Effects of transfusion of emboli and aged plasma on pulmonary capillary permeability

Effects of transfusion of emboli and aged plasma on pulmonary capillary permeability

J THORAC CARDIOVASC SURG 82:358-364, 1981 Effects of transfusion of emboli and aged plasma on pulmonary capillary permeability The effects of trans...

689KB Sizes 0 Downloads 1 Views

J

THORAC CARDIOVASC SURG

82:358-364, 1981

Effects of transfusion of emboli and aged plasma on pulmonary capillary permeability The effects of transfusion of whole blood clot emboli and aged citrated platelet-poor plasma on pulmonary capillary permeability were investigated in anesthetized sheep by continuous collection of pulmonary lymph. Changes in lymph flow and lymph-to-plasma ratios (C)C p) for albumin and globulin were utilized to detect changes in permeability. Infusion of 0.5 cc/kg offinely (-s;1 mm) diced autologous whole blood clot resulted in a 170% increase in lymph flow over control with no change in C)C pfor albumin or globulin. lrfusion of I cc/kg of autologous clot increased lymph flow 180% over control and increased C)C pfor albumin and globulin. Infusion of homologous platelet-poor plasma caused greater increases in lymph flow without changes in C)C p. Changes in each of these three groups were consistent with increased permeability. Balloon occlusion of one main pulmonary artery was induced without a fall in cardiac output and resulted in no change in lymph flow or C/jC pdespite a rise in pulmonary vascular resistance (PVR). Femoral arteriovenous fistulas were created to increase cardiac output, but no change in lymph flow or C)C poccurred. The results in these latter two experiments suggest that increased perfusion per unit lung capillary bed or increased PVR were not primarily responsible for the changes observed in the emboli-treated and plasma-infused animals. Since both emboli and aged platelet-poor plasma increased pulmonary capillary permeability, the permeability increasing factor appears to be humoral in origin. Similar humoral factors may be important in the pathogenesis of the adult respiratory distress syndrome in man.

John E. Mayer, Jr., M.D., MC, USAF, Thomas E. Kersten, M.D., and Edward W. Humphrey, M.D., Ph.D., Minneapolis. Minn.

DeSPite extensive investigation, the pathogenesis of the adult respiratory distress syndrome (ARDS) occurring after shock or nonthoracic trauma remains unsettled.'::" The major pathological finding in patients dying with this syndrome is the presence of increased amounts of protein-rich fluid in the pulmonary interstitial space.' This finding suggests that an increase in pulmonary capillary permeability to protein might have occurred. Multiple pulmonary microernboli,': 5 circulating "toxic" substances (including sepsis),": 8 and neurologically mediated pulmonary venospasm have From the Departments of Surgery, Minneapolis Veterans Administration Medical Center and University of Minnesota, Minneapolis, Minn. Supported partially by Grant HLl8762 from the National Heart and Lung Institute and partially by research funds from the Veterans Administration. Received for publication Oct. 10, 1980. Accepted for publication March 2, 1981. Address for reprints: Edward W. Humphrey, M.D., Veterans Administration Medical Center, 54th St. and 48th Ave. So., Minneapolis, Minn. 54417.

358

been proposed as causes of this fluid accumulation." Massive blood transfusion has been associated with the development of ARDS in man!' 10 and both humoral!' factors and microemboli': 5.12-19 have been proposed as the injurious agent in the transfused blood. The majority of previous studies have not directly assessed the effects of these agents on pulmonary capillary permeability. Rather, these studies have provided data on pulmonary hemodynamics, histologic changes in the lung, and changes in compliance and oxygenating capacity, all of which may indirectly reflect changes in protein concentration and fluid volume of the pulmonary interstitial space. Techniques for direct continuous sampling of pulmonary interstitial fluid are not currently available, but pulmonary lymph can be collected. This lymph is derived directly from interstitial fluid and probably is not altered significantly in its course through the lymphatics." Therefore, measurement of lymph flow and protein concentration can be used to detect changes in capillary permeability," Investigations from this laboratory have assessed the effect of hemorrhagic and endotoxic shock on the effluent

Volume 82 Number 3

from the right thoracic duct in dogs.t?: 21 Staub and associates" have described a similar model in sheep using the efferent lymph from the caudal mediastinal lymph node. They 23 . 24 have provided data suggesting that the lymph obtained is virtually pure pulmonary lymph. Utilizing the sheep model, we have asked the foIlowing questions to determine if blood transfusion could playa role in the development of ARDS: (1) Do whole blood clot emboli cause an increase in pulmonary capillary permeability? (2) Does citrated platelet-poor plasma contain factors which increase pulmonary capillary permeability? To answer these questions, we designed the foIlowing studies.

