Pulmoner Emboli

PHYSIOLOGICAL

REVIEWS

Vol. 63, No. 3, July

Printed

1983

in USA.

Pulmonary ASRAR

Microembolism B. MALIK

Department of Ph@oZogg,Albany MidicuZ College of Union University, Albany, New York I. Introduction ......................................................... A. Recent reviews ................................................... .................... B. Pulmonary microembolism and thrombogenesis ................... C. Methods of producing pulmonary microembolism II. Pulmonary Hemodynamic Response to Microembolization .............. A. Mechanisms of increase in pulmonary vascular resistance ........... ..................... B. Factors mediating pulmonary vasoconstriction III. Pulmonary Edema After Microembolization ............................ A. Effects of pulmonary microembolism on lung fluid balance .......... B. Effects of increasing degree of embolization ........................ C. Reversibility of increases in permeability ....................... .... D. Morphological alterations in lung endothelium ..................... E. Microvessels versus arteries as sites of fluid and protein exchange ... F. Hemodynamic mechanisms associated with increased permeability .. G. Lymphatic impairment ............................................ H. Effects of regional atelectasis ...................................... I. Cellular and humoral mechanisms ................................. J. Ischemia ......................................................... K. Neural factors .................................................... L. Role of bronchial circulation ...................................... M. Cellular edema ................................................... ................................... IV. Tachypnea After Microembolization A. Lung irritant or rapidly adapting receptors ........................ B. Juxtapulmonary capillary receptors (C fibers) ...................... C. Pulmonary arterial baroreceptors .................................. D. Pulmonary stretch receptors ...................................... E. Summary ........................................................ V. Airway Constriction After Microembolization .......................... A. Sites of bronchoconstriction ....................................... .............................. B. Factors affecting bronchoconstriction C. Mechanisms involved in producing bronchoconstriction ............. D. Homeostatic value of bronchoconstriction E. Alveolar dead space after embolization ............................. F. Alveolar collapse ................................................. VI. Mechanisms of Arterial Hypoxia ...................................... A. Time course of gas-exchange impairment ........................... ............................................. B. Diffusion impairment C. Increased venous admixture ....................................... D. Ventilation-perfusion imbalance ................................... .......................................... E. Alveolar hypoventilation VII. Bronchial Blood Flow ................................................. A. Anatomy and physiology of bronchial circulation ................... B. Effects of pulmonary microembolizatibn on bronchial blood flow ..... VIII. Conclusions .......................................................... ..........................

1114

0031-9333/83 $1.50 Copyright 0 1983 the American

Physiological

Society

1115 1115 1115 1116 1117 1117 1120 1127 112’7 1132 1134 1136 1138 1139 1140 1141 1141 1166 1166 116’7 1167 1169 1170 1170 1172 1173 1173 1174 1174 1175 1176 1178 1179 1179 1180 1180 1181 1181 1182 1183 1185 1185 1188 1191

July

1983

PULMONARY

MICROEMBOLISM

1115

I. INTRODUCTION

Because the pulmonary circulation receives the entire cardiac output, it has a major role in filtering emboli that may be present in the systemic venous blood. With this filtering function the pulmonary circulation protects the essential coronary, renal, and cerebral circulations during an embolic episode. However, pulmonary embolism has direct and unique effects on the lungs. Some effects of pulmonary embolism are homeostatic, whereas others disrupt lung function. For example, the bronchoconstriction occurring after embolization redistributes ventilation away from the embolized regions, thus improving the match betweeen ventilation and perfusion. On the other hand, pulmonary edema, which also follows embolization, can worsen the hypoxemia caused by the embolism. The purpose of this review is to examine the diverse alterations in pulmonary function associated with microembolization and to review critically the mechanisms proposed to explain these functional changes. This review deals with pulmonary microembolism, defined as obstruction of pulmonary arteries < 200 pm in diameter. Macroembolism, which refers to occlusion of larger pulmonary arteries, is considered only when the responses’ to small-vessel and large-vessel obstruction differ. A. Recent Reviews The proceedings of a symposium devoted to the pathogenesis of pulmonary microembolism and the resulting pulmonary and hemodynamic responses were published in 1973 (344). Moser (343) has reviewed the literature concerning the clinical manifestations of this problem, and Saldeen (437, 438) has summarized the extensive studies conducted in his laboratory on the pathophysiology of microembolism, particularly in regard to development of pulmonary edema. Certain aspects of the pulmonary response to microembolism have been discussed in reviews not dealing primarily with the problem of pulmonary vascular occlusion (297, 384, 470). B. Pulmonary

Microembolism

and Throwzbogenesis

A clot in the lumen of a pulmonary artery is usually caused by an embolism (197, 419). Two factors suggest the embolic nature of such clots: 1) their occurrence in previously normal pulmonary arteries (437) and 2) their common association with identifiable thrombi in systemic veins or in the right heart (419, 452). The reticuloendothelial system can modulate the degree of pulmonary embolism (427). Embolization of pulmonary microvessels results when the ability of the reticuloendothelial system to clear circulating microaggregates of fibrin and platelets is impaired (238, 430). Once lodged in pulmonary vessels, the microemboli may serve as sites

1116

ASRAR

B. MALIK

Vduww

63

of intravascular coagulation by directly activating the intrinsic coagulation system (166,437). Secondary pulmonary thrombosis may also occur if there is so much emboli-induced endothelial damage that the extrinsic coagulation pathway is also activated by exposure of subendothelial membrane to the blood components (237, 381, 457). Other causes of secondary thrombosis in the lungs after pulmonary embolism are states of blood hypercoagulability (192) and stagnant pulmonary blood flow caused by vascular occlusion (77, 237). C Methods

of Producing

Pulmonaw

Microembolism

Pulmonary embolism has been produced injecting substances of various sizes, shapes, weights, and compositions [e.g., pumice powder (539), lycopodium spores (115), starch granules (60), glass beads (302), fat emboli (433, 435), endotoxin (490), balloons ,(376), collagen suspension (510), and even gunshot (357)]. Although all these methods obstruct lung vessels, little consideration has been given to the possibility that they produce different alterations in pulmonary function. For example, substances such as glass beads cause intravascular coagulation by activating the Hageman factor (factor XII) (236, 237), whereas a balloon in a pulmonary artery causing the same degree of physical obstruction may not produce clotting. The possibility that different methods of embolization produce the same alteration in pulmonary function by different mechanisms has also been largely neglected. For example, the pulmonary edema after embolization with a balloon embolus may be due to a hemodynamic mechanism, whereas the edema associated with infusion of thrombin (EC 3.4.21.5) may be due to vascular injury resulting from the humoral mediators released because of this challenge (437). Instead of injecting foreign subtances (such as glass beads) to produce emboli, some workers have produced pulmonary microembolism by directly stimulating intravascular coagulation with thrombin (437) or by causing the release of clots previously formed in peripheral veins (536). These represent closer paradigms of real embolic disease, because these microemboli consist of fibrin, platelets, and white blood cells. With thrombin infusion, however, there is \disseminated intravascular coagulation (300). Intravenous injection of preformed autologous blood clots is also advantageous, because the embolism is usually localized in the lung. Another advantage of these methods is that dissolution can be studied (17, 108); this is not possible with foreign emboli. Embolization with glass beads of uniform size, however, allows better control over the size of the vessels being obstructed (219,556) than does the release of preformed venous clots, which usually vary in size and shape. Nevertheless the studies of Nelson and Smith (357) have shown that clumping of glass beads may occur in the larger arteries and that emboli advance into the smaller vessels. Another variable influencing the distribution of microemboli in the lung is specific gravity: emboli heavier than the formed

July

PULMONARY

1983

MICROEMBOLISM

1117

elements (e.g., glass beads) gravitate to the dependent lung regions, whereas lighter emboli (e.g., air emboli) are preferentially distributed to the upper lung regions (67). Finally, certain emboli may not remain static. For example, air bubbles probably do not remain at one point within the vessels but elongate within the vasculature and deform as they are pushed further into smaller vessels (67). Therefore the pulmonary response to air embolization may change over time as the emboli move from large to small pulmonary arteries. Thus when choosing a model of pulmonary microembolism one must consider whether the embolization is I) only obstructive, 2) associated with secondary thrombosis, 3) located in small or large vessels, 4) reversible or chronic, and 5) confined only to the pulmonary vascular bed. II. PULMONARY

HEMODYNAMIC

RESPONSE

TO MICROEMBOLIZATION

Microembolization clearly produces pulmonary hypertension without causing changes in left atria1 pressure (Pla) and pulmonary blood flow (265, 301, 376, 390). This observation indicates that the increased pulmonary arterial pressure (Ppa) primarily results from a rise in pulmonary vascular resistance (PVR). Mechanical obstruction, vasoconstriction, or a combination of both effects could cause this increase in PVR (530). The following sections deal with the relative contributions of these factors to the increased PVR associated with microembolization. A. Mechanisms

of Increase in Pulwwnarvy Vascular Resistance

I. Eflects of mechanical

obstruction

Mechanical obstruction of lung vessels after microembolization undoubtedly causes a rise in PVR, but the resistance begins to increase appreciably only after -50% of the pulmonary vascular bed has been embolized. In 1923 Haggart and Walker (172) reduced the cross-sectional area of the pulmonary artery of cats with a screw clamp and found that the Ppa increased only after the area had been reduced by more than 50%. Gibbon et al. (154, 155) restudied the problem in the same species by surgically reducing the cross-sectional area of the pulmonary vascular bed. They found that pressure did not begin to increase until more than one lung had been removed and that pressure increased thereafter in direct proportion to the amount of lung tissue removed. When only 29% of the lung remained, the pulmonary hypertension was so great that fatal pulmonary edema ensued. These authors did not measure PVR in their studies; however, because pulmonary blood flow and Pla do not change significantly until over 65% of the

1118

ASRAR

B. MALIK

vdume

63

pulmonary vascular bed is obstructed (103), the PVR in the experiments of Gibbon and co-workers must have increased (as did the Ppa) in proportion to the degree of mechanical obstruction. There are two reasons why it is necessary to obstruct more than 50% of the pulmonary vascular bed to produce any increase in the Ppa. First, the arteries and veins of the lung are compliant and canaccommodate twoto threefold increases in blood volume with only a 1-2 mmHg increase in pressure (136, 258, 259). Therefore Ppa cannot be expected to increase significantly until the pulmonary blood volume has risen 200-300%; this increase would not occur until more than one lung has been removed. After this point, small increments in volume produce relatively large increases in Ppa. Second, the recruitment of additional pulmonary vessels minimizes the pressure changes in lung vasculature after vascular obstruction (394). In most studies this factor was probably less important than the role played by vascular compliance: because the animals were supine in these experiments, the lungs were largely in the zone of total vascular recruitment, zone III (537). In zone III, Ppa is greater than the pulmonary venous pressure (Ppv), which is greater than the alveolar pressure (PA). However, where there is a discernible zone I (PA > Ppa > Ppv) and a zone II (Ppa > PA > Ppv) with a potential for greater, recruitment, these regions would minimize the rise in Ppa associated with vascular obstruction. z. Active

pul-q

vasocmtriction

a) EvW supporting vasocmttictim The increased Ppa seen after microembolization is caused by active pulmonary vasoconstriction as well as by the passive mechanical effect of obstruction described in section IIAI. Lee et al. (263) found that selective embolization of a portion of one dog lung with glass beads 100 pm in diameter caused significant increases in Ppa and PVR. These changes were not caused by hypoxic pulmonary vasoconstriction or mechanical obstruction, because the blood gases were normal and no reflux of beads occurred into the nonembolized lung. A similar increase in PVR has also been observed in patients with embolic occlusion of ~25% of the pulmonary vascular bed as determined by angiography (3). These results are in accord with the earlier finding of Dalen et al. (103) and Dexter et al. (115) that pulmonary arterial hypertension can be produced by embolization of ~50% of the pulmonary vascular bed only if the emboli are ~170 pm in diameter. The Ppa increased after embolization with particles 28-30 pm in diameter; of the 600-280,000 X lo6 vessels of this size present in the pulmonary circulation, only 22 X lo6 vessels were assumed to be blocked (Table 1; 15). On the other hand, as seen in Table 1, to produce the same elevation in Ppa with particles >170 pm in diameter, it was necessary to inject enough particles to block most of the vessels of this diameter. The pulmonary hypertension seen after microembolization with the smaller par-

July

PULMONARY

1983

1119

MICROEMBOLISM

Numbers and sizes of emboli needed to raise JX&WTUZ~ pressure 5-10 mmHg in dogs

1.

TABLE

arterial

Diameter of Emboli, m m

Artery

Type of Emboli

Lobar 1st order 2nd order

Blood clots Glass beads Polystyrene spheres or blood clots Polystyrene spheres

5.0 4.0 2.3

7 18 40

1.0

1,020

Lobular

Polystyrene

spheres

0.3

16,000

Atria1

Polystyrene

spheres

0.17

Arterioles

Lycopodium

spores

3rd order

From

Dalen

28-30

pm

Number

of Vessels

600-280,000

Number of Emboli Injected, mean f 1 SD

7tl 28t7 58 t 20 1,635 t 289

x lo6

28,125

t 10,270

89,140

t 4,260

22 x lo6

et al. (103).

titles is obviously due to some mechanism other than simple mechanical obstruction of the pulmonary vascular bed. Hyland et al. (202, 203) showed that constriction of pulmonary vessels by the microemboli (28 pm in diameter) was rapidly reversed by increasing blood flow through the lung. This decreased resistance could not be demonstrated when embolization was induced by particles 300 pm in diameter (202,203), however, indicating that the larger emboli did not constrict vessels. In another study the increased PVR observed in dogs after microembolization with glass beads 100 pm in diameter was greater than that predicted from the degree of vascular obstruction (302). The finding that a 50-60s obstruction of the pulmonary vascular bed caused a disproportionately large increase in PVR [i.e., an increase of 300~500% (302)] indicates microembolization-induced constriction of pulmonary vessels. Thus active pulmonary vasoconstriction occurs after microembolization but only when the small pulmonary arteries (those 470 pm in diameter) are obstructed. The homeostatic value of pulmonary vasoconstriction has been questioned because the response appears to constitute a positive feedback. Any additional vasoconstriction after pulmonary vascular obstruction would further increase the right ventricular afterload, and failure may occur if the increase in PVR is marked (137). Whitteridge (539) suggested that pulmonary vasoconstriction would be beneficial if constriction was confined to precapillary vessels because it would prevent the high Ppa from being reflected to the filtering microvessels. Experimentation has not yet established whether reversal of the active pulmonary vasoconstriction seen with microembolization can actually exacerbate the embolism-induced pulmonary edema. b) Site of pulmonary vasoconstriction. The pulsatile nature of pulmonary capillary blood flow, as measured by the nitrous’oxide method, has been used to determine the site of constriction after embolization (22, 246). The flow

1120

ASRAR

B. MALIK

Vohmw

6.9

wave generated by contraction of the right ventricle is propelled through the low-resistance pulmonary arterial system and normally arrives at the capillaries with only minor alterations in shape (246). Embolization of the left lung lobe, however, resulted in a decreased pulsatility of pulmonary capillary flow associated with an increased PVR (246). Active vasoconstriction occurring at precapillary sites explained the decreased pulse transmission. The flow response was only transient [lasting 30 min (246)], which suggests the effect was related to’ the release of a humoral substance(s), and the response waned as circulating levels of this substance(s) decreased or as tachyphylaxis occurred. The reversibility of the response indicated the effect was not simply due to mechanical obstruction. The value of the nitrous oxide technique for localizing the site of vascular resistance, however, is questionable-the approach is indirect and assumes an unaltered pulmonary vascular compliance (340). Nevertheless these findings suggest pulmonary arteries are the primary sites of constriction associated with microembolization. Researchers have not yet determined the size of the constricting pulmonary arteries after microembolization. The vessels most likely to constrict are the patent muscular vessels that are near the embolized lung units and that have diameters of 100-500 pm (410). The hypothesis that microembolism affects the ,nearby lung segments is supported by the observation that embolization of a portion of the lung causes vasoconstriction in an independently perfused lung segment (554). However, Daily et al. (100) have.challenged this conclusion. They showed that embolization of an isolated and perfused left lower lobe with glass beads 42 pm in diameter and small thrombi did not increase resistance in the nonembolized lung. The differences between these results could be explained by the different experimental protocols. In one case the isolated lobe was embolized, whereas in the other a much larger fraction of the lung was embolized (554). The release of the vasoconstrictor factor(s) after embolization of one lobe is likely to be less than that associated with embolization of a larger portion of the lung. The degree of embolization could explain why Daily and co-workers did not observe vasoconstriction in the nonembolized lung. However, the matter of constriction of pulmonary vessels close to the embolized regions needs to be reexamined. B. Factors

Mediating

1. Neural

mechanisms

Pulwwnaq

Vasoconstriction

Sympathetic nerve stimulation causes constriction of pulmonary vessels (104, 235, 478). There is a great deal of interspecies variation related to the intensity of the pulmonary vasoconstrictor response seen during sympathetic nerve stimulation; for example, the constrictor response during sympathetic nerve stimulation is weaker in dogs than in cats or sheep (104,174,175,206).

July

1983

PULMONARY

MICROEMBOLISM

1121

This lesser response appears to be related to the paucity of sympathetic motor innervation of pulmonary blood vessels and vascular smooth muscle (129, X34, 410). In all species, however, the pulmonary arteries are the primary sites of the sympathetically mediated constriction (206). The veins constrict less than the arteries (206), reflecting the sparse innervation and the relatively uneven distribution of smooth muscle in pulmonary veins (410). The role of sympathetic efferents in the pulmonary vasoconstrictor response associated with microembolization is poorly understood. Several studies have indicated that sympathetic mechanisms are not involved in the pulmonary vasoconstriction seen after microembolization (105,309,380,546). In an isolated and perfused dog lung preparation, pharmacological denervation with the ganglion-blocking agent hexamethonium did not affect the pulmonary hypertension resulting from embdlization with glass beads 60 pm in diameter (546). Surgical removal of the sympathetic nerves could not prevent the pulmonary vasoconstriction due to microembolization with lycopodium spores 28-30 pm in diameter (150). The pulmonary hypertension associated with the injection of autologous clots of different sizes was not altered after total lung denervation (309, 380). Bilateral cervical vagotomy did not influence the pulmonary hemodynamic response (105), indicating that vagal mechanisms are also not involved in the vasoconstriction. Finally, isolated and perfused lungs (a completely denervated preparation) embolized with glass beads demonstrated a steadily increasing PVR (14, 33, 120, 360). It can be deduced from these studies that there is little, if any, involvement of sympathetics or of the vagus nerve in the pulmonary vasoconstrictor response associated with microembolization. Although a few studies indicate that sympathetic mechanisms are involved in the vasoconstriction associated with emboli (360,401), it is unclear why their role was demonstrated in these studies. The increase in PVR after embolization was attenuated by thoracic sympathectomy (360,399), and the pulmonary vasoconstriction induced by embolization in an isolated hearthead-lung preparation was eliminated when the cephalic circulation was removed (401). With an intact cephalic circulation, the response was blocked either by bilateral removal of the stellate ganglion and thoracic sympathetic chains to T4 or T5 or by prior treatment with hexamethonium (401). In another study, embolization of one lung lobe with glass beads caused an immediate increase in perfusion pressure in another isolated and perfused lobe, an increase that could be prevented by sympathectomy and vagotomy (360). From these numerous studies it is apparent that the role of neural mechanisms in the pulmonary vasoconstriction after microembolization differs from preparation to preparation and among laboratories. A part of the problem in delineating neural mechanisms is related to the use of lung preparations that are isolated and perfused. The elaborate surgery necessary for isolated preparations interferes with the sensory and motor innervations of the lungs (104) and may produce conflicting results. Another experimental problem is that lungs are not always embolized to the same degree. In some

1122

ASRAR

B. MALIK

vdurne

63

cases the pulmonary artery stretch receptors, which may activate the sympathetic efferent pathways in the lung (84, 85,262), may not have been fully stimulated. Stretch receptors are believed to be stimulated by acute distension of the pulmonary artery, such as that produced by a nonocclusive balloon (84, 85, 262). The resultant reflex evokes pulmonary vasoconstriction and hypertension (230). The response was inhibited by surgical dissection of the pulmonary artery and procaine infiltration into the area of the bifurcation, suggesting that the reflex was associated with receptors situated at this site (230). Sympathectomy produced by 6-hydroxydopamine also abolished the balloon-induced increase in PVR (230), indicating that sympathetic efferents are activated. Phentolamine and propranolol did not abolish this reflex (230), suggesting that the responsible efferents are not the classic CY-and P-adrenergic receptors. 2. Humoral

mechanisms

In contrast to the ambiguities surrounding the involvement of sympathetic mechanisms in mediating the increased PVR seen after microembolization, there are many studies indicating that humoral factors are responsible fdr vasoconstriction. The same number of glass-bead microemboli caused greater increases in Ppa and PVR in nonheparinized than in heparinized animals, whereas the levels of hypoxemia and acidosis were similar in both groups (301). Because the experimental groups were identical except for the presence of heparin, the greater increase in PVR seen in the nonheparinized dogs is somehow related to the release of humoral factors after intravascular coagulation. Other studies have shown that to elevate PVR to similar levels in normal and thrombocytopenic animals, twice as many glass beads must be injected into the thrombocytopenic animals (36). Such studies support the notion that humoral factors are released from platelets after their aggregation and activation. The following section discusses the specific mediators that may be involved in increasing PVR after embolization. a) Specific mediators. One way to identify the humoral factors that mediate pulmonary vasoconstriction after microembolization is to examine the effects of pharmacological agents that inhibit the synthesis or release of pulmonary vasoactive mediators. Inhibition of prostaglandin synthesis by cyclooxygenase inhibitors (indomethacin, meclofenamate, or polyphloretinphosphate) and of histamine release (with chlorpheniramine and methamide) attenuated the increased PVR produced by glass-bead injection into dog lungs (500). The obvious conclusion is that prostaglandins and histamine are responsible for pulmonary vasoconstriction; these data do not provide definitive proof, however, because these antagonists may have independent pulmonary vasodilator effects (499). The finding that both histamine and the products of cyclooxygenase pathway [thromboxane A2 (TXA2) and prostaglandins] are present in in-

July

PULMONARY

1983

1123

MICROEMBOLISM

creased concentrations in the venous blood draining embolized lungs (432) nevertheless provides additional evidence involving these factors in the pulmonary vasoconstrictor response. Pulmonary vasoconstrictor agents such as TXAz are released after microembolization by aggregated platelets and leukocytes in the pulmonary circulation (207, 561). Histamine is released from the mast cells situated in the extravascular space (13, 269) and, to a lesser extent, from the basophils sequestered in the lungs after embolization (269, 272). In spite of these rather convincing data, a causal relationship between the release of these vasoactive factors and the pulmonary vasoconstriction associated with microemboli has not been firmly established. In fact, the observations that histamine (Fig. 1) and TXA2 appear to constrict primarily the pulmonary veins (47, 175, 374) would argue against these substances as primary mediators of the pulmonary arterial constriction associated with microembolization. Because serotonin (on a molar basis) is the most potent constrictor of pulmonary precapillary vessels (Fig. 1; 207), Comroe and colleagues (88) 1

7

HISTAMINE

)

(

(NS)

1

t

(0.003)

NOREPINEPHRINE

PROSTAGLANOIN

Fz,

k+k

(0.001)

SEROTONIN

U-1

t-i

t-:::14:.

HYPOXIA

STIMULATION

.

CONTROL

1

l-i

UPSTREAM

15

10

5 PRESSURE

l-i

I

DOWNSTREAM

0

5 DROP

10

15

(TORR)

FIG. 1. Effects of various pulmonary vasoconstrictor stimuli on upstream (arterial) and downstream (venous) pressure drops measured using the venous outflow occlusion method. Numbers in parentheses are P values comparing upstream and downstream pressure drops with control. Stimulation refers to stimulation of stellate ganglion. All stimuli except histamine caused greater increase in arterial resistance than in venous resistance. NS, not significant. [From Hakim et al. (175).]

1124

ASRAR

B. MALIK

Vohme

63

postulated this compound as the mediator of the smooth muscle contraction associated with emboli: a constant infusion of serotonin caused marked pulmonary hypertension that persisted as long as the infusion lasted. Knisley et al. (251) postulated a rather unique explanation for the pulmonary hypertensive effect associated with serotonin. They observed that serotonin infusions produced a fine white precipitate [possibly serotonin-mediated platelet aggregation (378)] within the pulmonary arteries; this precipitate disappeared soon after its formation. Because of its rapid disappearance, the precipitate cannot explain the sustained hypertension, which indicates that serotonin probably has a direct effect on pulmonary vascular smooth muscle. The finding that precapillary pulmonary vessels are the primary sites of constriction after both serotonin infusion (Fig. 1; 175) and microembolization (246) supports the hypothesis proposed by Comroe that serotonin is the primary mediator of the microemboli-vasoconstrictor response. The release of serotonin after embolization has been studied in a dog lung preparation in which a left lower lobe was isolated and perfused at a constant flow rate (554). Embolization of the remaining lung with autologous clots 150-250 pm in diameter caused an increased Ppa, followed by increased arterial and venous pressures in the perfused lobe at 45 s after the peak pressor response occurred in the embolized lung. These changes were accompanied by increased serotonin levels in pulmonary venous blood draining the nonembolized lung (544). Heparin prevented the increased vascularpressures and the release of serotonin (554), which indicates serotonin is released as a consequence of thrombosis. Prevention of serotonin release with heparin may explain why the increased PVR seen after glass-bead microembolization was greater in nonheparinized than in heparinized dogs, although the same degree of vascular obstruction was present in both conditions (301). Whether the increased serotonin concentration measured in the pulmonary venous effluent is a result of increased liberation or decreased breakdown by the pulmonary endothelium (138,229) has not been examined. Possibly serotonin, which is released after platelet aggregation (90, 346), is not degraded to the same extent as in normal lung during its passage through the pulmonary circuit (229). Serotonin may also be released in such large quantities that the enzyme system responsible for its inactivation is saturated (229). Serotonin also appears to reach the nonembolized lung via the bronchial circulation (112) and to constrict the vessels in the unobstructed lung as well as the vessels closer to the embolized blood vessels. Therefore the most likely humoral mediators of the pulmonary vasoconstrictor response after microembolization are TXA2,’ histamine, and serotonin. A definitive conclusion as to which is the most important cannot be made at this time. b) P&eZet aggregation, Because vasoactive substances such as serotonin and TXA2 are associated with the pulmonary vasoconstriction observed after microembolization (41, 88, 476, 507), it is not surprising that the role of platelet aggregation has been extensively studied. Platelets are the major

July

1983

PULMONARY

MICROEMBOLISM

1125

source of the serotonin released after microembolization (461,492,534), and platelet aggregation is associated with the release of other pulmonary vasoconstrictor substances such as the prostaglandins prostaglandin Fz, (PGF,,), TXA2 (511-513), and histamine (432, 448, 500, 504). The involvement of platelet aggregation in the pulmonary vasoconstrictor response has been examined by injecting collagen fibrils into one lung to produce unilateral platelet aggregation, which resulted in a diversion of flow to the normal lung (40). Flow redistribution was prevented after producing thrombocytopenia with antiplatelet serum (40), indicating a direct involvement of platelets in the response. Also, the increased PVR seen after microembolization was reduced by thrombocytopenia (36, 41, 335) and by both indomethacin (476, 500) and aspirin (304, 403), which inhibit platelet aggregation by preventing the generation of TXAz (144, 207). In another related study, twice as many glass-bead microemboli had to be injected into the pulmonary circulation of thrombocytopenic sheep to produce the same increase in PVR caused by the beads in normal sheep (36). These findings indicate the importance of platelets in producing the pulmonary vasoconstrictor response seen with microembolization. c) Active pulmonary vasodilation. Microembolization sometimes causes pulmonary vasodilation rather than constriction (244). Vasodilation has been observed in a lobe perfused at a constant flow rate when the remaining lung has been embolized (244). Vascular resistance may decrease over time in an embolized lung segment because small emboli may gradually be pushed into smaller vessels, resulting in a decreased obstruction of the larger vessels in the pulmonary vascular bed. The resultant decrease in total PVR may erroneously be thought to represent pulmonary vasodilation. However, in the study by Kealey and Brody (244), vasodilation occurred in the segment of the lung that was not embolized, and the vasodilation was prevented by cervical vagotomy. Atropine had no effect, indicating that the response to microemboli was due to vagal afferents. The response was also inhibited by hexamethonium, sympathetic denervation, and propranolol (244), indicating it was associated with a vasodilator component of the sympathetic efferents, possibly P-adrenergic receptors. Like the pulmonary vasoconstrictor response, vasodilation depends on the size of the emboli: vasodilation was not observed with beads averaging 130 pm in diameter but was seen with beads 64 pm in diameter (244). The receptors involved in this reflex have not been identified, although baroreceptors with vagal nerve afferents have been found in the bifurcation of the main pulmonary artery and in the proximal portions of right and left pulmonary arteries (81,84,364). But stimulation of pulmonary baroreceptors by distension of a pulmonary artery with a nonocclusive balloon produces pulmonary vasoconstriction, not vasodilation (204, 230). It is therefore necessary to invoke another set of receptors in the small precapillary vessels that are located upstream from the microemboli and that evoke the reflex vasodilation.

1126

ASRAR

B. MALIK

t%hne

6.9

d) Efects on pulmonary vascular compliance and blood volume. Few observations have been made concerning the effects of pulmonary embolization on overall or regional pulmonary vascular compliance. Alpert et al. (4) assessed “pulmonary vascular compliance” by measuring the ratio of pulmonary blood volume to mean Ppa. Embolization of a large pulmonary artery with a balloon catheter caused a greater decrease in the ratio than embolization of smaller pulmonary arteries with glass beads 100 pm in diameter. The ratio returned to base line 30 min after microemboli injection, whereas the decrease was sustained after balloon embolization. The differences in the compliance observed after microembolization might have been due to obstruction of small arteries downstream from the large capacitance vessels; thus compliance was not as severely restricted as after embolization of the large pulmonary arteries, which primarily obstruct the compliant portion of the pulmonary vascular bed (4), that is, the larger pulmonary arteries. The transient nature of the decrease in pulmonary vascular compliance after microembolization (4) suggests that the compliance change, like pulmonary vasoconstriction, is humorally mediated. The humoral factors responsible for the decreased compliance may be the same as those mediating the increased pulmonary vascular resistance associated with microembolism (i.e., TXA2, serotonin, and histamine). e) Regional pulmona77/ blood&w. Pulmonary microembolization alters the regional distribution of pulmonary blood flow, but the redistribution pattern depends on the region of the lung embolized. Glass beads tend to obstruct primarily the more dependent pulmonary vessels because the flow to this region is greater than that to the upper lung (301, 303). Also the beads tend to gravitate toward the dependent lung regions because they are heavier than the formed elements (206). The blood flow is redistributed away from the dependent lung regions owing to preferential embolization of the dependent lung and localized vasoconstriction occurring in regions of the lung receiving the emboli (39, 40). Air emboli, however, preferentially embolize the upper lung regions, increasing the blood flow to the dependent portions of the lung (67). Microemboli with specific gravities similar to those of formed elements (e.g., fat microemboli and microthrombi) are distributed relative to the regional blood flow without a redistribution of blood flow (24, 300). Therefore the patterns of blood flow observed with microemboli seem solely dependent on the specific gravity of the embolic material relative to that of the blood-formed elements. The major consequence of flow redistribution after emboli is the effect on matching (or mismatching) of alveolar ventilation @A) and perfusion (a). Because decreased ratios of ventilation to perfusion (i.e., decreased VA/ & values; see sect. VI) are the major cause of hypoxemia after microembolization (see sect. IV), it follows that flow-redistribution patterns resulting from different forms of microemboli produce regional VA/Q values that cause varying degrees of hypoxemia.

July

1983

PULMONARY

1127

MICROEMBOLISM

Decreased pulmonary perfusion persists even after dissolution of the microthrombi (101, 108, 417). This has been observed in coronary (76, 545), cerebral (64), and renal (143) vascular beds after periods of vascular occlusion. Swelling of capillary endothelial cells induced by short periods of occlusion (261, 286, 545) was hypothesized to cause the elevated resistance during reperfusion. A period of pulmonary vascular occlusion may similarly injure the microvascular endothelial cells and contribute to abnormal regional perfusion, but this has not been determined in the pulmonary circulation. III.

PULMONARY

EDEMA

AFTER

MICROEMBOLIZATION

In the lung an increased extravascular water content is commonly observed after embolization of pulmonary microvessels (302, 376, 437). The following section reviews the basis of the edema associated with microembolization, discusses the site of fluid exchange and accumulation, and presents the mechanisms believed to be responsible for edema formation. A. Efebts of Pulmonary

Microembolism

on Lung Fluid

Balance

Two factors have been proposed to explain the development of pulmonary edema after microembolization (376,437): I) an increased microvascular hydrostatic pressure and Z) an increased microvascular permeability to endogenous plasma proteins. The pressure is elevated in some microvessels if the increased Ppa associated with pulmonary vascular obstruction is transmitted to unobstructed microvessels (154,155). The increase in microvascular pressure (Pmv), if it occurs, has not been measured by direct techniques in either unobstructed or obstructed regions of the lung (286). Pulmonary edema may also result because the pulmonary microvessel permeability to proteins is increased either by mechanical injury of the microvessel wall or by the release of some humoral or cellular substance that acts directly on the microvascular permeability (50, 51, 471). 1. Pulmonary

fluid and protein

exchange

a) Shep studies. Pulmonary lymph (i.e., a measure of the net plasma filtrate) can be collected from the efferent duct of the caudal mediastinal node draining sheep lungs (126, 470). Because the changes in pulmonary lymph flow (&lYm) and the protein concentration of the lymph can subsequently be measured, this experimental model has been useful in describing the events that cause pulmonary edema. Microembolization of the lung invariably increases the transvascular filtration of fluid in lungs even though

1128

ASRAR

B. MALIK

vohne

6s

a significant portion of the pulmonary vascular bed has been obstructed and cannot contribute greatly to Qlyrn (303, 311, 376). An increased vascular surface area available for filtration cannot explain the increased capillary filtration, because microembolism obviously decreases filtering area (303). The increased pulmonary Pmv also fails to explain the increased transvascular fluid filtration, because Qlym increased after embolization, whereas the ratio of the lymph-to-plasma protein concentration (L/P) remained unchanged (376) or even increased (303). Figure 2 shows the effects of glassbead embolization on Qlyrn and on albumin and globulin L/P values in a sheep. A similar increase in Qlyrn induced by an increase in Pmv in a normal lung was invariably associated with a decrease in L/P (Fig. 3, A and B; 126, 331), a finding quite different from that noted after microembolization (Fig. 2). Therefore the relatively large increase in transvascular protein clearance (&,, X L/P) after microembolization probably represents an increased permeability of pulmonary microvessels to plasma proteins. That pulmonary microembolization increases lung vascular permeability was clearly demonstrated in other studies by increasing Pmv after microembolization (294, 375). When the capillary wall is more permeable to proteins, a given increase in Pmv results in greater increases in Qlym and in transvascular protein clearance than those seen in normal lung capillaries (488). This change occurs after pulmonary embolization induced by either air bubbles (375) or thrombin (294). Figure 4 shows that the increase in. Pmv induced by inflation of a left atria1 balloon after thrombin resulted in a large SHEEP ‘3 FIG. 2. Time course of effects of progressive pulmonary microembolization on pulmonary lymph flow (&,) and ratios of lymph-to-plasma concentrations (L/P) of albumin and globulin in sheep. Injections of glass-bead microemboli are indicated by El, EZ, and ES. Increase in pulmonary vascular resistance (PVR) after El is small, reflecting injection of few emboli, and PVR increases progressively as bed is further embolized (E2 and ES). Increase in transvascular protein clearance ( Qlym X L/P) is relatively large after El despite minimal increase in PVR. [From Malik and van der Zee (303).]

