Pulmonary microcirculation following full-thickness burns

Pulmonary microcirculation following full-thickness burns

Burns, 6, 227-235 227 Printed in Great Britain Pulmonary microcirculation full-thickness burns following M. Hayashi, T. P. Bond, M. M. Guest, H. ...

2MB Sizes 0 Downloads 0 Views

Recommend Documents

No documents
Burns, 6, 227-235

227

Printed in Great Britain

Pulmonary microcirculation full-thickness burns

following

M. Hayashi, T. P. Bond, M. M. Guest, H. Linares, C. H. Wells and D. L. Larson Divisions of Hematology, Physiology and Pathology, Shriners Burns Institute, and Department of Physiology and Biophysics, The University of Texas Medical Branch, Galveston, Texas

Summary

Direct observation, intermittent photographs and cinerecords of the pulmonary microcirculation in rats were made before, immediately after, and from 24 to 72 h after thermal injury by using a quartz rod transillumination technique, a dark field illumination method, and a combination of both methods. In the control (unburned) animals rapid and continuous blood flow occurred in arterioles, venules and capillaries with no relationship to the phases of the cardiac cycle; 10 to 60 min after thermal injury severe reduction of flow in capillaries was associated wth microaggregates and, in spite of constriction of most arterioles and venules, rapid and continuous blood flow occurred in these vessels; from 24 to 72 h after injury microaggregates were found in some capillaries with resultant stopped flow, many large shunts from arterioles to venules were observed, and rapid and continuous blood flow occurred in slightly dilated arterioles, venules and many capillaries; in histological sections of lung 72 h after thermal injury, evidence was obtained of interstitial oedema and aggregates of red cells. From our observations an increase in pulmonary vascular resistance occurs following a full-thickness

burn of 40 per cent of the body surface. This increase in resistance apparently results from aggregation of formed elements, especially red cells. The increased resistance may be responsible for the opening of shunts and consequent progressive pulmonary insufficiency.

INTRODUCTION numerous investigators (Stone et al., 1967; Zikria et al., 1968; Stone and Martin,

ALTHOUGH

1969; Pruitt et al., 1970; Achauer et al., 1973) have commented on the incidence of pulmonary complications in severely burned patients the precise mechanisms by which the pulmonary disability is produced are uncertain. Smoke inhalation is usually responsible for the primary damage to the pulmonary epithelium (Beal and Conner, 1970; Ambiavagar et al., 1974) but changes which occur in pulmonary rheology appear to be dependent, in part at least, upon blood damaged by heat (Guest and Bond, 1976) and/or substances released from other tissues damaged by heat (Guest and Bond, 1968). Recently, Pruitt et al. (1975) pointed out that severe thermal injury is similar to other forms of trauma with respect to development of progressive pulmonary complications. Arturson (1975) has reported that many patients have no pulmonary symptoms during the first 2 days post burn but on days 3-5 some of them, even in the absence of pulmonary or facial burns, develop a respiratory insufficiency which is frequently progressive. He has postulated that the delayed progressive pulmonary insufficiency following burns results from pulmonary microembolic sequences. Pulmonary microcirculatory abnormality can be inferred from measurements which permit calculation of the alveolar ventilation and capillary flow ratios (VA/Qc). If VA is within the range of normal or is subnormal and the ratio is abnormally large, Qc must have been reduced. Such indirect measurements, however, give no

228

Burns

information about the specific cause of the inadequate capillary flow and fail to identify abnormalities in circumscribed loci. Visualization and photographic recording of the pulmonary microcirculation in living animals, before and after burns to the surface of the body, should give semiquantitative information about the changes in microcirculatory flow and hopefully provide descriptive data which can be used in the interpretation of the causes of the altered rheology. Unfortunately, cinematographic evaluation of the pulmonary microcirculation is technically difficult and few laboratories are equipped for such studies. The earliest microscopic observation of the pulmonary circulation of which we have record was performed on a frog by Malpighi (1687). Hall technique, (1925), using a transillumination observed and described the pulmonary microcirculation in living cats and rabbits. More recent studies (Irwin et al., 1954; Webb et al., 1974; McNary et al., 1973; Knisely et al., 1957) demonstrated that vascular blockage by formed elements is a major cause of pulmonary insufficiency. Veith et al. (1968), following microscopic observation of the pulmonary microcirculation, reported that vasoconstriction occurred in pulmonary arterioles as a concurrent feature of acute haemorrhagic shock. Direct microscopic observation of the pulmonary microcirculation during and following burns to the surface of the body was performed by Knisely and Knisely (1954). They reported that immediately following distant somatic burns momentary pulmonary artery vasoconstriction occurred at the visible ends of these vessels and that the constrictions acted as ‘catch-traps’ for masses of aggregated red blood cells. Our study of the effects of full-thickness burns on the pulmonary microcirculation was designed to observe and semiquantitate contingent changes in blood flow in arterioles, capillaries and venules during and immediately after the infliction of thermal injury to the skin and at 48 and 72 h later. MATERIALS

