INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Reperfusion lung injury following orthotopic lung transplantation occurs in 10 to 20% of patients and is an important source of morbidity and mortality (1, 2). The clinical syndrome of reperfusion edema usually develops within 24 h of surgery and is characterized by hypoxemia, decreased pulmonary compliance, and patchy or diffuse alveolar infiltrates in the transplanted lung or lungs, with histologic evidence of diffuse alveolar damage (3). Although most patients recover, severe graft dysfunction can be fatal, contributing to 13% of early deaths after transplant (4). Even when less severe, reperfusion injury prolongs ventilator time and the duration of intensive care unit (ICU) stay (1), and also may be associated with an increase in the incidence of early rejection (5, 6). Furthermore, the increased risk of reperfusion injury and primary graft failure with longer graft ischemic times limits the availability of donor lungs, contributing to the current shortage of donor lungs for transplantation (7).

The mechanisms underlying reperfusion lung injury have been investigated in experimental models (8). The pulmonary endothelium appears to be particularly susceptible to injury from ischemia and reperfusion. This injury is mediated by a variety of mechanisms including neutrophil adhesion and transendothelial migration with release of proinflammatory mediators and reactive oxygen species (1, 9, 10). Increased endothelial permeability has been postulated to be a primary cause of reperfusion pulmonary edema (11). While the role of the endothelium in reperfusion lung injury has been increasingly well defined, the contribution of the alveolar epithelium is poorly understood.

Work from our laboratory has demonstrated the importance of intact alveolar epithelial transport to recovery from increased permeability edema associated with clinical acute lung injury (12). However, the capacity of the alveolar epithelium to remove edema fluid after lung transplantation has never been studied. Interestingly, recent data suggest that the human alveolar epithelium may be relatively resistant to injury from prolonged cold ischemia and rewarming (13). In an ex vivo human lung preparation, sodium and fluid transport were abolished by cooling to 4° C for 4 to 6 h, but returned to normal levels with rewarming to 37° C.

In order to better understand the role of the alveolar epithelium in both the initial and resolution phase of clinical reperfusion injury, we collected pulmonary edema fluid from eight patients with acute reperfusion edema after orthotopic lung transplantation. The first objective was to test directly the hypothesis that reperfusion edema is predominantly due to increased permeability of the alveolar-capillary barrier. By sampling undiluted pulmonary edema fluid and plasma at the onset of reperfusion edema, it was possible to measure the edema fluid to plasma protein ratio, a direct measure of permeability. The second objective was to determine if the alveolar epithelium was functionally intact by measuring changes in protein concentration in serial pulmonary edema fluid samples; a rise in protein concentration indicates net alveolar epithelial fluid transport (12, 14). Based on our studies in the ex vivo human lung, we hypothesized that the capacity of the alveolar epithelium to remove alveolar fluid would be preserved, even in the setting of pronounced reperfusion edema. In order to determine the relationship of alveolar fluid clearance to clinically relevant outcomes, evidence of early alveolar fluid clearance was correlated with radiographic and pathologic findings. To our knowledge, this represents the first human study of the alveolar epithelium in the initial and resolution phase of reperfusion pulmonary edema.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Selection and Clinical Data Collection

Between April 1993 and June 1998, eight patients were identified with acute clinical reperfusion edema within 24 h of single or bilateral sequential orthotopic lung transplantation. Clinical criteria for the diagnosis of reperfusion pulmonary edema included development of a ratio of arterial oxygen pressure to fraction of inspired oxygen (Pa_O2/ FI_O2) < 200 and diffuse infiltrates in the transplanted lung or lungs. Medical records were comprehensively reviewed for each patient and the following parameters were recorded: hemodynamic, respiratory, and laboratory data; graft ischemic time; length of cardiopulmonary bypass; cardiac function as assessed by echocardiographic or gated nuclear medicine studies; medications; duration of mechanical ventilation; length of ICU stay; hospital course; and survival. All chest radiographs for each patient were reviewed by two of the investigators in conjunction with a radiologist. Transbronchial biopsies were taken in five of the eight patients within 1 wk of transplant as part of the clinical transplant protocol and were reviewed with one of the authors, a pulmonary pathologist (W.F.). This study was approved by the Committee for Human Research at the University of California, San Francisco.