.8Op .60y

ALBUMIN

HOURS

July

1983

PULMONARY

1129

MICROEMBOLISM

ISHEEP C-3

1

10 ,-- ---a---0t -60

----r----l

_____

J-----=----1

______

i

I

0 4 Pla

60 MINUTES

120

I80

I”L-0

IO 20 30 Pulmonary mi~rovascular pressure, cmHzO

40

FIG. 3. A: effects of raising left atria1 pressure (Pla) in sheep by inflating left atria1 balloon. Qlym, pulmonary lymph flow; L/P, ratio of lymph-to-plasma concentrations of protein; CL, transvascular protein clearance ( Qlym X L/P); Ppa, mean pulmonary arterial pressure; Pla, mean Pla. Increase in Pla increased Qlyrn and decreased L/P because of greater movement of water than of proteins across microvessels. B: relationship between calculated pulmonary microvascular pressure (Pmv) and L/P; Pmv = Pla + 0.4 (Ppa - Pla) (where Ppa is mean left pulmonary arterial pressure); 0.4 = fraction of total resistance in downstream pulmonary vascular segment. [From Minnear et al. (331).]

increase in Qlym without a change in L/P, providing further proof that obstruction of pulmonary microvessels increases pulmonary vascular permeability. Because these observations in sheep involved lymph collected from the efferent lymph duct of the caudal mediastinal node (126,470), it is important to understand the assumptions involved in using this preparation to assess vascular permeability. A major assumption is that the lymph and interstitial fluid protein concentrations are similar. Guyton et al. (168) have argued that this assumption may not be tenable because a positive hydrostatic pressure in the terminal lymphatics could result in the movement of water out of the lymphatics into the interstitial tissue, thereby increasing the lymph protein concentration. Also the protein concentration can be altered as the lymph passesthrough the caudal mediastinal node (413), because water and solutes may be exchanged across nodal blood and lymph vessels (11,413). This mech-

1130

ASRAR

I I

B. MALIK

THROMBIN

vohne

6.9

CONTROL

,

-60

0 4 THROMBIN

60

120 MINUTES

180 f PIa

240

300

4. Effects of thrombin on pulmonary lymph flow (&,,), ratio of lymph-to-plasma protein concentration (L/P), transvascular protein clearance (CL), mean pulmonary arterial pressure @pa), mean left atria1 pressure @la), and pulmonary vascular resistance (PVR). Increase in PVR after thrombin-induced microembolization increased QIYm and L/P. Increase in Pla induced by balloon further increased &1, but without significant change in L/P. This contrasts with effects of left atria1 hypertension in normal lungs (Fig. 3, A and B), indicating that thrombin increases pulmonary microvascular permeability to proteins. FIG.

anism could be particularly crucial when lymph flow rates are normal or low. However, it is unlikely that the lymph concentration changes during high flows. If lymph protein concentration is altered by either of these mech-

July

1983

PULMONARY

MICROEMBOLISM

1131

anisms, any conclusions regarding endothelial permeability changes after microembolization may be in error. Despite these possibilities, there is no experimental proof to indicate that the pulmonary lymph protein concentration is different from interstitial fluid protein concentrations (11). The lymph and interstitial protein concentrations were similar when edema was produced either by increased Pla or by increased permeability associated with Pseudomonas bacteremia (522-524). There was, however, a great deal of scatter (522), possibly because the interstitial samples were from discrete regions, whereas lung lymph represents an integrated sample. The most reasonable conclusion from these studies is that afferent lung lymph and interstitial free fluids are identical. Autoradiographic assessment of albumin concentration in different-sized lymphatics in mouse lung (359) and in sheep lung (N. C. Staub, unpublished observations) indicate no major differences in protein concentrations between small and large lymphatics. There is another problem related to the interpretation of data obtained from the procedure described above. The efferent duct of the caudal mediastinal node in sheep drains only about two-thirds of the posterior lung (376, 470). However, as discussed in sect. IC, microemboli may obstruct selective regions that may be different from the lymph drainage sites (67,301, 303). If the site of embolization differs from the lung region that is drained by the pulmonary lymphatic, the lymph flux data does not reflect events in the embolized portion of the lung. But in the glass-bead studies, the posterior region of lung was embolized 301), and this was also the region from which lymph was collected (376,470), indicating that the lymph data reflect the permeability increase in the region of the lung receiving the emboli. A final problem in using the sheep preparation is possible contamination of pulmonary lymph by extrapulmonary sources, such as the diaphragm and esophagus (117). An increased systemic venous pressure induced by pulmonary vascular *obstruction could increase the lymph flow by increasing the diaphragmatic or esophageal inputs into the lymphatic; however, this should be associated with a decreased L/P rather than with the unchanged or increased L/P observed after microembolization (303, 376). Moreover an increased permeability was evident after microembolization even in paralyzed animals (303, 376); the effect of diaphragmatic contractions on lymph flow was minimized in these animals (117, 422). Also, raising Pmv should produce a greater pulmonary filtrate component in the lymph and thus should minimize any contribution from systemic sources (375); therefore the observation that the increased Qlym after an increase Pmv in embolized lungs was associated with an unchanged L/P clearly indicates an increased permeability (294, 375). b) Dog studies. An increased lung vascular permeability has also been demonstrated in dogs after thrombin-induced pulmonary microembolization. Some of these data are difficult to interpret, however, because the right duct lymph was used to assess lung vascular permeability (44). In dogs at least 40% of the lymph in the right duct comes from extrapulmonary sources (521).

1132

ASRAR

B. MALIK

Vohww

63

A more direct indicator of increased lung vascular permeability in this study was the finding that the protein concentration of tracheal edema fluid after thrombin embolization approached the plasma values (44). Minnear and colleagues (331a) showed by means of a prenodal hilar lung lymph preparation in the dog (488; A. E. Taylor and J. C. Parker, manuscript in preparation) that thrombin embolization produced an eightfold increase in protein-rich lymph flow. This finding supports the sheep data even though in the sheep studies Qlyrn never increased more than 300-400% after similar increments in PVR *with thrombin (303, 375). Thrombin decreased the plasma protein reflection coefficient [a measure of microvascular barrier restriction to flow of plasma proteins (471)] in dog lungs from a normal value of 0.65-0.70 to 0.49 (331a), indicating an increase in pulmonary microvascular permeability to proteins. In another study the extravascular lung water content was increased in dogs by 60-80s within 1 h after glass-bead embolization that produced a 35 mmHg increase in Ppa (302). Although the extent of alveolar flooding was not determined, this probably occurred in most of the animals, because increases in the extravascular water content of SO% are usually associated with accumulation of fluid in air spaces (295). The rise in Ppa was clearly not responsible for the rapid and severe edema because a similar increase in pressure did not produce edema within the same time period (167). The most likely explanation for the edema formation, which substantiates the data obtained from sheep lymph, is an increased endothelial permeability associated with microembolization. c) Studies of isolated and perfused lungs. The capillary filtration coefficient (CFC) and fluid filtration rates of isolated and perfused cat lungs were increased after microembolization induced by collagen fibrils (510). The increases were independent of Ppa alterations-the changes persisted when the pressure rise was prevented by pretreatment with a pulmonary vasodilator, papaverine (510). Because the increases in CFC and filtration rate were independent of any pressure rise and occurred even as vascular surface area decreased, it can be concluded that lung vascular permeability probably increases after microembolization. B. Eflects of Increasing

Degree of Embolixation

Malik and van der Zee (303) examined the effects of increasing amounts of pulmonary microembolization in sheep. Pulmonary lymph flow and L/P increased after only a 30% rise in PVR, as shown by the initial step increases in Qlym and L/P in each animal (Fig. 5; 303). This suggests that lung vascular permeability was increased and that it was independent of the increased Pmv because it was associated with a 2 to 5-mmHg increase in Ppa. This study also indicates that a minimal degree of pulmonary vascular obstruction is required to increase vascular permeability, because the increase was evident with only a 30% rise in PVR.

July

1983

PULMONARY

1133

MICROEMBOLISM

FIG. 5. Relationship between steady-state alterations in pulmonary lymph flow (QIYm) after glassbead pulmonary microembolization in 6 sheep and ratios of lymph-toplasma concentration (L/P) of globulin, albumin, and total protein. As . Qlym increased after first embolization (i.e., initial change in Qlyrn), L/ P also increased, but L/P did not change after further increases in . Qlym as lung was further embolized. After third embolization, Qlym actually decreased in some cases. [From Malik and van der Zee (303).]

1

I

I

1

I

I

0

J

5

IO

15

20

25

30

Q lym

3 ml/

hr

As the vascular bed was further embolized so that PVR was elevated by 130% from base line, Qlym increased to a higher level, while L/P remained elevated (303). Figure 5 shows this relationship between Qlym and L/P after the second embolization. The increase in Qlym was not due to the concomitant increase in pressure, because an increase in Pmv would have resulted in ultrafiltration and in a decreased L/P (126, 331). Ohkuda et al. (375) have also observed that increases in Qlym and transvascular protein clearance after air embolization were proportional to the degree of obstruction. After microembolization, however, a point is reached at which increasing the degree of obstruction does not produce a further rise in Qlyrn or protein clearance. In studies with glass beads this occurred when emboli obstructed -70% of the pulmonary vascular bed (303). In some cases Qlym actually decreased after the final embolization (Fig. 5). The failure of Qlym to increase with further vascular obstruction is probably due to a severe reduction in the vascular surface area, which counteracted the increased permeability effect. Ohkuda et al. (375) demonstrated this point by showing that obstruction of diaphragmatic lobes, which are primarily drained by the efferent duct lymph flow in the sheep, decreased the rate of Qlym, whereas the same degree of obstruction of upper lobes

1134

A

ASRAR

B. MALIK

Air Infusion f .-.-.f I

1

Vohme

63

t'i

Lting L mph i! ow (ml/hr)

20

0

1

0

1

I

2

I

I

I

4

1

6

1

1

8

1

I

10

HOUrS

FIG. 6A. Effects of pulmonary embolization with air on pulmonary lymph flow. In each case a similar degree of embolization was induced in the same sheep, whereas duration of air infusion was different. Increase in lymph flow is related to duration of air infusion, and increase in each case is reversible. Recovery time was prolonged in proportion to duration of air infusion. [From Ohkuda et al. (379.1

increased Qlym. Thus the lack of change or even the decreases in Qlyrn and protein clearance seen after severe microembolization are due to decreased vascular surface area. All studies must take into account the effect of decreased vascular filtration area associated with microembolization. C. Reversibility

of Increases in Permeability

Vaage et al. (510) observed that the increased fluid filtration measured in isolated and perfused cat lungs after collagen-induced microembolization lasted only 30 min, indicating the effect was reversible. In addition Ohkuda and . colleagues (375) have shown in sheep that the duration of the increased Qlym and transvascular protein clearance depended on both the duration of the embolization period and the severity of vascular obstruction. Figure 6A

July

1983

PULMONARY

1135

MICROEMBOLISM

PPA km Hz01

Air Infusion

-.-

50 40

0 1

0

1

1

2

I

1

4

1

1

6

1

1

8

I

1

10

Hours FIG. 6B. Dose response of lung lymph flow ,to pulmonary embolization with air infusion for 1 h. Both peak lymph flow and recovery period are proportional to obstruction. Dashed line, 60% increase in PVR; dotted and dush,ed line, 100% increase in PVR; solid line, 200% increase in PVR; Ppa, mean pulmonary arterial pressure. [From Ohkuda et al. (375).]

shows the effects of a similar degree of embolization (i.e., effects of similar increments in PVR) but of a varying duration of air infusion. Figure 6B shows the effects of varying the degree of obstruction but at a constant duration of air infusion. In Figure 6A the period of increased Qlym is different despite similar vascular obstruction, indicating that the duration of embolization is an important determinant of reversibility. In addition the degree of vascular obstruction also determines reversibility because a greater obincrease in Qiyrn than struction (Ppa = 50 cmHzO) produces a longer-lasting less-marked obstructions (Fig. 6B). The rapidity with which lung vascular permeability returned to normal after microembolizaGon suggests there were no gross morphological changes in the microvessel wall (92,201); in contrast, the endothelium is denuded and the vascular injury is irreversible when damage is induced by more toxic agents such as oleic aeid or alloxan (92, 201). The ability of the lung to recover from embolic damage is further sub-

1136

ASRAR

B. MALIK

Vohne

63

stantiated by the finding that pulmonary edema formed at 2 h after embolization was ~50% of that measured at 1 h after the same degree of embolization (295). The decreased extravascular lung water content was probably not a result of lowered Pmv because Ppa was the same. Also lymphatic removal of extravascular lung water during this period is small, considering that the maximum increase in the rate of QIYrn is only 20-25 ml/h (303, 375). The most plausible explanation for the reduction in edema is an actual time-dependent reduction in capillary wall permeability. Because cellular and humoral systems have been postulated as the factors responsible for the membrane changes (302, 437), dissipation or inactivation of these mediators may cause the reversible increase in permeability. The decreased pulmonary edema measured at 2 h postembolization appeared to be related to less intravascular coagulation, as compared with the first hour (295). This finding supports the notion that the decreased permeability with time is somehow related to local concentrations of blood-borne factors that alter vascular permeability. D. Murphological

Alterations

in Lung Endothelium

The interendothelial junctions in pulmonary microvessels were dilated after thrombin-induced pulmonary microembolization in dogs (92,240,447). This change in vascular leakage sites can explain the increased transvascular protein clearance observed after microembolization in dogs (92) and in sheep (303, 376). Moosavi et al. (338a), in an ultrastructural study, have also emphasized that interendothelial junctions are the primary sites of protein leakage after air microembolization. Moreover after air embolism the lesions were confined to small pulmonary arteries (338a), which are also the sites at which air emboli are lodged (375). Because thrombin-induced microemboli may lodge in arteries, capillaries, and veins, the alterations in interendothelial junctions may even extend downstream from the arteries. In fact, carbon-labeling studies in pulmonary edema induced by either cy-naphthylthiourea (ANTU) or pyrrolizidine indicated that protein leakage occurred only in pulmonary capillaries and veins and to a lesser extent in small pulmonary arteries (99). Carbon labeling was not observed, however, in the larger arteries and veins (99). The fluid accumulation within the spaces surrounding large vessels and airways in the hilum in the’ earlier stages of pulmonary edema reflects the high compliance of these loose connective tissues rather than vascular injury occurring at these sites (99, 486). The interendothelial gaps in capillaries and veins developed’during the period of edema formation but were not detectable when fluid accumulation had ceased (99). These temporary changes in junctional dimension are significant: Poisueille’s equation predicts that even small increases in junctional diameters will markedly alter fluid and protein movement, because fluid flow is directly proportional to the fourth power of the radius of the capillary

July

1983

PULMONARY

MICROEMBOLISM

1137

wall pathways for solute and solvent fluxes. Therefore the development of extensive pulmonary edema does not require wholesale destruction of the microvascular endothelial barrier but can result from the formation of discrete and transient focal lesions. Although at present there is no explanation for the dilation of endothelial junctions after microembolization, cellular and humoral factors are undoubtedly involved in the process. Majno et al. (291,293) described similar junctional widening in skeletal muscle venules after infusion of histamine or bradykinin. Cotran (93,94) noted a similar condition after thermal injury, as did Movat and Fernando (345) after endotoxin challenge. These workers suggested that endothelial open junctions are the sites of protein leakage and that they result from a reversible contraction of endothelial cells containing actin-myosin cellular microfilaments (291, 293). The venular endothelial cells should easily be distinguished from cells at other vascular sites because a) their cell bodies bulged into the lumen, b) their nuclei changed from ovoid to round in appearance, c) folds usually appeared on the abluminal cell surfaces, and d) cellular microfilaments 40-70 A in diameter were present (290). Even though the sites of the small vessels were not localized, similar structural changes also occurred in the pulmonary endothelium after microembolization (92, 338a), indicating that the endothelial cells in the lung exhibit similar contractile responses. In this regard it would be important to examine whether &-adrenergic agonists (such as terbutaline and isoproterenol) reverse the microembolization-induced increase in pulmonary endothelial permeability: Svensjii and co-workers (480) have reported that these’ agonists decrease the histamineor bradykinin-induced permeability changes in systemic venules. Because cellular elements appear to be involved, the effects of other agents such as colchicine and cytochalasin B, which impair cell contraction by disrupting cellular microtubules (25 A in diameter; 552) and subplasmalemmal microfilaments (5-7 A in diameter; 484), respectively, would also be helpful in elucidating the mechanisms responsible’of the increased permeability changes. Another alteration seen in the microvascular wall with microembolization is an increased number of pinocytotic vesicles in the endothelial cells (92, 412). Although several investigators have proposed that protein transport occurs primarily via plasmalemmal vesicles (99, 525), the increased number of vesicles seen during microembolization does not necessarily imply an increased endothelial permeability. A pulmonary Pmv elevation of 15-20 mmHg doubled the density of endothelial cell vesicles (111) but did not increase transvascular protein permeability (470). Because vesicle formation should be an energy-dependent process (70,415,416), studies have also been made in hindlimbs (416) and in lungs (70) that were perfused at 15OC to determine the effect of cooling on edema formation and protein transport. In these experiments the transfer of albumin across capillaries was similar at 36OC and at 15°C suggesting that an energy-dependent capillary pinocytosis is an unimportant component of macromolecular transport across

1138

ASRAR

B. MALIK

vdume

63

capillary walls. In addition pulmonary edema formation was not depressed at lower temperatures (70). A more likely mechanism by which pinocytotic vesicles contribute to increased protein transport is by formation of transcellular channels. These channels (100-1000 A in diameter) are produced by fusion of two or more vesicles (412,459). In isolated and perfused frog mesenteric capillaries it has been shown that the pathways are long-lived and that they provide a route for the increased transendothelial protein transport (75). The major difficulty with using the transcellular hypothesis to explain protein leakage sites is the extreme rarity of these structures in capillary walls of mammals. Many investigators have not found transendothelial channels in their preparations (57). In one study Bundgaard (57) did not find vesicular channels even in the very thin sections of the endothelium in over 700 sections! Against this background it seems reasonable to conclude that the increased passage of plasma proteins into the pulmonary interstitium after microembolization occurs primarily through dilated gap junctions and not through transendothelial channels. Nevertheless any future work should include serial sampling to demonstrate the transendothelial channels that may only appear as endothelial vesicles in the usual sectioning procedures. E. Microvessels

Versus Arteries

as Sites of Fluid

and Protein

Exchange

A greater surface area is available for filtration in the pulmonary microvessels (the vessels ~50 pm in diameter) than in the larger vessels; therefore it is not surprising that most of the fluid and solute should occur at this level of microcirculation (470, 473). The situation is altered, however, when pulmonary arteries are obstructed, as after microembolization. Iliff (208) raised either Ppa or Ppv in static blood flow to isolated lungs in which the alveolar vessels had been compressed by increasing alveolar pressure to values much greater than arterial pressure. Under these conditions, this rise in pressure increased filtration. This presumably reflected increased filtration at the level of larger vessels (i.e., extra-alveolar vessels) because the alveolar vessels were obstructed. In another study, fluid cuffs were observed around larger blood vessels after complete obstruction of the pulmonary artery with glass beads (449,538). Albert et al. (1) also demonstrated a steady weight gain in an isolated and perfused lung after increasing Ppa despite obstruction of the lobe with glass beads. The edema could only be explained by an increased transarterial filtration, because the lungs in these studies were completely obstructed by emboli. Although these studies demonstrated that fluid filtration can occur across pulmonary arteries after embolization, this does not mean that pulmonary arteries are the primary sites for fluid and protein leakage after microembolization. The studies cited above represent total obstruction of pulmonary arterial inflow. In most types of microembolization, however, the microvessels are still perfused through collateral branches originating up-

July

1983

PULMONARY

1139

MICROEMBOLISM

stream from the obstruction points (248,254). The small pulmonary vessels can thus participate in fluid exchange in spite of upstream obstructions. The permeability lesions appear to be confined to the small pulmonary vessels (92, 201). Studies in sheep demonstrated increased vascular permeability after embolization of vessels with beads 200 pm in diameter, but beads 500 pm in diameter did not alter vascular permeability (219). F. Hemodynumic

Mechanisms

Associated

With Increased

Permeability

1. Blood velocity The diversion of the entire cardiac output through unobstructed lung regions and the resultant decrease in the cross-sectional area of the pulmonary vascular bed increases the blood velocity after microembolization. Ohkuda et al. (376) proposed that the increased velocity and the resultant increases in tangential and shear stresses after microembolization could injure the pulmonary endothelium. This mechanism had previously been shown to damage walls of systemic vessels (110, 157). Damage to the aortic endothelium, which was subjected to high shear rates, was shown to be associated with “leaky endothelium” (119, 149, 150). Yet the evidence does not support this hypothesis as the cause of microvascular damage in the lung. The degree of pulmonary vascular obstruction, as reflected by increased PVR, was not correlated with the degree of pulmonary edema (59). Moreover the pulmonary vascular permeability increased with an increase in resistance as small as 30% above base line (303), where changes in blood velocity were minimal. The most complete study concerning velocity effects is that of Landolt et al. (256). These researchers examined the independent effect of blood velocity on lung fluid and protein exchange in sheep by resecting 65% of the lung and perfusing the remaining lung with the entire cardiac output. The . lym increased but L/P decreased, indicating that increased blood velocity Q produced hydrostatic rather than permeability edema. Therefore increased blood velocity and increased wall shear stress can be ruled out as causes of the increased permeability associated with microembolization. 2. Microvascular

pressure

A hemodynamically induced increase in capillary pressure can increase endothelial permeability to macromolecules (139, 165). Shirley et al. (458) used dextrans of known particle size and showed that only small-molecularweight dextrans escaped from the bloodstream at normal blood volumes; the larger particles were lost from the blood and appeared in the lymph when blood volume was increased. Schneeberger (446) observed alterations in interendothelial junctions in the mouse lung after injection of saline in a

1140

ASRAR

B. MALIK

Vdum

63

volume of 0.5 ml with horseradish peroxidase (Mr 40,000). The tracer passed rapidly between the endothelial cells of pulmonary capillaries with this volume but not with a volume of 0.1 ml. Fishman and Pietra (140) observed that raising Pla to 30 mmHg in isolated and perfused lungs produced a pressuredependent leakage of both horseradish peroxidase and hemoglobin into the interstitium; however, this may reflect a fragile, isolated lung preparation in which the endothelium may have been injured before the pressure elevation (68). Erdmann et al. (126) found no evidence of increased permeability in pulmonary vessels at Pla values up to 40 cmHzO with lymph flux data from intact sheep lungs. One can safely surmise that high Pmv values are required to produce damage. In the case of the hindlimb, an increased permeability was only evident at capillary pressures between 40 and 50 mmHg (414). The effects of such elevations in pulmonary Pmv have not been examined in intact animals because it is difficult to maintain adequate blood flows at the pressure levels required to produce damage. Although increased pressure could injure the endothelium in the lung, it is probably not the primary cause of the increased permeability associated with microembolization: increased permeability was evident even at low Ppa values (303, 375). Although small pressure elevations are clearly not associated with ,endothelial injury in the normal lung (126, 331), the effect of increased vascular pressures in lungs with an already damaged endothelium has not been examined. G. Lymphatic

Impairment

Rusznyak et al. (424a) and Staub (470) have outlined the role of the extensive pulmonary lymphatic system in removing fluid and plasma pro? teins that leak into the interstitium (354), and Halmagyi (177) has suggested that lymphatic failure plays a major role in the development of pulmonary edema. Surgically autotransplanted lungs (95) and lungs in which lymphatics had been removed (288) became edematous within days, indicating that the lymphatic loss prevents drainage of interstitial fluid and protein. Systemic venous hypertension also contributes to pulmonary edema (327), presumably owing to a decreased pressure gradient for QIYrn into the major veins. All these factors could operate after microembolization. The pulmonary lymphatic function could be compromised after microembolisation because electron-microscopic evidence indicated an accumulation of fibrin within the alveolar-capillary septum (92, 158). This accumulation may interfere with uptake of fluid and solutes by lymphatics [i.e., it may cause increased tissue can also occur resistance to QIYm (168, 322)]. Systemic venous hypertension as a result of pulmonary vascular obstruction and thus could decrease lymph drainage because of a smaller pressure head between lymph vessels and venous outflow. The base-line flow rate of pulmonary flow in sheep is -5 ml/ h (126). A 6-h period of total obstruction would therefore result in a 30-ml

July

1983

PULMONARY

MICROEMBOLISM

1141

increase in the extravascular fluid volume, which represents only 10% of the total extravascular fluid volume in a 20-kg sheep lung. Thus complete obstruction of the lymphatics would produce only slight interstitial edema within 6 h and cannot explain the near doubling of extravascular water content within l-2 h after microembolization in dogs (295, 302). H. Eflects of Regional

Atelectasis

Regional atelectasis occurs after pulmonary microembolization as a result of airway closure and surfactant loss (468). Pang et al. (387) observed that atelectasis reduced the degree of edema formation in the lung. The decreased fluid accumulation was believed to be caused by increased perimicrovascular pressure (468). Another consequence of regional atelectasis is that adjacent normal segments become hyperinflated owing to the interdependent effects (319, 393), and a more negative interstitial hydrostatic pressure develops. The latter provides a large pressure gradient for fluid movement into these areas from the adjacent atelectatic segments (387). Regional fluid shifts in the lung after microembolization have not been examined, but greater edema may be present in hyperinflated regions than in the collapsed regions. I. Cellular

and Huwwral

Mechanisms

Recent evidence has favored cellular and humoral factors as the mediators of the increased vascular permeability and pulmonary edema associated with microembolization. Before discussing the specific mediator, it is important to consider how the mediators reach the sites of fluid exchange in small pulmonary vessels. Although the sequence of events from mediator release to endothelial injury is poorly understood, there are three possible pathways that mediators could utilize to reach the sites of injury in pulmonary microvessels. First, terminal patent arteries ranging from 200 to 300 pm in diameter give off branches at right angles that form collateral pathways with downstream microvessels (as shown at left in Fig. 7; 248,249, 254,466). In this way blood flow can persist in these lung segments, and the putative factors may reach the microvessels despite obstruction of the main inflow vessels with microemboli. Second, small-molecular-weight factors can diffuse directly through the lung parenchyma to the nearby patent microvessels; therefore the adjacent vessels rather. than the obstructed vessels may be the primary sites of the increase in permeability. Third, substances released into the circulation downstream from the emboli may eventually be carried to pulmonary microvessels via the bronchial arterial flow. Because the bronchial and pulmonary vascular beds are connected (56), the putative factors in the bronchial circulation have a direct access to the pulmonary circulation.

1142

200Mm

ASRAR

Emboli

5OOym

Emboli

B. MALIK

vbluww

63

FIG. 7. Effects of emboli of different sizes on pulmonary microcirculatory hemodynamics. With microembolus caused by glass beads 200 pm in diameter (as shown at Z&), collateral vessels allow blood flow to persist in downstream lung segment because of collateral branching pattern. With macroembolus 500 pm in diameter (shown at tight), flow is obstructed upstream from point of vessel collateralization, and flow is diverted through nonembolized regions of lung. This model agrees with concepts of Kniseley (248) and Krahl (254) concerning pulmonary arterial “catch traps” that allow flow to persist through right-angle collateral vessels. Flow model explains the rapid lysis of microclots as well as vessel recanalization after microembolism and provides mechanism by which subtances released from thrombi, lung tissue, and plasma can increase permeability of downstream microvessels.

The first of the above mechanisms appears to be the most important because embolization with glass beads 500 pm in diameter, which presumably blocks any collateral flow, did not result in an increased pulmonary vascular permeability (219). Figure 7 compares the consequences of obstruction with a ZOO-pm embolus with those of obstruction wi th a 5000pm embol us Because of collate ral flow, downstream m.icrovessels are affected on1.y with th e smaller embolus. If the other two mechanisms discussed above were as important as the first, an increase in permeability should have been evident even after obstruction with the large embolus, but this was not the case (219); therefore the putative factors must reach the microvessels via collateral branches of the pulmonary artery. The following sections .discuss the specific cellular and humoral factors postulated as either mediators of or factors contributing to the pulmonary edema associated with the microembolization. 1. Fibrin

Fibrin “plugs” in pulmonary microvessels are usually seen with pulmonary microembolization that is associated with thrombosis (18, 65, 436438). Saldeen (437,438) hypothesized that fibrin entrapment was the factor primarily responsible for the increased permeability with microembolization (43). According to Saldeen (437, 438), the generation of fibrin-degradation products associated with the lysis of intravascular and extravascular fibrin mediates the endothelial cell injury. In addition to fibrin thrombi in vessels, fibrin deposits were sometimes found near the interendothelial junctions and in the perivascular spaces (15). In a somewhat puzzling observation in lungs obtained from patients who died from the adult respiratory distress syndrome, Bachofen and Weibel

July

1983

PULMONARY

MICROEMBOLISM

1143

(18) observed more fibrin in the interstitium than in the blood vessels. Rather than indicating a peculiar affinity of the interstitium for fibrin, this difference may reflect differences in the ability of the luminal and abluminal endothelial surfaces to lyse fibrin via the fibrinolytic system (418). The following sections review the evidence in support of the hypothesis that fibrin entrapment mediates increased permeability. a) Efects of defibrinogenation and heparin. Because fibrin clots have been implicated in the development of microembolization-induced edema (43’7; 438), several studies have examined the effects of fibrinogen depletion and intravascular coagulation inhibition on this phenomenon. Busch et al. (59) observed in dogs that defibrinogenation, which was produced with a purified extract of Malayan pit viper venom (Ancrod; 31,408), prevented the increase in lung weight associated with thrombin infusion. These data do not directly support the role of fibrin in edema formation because thrombin cannot produce intravascular thrombi without fibrinogen. Thus the absence of pulmonary edema in the defibrinogenated dogs may only reflect the absence of microemboli. To substantiate this interpretation, Malik et al. (298) demonstrated in defibrinogenated sheep that thrombin infusion failed to increase Ppa and PVR as well as Qlym and transvascular protein clearance. Therefore the protective effect of defibrinogenation in preventing increased permeability appears to be a result of the inability of thrombin to produce the microemboli rather than the result of any protective effect of defibrinogenation per se. Johnson and Malik (217) repeated Busch’s study (59) in dogs but produced emboli with glass beads. Glass beads not only produced a permanent obstruction of pulmonary microvessels but also activated the intrinsic coagulation cascade by a direct activation of factor XII (236, 237). Defibrinogenation prevented the pulmonary edema after glass-bead microembolization (217), indicating that the edema was the direct result of activation of intravascular coagulation and that it was not caused by obstruction of pulmonary vessels. Heparin pretreatment in dogs also prevented the pulmonary edema associated with glass-bead microembolization (148, 302). Beyond the inference that intravascular coagulation plays a role in mediating edema formation, these studies are not proof of the direct and primary role of fibrin, because defibrinogenation and anticoagulation prevent both the formation of fibrin microemboli and the subsequent fibrinolysis by plasmin (EC 3.4.21.7) (77,161,237,417). Therefore the inhibition of plasmin synthesis and the resultant failure to activate the complement system may well have brought about the protective effects of defibrinogenation or heparin pretreatment (127, 237). Because leukocyte margination in the pulmonary microcirculation may not occur without complement activation (96, 97, 212), the absence of edema could reflect a lack of leukocyte activation rather than a direct protective effect of defibrinogenation or heparin. Another criticism of this work in dogs is that lung vascular permeability was not assessed. The protective effect of heparin or defibrinogenation seen in dog lungs

1144

ASRAR

B. MALIK

Vdume

63

(217,302) is not reproducible in the sheep preparation (37). Binder et al. (37) showed that neither defibrinogenation nor heparin pretreatment prevented the increased Qlym and transvascular protein clearance observed after glassbead microembolization. The major difference in protocol was the use of siliconized glass beads in the sheep studies (37). Johnson and Malik (ZZOa), however, repeated Binder’s study in sheep and used nonsiliconized beads; and they obtained the same results. It is more likely that species differences exist relative to the coagulation and fibrinolytic systems (9, 116, 151). Fibrinolytic activity in dog lung is much greater than that of sheep lung (9, 151) because fibrinolysis in dogs must be inhibited with tranexamic acid or aminocaproic acid to induce microembolization after thrombin; this step is not necessary in sheep (9, 300). Therefore, for a given degree of pulmonary vascular thrombosis, the more potent fibrinolytic mechanism in dogs is expected to generate a greater increase in plasmin activity and in local levels of fibrin-degradation products, and therefore these products are more likely to produce injury of the dog pulmonary endothelium. The experiments involving heparin and defibrinogenation must be treated cautiously because the effects of heparin and defibrinogenation on fluid accumulation in the lung are more complex than was previously thought (30,216). Heparin pretreatment of dogs did not prevent the pulmonary edema induced by alloxan, but the edema was greater (323). Some researchers have suggested that fibrin in the interstitium of nonheparinized dogs prevented fluid accumulation by decreasing the available interstitial volume (323,486). In addition a recent study by Carr (66) has shown that heparin given intravenously at dosages of 50-100 units/kg suppresses in skin the increased vascular permeability induced by histamine, bradykinin, or prostaglandin Ez (PGE2). Thus the protective antipermeability effect of heparin may be independent of its anticoagulant properties. b) Fibrin-degmdatim pod~~ti. Fibrin-degradation products have been proposed as the primary mediators of the increased permeability associated with microembolization (278, 437, 438). Several experiments point to their importance. The increased vascular permeability observed after thrombininduced pulmonary microembolization is associated with the generation of fibrin-degradation products (221). Embolization with fibrin microaggregates, which do not activate intravascular coagulation or fibrinolysis (221), resulted in the filtration of protein-poor fluid (i.e., a decrease in L/P). The concentration of fibrin-degradation products did not increase to the same extent after injection of fibrin microaggregates, whereas the changes in leukocyte and platelet counts were comparable (221). Therefore the appearance of fibrin-degradation products correlates well with the time course of the increased pulmonary vascular permeability, although the causal relationship between fibrin-degradation products and increased permeability has not been established. Evidence also indicates that fibrin-degradation products of several different molecular weights increase vascular permeability (37, 438). Saldeen

July

1983

PULMONARY

MICROEMBOLISM

1145

(438) has isolated a pentapeptide (Ala-Arg-Pro-Ala-Lys) originating from the NH&erminal portion of the p-chain of fibrinogen and an undecapeptide (Ser-Gln-Leu-Gln-Lys-Val-Pro-Pro-Glu-Trp-Lys) originating from the midsection of the a-chain of fibrinogen; both these compounds increased macromoleculer permeability in the microvessels of hamster cheek pouch. However, the effects of these peptides on pulmonary vascular permeability has not been examined. Studies by Curreri and co-workers (284, 306, 307) have demonstrated that fragment D, a primary fibrin-degradation product, produced interstitial edema and alveolar hemorrhage when infused into rabbits. These changes were attributed to an increased capillary permeability to plasma proteins because greater amounts of labeled albumin accumulated in the extravascular space after the fragment D challenge. The effect of fragment D was specific because fragment E infusion did not evoke the same response (306). Possibly platelet aggregation rather than fragment D per se was responsible for the fluid accumulation and capillary leak in this study because the rabbits became thrombocytopenic during the infusion process (306). The finding that a histamine 1 (HI) blocker, diphenhydramine, prevented the pulmonary edema after fragment D infusion also suggests a role for histamine (306), which is known to be released in large amounts after aggregation of rabbit platelets (116). Fibrin-degradation products ranging from 15,000-25,000 iW” have also been reported to increase the permeability of macromolecules in rabbit skin vessels (479). Therefore, even though fibrindegradation products of varying values of molecular weight are apparently capable of altering endothelial integrity, investigators have not yet determined whether these products are responsible for the changes observed in lungs after microembolization. The mechanism by which these fibrin-degradation products increase vascular permeability to macromolecules is also unknown. Like other mediators of inflammation (such as bradykinin and histamine), they appear to cause endothelial cell contraction (233). A normal monolayer of cultured endothelial cells became rapidly disorganized after layering fibrin clots on the cells (98). The effect was’ reversed after removing the clots. Other cell types did not exhibit the same changes when exposed to fibrin (98), indicating that fibrin and/or its products have specific actions on endothelial cells. Fibrin-degradation products could also induce pulmonary leukostasis because they are known to be leukotactic substances (243,479)..The increased permeability seen with fibrin or its products may simply represent release of toxic substances, such as oxygen radicals and proteases, by activated leukocytes (290). In addition to direct and leukotactic effects, fibrin-degradation products may cause the release of other mediators such as histamine (306), which may then alter pulmonary vascular permeability. c) FibrinoZysti. Because fibrin-degradation products have been reported to mediate pulmonary edema associated with microembolization (437, 438), the rate at which fibrinolysis occurs should be related to the degree of edema. Fibrinolysis is particularly important in the pulmonary circulation because

1146

ASRAR

B. MALIK

Volume

63

pulmonary endothelial cells, like those of cerebral and coronary vessels, possess abundant amounts of tissue plasm inogen activator (417,497,526), which converts plasminogen to plasmin, the potent serine protease (237, 41.7). The finding that most fibr in was cleared from rabbit, rat, and dog lungs wi thin 1 h after intravascular coagulation is a measure of the remarkable ability of the pulmonary fibrinolytic system (60, 61, 295, 434). The role of fibrinolysis in the pulmonary edema formation associated with microemboli has been studied by inhibiting fibrinolysis with either tranexamic acid or e-aminocaproic acid (418). This caused greater fibrin deposition and a greater amount of pulmonary edema (59,295,300), suggesting that depression of fibrinolytic activity enhances the amount of edema formation. Because pulmonary vascular resistance did not increase after administration of the tranexamic acid, the results could not be explained by more vascular obstruction caused by fibrin entrapment (295). Saldeen and co-workers (280,281,438) suggested that the edema formed after inhibition of fibrinolysis is mediated by a gradual gen erati on of fibrindegradation products in embolized regions. This assumes that fibrinolysis is not completely inhi .bited and that a small amou nt of the degradation enters the microcirculation. products continuously Nevertheless this hypothesis should be tested by comparing the effects of complete versus incomplete fibrinolysis inhibition. When fibrinolysis was completely inhibited in sheep (i.e., when generation of fibrin-degradation products is prevented), thrombin failed to incre Nase lung vascu lar permeability (221a), supporting the notion that gradual generation of fibr in-degradation products is necessary. Unlike complete inhibition, microemboli-induced pulmonary edema can be enhanced by partial inhibition of fibrinolysis (59, 295, 300). The same effect of partial inhibition can also occur spontaneously after microembolization (277,279,437,438). Inhibition of fibrinolysis occurring over a period of several days has been proposed as a mechanism responsible for the delayed edema developing in patients after pulmonary thromboembolism (279, 336, 437, 438) and in dogs after thrombin injection (277, 437). The inhibitor is a stable ae-globulin that binds to both plasminogen and plasmin in vitro (438), activator for the suggesting that it acts by competing with the plasminogen binding sites on plasmi .nogen and plasmin. Circulating free fatty acids appear to regulate the synthesis of this inhibitor because the increased levels of free fatty acid caused by norepinephrine administration raised the inhibitor levels (438). Conversely, decreasing the levels of free fatty acid by nicotinic acid administration prevented the increase in inhibitor activity (438). The inhibitor is synthesized in the liver because fibrinolytic activity did not increase after thrombin injection in hepatectomized rats (438). The findi WV that spontaneous and pharmacological partial inhibition of fibrinolysis resulted in pulm .onary edema support the hypothesis that fibrinoly tic acti vity . modulates the degree of edema, perhaps by control .ling the local concen trations of fibrin-degradation products in the pulmonary microcirculation.

