveoli is significantly reduced, and pulmonary vascular obstruction and obliteration is evident by electron microscopy. A higher proportion of intraacinar and preacinar arteries have muscle in their wall than in the normal lung: in alveolar wall and duct regions, the proportion of partially muscular and muscular intraacinar arteries increases at the expense of nonmuscular ones (for both regions P chi 2 less than or equal to 0.001); and in arteries associated with terminal bronchioli and bronchioli the proportion of muscular arteries increases at the expense of partially muscular ones (for both regions P chi 2 less than or equal to 0.001). Both after weaning and after return to breathing air lumen size increases; but, even after 8 weeks, the %MT remains significantly increased, and the ratio of intraacinar arteries to alveoli is less than normal. After weaning, the proportion of muscularized intraacinar and preacinar arteries is similar to that after hyperoxia. Two weeks after return to breathing air, the proportion of muscularized alveolar wall and duct arteries is greater (for both regions P chi 2 less than or equal to 0.001). Even 8 weeks after return to breathing air more arteries are muscularized than normal (for both alveolar wall and duct regions P chi 2 less than or equal to 0.001), and within the alveolar wall still more are muscularized than after hyperoxia (P chi 2 less than or equal to 0.001). Hyperoxia causes right ventricular hypertrophy, reducing the ratio of the weight of the left ventricle and septum to that of the right ventricle (P chi 2 less than or equal to 0.001). Weaning further increases the hypertrophy, the ratio being further reduced (P chi 2 less than or equal to 0.001, compared with both hyperoxia and control values). On return to breathing air the degree of hypertrophy is less, but it persists, and even after 8 weeks the ratio is still less than normal (P chi 2 less than or equal to 0.01).
Animals↗
Technetium-fibrinogen lung scanning in canine lung contusion.
To detect experimentally induced acute lung contusion in anesthetized dogs, serial radionuclide images of the lung were recorded following intravenous infusion of 99mTc-labelled human fibrinogen (Tc-HF). The accumulation of Tc-HF in canine lungs was serially quantitated for up to 20 hours after lung contusion. A contusion (#1) was produced in one lung, Tc-HF was injected IV after 15 minutes, and 75 minutes later a contralateral lung contusion (#2) was produced in a series of 14 dogs. At autopsy the excised lungs were scanned, sectioned, and counted for radioactivity. Radiolabelled fibrinogen accumulated within 2-4 minutes of contusion #2 and remained stable over the next 20 hours in 14 dogs; contusion #1 was barely visible in four dogs. Lung Tc-HF activity in the central region of contusion #2 remained sixfold higher than in normal lung tissue. These data suggest that following lung contusion, fibrinogen deposition occurs rapidly and remains stable over a 20-hour interval of observation.
Animals↗
Pulmonary artery remodeling and pulmonary hypertension after exposure to hyperoxia for 7 days. A morphometric and hemodynamic study.
This study shows by morphometric and hemodynamic techniques that exposure to hyperoxia at normobaric pressure causes rapid structural remodeling of rat pulmonary arteries and pulmonary hypertension. After 7 days of 90% O2, pulmonary artery cross-sectional area is reduced by a striking loss of intraacinar arteries (control, 13 +/- 1 sq mm; exposed, 8 +/- 1 sq mm; P less than 0.001), the ratio of arteries to alveoli being 4:100 in control rats and 2.5:100 after hyperoxia. The lumen of preacinar and intraacinar arteries is narrowed by a reduction of vessel external diameter (ED) and an increased medial wall thickness (MT). There is a significant reduction in the percent medial thickness [( 2 X 100 X MT]/ED) in both regions. The proportion of muscular and partially muscular intraacinar arteries increases at the expense of nonmuscular ones (P [chi 2] less than 0.01), and fully muscular arteries appear in the alveolar wall where they are not normally found. Intimal thickening occurs in 19% of alveolar duct and 34% of alveolar wall nonmuscular arteries. Right ventricular hypertrophy occurs, the ratio of the left ventricle plus the septum to the right ventricle being significantly reduced (control, 4.07 +/- 0.26; exposed, 3.23 +/- 0.10; P less than 0.02). After 3 days of 87% O2, pulmonary artery pressure is still normal (17.0 +/- 0.9 mmHg) but after 7 days it is significantly increased (26.2 +/- 0.9 mmHg; P less than 0.01), as is pulmonary vascular resistance (control, 0.033 +/- 0.003; exposed, 0.065 +/- 0.015 U/kg; P less than 0.05). Return to air breathing (after 7 days at 87% O2) causes pulmonary vasoconstriction and a further rise of the pulmonary artery pressure (to 38.3 +/- 3.3 mmHg after 60 minutes).
Animals↗