by Increased Membrane Permeability

Transcription

1 Effect of Positive End-Expiratory Pressure on Lung Water in Pulmonary Edema Caused by ncreased Membrane Permeability Claudia Helbert, M.D., ndrew Paskanik, and Carl E. Bredenberg, M.D. BSTRCT Pulmonary edema caused by increased membrane permeability was created in dogs by alloxan and infusion of saline solution. Pulmonary extravascular water volume was measured gravimetrically using the supernatant hemoglobin concentration to estimate red cell mass in the calculation of residual pulmonary blood volume. Three groups were studied for two hours: a control group, a group given alloxan and mechanical ventilation without positive end-expiratory pressure (PEEP), and a group given alloxan and mechanical ventilation with 10 cm H20 of PEEP. fter two hours, alloxan caused moderately severe pulmonary edema in the two experimental groups, but PEEP had no effect on the accumulation of pulmonary extravascular water volume. No sustained differences in pulmonary or systemic hemodynamics were present throughout two hours of pulmonary edema. The pulmonary shunt was increased in the group without PEEP but was similar in the control group and the group with PEEP. No significant changes in alveolar dead space were noted among the three groups. Mechanical ventilation with positive endexpiratory pressure (PEEP) has been an effective treatment for patients with pulmonary edema resulting from congestive heart failure or from a variety of noncardiac causes frequently categorized as acute respiratory distress syndromes. t has been suggested that the observed clinical improvement is in part the consequence of the increased alveolar pressure being transmitted to the perivascular tissue, where it is presumed to drive fluid out of the pulmonary interstitium back into the vascular compartment [ 1- From the Department of nesthesia and Division of Cardiopulmonary Surgery, State University of New York, Upstate Medical Center, Syracuse, NY. ccepted for publication Sept 9, ddress reprint requests to Dr. Bredenberg, Department of Surgery, Upstate Medical Center, 750 E dams St, Syracuse, NY There has been little confirmation of this suggestion in experimental studies in which the mechanism for pulmonary edema is isolated from the many complicating factors that inevitably accompany observations of patients. Pulmonary edema caused by increased hydrostatic pressure in the pulmonary microvasculature has been we1 studied experimentally, and the consensus is that PEEP does not reduce lung water either in the intact animal or in the isolated lobe [4-91. Since, in the so-called acute respiratory distress syndromes, increased lung water is often present without a substantial increase in the hydrostatic pressure of the pulmonary microvasculature,. attention has focused on alterations in microva xular membrane permeability as a cause of pulmonary edema [lo, 111. The purpose of the prlesent experiments was to determine whether PEEP added to mechanical ventilation prior to lung injury can retard the accumulation of lung water in pulmonary edema caused by vascular membrane permeability in which the rnicrovascular hydrostatic pressure was not elevated. dog model was used in which pulmonary edema was induced with alloxan and a moderate infusion of normal saline solution. Material and Methods n 26 mongrel dogs weighing between 15 and 20 kg, general anesthesia was induced with pentobarbital, 25 mm per kilogram of body weight, administered intravenously. nimals were intubated and ventilated with room air using a piston ventilator' with, a tidal volume of 15 cc/kg and a rate of 15 breaths per minute. femoral artery and vein were cannulated by cutdown to measure systemic arterial pressure, obtain blood samples, and administer fluids. flowdirected balloon-tipped 7F catheter with ther- *Model 607, Harvard nstrument Company, Millis, M 42

2 43 Helbert, Paskanik, and Bredenberg: PEEP and Pulmonary Edema mistor tip was introduced through the left external jugular vein and passed into the pulmonary artery to measure pulmonary artery pressure and cardiac output and to collect mixed venous blood samples. similar catheter was advanced retrograde from the femoral artery into the left ventricle; its position was confirmed by obtaining a characteristic left ventricular pressure tracing. Electrocardiogram limb leads were placed, and the electrocardiographic tracing and left ventricular pressure tracing were displayed on the same recording paper running at 100 mm per second. Left ventricular end-diastolic pressure (LVEDP) was measured using the peak of the R wave of the ECG to determine the timing of end-diastole [12]. irway pressure was measured from the catheter connected to a sidearm of the endotracheal tube. irway and intravascular pressures were measured at endexpiration with Hewlett-Packard 267 BC transducers