July

1983

PULMONARY

MICROEMBOLISM

1147

d) Fibrinopeptides. Other potentially injurious factors released during fibrin deposition are fibrinopeptides A and B, which contain peptides having 16 and 14 amino acids, respectively (25, 365). These peptides are liberated from fibrinogen when it is converted to fibrin by the action of thrombin (379). Bayley et al. (27) demonstrated that bovine fibrinopeptide B and human fibrinopeptide A injected into dogs, rabbits, and lambs caused a decreased lung compliance and an increased resistance in airways and pulmonary vessels; bovine fibrinopeptide A, however, did not have these effects. Although fibrinopeptides have not been investigated relative to lung fluid balance, they increase endothelial permeability of systemic microvessels (91). This permeability-increasing effect of fibrinopeptides may be indirect, because fibrinopeptides enhance the edemagenic effects of bradykinin (290, 142) and also cause chemotaxis of leukocytes (290).

2. Pulmonary

leukostasis

Sequestration of leukocytes in the pulmonary circulation has been implicated as another primary factor in the development of lung vascular injury (96, 97, i42, 195, 211, 218, 407). Embolization of rabbit lung with autologous leukocytes produced pulmonary vascular lesions that were characterized by constriction of pulmonary arteries, endothelial swelling, and periarterial infiltration of polymorphonuclear leukocytes (6). In studies conducted in sheep, lung vascular permeability to proteins was increased after pulmonary leukostasis induced by infusion of zymosan-activated plasma (96, 97). Zymosan activates the complement system and causes the generation of leukotactic fragments, the most potent being the complement fragment C5a (96, 97). Either leukocyte or complement depletion prevented the permeability-increasing effect of zymosan-activated plasma (96, 97). Complement activation and the resultant generation of C5a appears to be a necessary step for pulmonary leukostasis (127, 312, 424). Polymorphonuclear leukocytes are attracted into the air space after instillation of C5a into airways either by a direct chemotactic action or by production of another chemotactic factor from pulmonary macrophages (146). Moreover the infusion of C5a into mice increased the pulmonary microvessel permeability to proteins, and neutropenia prevented the increased permeability (199), indicating that the response depends on the neutrophils. The pathway by which C5a is generated after pulmonary microembolization is not known. The most likely mechanism of C5a generation is via activation of the coagulation and fibrinolytic cascades that occur after pulmonary embolism (236, 237). Both thrombin and plasmin (which are activated during intravascular coagulation and fibrinolysis, respectively) have proteolytic activities (236,237) and thus can cleave the complement proteins. Plasmin is probably more important than thrombin in complement activation, however, because lung vascular injury did not occur after thrombin

1148

ASRAR B.MALIK

Vohrne

6.9

infusion in defibrinogenated sheep (221a). Therefore thrombin per se does not cause the injury; rather the deposition of fibrin and subsequent fibrinolysis are required. Although the evidence presented above implicates leukocytes in lung vascular injury, the question remains -are leukocytes necessary for the increased pulmonary vascular permeability and edema associated with microembolization? When leukocytes were depleted by nitrogen mustard (142) or when granulocytes were depleted by hydroxyurea (22), either transvascul .ar pro ltein clearance did not increase or the increase was attenuated after pulmonary microembolization (22, 142). Depletion of granulocytes by hydroxyurea also prevented the increase in extravascular lung water content observed after glass-bead microembolization (218). The effects of granulocytopenia have been further examined after thrombin (483a), a model known to induce intravascular coagulation and to activate the complement system in (294). In this study, neither Qlyrn nor L/P changed after embolization granulocytopenic sheep, whereas QIYrn increased and L/P was unchanged in control sheep that were embolized to the same degree. Raising the Pla in the granulocytopenic group after thrombin infusion further increased Qlym but decreased L/P (Fig. 8), whereas the increase in Qlym after thrombin administration in control animals was associated with an unchanged L/P (Fig. 8). The slope of the &, vs. L/P line in the granulocytopenic group was not different from the slope in normal animals after left atria1 hypertension (Fig. 8). These studies indicate a protective effect of granulocyte depletion in preventing the increased permeability. Lung vascular injury associated with microembolization may be similar to the injury seen with endotoxemia (186), O2 toxicity (98,146), and acute pancreatitis (483) because these pathological states are also dependenton the presence of white blood cells. Treatment of leukocytes with cytochalasin B, which causes them to their ability to adhere round up and retract their pseudopodia, minimizes to vascular walls and to release superoxide anions (431, 560). In such cases endothelial damage did not occur (560). Therefore vessel wall adherence must be another step necessary for the development of pulmonary endothelial injury associated with leukocytes. of leuhxxyte-mediated a) A4echantim 1) PROTEASES. The intravenous injection of leukocytes produced both lung endothelial lesions (6) and basement membrane changes (215), suggesti ng that lysosom al enzymes I released from leukocytes trapped within the lung contribute to lung vascular injury. The release of proteases from polymorphonuclear neutrophils is believed to occur during phagocytosis (533,535), and release may be pronounced when large numbers of neutrophils are trapped in the pulmonary circulation after embolization. The major enzymes released are neutral proteases, collagenase (EC 3.4.24.3), elastase (EC 3.4.21.11), and cathepsin G (EC 3.4.21.20) (535). The release of proteases is mediated via the activation of the alternate complement pathway (115), which may occur after the microemboli-induced coagulation process. A ntiproteases failed to protect against the immune com-

July

PULMONARY

1983

MICROEMBOLISM

1149

0.90.8-

0 P

aa

0, z

0.70.8OS0.4-

5

20

15 PULMONARY

LYMPH

FLOW

( ml I hr )

FIG. 8. Relationship between pulmonary lymph flow and ratio of lymph-to-plasma protein concentration (L/P) after raising pulmonary microvascular pressure (Pmv) by inflation of left atria1 balloon. Responses are shown for normal lungs, for lungs after embolization with thrombin, and for lungs after embolization with thrombin in granulocyte-depleted sheep. In normal lungs, L/P decreased after raising Pmv; however, L/P did not decrease when Pmv was increased after thrombin. Raising Pmv in granulocyte-depleted group after thrombin administration resulted in increase in lymph flow and associated decrease in L/P. The latter changes were consistent with normal sieving of proteins in these lungs and differed from increased microvascular permeability to proteins evident in untreated animals after thrombin administration.

plex-mediated lung vascular injury, whereas superoxide dismutase (EC 1X.1.1) completely prevented the injury (224), indicating the importance of superoxide anions rather than proteases. The release of myeloperoxidase from polymorphonuclear sites triggered by C5a was minimal, and no correlation existed between the degree of myeloperoxidase release and the extent of endothelial damage in cultured endothelial cells (431). These observations suggest that this enzyme is not a factor in the genesis of vascular injury. , II) SUPEROXIDE ANIONS. Sacks et al. (431), McCord and co-workers (395, 439), and Del Maestro et al. (113) have suggested that leukocytes injure the microvessels primarily by the production of superoxide anions during bursts of metabolic activity associated with phagocytosis. The pathways of O2 reduction are shown in Figure 9 (147, 395). It is well established that oxygen radicals serve a bactericidal role (147, 395); however, the concept that oxygen radicals are important in the pathogenesis of endothelial cell damage is relatively new. In a preliminary report, Flick et al. (141) showed that superoxide dismutase, which protects against deleterious actions of the oxygen radical 0; on cell walls by catalyzing its

1150 02

ASRAR e-

-

02-

e-

+ 2H+

B. MALIK

vi.&mte

.“202.22GTo”.z, H20

H20

02 - + 02 -+ H202

+ Hz02

H202

+ RH2

FIG. 9. Univalent

radicals

OH

l

2H+

6.9

-

t

Hz02

+ 02

Superoxide

2H20

+ 02

Catalases

2H20+R

pathways of oxygen reduction , 02, and H202. [From Fridovich (147).]

dismutases

Peroxidases

and enzymatic

defenses

against

the oxygen

dismutation to Hz02 and 02, prevented the increase in transvascular protein clearance associated with air embolization (Fig. 9). However, superoxide dismutase was protective only when it was combined with heparin (141). The reason for this unusual action of heparin is unclear, but the finding nevertheless suggests that 02 contributes to the microemboli-induced increase in lung vascular permeability. Another toxic species, H202, is normally removed by catalase, which converts it to HZ0 and 02, and by peroxidase (EC 1.11.1.7), which reduces it to HZ0 (Fig. 9). If the first two intermediates of O2 reduction, 0, and,HzOz, are effectively removed, the formation of a third potentially more toxic species, the hydroxyl radical OH. [generated by the Haber-Weiss reaction 0, + H202 4 O2 + OH. t OH- (147)] is preventable because the Haber-Weiss reaction depends on 0, + H202 substrates. Because of the potential of generation of these toxic oxygen radicals, there is a need to determine the relative contributions of 02, H202, and OH in mediating lung vascular injury after microembolization. The enzyme responsible for superoxide production by neutrophils is the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase located within the membrane of the cells (16,313,318). This enzyme oxidizes NADPH on the cytoplasmic side and converts O2 to 0, on the-exterior surface of the membrane (16,313,318). Therefore most of the 0, is produced on the exterior plasma membrane (147,313), where there is little or no superoxide dismutase, and thus may injure endothelial cells if produced in large quantities. Superoxide anions (OS, H202, and OH.) interact directly with tissue components by oxidizing or reducing the structurally important tissue elements such as cell walls (147). In the intracellular environment, the level of 0, and H202 are controlled by superoxide dismutase, catalase, and peroxidase (147, 314). Because these enzymes are not normally present in any appreciable quantities in the extracellular space, the rapid generation of 02 or H202 into the extracellular fluid may be an important determinant of lung vascular injury. There is ample evidence that superoxide anions injure cell membranes when their local concentrations are not reduced by scavenger enzymes such l

July

1983

PULMONARYMICROEMBOLISM

1151

as superoxide dismutase or catalase (EC 1.11.1.6). The instillation into rat lungs of xanthine and xanthine oxidase (EC 1.2.3.2), which generate superoxide anions, produced acute pulmonary injury (223). The injury was prevented in the presence of superoxide dismutase (223), which indicates that the injury is the result of 02 generation. The causative factor for the production of 0, by leukocytes is not known. Superoxide production may occur by the direct action of C5a (19,313) or by the action of serine proteases such as plasmin (247). Administration of C5a or plasmin in vitro caused the generation of 0; by polymorphonuclear leukocytes, and the effect of plasmin was inhibited by serine protease inhibitors (247). Because both C5a and plasmin are generated after microembolization as a res ult of complement and coagulation activation, they may be important factors in mi croembolism-ind uced lung vascular injury. Therefore plasmin activation and the generation of C5a appear to be important not only in leukocyte margination after microembolism but also in leukocyte activation and generation of superoxide anions. III) OTHER MECHANISMS. Leukocyte-mediated endothelial injury may also cause release of vasoactive hormones such as histamine from basophils (46,163). However, histamine does not produce a large increase in pulmonary induced by glassvascular permeability (53). A lso the increased permeability unaltered with a ntihistamines (257), indicating bead microembolization was the relative unimportance of histamine. Finally, leukocytes are often seen between endothelial cells and the basement membrane and even between the interendothelial junctions, probably caught in the process of transmigration (527). Because they can migrate into these strategic sites, any alterations in leukocyte shape that are due to swelling or shrinkage may mechanically disrupt the junctions and increase the permeability of the vessels to fluid and macromolecules. IV) LEUKOCYTE-PROSTAGLANDININTERACTION. Prostaglandins are active participants in the inflammatory response of leukocytes in the systemic microcirculation. Wedmore and Williams (527) have formulated a two-mediator hypothesis to explain the leukocyte-mediated permeability increase of skin blood vessels. The first mediator, probably C5a, increased vascular permeability by inducing leukostasis and leukocyte activation. The second mediator produced by endothelial cell membranes, either prostacyclin [prosthe plasma protein leakage by taglandin I2 (PG12)] or PGE2, potentiated inducing relaxation of arteriolar smooth muscle and thereby increasing capillary pressure (253, 527). Little protein exudation was observed when prostaglandin synthesis was inhibited with indomethacin (256). This was explained by the fact that synthesis of PG12 and PGEz was inhibited and no vasodilation occurred. Prostaglandins may interact in a similar fashion in the pulmonary microcirculation because inhibition of prostaglandin synthesis with indomethacin or ketoprofen prevented the thrombin-induced increases in pulmonary vascular permeability (514). It is difficult, however, to reconcile the reported enhancement of the leukocyte-mediated inflammatory

1152

ASRAR B.MALIK

Vohme

63

response by PG12 or PGE2, because the same prostaglandins depress leukocyte chemotaxis (369). A more likely mechanism by which prostaglandins contribute to the leukocyte-dependent increase in lung vascular permeability may be by the ability of some eicosanoids such as TXA2 to induce leukocyte adherence and aggregation (466a). Garcia-Szabo et al. (15la) recently observed that inhibition of TXA2 generation prevented the thrombin-induced increase in lung vascular permeability as well as the immediate decrease in leukocyte count that occurred with thrombin. This observation supports the notion that TXA2 interacts with leukocytes and thereby plays a permissive role in the development of lung vascular injury after intravascular coagulation. The specific roles of prostaglandins, thromboxanes, and leukotrienes in pulmonary edema after microembolism are discussed in section 11114. V) EFFECTS OF CORTICOSTEROIDS. In vitro granulocyte aggregation induced by zymosan-activated plasma was inhibited by methylprednisolone and hydrocortisone but not by dexamethasone (180). This was also true after leukocyte aggregation induced in vivo by injection of zymosan-activated plasma or endotoxin (20,179,180). The disaggregating effect of steroids was nonspecific because the effect of other leukotactic agents, N-formyl-MethLeu-Phe and ionophore A23187, was also inhibited by methylprednisolone (180, 462). Corticosteroids may disaggregate granulocytes by an effect on receptor sites because methylprednisolone has been shown to inhibit the binding of the chemotactic peptide N-formyl-Meth-Leu-Phe to the granulocyte receptors (180, 462). Methylprednisolone pretreatment or treatment soon after challenge of sheep with Escherichia coli (E. COZQendotoxin prevented increased vascular permeability, as measured by lack of increases in Qlym and transvascular protein clearance (52). Because studies of granulocyte depletion indicated that the increased permeability with endotoxemia is mediated by granulocytes (186), corticosteroids may act by inhibiting granulocyte aggregation (180). The effects of methylprednisolone on the increase in lung vascular permeability and on the pulmonary leukostasis after pulmonary microembolism have not been examined. VI) LEUKOCYTE-FIBRIN INTERACTION. Granulocyte removal and fibrinogen depletion prevented pulmonary edema formation after glass-bead microembolization in dogs (217, 218). The question arises whether fibrin entrapment and pulmonary leukostasis are related, because both appear to be necessary for pulmonary edema development. As shown in Figure 10, fibrin entrapment may be coupled to leukocyte mobilization and activation. Intravascular fibrin results in the activation of plasminogen and its conversion to plasmin, which activates the Hageman factor (factor XII) in plasma (236, 237, 341). The activated Hageman factor initiates the intrinsic coagulation, fibrinolytic, and complement-system activation; therefore the consequent generation of the chemotactic factor C5a is linked to the clotting mechanism. Plasmin can also directly activate the complement system independently of

July

PULMONARY

1983

1153

MICROEMBOLISM

t PLATELET AGGREGATION

THROMBIN -e

Fibrinogen

Complement . . Act,v>~osTAs’s

e

FIBRINe

w

FDP

I

Plasminogen Activator

Plasminogen

’

w Plasmin

FIG. 10. Consequences of thrombin-induced microembolization. Thrombin activates platelet aggregation, polymerizes fibrin, and activates the complement system directly by proteolytic cleavage and indirectly by activation of plasmin. Pulmonary leukostasis secondary to intravascular coagulation may occur by 3 mechanisms: 1) generation of chemotactic fragments (e.g., C5a) after complement activation, 2) entrapment of leukocytes in fibrin meshwork, and 8) chemotactic effects of generation of fibrin-degradation products secondary to plasmin-induced fibrinolysis.

Hageman factor (Fig. 10; 237). Fibrin-degradation products produced as a consequence of plasmin-induced fibrinolysis are also chemotactic for leukocytes (243, 479). Moreover the fibrin meshwork immobilizes granulocytes within intravascular thrombi and enhances their contact with the vessel wall (453). Thus intravascular coagulation is an important step in initiation as well as in amplification of the leukocyte-mediated pulmonary vascular injury. Leukocyte activation may in turn lead to further intravascular coagulation, because the release of lysosomal enzymes from leukocytes activates clotting factors (162,187). In addition leukocyte-mediated endothelial injury results in the release of tissue thromboplastin due to cellular injury, which then activates the extrinsic coagulation (thromboplastin-dependent) cascade (236, 237), and the exposed collagen substratum activates the intrinsic coagulation (Hageman factor-dependent) cascade (236, 237). Leukopenia induced by nitrogen mustard prevented the entrapment of fibrin (368), supporting the hypothesis that fibrin deposition also occurs as a direct result of leukocytes. Therefore activation of the clotting cascade produces pulmonary leukostasis, and leukocyte activation can produce intravascular coagulation; as a consequence a positive-feedback system is initiated. These different interactions between leukocytes and intravascular coagulation explain why both defibrinogenation and leukopenia protected against the microvascular damage associated with microembolization in dogs (217, 218).

1154

ASRAR

3. Platelet

B. MALE

vohrne

6.9

aggregation

The factors that cause platelet

aggregation are summarized in Figure of thrombin (193, 507), 2) the release of adenosine diphosphate (ADP) from injured endothelial cells (534), and 3) the exposure of the subendothelial collagen after vessel injury (185, 534). However, the role of platelet aggregation in mediating the increased pulmonary vascular permeability after embolization is probably minor. Collagen fibrils were injected into rabbit lungs to induce platelet aggregation; the result was only a 20% increase in fluid filtration that lasted 30 min (510). Increased filtration was prevented by perfusing the lungs with a plateletdeficient plasma before the collagen injection (510), indicating that the small transient increase on filtration is dependent on platelets. The effect of platelet aggregation on the change in lung vascular permeability was examined in sheep by determining the relationship between ADP-induced platelet aggregation and lung fluid balance (333). In these experiments, Qlym increased by 30% and the L/P did not change from base line after the ADP injection. As shown in Figure 12, the small filtration increase was not caused by increased endothelial permeability to plasma proteins, because an increase in Pmv produced an increased Qlyrn but decreased L/P, a finding consistent with hydrostatic edema. Therefore platelet aggregation per se does not seem to increase pulmonary vascular permeability to proteins, ruling it out as a major factor in initiating the increase in vascular permeability associated with microembolization. The role of platelets after microembolization has been examined by depleting platelets in dogs (59, 436) and sheep (36) with antiplatelet serum and then studying the response of the endothelium. Platelet depletion prevented neither the pulmonary edema (59) nor the increased pulmonary vascular permeability (36) associated with microembolization, indicating that the increased permeability and edema occur independently of platelets. There is a prevalent hypothesis that platelets support the integrity of the vasculature by forming “plugs” that close large gaps that develop in the endothelial lining of the capillary wall (106, 107). Petechiae and purpuric hemorrhages that characterize thrombocytopenia have been explained by postulating that large openings occur in the vasculature when platelet numbers are low (156). Roy and Djerassi (423) demonstrated an increased number of red blood cells in thoracic duct lymph after thrombocytopenia and a reduced number when small amounts of platelets were infused. In fact the amount infused did not appreciably increase the circulating platelet count (423). Dannelli (107) observed that the rate of edema formation decreased when frog hindlimb was perfused with a platelet-saline suspension. The mechanism by which vascular integrity is restored after platelet transfusion is poorly understood, but platelets may simply plug endothelial gaps, or they may release mediators that subsequently reduce permeability (106). Visscher (518) observed that the addition of platelets to the perfusion fluid caused a 11. These are: 1) the activation

INJURY-ADP

PHOSPHODIESTERASE

v

RELEASE)

MITOGENIC FACTOR PROTEOLYTIC ENZYMES PERMEABILITY FACTORS SEROTONIN ANTI-HEPARIN -THROMBOGLOBULIN B

(GRANULE

PLATELET CYCLIC-AMPI PHOS PHOLI PID

ADENYLCYCLASE

HROMBOXANE

t

PGFza

WALL)

A2 -VESSEL

(IN VESSEL

1 ‘PGG$H2)bPGE2, I

1 Pt$

SPASM

PLAT ELET A TP

FIG. 11. Platelets are aggregated secondary to tissue injury resulting in release of adenosine diphosphate (ADP), injury causing release of epinephrine, and coagulation-induced by thrombin. Aggregation is amplified after production of thromboxane A2 and release of endogenous ADP from platelets subsequent to their aggregation. S&d line, inhibitory reaction; AA, arachidonic acid; CO, cyclooxygenase; PG, prostaglandin. Platelet PGGz as shown or endogenous arachidonic acid causes PGIz (prostacyclin) to be synthesized in endothelium. Platelet aggregation is inhibited by PGIz because it increases the concentration of platelet cyclic adenosine monophosphate (CAMP). Platelet adhesion occurs after exposure of subendothelium and contact with collagen secondary to endothelial injury. Platelet-release reaction is associated with release of listed factors, some having potential to increase lung vascular permeability after microembolization. [From Weiss (534).]

COLLAGEN

COAGULATION-

?TISSUE

AMP+

z

ASRAR B. MALIK

1156

bhmte

6.9

.90

.40

I”.

4

I

’

6

8

L

lo

Pulmonary

12 lymph

1

14

1

16

A

FIG. 12. Relationship between pulmonary lymph flow and ratio of lymph-to-plasma protein concentration (L/P) after ADP-induced platelet aggregation @r&en liine with closed circles). Increased pulmonary microvascular pressure after inflation of left atria1 balloon increased pulmonary lymph flow but decreased L/P in normal lungs (broken line with open circles). [From Minnear et al. (333).]

18

flow.

ml / h

persistent loss of weight in isolated and perfused dog lungs. The same effect could be produced by injecting the supernatant of a platelet suspension (518), which indicates that a humoral factor was released from the platelets. Serotonin may be this weight-reducing factor, because infusions of low concentrations of serotonin into the lung caused a loss in lung weight, whereas a high concentration (which produces pulmonary hypertension) produced edema (54, 518). These findings agree with the recent observation of Sweetman et al. (481) that serotonin prevented the formation of petechiae in thrombocytopenic hamster cheek pouch microvessels. If this finding can be extrapolated to lung tissue, it is possible that platelet serotonin release inhibits the development of pulmonary edema by a direct protective effect on the microvessel endothelium. The other possibility is that serotonin produces intense precapillary constriction, thereby reducing the edema by decreasing the filtration pressure 4 Products of amchidonate a) Prostagkmdins. Prostaglandins of both the E and F series are released after pulmonary microembolization induced by injection of collagen fibrils (5,227,282,432,512). The normal pulmonary circulation rapidly inactivates prostaglandins of the E and F series (138, 398), indicating that these prostaglandins must have been generated either in large quantities or distal to the inactivating mechanisms. Prostaglandins are apparently released as a nonspecific response to microembolization: increased PGEz concentrations have been measured in pulmonary venous blood from isolated guinea pig lungs after embolization with substances as diverse as fat emulsion, microspheres, and colloidal emboli (227,282, 432, 511, 512). The PGEz release per-

July

1983

PULMONARY

MICROEMBOLISM

1157

sisted when the lungs were perfused with platelet-free plasma (511, 512), indicating a lung tissue source. However, this finding does not rule out the generation of prostaglandins from platelets and leukocytes directly embolized in the pulmonary vascular bed (337). In addition to the classical prostaglandins, TXAz and PG12 are also generated by the cyclooxygenase pathway (337), as shown in Figure 13. Thrombin is a potent stimulus for platelet aggregation and leukostasis, which results in the liberation of TXAz from platelets and possibly from leukocytes as well as the liberation of PG12 from pulmonary endothelial cells (337). Both TXA2 and PGIz have been detected in pulmonary venous effluent after endotoxemia (463), cardiopulmonary bypass (176), injection of zymosan-activated plasma (89), and thrombin-induced pulmonary microembolization (15lb). Thus TXA2 and PGIB release is a nonspecific response to various pulmonary insults. Bowers et al. (47) have examined the effects of PGE2, PGFz,, and an cu-methylene ether analogue of endoperoxide (PGH2) on pulmonary fluid and protein exchange in sheep. These agents caused an increased QIYrn and a decreased L/P, indicating that these prostaglandins in the normal lung increase Pmv and not microvascular permeability. These findings are consistent with the known pulmonary vasoconstrictor effects of these agents (234). Infusion of arachidonic acid also increased capillary filtration by increasing Pmv (374). It appears from these studies that prostaglandins in the normal lung increase pulmonary vascular pressures but not macromolecule permeability. Prostacyclin (i.e., PG12) caused a dose-dependent increase in Qlym, and the L/P was unchanged (373). Even though this finding could be interpreted as representing an increased vascular permeability, it more probably reflects an increased surface area for vascular exchange, because PGIz is a pulmonary vasodilator (234). The role of TXA2 has not been examined because it is degraded within 30 s (337); it is unlikely, however, that TXAz increases macromolecular permeability in the normal lung: ADP-induced platelet aggregation releases TXA2 (337) without any change in lung vascular permeability (333). It is conceivable that prostaglandins affect permeability but only in the presence of other humoral mediators. This synergistic effect was indicated by the enhanced leakage of plasma proteins induced by intradermal injections of bradykinin and histamine in the presence of prostaglandin El (PGE1) (493). The increased protein movement did not appear to be caused by the PGE1-induced vasodilation because papaverine did not potentiate edema formation in the presence of bradykinin and histamine (493). Infusion of PGIz blunted the increased Qlym and protein clearance associated with endotoxemia (114,372) and with pulmonary microembolization induced by either air (474) or clots (503). However, these effects do not necessarily represent a protective role for PGIz on permeability, as Demling et al. (114) have suggested. Infusion of PGIz decreases Ppa (234), and thus the protective effect is probably the result of a decreased Pmv after embolization.

1158

ASRAR

B. MALIK

Vohme

63

July

1983

PULMONARY

MICROEMBOLISM

1159

As discussed in section III..&?UIV, the generation of prostaglandins and TXAz associated with intravascular coagulation may contribute to the thrombin-induced increase in lung vascular permeability (38a, 151a, 151b). The evidence comes from the finding that cyclooxygenase inhibition with indomethacin or ketoprofen prevented the increased permeability after thrombin (38a). The protective effect was narrowed to inhibition of TXAz generation because inhibition of thromboxane synthetase with dazoxiben (an imidazole derivative) prevented the increased permeability (151a, 151b). Thromboxane A2 may contribute to the increased permeability by promoting leukocyte aggregation and adherence (466a). In fact, thromboxane synthetase also prevented the decrease in leukocyte count after thrombin administration (151a, 152b), supporting the theory of such TXAz promotion. Therefore TXA2 generation in the presence of pulmonary .leukostasis appears to be an important determinant of lung vascular injury after microembolization, whereas TXAz in the normal lung does not appear to produce injury. b) Leukotrienes. Arachidonic acid released from cellular phospholipids is metabolized to 5=hydroperoxy=6,8,11,14=eicosatetraenoic acid (50HPETE) by 5-lipoxygenase (EC 1.13.11.12) and subsequently to the leukotrienes (Fig. 13; 121, 270, 342). Leukotriene A is converted either to B4 or to the slowreacting substance of anaphylaxis components, leukotrienes C4 and D4 (121, 270,342). Arachidonic acid can also be metabolized by cyclooxygenase to the classic prostaglandins, TXAz, and PG12 (121, 342). The factors determining the dominant pathway for a given condition are unknown. Leukotrienes, in particular leukotriene B4 and 5-hydroxy=6,8,11,14=eicosatetraenoic acid (50HETE), are potent chemotacticand leukocyte-aggregating factors (370). These compounds either may have a direct effect or indirectly by affecti ,ng th .e white blood may increase vascular Pe rmeability cells. Their effect on lu w!z vascular permeability h as not been S tudied. Leukotriene D4 produces constriction of airways and skin microvessels, and it appears to be more potent than leukotriene C4 (182). Leukotriene D4 was as potent as bradykinin on a molar basis in increasing the permeability of skin vessels of m inea pigs and was 5-100 times more potent than leukotriene Cd. However, In rat skin vessels, leukotriene D4 was as potent as leukotriene C4 (502). In rabbits, however, nei .ther leuko triene C4 nor D4 increased the permeability of skin vessels (502) . Thus the permeability-increasing effects are species dependent. The comparative effects of the leukotriene series on lung vascular permeability have not been examined. The evidence involving leukotrienes in the formation of pulmonary edema associated with microembolization is at best circumstantial. Ogletree and Brigham (371) have made the intriguing observation that inhibitors of the cyclooxygenase system (such as indomethacin and meclofenamate) enhanced the pulmonary vascular permeability changes associated with endotoxemia. This finding suggests generation of leukotrienes as the pathway shifts from the cyclooxygenase system tofthe lipoxygenase system. It is also possible, however, that PG12 or PGE2 generation normally minimizes the

1160

ASRAR

B. MALIK

vohne

63

increase in pulmonary microvascular permeability; thus a permeability increase is expected during cyclooxygenase inhibition because synthesis of PGIz and PGE2 is blocked. Studies in sheep indicate that PGEl blunts the increases in QIYm and transvascular protein clearance associated with air microembolization (474) and endotoxemia (114, 372), although these effects may be explained by decrease in capillary pressure rather than by a direct effect on permeability. In vitro studies indicated that leukotriene D4 generation from arachidonic acid is enhanced by indomethacin and completely reversed by nordihydroguairetic acid or eicosatetraenoic acid, which are nonspecific inhibitors of the lipoxygenase pathway (370). These findings suggest that leukotriene production is increased by changing the pathway of arachidonate metabolism from cyclooxygenase to lipoxygenase, but the direct effect of leukotrienes on pulmonary endothelial permeability and their role in producing the permeability changes associated with microembolism are not known. 5. Serotonin

In 1954 Comroe et al. (88) observed that serotonin is a potent pulmonary vasoconstrictor and bronchoconstrictor (530). In 1917 Starling and Verney (469) were the first to appreciate that some substance was inactivated in its passage through the pulmonary circulation. They found that isolated kidneys could not be perfused because of intense renal vasoconstriction and embolization unless a heart-lung preparation was inserted into the perfusion circuit to remove particulate and vasoactive substances. One of these substances may have been serotonin; it is now known that serotonin is almost completely inactivated in blood passing through the pulmonary circulation (299). Serotonin appears in pulmonary venous blood after pulmonary microembolization, indicating either a release from the lungs into the circulation (171, 554) or an inhibition of its catabolism. The main sources of serotonin are platelets aggregated in the pulmonary circulation (534) and in i.nterstitial mast cells (13,363). The effect of serotonin may be species specific: mast cells in the rat, rabbit, and mouse contain higher concentrations than sheep or dog mast cells (269). Evidence obtained from the sheep preparation indicates that serotonin is not involved in the development of lung vascular injury. Serotonin produced an increased capillary filtration that was caused by an increased pulmonary Pmv (529); that is, the increased Qlym was associated with a concomitant decrease in L/P. Although serotonin may not increase pulmonary vascular permeabi .lity, this does not preclude a contribution of serotonin in edema formation associated with platelet aggregation. Because the lung vascular permeability to proteins is increased after microembolization (303), any increase of Pmv caused by serotonin enhances fluid movement into tissues and decreases the role of the edema safety factors restricting fluid accumulation (486).