AND

METHODS

Healthy, young Sprague-Dawley rats, both male and female, within the weight range of 150 to 250 g, were used in this investigation. They were anaesthetized with sodium pentobarbital, 6 mg/ 100 g body weight infused into the peritoneal cavity. A tracheostomy was performed to permit insufhation of the lungs with oxygen and to

Vol. ~/NO. 3

prevent collapse of the lungs following pneumothorax. The animal was placed on its left side. A skin incision was made on the right side and the external intercostal muscles directly over the space between the 4th and 5th, or 5th and 6th ribs were dissected away from the sternocostal margin to near the vertebral column, permitting transsection of the pleura and the spreading apart of the ribs to expose the lungs. Every precaution was taken to prevent bleeding. The lungs and thoracic cavity were kept moist by frequent irrigation with small amounts of O-9 per cent saline solution. A modification of the method described by Irwin et al. (1954) for observing and cinematographically recording events occurring in the pulmonary microcirculation was used. In the Irwin method the lungs are insufflated with oxygen, delivered via a tube which is inserted into the trachea. The outside diameter of the tube is smaller than the internal diameter of the trachea permitting gas to escape from the lungs. The diameter of the oxygen delivery tube relative to the inside diameterofthe trachea, together with the rate of oxygen delivery, determines the degree of pulmonary inflation. The anaesthetized animal has no movements of its respiratory muscles if sufficient flow of oxygen is maintained. However, survival was improved if the animals were allowed to make small respiratory movements at a rate of not more than 15 per minute. Irwin’s method is effective in preventing ventilatory movements but it does not eliminate movements of the lungs which are transmitted to them by cardiac contractions and by inertial forces imparted by movement of blood in pulmonary arteries during ventricular systole. To eliminate movement transmitted into the segment of the lung under observation, the lung was fastened to a cover glass with tissue glue (Eastman Kodak-910). The glue was spread on the cover glass in a thin film on two separated areas leaving the centre of the cover glass free from glue. The cover glass was gently brought into contact with the lower surface of the lung by means of a micromanipulator. The cover glass was then clamped with the lung surface on the horizontal plane. When the position of the lung had been fixed, the tip of a quartz rod was adjusted below the cover glass (Fig. 1). In the current investigation the caudal edge of the anterior lobe of the right lung was observed and photographed. In all experiments the lung was transilluminated, using the quartz rod technique for transferring and directing the light beam as described by Knisely (1936). A dark field

229

Hayashi et al. : Pulmonary Microcirculation

All animals in group 3 received the same type and degree of burn as that inflicted on the group 2 animals. Rats were housed in cages with water and dry food. No intravenous fluids were administered. Only 2 deaths occurred in the period between the thermal injury, which was performed under pentobarbital anaesthesia, and readminisrration of the anaesthetic prior to surgery in preparation for the observation of the pulmonary microcirculation. No observations were made on animals in which there was evidence of infection or on moribund animals. When the group 3 rats were terminated after the observations of the microcirculation had been completed, pulmonary tissue was fixed and prepared for histological examination. RESULTS Fig. 1. Sketch

transmittance

showing relationship between light and the thin edge of a lobe of the lung.

