INFLUENCE OF PULMONARY ARTERIAL PRESSURE DURING FLUSHING ON LUNG PRESERVATION

Abstract

The effect of changing pulmonary artery pressure during flushing(flushing pressure) on lung preservation was examined in an ex vivo rabbit lung perfusion model. Both lungs were flushed in situ with 200 ml of a preservation solution (extracellular fluid [ECF] type) at 8°C, at a constant flushing pressure maintained by regulating the flushing flow rate from 20 to 120 ml/min. In the controls, the flushing pressure was maintained at 15 mmHg and the lungs were assessed without storage. In the other 5 groups(n=7 in each group), the lungs were flushed at pressures of 5, 10, 15, 20, and 25 mmHg, respectively. The heart-lung block was then harvested, and immersed in the same solution at 8°C for 24 hr. After 24 hr of storage, the block was reperfused with pooled venous blood for 10 min, and then with oxygenated blood for another 60 min in a closed circuit. Assessment of lung function included blood gas analysis of effluent blood, pulmonary artery pressure, airway pressure, wet:dry weight ratio, and histologic study.

At flushing pressures of 5, 20, and 25 mmHg, uniform and clear flushing out of the pulmonary vascular beds was not obtained, resulting in postperfusion pulmonary hypofunction and a high incidence of pulmonary edema. However, at flushing pressures of 10-15 mmHg, we succeeded in completely flushing out the pulmonary vascular beds, and managed to preserve good pulmonary function. In conclusion, we determined the optimal flushing pressure for rabbit lung preservation to be 10-15 mmHg.

The single-flush perfusion method is currently the most widespread method of perfusion used for clinical lung transplantation ^{(1, 2)}. The Stanford group has demonstrated the clinical usefulness of this method using Euro-Collins solution in the preservation of distant donor grafts for combined lung and heart transplantations ⁽¹⁾. There have been a number of lung preservation procedures proposed, with many different viewpoints regarding the optimal composition of the flushing solution ^(3-5), the optimal preservation temperature ^(6-8), and the optimal temperature of the flushing solution ⁽⁹⁾. The efficacy of hyperinflation has also been evaluated ⁽¹⁰⁾. Currently, it is believed that, because of the structural peculiarities of the lungs, a low flushing pressure is desirable for lung preservation. However, there have been no detailed reports describing the effects of various flushing pressure on lung preservation. This study was designed to examine the optimal flushing pressure that minimizes perfusion-induced damage to the donor grafts.

MATERIALS AND METHODS

Lung perfusion and heart-lung block removal. Japanese male white rabbits weighing 2.8-3.2 kg were premedicated with an intramuscular injection of ketamine hydrochloride (30 mg/kg) and atropine sulfate (0.10 mg/kg). The animals were then anesthetized with intravenous sodium pentobarbital (20 mg/kg), and anticoagulated with intravenous heparin sodium (700 IU/kg). An endotracheal tube (I.D. 3.0 mm) was introduced through a cervical tracheotomy, and the animals were ventilated with room air using a volume-cycled respirator(Shinano Factory Co., Tokyo, Japan) at a tidal volume of 10 ml/kg, a respiratory rate of 40 breaths/min, and a positive end-expiratory pressure of 1 cm of H₂O. A median sternotomy was performed. Next, the pericardium was opened and the superior and inferior vena cavae were ligated. The pulmonary artery (PA^*) and the left atrium were cannulated with a double lumen and a single lumen plastic catheter, respectively(originally modified from Tygon I.D. 1/8). Following ligation of the ascending aorta, both lungs were flushed with 200 ml of an extracellular fluid(ECF)-type solution (Table 1) at 8°C through one of the PA cannulas; the other cannula was used to measure the flushing pressure. The flushing pressure was controlled at different levels in 6 subject groups by regulating the flushing flow rate between 20 and 120 ml/min with a tubing pump (Taiyo tubing pump, type 1500N, Tokyo, Japan). During the flushing procedure, the lungs were ventilated continuously; cannulation time, flushing volume, flushing time, flushing pressure, and surface temperature of the preserved lungs before and after the flushing were recorded. After this flushing procedure, the heart-lung block was removed with the lung inflated at end-inspiration levels and then immersed in the same solution at 8°C for 24 hr. The pulmonary vascular resistance (PVR) and flushing flow rate were calculated as follows: PVR (mmHg/L/min) = flushing pressure(mmHg)/0.2 L/flushing time (min), and flow rate (ml/min) = 200 ml/flushing time (min).