3 59

Pulmonary capillary permeability

September, 1981

Table I. Group I: Unilateral pulmonary artery occlusion

I

No. Lymph flow (ml/min) PAP (mm Hg) LAP (rnm Hg) Cardiac output (L/min) PVR (mm Hg/L/min) VD/VT (experimentall control)

Baseline 7 0.084 15 7 2.5 3.2 1.0

I

First hour

I

Second hour

7 0.079 23* 7 2.7 5.9* 1.50*

7 0.075 22* 7 2.6 5.8* 1.49*

Legend: PAP. Mean pulmonary artery pressure. LAP, Mean left atrial pressure. PVR. Pulmonary vascular resistance. Ratio of dead space ventilation to total tidal volume. *p < 0.05; t test for paired samples (experimental group versus baseline).

v.rv-,

Methods

Lambs weighing between 25 and 40 kg were used for all experiments, and all experiments were carried out under sterile conditions. Anesthesia was induced with thiopental and, after intubation, maintained with methoxyflurane. Ventilation was by a volume-cycled respirator at 12 cc/kg tidal volume and 2 to 3 em H 20 positive end-expiratory pressure. Respiratory rate was varied to maintain the arterial Pco, between 30 and 40 torr. The technique used for the collection of pulmonary lymph was that described by Staub and associates." Briefly, the technique consists of division of the lower end of the caudal mediastinal lymph node below the pulmonary ligament, a period of 2 to 7 days for recovery, and then cannulation of the efferent lymphatic from the cranial end of the node with a silicone rubber catheter. A catheter was placed directly into the left atrium for measurement of left atrial pressure. An arterial catheter was placed in the carotid artery and a SwanGanz catheter was floated into the pulmonary artery via the jugular vein. After instrumentation all animals were studied supine with the right thorax open. Mean pressures were measured every 15 minutes in the left atrium (LAP), pulmonary artery (PAP), carotid artery, and airway by means of Statham pressure transducers and a Sanborn 964 recorder. Cardiac output was computed by a Lexington COCSD cardiac output computer from indicator-dilution curves of indocyanine green dye. Every 30 minutes, arterial and mixed venous blood samples and expired air were coIlected and analyzed for Po 2 , Pco., and pH. Physiological dead space (V D ) was calculated from the Bohr equation. Alveolar Pco, was assumed to equal arterial Pco, and expiratory tidal volume (V T ) was measured from the setting on the respirator which had been previously calibrated to deliver a known tidal volume. Pulmonary

Table II. Group II: Autologous clot embolus (0.5 cc/kg) Third hour No. Lymph flow (mil min) PAP (mm Hg) LAP (mm Hg) Cardiac output (L/min) PVR (rnm Hg/L/min) V"/V T (experimentall baseline)

4 0.104 17 9 2.6 3.1 1.0

4 0.173* 26t 9 3.5

4 O.I44t 23t 8.5 3.0

4.9 1.51t

4.8 1.59t

3 0.129t 23t 8.5 2.2

6.6

For legend see Table I. *0.05 < P < 0.10. tp < 0.05 (experimental group versus baseline).

vascular resistance (PVR) was calculated from the relation PVR =

PAP - LAP

Cardiac output

Lymph flow was determined by continuously collecting lymph in graduated test tubes. Lymph samples were then centrifuged. No lymph samples were chylous. Arterial plasma samples were drawn every 30 minutes and analyzed for albumin and globulin concentrations by the bromocresol green method 25 and GoldenburgDrewes " method, respectively. Lymph-to-plasma ratios (CJC p ) were calculated from the protein concentrations of lymph and plasma samples coIlected in the same interval. In all animals, a minimum of 2 hours of stable lymph flow rates and hemodynamic measurements were obtained prior to any experimental intervention.