and recorded on a Hewlett-Packard recorder. Cardiac output was measured by thermodilution using the thermistor of the pulmonary artery catheter and a KM Model 3500 cardiac output computer. Measurements of cardiac output were performed in triplicate, and the average value was used. The ratio of dead space to tidal volume was calculated using partial pressure of arterial carbon dioxide (PaC02) and mixed expired partial pressure of carbon dioxide [13]. Expired partial pressure of carbon dioxide was measured from a mixing chamber attached to the expiratory port of the respirator using a Godart capnograph." rterial and mixed venous blood gases were measured on freshly drawn heparinized blood using conventional electrodes. Hemoglobin concentration was measured by standard techniques, and rectal temperature was maintained at 37" k 1 C by heating pad and electric lamps. Pulmonary shunt was measured after 15 minutes of ventilation with 100% oxygen. Oxygen contents were calculated from the partial pressure of arterial oxygen, partial pressure of mixed venous oxygen, and hemoglobin concentration [13]. Static compliance was measured by disconnecting the ventilator and measuring the "Type 110; Godart, De Bilt, The Netherlands. change in airway pressure after injecting 500 cc and 1,000 cc of room air using a calibrated syringe. Colloid oncotic pressure was measured using a Wescor colloid osmoter. Statistical comparison was by analysis of variance using the F table to test the hypothesis of homogeneity. Results are reported as mean k standard error of the mean. Protocol ll animals received 15 ml per kilogram of normal saline solution during instrumentation. Following induction of anesthesia, all animals had an arterial base deficit which was corrected with sodium bicarbonate prior to control measurements. Minor adjustments in the respirator rate were made to obtain PaC02 between 35 and 44 mm Hg, and subsequent minor rate changes were made at hourly intervals to maintain PaC02 within this range. fter it was ensured that hemodynamic and physiological stability were maintained for 30 minutes, control measurements were made and dogs were assigned to one of three groups using a random-number table. n Group 1 (control), 6 dogs were mechanically ventilated without PEEP for two hours and did not receive alloxan or additional intravenous fluids. Measurements were recorded at 15 minutes, one hour, and two hours. n Group 2 (alloxan), 10 dogs received 85 mm per kilogram of alloxan* in 10 ml of saline solution injected intravenously followed by the administration of 40 ml per kilogram of normal saline solution infused over 20 minutes. These animals were ventilated without PEEP. Measurements were made immediately following alloxan administration (during infusion of saline solution) and one and two hours following administration of alloxan. n Group 3 (alloxan with PEEP), 10 dogs were given 10 cm H20 of PEEP after control measurements using an Emerson 3 V-WR valve previously calibrated against a 10 cm column of water. Measurements were repeated after the application of PEEP, following which alloxan and saline solution were given as in Group 2. Measurements were repeated immediately fol- 'Eastman Kodak Company, Rochester, NY

3 44 The nnals of Thoracic Surgery Vol 36 No 1 July 1983 lowing the administration of alloxan (during infusion of saline solution) and at one and two hours after alloxan administration. ll animals were made to sigh with a breath three times tidal volume 15 minutes before control measurements and at 30-minute intervals subsequently. The sighs occurred 5 minutes before blood gas measurements. fter two hours of observation, the animals were killed with an overdose of intravenously administered pentobarbital (Beuthansasia Special). The chest was opened immediately after death. The hili were promptly clamped, and the lungs were removed and passively drained of blood. Residual hilar structures were excised, and both lungs were weighed in their entirety and processed for measurements of pulmonary extravascular water volume using the technique of Pearce and colleagues [14] with minor modifications [7]. Use of supernatant hemoglobin concentration with this method to estimate residual blood volume is accurate if there is minimal hemoglobin in the edema fluid [15]. The extravascular edema fluid in alloxaninduced pulmonary edema is not hemorrhagic [16, 171 in contrast to oleic acid-induced pulmonary edema. Results plus infusion of saline solution (Groups 2 and 3) produced moderately severe pulmonary edema. n both these groups, the lungs appeared grossly similar at thoracotomy. Two of the dogs manifested a gross white foam in the airways. n the remaining 18 dogs, white foam and clear lung fluid were apparent only on the cut surface after sectioning of the lobes. No hemorrhagic pulmonary edema fluid was seen grossly. The