Experimental groups. The animals were randomly divided into a control and 5 experimental groups. In the control group (n=7), the flushing pressure was controlled at 15 mmHg. Upon completion of the flushing, the heart-lung blocks were removed and a functional assessment was performed 10 min after harvesting. In groups 1-5 (n=7 in each group), the flushing pressure was maintained at 5, 10, 15, 20, and 25 mmHg, respectively. Upon completion of the flushing, the heart-lung blocks were excised and stored in the same preservation solution at 8°C for 24 hr, followed by a functional assessment.

Assessment. The functional assessment was performed for 70 min according to the isolated perfused rabbit lung model proposed by Wang et al.⁽⁶⁾. After 24 hr of storage, the right main bronchus, right pulmonary artery, and right pulmonary vein were ligated, and the right lung was removed for the wet:dry weight (W:D) ratio measurement and histologic study. The left lung was mounted in a prewarmed (37°C) and humidified resin-glass chamber, and then ventilated with room air at a tidal volume of 5 ml/kg, a respiratory rate of 30 breaths/min, and a positive end-expiratory pressure of 1 cm of H₂O. The left lung was perfused through the PA cannula with pooled homologous venous blood (200 ml) at 37°C for 10 min at a constant flow rate of 20 ml/min using a tubing pump. The graft was then perfused with oxygenated blood in a closed circuit for 60 min.

Measurement Indices. Arterial blood gas analysis: A sample of pooled venous blood was taken at the onset of reperfusion. The pulmonary venous effluent blood samples were taken from the left atrium cannula at 6 and 10 min after the initiation of reperfusion. The blood gases were analyzed using a gas analyzer (type 288, CIBA-Corning, Tokyo, Japan).

Airway pressure and PA perfusion pressure: The airway pressure and mean PA perfusion pressure were monitored continuously and recorded during reperfusion through the endotracheal tube and one of the double lumen plastic catheters, respectively.

Pulmonary W:D ratios: The right lung was weighed after 24 hr of storage and the left lung was weighed immediately after the reperfusion. The lungs were desiccated in a heating chamber at 80°C until they reached a constant weight, and then were reweighed (dry weight) for the calculation of the W:D ratios.

Histologic study: Portions of the right upper lobe and of the left lower lobe apex were fixed with 10% formalin, and then thinly sectioned. These sections were stained with hematoxylin-eosin for examination by light microscopy. Pulmonary edema was defined by the following findings: the presence of an exudate or blood in the alveoli and edematous changes in the perivascular tissues and/or in the stroma.

All results were expressed as the mean ± SEM. Statistical analyses were performed using the unpaired t test and the Mann-Whitney test. Statistical significance was assumed for a P-value less than 0.05.

All animals received humane care in compliance with the Principles of Laboratory Care formulated by the National Society for Medical Research, and the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences, and published by the National Institute of Health (NIH publication no. 85-23, revised 1985).

RESULTS

The 6 groups of animals did not differ significantly from each other with respect to the mean weight or blood gas and hemoglobin determinations from samples of homologous venous blood used for perfusion of the lungs. In the controls and in groups 1, 2, and 3, the functional assessment was feasible for 70 min in all of the lung specimens. However, in 3 of the 7 specimens in group 4 and 4 of the 7 specimens in group 5, pulmonary edema developed immediately after reperfusion; hence, the assessment was discontinued within 12 min in these cases due to the appearance of a foamy exudate in the airways. Therefore, by 12 min or more after reperfusion, the functional assessments were confined to only 4 specimens from group 4 and 3 specimens from group 5.

Pulmonary flush parameters(Table 2). The 6 groups did not differ significantly from each other in their PVR or cannulation times. As the flushing pressures increased, shorter flushing times and faster flushing flow rates were obtained, with significant differences between each group. The flushing times were 9.0±0.6, 4.1±0.4, 2.7±0.3, 2.4±0.2, and 1.7±0.2 min for groups 1, 2, 3, 4, and 5, respectively. The flushing flow rates were (ml/min) 23.4±2.4 for group 1, 54.1±6.4 for group 2, 75.1±6.5 for group 3, 91.8±9.1 for group 4, and 115.5±8.7 for group 5. The surface temperature of the lungs immediately after the flushing was significantly higher in group 1 than in the other groups (P<0.05). In the controls and groups 2 and 3 (Fig. 1A), the lungs appeared uniformly pale because of sufficient flushing of the blood out of the tissues. However, lungs from groups 1 (Fig. 1B), 4(Fig. 1C), and 5 demonstrated patchy patterns on their surfaces due to incomplete flushing out of the blood in the peripheral regions.