3 60

The Journal of Thoracic and Cardiovascular Surgery

Mayer, Kersten, Humphrey

Table III. Group III: Autologous clot embolus (l cc/kg)

Fourth hour No. Lymph flow (mI/min) PAP (mm Hg) LAP (mm Hg) Cardiac output (L/min) PVR(mm Hg/L/min) VO/V T (experimentall control)

6 0.080

6 0.143*

6 0.146*

5 0.136*

5 0.126*

12 4 2.1

32* 4 2.4

29* 4 2.5

28* 4 2.2

26* 4 1.9

3.8

11.6*

10*

10.9*

11.6*

1.0

2.55*

2.84*

2.77*

2.64*

For legend see Table I. • P < 0.05; t test for paired samples (experimental group versus baseline).

Table IV. Group IV: Homologous plasma infusion Third hour No. Lymph flow (mI/min) PAP (mm Hg) LAP (mm Hg) Cardiac output (L/min) PVR (rnm Hg/L/min)

6 0.116 14 7 2.7 2.6

6 0.178* 24* 10· 4.4* 3.2

6 0.267* 23* 13* 4.5* 2.2

6 0.291* 21* 11* 4.5* 2.2

Forlegend see Table I. .p < 0.05; t test for paired samples.

Whole blood clot was prepared by placing autologous blood into sterile glass tubes. As soon as a firm clot was formed, the test tubes were placed in ice. Prior to injection, the clot was removed from the test tubes and finely diced into particles 1 mm in diameter or less. The diced clot was then mixed with an equal volume of iced saline and injected into the jugular venous catheter. Homologous plasma was prepared by centrifuging whole blood from donor sheep at 350 g for 15 minutes. Platelet counts were measured on five of the freshly prepared plasma units with a Coulter counter. A qualitative assay for fibrin split products was carried out on the same five donor units." Plasma then was stored for 7 to 10 days at 4° C before infusion. Each unit of plasma was cross matched against recipient cells prior to infusion. Five groups of animals were studied. In Group I, after instrumentation and the baseline period, a balloon catheter was positioned fluoroscopically in the left or right pulmonary artery and inflated to obstruct this ves-

sel for the 2 hours during which the lymph studies were conducted. In all aninals the position of the balloon was confirmed at autopsy. In Group II, 0.5 cc/kg of autologous whole blood clot was injected intravenously over 1 to 2 minutes. This volume was chosen because the change in V0 after this dose of emboli approximated that resulting from occluding one pulmonary artery in Group I. Lymph samples were collected and hemodynamic measurements were performed for 3 hours after clot was injected. In Group III, a similar procedure was followed except that 1 cc/kg of autologous clot was used. In Group IV, homologous plasma was infused intravenously at a rate of 500 cc/hr. Lymph collections and hemodynamic measurements were continued for the 2 to 4 hours of infusion of plasma. In Group V, after the baseline periods, bilateral arteriovenous fistulas were created between the femoral artery and vein to increase cardiac output. Studies were continued for 2 to 4 hours after the fistulas were opened. The groin incisions and fistulas were created with the addition of regional anesthesia (procaine HC1) to ensure no change in central nervous system stimulation. All results are expressed in similar fashion. Each animal served as its own control for each variable. The average of values obtained during the last baseline hour was used as the baseline value. All results observed after the experimental intervention were averaged for 1 hour intervals. A t test for paired data was used to compare results after intervention with baseline values.