Collection of Pulmonary Edema Fluid

The first pulmonary edema fluid sample was collected at the onset of clinically evident reperfusion pulmonary edema. In three patients, reperfusion edema was evident within 90 min of postoperative ICU admission. The remaining five patients had initial sampling at the onset of pulmonary edema, 3.5 to 19 h after ICU admission. Serial edema fluid samples at sequential time points up to 24 h after the initial sample were collected in six of the eight patients. Pulmonary edema fluid was collected by the authors or trained respiratory therapists as previously reported (12). Briefly, a 14-French suction catheter was passed through the endotracheal tube and wedged in a distal airway. Gentle suction was applied to obtain at least 1 to 2 ml of undiluted edema fluid in an attached Luken suction trap (12). One hundred units of heparin was added to each sample. A simultaneous plasma sample was obtained in an ethylenediaminetetraacetic acid (EDTA)-treated tube at the time of each edema fluid sampling. Edema fluid and plasma samples were immediately centrifuged at 3,000 × g for 10 min; the supernatants were removed and frozen at 70° C.

Protein Determination and Cell Counts

Protein concentrations were measured in duplicate in thawed edema fluid supernatants and plasma by the Biuret method, as previously described (12, 15). Edema fluid cell and differential counting were done by standard methods in the hospital laboratory.

Calculation of Edema Fluid to Plasma Protein Ratios and Alveolar Liquid Clearance

The ratio of the protein concentration in pulmonary edema fluid to simultaneous plasma protein was calculated as previously described (12, 16). The rate of alveolar fluid clearance was calculated by comparing the final and initial pulmonary edema fluid protein concentrations in patients with sequential samples to calculate percent clearance per hour. These methods have been previously validated in dog and sheep lungs (14, 19, 20), as well as in clinical studies of human pulmonary edema (12, 16).

Statistical Analysis

A paired t test was used to compare the initial and final alveolar edema fluid protein concentration in the six patients with sequential sampling. An unpaired t test was used to compare the presence or absence of alveolar fluid clearance with duration of mechanical ventilation and ICU stay. A p value < 0.05 was accepted as statistically significant. Linear regression analysis was used to correlate graft ischemic time to edema fluid to plasma protein ratio, duration of mechanical ventilation, and ICU stay.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

A summary of patient clinical characteristics is presented in Table 1. Eight patients with clinically significant reperfusion edema were identified during the study period, representing 15% of a total of 55 lung transplants done at our institution during this time period. Prior to transplantation, all patients were clinically stable as outpatients. All patients underwent orthotopic single or bilateral sequential lung transplantation for severe chronic progressive lung disease using standard surgical techniques. The lung preservation protocol included a single flush of University of Wisconsin solution (21) and cooling to 4° C. Intraoperative cardiopulmonary bypass was utilized in two patients (Patients 2 and 4). Overall, there were no intraoperative complications. Patient 2 required three reoperations for left hemothorax during the first three postoperative days. A standard postoperative immunosuppression protocol was administered in all patients including high-dose intravenous methylprednisolone, cyclosporine, and azathioprine. All eight patients recovered from reperfusion edema and survived to hospital discharge. Currently seven of the eight patients are alive; Patient 4 died of overwhelming infection 1 yr after transplant.