technique (Sherman et al., 1971) was also used in many experiments. Observations were made with a Leitz monocular microscope at magnifications of 88 to 165 times. Blood flow during and following each experimental procedure was recorded on film. For recording at 24 frames per second a Doiflex 16 mm camera was employed. When high speed photography was undertaken a Mitchell Monitor 600 camera was used. The experimental animals were divided into three groups. Group 1 consisted of 10 rats (controls) which were not burned. Their pulmonary microcirculations were observed and cinerecorded for periods of up to 3 h after the lung had been exposed. Group 2 consisted of 30 rats in which acute changes in pulmonary circulation were evaluated after a third-degree burn over 40 per cent of the body surface was produced with a heated electric soldering iron. Observations were made prior to burning the surface of the body and for one hour after the injury had been inflicted. Only animals with a normal pulmonary circulation prior to burning were included in the acute burn study. In approximately one-third of group 2 animals, the vessels of the pulmonary circulation which were examined and followed were arterioles and venules, while in the remainder attention was directed to the capillary circulation. Group 3 consisted of 32 rats in which the pulmonary circulation was studied: (a) 24 h after injury (10 rats); (6) 48 h after injury(lOrats);and (c) 72 h after injury (12 rats).

illumination

Group 1 (unburned controls) During the 3-h observation period the pattern of blood flow did not change. Velocity of blood in arterioles, capillaries and venules remained essentially constant throughout the observation period. Flow in arterioles was rapid. No changes in velocity related to the phases of the cardiac cycle occurred. Diameters of arterioles remained constant. Fig. 3a is a frame from the cinefilm showing the typical appearance of the microcirculation of the lung in an unburned animal. Capillaries were observed on the alveolar surfaces and between septa. The capillaries anastomosed with other capillaries forming networks on the surface of alveoli (Fig. 2~). Blood flow in capillaries was rapid and nonpulsatile. Filming at 24 frames per second and projection of the film at 24 frames per second usually did not permit identification of individual erythrocytes. Several long capillaries could be traced, extending over one or two alveoli. Short capillaries were also common. Typically, several capillaries (2 to 5) emptied into a venule. Venules anastomosed with other venules. The larger venules disappeared into deeper tissues. Flow in venules was rapid, smooth and nonpulsatile. No changes in diameter of capillaries and venules were observed during the 3-h period. Group 2 Prior to thermal damage to the skin the appearance of the pulmonary microcirculation was the same as in rats in group 1. The time required to inflict thermal injury was about 5 min and during this period no changes in pulmonary microcirculation were noted (Fig. 2a).

Fig. 2. a, Before thermal injury: rapid flow observed in all capillaries. Capillary networks can readily be identified on alveolar surfaces (magnification approx. x 141). b, 2.5 min after thermal injury: aggregates of red cells observed in capillaries in which flow had stopped (magnification approx. x 141).

a

b

Fig. 3. a, Before thermal injury: typical lung structure; capillaries seen on surface of alveoli; large arteriole originates in deep tissue and extends to surface where it divides into small arterioles (magnification approx. x 75). b, 30 min after thermal injury: the large arteriole has constricted; compare with diameter of arteriole prior to thermal injury (magnification approx. x 75).

Within 10 min of the injury flow was absent in some capillaries. The number of inactive capillaries gradually increased over the 60-min observation period. In Fig. 2b reduction in capillary flow is apparent. Red cell aggregates were seen in inactive capillaries, but no significant increases in aggregation over that during the control period were observed in arterioles and venules. In the period from 10 to 60 min following infliction of thermal damage to the skin, decreases in diameter of most arterioles and venules occurred (Fig. 3) but a change in the velocity of flow in these vessels was not detected. No indication of opening of shunts was observed during the immediate post burn period.

Group 3 In group 3a (10 rats, 24 h after burning) 2 of 10 rats had pulmonary microcirculatory flow which could not be distinguished from that observed in group 1 (control rats) or from preburn observations in group 2 rats. The other 8 rats all showed some degree of reduced capillary flow. Red cell aggregates were present in capillaries without ffow and in some cases these static capillaries appeared to be filled with clotted blood. Arterioles and venules appeared to have dilated; flow in these vessels was rapid. Large shunts from arterioles to venules were noted in I of the 8 rats with reduced or stopped capillary flow (Fig. 4).

Fig. 4. 24 h after thermal injury: a large shunt from an arteriole to a venule is present (magnification approx. Y 70).

Fig. 5. 48 h after thermal injury: small, dilated vessels have anastomosed with other small, dilated vessels (magnification approx. X 141).

Fig. 6. 72 h after thermal injury: microaggregates seen in capillaries (magnification approx. X 141).

Fig. 7. 72 h after thermal injury: large shunts and many anastomoses of dilated vessels are characteristic feature (magnification approx. x 75).