Blood gas analysis during reperfusion(Table 3). At 6 and 10 min after reperfusion in group 4, and at 10 min after reperfusion in group 1, the pulmonary venous effluent blood oxygen tensions were significantly lower than in controls (P<0.05). Group 5 demonstrated a significantly lower pulmonary venous effluent blood oxygen tension than either the controls or groups 2 and 3 (P<0.05). The pulmonary venous effluent blood carbon dioxide tensions obtained at 6 min after reperfusion in groups 1 and 5 were significantly higher than in controls(P<0.05). Groups 1 and 5 also each demonstrated a significantly higher pulmonary venous effluent blood carbon dioxide tension at 10 min after reperfusion than either the controls or groups 2 and 3(P<0.05).

PA pressure during reperfusion(Table 4). During the experiment (70 min), the mean PA pressure was significantly higher in group 4 than in either the controls or groups 1, 2, and 3(P<0.05). Group 5 demonstrated a significantly higher PA pressure at 3 and 6 min after reperfusion than either the controls or groups 1, 2, and 3 (P<0.05).

Airway pressure. No significant differences in the peak airway pressures recorded during reperfusion existed between any of the groups.

W:D ratios and the incidence of pulmonary edema(Fig. 2). No significant differences in the W:D ratios of the unperfused right lungs were detected among any of the groups. The postreperfusion W:D ratios of the left lungs in group 5 were significantly higher than in the other groups (P<0.05). Groups 1 and 4 also showed significantly higher postreperfusion W:D ratios than the controls(group 1 vs. control: P<0.01, group 4 vs. control:P<0.05). Pulmonary edema was histologically detected in 0/7 specimens from the controls, 3/7 from group 1, 1/7 from group 2, 1/7 from group 3, 3/7 from group 4, and 7/7 from group 5.

DISCUSSION

The single-flush perfusion method, which involves rapid cooling of the lung followed by static hypothermic storage, is the most widely used method in clinical lung preservation. This method provides several advantages: (a) the procedure is simple and involves the minimum amount of special equipment, (b) it shortens the warm ischemic time by its use of rapid cooling, (c) it allows satisfactory flushing out of the intrapulmonary blood, thus preventing the no-reflow phenomenon and the production of toxic substances from residual blood components, and (d) it minimizes the adverse effects of cooling on cell physiology by the use of an appropriate preservation solution^{(11, 9)}. On the other hand, there are several potential disadvantages, including: (a) thrombogenesis in the pulmonary capillaries, (b) difficulty in obtaining uniform cooling, because of disturbances in the flow distribution of the lungs due to an insufficient volume or a low flow rate of the perfusion solution, and (c) direct damage to the endothelial cells by the perfusion process ⁽¹¹⁾. Therefore, knowledge of the optimal flushing pressure for lung preservation may help prevent organ damage, since a satisfactory flushing out of the intrapulmonary blood and a shortening of the warm ischemic time can be achieved.

There have been a number of reports describing flushing pressures used for lung preservation. In a lung preservation study in adult chacma baboons reported by Sundaresan ⁽¹²⁾, the lungs were flushed with a low-potassium dextran solution at a height of 40 cm without monitoring flushing pressures. During lung transplantation performed in adult mongrel dogs, Puskas et al. ⁽¹³⁾ perfused the grafts at flushing pressure of less than 12 mmHg at a height of 30 cm. In a comparative experiment on preservation solutions using New Zealand White rabbits, Yamazaki et al. ⁽¹⁴⁾ controlled the flushing pressure to less than 25 mmHg. Bresticker et al. ⁽¹⁵⁾ have reported the preservation of mongrel dog lungs at a flushing pressure of 15 mmHg. Haverich et al. ⁽¹⁶⁾ have reported the efficacy of using a large volume and high flow rate of flushing solution in lung preservation, but did not report the optimal flushing pressure or flow rate. In clinical lung transplantation, Colquhoun and co-workers have perfused grafts with Euro-Collins solution (60 ml/kg) by gravity drainage for more than 5 min; however, the flushing pressures were not determined during this procedure⁽¹¹⁾. In addition, Baldwin et al. ⁽¹⁾ have proposed that the flushing pressure should be less than 20 mmHg. Thus, a consensus on the optimal flushing pressure has not been reached by investigators in either experimental or clinical fields.

In this study, the flushing pressure of each group was regulated at either 5, 10, 15, 20, or 25 mmHg, with a mean pressure of 15 mmHg. This mean was selected since the normal mean pulmonary arterial pressure upon intubation following the median sternotomy was 17.5±1.0 mmHg (n=10). Instead of the widely used method of flushing by gravity drainage, we used a roller pump in order to dynamically change the flushing pressure. In addition, we flushed the lungs with 200 ml of a preservative solution based upon clinical requirements as determined by weight (60 ml/kg) ^{(1, 2)}.