Pulmonary Edema Fluid Analysis

The pulmonary edema fluid to plasma protein ratio was calculated from the initial edema fluid and plasma sample for each patient (Table 1). Six of the eight patients with clinical reperfusion edema had an edema fluid to plasma protein ratio greater than 0.75, consistent with increased permeability pulmonary edema (12, 15). The remaining two patients (Patients 1 and 2) had edema fluid to plasma protein ratios of 0.66 and 0.73 consistent with pulmonary edema due to a combination of hydrostatic and increased permeability forces (12, 17).

The pulmonary edema fluid to plasma protein ratio, an indicator of the relative permeability of the alveolar-capillary barrier, correlated significantly with graft ischemic time (r = 0.77, p < 0.05) (Figure 1). Patient 8 was excluded from this analysis because the initial edema fluid to plasma protein ratio was greater than one, indicating that some alveolar fluid resorption had already taken place before collection of the initial edema fluid sample. Interestingly, the edema fluid to plasma protein ratio did not correlate with clinical parameters such as the degree of hypoxemia as measured by initial alveolar-arterial oxygen gradient, severity of pulmonary infiltrates, duration of mechanical ventilation, or length of ICU stay.

View larger version (14K): [in this window] [in a new window]   Figure 1.   Correlation of graft ischemic time with initial pulmonary edema fluid to plasma protein ratio in patients with reperfusion edema after orthotopic lung transplantation (r = 0.77, p < 0.05). Undiluted edema fluid and plasma were collected simultaneously at the time of onset of the reperfusion pulmonary edema. Graft ischemic time refers to the time from organ harvest until reperfusion of the lung in the recipient in the operating room. When bilateral sequential lung transplantation was carried out, the longer of the two graft ischemic times was used.

In five patients, sufficient pulmonary edema fluid was obtained to analyze cell counts. The edema fluid white blood cell counts were 1,013 to 66,000 leukocytes/µl with a marked predominance of neutrophils (71 to 99%).

Resolution of Alveolar Edema

Serial sampling of pulmonary edema fluid allowed calculation of alveolar fluid clearance rates in six of the eight patients. As shown in Table 2 and Figure 2, four of the six patients had high rates of clearance, ranging from 12 to 32% per hour. Patients 2 and 4 had less than 1% clearance per hour, consistent with no net alveolar fluid clearance. The four patients with intact alveolar fluid clearance had rapid improvement in alveolar-arterial oxygen difference (AaPO₂) over the first 10 h, whereas Patients 2 and 4 had persistent hypoxemia (Figure 3). The presence or absence of clearance was not associated with any change in ventilator parameters such as positive end-expiratory pressure (PEEP). All patients were ventilated postoperatively in an intermittent mandatory ventilation mode. PEEP levels ranged from 5 to 8 cm H₂O and the level of PEEP was not changed by more than 3 cm H₂O in any patient during the time of sequential edema fluid sampling. There was a trend toward more prolonged ICU stay (p = 0.10) and longer duration of mechanical ventilation in patients with no net alveolar fluid clearance (p = 0.058).

View larger version (20K): [in this window] [in a new window]   Figure 2.   Sequential edema fluid protein measurements in six patients with reperfusion edema after orthotopic lung transplantation. Collection of pulmonary edema fluid began immediately after the onset of clinically apparent reperfusion edema. Note that net alveolar fluid clearance occurred in four patients (open symbols) with a rise in protein concentration over 4 h, whereas there was no net alveolar fluid clearance in two patients (closed symbols) over 11 h.

View larger version (11K): [in this window] [in a new window]   Figure 3.   Time course of mean AaPO2 after the onset of reperfusion pulmonary edema. Comparison of mean AaPO2 in four patients with intact alveolar epithelial fluid clearance (open squares) to the patients with no net alveolar epithelial fluid clearance (solid squares). The data for Patients 3, 5, 6, and 8 are expressed as mean ± standard deviation. The data for Patients 2 and 4 are expressed as the average of the AaPO2 at each time point.