Results In Group I (Table I), occlusion of either the right or left pulmonary artery resulted in an increase in the ratio of dead space ventilation to total tidal volume (Vol VT) of 150% of the baseline value, but lymph flow fell slightly despite a rise in PAP and no change in cardiac output. PVR increased from 3.2 to 5.9 mm Hg/Limin. CL/Cp for albumin and globulin did not change. The results of Group II animals receiving 0.5 cc/kg of diced autologous blood clot are presented in Table II. Lymph flow increased to 170% of baseline values immediately following embolization. CL/Cp for albumin increased from 0.70 in the baseline period to 0.73 and 0.82 at the second and third hours, respectively, after embolization (p < 0.05). CL/Cp for globulin also increased, but not to statistically significant levels. PAP increased, LAP remained the same, and cardiac output increased slightly but not significantly. PVR increased from 3.1 to 4.8 mm Hg/Limin. Vo increased to 154% of the baseline value of 186 cc. In Group III (Table III), with a larger dose of emboli, the changes were similar to those in Group II except

Volume82 Number 3 September, 1981

that the increases in VD , PAP, and PVR were greater.

Pulmonary capillary permeability

Table V. Group V: Arteriovenous fistula

CL/C p for albumin increased from a baseline of 0.63 to 0.66 at 4 hours (p < 0.05). CfC; for globulin in-

creased from 0.54 to 0.58 at 4 hours (p < 0.05). In Group IV (Table IV), homologous plasma infusions caused lymph flow to increase to 2.5 times baseline levels by the third hour of infusion. There was no change in CL/C p for albumin (0.69 baseline, 0.69 at 3 hours) or globulin (0.57 baseline, 0.57 at 3 hours). Cardiac output increased from 2.7 to 4.5 Llmin, and PAP and LAP increased. In contrast, Group V animals (Table V) showed no increase in lymph flow despite a 40% increase in cardiac output when arteriovenous fistulas were opened. PAP and LAP, PVR, and CL/C p for albumin and globulin did not change significantly. Platelet counts in the five units of plasma averaged 6,500/mm 3 • Qualitative assay for fibrin split products was positive in all five units.

Discussion Many previous reports have suggested a relationship between blood transfusions and ARDS. In 1964, Jenevin and Weiss" suggested that material in transfused blood might damage the lung, and Moseley and Doty'" reported two cases in which massive blood transfusion seemed to be related to pulmonary failure. McNamara and associates!" showed in dogs that the infusion of 3-week-old heparinized blood caused increased PVR. Barrett and associates'? infused 5-day-old heparinized blood and found increases in PVR and shunt fraction. In vitro perfusion of lungs with stored blood caused edema, increased PVR, and decreased compliance in both dogs and baboons in the reports of Veith," Bennett," and their colleagues. Subsequent investigations have attempted to define which blood component is important in producing the changes in pulmonary structure and function seen with massive transfusion. Several studies suggest that the microthrombi which form in stored blood cause changes in pulmonary function.v :": 16, 19, 29,:11 Connell and Swank" found microvascular occlusion and endothelial damage in the lungs of dogs after 25% exchange transfusions. These changes were prevented by Dacron wool filtration which presumably removed emboli. Brown and co-workers" showed that partial exchange transfusion with 5-day-old blood in dogs caused increased shunt fraction and pulmonary edema, but these changes could be prevented by Dacron wool (Swank) filtration of the transfused blood. Berman and associates" studied exchange transfusion in rats and found that infusions of stored blood caused increased pulmonary capillary permeability, After infusing vari-

36 1

Third hour

No. Lymph flow (ml/rn in) PAP (mm Hg) LAP (mm Hg) Cardiac output (Umin) PVR (mm Hg/Umin)

5 0.125 16 8

2.6 3.1

5 0,125 16 8 3.0* 2,7

5 0.132 16 8 3.6* 2.2

5 0.129 17 9 3,6* 2,2

For legend see Table I. • P < 0.05; t test for paired samples.