July

1983

PULMONARY

MICROEMBOLISM

1161

6. Histamine Histamine has long been proposed as a possible mediator of lung vascular injury because it is known to increase vascular permeability (44). The concentration of histamine in the pulmonary venous blood increases after embolization with thrombi (432). The primary sources of histamine are the mast cells, which are situated in the perivascular and interstitial spaces of the lung (13, 363, SO), and the basophils (45) and platelets (534), which are trapped in pulmonary microvessels after microembolization. The details concerning the degranulation of mast cells are presently not known, but several factors associated with microembolization c:an cause the release of histamine: 1) mechanical injury (28, 110) [e.g., damage caused by an increase in blood velocity associated with microembolization (376)], 2) chemical agents (377) [e.g., the release of cationic proteases after leukocyte activation (405)], 3) activation of the lipoxygenase pathway (377), and 4) immunological processes (13, 46) (e.g., complement activation) associated with intravascular coagulation (13, 377). At doses that did not affect pulmonary hemodynamics, histamine infusion i nto th .e pu .lmonary artery in sheep increased Qlym and protein flux measurements, suggesting an increased vascular permeability (51, 53). Pretreatment with the histamine 1 (HI)-receptor antagonist diphenhydramine prevented such increases in Qlyrn and protein flux, suggesting that ,a specific histamine receptor mediated the increased permeability (53). The data in dogs, however, do not agree with the findings of the sheep lymph studies. Morphological studies in dogs indicated that histamine acted primarily on the bronchial venules-the permeability of these vessels to colloidal carbon increased, whereas the pulmonary vessels were unaffected (396). Drake and Gabel (118) could not demonstrate that histamine infusion increased pulmonary microvascular permeability in dog lung preparations. In an attempt to solve the apparent controversy, Nakahara et al. (355) demonstrated that histamine infusion into the bronchi .a1 artery supply of sheep caused increases . in pulmonary Qlym and lYmPh protein flux that were comparable to tho$e observed after infusion into the pulmon .ary artery, indicating that h istamine produced the same effect on lungs regardless of its route of infusion. Because histamine is normally not degraded in its passage through the pulmonary circulation (432), one possible explanation is that histamine entering the bronchial arteries caused the increased permeability of bronchial vessels in sheep. The other possibility is that vascular permeability was not adequately assessed (488). Raising pulmonary capillary pressure during histamine infusion to assess lung vascular permeability has failed to show a sustained increase in permeability in sheep (32a, 333a). However, there may well be a transient increase in lung vascular permeability that has also been described after histamine infusion in the skel .etal muscl .e vessels (333a). Another factor weigh .ing heavily aga inst histamine is that it has been difficult to produce pulmonary edema with histamine, except when it is ad-

1162

ASRAR

B. MALIK

v&.me

63

ministered in extremely large dosages. [A summary of the negative findings can be found in the excellent review of Visscher et al. (519).] Moon and Morgan (338) showed that pulmonary edema in dogs developed only after administration of 7.5-15.0 mg/kg of histamine twice daily for 9 days. Bariety and Kohler (25) were able to produce pulmonary edema only when both histamine (in doses of 0.2-0.5 mg/kg body wt) and epinephrine (0.5 mg/kg body wt) were administered; histamine given alone was ineffective. Even in isolated and perfused rat lungs, which are more susceptible to edema (68), Born (45) was unable to produce pulmonary edema by addition of histamine to the perfusion fluid. In intact dogs, Pietra and co-workers (396, 397) observed mild interstitial edema after a 190-min infusion of histamine at doses of 0.7 mg/kg. The increase in extravascular water content in sheep after histamine infusion averaged only lo-20% after several hours of infusion (396, 397). The variability in the edema response may be related to a variable constriction of the postcapillary pulmonary vessels (175) and the resulting lack of change of the capillary pressure in some experiments, which could minimize the increased fluid accumulation. More probably, however, histamine produces only a trivial *increase in lung vascular permeability. Pretreatment with diphenhydramine did not prevent increased lung vascular permeability when pulmonary microembolization was induced by glass beads (257). This study can be criticized, however, because activation of the fibrinolytic and complement systems may be necessary for the degranulation of mast cells, and the status of these systems after glass-bead emboli was not assessed. When these systems were activated (e.g., with E. coli in sheep), diphenhydramine partially prevented the endotoxin-induced increase in pulmonary vascular permeability (55). If histamine is released locally after pulmonary microembolization in large quantities (secondary to activation of coagulation and complement systems), it may similarly contribute to the increased pulmonary vascular permeability after pulmonary microembolization. Nevertheless this effect, if it does indeed occur, is probably small and transient. 7. Bradykinin

Bradykinin is generated within the plasma by activation of the intrinsic coagulation system (409,444). Activation of the Hageman factor (factor XII) causes the conversion of plasma prekallikrein to kallikrein, which in turn cleaves circulating plasma kininogen to bradykinin (444). Bradykinin is not generated in plasma deficient of either prekallikrein or Hageman factor (444). The pulmonary circulation has an important role in regulating circulating bradykinin concentrations because the converting enzyme, which is situated at the luminal surface of the pulmonary endothelium, is responsible for bradykinin catabolism (425). Although the converting enzyme is present on endothelial surfaces of all blood vessels (425,467), the pulmonary

July

1983

PULMONARY

MICROEMBOLISM

1163

circulation is by far the most important organ system for bradykinin inactivation, because lungs have a tremendous surface area and receive the entire cardiac output. Therefore a decreased vascular surface area after microembolization can impair the inactivation of bradykinin to less vasoactive products. Decreasing alveolar Po2 (PA%) to levels associated with arterial partial pressure of O2 (Pas) below 40 mmHg also interferes with the abil ity of the converting enzyme to inactivate bradykinin and convert angiotensin I to angiotensin II (467). Decreased surface area and hypoxemia both’occur after microembolization and thus can potentially decrease the catabolism of bradykinin, resulting in high circulating levels of this compound. An extensive literature exists on the role of bradykinin in the inflammatory response of systemic vessels (409) [in particular skin vessels (29a)], but the effect of bradykinin in altering the permeability of pulmonary microvessels is poorly understood. Morphological studies in dogs indicated that bradykinin increases the permeability of the bronchial venules (396), but other studies with dog right lymph duct (which contains a very mixed lymph consisting of lymph fluid from heart, intestine, and liver) indicated an increased pulmonary Pmv (325). To further confuse the issue, studies in sheep indicated that bradykinin infusion produces only a small increase in Qlym with an unchanged L/P (386), although QIYm increased significantly when hypoxia was induced during the bradykinin infusion. These data suggest an increased pulmonary endothelial permeability when higher levels of circulating bradykinin were present. However, increased pulmonary Pmv during bradykinin infusion increased C&m and decreased L/P (33la, 332); such occurring across normal microvessels rather changes indicate ultrafiltration Also the i nhibiti ,on of the converting enzyme than an increased permeability. with captopril in dogs did not produce more edema after glass-bead microembolization (247a). These findings suggest that higher concentrations of bradykinin in the pulmonary circulation associated with inhibition of the converting enzyme do not contribute to the edema. 8. Impaired

pulrrumar2/

endothxdial function

a) Alterations in mdzbolic function The pulmonary endothelium is capable of synthesizing, releasing, and inactivating several vasoactive and permeability-increasing mediators that affect lung fluid balance (290, 494). A complete discussion of these properties is beyond the scope of this review, although the factors anabolized or catabolized by the pulmonary endothelium are summarized in Table 2. Any alterations in the concentrations of these humoral factors resulting from an impairment in endothelial function may contribute to increased pulmonary vascular permeability and to alterations in capillary pressure associated with microembolization. The effectiveness of lung endothelium in performing its metabolic func-

1164 TABLE

ASRAR

2. Handling

B. MALIK

of vasoactive substances in the pulmonaw

Vohm

63

vascular bed

I. Metabolized at endothelial surface Bradykinin-inactivated Adenine nucleotides (adenosine monophosphate, adenosine triphosphate)-inactivated II Angiotensin I-converted to angiotensin II. Metabolized intracellularly Serotonin 1-Norepinephrine Prostaglandin E, prostaglandin F, prostaglandin A III. Synthesized within lung cells Prostaglandin E and prostaglandin F Prostaglandin I2 Thromboxane A2 Plasminogen activator IV. Discharged from intrapulmonary stores Histamine Leukotriene B4 Slow-reacting substance of anaphylaxis (leukotrienes C4 and D4) Kallikreins

tion on biologically active substances is related more to its enormous endothelial surface area and location than to any unique characteristic of the endothelium (425). Figure 14 shows the metabolic processing of vasoactive agents. The vasoactive prostaglandins synthesized de novo by the lung endothelium include PGFzc,, PGEz, and PGI,; prostaglandins of the E and F series are inactivated by conversion to dihydro-15ketoprostaglandin derivatives (337, 425). In addition, lung endothelial cells contain plasminogen activator, which is capable of initiating fibrinolysis (425, 526, 548). Because fibrin-degradation products generated by fibrinolysis are believed to contribute to increased lung vascular permeability (438), any alteration in endothelial fibrinolytic activity may thus contribute to the development of pulmonary edema. The luminal surface of the endothelial cells is also the site of the converting enzyme responsible for inactivating bradykinin and for converting angiotensin I to angiotensin II (426). Serotonin, another substance that may alter lung fluid filtration after microembolization, is also largely inactivated during a single passage through the pulmonary circuit (229). As previously discussed, neither bradykinin nor serotonin appears to be a primary mediator of the increased permeability; nevertheless they may enhance fluid accumulation because of their effects on Pmv and vascular surface area. b) Changes in ceZZshap. The endothelial monolayer serves as an effective barrier against fluid accumulation in the interstitial spaces. Alterations in endothelium shape, which may be related to changes in cell volume (123, 406) and/or to contraction of the intracellular microfilaments (29,555, 562), can increase the permeability of the microvascular wall to proteins. In addition, the loss of fibronectin, an adhesive protein that “glues” the endothelial

July

1983

PULMONARY

INTRAVASCUCAR

1165

MICROEMBOLISM

SPACL

L-NE

saturable, Na* - dependent,

5-HT

carrier -mediated

r

L-NE

transport

carrier-mediated, OH-PG-dehydrogena 7d other enzymes 15 Keto. Lm13-14dihy&o.

a

iznverting

I metabo’ites

enzyme

sdykininase

deaminated dephosphorylated FIG. 14. Handling of circulating vasoactive substances in pulmonary microcirculation. LNE, L-norepinephrine; 5-HT, serotonin; COMT, catechol-O-methytransferase; MAO, monoamine oxidase; PG, prostaglandin; AI, angiotensin I; AII, angiotensin II; ATP, adenosine triphosphate; AMP, adenosine monophosphate. Angiotensin-converting enzyme and bradykinase are same enzyme. [From Junod (ZBa).]

cells to the collagen substratum and maintains cell-to-cell contacts, may also increase capillary endothelial permeability (49,485,555). The release of proteases from granulocytes causes the breakdown of cell-surface fibronectin (430). Saba and co-workers (428-430) believe this is an important means of increasing vascular permeability. Fibronectin administration after pulmonary injury improved cardiopulmonary function, as evidenced by decreased venous admixture and alveolar dead space (428,429). These findings suggest fibronectin is important in maintaining the integrity of the pulmonary microcirculation by promoting the adhesion of endothelial cells to basement membranes and collagen. An experiment performed by Niehaus et al. (361)

1166

ASRAR

B. MALIK

i?&mt-e

6.9

supports this hypothesis. In this study the increased QIYrn and protein clearance associated with Pseudomonas sepsis were enhanced. in sheep with fibronectin deficiency. The effect of fibronectin repletion was not studied.

The hypoxemia and loss of essential substrates occurring distal to the obstruction of pulmonary vessels have been proposed as factors contributing to the increase in lung vascular permeability associated with emboli (22, 340). Hypoxemia per se cannot be an important factor because decreasing Paoz in sheep to 40 mmHg did not alter lung vascular permeability to proteins (39). The decreased substrate delivery, however, appears to be an important but poorly understood phenomenon associated with endothelial injury. Barie et al. (22) found that obstruction of one pulmonary artery for 3 h followed by a 2-h period of reperfusion caused greater edema in ischemic lungs. The greater edema formation in the ischemic lung could be due to an increased interstitial protein concentration resulting from the increased microvascular permeability. That pulmonary vascular permeability was increased (22,225) after pulmonary ischemia was also demonstrated in sheep studies (23). Occlusion of the pulmonary artery for 3 h resulted in steadily increasing Qlym and transvascular protein clearance despite a decrease in the vascular surface area. The increases in Qlyrn and clearance persisted even after reperfusion (23), indicating the increased permeability was caused by a period of pulmonary ischemia rather than by reperfusion per se. The pulmonary vessels and the coronary vessels behave in a similar manner: coronary vascular occlusion also increases coronary QIYm and lymph protein concentrations (128). The edema in the ischemic lung was not related to intravascular coagulation and the ensuing microembolization, because the ischemic lung became edematous even in heparinized animals. Johnson and co-workers (225) noted that granulocytes marginated in the pulmonary microcirculation, suggesting their involvement in edema formation after ischemia. Another possibility is the direct generation of oxygen radicals during ischemia from the ischemia-induced breakdown of adenosine triphosphate (ATP) to hypoxanthine and the conversion of xanthine dehydrogenase (EC 1.2.1.37) to xanthine oxidase (162a, 313). The reaction between hypoxanthine and xanthine oxidase results in the formation of oxygen radicals, which have been implicated in intestinal vascular injury after intestinal ischemia (162a). K. Neural

Factors

Several studies have examined the role of neural mechanisms in the development of pulmonary edema after microembolization (231, 232, 460). Localized pulmonary edema produced by injection of starch suspension into

July

PULMONARY

1983

MICROEMBOLISM

1167

a pulmonary artery wedge catheter was partially inhibited by the adrenergic antagonists benanserin and dihydroergotamine (231, 232, 460). The notion that sympathetics were involved was confirmed by bilateral removal of the sympathetic chain in dogs (231). The response was largely unaffected by lysergide, however, suggesting it was not due to serotonin release. When procaine was injected into the embolized lobe, it also prevented edema in all lung lobes (232). It was suggested that procaine anesthetized the local endorgans, which initiated the chain of events leading to edema, but the nature and site of these receptors were not defined. The decreased water content of the lung may have been due to a decreased pulmonary Pmv, which can occur after sympathectomy or procaine injection (104, 206). Edema by neurogenic mechanisms need not be invoked. The neurogenic theory of pulmonary edema has also been supported by the finding in dogs that embolization of a lung lobe with a suspension of starch granules (lo-15 pm in diameter) produced edema not only in the lobe receiving the emboli but also in the remaining lung lobes (231, 232, 460). Studies have clearly demonstrated, however, that injection of these emboli into one lobe does not necessarily rule out emboli entering other parts of the lung (357). When care is taken to ensure that the emboli do not disperse, edema develops only in the region of the lung receiving the emboli (263). L Role of Bronchial

Circulation

Because bronchial arteries anastomose with pulmonary microvessels (56), bronchial blood flow persists through segments of the lung in which the pulmonary arteries have been obstructed (102). Pulmonary vascular lesions and infarctions are more common after obstruction of small pulmonary arteries than after obstruction of large pulmonary arteries (102). Because the capacitance of the pulmonary vascular bed distal to the point of obstruction is lower after microembolization than after macroembolization, bronchial blood flow entering the obstructed region results in a greater increase in pressure after microembolization. Figure 15C shows how a greater rise in pulmonary capillary pressure enhances the degree of filtration in the segment obstructed in a manner similar to obstruction of a vessel with normal Ppv (as shown in Fig. 15B). In the model shown in Figure 15, the increased pressure is due to pulmonary venous hypertension, but it could also occur as a result of obstruction of small pulmonary arteries in which influx of bronchial blood flow results in a greater increase in capillary pressure than would occur if the obstruction were in the large pulmonary artery. M Cellular

Edema

Although 65% of the extravascular water content in the lung is contained within cells, little is known about the factors responsible for cell fluid balance (465, 489). Endothelial cells represent ~35% of the total cell population of

1168

ASRAR B. MALIK

Pulmomry

vohne

63

Arteriole

Broncho-pulmonary Anaatomoda Bronchial

Direction

of Blood

Flow-

Arteriole

CLOT

-Direction Bronchial

of Blood

Flow 4

At teriole

C

Venou8

Pressure

Broncho-pulmonary

8ronchiol

Direction

of Blood

Flow e

Arteriole

FIG. 15. A: relationship between bronchopulmonary anastomoses and pulmonary microcirculation. B: possible sequence of events when pulmonary artery is occluded but without rise in pressure in occluded segment (e.g., without an increase due to macroembolism or obstruction of larger pulmonary artery). C embolization of pulmonary artery but with associated rise in pressure in occluded segment, possibly caused by embolization of microvessels downstream from more compliant pulmonary arteries or pulmonary venous hypertension. Events in C would lead to greater edema because of greater rise in capillary pressure in occluded lung segment. [From Dalen et al. (102).]

the lung (471,489), and a potential exists for considerable fluid accumulation in these cells. The total amount of free solutes and consequently their osmotic activity within the cells are regulated by pumps in the membrane. The os-

July

1983

PULMONARY

MICROEMBOLISM

1169

motic activity difference between cell and environment determines intracellular water content, and any impairment in the function of the Na+-K+ pump determines how much the cell shrinks or swells (287). During hemorrhagic shock, the Na+-K+ pump fails because of low O2 availability, and the result is an increased influx of sodium and water into vascular smooth muscle (lOO), liver (442,443), and lung cells (421). Moreover, as cell swelling occurs, a further decrease in O2 consumption occurs (271), causing a further impairment in the pump function in a positive-feedback fashion. Morphological evidence indicates that cellular edema is a common feature of pulmonary vascular injury caused by many different substances (489). The work of Powers et al. (400) indicates the importance of cells as sites of fluid accumulation in pulmonary injury. This study shows that mannitol injection produced a rapid fall in PVR and alveolar dead space and an increased Pa%. Because these changes occurred rapidly and were associated with decreased plasma sodium concentration, withdrawal of fluid from pulmonary cells (probably endothelial cells) may be responsible for the increased oxygenation capacity (69,123,464). Although the evidence of cellular swelling is indirect, it does indicate that cell volume changes should be considered after microembolization. Mannitol was ineffective in reducing the pulmonary edema associated with high hydrostatic pressures (411), presumably because the edema was confined to the interstitium. An important ramification of endothelial swelling is the concept of no reflow or impaired reflow (143), which refers to reperfusion after stasis or impaired blood flow. If the endothelial injury in the microcirculation causes endothelial cell swelling, the return of flow to base-line levels may be impaired, and a vicious cycle results in which low O2 delivery increases cell volume, which increases resistance and further decreases O2 delivery.

IV.

TACHYPNEA

AFTER

MICROEMBOLIZATION

One of the classic signs of pulmonary microembolization is the rapid shallow breathing that can persist for several hours and that may be preceded by apnea with rapid embolization (62, 320, 539, 541). The altered respiratory pattern could be due to hemodynamic or blood gas changes, such as the arterial hypotension and hypoxemia that are the inevitable consequences of microembolization. However, tachypnea occurred even when the Pa%, the arterial partial pressure of CO2 (Pa& and the arterial pressures are held constant after microembolization (539, 541). Because this response does not occur after bilateral vagotomy, the tachypnea is associated with an unidentified vagal reflex (539, 541). The mechanisms mediating the tachypneic response after microembolization are not as obscure as when Whitteridge (539) reviewed this subject over 30 years ago. The following section reviews this phenomenon with particular emphasis placed on the lung receptors responsible for it.

1170

A. Lung Irritant

ASRAR

or Rapidly

Adapting

B. MALIK

Vobne

63

Receptors

Rapidly adapting fibers were identified by Knowlton and Larrabee (252), who suggested that these fibers provided an excitatory input into the respiratory centers and that this input balanced the inhibitory effects of slowly adapting stretch receptors. The receptors consist of nerve endings situated beneath the epithelial cells of larger airways (130,252,384,385), particularly near the carina, and in the smaller bronchi (384, 441, 542). These rapidly adapting receptors were termed irritant receptors because they are activated by intravenous injections of irritants such as ammonia as well as by histamine (348, 384, 542). The stimulation of irritant receptors induces rapid shallow breathing that is periodically interrupted by deep breaths (62, 320, 539, 541). The nerve endings are directly stimulated by substances such as histamine, prostaglandin FZa, and serotonin (384, 541, 542), which are known to be released during embolization (see sect. 1111). Mechanical distortion of the nerve endings, such as with airway constriction, also causes activation of the fibers (329, 384, 541). The increased nerve activity induced by intravenous injection of histamine was attenuated and sometimes abolished by prior injections of bronchodilator agents (348); this finding suggests that the histamine effect is not a direct one but that it occurs secondary to its bronchoconstrictor action. It is unlikely, however, that the irritant receptors are responsible for the sustained tachypnea associated with embolization (62, 320) because these receptors adapt rapidly (252). B. Juxtapulwwnary

Capillary

Receptors (C Fibers)

Paintal (384,385) hypothesized that juxtapulmonary capillary receptors (J receptors) mediated tachypnea when activated by the pulmonary congestion produced by pulmonary microembolization. This idea stems from the notion that J receptors are positioned to sense changes in the interstitial fluid volume (Fig. 16); that is, the receptors are sandwiched between the pulmonary capillary wall and the alveoli (324). The afferent fibers are slowly conducting, unmyelinated fibers contained in the vagus and thus are easily differentiated from the stretch receptor afferent fibers, which respond to lung distension, and the rapidly adapting irritant receptor fibers (384, 385). Because the slowly conducting fibers are found in other lung regions, Paintal’s concept has been modified to distinguish between afferent vagal fibers supplying the respiratory exchange area (i.e., pulmonary C fibers) from those supplying the conducting airways (i.e., bronchial C fibers) (79,83,242). The terms puhonaqy and bronchiaL therefore refer to the vascular regions where the endings originate. Type J receptors or C fibers do not exhibit a firing pattern associated with normal respiration (32), but they do become activated when foreign substances such as phenyldiguanide or capsaicin are injected intravenously

July

1983

PULMONARY

16. Schematic representation of (i.e., juxtapulmonary capillary) receptor lining interstitial tissue between capillary and alveolus containing collagen fibrils (vertical lines). Nerve ending is stimulated by increase in interstitial fluid volume or interstitial fluid pressure produced during accumulation of fluid in interstitial space. Substances such as volatile anesthetics or phenyldiguanide (pdg) act on regenerative region (R) or receptor, whereas increase in interstitial fluid volume or pressure act on generator region (G) or ending. [From Paintal (384).]

1171

MICROEMBOLISM

FIG.

type

J

CAPILLARY TYPE Nervefi

pdg etc. I

J RECEPTOR INTERSTITIAL VOLUME I I

I

ANAESTHETICS

ALVEOLUS (384,385). The activation of C fibers caused a reflex tachypnea in all species studied (382,384,285); however, there were species differences. For example, an intravenous injection of phenyldiguanide or capsaicin in cats caused apnea, which was followed by rapid shallow breathing with a latency compatible with that of J receptors (or pulmonary C fibers) (382, 385). This response is identical to the ventilatory response observed after pulmonary microembolism in cats (384, 541), suggesting that the ventilatory response to microembolism in cats may be mediated by pulmonary C fibers. Phenyldiguanide had no effect in dogs (83), but capsaicin (82) produced a rapid apnea and rapid shallow breathing similar to that observed in cats. These species differences may reflect differences in stimulation of pulmonary C fibers and of bronchial C fibers. The mechanism of stimulation of C fibers after microembolization remains unclear. If microembolization stimulates these slowly conducting vagal afferents, it is unlikely that this is to an increased interstitial fluid volume [as Paintal (385) has suggested]-tachypnea occurs within seconds after embolization when little, if any, interstitial edema would be present. In addition no relation between pulmonary edema and the activity in slowly conducting vagal afferents was found with alloxan-induced edema (83). Microemboli in the small pulmonary arteries may mechanically deform the interalveolar septa, which distorts the interstitium and may thereby stimulate the nerve endings (169, 541). Sufficient numbers of nerve endings may not be present in the interstitium, however, to account for the sustained tachypnea associated with microembolization (324). Both pulmonary and bronchial C fibers are also stimulated by the release of humoral factors such as bradykinin, histamine, serotonin, and prostaglandins after microembolization (80, 83, 242). Histamine activates pulmonary C fibers in rabbit (541) but not in cat (384). Intravenous histamine in cats primarily stimulated the bronchial C fibers because latency of response was slow and apnea rarely occurred (384). Therefore it seems reasonable to conclude that the reflex effects of histamine on respiration in cats are due

1172

ASRAR

B. MALIK

vohne

63

to its entrance into the blood supply of the airways. Serotonin injected into the pulmonary artery of cats produced a prompt apnea followed by rapid shallow breathing (384), indicating a direct stimulation of the pulmonary C fibers. In dogs, serotonin did not stimulate any C fibers (87,88,384). Both bradykinin and the prostaglandins have also been shown to be potent stimulators of C fibers, primarily the bronchial ones (80,242). Therefore several humoral mediators released after microembolization are capable of activating C fibers. A humorally mediated tachypnea is an attractive hypothesis because substances such as histamine, serotonin, prostaglandin, and bradykinin are released after microembolization and are known to stimulate both pulmonary and bronchial C fibers (83,159,384). Stimulation of pulmonary C fibers can occur by direct stimulation of the receptors in the alveolar-capillary septum, whereas stimulation of bronchial C fibers occurs by transmission of these substances via the bronchial circulation and by subsequent stimulation of the C fibers contained in bronchial tissues (83). Platelet depletion prevented the increased activity of both irritant and C fibers after glass-bead microembolization (10). According to this evidence, the activation of both types of fibers is due to serotonin, histamine, and prostaglandins, which are released as a consequence of platelet aggregation. Platelet aggregation associated with pulmonary microembolization therefore appears to be an important component of the tachypneic response. C Pulm~

Arterial

Baroreceptors

Tachypnea during embolism can also be triggered by activating pulmonary arterial baroreceptors (320,321). Tachypnea is associated with vascular congestion upstream from the site of obstruction, which suggests that distension of pulmonary arteries mediated the response (320, 321). A dense network of myelinated and nonmyelineated afferent fibers terminates in the adventitia and outer layers of the media of extrapulmonary portions of the right and left pulmonary arteries (83-85, 456). The terminations of these fibers were found to be most abundant at the bifurcation portion of the main pulmonary artery. There appear to be two distinct sets of vagal afferent fibers: 1) a set of myelinated afferent fibers in the vagus that have normal action potentials at Ppa values of lo-50 mmHg (35, 83) and 2) a set of receptors that are activated at very high pulmonary pressures (60-100 mmHg; 81). In addition to these vagal afferents, afferent fibers originating from the pressure receptors in the walls of pulmonary arteries and contained in the left cardiac sympathetic nerve discharge at Ppa values of ~40 mmHg (364, 501). Raising the arterial pressure of one lung after ligation of the pulmonary veins sometimes resulted in tachypnea (84, 85). This response only occurred with intact vagi; ventilation increased when the Ppa was increased to 80

July

1983

PULMONARY

MICROEMBOLISM

1173

mmHg, but the response was variable (84,85). The possibility that pulmonary C fibers were stimulated by the pulmonary edema cannot be ruled out. With an ingenious preparation in which a pulmonary artery pouch was created to prevent the edemagenic effects of pulmonary hypertension and the stimulation of J receptors, increased phrenic nerve activity and respiratory rate were found to be directly related to the increased pouch pressure (262). Vagotomy abolished these changes, suggesting a role for the pulmonary arterial baroreceptors in the microemboli-induced tachypneic response. However, the pulmonary arterial baroreceptors are probably not essential in mediating tachypnea after microembolization, because the tachypnea associated with microembolization occurs at Ppa
D. Pulmonary

Stretch Receptors

Pulmonary stretch receptors are thought to be located within the airway smooth muscle (334). The primary stimulus for their activation is distension of the lungs, and their activity persists as long as the lungs are inflated (543). Pulmonary stretch receptors are probably not activated during pulmonary microembolization because obstruction of pulmonary microvessels usually causes areas of airway collapse and lung deflation (48, 350). Moreover the primary effect of activation of stretch receptors is a slowing of the respiratory rate due to an increased expiratory time (384) rather than tachypnea. The other reflex response of stimulation of stretch receptors is bronchodilation (384,385), whereas bronchoconstriction is the primary response in the airways after microembolization (see sect. v). For these reasons, it is unlikely that stretch receptors play a role in the tachypneic response associated with microemboli.

E. Summary From the above discussion it is clear that activation of any one set of pulmonary receptors does not explain the rapid shallow breathing associated with microembolization. Irritant receptors, C fibers, and pulmonary arterial baroreceptors can all contribute to the tachypneic response. The irritant receptors may be involved in initiating the response, whereas C fibers may be involved in sustaining it. The release of mediators such as histamine, bradykinin, serotonin, and prostaglandins after microembolization appears to activate both receptors. Baroreceptors may also be activated after embolization but only when Ppa increases to very high levels.

1174 V. AIRWAY

ASRAR CONSTRICTION

AFTER

B. MALIK

vohme

63

MICROEMBOLIZATION

A. Sites of Brmchoconstrictim Pulmonary microembolization also causes bronchoconstriction, which occurs primarily in the terminal airways (48,350). Histological examination of rapidly frozen lungs after pulmonary microembolization showed constriction of alveolar ducts and terminal airways (350). The diameters of the larger airways did not change significantly (350), but because of their greater wall thickness the larger airways may have frozen too slowly to permit assessment of the degree of their constriction (73, 349). Tantalum was used after barium sulfate microembolization to indicate a marked narrowing of airways with diameters ~1 mm. Progressively smaller changes were seen in airways up to 3 mm in diameter, and larger airways demonstrated no change in diameter (73, 349). Constriction of the terminal airway was always greater in cats than in dogs (73); these findings reflect either a greater amount of bronchial smooth muscle or a greater reactivity of such muscle in the cat (73, 347). The radiological evidence of constriction of small airways (73,347, 350) also agrees with the finding that the total airway resistance does not increase significantly after microembolization (350). Constriction of the small airways has little effect on the total airway resistance because ~80% of the resistance normally resides in the larger airways (i.e., those >3 mm in diameter) (285, 553). Stein and co-workers (191,475) have challenged the conclusion that only the peripheral airways constrict. They observed a time-dependent increase in airway resistance, which suggested that the large airways also constricted. A sharp decrease in the dynamic lung compliance (CL) was observed in the first 30 s after microembolization, and this was not accompanied by a change in airway resistance (RL). In the next 30 s, however, RL increased to levels 40% > control. The finding that CL initially decreased without any change in RL indicates peripheral airway constriction (73, 347), but the increased RL as CL continued to decrease also suggests central airway constriction. The difference in findings may be related to a greater degree of embolization with the autologous microthrombi used in Stein’s studies (191, 475), in contrast to the degree of embolization associated with the barium sulfate microemboli used by others (73,347,350). The platelet-fibrin microthrombi may also cause a greater release of bronchoactive substances than barium sulfate embolization (475, 505). The degree of embolization also appears to be an important determinant of the airway changes (74). The CL decreased only at higher doses of barium sulfate particles in guinea pigs, whereas the RL increased at lower doses (74), indicating that the degree of embolization somehow determines the bronchoconstriction site.

July

19w

PULMONARY

MICROEMBOLISM

1175

Selective embolization has been used to localize the site of bronchoconstriction (191,495). Embolization of one lung with autologous microthrombi resulted in a significant decrease in & and an increase in RL only in the embolized lung, indicating that the effects were localized (191). The Ci, decreased in both lungs, however, after a unilateral. microembolization produced by injection of iodized oil into one pulmonary artery, and the change in CL could be reversed with isoproterenol (495). This finding suggests that peripheral airways were constricted because isoproterenol primarily dilates the small airways (350, 495). One explanation for the contralateral bronchoconstriction may be that the iodized oil was not confined only to the injected lung. If the emboli were localized to the injected lung, the bilateral constriction could also be a result of the release of bronchoconstrictor substances from the embolized lung into the blood, which then arrives at the airway smooth muscle of the contralateral lung via the bronchial arteries (112). B. Factors Aflecting

Bronchoconstriction

The bronchoconstriction sites associated with emboli appear to be quite variable from laboratory to laboratory. There are several reasons for these differing results. Alveolar hypocapnia, which results from the reduction of CO2 exchange in embolized lung segments, may constrict peripheral airways (209, 450), and the arterial hypercapnia associated with embolization (108) causes the central airways to constrict (475,477). The Pacoz and airway Pco2 have not been controlled in embolization studies; these factors may have caused some of the variability among studies. The possibility of different interactions occurring between various bronchoactive mediators (histamine, serotonin, bradykinin, prostaglandins, and catecholamines) in individual experiments may also be responsible for some of the variability. For example, the simultaneous infusion of catecholamines can reverse the constrictor effect of histamine or bradykinin on bronchioles (352; 399); therefore the typical bronchoconstrictor response induced by microembolization would be markedly attenuated if the dilating substances are released by the emboli to a greater extent than are the constricting substances. In addition the degree and site of airway constriction depend on the prevailing smooth muscle tone in the airways (516), and the use of anesthetics and muscle relaxants may reduce bronchial smooth muscle tone (540). There was a smaller increase in RL after embolization if large airways were previously constricted because of either increased vagal tone (173) or decreased lung volume (516). Therefore RL would presumably increase more if larger airways were dilated before the embolization occurred. Finally, differences in the degree of embolization and in the methods of producing it (which result in different concentrations of bronchoconstrictor agents) may also explain the lack of agreement.

1176

ASRAR

C. Mechanisms

Involved

in Producing

B. MALIK

Vohme

63

Bronchoconstriction

The existing physiological evidence indicates that primarily the small airways constrict and that this is mediated by humoral mechanisms. The evidence supporting the constriction of large airways is equivocal; when large .airway constriction does occur, however, it is mediated by neural mechanisms. 1. Humoral

mechanisms

The constriction of small airways appears to be chiefly mediated through humoral factors, because the decreased CL associated with emboli can be inhibited by antagonists of serotonin, histamine, and prostaglandins (‘78,191, 350). Also, 2 min are required for small airways to become maximally constricted after embolization (350, 475, 476), whereas vagally mediated constriction occurs in seconds (73,347,350). In addition the constriction of small airways after barium sulfate embolization cannot be blocked by atropine or vagotomy (350). Finally, powerful bronchoconstrictor agents such as histamine, bradykinin, serotonin, and prostaglandins are known to be released from aggregated platelets and leukocytes in plasma and by mast cells after microembolization (363, 432). Vaage and co-workers (509, 511, 512) have assessed the relative contributions of each humoral factor to bronchial constriction. A temporal relationship was established between the increased airway pressure in the cat after collagen-induced embolization and the contraction of in vitro smooth muscle bioassay preparations superfused with pulmonary venous blood from the same animal. The smooth muscle samples were pretreated with antagonists of catecholamines, histamine, serotonin, and acetylcholine (330, 509, 511, 512), which eliminated these compounds as bronchoconstrictive agents. Prostaglandin-like activity was demonstrated with this bioassay system, and indomethacin inhibited the contraction of smooth muscle (511). Clay and Hughes (14) also inhibited the embolization-induced increase in airway resistance with indomethacin in intact guinea pigs. Therefore prostaglandins significantly mediate the bronchoconstriction associated with embolization. The humoral factors responsible for constriction of small airways are released as a consequence of intravascular coagulation or platelet aggregation (437, 492, 507), because airway changes are inhibited or at least attenuated by either platelet depletion (41,403,505,507,509) or heparin (267). If this is the case, the number of mediators released should be reduced in thrombocytopenic animals. Radegran (403, 404) demonstrated this effect in a cat preparation. Once formed, the mediators probably diffuse from the blood vessels into the lung interstitium and then to the smooth muscle of small adjacent airways. This method of mediator dispersion is supported by the finding that constriction of the larger airways after microembolization (when it occurs)

July

1983

PULMONARY

is delayed and is much smaller in magnitude sponse of small airways (475, 476). 2. Neural

1177

MICROEMBOLISM

when compared

with the re-

factors

When microembolization produces increased RL (191,475,476), it occurs rapidly, suggesting a reflex arc. Sectioning of the cervical vagi or administration of atropine partially or completely reduced the change in RL (191). The decreased CL was unaffected by these maneuvers (191, 403, 505). Vagal stimulation caused constriction of airways 3-8 mm in diameter in dogs (160, 191, 476, 505, 553). a) Irritant receptors. The reflex may be mediated by histamine acting on the irritant receptors in the mucosa of large airways (385,543). The reflex elicited by stimulation of these receptors can be blocked by atropine and by cooling or cutting the vagi (112), but the effects of antihistamine are not known. A 118% increase in RL and a 14% decrease in & occur within 14-18 s after injection of histamine directly into bronchial arteries in dogs (112). The degree of reflex activation may therefore depend on the concentration of histamine after embolization; the failure to observe an unequivocal constriction of large airways may reflect the degree of embolization and the amount of histamine released (138, 432). The irritant or rapidly adapting receptors responsible for the reflex phenomena just described are situated beneath the epithelial layer of the trachea, and the major airways are supplied by myelinated vagal fibers (385, 543). They are chemical-type receptors stimulated by substances (such as histamine) that are liberated by microembolization (83,384,385). In addition, they are also mechanically stimulated by alterations in lung volume, contraction of adjacent smooth muscle, and pulmonary congestion (384, 385), changes that also occur as a consequence of microembolization. These receptors may respond differently in different species in terms of the magnitude (21,73,239), but it is not certain that the lungs were embolized to the same extent in all studies. Another property of irritant receptors is that their stimulation has very diffuse effects. Stimulation of receptors in the lower airways with histamine causes constriction to occur in the upper airways (351, 353), whereas mechanical irritation of upper airway receptors constricts lower airways (160). Furthermore, a stimulus localized to one lung causes constriction within the other lung (112). These widespread effects may explain the constriction seen in both lungs after microembolization of one,lung (495). b) Juxtapulmonary capdlary receptors. Paintal (383-385) believes that the bronchoconstriction that follows microembolization is due to stimulation of the J receptors (i.e., pulmonary C fibers). Furthermore, because these receptors are found in the alveolar septal space, it is hypothesized that embolization of small pulmonary arteries stimulates them by distorting the

1178

ASRAR

B. MALIK

alveolar septa (324). The role of pulmonary constriction associated with microembolization choconstriction can be produced by simply tery with a balloon when there is not likely alveolar septum (451). Stimulation of the taglandins, and other mediators released likely mechanism of their activation. 3. Airway

Vohrne 63 C fibers in producing bronchois uncertain, however; bronoccluding a large pulmonary arto be a marked distortion of the J receptors via histamine, prosafter embolism may be a more

hypocapnia

Some changes associated with pulmonary microembolization may be related to decreases in the partial pressure of CO2 (Pco~) of lung units not receiving CO2 from mixed venous blood. Ingram (209) showed that a decreased alveolar Pco2 produces a large decrease in static lung compliance and only a small increase in R L, a response similar to microembolization. Severinghaus et al. (451) demonstrated that the bronchoconstriction associated with occlusion of a pulmonary artery was prevented by adding an 8% mixture of C02. Because the tone of the small airways was unchanged in the normal lung after breathing this mixture, the effect of CO2 in preventing the bronchoconstriction was only the reversal of alveolar hypocapnia (164, 352). Alveolar hypocapnia may therefore be an important factor in producing bronchoconstriction after microembolization only when portions of the’ circulation have been totally blocked by the procedure (145, 228, 259, 495). D. Homeostatic

Value of Brmchoconstriction

Studies have demonstrated that airflow to an embolized lung decreases within a few breaths, whereas airflow to the unobstructed lung increases (247, 451, 515). This tends to maintain a near-normal VA/(). In addition, hypoxemia is greater when bronchoconstriction is prevented by ventilating with a C02-rich gas mixture at the time of embolization (451,482). Thus the bronchoconstriction associated with pulmonary vascular obstruction is a homeostatic mechanism similar to the pulmonary vasoconstriction that occurs with hypoxia. Both mechanisms serve to improve the VA/Q inequality that would otherwise result. The redistribution of airflow from the embolized lung occurs in two stages (266) resulting from two distinct mechanisms. There appears to be an initial nonuniform bronchoconstriction caused by humoral and neural mechanisms followed by a bronchoconstriction in the embolized lung regions related to the airway hypocapnia. The initial airflow shift from the obstructed lung units is followed by an even greater shift. During the initial stage, the airflow shift has been related to the release of various bronchoconstrictor substances that cause diffuse constriction in nonembolized and embolized segments. Therefore the redistribution away from embolized regions was not as marked as would have been expected had the constriction

July

PULMONARY

1983

MICROEMBOIJSM

1179

been confined solely to the embolized regions (266, 491, 492). The diffuse bronchoconstriction is due to the release of humoral factors (which are associated with intravascular coagulation), because heparin resulted in a more uniform constriction that was confined to the embolized regions (266). The second stage of the airflow response, in which bronchoconstriction is localized to the embolized lung units, is mediated by airway hypocapnia, because it could be prevented by adding CO2 to the inspirate (266).

E. Alveolar

Dead Space After Embolixation

Alveolar dead space either increases (134, 301) or remains unchanged after pulmonary embolization (108). The variability of the response reflects the degree of redistribution in airflow associated with the emboli. If the ventilation distribution remains unchanged, alveolar dead space increases by an amount equal to the original VA of the occluded lung segment; but if total ventilation of the occluded lung is diverted into the perfused lung, the dead space increases by the lung volume of the unperfused lung. Because ventilation is only partly redistributed from embolized lung units (108) and because .the degree of redistribution is a time-dependent phenomenon (451), any measured change in dead space after pulmonary embolism reflects the amount of airflow redistribution. Therefore the measurement of dead space is not a reliable index of the degree of vascular obstruction (450). Another complicating factor is that emboli producing incomplete vascular obstruction result in a lower dead-space value than emboli producing total obstruction (134), presumably because the incomplete obstruction allows some perfusion to continue to the lung segments.