Charts were reviewed for the use of medications that may upregulate alveolar epithelial fluid clearance, including -agonists, dopamine, dobutamine, and glucocorticoids (22). All patients except Patient 6 received postoperative dopamine at low doses (3 to 10 µg/kg/min) and all patients received high-dose intravenous methylprednisolone (250 to 450 mg/24 h). There was no correlation of alveolar fluid clearance with use or dosage of inhaled -agonists, which were administered to six of the eight patients. Patients 2, 3, and 4 received a pure -agonist, neosynephrine, in the first 24 h postoperatively for hypotension. Interestingly, the fastest rate of clearance (32%/h) occurred in a patient who received dobutamine, Patient 4. However, the other patient who received dobutamine, Patient 2, had no net alveolar fluid clearance.

Cardiac and Hemodynamic Data

Normal left ventricular systolic function was documented in seven of eight patients by a combination of preoperative echocardiogram or gated nuclear medicine study (n = 3), intraoperative transesophageal echocardiogram (n = 3), or postoperative transthoracic echocardiogram (n = 2). Only one patient (Patient 7) had no cardiac imaging at any point in her clinical course, but her pulmonary artery wedge pressure (Ppaw) was 12 mm Hg at the time of edema fluid sampling (Table 1). Hemodynamic monitoring also confirmed the absence of left ventricular dysfunction as a cause of pulmonary edema in this group of patients. Three patients had pulmonary artery catheters inserted with Ppaw of less than 12 mm Hg at the time of edema fluid sampling (Table 1). Central venous pressure was less than 12 mm Hg at the time of the onset of reperfusion edema in all but Patients 2 and 3.

Examination of serial hemodynamic measurements showed that resolution of alveolar edema was not coincident with a major decline in pulmonary arterial wedge or central venous pressures (Pcv) (Table 2). Diuresis or fluid overload also did not appear to be responsible for the presence or absence of resolution of alveolar edema. While four of the eight patients received furosemide (20 to 80 mg) during the first 24 h after surgery, only one patient (Patient 8) had a negative fluid balance (output > input by 1,550 ml), and this patient had a stable Pcv during the period of resolution of pulmonary edema (Table 2). Of the remaining patients, five had balanced fluid intake and output during the first 24 h after transplantation. Patient 5, the patient with the highest rate of alveolar fluid clearance, had a modestly positive fluid balance postoperatively (intake > output by 650 ml). Patient 2 had intake greater than output of 3,000 ml during the first postoperative day in the setting of fluid resuscitation for postoperative bleeding and hypotension. No patient received amiloride, a diuretic which has been shown experimentally to inhibit alveolar epithelial fluid transport (19, 24).

Radiologic Data

All six patients had patchy or diffuse infiltrates in the transplanted lung or lungs consistent with reperfusion edema. Review of serial chest radiographs showed that improvement in radiographic infiltrates lagged behind clinical improvement, but was usually evident by postoperative Day 2 or 3 (Figure 4A and 4B). However, in the patients who had prolonged clinical reperfusion edema and no net alveolar fluid clearance (Patients 2 and 4), the chest radiograph did not improve during the first three postoperative days (Figure 4C and 4D).

View larger version (148K): [in this window] [in a new window]   View larger version (144K): [in this window] [in a new window]   View larger version (140K): [in this window] [in a new window]   View larger version (128K): [in this window] [in a new window]   Figure 4.   Serial anteroposterior chest radiographs of two patients with reperfusion edema after lung transplantation. (A and B) Radiographs taken of Patient 8 (no net alveolar fluid clearance) on postoperative Days 0 and 2, respectively, after bilateral sequential lung transplantation. The initial radiograph (A) demonstrates patchy alveolar infiltrates bilaterally that have decreased substantially by Day 2 with interval removal of the endotracheal tube (B). (C and D) Patient 4 (intact alveolar fluid clearance) on postoperative Days 0 and 2, respectively. The initial radiograph (C ) demonstrates dense infiltrates on the left, the side of the allograft, with hyperinflation on the right. On Day 2 (D) there is no improvement in the diffuse infiltrates, a radiographic finding that correlated with lack of clinical improvement.