In group 3b (10 rats, 48 h after burning) capillary stasis was observed in 7 rats. Aggregates of red cells and apparent clots were noted in capillaries without flow. Arterioles and venules were dilated and flow was rapid. Many anastomosed, dilated small vessels were observed (Fig. 5). In 6 of the 7 rats with abnormal microcirculatory flow, large shunts were observed. In the 3 rats in which capillary stasis was not observed, arterioles, capillaries and venules appeared to be dilated. In one of these rats a large shunt was found. In group 3c (12 rats, 72 h after burning) the flow in all capillaries was abnormally slow and capillary stasis was common in 9 of theserats. Red cell aggregates and apparent fibrin clots were observed in non-flowing capillaries (Fig. 6). Arteriolar flow was rapid through dilated and frequently anastomosed vessels. Large shunts were present in 6 of the 9 rats with abnormal capillary flow (Fig. 7). Two of the 6 rats had well

haemorrhages in the lung but it was not possible to identify haemorrhaging vessels. Three rats in which slowed flow or capillary stasis was not observed had relatively rapid flow through dilated arterioles and venules. In one of these rats a large shunt was found. delineated

Histological

findings

Pulmonary tissue of 4 rats from group 3 was examined histologically (72 h after burning). All lungs examined showed intense congestion and red cell aggregation. The sections also showed evidence of interstitial oedema and some degree of small blood vessel dilation. DISCUSSION Technical Aspects All techniques devised to date for direct observation and photographic recording of the pulmonary

232 microcirculation either interfere with the mechanics of ventilation or are limited in value because of inadequate magnification and resolution. If pulmonary inflation and deflation are permitted, the portion of the lung being observed is continually changing position with reference to the microscope objective and, in consequence, precision focusing is difficult or impossible. McNary et al. (1973) have emphasized the traumatic effects and abnormalities introduced by producing an open pneumothorax. Terry (1939), Krahl(l963) and Klausner et al. (1971) eliminated the open pneumothorax by viewing the lung through a previously implanted thoracic window. After a number of preliminary experiments in which several different techniques were tested we elected to carry out the current study by utilizing a modification of the technique of Irwin et al. (1954). In addition to diffusion ventilation we used a tissue glue to help stabilize the field for observation. Although tissue in contact with the glue was damaged, the tissue observed was kept completely free of the glue. Admittedly these procedures subjected the lung and its circulation to non-physiologic conditions; however, a stable field was insured and better lighting, greater magnification and higher resolution can be attained than with techniques employing the thoracic window. For observation and photography of flow in microvessels of the lung and for estimating rare of flow we have found that transillumination, using the quartz rod technique described by Knisely (1936) for transmitting the light to the object, is more effective than other methods. On the other hand, the dark field illumination technique, described by Sherman et al. (1971), gives better images of structural relationships within the lung. Physiologic flow in pulmonary microvessels With respect to pulsating flow v. steady flow in the pulmonary microcirculation, we have observed a relatively smooth and continuous flow in control experiments (Fig. 2a). Furthermore, blood flow in the pulmonary microcirculation after thermal injury to skin was also non-oscillatory. In contrast, Irwin et al. (1954) reported that, in the rabbit, oscillations and intermittencies in blood flow within small pulmonary vessels were characteristic features and that they observed contractions and dilations of arterioles and venules. However, in studies on cats and rabbits, reported by Hall (1925), oscillations in flow in pulmonary microvessels and contractions or dilations of these vessels were not observed.