F. Alveolar

Collapse

Collapse of alveoli is usually observed with pulmonary microembolization (48,350). The blockage of terminal airways by bronchoconstriction may explain segmental atelectasis occurring within 1 h after embolism (350). The alveoli collapse as a result of the airway closure because the gases trapped in the alveoli diffuse down concentration gradients into the blood. The most common site of alveolar collapse is in the dependent lung (350), because the lung parenchyma is less expanded; thus microembolism-induced bronchoconstriction can more easily produce airway closure. Atelectasis can also be related to the loss of surfactant activity associated with pulmonary vascular obstruction (69). Finley et al. (131) found that fluid extracts from lungs whose arteries had been ligated for 12-16 h had high surface tensions. Giammona et al. (153) observed an increased tension as early as 4 h after ligation of a pulmonary artery. Morphological and biochemical studies indicated that alterations in phospholipid and dipalmityl phosphatidylcholine levels occurred only within atelectatic areas

1180

ASRAR

B. MALIK

l%hme

63

of the occluded lung (122,200,339). The studies of Finley et al. (131), however, demonstrated a more generalized type of alteration in surfactant activity, severe congestion, and atelectasis after occlusion of a pulmonary artery. The different findings may represent varying degrees of collateral bronchial blood flow (528,547). Although the severity of the lesions was variable, both studies indicated that lung vascular obstruction somehow interferes with surfactant metabolism and that it results in alveolar instability. These changes in altered surfactant metabolism ‘are not easily reversed: the obstructed lung differed from a contralateral lung in having a smaller fraction of total air volume returned at each transpulmonary pressure during deflation, even after 2 wk (153). The mechanisms responsible for surfactant inactivation are unknown. Perhaps the leakage of plasma proteins into the alveolar spaces (388) somehow inactivates surfactant (222). Although pulmonary embolism is known to cause an increased microvascular permeability, its effect on the epithelium may be less severe unless the embolism is severe (92,131,153,303). It is also possible that a direct ischemic injury to the metabolically active surfactantproducing type II alveolar epithelial cells could result from decreased pulmonary and bronchial perfusion of the obstructed segments (72, 302, 303). The return, of surface activity toward normal values several weeks after embolization also parallels the development of nutrient collateral circulations (153, 200, 339). The hypothesis that pulmonary ischemia could deplete surfactant has been supported by the observation that pulmonary arterial occlusion reduces the density of lamellar bodies in type II alveolar epithelial cells by 80%; these bodies are the intracellular storage sites for surface-active phospholipids (26, 454). Reperfusion for 6 h was not sufficient to reestablish normal type II cell morphology (26, 454). VI.

MECHANISMS

OF ARTERIAL

HYPOXIA

Rapid development of hypoxemia is a major consequence of pulmonary embolization (420). Hypoxemia is not caused by pulmonary edema-the Paoz decreases immediately after pulmonary microembolization (301), and no substantial accumulation of edema fluid occurs during this time. The hypoxemia is caused by right-to-left shunting (&,/&,; 245), alveolar-capillary impairment of O2 diffusion (58, 226, 455), regional TjA/t$ inequalities (108, 267, 268), and alveolar hypoventilation (108, .420). The relative contribution of each of these factors is discussed in the following section. A. Time Course of Gas-Exchange

Impakvnent

A decrease in Paoz occurs within minutes after pulmonary bolization induced by emboli of varying sizes and compositions

microem(301). In

July

1983

PULMONARY

MICROEMBOLISM

1181

experiments in which blood clots were used to produce microembolization, hypoxemia was associated with alterations in the distributions of VA and &, which caused a VA/Q mismatch (108). The VA/Q inequality after embolization with thrombi was transient, however, because recovery began 15 min after embolization and was complete after 2 h (108). Restoration of the normal VA/Q was probably caused by dissolution of the blood clot, resulting in a gradual return of blood flow and ventilation to normal values (101,417). The Paoz decreased immediately after embolization with glass beads; nevertheless, in contrast to the effect of embolization with blood clots, Pa% continued to decrease steadily (301). The latter decrease in Pao, in these experiments was associated with interstitial edema and, in some cases, with alveolar flooding (301). This steady decline in Paoz did not occur when pulmonary edema was prevented by prior treatment with heparin (301). B. fi$‘jFu.sion Impairment Because Paoz decreases immediately after microembolization when there is no evidence of alveolar flooding, a diffusion impairment across the alveolarcapillary membranes cannot explain the observed hypoxemia. Moreover the small decreases in diffusion capacity of carbon monoxide (2533%) that have been observed after microembolization (58, 267) indicate no appreciable alveolar-capillary barrier to O2 diffusion (226,455). Other emboli studies have shown no change in the diffusion capacity of carbon monoxide but a considerable degree of hypoxemia (245); therefore a diffusion impairment can be ruled out as being primarily responsible for the hypoxemia associated with microembolization. This does not preclude the increase in alveolar-arterial PO, gradient that may occur after marked pulmonary arterial obstruction due to an increased blood velocity through the remaining lung segment and the resulting decreased transit time. C. Iweased

Venous Admixture

1. Acute increase A rapid increase in the Qs/@ of blood through the lungs could be caused either by the opening of arteriovenous communications or by an increased percentage of blood flow occurring through collapsed lung units. Such an increase could explain the rapid development of hypoxemia. But @/Qt does not change in a predictable fashion (266,267); in fact Qs/@ sometimes does not change at all, despite the presence of hypoxemia (108). These differences could be explained if the embolization caused airway closure or constriction and alveolar collapse. Caldini (63) reversed the increase Qs/Qt after embolization in dogs by applying positive end-expiratory pressure and reexpanding the alveoli. Moreover an increased venous admixture was not ob-

1182

ASRAR

B. MALIK

vdume

63

served after glass-bead microembolization in dogs pretreated with heparin (267,.301), suggesting that inhibition of intravascular clotting prevents airway constriction and the resulting alveolar collapse. A sudden increase in pulmonary arteriovenous shunting produced either by an increased collateral bronchial blood flow (15, 56) or by an opening of parallel pulmonary arteriovenous shunts (135, 136) may be another explanation for the rapid increase in @/Qt. The bronchial blood flow, however, does not increase immediately after embolization (299), and therefore an increased collateral bronchial blood flow is probably not a cause of the increased shunting. The opening of parallel pulmonary arteriovenous pathways is also not likely because no evidence indicates their existence in normal or embolized lungs (108,135,136,296). Another latent (but unproven) cause of an acvte intx!wksst& C&$$Ld~kkhR a yi&wxCtiy y3hn!t G3mmsec wake tkat may open if the right atria1 pressure exceeds the left atria1 pressure after pulmonary vascular obstruction. 2. Delayed increase In addition to the acute change, Qs/@ also increased slowly after embolization (549). The delayed increase observed 19 days after an embolic episode could be reversed by deep breathing, indicating alveolar collapse (549). This increase may be related to alveolar collapse from loss of surfaceactive material (549) as a consequence of prolonged pulmonary hypoperfusion and impaired surfactant production (72, 498, 528). Possibly the fibrinolysis occurring after embolization clears emboli before the surface-active properties are restored (101); therefore Qs/Qt would increase because the vessels are perfused, but the alveoli would not exchange gas. The increased collateral bronchial blood flow, which occurs over a period of days after embolization (15), may also be responsible for the delayed &s/ &t increase. Bronchopulmonary collateral blood flow shunts venous blood into the pulmonary circuit (15). Collateral blood flow increases markedly over a period of months (528), and a two to threefold flow increase has been observed as early as 2 wk (245). Because the O2 saturation of the bronchopulmonary effluent blood averages only 50% (15), any increase in the bronchopulmonary flow would result in an increase in @/Qt. D. Ventilation-Perfusion

Imbalance

The most important cause of arterial hypoxemia occurring immediately after pulmonary microembolization is an imbalance in regional ventilation and perfusion. This VA/Q imbalance is caused by embolization of pulmonary arteries, because embolization alters the distribution of pulmonary perfusion (265, 301, 302). This redistribution of & produces regions havinglow VA/Q values (log), which is the primary defect responsible for the hypoxemia associated with microembolization (133, 283, 349).

July

1983

PULMONARY

1183

MICROEMBOLISM

A two-compartment lung model (described in Fig. 17) has been used to explain the VA/Q imbalance (108). In this example, one compartment initially receives 83% less of the ventilation and blood flow and the other receives 17% of each. If the pulmonary artery of the smaller segment is embolized so that 90% of its base-line blood flow is diverted, the VA/t& of the smaller segment would be increased to 10, whereas the VA/Q of the larger segment would only decrease from 1.0 to 0.9. But if the larger segment is embolized and only 50% of its base-line flow is diverted to the other lung, the VA/t) of the smaller segment would decrease from 1.0 to 0.3, whereas the VA/Q of the larger segment would increase to 2.0. Thus in the second case there is a region of the lung with a VA/Q value
mechanism

of ventilation-perfusion

imbalance

Platelet aggregation plays a major role in mediating the VA/Q irregularities associated with microembolization. The development of hypoxemia after ADP-induced platelet aggregation was comparable in both time course and severity to the hypoxemia associated with microembolization (333). The hypoxemia was prevented by platelet depletion before the ADP injection (333); therefore platelet aggregation occurring after pulmonary microembolization and the concomitant vasoactive and bronchoactive substances released from activated platelets are primary factors in producing the VA/Q imbalance (506, 534). E. Alveolar

Hgpventdation

Sudden embolization is commonly associated with a period of rapid shallow breathing (48, 320, 539, 541). Although minute ventilation may be increased (62, 320, 41?), the rapid shallow breathing could produce alveolar

1184

ASRAR B. MALIK SMALL COMPARTMENT

vohme

63

LARGE COM~%RTMENT

HOMOGENEOUS LUNG PRE-EMBOLIZATtON ARTERIAL

PO, 104

ARTERIAL

Pco, 39

SMALL COMPARTMENT EMBOLIZED ARTERIAL pop

pco,

.

142

98

I6

40

v*

1.0

0

3.5

it*/6

0.3

PO2

58

pco,

44

ARTERIAL

PO, 98 PC02 40

LARGE COMPARTMENT EMBOLIZED ARTERIAL

Po2 66

ARTERIAL

PC02 40

FIG. 1’7. Two-compartment lung model with pulmonary blood flow (&) and alveolar ventilation (VA) in liters/min. Both VA and & are matched in top pan@ & to smaller compartment is reduced by 90% in middle pan& and & to larger compartment is reduced by 50% in bottom panel. Changes in VA, &, VA/Q, and POT and PCop of blood from each compartment (Zefi) are indicated along with mired arterial partial pressure of O2 (POT) and partial pressure of CO2 (PC@) values (right). [From Dantzker et al. (lOS).]

hypoventilation and arterial hypoxemia (326). Because hypoxemia also occurred after embolization in conditions of controlled ventilation (301), alveolar hypoventilation probably is not responsible for the observed arterial

July

1983

PULMONARY

MICROEMBOLISM

1185

hypoxemia. Alveolar hypoventilation occurs in spontaneously breathing animals after embolization (log), however, and in these instances it may contribute to hypoxemia. VII.

BRONCHIAL

BLOOD

FLOW

Morphological studies indicate that the bronchial arteries in most mammals supply the bronchial tree, including the terminal bronchioles and pulmonary vasa vasorum (104, 328). The pulmonary arteries supply the small respiratory bronchioles, alveolar ducts, and alveoli (316, 317). Injection of histamine into the pulmonary artery results in constriction of the alveolar ducts only, whereas injection into a bronchial artery causes constriction of the larger airways (112). Von Luschka (520) referred to the bronchial arteries as the “vasa privata” to distinguish them from the pulmonary arteries, the “vasa publica.” A century ago Virchow and Gesammelte (517) demonstrated that pulmonary infarction rarely occurred after ligation of a main pulmonary artery. An increased amount of collateral bronchial blood flow entering embolized portions of the lung prevents necrosis, but the mechanism of this protective effect is not well understood. Bronchial blood flow normally represents only l-3% of the cardiac output (104), and it is markedly increased after embolization (124,299,517,528). Although there is good evidence that bronchial blood flow is increased (275, 389, 547), the rapidity with which the flow increases after vascular obstruction and the factors that mediate the increase are not yet well understood. The following section contains a review of the normal pattern of bronchial perfusion, the factors that regulate the bronchial blood flow, and the extent to which these factors participate in bronchovascular regulation after pulmonary embolization. A. Anatomy 1. Pattern

and Physiology of normal

bronchial

of Bronchial

Circulation

perfikm

There is much interspecies variability in the origin and distribution of the bronchial circulation (104, 316, 317, 366). The most thoroughly studied system has been that of the dog (56,104,366), and as seen in Figure 18, the pattern in this animal is quite complex. Some generalizations can be made. 1) In all mammals the bronchial artery forms a capillary bed (Fig. 18), which primarily drains the airways to the level of the terminal bronchioles (104). 2) The bronchial capillary bed drains into the bronchial veins, which then return the blood to the right atrium either directly or via the superior azygos vein (Fig. 18); the alternate pathway is through the bronchopulmonary anastomoses, which drain blood directly into the precapillary and postcapillary pulmonary microvessels (Fig. 18; 104, 316, 317, 366).

SUPERIOR

VEIN

BED

VENA CAVA

BRONCHIAL

BRONCHIAL

4

PULMONARY VASAE VASORUM

'-\

AORTA

O--A-

PULMONARY

BED

ARTERY

VEIN

BRONCHIOLAR

RIGHT BRONCHIAL

1

n

FIG. 18. Vascular connections of right posterior bronchial artery in dog. Heaw line indicates arteries. Components of bronchial circulation are: 1) alveolar-capillary bed of pulmonary circulation; 2) capillary bed of respiratory bronchioles showing anastomoses with alveolar vessels that are supplied by pulmonary artery; 8) capillary bed supplying lung parenchymal tissue; 4) bronchial capillary bed supplying. larger airways that drain into true bronchial veins and in turn drain into azygos vein and superior vena cava; 5) pulmonary vasa vasorium, i.e., nutritional vessels of pulmonary vessels; and 6) visceral pleural capillary bed supplied by both pulmonary arterial system and bronchial arterial system. [Adapted from Bruner and Schmidt (56).]

u

5

0

ARTERY

VISCERAL PLEURAL BED

PULMONARY

6

0

July

PULMONARY

1983

1187

MICROEMBOLISM

The percentage of total bronchial arterial flow that enters pulmonary vessels relative to that draining into the bronchial veins is highly variable. Apparently 30-70s of the total bronchial arterial inflow passes through the bronchial veins, and the remaining flow drains into the pulmonary vessels via the anastomoses (7, 8). However, the techniques used to measure the anastomotic flow are crude, and the flow values are not known quantitatively (547, 558). Although anastomoses have been demonstrated in rat and guinea pig lung (316,317), they do not seem to be well developed in dog, cat, and primate lungs (316, 317). [In fact, they are nonexistent in rabbit lungs (317).] Their function has not yet been documented, but the prominence of anastomoses in some species may be related to greater requirements of nonrespiratory tissue for oxygenated blood. 2. N-1

bronchial

bloud&w

values

Attempts have been made to quantify bronchial blood flow, but in most instances extensive surgery is required to conduct the measurements (366). In addition the bronchial blood flow appears to vary spontaneously in a wavelike fashion, having a period of ~2 min (56). The basis for this phenomenon is not clear, but it may be caused by bronchial spasms associated with the extensive surgery. As measured with a bubble flowmeter in dogs, the bronchial arterial blood flow in the right posterior bronchial artery averaged 4.8 ml/min (56). This measurement did not include flow to the left lung or contributions to the right lung made by minor bronchial arteries. Total bronchial blood flow was estimated to* be 1% of the cardiac output (56, 104). Table 3 provides measurements of bronchial blood flow in dogs. These measurements were obtained by using labeled microspheres 15 pm in diameter. Bronchial blood flow represents Z-3% of cardiac output in these studies (299), which probably reflects a more realistic estimate of bronchial blood flow. The microsphere TABLE

3. Bronchial

blood flow afier ernbolizatim

Bronchial Blood Flow, ml/min

Group I Base line 5 min PE 60 min PE Group II 2wkPE

31.2 AI 6.2 42.5 t 9.5 8.6 AI 2.2* 178.9 t 7’7.6’

Bronchial Blood Flow, ml min-’ 100 g dry lung-’ l

l

Bronchial Flow, Cardiac

Blood 5%~of Output

Airway Blood Flow, ml min-’ 100 g dry wt-’ l

l

83.4 k 23.1 111.2 k 32.3 23.5 t 4.2”

1.6 k 0.5 1.9 t 0.8 0.7 t 0.2”

33.6 t 6.8 46.0 k 15.3 35.7 k 8.4

359.2

7.0 t 2.8*?

50.8 t 16.3

t 145.1*-t

Values are means t SE. PE, postembolization times. * Different from base line (P t Different from 5 and 60 min PE (P < 0.05). [From Malik and Tracy (299).]

< 0.05).

1188

ASRAR

B. MALIK

Volume

63

technique requires relatively little surgery, but it is necessary to correct for the number of tissue microspheres that shunt through peripheral arteriovenous anastomoses, and the correction factor is substantial (50-100%). The total bronchial arterial blood flow has also been estimated by creating a pouch in the portion of the descending aorta from which the bronchial arteries orginate (198). All arteries originating from the aorta except the bronchial arteries are ligated, and blood flow in the ascending aorta is shunted to the descending aorta. The flow of blood through this pouch was then measured with a rotameter. In dogs with intact nervous systems the bronchial arterial flow averaged 30 ml/min, which is similar to the microsphere estimates (198). Another technique for estimating bronchial flows involves ligating the pulmonary artery and vein in the left diaphragmatic lobe and sampling the blood flow coming through an outflow cannula placed proximally to the venous ligation (8, 15, 547). The flow represents the bronchopulmonary anastomotic flow because no pulmonary blood flow exists. The mean value for the bronchopulmonary flow in the left diaphragmatic lobe ranged from 3.8 to 6.5 ml/min (15, 547). Because the lobe represents 26.2% of lung weight, the total bronchopulmonary flow reaching the left atrium was estimated to be 18.3 ml/min. The flow may not represent true anastomotic flow, however, because ligation of the pulmonary artery may in itself alter the flow distribution. The bronchoesophageal artery in the intact sheep has also been used to measure bronchial blood flow, because it is believed to be the major source of bronchial blood flow in sheep (287). Estimates of flow measured with electron magnetic flow probes ranged from 5 to 14 ml/h, or only 0.4% of the cardiac output (287). These estimates suggest that there is a low bronchial blood flow in sheep or that other arteries supply much of the bronchial blood flow. Bronchial blood flow has been estimated in sheep with the aortic-pouch technique (558), and valueslare twice as high as those reported in the bronchoesophageal artery (287). Some generalizations can be made about the normal bronchial blood flow. 1) The flow is an extremely small portion of total cardiac output (representing only l-3%), but in absolute units the flow is 30 ml/min. 2) The microsphere method is less traumatic and is the most physiological method for measuring bronchial blood flow. 3) The relationship between the relative fractions of bronchial blood flow entering the true bronchial veins and the fractions entering the anastomoses needs further quantification. B. Eflects of Pulmonary

lkficnwnbolization

on Bronchial

Bloud Fbw

I. Acute alterations

jetted

Branch i al blood flow measured in dogs with labeled microspheres ininto the left atrium indicates th at flow did not decrease from base-

July

1983

PULMONARY

MICROEMBOLISM

1189

line levels at 5 min after glass-bead microembolization but was decreased at 60 min (see Table 3). Moreover the decreased flow was associated with an increased bronchovascular resistance (299), implying vasoconstriction at some vascular element within the bronchial circulation. The time lag probably indicates some systemic vasoconstrictor agents being slowly released after microembolization rather than some rapidly acting neural mechanisms (88). 2. Consequences of bronchial

hypoperfidm

The decreased bronchial blood flow observed after microembolization is of pathological significance because normal bronchial perfusion is necessary to maintain the tissue integrity of the lung (389, 517). Alterations in morphology, surface-tension properties, and regional ventilation occurred as early as one day after ligation of a pulmonary artery (389), presumably because collateral bronchial perfusion cannot maintain adequate oxygenation and substrate delivery to the bronchial structures. In the dog, the minimal blood flow requirement was estimated to be 5 ml min-l *kg-l body wt, which is -5% of cardiac output (87, 498). When pulmonary blood flow is stopped after microembolization, the flow requirements must be met by an increased bronchial blood flow; however, because bronchial blood flow was reduced to one-third of the base-line value within 60 min after embolization, this system may not provide adequate O2 and substrates to the alveolar structures and small airways (299). The cells likely to be affected by poor oxygenation and substrate delivery are the metabolically active cells, such as the type II alveolar epithelial cells that synthesize surfactant and the endothelial cells that modify several vasoactive factors (see Table 2; 138). l

3. Regulation

of bronchial

blood jlow

This section discusses the. neurohumoral and mechanical factors involved in the regulation of bronchial blood flow and how these factors contribute to the observed decrease in bronchial blood flow after microembolization (299). a) N&urohuwwraZ factors. The bronchial circulation is under greater neurogenic and humoral control than the pulmonary circulation (104, 129). Sympathetic stimulation results in bronchial vasoconstriction, and vagal stimulation results in vasodilation; these effects can be inhibited by appropriate antagonists (56). The bronchial circulation also demonstrates tonic neural control, because flow increases after sectioning of the sympathetic nerves and decreases after sectioning of the vagus (56). Epinephrine and norepinephrine cause vasoconstriction and decrease the bronchial blood flow (56). Serotonin and histamine, which are released after pulmonary microembolization, are bronchial vasodilators (299). Sympathetic factors may there-

1190

ASRAR B.MALIK

Vdume

63

fore contribute to the postembolization bronchial hypoperfusion, because sympathetic stimulation is one of the few interventions that clearly produces bronchial vasoconstriction (56). Sympathetic nerve stimulation and catecholamine infusions decreased bronchial arterial flow and bronchial venous outflow, but the bronchopulmonary anastomotic flow increased (56,310). These studies suggest that the primary sites of constriction associated with sympathetic stimulation (and by inference with microembolization) are the bronchial veins (104). Hypoxemia and hypercapnia of the bronchial artery blood flow have also been shown to cause increased bronchial vascular resistance (56, 283, 299). However, breathing a gas low in O2 (7.5 or 10%) and high in CO2 (5 or 7%) appears to cause vasodilation and increased bronchial blood flow (56). Thus the responses to hypoxemia and hypercarbia in the bronchial circulation are not clear at this time, nor is it clear to what extent the’changes in blood gases influence bronchomotor tone after microembolization. b) lMechanicaZjii&xw. I) PERFUSIONPRESSURE. In addition to the neural and humoral factors already discussed, a decrease in bronchial perfusion pressure can also produce bronchial hypoperfusion (12440,531). In a preparation in which the pulmonary and bronchial vascular beds were separately perfused so’that the systemic arterial pressures or the Ppa values could be independently varied, the bronchial blood flow decreased as the systemic artery-pulmonary artery pressure gradient was decreased, and the flow increased as the gradient was increased (12, 440, 531). Thus bronchial blood flow is dependent on the perfusion pressure and does not appear to autoregulate when pressure is altered. Aramendia et al. (7), however, have demonstrated that flow in true bronchial veins (i.e., the veins draining the bronchial blood flow into the right atrium) was constant in half of the animals at perfusion pressure >lOO mmHg, yet total bronchial blood flow did not remain constant. Therefore a portion of the bronchial circulation (bronchial veins) may regulate its blood flows by other means (i.e., by humoral mechanisms such as adenosine). The decreased bronchial blood flow associated with embolization may be explained by the perfusion pressure because the systemic arterial-pulmonary arterial pressure gradient was decreased (299). II) SYSTEMIC VENOUS HYPERTENSION. Systemic venous hypertension, which is associated with severe pulmonary microembolization (438), can also reduce bronchial venous flow and can thus contribute to the decreased bronchial blood flow observed after microembolization (56). Nevertheless the autoregulation of the bronchial veins may prevent significant changes in bronchial venous pressure resulting from systemic venous hypertension (7). A Long-term

alterations

One year after pulmonary ar ltery ligation, the left lower lobe had increased dramatically

the collateral blood flow in from the base-line range of

July

1983

PULMONARY

MICROEMBOLISM

1191

4-9 ml/min to 38-39 ml/min. The value of the compensatory enlargement of the bronchial arteries is obvious because infarcts were never seen in the obstructed vessels (122, 125, 183, 200, 274), although necrosis and infarcts were common when collateral flow was not allowed to increase by chronically reducing the systemic arterial pressure (389). In addition to maintaining substrate-delivery and preventing tissue injury, another important consequence of increased bronchial circulation is that it allows gas exchange to occur to some extent in the obstructed lung segments (276). Three years after ligation of the left pulmonary artery in dogs there was only a slight decrease in ventilation of the left lung (from the preligation value of 42.5% of the total ventilation to 34.5%), and O2 uptake in the left lung was reduced to only 11% of the total (276). The increased bronchial circulation after chronic pulmonary arterial obstruction allowed O2 uptake to continue, although at a reduced rate. Weibel(528) has made a quantitative assessment of the long-term bronchial vascular alterations after pulmonary vascular occlusion. Proliferative changes in the endothelium and media of bronchial arteries were observed 5 days after ligation of a pulmonary artery. These changes were accompanied by the formation of new bronchial arteries as well as by enlargement of preexisting vessels. An increase in bronchial blood flow was also noted with the microsphere technique after glass-bead microembolization (299); the flow increased from 2% of cardiac output to 7% of the cardiac output within a relatively short period of 2 wk (Table 3). Chronic hypoxemia cannot be implicated as the mechanism of the increased vascularity, because bronchial blood flow did not increase in sheep born and raised under hypoxic conditions at high altitude (558). Because new vessels were found in both ligated and nonligated regions of the lung, Weibel’s (528) suggestion that vessel proliferation after pulmonary artery obstruction occurs as a result of release of “antiogentic factor(s)” is a reasonable one. VII.

CONCLUSIONS

Recent studies have reaffirmed the view of Starling and Verney (469) that the pulmonary circulation serves as a filter for the removal of humoral substances such as serotonin and also for the removal of particulate matter originating in the peripheral circulation. Owing to this filtering function there are certain homeostatic adjustments (such as the bronchoconstriction and vasoconstriction occurring primarily in areas in which vessels are obstructed) that tend to minimize the deleterious effects of vascular obstruction. Bronchoconstriction and vasoconstriction lessen the ventilation and perfusion imbalance that would otherwise occur. Enlargement of the bronchial circulation is an example of a long-term alteration that occurs after pulmonary vascular obstruction. The anatomical and physiological change minimizes necrosis within pulmonary tissue and allows for some gas ex-

1192

ASRAR

B. MALIK

vohme

63

change to occur even though pulmonary arteries are obstructed. The ability of the lung to receive particulate matter from the periphery has limits, however, because a point is reached at which the compensatory mechanisms are overwhelmed and the reserves are fully utilized; consequently pulmonary edema develops and gas exchange becomes inefficient. Another major conclusion drawn from previous studies is that the pulmonary responses occurring after microembolism appear to be mediated primarily through humoral mechanisms. These include the pulmonary vasoconstrictor, bronchoconstrictor, and tachypneic responses and the increased pulmonary vascular permeability. The humoral substances are activated or released as a direct result of obstruction, intravascular coagulation, and leukocyte and platelet activation. Some of the responses (e.g., tachypnea) also have a major neural component, but the neural mechanisms may be triggered by humoral factors such as histamine and prostaglandins. Although intravascular coagulation and leukocyte and platelet activation are important in mediating the various pulmonary responses to microembolization, the precise mediators released after the activation of these blood components and the steps that produce the aggregation and activation have not been clearly delineated. There is a need for a more complete understanding, of the humoral control of pulmonary vessels, airways, ventilation, and lung fluid balance. Pulmonary microembolism offers a unique model because humoral factors appear to play such a dominant role in mediating the pulmonary responses. I deeply appreciate the generous comments of Dr. Aubrey Taylor and Dr. Leonard Grumbath, who reviewed this manuscript. I thank Kathleen Roche for her patient efforts in typing this manuscript. The author’s research was supported by National Institutes of Health Grants HL-17355, HL-26551, and HL-27016 and by Research Career Development Award HL-00363.

REFERENCES 1. ALBERT, R. K., S. LAKSHMINARAYAN, T. W. HUANG, AND J. BUTLER. Fluid leaks from extra-alveolar vessels in living dog lungs. J. Appl Phyaiol: Reqimk Environ. Exercise Ph&cd 44: 759-762, 1978. 2. ALLEY, R. D., L. H. S. VAN MIEROP, A. S. PECK, H. W. KAUSEL, AND A. STRANAHAN. Bronchial arterial collateral circulation. Am Rev. Reap+. Dis. 83: 3137, 1961. 3. ALPERT, J. S., J. GODTFREDSEN, I. S. OCKENE, J. ANAS, AND J. E. DALEN. Pulmonary hypertension secondary to minor pulmonary embolism. Chest 73: 795-797, 1978. 4. ALPERT, J. S., F. W. HAYNES, J. E. DALEN, AND L. DEXTER. Experimental pulmonary embolism: effect on pulmonary blood volume and vascular compliance. Circulutiorr 49: 152-157, 1974. 5. ANDERSON, F. L., W. JUBIZ, T. J. TSAGARIS, AND H. KUIDA. Endotoxin-induced prostaglandin E and F release in dogs. Am J. Physid 228: 410-414,1975. 6. ANURAS, J., F. H. F. CHENG, AND H. B. RICHERSON.

7.

8.

9.

10.

11.

Experimental leukocyte-induced pulmonary vasculitis with inquiry into mechanism. Cheat 71: 383-387, 1977. ARAMENDIA, P., J. MARTINEZ, L. DE LETONA, AND D. M. AVIADO. Responses of the bronchial veins in a heart-lung-bronchial preparation with special reference to a pulmonary to bronchial shunt. Circ Rea 10: 3-10, 1962. ARAMENDIA, P., J. MARTINEZ, L. DE LETONA, AND D. M. AVIADO. Exchange of blood between pulmonary and systemic circulations via bronchopulmonary anastomoses. Circ &.a 11: 870-879, 1962. ARFORS, K. E., C. BUSCH, S. JAKOBSSON, 0. LINDQUIST, P. MALMBERG, L. RAMMER, AND T. SALDEEN. Pulmonary insufficiency following infusion of thrombin and AMCA (tranexamic acid) in the dog. Actu Chir. Scand 138: 445-452, 1972. ARMSTRONG, D. C., AND S. A. MILLER. Lung irritant and C-fibre response to embolism in thrombocytopenic rabbits (abstr.). J. Physid London 363: 41P-42P, 1980. AUKLAND, K., AND G. NICOLAYSEN. Interstitial fluid

July

19&3

PULMONARY

MICROEMBOLISM

volume: local regulatory mechanisms. Physid Rev. 61: 556-643,1981. 12. AULD, P. A. M., A. M. RUDOLPH, AND R. J. GOLINKO. Factors affecting bronchial collateral flow in the dog. Am J. Physid 198: 1166-1170,196o. 13. AUSTEN, K. F. Systemic analphylaxis in the human being. N. ET& J. Med 291: 661~664,1974. 14. AVIADO, D. M. Pulmonary embolism. In: The Lung Circulation. Oxford, UK: Pergamon, 1965, pt. 2, p. 933-965. 15. AVIADO, D. M., M. DE B. DALY, C. Y. LEE, AND C. F. SCHMIDT. The contribution of the bronchial circulation to the venous admixture in pulmonary venous blood. J. Phgsid London 155: 602-622,196l. 16. BABIOR, B. M. Oxygen-dependent microbial killing by phagocytes. N. EngL .l Med 298: 659-668,1978. 1’7. BACHOFEN, M., AND E. R. WEIBEL. Basic pattern of tissue repair in human lungs following unspecified injury. cheet 65: 145-195, 1974. 18. BACHOFEN, M., AND E. R. WEIBEL. Alterations of the gas exchange apparatus in adult respiratory insufficiency associated with septicemia. Am Rev. Respir. Dis. 116: 589-615, 1977. 19. BADWEY, J. A., J. T. CURNUTTE, AND M. L. KARNOVSKY. The enzyme of granulocytes that produces superoxide and peroxide. N. Engl. J. Meti 300: 1157-1166, 1979. 20. BALIS, J. U., E. S. RAPPAPORT, L. GERBER, J. FAREED, F. BUDDINGH, AND H. L. ME&ORE. A primate model for prolonged endotoxin shock. Blood-vascular reactions and effects of glucocorticoid treatment. tab. Znvest. 38: 511-523, 1978. 21. BARER, G. R., AND E. NUSSER. The part played by bronchial muscles in pulmonary reflexes. Br. J. Phurmad 8: 315-320,1953. 22. BARIE, P. S., T. S. HAKIM, AND A. B. MALIK. Effect of pulmonary artery occlusion and reperfusion on lung fluid and protein exchange. J. AppL Physid: Reqdmt. Envirm Exerbe PhydoL 50: 102-166, 1981. 23. BARIE, P. S., AND A. B. MALIK. Metabolic factors as mediators of pulmonary vascular injury. Ann. NY Acad sci 384: 344-354,1982. 24. BARIE, P. S., F. L. MINNEAR, AND A. B. MALIK. Increased pulmonary vascular permeability after bone marrow injection in sheep. Am Rev. R&r. Dis. 123: 64%653.1981. 25. BARIETY, M., AND K. KOHLER. Sur l’existence d’un facteur histaminique dans le d&erminisme de I’oed&me pulmonaire aigu adrbnalinique chez le lapin. C. R Ad sci 236: 1389-1391, 1953. 26. BARITUSSO, A., AND J. CLEMENTS. Acute effects of pulmonary artery occlusion on the pool of alveolar surfactant. &upir. Phyaid 45: 3%336,198l. 27. BAYLEY, T., J. A. CLEMENT& AND A. J. OSBAHR. Pulmonary and circulatory effects of fibrinopeptides. Circ Rea 21: 469-485, 1967. 28. BEAVER, M. A. Histamine. N. EngL J. Med 294: 30-36, 320-325.1976. 29. BEEBE, D. C., D. E. FEAGANS, E. J. BLANCHETTEMACKIE, AND M. E. NAU. Lens epithelial cell elongation in the absence of microtubules: evidence for a new effect of colchicine. science 206: 836-837, 1979. 29a.BELL, D. R., R. J. MULLINS, AND M. R. POWERS. Extravascular distribution of albumin and IgG during highpermeability edema in skin. Am J. Phyd 244 (Heart Cirr: Physid 13): H599-H606,1983. 30. BELL, W. R., N. D. ANDERSON, AND A. 0. ANDERSON. Heparin-induced coagulopathy. J. L&A C&r. Med 89: 741759,1977.