Pathologic Specimens

Transbronchial biopsies were obtained within the first 7 d after transplant in five of the eight patients as part of the clinical lung transplant protocol. Four of the five biopsies showed only mild abnormalities including focal organization, focal type II cell hyperplasia, and absent or minimal neutrophil infiltration. These findings were seen in patients who had rapid clinical improvement of their reperfusion pulmonary edema (Patients 1, 3, 6, and 7). A representative biopsy is shown in Figure 5A. By contrast, in one of the patients who had no measurable net alveolar fluid clearance and prolonged reperfusion edema (Patient 4), there was pathologic evidence of diffuse alveolar damage with intraalveolar edema and extensive neutrophil infiltration on a biopsy obtained 6 d after transplantation (Figure 5B).

View larger version (115K): [in this window] [in a new window]   View larger version (130K): [in this window] [in a new window]   Figure 5.   Transbronchial biopsies of pulmonary allograft obtained 6 d after lung transplantation. (A) Biopsy from Patient 3 who had rapid clinical resolution of reperfusion edema and intact alveolar fluid clearance. There is minimal neutrophilic infiltration with mild focal alveolar epithelial type II cell hyperplasia. (B) Biopsy from Patient 4, who had prolonged clinical reperfusion edema without evidence of net alveolar fluid clearance. There is extensive interstitial inflammatory infiltrate, intraalveolar edema, and fibrin consistent with resolving diffuse alveolar damage. Bar = 200 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The goal of this study was to better define the function of the alveolar epithelium in both the initial and resolution phases of severe reperfusion edema after orthotopic lung transplantation. The first objective was to test the hypothesis that reperfusion pulmonary edema after lung transplantation is due to increased permeability of the alveolar-capillary barrier. This was accomplished by measuring the ratio of pulmonary edema fluid to plasma protein concentration at the onset of reperfusion edema. The second objective was to study the alveolar fluid transport capacity of the grafted lung by measuring protein concentration in serial pulmonary edema fluid samples. The third objective was to test for a correlation between clinical parameters and the presence or absence of alveolar fluid clearance.

Summary of Results

The major findings of the study can be summarized as follows. First, a high initial edema fluid to plasma protein ratio was found in six of eight patients, proving that the usual cause of pulmonary edema after lung transplantation is increased permeability. This is the first report to directly confirm increased permeability edema in this setting. Second, the degree of permeability correlated well with graft ischemic time. This lends support to the observation by some investigators that longer ischemic times may be associated with reperfusion lung injury (3, 7). However, the degree of alveolar-capillary barrier permeability, as measured by edema fluid to plasma protein ratio, did not correlate with clinical severity of reperfusion injury as measured by degree of hypoxemia or duration of mechanical ventilation. Third, despite the clear evidence of increased alveolar-capillary barrier permeability, alveolar epithelial fluid transport was preserved in four of the six patients in whom it was measured. This is a remarkable finding, particularly since all patients had evidence of severe reperfusion edema with radiographic infiltrates and the need for a high FI_O2. Finally, preservation of net alveolar fluid transport did correlate with more rapid improvement in oxygenation, a decrease in radiographic infiltrates, and less severe lung injury, as assessed by direct tissue histology. These findings indicate that intact alveolar epithelial fluid transport is critically important to a timely recovery from post-transplantation reperfusion pulmonary edema.

Pulmonary Edema Fluid Sampling

The primary method used in this study, collection of undiluted pulmonary edema fluid, is a useful method to study mechanisms of human pulmonary edema. When sampling is done early after an insult, this method allows differentiation of edema due to hydrostatic causes from increased permeability of the alveolar-capillary barrier (12, 16). A pulmonary edema fluid to plasma protein ratio less than 0.65 is consistent with edema from hydrostatic causes, whereas a ratio greater that 0.75 is consistent with increased permeability edema (12). In the past, we have used this technique in clinical studies to differentiate patients with pulmonary edema caused by congestive heart failure from those with acute lung injury (12), and to better characterize the mechanism underlying neurogenic pulmonary edema (16). Other investigators have also confirmed the utility of this technique (17, 18).