Burns Vol. ~/NO. 3 Schlosser et al. (1965), using a Starling pump to maintain an artificial ventilatory cycle, measured the rate of transit of erythrocytes through pulmonary capillaries and demonstrated that during inspiration blood velocity increased. Sherman et al. (1971) reported that pulmonary capillary oscillations are related to both the phases of the cardiac cycle and the ventilatory cycle. McNary et al. (1973) have stated, based upon studies in the rat, that volume flow and velocity in pulmonary capillaries are dependent upon the phases of the ventilatory cycle; however, volume flow and velocity in arterioles and venules are largely independent of the phase of the respiratory cycle, but are influenced by the relative length of the cardiac cycle. They concluded, on the basis of marked reduction in capillary flow on the surface of deflated alveoli when compared with better flow during maintenance of an end-expiratory pressure greater than atmospheric, that positive pressure breathing should be beneficial in maintaining pulmonary capillary flow during pathological states in which pulmonary circulation is compromised. From our observations, the flow rate under control conditions in the pulmonary capillaries of rats is relatively rapid and non-oscillatory. The only conditions under which we observed oscillatory flow in capillaries, arterioles and venules was when the heart rate was abnormally slow (overdose of anaesthesia and preceeding death) or following unnecessary surgical trauma. Flow in pulmonary microvessels after thermal injury Two abstracts from the Knisely laboratory (Knisely and Knisely, 1954; Knisely, 1955) are the only reports we have found of studies in which the pulmonary microcirculation has been directly observed and photographed during and following thermal injury to the skin. Using the Irwin technique in cats, dogs and rabbits, Knisely and Knisely (1954) reported that at the moment of burning constriction occurred in pulmonary arteries, apparently at or near their tips. Red blood cell aggregates entered the artery tips about 30 s after the burn was inflicted and were trapped. These aggregates slowly broke up. The Kniselys termed the artery tips, ‘catch-traps’, with the apparent function of sequestering cell aggregates until they disintegrated. No mention was made of aggregates in pulmonary capillaries. In contrast to the Knisely study, our observations of the response of the pulmonary vasculature to thermal injury of the skin have extended over a

Hayashi

et al.

: Pulmonary

Microcirculation

longer period, i.e. 72 h and we have not critically evaluated events occurring at the onset of burning. The burn we have inflicted is more extensive than that induced by the Kniselys and it has taken longer to produce. The early changes which they described may also occur in the rat but such changes probably have little effect on pulmonary function hours to days following the thermal injury. In our current experimental study we have not attempted to determine the history or precise composition of the microaggregates. Red cell aggregates can be generated during the slow intravascular release of thrombin (Guest and Bond, 1968). Arturson and Rammer (1974) and Arturson (1975) have suggested that the cause of most of the symptoms in post-burn pulmonary insufficiency is the development of disseminated intravascular coagulation (DIC), which they believe develops because of the endogenous inhibition of fibrinolysis. Meyers et al. (1957) and Olow (1963) have reported that fibrinolytic inhibition occurs at about 24 h after the initial traumatic episode. Microaggregates have been found in the systemic circulation during and following thermal injury to the skin (Bond and Guest, 1976). It is uncertain whether the red cell aggregates are formed in the systemic veins and transported to the lungs or whether they form de now in pulmonary vessels. A reduction in the number of patent capillaries secondary to blockage by cell aggregates, as observed in our studies, must cause an increase in pulmonary vascular resistance. Although we were unable to find a report of a measured increase in pulmonary vascular resistanceafterthermal injury to the skin, an increased resistance has been found during haemorrhagic shock (Eaton, 1947). Eaton suspected that the increased resistance resulted from pulmonary vessel obstruction. Other investigators have also reported an increase in pulmonary vascular resistance during haemorrhagic shock (Sealy et al., 1966; Henry etal., 1967; Cook and Webb, 1968; Veith et al., 1968) and during endotoxin shock (Keller et al., 1967; Harrison et al., 1969). We observed that the most commonly ‘stopped capillaries were longer capillaries. Flow usually continued in short capillaries. Sherman et al. (1971) also observed that, under conditions of reduced flow, longer pulmonary capillaries were more vulnerable to conditions which inhibited flow. These authors postulated that, with blockage of longer capillaries, short capillaries dilate and form arteriovenous shunts. Their postulate is