ous combinations of blood components, these authors concluded that platelets were an important factor in causing permeability changes. Nachman, Weksler, and Ferris 32 have shown that platelets contain a factor which increases capillary permeability, and platelet serotonin has been shown to be important in the hemodynamic changes in pulmonary embolism." However, Brigham and Owen'" found that serotonin infusion did not alter pulmonary capillary permeability in sheep. Wilson and colleagues:" have stressed the importance of leukocytes in pulmonary changes after shock and reinfusion. Others have suggested that fibrin may be the important factor in emboli.l": 18 Malik and van der Zee:!6 found that glass bead emboli caused increases in pulmonary extravascular water and that these changes could be prevented by heparin. However, Binder and Staub:" reported that defibrination or heparinization prior to embolism did not prevent increases in lung lymph flow. Experimental work from other investigators suggests that humoral factors may playa role in the production of pulmonary damage after transfusion. Veith and associates" proposed a humoral mechanism to explain pulmonary damage associated with shock, contact of blood with foreign surfaces, and transfusion. Luterman, Manwaring, and Curreri'" presented evidence in rabbits that fragment D (a fibrin split product) caused pulmonary changes characterized by increased permeability to labeled albumin, interstitial edema, and hypoxemia. Bayley, Clements, and Osbahr? had shown previously that fibrin split products produced pulmonary hypertension, decreased lung compliance, and decreased gasexchanging function in different species. Other studies suggest that the hemodynamic response to pulmonary embolism depends on humoral factors.f": :!9,40 Craddock and co-workersf ': 42 have presented evidence that activated complement is formed when plasma contacts a foreign surface and this activated complement causes increased pulmonary capillary permeability and im-

The Journal of

362 Mayer, Kersten, Humphrey

Thoracic and Cardiovascular Surgery

paired gas exchange in sheep. Others have suggested that prostaglandins are released when emboli reach the lungs.": 44 The results of these earlier studies have suggested that both emboli and humoral factors may playa role in the pulmonary changes associated with massive transfusion. The relative importance of emboli and humoral factors in the production of these changes and the possible interrelationships between emboli and the release of humoral factors remain uncertain. The interpretation of data on lymph flow and lymph to plasma ratios (CL/C p ) for protein is based on the Kedem and Katchalsky " equations governing the transport of solutes (L) and solvent (Jv) across a semipermeable membrane of minimal thickness. Jp = wA li'TT + JvC p (1 - (T)

(I)

J, = LpA (liP - (Tli'TT)

(2)

where Jp is the protein flux, J, is the volume flux, w is the permeability coefficient, A is the membrane area, Cp is the average membrane concentration of protein, (T is the reflection coefficient, L, is the filtration coefficient, 'TT is the osmotic pressure due to protein, and AP is the hydrostatic pressure gradient. Dividing equation I by equation 2 yields equation 320 : C L = (JplJ v ) = (I -

IT)

Cp

+

wAli'TT

LpA(liP - (Tli'TT)

(3)

At steady state, Jpl J, will equal the concentration of protein in the interstitial fluid of the perfused capillary bed. Since lymph is directly derived from interstitial fluid, lymph flow is assumed to be equal to J, and Jp/J v equals the concentration of protein in the lymph (CL ) . Equation 3 predicts that an increase in AP will result in an increase in J, and a fall in CJC p if C p is constant; i.e., as hydrostatic pressure is raised, relatively more water than protein is forced out of the capillary into the interstitium, and the result is a fall in lymph protein concentration. Erdmann and associates," using the sheep model, found that increasing microvascular pressure by inflating a left atrial balloon did cause an increase in J, and a fall in CL/C p as predicted. An increase in wand a fall in (T (increase in permeability) results in increases in both Jp (protein transport) and J, (water transport), so that CL/Cp will either rise or remain the same. Inspection of equation 3 shows that for any given level of Jv , an increase in J, without a fall in C L must be accompanied by an increase in wA or a decrease in (T, i.e., increased permeability to protein. In sheep, infusions of both Pseudomonas aeruginosa and histamine have resulted in increase in J, without a fall in CL/C p , and these changes have been interpreted by Brigham and