In the current study, the initial pulmonary edema fluid samples were collected from patients at the onset of reperfusion edema after lung transplantation. The timing of the first sample, immediately upon development of reperfusion edema, is important, because alveolar fluid clearance will result in an increase in edema fluid protein concentration over time (12). If the first edema fluid sample were obtained many hours after the onset of reperfusion edema, a high edema fluid to plasma protein ratio might reflect an increase in protein concentration from alveolar fluid transport, rather than increased permeability. In this sense, lung transplant patients represent an ideal study population because unlike usual clinical acute lung injury, a syndrome with multiple and sometimes unknown or unrecognized causes, the mechanism of injury after lung transplantation is uniform, and the time of onset can be pinpointed accurately, almost always occurring within the first 24 h after transplantation.

Alveolar-Capillary Barrier Permeability

In six of the eight patients studied, the initial edema to plasma protein ratio was greater than 0.75, consistent with increased permeability edema. The initial samples were obtained at zero to 19 h after ICU admission, as soon as reperfusion edema was clinically evident. There was no correlation between the time from ICU admission to the first sample and the edema to plasma protein ratio (r = 0.003), indicating that alveolar fluid transport prior to initial sampling is insufficient to explain the high ratios. Patient 8 did have an edema fluid to plasma ratio greater than one, which suggests a component of early transport and concentrating of alveolar protein before sampling. However, it is unlikely that this patient initially had hydrostatic edema with a low ratio (< 0.65) because edema fluid was sampled within 90 min of ICU admission postoperatively, and the remainder of the clinical data in this patient, including hemodynamic parameters, did not indicate an increase in left atrial pressure.

The finding of a high edema fluid to plasma protein ratio in the majority of patients with reperfusion edema confirms that reperfusion edema after lung transplantation is usually the result of increased permeability of the alveolar-capillary barrier. To our knowledge, this is the first study to definitively make this measurement. A previous study of patients with reperfusion edema, or pulmonary reimplantation response, sampled pulmonary edema fluid in five of 40 patients who were studied (27). However, the investigators only reported that the edema fluid to plasma protein ratio was > 0.5 in all five patients, a finding that could be consistent with either hydrostatic or increased permeability pulmonary edema (12). Pulmonary endothelial permeability has also been measured indirectly after lung transplantation using a double isotope technique. In a study of 10 patients within 36 h of lung transplantation the protein accumulation index, as measured by scintillation counting over the chest, was nearly threefold higher in transplant patients compared with normal control subjects (28).

Although this study provides clear evidence of increased endothelial permeability in the setting of reperfusion injury, it is not possible to directly quantify the amount of alveolar epithelial injury and permeability. It seems likely, based on experimental evidence, that endothelial injury is the first step in ischemia reperfusion injury (1, 8). Alveolar flooding could then occur by one of two mechanisms, either direct alveolar epithelial injury, or bulk flow of fluid from the interstitium into the airspaces resulting from accumulation of large quantities of interstitial edema (29, 30). The finding that alveolar epithelial fluid clearance is intact in the majority of patients supports the hypothesis that the injury to the alveolar epithelium was not severe in most cases of reperfusion edema, despite a marked increase in lung endothelial permeability. Indeed, the rapid clearance that occurred in many patients implies not only a functionally intact epithelium, but also suggests that endothelial injury may be self-limited; a decrease in the magnitude of transendothelial fluid flux may also have contributed to the rapid alveolar fluid clearance.