mean pulmonary extravascular water volume of both alloxan groups differed significantly from the control value ( p < 0.005), but there was no difference in this variable between Groups 2 and 3, regardless of whether PEEP was used (Fig 1). mmediately following injection of alloxan, there was a rapid spike in mean pulmonary artery pressures in Groups 2 and 3. t peaked at 2 minutes and returned to a value within normal range 10 minutes following the injection. Pulmonary artery pressures remained relatively *.O 0 f.v c L E 0 a, a 3.0 0, E n - n =6 n=o n=o without with PEEP PEEP Fig 1. Pulmonary extravascular water volume for the three groups. Data are shown as the mean 5 standard error of the mean. Both alloxan groiqs differ from the control group (p < 0.005), but do not differ significantly between each other. constant throughout the remainder of the two hours (Fig 2). n Group 3 (alloxan plus PEEP), mean pulmonary artery pressure (measured relative to atmosphere) was raised 3 to 5 mm Hg by 10 cm H20 of PEEP. The same difference in pulmonary artery pressure (relative to atmosphere) between the two alloxan groups was maintained throughout the two hours of observation. The only significant change in LVEDP in any group was an increase of 2 to 3 mm Hg (relative to atmosphere) caused by the addition of 10 cm H20 of PEEP (see Fig 2). Cardiac output fell with the addition of PEEP, but was increased in both experimental groups by injection of alloxan and infusion of saline solution. There was no significant difference in cardiac output among the three groups following the administration of alloxan and saline solutions (see Fig 2). There was no statistically significant change in systemic arterial pressure in any of the groups. Colloid oncotic pressure did not change in the control group. n Group 2, it fell from 16 t 0.7 to mm Hg over two hours, and in Group 3, from 17? 0.5 to mm Hg. There was no difference between the two alloxan groups, and both groups differed from the control (p < 0.005). n contrast, hemoglobin concentration did not alter significantly relative to the control in either group.

4 45 Helbert, Paskanik, and Bredenberg: PEEP and Pulmonary Edema w 30r z p lot 5r 0 + PEEP f 1 T P ence among the groups. The pulmonary shunt in Group 2 was significantly greater than that in both the control group and Group 3 (p < 0.005) (Fig 3). The shunt in Group 3 (alloxan plus PEEP) did not differ significantly from that in the control group. The slight tendency for the ratio of dead space to tidal volume to increase in the alloxan group ventilated with PEEP did not approach statistical significance (see Fig 3). Static compliance at both 500 cc and 1,000 cc ranged between 0.08 k 0.01 and 0.10 k 0.01 L/ cm H20. Static compliance did not change significantly in any of the groups over the course of the experiment, and neither alloxan group differed significantly from control at any time. Y a > a 2 P 1' 1 C P ' 2' Fig 2. Measurement of pulmonary hemodynamics at control (C), after alloxan (), at one hour ( '), and at two hours (2'). The group given alloxan plus PEEP has additional measurements (P) immediately after application of 10 cm H20 of PEEP. rrow at indicates alloxan injection and beginning of 20-minute infusion of 40 ml per kilogram of normal saline solution. Mean pulmonary artery (P) and left ventricular end-diastolic pressure (LVEDP) measurements are made relative to atmosphere. Data are given as mean? standard error of the mean. (PVR = pulmonary vascular resistance.) The Pa02 was significantly lower in the alloxan group ventilated without PEEP (Group 2) relative to either of the other groups ( p < 0.005). n the alloxan group ventilated with PEEP (Group 3), Pa02 did not differ significantly from that in the control group (Table). There was no significant variance in PaC02 in any group. metabolic acidosis developed in all three groups, with no statistically significant differ- T Comment nfusion of alloxan is one of the standard experimental techniques for inducing membrane permeability pulmonary edema [ll, 16-18]. Quantitatively more consistent edema is obtained if infusion of saline solution is combined with administration of alloxan. The findings in both alloxan groups of a fall in colloid oncotic pressure with an unchanged hemoglobin concentration are consistent with protein loss in the edema fluid secondary to the alloxan-induced increase in microvascular permeability [ 171. The transient rise in pulmonary artery pressure is a recognized hemodynamic consequence of alloxan infusion [16, 191, and it occurred in both experimental groups. Ten cm H20 of PEEP alone elevated LVEDP and mean pulmonary artery pressure measured relative to atmosphere. Since with unchanged compliance, it can be anticipated that about 50% of airway pressure will be transmitted to pulmonary interstitium and pleural space [9, 201, the apparent increase in LVEDP of 2 to 3 mm Hg does not represent any real increase in transmural pressure. Thus, in these experiments, alloxan, which is known to increase pulmonary microvascular permeability, plus infusion of saline solution created a nonhemorrhagic pulmonary edema in which there was no sustained increase in transmural pulmonary microvascular hydrostatic pressure and in which total pulmonary blood flow (cardiac output) remained constant in all