In this study, we used an ECF-type preservation solution. The efficacy of using this solution as a preservative has been well established^{(3-5, 14)}. When compared with flushing with intracellular fluid-type solutions, a larger volume of ECF solution can be perfused in a shorter period of time at lower flushing pressures, thus achieving more satisfactory preservation conditions ^{(3, 14)}. With an ECF-type solution, we were able to flush a large volume of solution in a shorter period of time, and increase the flushing pressure by raising the flushing flow rates.

The isolated perfused rabbit lung model used in this study is a valuable screening method for evaluating the effects of various factors on lung preservation. This method has been used in several experiments with satisfactory outcomes ^{(6, 14, 15)}. In this model, the flow distribution of the preserved lungs and the degree of lung inflation, which are important parameters in the evaluation of perfusion patterns, were immediately visible after reperfusion. In addition, this modified model allowed for the assessment of longer-term lung preservation as compared with the routine model consisting of a 10-min perfusion period. We used a flow rate of 20 ml/min to reperfuse preserved lungs. In a preliminary experiments (without preservation, n=5), when the both lungs were perfused by venous blood in an ex vivo circuit at moderate perfusion flow rates (50-80 ml/min), a moderate pulmonary artery perfusion pressure (10-15 mmHg) was obtained. Therefore, we used a flow rate of 50 ml/min at bilateral lung perfusion to minimize direct damage to the pulmonary vessels caused by the mechanical perfusion (rather than 80 ml/min). Because the left lung is smaller than the right lung, we used a flow rate of 20 ml/min for the left lung. In this study, we obtained 11-14 mmHg (pulmonary artery perfusion pressure) at a constant flow rate of 20 ml/min in the control group for 70 min.

During organ preservation, the immediate reduction of warm ischemic conditions by rapid cooling is an important factor. It would be ideal to perfuse a large volume of perfusate at a low flushing pressure for a short time ⁽¹⁶⁾. However, a large volume may increase the flushing pressure, and thus cause mechanical damage to the preserved lungs. In this study, the high-pressure flushing groups (20 and 25 mmHg) demonstrated unsatisfactory flushing patterns on the lung surface after flushing(Fig. 1C), despite the higher flow rates (90-120 ml/min) and the shorter flushing times (2-2.5 min). In the low-pressure flushing group(5 mmHg), the surface temperature of the lungs immediately after flushing was higher due to the low flow rate (20-30 ml/min), and a uniform and clear flushing out of the blood could not be obtained (Fig. 1B). However, at flushing pressures of 10-15 mmHg, and at moderate flushing flow rates (50-80 ml/min), uniform flushing out of the blood was achieved(Fig. 1A) without compromising the functional capacity of the preserved lungs. The high pressure flushing at 20 or 25 mmHg induced vasoconstriction of the pulmonary artery due to the rapid inflow, resulting in unsatisfactory flushing out of the pulmonary vascular beds and dysfunction of the preserved lungs. Low pressure flushing at 5 mmHg failed to reach the peripheral vascular beds due to the slow flow rate, and resulted in incomplete flushing out of the blood. This led to poor gas diffusion and oxygenation with a high incidence of pulmonary edema. In the heart, with regard to perfusion pressures involving a cardioplegic solution, Buckberg and co-workers have demonstrated that a perfusion pressure of higher than 100 mmHg causes myocardial edema ⁽¹⁷⁾. In a report by Becker et al.⁽¹⁸⁾, perfusion pressures of less than 50 mmHg failed to reach the periphery, thereby causing myocardial injury secondary to ischemia.

The pulmonary vascular bed is quite barosensitive, and an increase in pressure induces reflex vasoconstriction which can lead to endothelial injury. However, at low pressures, the vascular beds cannot be fully flushed. In this study, by flushing the rabbit lungs at a flushing pressure of 10-15 mmHg, which is slightly lower than the normal pulmonary arterial pressure, we succeeded in completely flushing out the pulmonary vascular beds, and managed to preserve good pulmonary function. The normal mean PA pressure of the rabbit lungs used in this study was 17.5±1.0 mmHg, which is close to the human mean PA pressure of 10-20 mmHg ⁽¹⁹⁾. Therefore, we hope that the results obtained in this study will provide some useful information for clinical human lung transplantation.

Figure 1: (A) Pulmonary surface following perfusion in a specimen from group 2 (10 mmHg) demonstrates satisfactory flushing out of the pulmonary vascular beds. (B and C) In contrast, in a specimen from groups 1 (5 mmHg; B) and 4 (20 mmHg; C), a patchy pattern was found on the pulmonary surface, indicative of incomplete flushing out of the peripheral regions.

Figure 2: Pre- and postreperfusion W:D ratio. ^* P<0.05 group 5 vs. control and groups 1, 2, 3, and 4; ^** P<0.05 group 4 vs. control. †P<0.01 group 1 vs. control.