co-workers't''<" as defining an increase in permeability. In the current experiments, emboli in Groups II and III caused increases in J, with no change or slight increases in CL/C p • These changes are consistent with an increase in permeability. Pulmonary hypertension did occur in the animals subjected to emboli, but an increase in microvascular pressure alone causes increased J, with a fall in C t/C p •2:1 . 47.48 Since cardiac output did not fall after embolization but perfused capillary area decreased, as evidenced by increases in Vo and PVR, blood flow through the remaining capillary bed must have been increased. This increased flow per unit capillary area at increased pressure could have resulted in pore stretching, as postulated by Shirley and colleagues." In Group I, however, balloon occlusion of the right or left pulmonary artery also produced a V, increase of approximately 50% and an increase in PVR, both similar to those in the animals subjected to emboli in Group II. In the Group I animals, however, no increase in permeability occurred. Therefore, pore stretching probably did not play a role in the permeability changes observed in the animals having emboli. We conclude that the permeability changes in the animals subjected to emboli resulted from some direct effect of the emboli other than from the mechanical occlusion of the pulmonary vasculature. In Group IV, infusions of aged homologous platelet-poor plasma also resulted in significantly increased lymph flow without a fall in CJC p , These changes may result from either an increase in permeability or an increase in perfused area. However, somewhat smaller increases in cardiac output in Group V animals with arteriovenous fistulas were not associated with increased lymph flow. Levels of PVR were similar in these two groups. In addition, since previous studies have shown that two thirds of the pulmonary capillary bed is perfused under basal conditions," a maximal increase in perfused capillary area could account for only a 50% increase in lymph flow. The increased lymph flow in Group IV animals was almost three times that of the baseline period. Although left atrial pressure rose somewhat during infusion, CL/C p did not fall. If the lymph flow changes were due to increased left atrial pressure alone, CL/Cp would fall, as demonstrated by Erdmann and colleagues." In addition, these authors found that lymph flow rose linearly with left atrial pressure. In contrast, in the Group IV animals, lymph flow rose 2.5 times when left atrial pressure was only 1.5 times the baseline value. Therefore, we conclude that the changes in lymph flow in Group IV animals resulted from increased permeability rather than from increased perfused capillary area secondary to

Volume82 Number3 September, 1981

increased cardiac output or from increased microvascular pressure, The findings in the animals subjected to emboli in the present study are remarkably similar to those reported by Ohkuda and associates." They reported increased permeability with a variety of "inert" emboli. However, they also found that occlusion of 70% of the pulmonary vascular bed with balloon catheters increased pulmonary capillary permeability. This finding contrasts with our results showing no permeability change when a smaller fraction of the pulmonary vascular bed was occluded by balloon catheters. They found the best correlations between increases in lymph flow and increases in PVR. These authors offered two explanations for their results: The first was that the emboli caused the release of a humoral substance, which resulted in an increase in permeability in the portions of the pulmonary vascular bed which remained perfused; the second was that embolic obstruction caused increased flow at increased pressure through the areas of the pulmonary vascular bed that remained perfused. This increased physical stress on the endothelial surfaces then could result in increased permeability. Our data suggest that a humoral factor or factors was responsible for the permeability changes observed in our animals. Both emboli and platelet-poor plasma were associated with evidence of increased penneability. Permeability changes did not occur after pulmonary arterial occlusion or creation of arteriovenous fistulas. These findings suggest that within the limits used, the physical effects of increased flow per unit area and increased pressure were not primarily important in causing the permeability changes observed after emboli and plasma infusion. Fibrin split products were found in the aged plasma, as previously described by Honig and associates.f and these split products may playa role in the permeability changes observed. The importance of other factors such as prostaglandins, complement, or vasoactive amines cannot be ruled out.

REFERENCES I Blaisdell FW, Lewis FR: Respiratory Distress Syndrome of Shock and Trauma: Post-traumatic Respiratory Failure, New York, 1977, W. B. Saunders Company 2 Humphrey E, Schwartz M, Northrup W, Murray C: The adult respiratory distress syndrome, Trauma: Clinical and Biological Aspects, SB Day, ed, New York, 1975, Plenum Press, Inc. 3 Staub N: Pulmonary edema. Physiol Rev 54:678, 1974 4 Teplitz C: The core pathobiology and integrated medical science of adult acute respiratory insufficiency. Surg Clin North Am 56:1091,1976