5 46 The nnals of Thoracic Surgery Vol 36 No 1 July 1983 rterial Blood Gas Dataa Variable Group Baseline First Hour Second Hour PH ~~~ ~ ~ ~ adata are shown as mean 5 standard error. 102 t 4 95 t 2 97 t 4 39 * ? t ? t * 1 o t o -1 t 0 94 t 5 81 t 7 99 t 6 38 t 1 37 * 2 41 t t t ? t 1-8? 1-8 t 1 98 * 5 73 * t 5 36 t 2 38? 2 36 * t t t 1-8 t 1-8 t 1 PaOZ = partial pressure of arterial oxygen; PEEP = positive end-expiratory pressure; PaC:OZ = partial pressure of arterial carbon dioxide. 5 - three groups. Positive end-expiratory pressure was applied prior to the induction of pulmonary edema in order to maximize the potential for - * demonstrating an effect of PEEP by retarding from the outset the accumulation of lung water, rather than attempting to demonstrate reversal of a process already far advanced. Significant 5 7 pulmonary edema was observed in both alloxan groups, however, and 10 cm H20 of PEEP 0 \ 1.0 r 0 lloxon added to controlled mechanical ventilation had no measurable effect on the accumulation of pulmonary extravascular lung water. The speculation tlhat PEEP may drive water + from the lung has been based largely on obser- PEEP vation of patients in complex clinical settings without direct measurement of lung water Many authors [4--91, however, have demonstrated experimentally that PEEP does not decrease lung water in pulmonary edema caused by increasing microvascular hydrostatic pres- C ' 2' sure. Less experimental work has been directed '----@' $ - at the effect of PEEP on membrane permeabil- 3. (Qs~QT) and dead space to tzdul volume (VDV,) ratios. (C = control; 1' = one hour; 2' ity-induced edema, which may be the form of = two hours; 2 = < 0,005, Group 2 compared with edema particularly relevant to the acute respira- Groups 1 and 3.) tory distress syndromes. Dunegan and coworkers [21] studied membrane permeability pulmonary edema in dogs that was induced by oleic acid, and measured lung water both by lung thermal volume and by a gravimetric technique that made no correction for intravascular

6 47 Helbert, Paskanik, and Bredenberg: PEEP and Pulmonary Edema water. They found that PEEP reduced lung thermal volume and concluded that PEEP effects a small reduction in the rate of accumulation of pulmonary edema. Their gravimetric estimate of lung water showed no effect of PEEP, but they believed that the gravimetric results were compromised by not allowing for changes in intravascular water volume. n alternative explanation of their data is that PEEP did not substantially alter lung water and that the observed decrease in lung thermal volume may reflect an inaccuracy of the double-indicator method. Weisel and associates [22], using a doubleindicator dilution technique in dogs, also showed a decrease in pulmonary extravascular water volume with PEEP in oleic acid-induced pulmonary edema. They thought, however, that the lower measurement of this variable in animals given PEEP was due to the effect of a decreased pulmonary perfusion bed on the double-indicator method, and they concluded that PEEP did not reverse interstitial edema. Hopewell [23] studied the effect of PEEP in alloxan-induced pulmonary edema using a gravimetric method for measuring pulmonary extravascular water volume. He found no reduction at one hour in dogs in which 10 cm H20 was added to mechanical ventilation following the administration of alloxan and saline solution. n 1981, Peitzman and colleagues [24] reported the results of a study of the effect of 10 and 15 cm H20 of PEEP applied three hours after oleic acid-induced pulmonary edema. Pulmonary extravascular water volume was measured serially using the thermallgreen dye double-indicator technique and terminally, using gravimetric methods. The double-indicator technique showed no reduction in pulmonary extravascular water volume over two hours following the addition of PEEP. Excellent correlation (r = 0.94) was demonstrated between the final double-indicator measurements of pulmonary extravascular water volume and the terminal gravimetric measurement. Our own experiments agree with the results of Hopewell [23] and Peitzman and co-workers [24] and demonstrate that 10 cm H20 of PEEP added to mechanical ventilation not only fails to reverse an established edema process, but also fails to retard the accumulation of pulmonary extravascular water volume when it is applied before the injury to the pulmonary