31. BELL, W. R., S. S. SHAPIRO, J. MARTINEZ, AND H. L. NOSSEL. The effects of ancrod, the coagulating enzyme from the venom of Malayan pit viper (A. rhoddoma) on prothrombin and fibrinogen metabolism and fibrinopeptide A release in man. J. Lab. Clin, Msd 91: 592~604,1978. 32. BERGER, A. J., R. A. MITCHELL, AND J. W. SEVERINGHAUS. Regulation of respiration. N. EngL J. Med 297: pt. 1, 92-97; pt. 2, 138-143; pt. 3, 194-201; 1977. 32a.BERNARD, G..R., J. R. SNAPPER, A. A. HUTCHINSON, AND K. L. BRIGHAM. Effects of left atria1 pressure elevation during histamine infusion on lung fluid balance in sheep (abstr.). Federation Proc 41: 7199, 1982. 33. BERNTHAL, T., A. D. HORRES, AND J. T. TAYLOR III. Pulmonary vascular obstruction in graded tachypneagenic diffuse embolism. Am. J. Physid 200: 279-286, 1961. 34. BHATTACHARYA, J., AND N. C. STAUB. Direct measurements of microvascular pressures in the isolated perfused dog lung. * 210: 327-328, 1980. 35. BIANCONI, R., AND J. H. GREEN. Pulmonary baroreceptors in the cat. Arch, ZtaL Bid 97: 305-310, 1959. 36. BINDER, A. S., W. KAGELER, A. PEREL, M. R. FLICK, AND N. C. STAUB. Effect of platelet depletion on lung vascular permeability after microemboli in sheep. J. AppL Physid: Reapirat. Envirtm Exercise PhysiaL 48: 414-420,198o. 37. BINDER, A. S., K. NAKAHARA, K. OHKUDA, W. KAGELER, AND N. C. STAUB. Effect of heparin or fibrinogen depletion on lung fluid balance in sheep after emboli. J. AppL Ph@oL: Reqirat Environ Exercise Physid 47: 213-219, 1979. 38. BINDER, A. S., AND N. C. STAUB. Increased lung vascular permeability after microemboli does not depend on fibrin deposition (abstr.). Am Rev. Reqir. Dis. 117: 311, 1978: 38a.BIZIOS R., F. L. MINNEAR, H. VAN DER ZEE, AND A. B. MALIK. Effects of cyclooxygenaae and lipoxygenase inhibition on lung fluid balance after thrombin. J. AppL Physid in press. 39. BLAND, R. D., R. H. DEMLING, S. L. SELINGER, AND N. C. STAUB. Effects of alveolar hypoxia on lung fluid and protein transport in unanesthetized sheep. Circ Rea 40: 269-274, 1976. 40. BO, G., AND J. HOGNESTAD. Pulmonary vascular effects of suddenly induced unilateral blood platelet aggregation. Acttz PhyaiuL Scud 90: 237-243,1974. 41. BO, G., J. HOGNESTAD, AND J. VAAGE. The role of blood platelets in pulmonary responses to microembolization with barium sulphate. Actu PhgsiuL Scud 90: 244-251, 1974. 42. BOHM, G. M. Vascular permeability changes during experimentally produced pulmonary edema in rats. J. Pathd BacteriaL 92: 151-161, 1966. 43. BONE, R. C., P. B. FRANCIS, AND A. K. PIERCE. Intravascular coagulation associated with adult respiratory distress syndrome (abstr.). Am Rev. Respir. Dis. 111: 920, 1975. 44. BONE, R. C., S. LERNER, D. STECHSCHULTE, E. MURPHY, J. WOLFE, R. SOBONYA, AND J. HOOD. Vasoactive mediators in the pathogenesis of experimental acute respiratory failure. Chest 75, Suppl.: 218-220, 1979. 45. BORN, G. V. R. Acute pulmonary oedema in isolated perfused lungs of rabbits. J. Ph@oL London 124: 502514, 1954. 46. BOURNE, H. R., L. M. LICHTENSTEIN, K. L. MELMON, C. S. HENNEY, Y. WEINSTEIN, AND G. M. SHEARER. Modulation of inflammation and immunity by cyclic

ASRAR

47.

48.

49.

50. 51.

52.

53.

54.

55.

56.

57. 58.

59.

69.

61.

62.

63.

64.

AMP: receptors for vasoactive hormones and mediators of inflammation regulate many leukocyte functions. Science 184: 19-28,1974. BOWERS, R. E., E. F. ELLIS, K. L. BRIGHAM, AND J. A. OATES. Effects of prostaglandin cyclic endoperoxides on the lung circulation of unanesthetized sheep. J. clin Invest. 63: 131-137, 1979. BOYER, N. H., AND J. J. CURRY. Bronchospasm associated with pulmonary embolism. Arch, Intern Med 73: 403-499,1944. BRAY, B. A. Cold insoluble globulin (fibronectin) in connective tissues of adult human lung and trophoblast basement membrane. J. CZin. Iwe& 62: 745-752, 1978. BRIGHAM, K. L. Factors affecting lung vascular permeability. Am Rev. Respir. Dia. 115: 165-172, 1977. BRIGHAM, K. L. Lung edema due to increased vascular permeability. In: Lurrg Waler and SoZ&, edited by N. C. Staub. New York: Dekker, 1978, p. 235-276. BRIGHAM, K. L., R. E. BOWERS, AND C. R. McKEEN. Methylprednisolone prevention of increased lung vascular permeability following endotoxemia in sheep. J. Cl&. Invest. 67: 1103-1110,198l. BRIGHAM, K. L., R. E. BOWERS, AND P. J. OWEN. Effects of antihistamines on the lung vascular response to histamine in unanesthetized sheep. Diphenhydramine preventions of pulmonary edema and increased permeability. J. clin Invest. 58: 391-398, 1976. BRIGHAM, K., AND P. J. OWEN. Mechanisms of the serotonin effect on lung transvascular fluid and protein movement in unanesthetized sheep. Circ Res. 36: 761770, 1975. BRIGHAM, K. L, S. PADORO, D. BRYANT, C. McKEEN, AND R. BOWERS. Diphenhydramine reduces the endotoxin effects on lung vascular permeability in sheep. J. Appl Physid: Resp&rat. Envirm Exercise Physid 49: 516-520, 1980. BRUNER, H. D., AND C. F. SCHMIDT. Blood flow in the bronchial artery of the anesthetized dog. Am. .L Physid 148: 648-666, 1947. BUNDGAARD, M. Transport pathways in capillariesin search of pores. Anna Rev. Physid 42: 325-336,198O. BURGESS, J. H., J. GILLESPIE, P. D. GRAF, AND J. A. NADEL. Effect of pulmonary vascular pressures on single-breath CO diffusing capacity in dogs. J. Ad Physid 24: 692-696,1968. BUSCH, C., 0. LINDQUIST, AND T. SALDEEN. Respirtory insufficiency in the dog induced by pulmonary microembolism and inhibition of flbrinolysis. Effect of deBbrinogenation, leukopenia and thrombocytopenia. Acta Chir. Scund 140; 255-266, 1974. BUSCH, C., AND L. RAMMER. Quantitation of fibrin deposition and elimination in organs of rats injected with labelled fibrinogen by isolation of the labelled fibrin from water soluble tracer. Thti Diath. Haemow h 29: 108114, 1973. BUSCH, C., L. RAMMER, AND T. SALDEEN. Quantitation of fibrin deposition and elimination in organs of rats injected with labelled fibrinogen and albumin. Thmmbin LXath H aemmrh 29: 94-106,1973. CAHILL, J. M., E. 0. ATTINGER, AND J. J. BYRNE. Ventilator-y responses to embolization of lung. J. AppL Physid 16: 469-472, 1961. CALDINI, P. Pulmonary hemodynamics and arterial oxygen saturation in pulmonary embolism. J. Appl. Physid 20: 184-199,1965. CANTER, R. C., AND A. AMES III. Experimental prevention of cerebral vasculature obstruction produced by ischemia. J. Nacroeurg. 30: 50-54, 1969.

B. MALIK

Vohne

63

65. CARLSON, R. W., R. C. SCHAEFFER, JR., M. CARPIO, AND M. H. WEIL,. Edema fluid and coagulation changes during fulminant pulmonary edema. Chest 79: 43-49, 1981. 66. CARR, J. The anti-inflammatory action of heparin: heparin as an antagonist to histamine, bradykinin and prostaglandin El. Thromb. Res. 16: 507-516, 1979. 67. CHANG, H. K., L. DELAUNOIS, R. BOILEAU, AND R. R. MARTIN. Redistribution of pulmonary blood flow during experimental air embolism. .I. AppL Physid: Respirat Environ. Exercise Phpid 51: 211-217, 1981. 68. CHANG, R. S. Y., P. SILVERMAN, AND R. M. EFFROS. Vascular and extravascular compartments of the isolated perfused rabbit lung. J. Appl Physid: Respirat. Environ. Exercise Physid 46: 74-78,1979. 69. CHERNICK, V., W. A. HODSON, AND L. J. GREENFIELD. Effect of chronic pulmonary artery ligation on pulmonary mechanics and surfactant. J. Appl Physid 21: 1315-1320, 1966. 70. CHINARD, F. P., AND D. 0. DEFOUW. Ultrastructural changes of the air-blood barrier after spontaneous development of edema in isolated dog lungs perfused at 15’. Microvasc. Res. 21: 48-56,198l. 71. CHRISTODOULOU, J., R. ERLANDSON, C. SMITHEN, T. KILLIP, AND N. BRACHFELD. Effects of mannitol on cardiac ultrastructure and microcirculation following anoxia. Am J. Physid 229: 853-860, 1975. 72. CHU, J., J. A. CLEMENT& E. K. COTTON, M. H. KLAUS, A. Y. SWEET, AND W. H. TOOLEY. Neonatal pulmonary ischemia. I. Clinical and physiological studies. Pediatrics 40, Suppl.: 799-782, 1967. 73. CLARKE, S. W., P. D. GRAF, AND J. A. NADEL. In vivo visualization of small-airway constriction after pulmonary microembolism in cats and dogs. J. AppL Physid 29: 646-650, 1970. 74. CLAY, T. P., AND J. M. B. HUGHES. The role of prostaglandins in the bronchoconstriction induced by pulmnary micro-embolism in the guinea-pig. J. Physid Lvndim 308: 427-437, 1980. 75. CLOUGH, G., AND C. C. MICHEL. The labelling of endothelial cell vesicles with ferritin in individually perfused capillaries. Micrmmc Res. 17, Suppl.: S159, 1979. 76. COBB, F. R., P. A. McHALE, AND J. C. REMBERT. Effects of acute cellular injury on coronary vascular reactivity in awake dogs. Circulation 57: 962-968, 1978. 77. COE, N. P., AND E. W. SALZMAN. Thrombosis and intravascular coagulation. Suq~ Clin. North Am 56: 875890,1976. 78. COLEBATCH, H. J. H., C. R. OLSEN, AND J. A. NADEL. Effect of histamine, serotonin, and acetylcholine on the peripheral airways. J. AppL Phgsid 21: 217-226, 1966. 79. COLERIDGE, H. M., AND J. C. G. COLERIDGE. Impulse activity in afferent vagal C-fibres with endings in the intrapulmonary airways of dogs. Respir. Physid 29: 125142, 1977. 80. COLERIDGE, H. M., J. C. G. COLERIDGE, D. G. BAHER, K. H. GINZEL, AND M. A. MORRISON. Comparison of the effects of histamine and prostaglandin on afferent C-fiber endings and irritant receptors in the intrapulmonary airways. Adv. Exp. Med Bid 99: 291-305, 1978. 81. COLERIDGE, H. M., J. C. G. COLERIDGE, A. DANGEL, C. KIDD, J. C. LUCK, AND P. SLEIGHT. Impulses in alowly conducting vagal fibers from afferent endings in veins, atria and arteries of dogs and cats. Circ Rsa 33: 87-97, 1973. 82. COLERIDGE, H. M., J. C. G. COLERIDGE, AND J. C. LUCK. Pulmonary afferent flbres of small diameter atim-

July

83.

84.

85.

86.

8’7. 88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

1983

PULMONARY

MICROEMBOLISM

ulated by capsaicin and by hyperinflation of the lungs. J. Physid London 179: 248-262, 1965. COLERIDGE, J. C. G., AND H. M. COLERIDGE. Afferent C-fibers and cardiorespiratory chemoreflexes. Am. Rev. Re8pir. Dia 115, Suppl.: 251-260, 19’77. COLERIDGE, .I. C. G., AND C. KIDD. Electrophysiological evidence of baroreceptors in the pulmonary artery of the dog. J. PhysioL London 150: 319-331, 1969. COLERIDGE, J. C. G., AND C. KIDD. Reflex effects of stimulating baroreceptors in pulmonary artery. J. Phy&id Lc-mion 166: 19%210,1963. COLLIER, H. 0. J. Role of kallikrein-kinin system in lung diseases. In: Chemigtry and Bidogy of the KaUikreinKinin System in Health and Disease, edited by J. J. Pisano and K. F. Austen. Washington, DC: US Govt. Printing Office, 1977, p. 495-503. COMROE, J. H., JR. The main functions of the pulmonary circulation. Cirfxulutifm 33: 146-158, 1966. COMROE, J. H., JR., B. VAN LINGEN, R. C. STROUD, AND A. RONCORONI. Reflex and direct cardiopulmonary effects of 5-OH-tryptamine (serotonin). Their possible role in pulmonary embolism and coronary thrombosis. Am. J. Physid 173: 379-386, 1953. COOPER, J. D., J. W. D. MCDONALD, M. ALI, E. MENKES, J. MASTERSON, AND P. KLEMENT. Prostaglandin production associated with the pulmonary vascular response to complement activation. Surgery 88: 215~221,198O. COPLEY, A. L. Platelets and physiological defense mechanisms. In: Pla&Zetx A iUu&idisc@lina~ Approach, edited by G. de Gaetano and S. Garatini. New York: Raven, 1978, p. 45-59. COPLEY, A. L., J. P. HANIG, B. W. LUCHINI, AND R. L. ALLEN, JR. Capillary permeability enhancing action of fibrinopeptides isolated during fibrin monomer formation (abstr.). Federation Pruc 25: 446, 1966. COSTABELLA, P. M., 0. LINDQUIST, Y. KAPANCI, AND T. SALDEEN. Increased vascular permeability in the delayed microembolism syndrome. Micrauasc, Rea 15: 275-286, 1978. COTRAN, R. S. The delayed and prolonged vascular leakage in inflammation. II. An electron microscopic study of the vascular response after thermal injury. Am. J. Path& 46: 589-620, 1965. COTRAN, R. S. Studies on inflammation: ultrastructure of the prolonged vascular response induced by Cdstridium oedmatierrs toxin. Lab. Invest. 17: 39-60, 1967. COWAN, G. S. M., JR., N. C. STAUB, AND L. H. EDMUNDS, JR. Changes in the fluid compartments and dry weights of reimplanted dog lungs. J. AppL Physid 40: 962-970, 1976. CRADDOCK, P. R., J. FEHR, K. L. BRIGHAM, R. S. KRONENBERG, AND H. S. JACOB. Complement and leukocyte-mediated pulmonary dysfunction in hemodialysis. N. Engl J. Med 296. 769-774,1977. CRADDOCK, P. R., J. FEHR, A. P. DALMASSO, K. L. BRIGHAM, AND H. S. JACOB. Hemodialysis leukopenia: pulmonary vascular leukostasis resulting from complement activation by dialyses cellophane membranes. J. Clin InveaL 59: 879-888,1977. CRAPO, J. D., M. PETERS-GOLDEN, J. MARSH-SALIN, AND J. S. SHELBURNE. Pathologic changes in the lung of oxygen-adapted rats-a morphometric study. Lab Invest. 39: 6404x3,1978. CUNNINGHAM, A. L., AND J. V. HURLEY. Alpha-napthyl-thiourea-induced pulmonary edema in the rat: a topographic and electron microscopic study. J. Path& 106: 25-42,1972.

1195

160. DAILY, P. O., A. GRIFFIN, E. BLACKSTONE, AND P. V. MOULDER. The hemodynamic effects of localized pulmonary embolism. Surg CyLeod obstet. 123: 765-773, 1966. 101. DALEN, J. E., J. S. BANAS, JR., H. L. BROOKS, G. L. EVANS, J. A. PARASKOS, AND L. DEXTER. Resolution rate of acute pulmonary embolism in man. N. EngL J. Med 286: 1194-1199,1969. 102. DALEN, J. E., C. I. HAFFAGEE, J. S. ALPERT, J. P. HOWE, I. S. OCKENE, AND J. A. PARAKOS. Pulmonary embolism, pulmonary hemorrhage and pulmonary infarction. N. Engl J. Med 296: 1431-1435.1977. 103. DALEN, J. E., F. .W. HAYNES, F. G. HOPPIN, G. L. EVANS, P. BHARDWAJ, AND L. DEXTER. Cardiovascular responses to experimental pulmonary embolism. Am J. Cardid 20: 3-9,1967. 104. DALY, I. DE B., AND C. HEBB. Pulnuma7y and Bronchial Vascular Systema Baltimore, MD: Williams & Wilkins, 1966. 105. DALEY, R., J. D. WADE, F. MARAIST, AND R. J. BING. Pulmonary hypertension in dogs induced by injection of lycopodium spores into the pulmonary artery, with special reference to the absence of vasomotor reflexes. Am J. PhysioL 164: 380-390,195l. 106. D’AMORE, P. Platelet-endothelial interaction and the maintenance of the microvasculature. Miowuaac Rtx 15: 137-145, 1978. 107. DANIELLI, J. F. Capillary permeability and oedema in the perfused frog. J. Phyaid London 98: 169-129, 1940. 108. DANTZKER, D. R., P. D. WAGNER, V. W. TORNABENE, N. P. ALAZRAKI, AND J. B. WEST. Gas exchange after pulmonary thromboembolism in dogs. Circ R&a 42: 92-103, 1978. 109. DAY, B., AND S. M. FRIEDMAN. Intracellular sodium and potassium changes in vascular smooth muscle during hemorrhagic shock. Surg. Gylecd obstet. 147: 25-26, 1978. 110. DEFORREST, J. M., AND T. M. HOLLIS. Shear stress and aortic histamine synthesis. Am. .I. Phy& 234 (Heart Circ Physid 3): H701-H705,1978. 111. DEFOUW, D. O., AND P. B. BERENDSON. Morphological changes in isolated perfused dog lungs after acute hydrostatic edema. Circ Re.a 43: 72-82, 1978. 112. DEKOCK, M. A., J. A. NADEL, S. ZWI, H. J. H. COLEBATCH, AND C. R. OLSEN. New method for perfusing bronchial arteries: histamine bronchoconstriction and apnea. J. AppL Phpsid 21: 185-194, 1966. 113. DEL MAESTRO, R. F., H. H. THAW, J. BJORK, M. PLANKER, AND K. E. ARFORS. Free radicals as mediators of tissue injury. Acta Physid &and SuppL 492: 43-58, 1980. 114. DEMLING, R. H., M. SMITH, R. GUNTHER, M. GEE, AND J. FLYNN. The effect of prostacyclin infusion on endotoxin-induced lung injury. Surgery 89: 257-263,198l. 115. DEXTER, L., AND G. T. SMITH. Quantitative studies of pulmonary embolism. Am. J. Med Sci 35: 641~648,1964. 116. DODDS, W. J. Platelet function in animals: species specificaties. In: Platelets A Mu&disciplinary Approach, edited by G. de Gaetano and S. Garatini. New York: Raven, 1978, p. 45-59. 117. DRAKE, R., T. ADAIR, D. TRABER, AND J. GABEL. Contamination of caudal mediastinal node efferent lymph in sheep. Am. J. Physid 241 (Heart Circ Phyti 10): H354-H357,1981. 118. DRAKE, R. E., AND J. C. GABEL. Effect of histamine and alloxan on canine pulmonary vascular permeability. Am J. PhysioL 239 (Heart Circ Ph+L 8): H96-HlOO, 1986.

1196

ASRAR

119. DUNCAN, E. E., JR., J. CORNFIELD, AND K. BUDE. Effects of blood pressure in passage of labeled plasma albumin in canine aortic wall. J. Clin Invest. 41: 153’71542, 1962. 120. DUNN, J. S. The effects of multiple embolism of pulmonary arterioles. Q. J. Med 13: 129-147, 1920. 121 DUSTING, G. J., S. MONCADA, AND J. R. VANE. Prostaglandins, their intermediates and precursors: cardiovascular actions and regulatory roles in normal and abnormal circulatory systems. Prq7. Curdioumc Di.s 21: 405-430, 1979. 122 EDMUND& L. H., JR., AND G. L. HUBER. Pulmonary artery occlusion. I. Volume-pressure relationships and alveolar bubble stability. J. AppL PhytioL 22: 990-1091, 1967. 123 EFFROS, R. M. Small solutes and water. In: tune Water and salute Exchange, edited by N. C. St&b. New York: Dekker, 1978, p. 183-231. 124 ELLIS, F. H., JR., J. H. GRINDLAY, AND J. E. EDWARDS. The bronchial arteries. I. Experimental occlusion. Surgery 30: 810-826, 1951. 125 ELLIS, F. H., JR., J. H. GRINDLAY, AND J. E. EDWARDS. The bronchial arteries. II. Their role in pulmonary embolism and infarction. Surgery 31: 167-179, 1952. A. J., III, T. R. VAUGHAN, K. L. BRIGHAM, 1%. ERDMANN, W. C. WOOLVERTON, AND N. C. STAUB. Effects of increased vascular pressure on lung fluid balance in unanesthetized sheep. Circ. &a 37: 271-284, 1975. 127. FEARON, D. T.‘, AND F. AUSTEN. The alternative pathway of complement-a system of host resistance in microbial infection. N EngL .I. Med 303: 259-263, 1980. 127a.FEIGEN, L. P., AND A. L. HYMAN. Vascular influences of prostaglandina. Federation Proc 40: 1985-1986, 1981. 128. FEOLA, M., AND G. GLICK. Cardiac lymph flow and composition in acute myocardial ischemia in dogs. Am. J. PhysioL 229: 44-48, 1975. 129. FILLENZ, M. Innervation of pulmonary and bronchial blood veseele of the dog. J. And 106: 449-461,197O. 130. FILLENZ, M., AND J. G. WIDDICOMBE. Receptors of the lungs and airways. In: Handbook of i3enaoy Physidogy. Enteroceptar8, edited by E. Neil. Berlin: SpringerVerlag, 1972, vol. III, pt. 1, p. 81-112. 131. FINLEY, T. N., W. H. TOOLEY, E. W. SWENSON, R. E. GARDNER, AND J. A. CLEMENTS. Pulmonary surface tension in experimental atelectasis. Am. Rev. Reapir. L3iu 89: 372-378,1964. 132. FISHER, A. B., H. KOLLMEIER, J. S. BRODY, R. W. HYDE, J. HANSELL, J. N. FRIEDMAN, AND J. A. WALDHAUSEN. Restoration of systemic blood flow to the lung after division of bronchial arteries. J. AppL Php sid 29: 839-846, 1970. 133. FISHER, J., W. H. NOBLE, AND J. C. KAY. Hypoxemia following pulmonary embolism. Anestheaidosy 54: 204209, 1981. 134. FISHER, S. R., A. BURANCEAU, R. D. FLOYD, AND W. G. WOLFE. Comparative changes in ventilatory dead space following micro and massive pulmonary emboli. J. Surg Rea 20: 195201,1976. 135. FISHMAN, A. P. Respiratory gases in the regulation of the pulmonary circulation. PhysioL Rev. 41: 214-280, 1961. 136. FISHMAN, A. P. Dynamics of the pulmonary circulation. In: Handbook of Phytiolugp. Circulatian, edited by W. F. Hamilton and P. Dow. Washington, DC: Am. Physiol. sot., 1963, vol. II, p. 1667-1743. 137. FISHMAN, A. P. Chronic car pulmonale. Am. Rev. Re.spir. Dia 114: 775-794, 1976.

B. MALIK

t’idume

63

138. FISHMAN, A. P., AND G. G. PIETRA. Handling of bioactive materials by the lung. N. EngL J. MecL 291: 884-889, 1974. 139. FISHMAN, A. P., AND G. G. PIETRA. Permeability of pulmonary vascular endothelium. In: Lung Lipids, edited by R. Porter and M. O’Connor. Amsterdam: Elsevier, 1976, p. 29-48. (Ciba Found. Symp. 38.) 140. FISHMAN, A. P., AND G. G. PIETRA. Hemodynamic pulmonary edema. In: Pdmmu7y Edemu, edited by A. P. Fishman and E. M. Renkin. Bethesda, MD: Am. Physiol. Sot., 1979, p. 79-96. 141. FLICK, M. R., J. HOEFFEL, AND N. C. STAUB. Superoxide dismutase prevents increased vascular permeability after microemboli (abstr.). Federation Proc 38: 549, 1979. 142. FLICK, M. R., A. PEREL, AND N. C. STAUB. Leukocytes are required for increased lung vascular permeability after microembolization in sheep. Circ Re-a 48: 344351, 1981. 143. FLORES, J., D. R. BIBONA, C. H. CECK, AND A. LEAF. The role of cell swelling in ischemic renal damage and the protective effect of hypertonic saline. J. Clin Invest. 51: 118-126, 1972. 144. FLOWER, R. J. Drugs which inhibit prostaglandin biosynthesis. PhmmmoL Rev. 26: 33-67, 1974. 145. FOLKOW, B., AND J. R. PAPPENHEIMER. Components of the respiratory dead space and their variation with pressure breathing and with bronchoactive drugs. J. AppL PhysioL 8: 102-110, 1955. 146. FOX, R. B., J. R. HOIDAL, D. M. BROWN, AND J. E. REPINE. Pulmonary inflammation due to oxygen toxicity: involvement of chemotactic factors and polymorphonuclear leukocytes. Am. Rev. Respir. L?is. 123: 521523, 1981. 147. FRIDOVICH, I. The biology of oxygen radicals. science 201: 875889,1978. 148. FRIEDMAN, M., AND G. C. WHITE. The effects of heparin pretreatment on the development of pulmonary edema following microembolization (abstr.). Physiologist 22(4): 41, 1979. 149. FRY, D. L. Acute vascular endothelial changes associated with increased blood velocity gradients. Circ. Rsa 22: 165197, 1968. 150. FRY, D. L. Certain histological and chemical responses of the vascular interface to acutely induced mechanical stress in the aorta of the dog. Circ Res 24: 93-108,1969. 151. GAJEWSKI, J., AND M. POVAR. Blood coagulation values of sheep. Am J. Vet. Rea 32: 405-409,197l. 15la.GARCIA-SZABO, R. R., F. L. MINNEAR, R. BIZIOS, A. JOHNSON, AND A. B. MALIK. Role of thromboxane in the pulmonary vascular response to microembolization. Chest. 83, Suppl.: 76S-78S, 1983. 15lb.GARCIA-SZABO, R. R., R. BIZIOS, M. B. PETERSON, W. D. WATKINS, AND A. B. MALIK. Thromboxane generation after thrombin microembolism: protective effect of thromboxane synthetase inhibition on lung fluid balance. Circ Res. in press. 152. GEE, M. H., AND D. 0. WILLIAMS. Effect of lung inflation on perivascular cuff fluid volume in isolated dog lung lobes. ikficrouasc Res. 17: 192-201, 1979. 153. GIAMMONA, S. T., I. MANDELBAUM, J. FOY, AND S. BONDURANT. Effects of pulmonary artery ligation on pulmonary surfactant and pressure-volume characteristics of dog lung. Circ Res. 18: 683-691, 1966. 154. GIBBON, J. H., JR., AND M. H. GIBBON. Experimental pulmonary edema following lobectomy and plasma infusion. Surgery 12: 694-704, 1942. 155. GIBBON, J. H., JR., M. HOPKINSON, AND E. D. CHUR-

July

1983

PULMONARY

MICROEMBOLISM

CHILL. Changes in the circulation produced by gradual occlusion of the pulmonary artery. J. Clin. Invest. 11: 543553, 1932. 156. GIMBRONE, M. A., R. H. ASTER, R. S. COTRAN, J. BROKERY, J. H. JANDL, AND J. FOLKMAN. Preservation of vascular integrity in organs perfused in vitro with platelet-rich medium. Nature London 222: 33-36, 1969. 157. GLAGOV, S. Mechanical stress on vessels and non-uniform distribution of atherosclerosis. Med Clin. North Am 57: 63-77, 1977. 158. GLENN, W. W. L., D. K. PETERSON, AND C. K. DRINKER. The flow of lymph from burned tissue, with particular reference to the effects of fibrin drainage and composition. Surgery 12: 685-693, 1942. 159. GOLD, W. M. Neurohumoral interactions in airways. Am Rev. Respir. Dk. 115: 127-137, 1977. 160. GOLD, W. M., G.-F. KESSLER, AND D. Y. C. YU. Role of vagus nerves in experimental asthma in allergic dogs. J. AppL Physid 33: 719-725, 1972. 161. GOLDSMITH, G. H., JR., H. SAITO, APJD 0. D. RATNOFF. The activation of plasminogen by Hageman factor (factor XII) and Hageman factor fragments. J. Clin. Invest. 62: 54-60, 1978. 162. GOLDSTEIN, I. M., M. BRAI, A. G. OSTER, AND G. WEISSMANN. Lysosomal enzyme release from human leukocytes; mediation by the alternate pathway of complement activation. J. Immunol 111: 33-37, 1973. 162a.GRANGER, D. N., G. RUTILI, AND J. M. McCORD. Superoxide radicals in feline intestinal &hernia. G&m enterdogy 81: 22-29, 1981. 163. GRANT, J. A., E. DUPREE, AND A. S. GOLDMAN. Complement-mediated release of histamine from human leukocytes. J. ZmmunoL 114: 1101-1106, 1975. 164. GREEN, M., AND J. G. WIDDICOMBE. The effect of ventilation of dogs with different gas mixtures on airway calibre and lung mechanics. J. PhytioL London 18: 363381, 1966. 165. GROTTE, G. Passage of dextran molecules across the blood lymph barrier. Acta Chir. Stand SuppL 211: l-84, 1956. 166. GUREWICH, V., T. NUNN, AND B. LIPINSKI. Activation of intrinsic or extrinsic blood coagulation in experimental venous thrombosis and disseminated intravascular coagulation: pathogenetic differences. Thrd. Res. 14: 931-940, 1979. 167. GUYTON, A. C., AND A. W. LINDSEY. Effect of elevated left atria1 pressure and decreased plasma protein concentration on the development of pulmonary edema. Circ. Res. 12: 649-657, 1959. 168. GUYTON, A. C., A. E. TAYLOR, R. E. DRAKE, AND J. C. PARKER. Dynamics of subatmospheric pressure in the pulmonary interstitial fluid. In: Lung Liquids, edited by R. Porter and M. O’Connor. Amsterdam: Elsevier, 1976, p. 77-96. (Ciba Found. Symp. 38.) 169. GUZ, A., AND D. W. TRENCHARD. The role of nonmyelinated afferent vagal fibres from the lungs in the genesis of tachypnoea. J Physid London 213: 345-371, 1971. 170. HADDY, F. J., AND J. B. SCOTT. Metabolically linked vasoactive chemicals in local regulation of blood flow. PhysioL Rev. 48: 688-707, 1968. 171. HAGEMAN, W. E., D. G. WENTLING, AND T. P. PRUSS. Interaction of vasoactive agents and microemboli on the pulmonary circulation. Eur. J. PhamnucoL 22: 295-303, 1973. 172. HAGGART, G. E., AND A. M. WALKER. The physiology of pulmonary embolism as disclosed by quantitative oc-

173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

1197

clusion of the pulmonary artery. Arch Surg. 6: 764-773, 1923. HAHN, H. L., P. D. GRAF, AND J. A. NADEL. Effect of vagal tone on airway diameters and on lung volume in anesthetized dogs. J. AppL Physid 41: 581-589, 1976. HAKIM, T. S., AND C. A. DAWSON. Sympathetic nerve stimulation and vascular resistance in a pump perfused dog lung lobe. Proc Sot Exp. Bid 160: 38-41, 1979. HAKIM, T. S., C. A. DAWSON, AND J. H. LINEHAN. Hemodynamic responses of dog lung lobe to lobar venous occlusion. J. AppL PhysioL: Respirat. Environ. Exercise PhysioL 47: 145-152, 1979. HALES, C. A., L. SONNE, M. PETERSON, D. KONG, M. MILLER, AND W. D. WATKINS. Role of thromboxane and prostacyclin in pulmonary vasomotor changes after endotoxin challenge in dogs. J. Clin. Invest. 68: 497-505, 1981. HALMAGYI, D. F. Role of lymphatics in the genesis of “shock lung.” In: Lung Water and Solute Exchange, edited by N. C. Staub. New York: Dekker, 1978, p. 423-435. HALMAGYI, D. F., B. STARZECKI, AND G. J. HORNER. Humoral transmission of cardiorespiratory changes in experimental lung embolism. Circ. Res. 14: 546-554,1964. HAMMERSCHMIDT, D. E., P. D. HARRIS, J. H. WAYLAND, P. R. CRADDOCK, AND H. S. JACOB. Complement-induced granulocyte aggregation in vivo. Am J PuthoL 102: 146-150, 1981. HAMMERSCHMIDT, D. E., J. G. WHITE, P. R. CRADDOCK, AND H. S. JACOB. Corticosteroids inhibit complement-induced granulocyte aggregation. A possible mechanism for their efficacy in shock states. J. Clin. Invat. 63: 7984303, 1979. HAMMERSON, F. Endothelial filaments and intercellular gaps: a sufficient evidence for contractility? Bibl. And 12: 159-164, 1973. HANNA, C. J., M. K. BACH, P. D. PARE, AND R. R. SCHELLENBERG. Slow-reacting substances (leukotrienes) contract human airway and pulmonary vascular smooth muscle in vitro. Nature London 290: 343-344, 1981. HAWORTH, S. G., M. DE LEVAL, AND F. J. MACARTNEY. How the left lung is perfused after ligating the left pulmonary artery in the pig at birth: clinical implications for the hypoperfused lung. Cardiovasc Res. 15: 214-226, 1981. HEBB, C. Motor innervation of the pulmonary blood vessels of mammals. In: The Pulmonary Circulation and Interstitial Space, edited by A. P. Fishman and H. H. Hecht. Chicago, IL: Univ. of Chicago Press, 1969, p. 195-222. HECHTMAN, H. B., E. A. LONERGAN, H. P. B. STAUNTON, AND R. C. DENNIS. Pulmonary entrapment of platelets during acute respiratory failure. Surgery 83: 277-283, 1978. HEFLIN, A. C., AND K. L. BRIGHAM. Prevention by granulocyte depletion of increased lung vascular permeability of sheep lung following indotoxemia. J. Clin. Invest. 68: 1253-1260, 1981. HENRIKSSON, P., AND I. M. NILSSON. Effects of leukocytes, plasmin and thrombin on clotting factors. A comparative in vitro study. Thromb. Res. 16: 301-312, 1979. HENSON, P. M., K. MCCARTHY, G. L. LARSEN, R. 0. WEBSTER, P. C. GICLAS, R. B. DREISIN, T. E. KING, AND J. 0. SHAW. Complement fragments, alveolar macrophages, and alveolitis. Am J. Pathd 97: 93-110, 1979. HIEBERT, L. M., AND L. B. JACQUES. The absorption of heparin on endothelium after injection. Thromb. Res. 8: 195-204, 1976.

1198

ASRAR

190. HIEBERT, L. M., AND L. B. JACQUES. Heparin uptake on endothelium. Artm 2: 26-37, 1976. 191. HIROSE, T., T. YASUTAKE, A. TARABEIH, AND M. STEIN. Location of airway constriction following acute experimental pulmonary thromboembolism. J. Appl Phyaid 34: 431-437, 1973. 192. HIRSH, J. Hypercoagulability. Semis Hem&d 14: 409425,1977. 193. HOAK, J. C., R. L. CZERVIONKE, G. L. FRY, AND J. B. SMITH. Interaction of thrombin and platelets with the vascular endothelium. Federation Proc 39: 2606-2609, 1980. 194. HOGG, J. C. Effect of pulmonary edema on distribution of blood flow in the lung. In: Lurrg Waterand Saute Exchange, edited by N. C. Staub. New York: Dekker, 1978, p. 167-182. 195. HOHN, D. C., A. J. MEYERS, S. T. GHERINI, A. BECKMAN, R. E. MARKISON, AND A. M. CHURG. Production of acute pulmonary injury by leukocytes and activated complement. Surgeq 88: 48-58,198O. 196; HOLMAN, E. Problems in dynamics of blood flow: conditions controlling collateral circulation in the presence of an arterio-venous fistulas and following ligation of an artery. Suqj’ery 26: 889-917, 1949. 197. HOPEWELL, P. D., AND J. F. MURRAY. The adult respiratory distress syndrome. Annu. Rev. Med 27: 343356, 1976. 198. HORISBERGER, B., AND S. RODBARD. Direct measurement of bronchial arterial flow. Circ Res. 8: 11491156, 1960. 199. HOSEA, S., E. BROWN, C. HAMMER, AND M. FRANK. Role of complement activation in a model of adult respiratory distress syndrome. J. CZin. Invest. 66: 375-382, 1980. 200. HUBER, G. L., AND L. H. EDMUNDS, JR. Pulmonary artery occlusion. II. Morphologic studies. J. Appl Ph&d 22: 1002-1011,1967. 261. HURLEY, J. V. Current views on the mechanisms of pulmonary oedema. J. Puthd 125: 59-79, 1978. 202. HYLAND, J. W., T. E. PEIMME, S. ALEXANDER, F. W. HAYNES, G. T. SMITH, AND L. DEXTER. Behavior of pulmonary hypertension produced by serotonin and emboli. Am J. Phpid 205: 591-597, 1963. 293. HYLAND, J. W., G. T. SMITH, L. B. McGUIRE, D. C. HARRISON, F. W. HAYNES, AND L. DEXTER. Effect of selective embolization of various sized pulmonary arteries in dogs. Am. J. PhysioL 204: 619-6251963. 294. HYMAN, A. L. Pulmonary vasoconstriction due to nonocclusive distension of large pulmonary arteries in the dog. Circ. Res. 23: 401-413, 1968. 265. HYMAN, A. L., W. D. MYERS, AND A. MEYER. The effect of acute pulmonary embolus upon cardiopulmonary hemodynamics. Am Heart J. 67: 313-323, 1964. 206. HYMAN, A. L., P. NANDIWANDA, D. S. KNIGHT, AND P. KADOWITZ. Pulmonary vasodilator responses to catecholamines and sympathetic nerve stimulation. Circ Ra 48: 407-415,198l. 207. HYMAN, A. L., E. W. SPANNHAKE, AND P. J. KADOWITZ. Prostaglandins and the lung. Am Rev. Reqir. Dis. 117: ill-273,1978. 268. ILIFF, L. D. Extra-alveolar vessels and edema development in excised dog lungs. Circ Rea 28: 524-532,197l. 209. INGRAM, R. H., JR. Effects of airway versus arterial CO, changes on lung mechanics in dogs. J. Appl Physid 38: 603~607,1975. 210. IRVIN, C. G., AND J. A. DEMPSEY. Role of Hi and Ht receptors in increased small airway resistance in the dog. Req.+. Ph@ol 35: 161-176.1978.