Pulmonary capillary permeability

363

5 Saldeen T: Trends in microvascular research. The microembolism syndrome. Microvasc Res 11:227, 1976 6 Staub N: Pulmonary edema due to increased microvascular permeability to fluid and protein. Circ Res 43: 143, 1978 7 Bayley T, Clements J, Osbahr A: Pulmonary and circulatory effects of fibrinopeptides. Circ Res 21:469, 1967 8 Esrig B, Fulton R: Sepsis, resuscitated hemorrhagic shock, and shock lung. Ann Surg 182:218, 1975 9 Moss G, Staunton C, Stein A: The centrineurogenic etiology of the acute respiratory distress syndromes. Am J Surg 126:37, 1973 10 Moseley R, Doty D: Death associated with multiple pulmonary emboli soon after battle injury. Ann Surg 171:336, 1970 II Veith F, Hagstrom J, Panossian A, Nehlsen S, Wilson J: Pulmonary microcirculatory response to shock, transfusion, and pump-oxygenator procedures. A unified mechanism underlying pulmonary damage. Surgery 64:95, 1968 12 Bennett S, Geelhoed G, Aaron R, Solis R, Hoye R: Pulmonary injury resulting from perfusion with stored bank blood in the baboon and dog. J Surg Res 13:295, 1972 13 Berman I, I1iescu H, Ranson J, Eng K: Pulmonary capillary permeability-a transfusion lesion. J Trauma 16: 471,1976 14 Brown C, Dhurendhar M, Barrett J, Litwin M: Progression and resolution of changes in pulmonary function and structure due to pulmonary microembolism and blood transfusion. Ann Surg 185:92, 1977 15 Bursch C, Lindquist 0, Saldeen T: Respiratory insufficiency in the dog induced by pulmonary microembolism and inhibition of fibrinolysis. Acta Chir Scand 140:255, 1974 16 Connell R, Swank R: Pulmonary fine structure after hemorrhagic shock and transfusion of aging blood, Microcirculatory Approaches to Current Therapeutic Problems, Basel, 1971, S Karger AG, p 49 17 Dawidson I, Barrett J, Miller E, Litwin M: Pulmonary microembolism associated with massive transfusion. Ann Surg 181:51, 1975 18 Lindquist 0, Saldeen T, Sandler H: Pulmonary damage following pulmonary microembolism in the dog. Acta Chir Scand 142:15, 1976 19 McNamara J, Burran E, Larson E, Omiya G, Suehiro G, Yamase H: Effects of debris in stored blood on pulmonary vasculature. Ann Thorac Surg 14:133, 1972 20 Northrup W, Humphrey E: The effect of hemorrhagic shock on pulmonary vascular permeability to plasma proteins. Surgery 83:264, 1978 21 Northrup W, Humphrey E: Pulmonary and systemic capillary permeability to protein following endotoxin. Surg Forum 27:65, 1976 22 Staub N, Bland R, Brigham K, Demling R, Erdmann AJ, Woolverton W: Preparation of chronic lung lymph fistulas in sheep. J Surg Res 19:315, 1975 23 Erdmann J, Vaughan T, Brigham K, Woolverton W,