microvascular membrane. lthough there is not absolute unanimity, a consensus emerges, particularly from the gravimetric measurements of lung water, that PEEP does not retard or reduce the accumulation of lung water in membrane permeabilityinduced pulmonary edema. Thus, PEEP does not appear to have a direct mechanical effect on lung water when either of the two mechanisms for pulmonary edema is experimentally isolated, that is, increased microvascular permeability or increased microvascular hydrostatic pressure. t is apparent that the proportion of airway pressure transmitted to the perivascular interstitium is balanced by the same pressure being transmitted to the intravascular compartment [15, 201, so the addition of PEEP does not alter the net balance of hydrostatic pressures affecting water transfer across the pulmonary microvascular membrane. The major benefits of PEEP in edematous lungs are increased functional residual capacity and reduced pulmonary shunt. By keeping the lung higher on its pressure volume curve, more alveoli remain open and ventilated despite interstitial and intraalveolar edema [16]. These benefits have been demonstrated clinically in a variety of acute respiratory distress syndromes [, 31 and experimentally in both high-pressure and membrane permeability edema [4, 7, 241. Removal of PEEP and loss of alveolar volume may cause a shift of fluid from alveoli to more proximal conducting airways where foaming may occur, leading to airway obstruction [16]. This has been observed experimentally [8] without changes in pulmonary extravascular water volume and may explain an increase in rales and frothy sputum, which sometimes are noted clinically when PEEP is removed prematurely from patients with pulmonary edema. References 1. shbaugh DG, Petty TL, Bigelow DB, et al: Continuous positive-pressure breathing (CPPB) in adult respiratory distress syndrome. J Thorac Cardiovasc Surg 5731, Barach L, Martin J, Eckman M: Positive-pressure respiration and its application to the treatment of acute pulmonary edema. nn ntern Med 12:754, 1938

7 48 The nnals of Thoracic Surgery Vol 36 No 1 July Kumar, Falke KJ, Geffen B, et al: Continuous positive-pressure ventilation in acute respiratory failure. N Engl J Med 283:1430, Bredenberg CE, Kazui T, Webb WR: Experimental pulmonary edema: the effect of positive endexpiratory pressure on lung water. nn Thorac Surg 26:62, Bredenberg CE, Webb WR: Experimental pulmonary edema: the effect of unilateral PEEP on the accumulation of lung water. nn Surg 189:433, Caldini PJ, Leith D, Brennan MJ: Effect of continuous positive-pressure ventilation (CPPV) on edema formation in dog lung. J ppl Physiol 39:672, Demling RH, Staub NC, Edmunds LH Jr: Effect of end-expiratory airway pressure on the accumulation of extravascular lung water. J ppl Physiol 38:907, Wagner E, Rieben P, Katsuhara K, et al: nfluence of airway pressures on edema in the isolated dog s lung. Circ Res 9:382, Woolverton WC, Brigham KL, Staub NC: Effect of positive-pressure breathing on lung lymph flow and water content in sheep. Circ Res 42:550, Demling RH: The pathogenesis of respiratory failure after trauma and sepsis. Surg Clin North m 60:1372, Staub NC: Pulmonary edema due to increased microvascular permeability to fluid and protein. Circ Res 43:143, Braunwald E, Fishman P, Cournand, et al: Time relationship of dynamic events in the cardiac chambers, pulmonary artery and aorta in man. Circ Res 4:100, Bendixen HH, Egbert LD, Hedley-Whyte J, et al: Respiratory Care. St. Louis, Mosby, Pearce ML, Yamashita J, Beazell J: Measurement of pulmonary edema. Circ Res 16:482, Staub NC, Hogg JC: Conference report of a workshop on the measurement of lung water. Crit Care Med 8:752, Staub NC, Magaro 13, Pearce ML: Pulmonary edema in dogs, espelzially the sequence of fluid accumulation in lungs. J ppl Physiol22:227, Vreim CE, Staub NC: Protein composition of lung fluids in acute alloxan edema in dogs. m J Physiol 230:376, Nelson RM, Mcntyre BR, Egan E: Solute permeability of the alveolar epithelium in alloxan edema in dogs. J ppl Physiol 44:353, viado DM: The Lung Circulation. New York, Pergamon, 1965, vol2, pp Chapin JC, Downs J13, Douglas ME, et al: Lung expansion, airway pressure transmission, and positive end-expiratory pressure. rch Surg 114:1193, Dunegan JL, Knight DC, Harken, et al: Lung thermal volume in pulmonary edema: effect of positive end-expiratory pressure. nn Surg 181:809, Weisel RD, Vito L, l latt H, et al: Treatment of acute respiratory insufficiency with cyclic expiratory pressure. Surg Forum 24:234, Hopewell PC: Failure of positive end-expiratory pressure to decrease lung water content in alloxan-induced pulmonary edema. m Rev Respir Dis 120:813, Peitzman B, Corbett W, Shires GT 111, et al: The effect of increasi~ng end-expiratory pressure on extravascular lung, water. Surgery 90:439, 1981