Cell Count Analysis

Cell count analysis of pulmonary edema fluid was also consistent with increased permeability edema. Two previous studies have compared pulmonary edema fluid cell counts in patients with hydrostatic edema to patients with acute lung injury (31, 32). In patients with hydrostatic pulmonary edema, edema fluid white blood cell counts ranged from 162 to 1,800 leukocytes/µl with 4 to 33% neutrophils. Patients with acute lung injury had more cellular edema fluid and a marked predominance of neutrophils, with 1,388 to 296,000 leukocytes/µl and 45 to 93% neutrophils. In the current study, edema fluid cell counts were 1,013 to 66,000 leukocytes/µl with 71 to 99% neutrophils, falling within the range described for acute lung injury. Interestingly, Patient 4, the patient with the most severe neutrophil infiltration histologically, also had the highest edema fluid neutrophil count (99%), and no net alveolar fluid clearance.

Correlation with Graft Ischemia Time

There has been controversy in the literature as to whether increased ischemic time is associated with increased risk of reperfusion injury and primary graft failure. One study from the University of Minnesota found no correlation between ischemic time and early or late outcome with ischemic times as long as 11.8 h (33). A small study from the University of Pittsburgh found no significant difference in ischemic time between five patients with severe primary graft failure and a group of lung transplant recipients who did well (34). More recently, a retrospective study from the University of Pennsylvania of 100 consecutive lung transplant patients found no difference in graft ischemic time between the 15 patients who developed primary graft failure and the 85 who did not (2). Several other studies have reported adverse outcomes with ischemic times over 4 to 5 h (7, 27). The current practice in most centers is to avoid ischemic times over 6 to 8 h (8). In this study, the graft ischemic times ranged from 2.3 to 6.7 h, and graft ischemic time did not correlate with severity of hypoxemia, duration of mechanical ventilation, or duration of ICU stay. Furthermore, graft ischemic time did not correlate with the presence or absence of alveolar fluid clearance.

There was, however, a strong correlation between graft ischemic time and the initial edema fluid to plasma protein ratio, with longer ischemic times associated with increased permeability. This lends support to the hypothesis that a longer duration of ischemia can lead to increased lung endothelial injury and permeability. However, since no comparable data are available in patients after lung transplantation who did not develop reperfusion edema, it is not possible to comment on any association between graft ischemic time and overall risk for development of reperfusion edema. Undoubtedly, factors other than endothelial protein permeability influence both the formation and resolution of pulmonary edema after lung transplantation. These factors include the ability to reabsorb pulmonary edema (discussed subsequently), the degree of epithelial injury, and other clinical problems related to lung transplantation including infection and rejection. In the current group, factors contributing to prolonged intubation included infection, bronchospasm, atrial fibrillation, pulmonary embolism, and encephalopathy.

Resolution of Pulmonary Edema

Despite a significant increase in alveolar-capillary barrier permeability to protein, the alveolar epithelial fluid transport capacity was remarkably intact in the majority of patients. In the group of six patients who had sequential edema fluid sampling, four of the six patients had intact alveolar fluid clearance, which correlated with improvement in oxygenation, radiographic infiltrates, and less severe lung injury histologically. The two patients who had no net alveolar fluid clearance had prolonged hypoxia, persistent radiographic infiltrates, and in one patient, diffuse alveolar damage on biopsy. There was a trend toward shorter duration of mechanical ventilation and ICU stay in the patients with intact clearance. Overall, these results suggest that the ability to transport fluid from the alveolar space is more important to recovery from reperfusion pulmonary edema than the initial degree of endothelial injury.

We have previously used this same method to study alveolar fluid clearance in patients with acute lung injury (12). In that group, the ability to concentrate protein in the alveolar space, a direct marker of alveolar fluid clearance, correlated with clinical improvement and survival. Lung injury in that study was predominantly from sepsis, drug toxicity, aspiration, and cardiopulmonary bypass. The present study supports the extension of these findings to another cause of acute lung injury, reperfusion lung injury.