The Journal of

3 64

Mayer, Kersten, Humphrey

Thoracic and Cardiovascular Surgery

24

25

26 27

28

29

30

31

32

33

34

35

36

Staub N: Effect of increased vascular pressure on lung fluid balance in anesthetized sheep. Circ Res 37:271, 1975 Vreim C, Snashall P, Demling R, Staub N: Lung lymph and free interstitial fluid protein composition in sheep with edema. Am J Physiol 230;1650, 1976 Westgard J, Paquette M: Determination of serum albumin with the SMA 12/60 by a bromocresol green dye binding method. Clin Chern 18:647, 1978 Goldenburg H, Drewes PA: Direct photometric determination of globulin in serun. Clin Chern 17:358, 1971 Latallo Z, Wegrzynowicz Z, Teisseyre E, Kopec M: Simple and rapid evaluation of the intravascular coagulation and fibrinolytic states by application of protamine sulfate and Reptilase R. Scand J Haematol Suppl 13:387, 1971 Jenevin E, Weiss D: Platelet microembolism associated with massive blood transfusion. Am J Pathol 45:313, 1964 Barrett J, Dawidson 1, Dhurandhar H, Miller E, Litwin M: Pulmonary microembolism associated with massive transfusion. Ann Surg 182:56, 1975 Veith FJ, Hagstrom JWC, Nehlsen SL, Karl RC, Deysine M: Functional hemodynamic and anatomic changes in isolated perfused dog lungs. The importance of perfusate characteristics. Ann Surg 165:267, 1967 Giordano J, Zinner M, Hobson R, Gervin A: The effect of microaggregates in stored blood on canine pulmonary vascular resistance. Surgery 80:617, 1976 Nachman R, Weksler B, Ferris B: Characterization of human platelet vascular permeability-enhancing activity. J Clin Invest 51:549, 1972 Ozdemir 1, Webb W, Wax S: Effect of neural and humoral factors on pulmonary hemodynamics and microcirculation in pulmonary embolism. J THORAC CAR0I0V ASC SURG 68:896, 1974 Brigham K, Owen P: Mechanism of serotonin effect on lung trans vascular fluid and protein movement in awake sheep. Circ Res 36:761, 1974 Wilson J, Ratliff N, Young W, Hackel D, Mikat E: Changes in the morphology of leukocytes trapped in pulmonary circulation during hemorrhagic shock, Microcirculatory Approaches to Current Therapeutic Problems, Basel, 1971, S Karger AG, p 41 Malik A, van der Zee H: Mechanism of pulmonary edema induced by microembolization in dogs. Circ Res 42:72, 1978

37 Binder A, Staub N: Increased lung vascular permeability after microemboli does not depend on fibrin deposition. Am Rev Respir Dis 117:311, 1978 38 Luterman A, Manwaring D, Curreri P: The role of fibrinogen degradation products in the pathogenesis of the respiratory distress syndrome. Surgery 82:703, 1977 39 McNamara J, Burran E, Larson E, Omiya G, Suehiro G, Yamase H: Time course of pulmonary vascular response to microembolization. Pulmonary vasculature. Ann Thorac Surg 14:133, 1972 40 Woolverton W, Human A: The pulmonary hemodynamic effects of lung thromboemboli in dogs. Surgery 73:572, 1973 41 Craddock P, Fehr J, Dalmasso AP, Brigham K, Jacob H: Hemodialysis leukopenia. J Clin Invest 59:879, 1977 42 Craddock P, Fehr J, Brigham K, Kronenberg R, Jacob H: Complement and leukocyte mediated pulmonary dysfunction in hemodialysis. N Engl J Med 296:769, 1977 43 Rdegram K, Olson P, Alme B, Granstrom E: Prostaglandin Fz as a a mediator of pulmonary changes during platelet aggregation. Acta Anaesth Scand 21:168, 1977 44 Vaage J, Piper P: The release of prostaglandin-like substances during platelet aggregation and pulmonary microembolism. Acta Physiol Scand 94:8, 1975 45 Kedem 0, Katchalsky A: A physical interpretation of the phenomenological coefficients of membrane permeability. J Gen Physiol 45:143, 1961 46 Brigham K: Effects of histamine on lung transvascular fluid and protein movement in awake sheep. Chest 67:50, 1975 (Suppl) 47 Brigham K, Woolverton W, Blake L, Staub N: Increased sheep lung vascular permeability caused by Pseudomonas bacteremia. J Clin Invest 54:792, 1974 48 Ohkuda K, Nakahara K, Weidner W, Binder A, Staub N: Lung fluid exchange after uneven pulmonary artery obstruction in sheep. Circ Res 43: 152, 1978 49 Shirley H, Wolfram C, Wasserman K, Mayerson H: Capillary permeability to macromolecules. Stretched pore phenomenon. Am J Physiol 190:189, 1957 50 Schwartz M, Murphy W, Nicoloff D, Humphrey E: Factors altering pulmonary capillary perfusion. Surgery 75:750, 1974 51 Honig G, Abildgaard C, Forman E, Gotoff S, Lindley A, Schulman 1: Some properties of the anticoagulant factor of aged pooled plasma. Thromb Diath Hem 22: 151, 1969