233

consistent with our observations that relatively large shunts appeared during the period between 24 and 72 h after thermal injury. Shunts, presumably, are opened because pulmonary capillary resistance increases to a critical level thereby augmenting the pressure and flow in patent vessels. During the period between 24 and 72 h after the burn was inflicted we observed a considerable number of large shunts (Figs. 4, 7). Lucas et al. (1968) have reported that 5 to 8 per cent of the pulmonary blood is shunted in normal human adults when the measurement of shunting is calculated from Berggren’s formula (1942) and that the percentage is increased to 40 or 50 following severe injury and during shock. They state that more than 50 per cent shunting is lethal. Achauer et al. (1973) have indicated that shallow tidal breathing after severe thermal injury results in gradual alveolar collapse, decreased compliance and increased shunting; when the shunting exceeds 25 per cent, tachypnoea and cyanosis occur. Whatever the cause of the shunting, the increased resistance to flow during its development may be involved in an augmented capillary permeability and consequent interstitial oedema. Histological sections of the lungs 72 h after the burn was inflicted gave evidence of interstitial oedema Furthermore, within 1 h of thermal injury we were able to observe the escape of fluorescein-tagged albumin into the interstitial space; no fluorescein could be detected prior to injury: the fluorescein-tagged albumin was given via an external jugular vein. Fluid resuscitation was not attempted in our thermally injured rats. Even without fluid administration by vein, many pulmonary capillaries were dilated in the period 24 to 72 h after thermal injury (Fig. 5) and evidence of pulmonary oedema was obtained. Apparently most of our experimental subjects were able to maintain a relatively normal blood volume. All rats had continuous access to drinking water. As pointed out by Cooper et al. (1967) oedema (seepage of plasma into alveoli) inhibits pulmonary surfactant resulting in reduced pulmonary compliance and increased physiologic dead space. Oedema fluid increases the alveolar diffusion distance and thereby promotes an oxygen diffusion dysfunction, reduction in haemoglobin saturation and lactacidaemia. The oedema-related malfunctions are additive to those directly engendered by pulmonary circulatory obstructions and opening of arteriovenous shunts. This investigation has not supplied a new and unique explanation for the delayed development

234

of a non-infectious pulmonary insufficiency in severely burned individuals, but it has brought to a reasonable focus the aetiological sequences which have been previously postulated and marginally demonstrated. We have directly observed in an experimental animal model the presence in pulmonary capillaries of cellular aggregates and stopped flow. Arteriovenous shunts develop during the period between 24 and 72 h after infliction of thermal injury to the skin; shunts are associated with blockage of capillaries and presumed resultant increase in pressure in other capillaries. Concurrent with or immediately sequential to the blockage of capillaries, interstitial pulmonary oedema develops. The oedema potentiates the inability of the damaged lung to perform an adequate respiratory function. REFERENCES

Achauer B. M., Allyn P. A., Furnas D. W. et al. (1973) Pulmonary complications of burns: the major threat to the burn patient. Ann. Surg. 177, 311. Ambiavagar M., Chalon J. and Zargham I. (1974) Tracheobronchial cytologic changes following lower airway thermal injury. J. Trauma 14, 280. Arturson G. (1975) Respiratory responses to burn injury. Burns 1, 254. Arturson G. and Rammer L. (1974) Endogenous inhibition of fibrinolysis in patients with severe burns. Acta Chir. Stand. 140, 181. Beal D. D. and Conner G. H. (1970) Respiratory tract injury: a guide to management following thermal and smoke injury. Laryngoscope SO,25 Bond T. P. and Guest M. M. (1976) Unpublished data. Cook W. A. and Webb W. R. (1968) Pulmonary changes in hemorrhagic shock. Surgery 64,85. Cooper N., Matsuura Y., Murner E. S. et al. (1967) Analysis of pathologic mechanisms in pulmonary surfactant destruction following thermal burn. Am. Surg. 33, 882. Eaton R. M. (1947) Pulmonary edema. Experimental observations on dogs following acute peripheral blood loss. J. Thorac. Surg. 16, 668. Guest M. M. and Bond T. P. (1968) Release of thromboplastin after thermal injury. Ann. NY Acad. Sci. 150, 528. Guest M. M. and Bond T. P. (1976) Effect of heatdamaged blood on vascular endothelium. Am. Burn Assoc. Eighth Annu. Meeting, p. 38. Hall H. A. (1925) A study of the pulmonary circulation by the trans-illumination method. Am. J. Physiol. 72, 446. Harrison L. H., Beller J. J., Hinshaw L. B. et al. (1969) Effects of endotoxin on pulmonary capillary permeability, ultrastructure and surfactant. Surg. Gynecol. Obstet. 129, 723. Henry J. N., McArdle A. H., Scott H. J. et al. (1967) A study of the acute and chronic respiratory pathophysiology of hemorrhagic shock. J. Thorac. Cardiovasc. Surg. 54, 666.