B. MALIK

Vohne

6.9

211. JACOB, H. S., P. R. CRADDOCK, D. E. HAMMERSCHMIDT, AND C. F. MOLDOW. Complement-induced granulocyte aggregation. N. EngL’ J. Med 302: 789-794, 1980. 212. JACOB, J. S. Complement-induced vascular leukostasis. Arch, PathoL Lab Med 104: 617-626,198Q. 213. JAFFE, E. A. Endothelial cells and the biology of factor VIII. N. En& J Med 296: 377-383, 1977. 214. JAIN, S. K., S. SUBRANIAN, D. B. JULKA, AND A. GUZ. Search for evidence of lung chemoreflexes in man: study of respiratory and circulatory effects of phenyldiguanide and lobeline. Clin. S& 42: 167-177, 1972. 215. JANOFF, A., R. WHITE, H. CARP, S. HAREL, R. DEARING, AND D. LEE. Lung injury induced by leukocytic proteases. Am J. Pathd 97: ill-136,1979. 216. JAQUES, L. B. Heparin: an old drug with a new para; digm. science 206: 528-533,1979. 217. JOHNSON, A., AND A. B. MALIK. Effect of defibrinogenation on lung water accumulation after pulmonary microembolism in dogs. J. Appl Phytiol: Respirat. En.viron, Exercise PhysioL 49: 841-845, 1980. 218. JOHNSON, A., AND A. B. MALIK. Effect of granulocytopenia on extravascular lung water content after microembolization. Am. Rev. Respir. Dis. 122: 561-566,198O. 219. JOHNSON, A., AND A. B. MALIK. Effects of differentsize microemboli on lung fluid and protein exchange. J. Appl Physiol: Respimt. Environ Exercke Physd 51: 461-464, 1981. 220. JOHNSON, A., AND A. B. MALIK. Pulmonary edema after glass bead microembolization: protective effect of granulocytopenia. J. AppL Phytiol: Reqirat Envimn. Exercise Physid 52: 155-161, 1982. 22Oa.JOHNSON, A., AND A. B. MALIK. Effect of defibrinogenation on lung fluid and protein exchange after glass bead embolization. J. Appl Physid: Respirat. Environ. Ewrcbe Physid 53: 895~900,1982. 221. JOHNSON, A., M. V. TAHAMONT, J. E. KAPLAN, AND A. B. MALIK. Lung fluid balance after pulmonary embolization: effects of thrombin vs. fibrin aggregates. J. Appl Physid: Respimt Envim Exemise Physid 52: 1565-1570, 1982. 22laJOHIUSON, A., M. V. TAHAMONT, AND A. B. MALIK. Thrombin-induced lung vascular injury: role of fibrinogen and fibrinolysis. Am Reu. Respir. Dia in press. 222. JOHNSON, J. W. C., S. PERMUTT, J. H. SIPPLE, AND E. S. SALEM. Effect of intra-alveolar fluid on pulmonary surface tension properties. J. AppL Physid 19: 769-777, 1964. 223. JOHNSON, K. J., J. C. FANTONE III, J. KAPLAN, AND P. A. WARD. In vivo damage of rat lungs by oxygen metabolites. J. Clin. Invest. 67: 983-993, 1981. 224. JOHNSON, K. J., AND P. A. WARD. Role of oxygen metabolites in immune complex injury of lung. J. Immund 126: 2365-2369, 1981. 225. JOHNSON, R. L., JR., S. S. CASSIDY, M. HAYNES, R. L. REYNOLDS, AND W. SCHULZ. Microvascular injury distal to unilateral pulmonary artery occlusion. J. AppL Physid: Respirat. Environ Exercbe Phgsitd 51: 845851.1981. 226. JOHNSON, R. L., JR., H. F. TAYLOR, AND A. (3. DEGRAFF, JR. Functional significance of a low pulmonary diffusing capacity of CO. J. CZia Inve& 44: 759~800,1965. 227. JONSSON, C. E., 0. LINDQUIST, AND P. MALMBERG. Efflux of prostaglandins in right lymphatic duct lymph following induced intravascular coagulation and inhibited fibrinolysis in the dog. Scar& J. Clin m Invest 33: 361370, 1974. 228. JULIAN, D. G., D. M. TRAVIS, E. D. ROBIN, AND C. H.

July

1983

PULMONARY

MICROEMBOLISM

CRIMP. Effect of pulmonary artery occlusion upon endtidal COs tension. J. Appl Physid 15: 87-91, 1969. 229. JUNOD, A. Metabolism in the intact pulmonary microcirculation. Ann NY Am& Sci 334: 66-71, 1982. 229aJUNOD, A. F. Metabolism of vasoactive agents in lung. Am Reu. Bespir. Dia 11s (Suppl.): 51-57,1977. 296. JURATSCH, C. E., f. A. LENGO, AND M. M. LAKS. Role of the autonomic nervous system and pulmonary artery receptors in production of experimental pulmonary hypertension. Cht 71, Suppl.: 265-268, 1977. 231. KABINS, S. A., J. FRIDMAN, M. KANDELMAN, AND H. WEISBERG. Effect of sympathectomy on pulmonary embolism-induced lung edema. Am J. Physid 202: 687689, 1962. 292. KABINS, S. A., J. FRIDMAN, J. NEUSTADT, G. ESPINOSA, AND L. N. KATZ. Mechanisms leading to lung edema in pulmonary embolization. Am J. Physid 198: 543~546,X%6.

233. KADISH, J. L., C. E. BUTTERFIELD, AND J. FOLKMAN. The effect of fibrin on cultured vascular endothelial cells. Tissue CkU 11: 99-108, 1979. 234. KADOWITZ, P. J., C. A. GRUE’M’ER, E. W. SPANNHAKE, AND A. L. HYMAN. Pulmonary vascular responses to prostaglandins. Fe&r&ion Proc 40: 1991-1996, 1981. 235. KADOWITZ, P. J., AND A. L. HYMAN. Effect of sympathetic nerve stimulation on pulmonary vascular resistance in the dog. Circ Bea 32: 221~227,1973. 296. KAPLAN, A. P. The Hageman factor dependent pathways of human plasma. Microuc~ Res. 8: 97-111, 1974. 237. KAPLAN, A. P., AND K. F. AUSTEN. Activation and control mechanisms of Hageman factor-dependent pathways of coagulation, flbrinolysis, and kinin generation and their contribution to the inflammatory response. J. Allergy C&L Immund 66: 491-596, 1975. 238. KAPLAN, J. E., AND T. M. SABA. Low-grade intravascular coagulation and reticuloendothelial function. Am J. Physid 234 (He& Ck Ph&d 3): H323-H329,1978. 239. KARCZEWSKI, W., AND J. G. WIDDICOMBE. The role of the vague nerves in the respiratory and circulatory responses to intravenous histamine and phenyl diguanide in rabbits. J. Physid Lvndm 201: 271-291, 1969. 249. KARNOVSKY, M. J. Morphology of capillaries with special reference to muscle capillaries. In: CapiUurg Per-i@, edited by C. Crone and N. A. Lassen. New York: Academic, 1970, p. 341-350. 241. KATZ, S., AND A. D. HORRES. Medullary respiratory neuron response to pulmonary emboli and pneumothorax. J. Appl Phw 33: 390-396.1972. 242. KAUFMAN, M. P., H. M. COLERIDGE, J. C. G. COLERIDGE, AND D. G. BAKER. Bradykinin stimulates afferent vagal C-fibers in intrapulmonary airways of dogs. J. Appl. Physid: Reqnbk Envirtm Extm.Ge Physid 48: 511~517,1989. 243. KAY, A. B., D. S. PEPPER, AND M. R. EWART. Generation of chemotactic activity for leukocytes by the action of thrombin on human fibrinogen. Nc&ure ti 243: !%-67,1973. 244. KEALEY, G. P., AND M. J. BRODY. Studies on the mechanism of pulmonary vascular response to miliary pulmonary embolism. Circ Rea 41: 867-814,1977. 245. KENTERA, D., C. R. WALLACE, W. F. HAMILTON, AND L. T. ELLISON. Venous admixture in dogs with chronic pulmonary hypertension. J. Appl. Physid 20: 919-921, 1965. 246. KIMBEL, P., AND M. NAGARRO. Pulmonary capillary blood flow and experimental pulmonary embolism in the dog. In: Pulmmarg Th PYAxxwW~.~~, edited by K. M.

1199

Moser and M. Stein. Chicago, IL: Year Book, 1973, p. 133143. 247. KITAGAWA, S., F. TAKADU, AND S. SAKAMOTO. Evidence that proteases are involved in superoxide production by human polymorphonuclear leukocytes and monocytes. J Clin, Invex& 65: 74-81, 1980. 247a.KIVLEN, C. M., A. JOHNSON, T. PITTMAN, A. GUILE, AND A. B. MALIK. Effect of converting enzyme inhibition on pulmonary edema after microembolization. J. AppL Physid: Respirat. Envirm Exercise Physid 53: 15461550, 1982. 248. KNISELY, W. H. Mcroswpic &wmatiom tithe Architecture and Behuvim of Smuller Blood Vessels in Living Lungs of Mammals and Related &wrvatians in Other Vemels of Eaymbnental Animals and Human Patiats (PhD thesis). Charleston: Med. Coll. of South Carolina, 1954. 249 KNISELY, W. H. Small blood vessels. Ann. Intern, Med 69: 909-917, 1968. 250 KNISELY, W. H., AND M. H. KNISELY. Preliminary observations of the catch-trap architecture of pulmonary artery tips in health and their responses following distant somatic burns (abstr.). Anat. Ret 118: 326, 1954. 251 KNISELY, W. H., J. M. WALLACE, AND W. A. ADDISON. Temporary pulmonary embolization caused by intravenous injections of 5-hydroxytryptamine (abstr.). Fe&ration Proc 17: 88,1958. 252 KNOWLTON, G. C., AND M. G. LARRABEE. A unitary analysis of pulmonary volume receptors. Am. J. Physid 147: 100-114, 1946. 253 KOMORIYA, K., H. OHMORI, A. AZUMA, S. KUORZUMI, AND Y. HASHIMOTO. Prostaglandin Is as a potentiator of acute inflammation in rats. Pr&u&mdi7u 15: 557-564, 1978. 254 KRAHL, V. E. The lungs as a target organ in thromboembolism. In: Pulmmary En&a& Disease, edited by A. A. Sasahara and M. Stein. New York: Grune Bt Stratton, 1965, p. 13-22. 255 LAKI, K. Fib-iNew York: Dekker, 1968. 256 LANDOLT, C. C., M. MATTHAY, K. A. ALBERTINE, P. J. ROOS, J. P. WIENER-KRONISH, AND N. C. STAUB. Overperfusion, hypoxia, and increased pressure cause only hydrostatic edema in anesthetized sheep. Circ Rea. 52: 335-341, 1983. 257 LANDOLT, C., M. MATTHAY, M. R. FLICK, AND N. C. STAUB. Failure of antihistamine to prevent increased lung vascular permeability after microemboli in sheep (abstr.). Microuusc Reu 17: 90, 1979. 258 LATEGOLA, M. T. Pressure-flow relationships in the dog lung during acute, subtotal pulmonary vascular occlusion. Am J Physid 192: 613-619, 1958. 259. LATEGOLA, M. T., AND H. RAHN. &dies in Respiratory Physidogy (2nd Ser.). Dayton, OH: Wright Air Dev. Center, 1955, p. 180. 260. LATOUR, J. B., AND F. BERNARD. Prevention by aspirin of the classic generalized Schwartzman reaction. Am ,I, Pathd 91: 595~605,1978. 261. LEAF, A. Cell swelling: a factor in ischemic tissue injury. Circuliztiun 43: 455-458, 1973. 262. LEDSOME, J. R. The reflex role of pulmonary arterial barorcceptors. Am Rev. Respir. Dia 115, Suppl.: 245-256, 1977. 263. LEE, B. C., H. VAN DER ZEE, AND A. B. MALIK. Site of pulmonary edema after unilateral microembolization. J. AppC Physid: Reqirat Environ E&e Ph@d 47: !556-560.1979. 264. LEUENBERGER, P. J., S. A. STALCUP, R. B. MELLINS, L. M. GREENBAUM, AND G. M. TURINO. Decrease in

1200

ASRAR

I conversion in dogs (40252). Proc Sot Exp. 158: 586-589,19’78. LEVY, S. E., B. J. SHAPIRO, AND D. H. SIMMONS. Pulmonary hemodynamics after autologous in vivo pulmonary thromboembolism. J. AppL PhysioL 27: 53-60,1969. LEVY, S. E., AND D. H. SIMMONS. Redistribution of alveolar ventilation following pulmonary thromboembolism in the dog. J. App! Physid 36: 60-68, 19’74. LEVY, S. E., AND D. H. SIMMONS. Mechanism of arterial hypoxemia following pulmonary thromboembolism in dogs. J. AppL Physid 39: 41-46, 1975. LEVY, S. E., M. STEIN, R. S. TOTTEN, S. WESSLER, I. BRUNDERMAN, AND E. D. ROBIN. Ventilation-perfusion abnormalities in experimental pulmonary embolism. J. Clin. Invest, 44: 1699-1’707, 1965. LEWIS, R. A., AND K. F. AUSTEN. Non-respiratory function of pulmonary cells: the mast cell. Federation Proc. 36: 2676-2683, 1977. LEWIS, R. A., AND K. F. AUSTEN. Mediation of local homeostasis and inflammation by leukotrienes and other mast cell-dependent compounds. Nature London 293: 103-108, 1981. LEWISTON, N. J., J. THEODORE, AND E. D. ROBIN. Intracellular edema and dehydration: effects on energy metabolism of alveolar macrophages. Science 191: 403464, 1976. LICHTENSTEIN. L. M. Mediators and the mechanism of their release. Chest 73, Suppl.: 919-925, 1978. LIEBOW, A. A,, M. R. HALES, W. E. BLOOMER, W. HARRISON, AND G. E. LINDSKOG. Studies on the lung after ligation of the pulmonary artery. II. Anatomical changes. Am. J. PuthoL 26: 177-195, 1950. LIEBOW, A. A., M. R. HALES, W. HARRISON, W. BLOOMER, AND G. E. LINDSKOG. Genesis and functional implications of collateral circulation of the lungs.

B. MALIK

angiotensin

BioL Med

265.

266.

267.

268.

269.

270.

271.

272. 273.

274.

Yale J. BioL Med

284.

285.

286. 287.

288.

289.

290.

291.

293.

22: 637~650,195O.

275. LIEBOW, A. A., M. R. HALES, AND G. E. LINDSKOG. Enlargement of bronchial arteries and their anastomoses with pulmonary arteries in bronchiectasis. Am J. PuthoL 25: 211-232, 1949. 276. LILKER, E. S., AND E. J. NAGY. Gas exchange in the pulmonary collateral circulation of dogs. Effects of alveolar hypoxia and systemic hypoxemia. Am Rev. Reqir. L?is. 112: 615-620, 1975. 277. LINDQUIST, O., L. BAGGE, AND T. SALDEEN. Induction of endogenous fibrinolysis inhibition in the dog. Actu Chir. Scud 142: 20-25, 1976. 278. LINDQUIST, O., B. GERDIN, T. SALDEEN, AND E. SVENSJO. Demonstration of a permeability increasing factor in interstitial fluid from lungs following thrombininduced pulmonary embolism in the dog and rat (abstr.). Mtim.mmc Res. 10: 238,1975. 279. LINDQUIST, O., L. RAMMER, AND T. SALDEEN. Pulmonary insufficiency microembolism and inhibition of the fibrinolytic system in post-traumatic autopsy material. Acta Chir. Sound 138: 545-549, 1972. 280. LINDQUIST, O., AND T. SALDEEN. On the importance of vasoactive mediators for the oedema formation in thrombin-induced pulmonary microembolism in the dog (abstr.). Microvaac Rea 256, 1974. 281. LINDQUIST, O., T. SALDEEN, E. SVENSJO, AND R. WALLIN. On the cause of the increased vascular permeability in the delayed microembolism syndrome (al&r.). Micrtnmc Res. 11: 427, 1976. 282. LINDSEY, H. E., AND J. H. WYLLIE. Release of prostaglandins from embolized lungs. Br. J. Surg. 57: 738-741, 1970. 283. LUNG, M. A. K. Y., J. C. C. WANG, AND C. K. CHENG.

294.

295.

298

299

300

301.

63

Bronchial circulation: an autoperfusion method for assessing its vasomotor activity and the study of a- and @-adrenoceptors in the bronchial artery. Life&i 19: 577589, 1976. LUTERMAN, A., D. MANWARING, AND P. W. CURRERI. The role of fibrinogen degradation products in the pathogenesis of the respiratory distress syndrome. Surgerg 82: 703-709, 1977. MACKLEM, P. T., AND J. MEAD. Resistance of central and peripheral airways measured by a retrograde catheter. J. AppL PhysioL 22: 395-401, 1967. MACKNIGHT, A. D. C., AND A. LEAF. Regulation of cellular volume. Physid Rev. 57: 510-573, 1977. MAGNO, M., AND A. P. FISHMAN. Origin, distribution, and blood flow of bronchial circulation in anesthetized sheep. J. AppL Physid: Respirat. Environ Exercise PhysioL 53: 272-279, 1982. MAGNO, M., AND J. P. SZIDON. Hemodynamic pulmonary edema in dogs with acute and chronic lymphatic ligation. Am J. PhysioL 231: 1777-1782, 1976. MAJNO, G. Ultrastructure of the vascular membrane. Handbook of Physiology. Circulation, edited by W. F. Hamilton and P. Dow. Washington, DC: Am. Physiol. Sot., 1965, sect. 2, vol. III, chapt. 64, p. 2293-2375. MAJNO, G., AND I. JORIS. Endothelium 1977: a review. In: Thrombotic P-8 in Atherogenmis, edited by A. B. Chandler, K. Eurenius, G. C. McMillan, C. B. Nelson, C. J. Schwartz, and S. Wessler. New York: Plenum, 1978, p. 169-526. MAJNO, G., AND G. E. PALADE. Studies of inflammation. I. The effect of histamine and serotonin on vascular permeability. An electron microscopic study. J. Biophys. Biochem. CytoL 11: 571605; 1961. MAJNO, G., S. M. SHEA, AND M. M. LEVANTHAL. Endothelial contraction induced by histamine-type mediators: an electron microscopic study. J. Cell Bid 42: 647-72, 1969. MALIK, A. B., A. JOHNSON, AND M. V. TAHAMONT. Mechanisms of lung vascular injury after intravascular coagulation. Ann NY Acad S& 384: 169-190, 1982. MALIK, A. B., B. C. LEE, H. VAN DER ZEE, AND A. JOHNSON. The role of fibrin in the genesis of pulmonary edema after embolization in dogs. Circ Res. 45: 120-125, 1979. MALIK, A. B., AND J. C. NEWELL. Pulmonary perfusion and gas exchange in hemorrhage and shock. J. AppL Physid:

297

Volume

Reapirat.

Environ.

Exercbe

Physid

42: 279-286,

1977. MALIK, A. B., AND N. C. STAUB. Mechanisms of lung microvascular injury. Ann. NYAcad Sci 384: l-562,1982. MALIK, A. B., M. V. TAHAMONT, AND A. JOHNSON. Increased lung vascular permeability after pulmonary intravascular coagulation (PIC): protective effect of defibrinogenation (abstr.). Circulation 64: 52, 1981. MALIK, A. B., AND S. E. TRACY. Bronchovascular adjustments after pulmonary embolism. J. AppL PhysioL: Respirat. Environ Exercbe PhysioL 49: 476-481, 1980. MALIK, A. B., AND H. VAN DER ZEE. Thrombin-induced pulmonary insufficiency. Thromb Rea 11: 497-566, 1977. MALIK, A. B., AND H. VAN DER ZEE. Time course of pulmonary vascular response to microembolization. J. AppL

PhpsioL:

Reapirat.

Environ

Exercise

PhpioL

43:

51-58, 1977. 362. MALIK, A. B., AND H. VAN DER ZEE. Mechanisms of pulmonary edema induced by microembolization in dogs. Circ Re8. 42: 72-79, 1978. 363. MALIK, A. B., AND H. VAN DER ZEE. Lung vascular

July

394.

395.

366.

307.

308.

369.

310.

311.

312. 313.

314.

315.

316.

317.

1983

PULMONARY

MICROEMBOLISM

1201

permeability following progressive pulmonary embolization. J. Appl Physid: Respirat. Envirm Exercise Phgsid 45: 5904597, 1980. MALIK, A. B., H. VAN DER ZEE, AND B. C. LEE. Pulmonary transvascular fluid dynamics in sheep during hemorrhage. Lymphobgy 12: 149-158,1979. MALIK, A. B., H. VAN DER ZEE, AND P. H. NEUMANN. Effects of pulmonary edema on regional pulmonary perfusion in the intact dog lung. J App! Phyaid: Respiru~ Environ. Exercise Physid 49: 834-840, 1980. MANWARING, D., D. THORNING, AND P. W. CURRERI. Mechanisms of acute pulmonary dysfunction induced by fibrinogen degradation product D. Surgery 84: 45-54, 1978. MANWARING, D., D. THORNING, AND P. W. CURRERI. Mechanism of acute pulmonary dysfunction. Ann. Surg. 192: 103-107, 1980. MARSHALL, B. E., L. R. SOMA, AND G. R. NEUFELD. Lung water volume at rest and exercise in dogs. J. Appl. Physid 39: 7-8, 1975. MARSHALL, R., D. C. SABISTON, P. R. ALLISON, A. R. BOSMAN, AND M. S. DUNHILL. Immediate and late effects of pulmonary embolism ,by large thrombi in dogs. Thorax 18: l-9,1963. MARTINEZ, J., L. DE LETONA, R. C. DE LA MATA, AND D. M. AVIADO. Local and reflex effects of bronchial arterial injection of drugs. J. Pharmucd Exp. Ther. 133: 295-303, 1961. MAYER, J. E., AND E. H. HUMPHREY. Effects of pulmonary emboli and pulmonary artery occlusion on pulmonary capillary permeability (abstr.). Federation Proc 37: 2622, 1978. MAYER, M. M. The complement system. Sci Am. 229: 54-69, 1973. McCORD, J. M.,. AND I. FRIDOVICH. The biology and pathology of oxygen radicals. Ann Intern Med 89: 122127, 1978. MCCORMICK, J. R., M. M. HARKIN, K. J. JOHNSON, AND P. A. WARD. Suppression by superoxide dismutase of immune-complex-induced pulmonary alveolitis and dermal inflammation. Am J. Pa.thd 102: 55-61,198l. McEVOY, R. K., R. A. HARDER, AND W. A. DALE. Respiratory and cardiovascular phenomena associated with pulmonary embolism. Sum. Gyteod obstet. 106: 271-287, 1958. MCLAUGHLIN, R. F., JR., W. S. TYLER, AND R. 0. CANADA. Subgross pulmonary anatomy in various mammals and man. J. Am. Med. A8soc 175: 694-697, 1961. MCLAUGHLIN, R. F., JR., W. S. TYLER, AND R. 0. CANADA. Subgross pulmonary anatomy of the rabbit, rat, and guinea pig, with additional notes on the human lung.

323. MEYER, E. C., AND R. OTTAVIANO. The effect of fibrin on the morphometric distribution of pulmonary exudative edema. L&X Invest 29: 320-328, 1973. 324. MEYRICK, B., AND L. REID. Nerves in rat intra-acinar alveoli: an electron microscopic study. ReqGr. Phykd 11: 367-377, 1971. 325. MICHAEL, J., H. SAKIO, W. MCFARLAND, AND S. I. SAID. Bradykinin increases pulmonary lymph flow in anesthetized dogs by increasing microvascular pressure (abstr.). Physiulugbt 22(4): 87, 1978. 326. MILIC-EMILI, J., J. A. M. HENDERSON, M. B. DOLOVICH, D. TROP, AND K. KANEKO. Regional distribution of inspired gas in the lung. J. Appl, Physid 21: 749-759,1966. 327. MILLER, W. C., W. W. SIMI, AND D. L. RICE. Contribution of systemic venous hypertension to the development of pulmonry edema in dogs. Circ Rea 43: 598-600, 1978. 328. MILLER, W. S. The Lung (2nd ed.). Springfield, IL: Thomas, 1948. 329. MILLS, J. E., H. SELLICK, AND J. G. WIDDICOMBE. Activity of lung irritant receptors in pulmonary microembolism, anaphylaxis and drug-induced bronchoconstrictions. J. Physid Lvndun 203: 337-357, 1969. 330. MILLS, J. E., AND J. G. WIDDICOMBE. Role of the vagus nerve in anaphylaxis and histamine-induced bronchoconstriction in guinea pigs. Br. J. Phurmucd 39: 724-731, 1970. 331. MINNEAR, F. L., P. S. BARIE, AND A. B. MALIK. Lung fluid and protein exchange in the acute sheep preparation. J. Appl Physid: Reap+& Envimn ExercGe Phvsid 50: 1358-1361, 1981. 33la.MINNEAR, F. L., C. M. KIVLEN, AND A. B. MALIK. Effect of bradykinin on lung vascular permeability in sheep. J. Appl Physid: lkspiti Envim Exe&w Phy8id in press. 332. MINNEAR, F. L., AND A. B. MALIK. Bradykinin does not increase lung vascular permeability (abstr.). Fe&ration Pruc 41: 7189, 1982. 333. MINNEAR, F. L., D. G. MOON, J. E. KAPLAN, AND A. B. MALIK. Effect of ADP-induced platelet aggregation on lung fluid balance in sheep. Am J. Physid 242 (Heart Circ Physid 11): H645-H651, 1982. 333a,MINNEAR, F. L., R. J. MULLINS, D. R. BELL, AND A. B. MALIK. Histamine-induced increase in pulmonary lymph flow: hemodynamics versus permeability (abstr.). Physidogist 25: 8.13, 1982. 334. MISEROCCHI, G., J. MORTOLA, AND G. SANT’AMBROGIO. Localization of pulmonary stretch receptors in the airways of the dog. J. Physid Lvndun 235: 775-782,

Am

335. MLCZOCH, J., A. TUCKER, E. K. WEIR, J. T. REEVES, AND R. F. GROVER. Platelet-mediated pulmonary hypertension and hypoxia during pulmonary microembolism. Reduction by platelet inhibition. Chest 74: 648-653, 1978. 336. MODIG, J., C. BUSCH, S. OLERUD, AND T. SALDEEN. Pulmonary microembolism during intramedullary orthopedic trauma. Acta Anesthe8id Scund 18: 133-143, 1974. 337. MONCADA, S., AND J. R. VANE. Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane An, and prostacyclin. Phawnucd Rev. 30: 293331, 1979. 338. MOON, V. H., AND D. R. MORGAN. Experimental pulmonary edema. Proc Sue Exp. Bid Med 35: 560-562, 1935-36. 338a.MOOSAV1, H., M. T. UTELL, R. W. HYDE, F. J. FAHEY,

Rev, Respir.

Dia

94: 3&l-387,1966.

318. McPHAIL, L. C., P. M. HENSON, AND R. B. JOHNSTON, JR. Reapiratory burst enzyme in human neutrophils. J. Clin. Invest 67: 710-716,198l. 319. MEAD, J., T. TAKISHIMA, AND D. LEITH. Stress distribution in lungs: a model of pulmonary elasticity. J. Appl

Physid

28: 596~608,197O.

329. MEGIBOW, R. S., L. N. KATZ, AND M. FEINSTEIN. Kinetics of respiration in experimental pulmonary embolism. Arch. InMed 71: 536-546, 1943. 321. MEGIBOW, R. S., L. N. KATZ, AND F. S. STEINLITZ. Dynamic changes in experimental pulmonary embolism. surgery 11: 19-32, 1942. 322. MEYER, E. C. Acute and chronic clearance of lung fluids, proteins and cells. In: Lund Water and Solute Ezhunge, edited by N. C. Staub. New York; Dekker, 1978, p. 277-

““I. 231

1973.

1202

339.

340.

341.

342.

ASRAR

B. T. PETERSON, J. DONNELLY, AND K. M. JENSEN. Lung ultrastructure in noncardiogenic pulmonary edema induced by air embolisation in dogs. Luh Inve& 45: 456464, 1981. MORGAN, T. E., AND L. H. EDMUND& JR. Pulmonary artery occlusion. III. Biochemical alterations. J. AppL Physid 22: 1012-1016, 1967. MORKIN, E., 0. R. LEVINE, AND A. P. FISHMAN. Pulmonary capillary flow pulse and the site of pulmonary vasoconstriction in the dog. Circ Rea 15: 146-160, 1964. MOROZ, L. A., A. D. SNIDERMAN, AND D. G. F. MARPOLE. Leukocyte-plasma interaction in fibrinolysis. N. En& J. Md 301: 119%1104,1979. MORRIS, H. R., P. J. PIPER, G. W. TAYLOR, AND J. R. TIPPINS. The role of arachidonate lipoxygenase in the release of SRS-A from the guinea pig chopped lung. Prm-

taglandins 19: 371-377,198O. 343. MOSER, K. M. Pulmonary embolism. Am. Rev. Respir. mu 115: 829~852,1977. 344. MOSER, K. M., AND M. STEIN. Pulmmarg Thrmnboem-

holism. Chicago,

IL: Year Book, 1973.

Acute inflammation: the earliest fine structural changes at the bloodtissue level. Lo& Invest. 12: 895-910, 1963. 346. MUSTARD, J. F., H. Z. MOVAT, D. R. C. MACMORINE, AND A. SENYI. Release of permeability factors from blood platelets. Proc Sot Exp Bid M& 119: 988-991,

345. MOVAT,

H. Z., AND N. V. P. FERNANDO.

1965.

347. NADEL, J. A. Airway responses to pulmonary embolism: structural changes. In: Pulnumary Thmmboe&&ism, edited by K. M. Moses and M. Stein. Chicago, IL: Year Book, 1973, p. 160-165. 348. NADEL, J. A. Autonomic control of airway smooth muscle and airway secretions. Am. Rev. ReqGr. Dia 115, Suppl.: 117-126, 1977. 349. NADEL, J. A., G. A. CABEZAS, AND J. H. M. AUSTIN. In uiw, roentgenographic examination of parasympathetic innervation of small airways. Use of powdered tantalum and fine focal spot x-ray tube. Invest. Radid 6: 914, 1971.

350. NADEL, J. A., H. J. H. COLEBATCH, AND C. R. OLSEN. Location and mechanism of airway constriction after barium sulfate microembolism. J. Appl Ph@oL 19: 387394,1964. 351. NADEL,

352.

353.

354. 355.

356.

357.

J. A., H. SALEM, B. TAMPLIN, AND Y. TOKIWA. Mechanism of bronchoconstriction during inhalation of sulfur dioxide. J. AppL Phgsid 20: X4-167, 1965. NADEL, J. A., AND J. G. WIDDICOMBE. Effect of changes in blood gas tensions and carotid sinus pressure on tracheal volume and total lung resistance airflow. J. Physid Lo&m 163: 13-33,1962. NADEL, J. A., AND J. G. WIDDICOMBE. Reflex effects of upper airway irritation on total lung resistance and blood pressure. J. Appl Ph@d 17: 861-865,1962. NAGAISHI, C. Fmctimal Anatomy and Histdagy of the Lzcng. Baltimore, MD: University Park, 1972. NAKAHARA, EC., K. OHKUDA, m N. C. STAUB. Effect of infusing histamine into pulmonary or bronchial artery on sheep pulmonary fluid balance. Am Rev. Respir. Lxu 120: 877-882,1979. NAKANO, J., AND R. B. McCLOY, JR. Effects of indomethacin on pulmonary vascular and airway resistance responses to pulmonary microembolisation. Pmc Sot Exp Bid Med. 143: 218-221, 1973. NELSON, J. R., AND J. R. SMITH. The pathologic physiology of pulmonary embolism: a physiologic discussion of the vascular reactions following arterial obstruction by emboli of various size. Am Heart J. 58: 911-932,1959.

3. MALIK

vi&me

6.9

358. NEWBALL, H. H., H. R. KEISER, AND J. J. PISANO. Bradykinin and human airways. Respir. Physid 24: 139146, 1975. 359. NICOLAYSEN, G., A. NICOLAYSEN, AND N. C. STAUB. A quantitative radiographic comparison of albumin concentration in different sized lymph vessels in normal mouse lung. Micravasc &a 10: 138-152, 1975. 360. NIDEN, A. H., AND D. M. AVIADO. Effects of pulmonary embolism on the pulmonary circulation with special reference to arteriovenous shock in the lung. Circ Res. 4: 67-73, 1956. 361. NIEHAUS, G. D., P. T. SCHUMACKER, AND T . M. SABA. Influence of opsonic fibronectin deficiency on lung fluid balance during bacterial sepsis. J. Appl Physid; Respirat. Envirm. Exercise Physid 49: 693-699,198O. 362. NIELSEN, T. M., AND 0. F. PEDERSEN. The effect of CO, on peripheral airways. Actu Physid Scund 98: 192199, 1976.

363. NISAM, M. R., A. ZBINDEN, S. CHESROWN, D. BARNETT, AND W. M. GOLD. Distribution and pharmacological release of histamine in canine lung in vivo. J. Appl Physid: Reqirat. Environ, Exercise Physid 44: 455-463, 1978. 364. NISHI, K., T. SAKANASHI, AND F. TAKENAKA. Afferent fibers from pulmonary arterial band receptors in the left cardiac sympathetic nerves of the cat. J. Physid L#ndon 240: 53-60,1974. 365. NOSSEL, H. L. Radioimmunoassay of fibrinopeptides in relation to intravascular coagulation and thrombosis. N. Errgl. J. Med 295: 428-432,1976. 366. NOTKOVICH, H. The anatomy of the bronchial arteries of the dog. J. Thmac Surg. 33: 242-253, 1957. 367. O’BRODOVICH, H. M., S. A. STALCUP, L. M. PANG, J. S. LIPSET, AND R. B. MELLINS. Bradykinin production and increased pulmonary endothelial permeability during acute respiratory failure. J. Clin, Inue8t. 67: 514523, 1981. 368. O’FLAHERTY, J. T., S. COUSART, A. S. LINEBERGER, E. BOND, D. A. BASS, L. R. DECHATELET, E. S. LEAKE, AND C. E. MCCALL. Phorbol myristate acetate: in vivo effects upon neutrophils, platelets and lung. Am J Pathd 101: 79-92, 1980. 369. O’FLAHERTY, J. T., D. L. DREUTZER, AND P. A. WARD. Effect of PGEl, PGEs and Fs, on neutrophil aggregation. Prosta&zn&ns 17: 201-210, 1979. 370. O’FLAHERTY, J. T., M. J. THOMAS, C. J. LEES, AND C. E. MCCALL. Neutrophil-aggregating activity of monohydroxyeicosatetraenoic acids. Am. J. Puthd 104: 55-62, 1981. 371. OGLETREE, M., AND K. BRIGHAM. Indomethacin and meclofenamate augment endotoxin induced increase in lung vascular permeability in sheep (abstr.). Am Rev. Respir. Dis 119: 374, 1979. 372. OGLETREE, M. L., AND K. L. BRIGHAM. PGE1 reduces lung transvascular filtration during endotoxin induced high permeability in sheep (abstr.). Clin Rea 27: 402A, 1979. 373. OGLETREE, M. L., AND K. L. BRIGHAM. Prostacyclin (PGIs) and PGEi produce opposite effects on lung vascular permeability (abstr.). Federation Pruc 38: 5494, 1979.

374. OGLETREE, M. L., AND K. L. BRIGHAM. Arachidonate raises vascular resistance but not permeability in lungs of awake sheep. J. Appl Physid: Reqkzt. Envim Exercise Physid 48: 581-586, 1980. 375. OHKUDA, K., K. NAKAHARA, A. BINDER, AND N. C. STAUB. Venous air emboli in sheeD: reversible increase l

~-

July

1983

PULMONARY

MICROEMBOLISM

in lung microvascular permeability. J. AppL Physid: Reap’rat. Environ. Exer&e Physid 51: 887-894, 1981. 376. OHKUDA, K., K. NAKAHARA, W. J. WEIDNER, A. BINDER, AND N. C. STAUB. Lung fluid exchange after uneven pulmonary artery obstruction in sheep. Circ Rea

395.

43: 152-161,1978. 377. ORANGE, R. P., AND K. F. AUSTEN.

378.

379.

380.

381. 382. 383.

384. 385. 386.

The immunological release of chemical mediators of immediate type hypersensitivity from human lung. In: Progress in Immunology, edited by B. Amos. New York: Academic, 1971, p. 173-186. (Int. Congr. Immunol., 5th.) ORLOFF, K. G., AND D. MICHAELI. Fibrin-induced release of platelet serotonin. Am J. Physid 231: 344350, 1976. OSBAHR, A. J., J. A. GLADNER, AND K. LAKI. Studies on the physiological activity of the peptide released during the fibrinogen-fibrin conversion. Biochem. Biophya Acta 86: 535-542, 1964. OZDEMIR, I. A., W. R. WEBB, AND S. D. WAX. Effects of neural and humors1 factors on pulmonary hemodynamics and microcirculation in pulmonary embolism. J. Thorn Cardimasc. Surg. 68: 896-924,1974. PACKHAM, M. A., AND J. F. MUSTARD. Clinical pharmacology of platelets. Blood 50: 555-573, 1977. PAINTAL, A. S. Mechanism of stimulation of type J pulmonary receptors. J. Physid Lon&an 203: 511-532, 1969. PAINTAL, A. S. The mechanisms of excitation of type J receptors and J reflex. In Breathing: HerinpBreuer Centenary Symposium, 1969. London: Churchill, 1970. (Ciba Found. Symp.) PAINTAL, A. S. Vagal sensory receptors and their reflex effects. Physid Rev. 53: 159-227.1973. PAINTAL, A. S. Thoracic receptors connected with sensation. Br. Med Bull 33: 169-174, 1977. PANG, L. M., H. M. O’BRODOVICH, R. B. MELLINS, AND S. A. STALCUP. Bradykinin-induced increase in pulmonary vascular permeability in hypoxic sheep. J. Appl Physid: Respirat Environ. Exercbe Phyeid 52:

370-375, 1982. 387. PANG, L. M., F. RODRIQUEZ-MAkTINEZ, S. A. STALCUP, AND R. B. MELLINS. Effect of hyperinflation and

atelectasis on fluid accumulation AppL Phygid: Respirat Envim

in the puppy lung. J. Exercbe Physid 45:

396.

397.

398.

399.

400.

401.

402.

403.

404.

405.

466.

284-288,1978.

388. PANOSSIAN, A., J. W. C. HAGSTROM, S. L. NEHLSEN, AND F. J. VEITH. Secondary nature of surfactant changes in post-perfusion pulmonary damage. J. Thorax: Cc&i+ vaac Suqg 57: 628-634,1969. 389. PARKER, B. M., AND J. R. SMITH.

390.

391.

392.

393.

394.

Studies of experimental pulmonary embolism and infarction and the development of collateral circulation in the affected lung lobe. J. La.& Ctia Msd 49: 850-857,1957. PARMLEY, L. F., R. L. NORTH, AND B. S. OH. Hemodynamic alteration of acute pulmonary thromboembolism. Circ &a 11: 450-465, 1962. PEARCE, M. L., B. YAMASHITA, AND J. BEAZELL. Measurement of pulmonary edema. Circ Rea 16: 482-488, 1965. PERL, W. Modified filtration-permeability model of transcapillary transport-a solution of the PappenheiRea 3: 233~251,197l. mer pore puzzle? Mi0wta3c PERMUTT, S. Mechanical influences on water accumulation in the lungs. In: Pulmaary Edema, edited by A. P. Fiahman and E. M. Renkin. Bethesda, MD: Am. Physiol. Sot., 1979, p. 175-193. PERMUTT, S., P. CALDINI, A. MASERI, W. H. PALMER, T. SASAMORI, MD K. ZIERLES. Recruitment versus distensibility in the pulmonary vascular bed. In:

497.

408.

499. 410.

1203

The Pulmonu~ Circulation and Interstitial Space, edited by A. P Fishman and H. H. Hecht. Chicago, IL: Univ. of Chicago Press, 1969, p. 375-390. PETRONE, W. F., D. K. ENGLISH, K. WONG, AND J. M. McCORD. Free radicals and inflammation: superoxide-dependent activation of a neutrophil chemotactic factor in plasma. Proc N&l. Acad. Sci USA 77: 11591163, 1980. PIETRA, G. G., AND A. P. FISHMAN. Bronchial edema. In: Lung Water and Solute Exchunge, edited by N. C. Staub. New York: Dekker, 1978, p. 407-421. PIETRA, G. G., J. P. SZIDON, H. A. CARPENTER, AND A. P. FISHMAN. Bronchial venular leakage during endotoxin shock. Am J. Path& 77: 387-406, 1974. PIPER, P. J., AND J. R. VANE. The release of prostaglandins from lung and other tissues. Ann. NYAcad Sci 18: 363-383, 1971. PLOY-SONG-SANG, Y., R. P. CORBIN, AND L. A. ENGEL. Effects of intravenous histamine on lung mechanics in man after beta-blockade. J. Appl Physid: Req&at Environ. Exercise Phyaid 44: 690-695,1978. POWERS, S. R., JR., D. SHAH, D. RYON, J. NEWELL, C. RALPH, W. SCOVILL, AND R. DUTTON. Hypertonic mannitol in the therapy of the acute respiratory distress syndrome. Ann. Surg. 185: 619-625, 1977. PRICE, K. C., D. HATA, AND J. R. SMITH. Pulmonary vasomotion resulting from miliary embolism of the lungs. Am J. Physid 182: 183-190,1955. PROPST, K., J.E. MILLEN, AND F. L. GLAUSER. The effects of endogenous and exogenous histamine on pulmonary alveolar membrane permeability. Am. Rev. Respir. I.& 117: 1063-1068,1978. RADEGRAN, K. Circulatory and respiratory effects of induced platelet aggregation: an experimental study in dogs. Acta Chir. Scan& Suppl 420: 7-24,197l. RADEGRAN, K. The effect of acetylsalicylic acid on the peripheral and pulmonary vascular responses to thrombin. Acta An&he&d Scund 16: 140-146, 1972. RANADIVE, N. S., AND C. G. COCHRANE. Mechanism of histamine release from mast cells by cationic protein (band 2) from neutrophil lysosomes. J. Immund 106: 506-516, 1971. RAPOPORT, S. I. B&z&-Brain Barrier in Phvw and iUe&ci~ New York: Raven, 1976. RATLIFF, N. B., J. W. WILSON, E. MIKAT, D. B. HACKEL, AND T. C. GRAHAM. The lung in hemorrhagic shock. IV. The role of neutrophilic polymorphonuclear leukocytes. Am. J. Pathd 65: 325-334,197l. REGOECZI, E., J. GARGELY, AND A. S. McFARLANE. In vivo effects of agkistrodon rhodoatoma venom studies with fibrinogen- ltlI. J. Clin Inve8t 45: 1202-1212, 1966. REGOLI, D., AND J. BARABE. Pharmacology of bradykinin and related kinins. Phurmucul Rev. 32: l-46,1980. REID, L. M. Definition and organization of the pulmonary microcirculation at the tissue level. Ann NYAcad

sci 384: 3-20,1982. 411. REIF, M. E., J. R. McCURDY,

J. J. COALSON, AND L. J. GREENFIELD. Effects of mannitol and dextran on the interstitial pulmonary edema. J. SUTQ. Rea 12: 234239,1972. 412. RENKIN, E. M. Multiple pathwaya of capillary permeability. Circ Rea 41: 735-743.1977. 413. RENKIN, E. M. Lymph as a measure of the composition of interstitial fluid. In: Pulmonary Edemu, edited by A. P. Fishman and E. M. Renkin. Bethesda, MD: Am. Physiol. Sot., 1979, p. 145-159. 414. RENKIN, E. M., W. J. JOYNER, C. H. SLOOP, AND P. D. WATSON. Influence of venous pressure on plasma-

1204

ASRAR

lymph transport in the dogs paw: connective and dissipative mechanisms. Microvasc Res. 14: 191-204, 197’7. 415. RIPPE, B., A. KAMIYA, AND B. FOLKOW. Is capillary micropinocytosis of any significance for the transcapillary transfer of plasma proteins? Acta Physid Stand loo: 258-260, 1977. 416. RIPPE, B., A. KAMIYA, AND B. FOLKOW. Transcapillary passage of albumin, effects of tissue cooling and of increases in filtration and plasma colloid osmotic pressure. Acta PhysioL Scati 105: 171-187, 1979. 417. RISBERG, B. Fibrinolysis in the lung. Acta Chir. Stand Suppl 458: l-36, 1975. 418. RISBERG, B. Effect of tranexamic acid on pulmonary tissue fibrinolysis in the rat. Micr~asc Res. 11: 307-314, 1976. 419. ROBERTS, W. C. Coronary embolism: a review of causes, consequences, and diagnostic considerations. Cardiovasc Med 3: 699-710, 1978. 420. ROBIN, E.a D., C. E. FORKNER, P. A. BROMBERG, J. R. COROTEAU, AND D. M. TRAVIS. Alveolar gas exchange in clinical pulmonary embolism. N. Engl .I. Med 262: 283-287, 1960. 421. ROHATAGI, P., I. TAUBER, AND D. MASORO. Hemorrhagic hypotension and the lung: in vitro respiration. Am. Rev. Reqir. Dk 113: 763-767, 1976. 422. ROOS, P., J. WIENER-KRONISH, K. ALBERTINE, M. MATTHAY, AND N. STAUB. Removal of subdiaphragmatic contamination of caudal mediastinal node (CMN) efferent lymph, in anesthetized sheep (abstr.). Physiolu gist 24(4): 281, 1981. 423. ROY, A. J., AND I. DJERASSI. Effects of platelet transfusions: plug formation and maintenance of vascular integrity. Proc Sot Exp. Bid Med 139: 137-142, 1972. 424. RUDDY, S., I. GIGLI, AND K. F. AUSTEN. The complement system of man. N. Engl. J. Med 287: 489-494,1972. 424a.RUSZNYAK, J., M. FOLDI, AND G. SZABO. Lgmphutics and Lymph Circulation: Physiodogy and Pat(2nd ed.). Oxford, UK: Pergamon, 1967. 425. RYAN, J. W., AND U. S. RYAN. Pulmonary endothelial cells. Federation Proc 36: 2683-2691, 1977. 426. RYAN, J. W., AND U. S. RYAN. Humoral control of arterial blood pressure: a role for the lung? Cardiouasc Md 3: 531-550, 1978. 427. SABA, T. M. Physiology and physiopathology of the reticuloendothelial system. Arch, Zntemz Med. 126: 10311052, 1970. W. A. SCOVILL, 428. SABA, T. M., F. A. BLUMENSTOCK, AND H. BERNARD. Cryoprecipitate reversal of opsonic aa-surface binding glycoprotein deficiency in septic surgical and trauma patients. science 201: 622-624, 1978. 429. SABA, T. M., F. A. BLUMENSTOCK, P. WEBER, AND J. E. KAPLAN. Physiologic role of cold-insoluble globulin in systemic host defense: implications of its characterization as the opsonic as-surface binding glycoprotein. Ann. NY Acad Sci 312: 43-55, 1978. 430. SABA, T. M., AND E. JAFFE. Plasma fibronectin (opsonic glycoprotein): its synthesis by vascular endothelial cells and role in cardiopulmonary integrity after trauma as related to reticuloendothelial function. Am .I. Med 68: 577-594, 1980. 431. SACKS, T., C. F. MOLDOW, P. R. CRADDOCK, T. K. BOWERS, AND H. S. JACOB. Oxygen radicals mediate endothelial cell damage by complement-stimulated granulocytes. J. Clin. Znved 61: 1161-1167, 1978. 432. SAID, S. I. Release induced by physical and chemical stimuli. In: Metubolic Functions of the Lung, edited by Y. S. Bakhle and J. R. Vane. New York: Dekker, 1977, p. 297-320.

B. MALIK

vohmne

63

433. SALDEEN,

bolism

T. Experimental investigation on fat emand intravascular coagulation (abstr.). Acta Pa-

th& Microbid &and 66: 271,1966. 434. SALDEEN, T. The disappearance of fibrin

from the pulfat embolism. Thromb. Death. H a.emodL 22: 360-371, 1969. SALDEEN, T. Intravascular coagulation in the lungs in experimental fat embolism. Acta Chir. Stand 135: 653662, 1969. SALDEEN, T. Fat embolism and signs of intravascular coagulation in a posttraumatic autopsy material. J. Trauma 10: 273-286, 1970. SALDEEN, T. The microembolism syndrome. Microva.w Res. 11: 227-259, 1976. SALDEEN, T. The microembolism syndrome: a review. In: The Microembolism Syndrome, edited by T. Saldeen. Stockholm: Almqvist & Wiksell, 1979, p. 7-44. SALIN, M. L., AND J. M. McCORD. Free radicals and inflammation. Protection of phagocytosing leukocytes by superoxide dismutase. J. Clin. Invest. 56: 1319-1323,1975. SALISBURY, P. F., P. WEIL, AND D. STATE. Factors influencing collateral blood flow to the dog’s lung. Circ Res. 5: 303-309, 1957. SAMPSON, S. R. Sensory neurophysiology of airways. Am. Rev. Reqir. Dis. 115, Suppl.: 107-115, 1977. SAYEED, M. M., AND A. E. BAUE. Na-K transport in rat liver slices in hemorrhagic shock. Am J Physid 224: 1265-1270, 1973. SAYEED, M. M., I. H. CHAUDRY, AND A. E. BAUE. Na+-K+ transport and adenosine nucleotides in lung in hemorrhagic shock. Surgery 77: 395-402, 1975. SCHACHTER, M. Kallikreins (kininogenases)-A group of serine proteases with bioregulatory actions. Phurmad Rev. 31: 1-17, 1986. SCHLEIMER, R. P., S. M. LICHTENSTEIN, AND E. GILLESPIE. Inhibition of basophil histamine release by anti-inflammatory steroids. Nature London 292: 454-457, 1981. SCHNEEBERGER, E. E. Ultrastructural basis of alveolar-capillary permeability to proteins. In: Lung Liquids. Amsterdam: Elsevier, 1976, p. 3-28. (Ciba Found. Symp. 38.1 SCHNEEBERGER-KEELEY, E. E., AND M. J. KARNOVSKY. The ultrastructural basis of alveolar capillary membrane permeability to peroxidase used as a tracer. J. CeU Biol 37: 781-793, 1968. SELYE, H. The Mast CeU Washington, DC: Butterworths, 1965. SEVERINGHAUS, J. W. Pulmonary vascular function. Am. Rev. Respir. Dia. 115, Suppl.: 149-158, 1977. SEVERINGHAUS, J. W., AND M. A. STUPFEL. Alveolar dead space as an index of distribution of blood flow in pulmonary capillaries. J. Appl Physid 10: 335-348,1957. SEVERINGHAUS, J. W., E. W. SWENSON, T. N. FINLEY, M. T. LATEGOLA, AND J. WILLIAMS. Unilateral hypoventilation produced in dogs by occluding one pulmonary artery. J. Appl Physid 16: 53-60, 1961. SEVITT, S. Pathology and pathogenesis of dog’s vein thrombosis. In: Pulmonary Thr omboembdism, edited by K. M. Moser and M. Stein. Chicago, IL: Year Book, 1973, p. 93-103. SHAW, J. O., P. M. HENSON, J. HENSON, AND R. 0. WEBSTER. Lung inflammation induced by complementderived chemotactic fragments in the alveolus. &zb. Znvest. 42: 547-558, 1980. SHEPARD, J. W., JR., D. HAUER, K. MIYAI, AND K. M. MOSER. Lamellar body depletion in dogs undergoing monary

435.

436.

437. 438.

439.

440.

441. 442.

443.

444.

445.

446.

447

448 449 450.

451.

452.

453.

454.

vessels

in experimental

July

1983

PULMONARY

pulmonary artery occlusion. J. C&n. Invest. 66: 36-42, 1980. 455. SHEPARD, R. H. Effect of pulmonary diffusing capacity on exercise tolerance. J. Appl Physid 12: 487-488,1958. 456. SHEPHERD, J. T. The lungs as receptor sites for cardiovascular regulation. Circulation 63: l-10, 1981. 45’7. SHERRY, S. The problem of thromboembolic disease. Semin. NUCG Med 8: 205-211, 1977. 458. SHIRLEY, H. H., JR., C. G. WOLFRAM, K. WASSERMAN, AND H. S. MAYERSON. Capillary permeability to macromolecules: stretched pore phenomenon. Am J. Phy&d 190: 189-193, 1957. 459. SIMIONESCU, N., M. SIMIONESCU, AND G. E. PALADE. Permeability of muscle capillaries to small hemepeptides. Evidence for the existence of patent transendothelial channels. J. Cell Bid 64: 586-607, 1975. 466. SINGER, D., C. HESSE, R. PICK, AND L. N. KATZ. Diffuse bilateral pulmonary edema associated with unilobar pulmonary embolism in dogs. Circ Rea 6: 4-9, 1958. 461. SKAER, R. J. Elemental composition of platelet dense bodies. In: Biochemistry and Phwm.acologg of Platelets. Amsterdam: Elsevier, 1975, p. 239-252. (Ciba Found. Symp. 35.) 462. SKUBITZ, K. M., P. R. CRADDOCK, D. E. HAMMERSCHMIDT, AND J. T. AUGUST. Cortisteroids block binding of chemotactic peptide to its receptor on granulocytes and cause dieaggregation of granulocyte aggregates in vitro. J. Clin, Invest 68: 13-20, 1981. 463. SMITH, M. E., R. GUNTHER, M. GEE, J. FLYNN, AND R. H. DEMLING. Leukocytes, platelets, and thromboxane As in endotoxin-induced lung injury. Surgm 96: 102107, 1981. 464. SMITHEN, C., J. CHRISTODOULOU, T. KILLIP, AND N. BRACHFELD. Metabolic and hemodynamic consequences of mannitol following myocardial anoxia. Am. J. Physid 229: 847-852, 1975. 465. SNASHALL, P. D., W. J. WEIDNER, AND N. C. STAUB. Extravascular lung water after extracellular fluid volume expansion in dogs. J. Appl Physid: Respirat. Environ. Exercise Physid 42: 624-629, 1977. 466. SOBIN, S. S., M. INTALIETTA, W. FRASHER, JR., AND J. M. TREMER. The geometry of the pulmonary microcirculation. Angidogy 17: 24-30, 1966. 466a.SPAGNUL0, P. J., J. J. ELLNER, A. HASSID, AND M. J. DUNN. Thromboxane As generation mediates augmented polymorphonuclear leukocyte adhesiveness. J. Clin. Inve8t. 66: 406-414,1986. 467. STALCUP, S. A., J. S. LIPSET, J. WOAN, P. LEUENBERGER, AND R. B. MELLINS. Inhibition of angiotensin converting enzyme activity in cultured endothelial cells by hypoxia. J. Clin Invest. 63: 966-976, 1979. 468. STALCUP, S. A., AND R. B. MELLINS. Mechanical forces producing pulmonary edema in acute asthma. N. Engl J. Med

1205

MICROEMBOLISM

297: 592-600,1977.

469. STARLING, E. H., AND E. B. VERNEY. The secretion of urine as studied in the isolated kidney. ti R Sot lkmdim Ser. B 97: 321-363, 1925. 470. STAUB, N. C. Pulmonary edema. Phgsid Rev. 54: 678811,1974. 471. STAUB, N. C. Pulmonary edema due to increased microvascular permeability to fluid and protein. Circ Res. 43: 143-151, 1978. 472. STAUB, N. C., M. GEE, AND C. VREIM. Mechanism of alveolar flooding in acute pulmonary edema. In: Lung Liquids Amsterdam: Elsevier, 1976, p. 255-263. (Ciba Found. Symp. 38.) 473. STAUB, N. C., H. NAGANO, AND M. L. PEARCE. Pul-

monary edema in dogs, especially the sequence of fluid accumulation in lung. J. AppL Physid 22: 227-240,1967. 474. STAUB, N. C., AND K. OHKUDA. PGEi reverses increased lung vascular permeability during air emboli. Microvasc Rea 15: 271-272, 1978. 475. STEIN, M., T. HIROSE, T. YASUTAKE, AND A. TARABEIN. Airway responses to pulmonary embolism: pharmacologic aspects. In: Pulmonary Thmmboembdigm, edited by K. M. Moser and M. Stein. Chicago, IL: Year Book, 1973, p. 166-177. 476. STEIN, M., AND S. R. LEVY. Reflex and humoral response to pulmonary embolism. Prug. Curdiova8c LG. 17: 167174, 1974. 477. STEPHENS, N. L., J. L. MEYERS, AND R. M. CHERNIACK. Oxygen, carbon dioxide, H+ ion, and bronchial length-tension relationships, J. Appl. Physid 25: 376383,1968. 478. SU, C., R. D. BEVAN, S. P. DUCKLES, AND J. A. BEVAN. Functional studies of the small pulmonary arteries. Mimast Res. 15: 37-44, 1978. 479. SUEISHI, K., S. NANNO, AND K. TANAKA. Permeability enhancing and chemotactic activities of lower molecular weight degradation products of human fibrinogen. Thromb. Haemu& 45: 90-94, 1981. 480. SVENSJO, E., C. G. A. PERSSON, AND G. RUTILI. Inhibition of bradykinin induced macromolecular leakage from post-capillary venules by a /3s-adrenoreceptor stimulant, terbutaline. Actu Physid Scund 101: 504~566,1977. 481. SWEETMAN, H. E., D. SHEPRO, AND H. B. HECHTMAN. Inhibition of thrombocytopenic petechiae by exogenous administration. Haemostaaia 10: 65-78, 1981. 482. SWENSON, E. W., T. N. FINLEY, AND S. V. GUZMAN. Unilateral hypoventilation in man during temporary occlusion of one pulmonary artery. J. C&n, Invest. 40: 828835, 1961. 483. TAHAMONT, M. V., P. S. BARIE, F. A. BLUMENSTOCK, M. V. HUSSAIN, AND A. B. MALIK. Increased lung vascular permeability after pancreatitis and trypsin. Am J. Puthd 109: 15-26, 1982. 483a.TAHAMONT, M. V., AND A. B. MALIK. Granulocytes mediate the increase in pulmonary vascular permeability after thrombin embolism. J. AppL Physid: Respirut Environ.

Exer&e

Physid

54: 1489-1495,

1983.

484. TAKAYAMA, H., T. KATSUMOTO, AND A. TAKAGI. Electron microscopic studies of macrophages. II. Effect of colchicine and cytochalsin B on the cell cortex as revealed by ultrathin sectioning. J. Electron Mimwc 26: 718, 1977. 485. TAYLOR, A. E., AND K. A. GAAR, JR. Estimation of equivalent pore radii of pulmonary capillary and alveolar membranes. Am. J. Physid 218: 1133-1140, 1970. 486. TAYLOR, A. I&, AND R. E. DRAKE. Fluid and protein movement across the pulmonary microcirculation. In: Lung Water and So&% Exchange, edited by N. C. Staub. New York: Dekker, 1978, p. 129-166. 488. TAYLOR, A. E., J. C. PARKER, D. N. GRANGER, N. A. MORTILLARO, AND G. RUTLI. Assessment of capillary permeability using lymphatic protein flux: estimation of the osmotic reflection coefficient. In: The Microcirculation, edited by R. M. Effros, H. Schmid-Shanbein, and J. Ditzel. New York: Academic, 1981, p. 19-31. 489. TEPLITZ, C. Pulmonary cellular and interstitial edema. In: Pulmmmq Edema, edited by A. P. Fishman and E. M. Renkin. Bethesda, MD: Am. Physiol. Sot., 1979, p. 97-111. 496. THEISS, W., P. HILGARD, AND H. HEYES. Induction of disseminated intravascular coagulation by endotoxin

1206

491.

492.

493.

494.

495.

ASRAR

and saline loading in rats. II. Fibrin deposition and removal. Thmmb. Rea 7: 47-58.1975. THOMAS, D., M. STEIN, G. TANABE, V. REGE, AND S. WESSLER. Mechanism of bronchoconstriction produced by thromboemboli in dogs. Am. J. Physid 206: 1207-1212,1964. THOMAS, D. P., V. GUREWICH, AND T. P. ASHFORD. Platelet adherence to thromboemboli in relation to the pathogenesis and treatment of pulmonary embolism. N. EngL J. Med 274: 953-956,1966. THOMAS, G. Characteristics of prostaglandin El potentiation of inflammatory activity of some agents. Prost4zghndim -19: 39-45, 1986. THORGEIRSSON, G., AND A. J. ROBERTSON, JR. The vascular endothelium-pathobiologic significance. A review. Am. J. Path& 93: 803-848, 1978. TISI, G. M., W. G. WOLFE, R. J. FALLAT, AND J. A. NADEL. Effects of 00 and CO% on airway smooth muscle following pulmonary vascular occlusion. J. AppL Physid 28: 570-573,197o.

496. TOBIN, C. E. Some observations concerning the pulmonic vasa vasorum. Surg. Gyecd &tet. 111: 297-363,X%0. 497. TODD, A. S. Some topographical observations on fibrinolysis. J. Clin Path& 17: 324-327, 1964. 498. TOOLEY, W. H. Hyaline membrane disease. Am Rev. Respir. Dis. 115, Suppl.: 19-28, 1977. 499. TUCKER, A., E. K. WEIR, J. T. REEVES, AND R. F. GROVER. Histamine H1- and Hz-receptors in pulmonary and systemic vasculature of the dog. Am J. Physid 229: 1008-1013, 1975. 500. TUCKER, A., E. K. WEIR, J. T. REEVES, AND R. F. GROVER. Pulmonary microembolism: attenuated pulmonary vasoconstriction with prostaglandin inhibitors and antihistamines. Proetagtandins 11: 31-41, 1976. 501. UEDA, H., Y. UCHIDA, AND K. KAMISAKA. Distribution and response of the cardiac sympathetic receptors to mechanically induced circulatory changes. Jm Heart J. lo: 70-81,1969. 592. UENO, A., K. TANAKA, M. KATORI, M. HAYASHI, AND Y. ARAI. Species difference in increased vascular permeability by synthetic leukotriene Cr and D,. Prostuglundim 21: 637-647, 1981. 503. UTSUNOMIYA, T., M. M. KRAUSZ, C. R. VALERI, D. SHEPRO, AND H. B. HECHTMAN. Treatment of pulmonary embolism with prostacyclin. Surgery 88: 25-30, 1980. 564. UVNAS, B. Mast cells and histamine release. Indian J. Phurmucd 1: 23-32, 1969. 595. VAAGE, J. Vagal reflexes in the bronchoconstriction occurring after induced intravascular platelet aggregation. Actu Physid Scund 97: 94-103, 1976. 596. VAAGE, J. Intravascular platelet aggregation and acute respiratory insufficiency. Circ. Shock 4: 279-290, 1977. 567. VAAGE, J., G. BO, AND J. HOGNESTAD. Pulmonary responses to intravascular platelet aggregation in the cat. AC&L Physid Scund 92: 546-556, 1974. 568. VAAGE, J., AND K. GJESDAL. Lung responses and platelet release reaction following induced intravascular platelet aggregation in the cat. Scund J. CZin Lob. Invest. 36: 641-648,1976. 509. VAAGE, J., AND A. HAUGE. Small airway constriction and closure after induced intravascular platelet aggregation. Actu Phpid Scund 100: 221-230, 1977. 510. VAAGE, J., G. NICHOLAYSEN, AND B. A. WAALER. Aggregation of blood platelets and increased hydraulic conductivity of pulmonary exchange vessels. Acta Phyt&t! Scand 98: 175-184, 1976.

B. MALIK

Vohme

63

511. VAAGE, J., AND P. J. PIEPER. The release of prostaglandin-like substances during platelet aggregation and pulmonary microembolism. Acta Physid Scund 97: 94103,1975. 512. VAAGE, J., T. WILBERG, AND R. MUNDAL. Microembolization and prostaglandin (PG) synthesis in isolated, perfused rabbit lungs. Pm&zglandina 15: 714, 1978. 513. VAAGE, J., T. WILBERG, AND E. SCOTT.. Release of prostaglandin-like substances and lung reactions to induced intravascular coagulation. Scund J. Clin. Lab. Invest 38: 337-347, 1978. 514. VAN DER ZEE, H., N. GERTZBERG, W. JURACKA, AND A. B. MALIK. Indomethacin attenuates the increase in lung vascular permeability after pulmonary vascular thrombosis (abstr.). Am Rev. Respir. LXX 125: 283, 1982. 515. VENRATH, H., R. ROTTHOFF, H. VALENTIN, AND W. BOLT. Bronchospirographische untersuchungen bei Durchblutungs Stoeungen in Kleiner Kreislanf. Beitr. Klin Tube& Spez(&&en Tube& Fmch. 197: 291-294, 1952. 516. VINCENT, N. J., R. KNUDSEN, D. E. LEITH, P. T. MACKLEM, AND J. MEAD. Factors influencing pulmonary resistance. J. AppL Physid 29: 236-243, 1970. 517. VIRCHOW, R. L., AND K. GASAMMELTE. Abhmdlungen zur tiensch&lichen Medicin Frankfurt: Meidinger, 1856, p. 285,294~296. 518. VISSCHER, M. B. Studies on embolization of lung vessels. Med Z’hmc 19: 334340, 1962. 519. VISSCHER, M. B., F. J. HADDY, AND G. STEPHENS. The physiology and pharmacology of lung edema. Phurmacd Rev. 8: 389-443, 1956. 529. VON LUSCHKA, H. Die anatomie des Menschen I, pt. 2. In: R&k&& aqfdie Bediirtnisse der Prakhsch.eq’Heilku& Ttibingen, West Germany: Laupp, 1963. 521. VREIM, C. E., K. OHKUDA, AND N. C. STAUB. Proportions of dog lung lymph in the thoracic and right lymph ducts. J. AppL Physid: Reap’rat. Environ Exercise Phy8id 43: 894-898. 1977. 522. VREIM, C. E., P. D. SNASHALL, R. H. DEMLING, AND N. C. STAUB. Lung lymph and free interstitial fluid protein composition in sheep with edema. Am J. Physid 230: 1650-X53,1976. 523. VREIM, C. E., P. D. SNASHALL, AND N. C. STAUB. Protein composition of lung fluids in anesthetized dogs with acute cardiogenic edema. Am J. Physid 231: 14661469, 1976. 524. VREIM, C. E., AND N. C. STAUB. Protein composition of lung fluids in acute alloxan edema in dogs. Am J. Physid 230: 376-379, 1976. 525. WAGNER, R. C., AND J. R. CASLEY-SMITH. Endothelial vesicles. Microvasc Res. 21: 267-298, 1981. 526. WARREN, B. A. Fibrinolytic properties of vascular endothelium. Br. J. Exp. P&d 44: 365-372, 1963. 527. WEDMORE, C. V., AND T. J. WILLIAMS. Control of vascular permeability by polymorphonuclear leukocytes in inflammation. Nature Iondon 289: 646-650, 1981. 528. WEIBEL, E. R. Early stages in the development of collateral circulation to the lung in the rat. Circ Res. 8: 353376, 1960. 529. WEIBEL, E. R. Morphological basis of alveolar-capillary gas exchange. Physid Rev. 53: 419-495, 1973. 530. WEIDMAN, W. H., R. J. MARSHALL, C. A. WAGENVOORT, AND J. T. SHEPHERD. Relation in dogs of pulmonary vascular obstruction to pulmonary vascular “resistance.“Lub. Inveet. 12: 821-829, 1963. 531. WEIL, P., P. F. SALISBURY, AND D. STATE. Physiological factors influencing pulmonary artery pressure

July

532.

533.

534. 535.

536.

1983

PULMONARY

during separate perfusion of the systemic and pulmonary circulations in the dog. Am J. Physid 191: 453-460,1957. WEISBACH, H., T. P. WAALKES, AND S. ALDENFRIEND. Presence of serotonin in lung and its implications in the anaphylactic reaction, Science 125: 235236, 195’7. WEISMANN, G., R. B. ZURIER, AND S. HOFFSTEIN. Leukocyte proteases and the immunologic release of lysosomal enzymes. Am. J. Pathd 68: 539-559, 1972. WEISS, H. J. Properties of platelets. N. Engl J. Med 298: 1344-1347,197s. WEISSMANN, G., J. E. SMOLEN, AND H. M. KORCHAK. Release of inflammatory mediators from stimulated neutrophils. N. En& J. &fed 303: 27-34, 1980. WESSLER, S., D. C. FREIMAN, .I. D. BALLOU, J. H. KATZ, R. WOLFF, AND R. WOLFE. Experimental pulmonary embolism with serum induced thrombi. Am J. Pas&l

the lung after

548. 549.

550.

551.

552.

553

554

25: 729-732,1968.

539. WHITTERIDGE, D. Multiple embolism of the lung and rapid shallow breathing. Physid Rev. 30: 475-486,195O. 540. WIDDICOMBE, J. G. Regulation of tracheobronchial smooth muscle. Phgtid Rev. 43: l-37, 1963. 541. WIDDICOMBE, J. G. Reflex mechanisms in pulmonary thromboembolism. In: Pulmonary Thromboembdism, edited by K. M. Moser and M. Stein. Chicago, IL: Year Book, 1973, p. 178-186. 542. WIDDICOMBE, J. G. Studies on afferent airways innervation. Am Rev. Respir. Dis. 115, Suppl.: 99-105, 1977. 543. WIDDICOMBE, J. G., AND M. GLOGOWSKA. Relative roles of irritant, type-J and pulmonary stretch receptors in lung reflexes. Acta New&d Exp 33: 21-31,1974. 544. WILHELM, D. L. Chemical mediators. In: The Immmutory Procea (2nd ed.), edited by B. W. Zweifach, L. Grant, and R. T. McCluskey. New York: Academic, 1973, vol. 2, p. 251301. 545. WILLERSON, J. T., W. J. POWELL, JR., T. E. GUINEY, J. J. SHARK, C. A. SANDERS, AND A. LEAF. Improvement of myocardial function and coronary blood flow in ischemic myocardium after mannitol. J. C&n Invest. 51: 2989-2998.1972. 546. WILLIAMS, M. H., JR. Mechanical vs. reflex effects of diffuse pulmonary embolism in anesthetized dogs. Circ Res. 4: 325~331,1956. 547. WILLIAMS, M. H., JR., AND E. J. TOWBIN. Magnitude and time of development of the collateral circulation to

occlusion of the left pulmonary artery. 1955. WILLIAMS, W. J. The activity of lung microsomes in blood coagulation. J. Bid Chem, 239: 933-942, 1964. WILSON, J. E., A. K. PIERCE, R. L. JOHNSON, JR., E. R. WINGA, W. R. HARRELL, G. C. CURRY, AND C. B. MULLINS. Hypoxemia in pulmonary embolism, a clinical study. J. clin Invest. 50: 481-490, 1971. WILSON, J. W. Mast cell morphologic alterations after acute pulmonary injury or respiratory distress. Chest 65, Suppl.: 5-9, 1974. WILSON, J. W., N. B. RATLIFF, AND E. MIKAT. Leukocyte changes in the pulmonary circulation. A mechanism of acute pulmonary injury by various stimuli. Chest 59, Suppl.: 365-395, 1971. WILSON, L. Microtubules as drug receptors: pharmacological properties of microtubule proteins. Ann NY Acad Sci 253: 213-231, 1975. WOOLCOCK, A. J., P. T. MACKLEM, J. C. HOGG, N. J. WILSON, J. A. NADEL, N. R. FRANK, AND J. BRAIN. Effect of vagal stimulation on central and peripheral airways in dogs. J. Appl Physid 26: 806-813, 1969. WOOLVERTON, W. C., AND A. L. HYMAN. The pulmonary hemodynamic effects of lung thromboemboli in dogs. Surgery 73: 572-578, 1973. YAMADA, K. M., AND K. OLDEN. Fibronectins-adhesive glycoproteins of cell surface and blood. Nature London 275: 179-184, 1978. YOUNG, I., R. W. MAZZONE, AND P. D. WAGNER. Identification of functional lung unit in the dog by graded vascular embolization. J. Appl Physid: Respirak Environ Ezerciae Phpid 49: 132-141, 1980. YUEN, T. G. H., AND R. P. SHERWIN. ‘Hyperplasia of type II pneuminocytes and nitrogen dioxide (100 ppm) exposure: a quantition based on electron micrographs. Arch. Environ Health 22: 178-188,197l. ZAPATA-O/RTIZ, V., R. C. DE LA MATA, E. FERNANDEZ, A. GEU, AND L. BATALLA. Bronchial circulation in high altitudes. Am J. Physid 212: 1464-1468, 1967. ZAPOL, W. M., K. KOBAYASHI, M. T. SNIDER, R. GREENE, AND M. B LAVER. Vascular obstruction causes pulmonary hypertension in severe acute respiratory failure. Che8t 71, Suppl.: 306-307, 1977. ZIGMOND, S. H., AND J. G. HIRSCH. Effects of cytochalsin B on polymorphonucIear Ieukocyte Iocomotion, phagacytosis and glycolysis. Exp. CeU R~JX 73: 383-393, 1972. ZURIER, R. B., AND D. M. SAYADOFF. Release of prostaglandins from human polymorphonuclear leukocytes. Iq/kmmution 1: 93-101,1975. ZWEIFACH, B. W. Integrity of vascular endothelium. Circ

38: 89-99,196l.

537. WEST, J. B., AND C. T. DOLLERY. Distribution of blood flow and the pressure-flow relations of the whole lung. J. Appl Physid 20: 175-183,1965. 538. WHAYNE, T. F., JR., AND J. W. SEVERINGHAUS. EXperimental hypoxic pulmonary edema in the rat. J. Appl Physid

1207

MICROEMBOLISM

555

556

557.

558.

559.

560.

561.

562.

Rea 3: 422-424,

Adv. Microcityc

9: 206-225,198O.