Burns Vol. ~/NO. 3

Irwin J. W., Burrage W. S., Aimar C. E. et al. (1954) Microscopical observations of the pulmonary arterioles, capillaries and venules of living guinea pigs and rabbits. Anat. Rec. 119, 391. Keller C. A., Schramel R. J., Hyman A. L. et al. (1967) The cause of acute congestive lesions of the lung. J. Thorac. Cardiovasc. Surg. 53, 743. Klausner S. K., Sherman H. and Cook W. A. (1971) A thoracic window for incident light microscopy. Vast. Surg. 5, 215. Knisely M. H. (1936) A method of illuminating living structures for microscopic study. Anat. Rec. 64, 499. Knisely W. H. (1955) Illustrations of the architecture of visible pulmonary blood vessels and responses of these vessels to experimental stimuli. Anut. Rec. 121, 438.

Knisely W. H. and Knisely M. H. (1954) Preliminary observations of the catch-trap architecture of pulmonary artery tips and their responses following distant somatic burns. Anat. Rec. 118, 320. Knisely W. H., Wallace J. M., Mahaley M. S. et al. (1957) Evidence, including in vivo observations, suggesting mechanical blockage rather than reflex vasospasm as the cause of death in pulmonary embolization. Am. Heart J. 54, 483. Krahl V. E. (1963) A method of studying the living lung in the closed thorax, and some preliminary observations. Angiology 14, 149. Lucas C. E., Ross M. and Wilson R. F. (1968) Physiologic shunting in the lungs in shock or trauma. Surg. Forum 19, 35. McNary W. F., Berman H. H., Daly B. et al. (1973) The use of videotape to record and review data obtained by incident light darkfield microscopy of the living lung. Bibl. Anat. 11, 44. Malpighi M. (1687) De Pulmonibus. Epistola II. (Praeclarissimo et eruditissiomo Viro 10, Alphonso Borellio) In: Opera Omnia, Lugduni Batatvorum 2, 327. Meyers W. M., Burdon K. L. and Riley M. N. (1957) Proteolytic activity of whole human serum following surgery. J. Lab. Clin. Med. 49, 377. Olow B. (1963) Effects of streptokinase on postoperative changes in some coagulation factors and the fibrinolytic system. Acta Chir. Stand. 126, 197. Pruitt B. A., DiVincenti F. C., Mason A. D. et al (1970) The occurrence and significance of pneumonia and other pulmonary complications in burned patients: comparison of conventional and topical treatments. J. Trauma 10, 519. Pruitt B. A., Erickson D. R. and Morris A. (1975) Progressive pulmonary insufficiency and other pulmonary complications of thermal injury. J. Trauma 15, 369. Schlosser D., Hyse E. and Bartels H. (1965) Flow rate of erythrocytes in the capillaries of the lung. J. Appl. Physiol. 20, 110. Sealy W. C., Ogino S., Lesage A. M. et al. (1966) Functional and structural changes in the lung in hemorrhagic shock. Surg. Gynecol. Obstet. 122, 754.

Hayashi

et al.

: Pulmonary

Sherman H., Klausner S. and Cook W. A. (1971) Incident dark-field illumination: a new method for microcirculatory study. Angiology 22, 295. Stone H. H. and Martin J. D. (1969) Pulmonary injury associated with thermal burns. Surg. Gynecol. Obstet. 129, 1242. Stone H. H., Rhame D. W., Corbett J. D. et al. (1967) Respiratory burns: a correlation of clinical and laboratory results. Ann. Surg. 165, 157. Terry R. J. (1939) A thoracic window for observation of the lung in a living animal. Science 90, 43. Veith F. J., ffagstrom J. W. C., Panossian A. et al. (I 968) Pulmonary microcirculatory response to

Requests for reprintsshould Biophysics. The University

235

Microcirculation

be addressed lo: of Texas Medical

shock, transfusion and pump-oxygenator procedures: a unified mechanism underlying pulmonary damage. Surgery 64, 95. Webb W. R., Kamiyama T. M., Wax S. 0. et al. (1974) Microscopic studies of the pulmonary circulation in situ. Surg. Clin. North Am. 54, 1067. Zikria B. A., Sturner W. Q., Astar Jian N. K. et al. (1968) Respiratory tract damage in burns: pathophysiology and therapy. Ann. NY Acad. Sci. 150, 618.

Paper accepted 12 August 1978.

Mr M. M. Guest, Ashbel Smith Professor Branch, Galveston, Texas, USA.

of Physiology,

Department

of Physiology

and