Interestingly, the measured rates of alveolar fluid clearance were similar to previous data in patients with other forms of acute lung injury. In our prior study of patients with acute lung injury, mean alveolar liquid clearance was 18 ± 15% (12), faster than basal rates seen in dogs (4 ± 2%/h) (19), sheep (8 ± 3%/h) (14), and similar to basal clearance in rabbits (15 ± 4%/h) (35). In the ex vivo human lung, the rewarmed lung had a basal alveolar liquid clearance rate of 3 ± 1%/h (13, 23). Overall, 67% of the patients in this study had intact alveolar fluid clearance. This is a higher percentage than previously reported in a group of 18 patients with acute lung injury (47%) (12), but lower than previous observations in two separate studies of patients with hydrostatic pulmonary edema wherein 78 to 79% of patients had intact clearance (12, 36).

The mechanisms responsible for the high rates of alveolar fluid clearance are not known, but may include catecholamine-dependent and -independent effects from medications such as glucocorticoids, -agonists, and vasopressors, as well as elevated levels of endogenous catecholamines. All of these agents have been shown in experimental studies to increase vectorial sodium and fluid transport by the alveolar epithelium (22, 37, 38). In view of the small number of patients, it is not possible to make definitive conclusions about the contribution of these drugs. It is interesting to note, however, that all patients received at least one, and often two or more medications, that have been shown experimentally to increase alveolar fluid transport. All but one patient received dopamine, a combined - and -adrenergic receptor agonist that has been reported in one experimental model to increase alveolar fluid clearance (22). All patients received high doses of glucocorticoids that have also been shown experimentally to increase alveolar fluid clearance (25, 26). One of the patients who received dobutamine had an extremely high rate of clearance, but the other did not. Dobutamine has also recently been reported in an experimental rat model in our laboratory to increase alveolar fluid clearance (24). Endogenous catecholamines were not measured in this study, but also may have been increased in the post-transplantation setting.

The finding of intact alveolar epithelial fluid clearance in the majority of patients with reperfusion pulmonary edema seems to concur with the very limited experimental data available on the transport function of the alveolar epithelium in the setting of ischemia reperfusion injury. In the ex vivo human lung, cooling to 4° C followed by rewarming had no effect on alveolar fluid clearance in the rewarmed lung (13). Both alveolar fluid clearance and Na-K-ATPase activity were intact in rat lungs preserved (but not reperfused) for 24 h in an extracellular type phosphate-buffered solution but not with Euro-Collins solution (39, 40). The lack of intact alveolar epithelial fluid clearance, as seen in two patients in this study, may reflect oxidant-mediated modulation of alveolar fluid clearance. Unpublished data from our laboratory have shown that oxidant-dependent mechanisms may modulate alveolar fluid clearance in rat and rabbit models of ischemia reperfusion injury (Sakuma and colleagues).

Effect of Cardiac and Fluid Balance on Pulmonary Edema

The possible contribution of fluid balance and cardiac function to the findings in this study were considered carefully. It appears unlikely that fluid overload contributed substantially to the development of pulmonary edema in these patients. Measurements of the Ppaw and the Pcv were not consistent with fluid overload or elevated left atrial pressure. Furthermore, cardiac imaging consistently showed normal left ventricular function. In addition, the high rates of alveolar fluid clearance did not coincide with a large decline in Pcv, nor with negative fluid balance. Similarly, there was no correlation between changes in ventilator settings such as the level of PEEP and the resolution of pulmonary edema.

Conclusions

In conclusion, the finding of intact and rapid alveolar fluid clearance in a group of patients with reperfusion pulmonary edema confirms that a functionally intact alveolar epithelium may be preserved despite clinically severe reperfusion lung injury. Furthermore, alveolar epithelial fluid transport was preserved in most patients despite evidence of marked lung endothelial injury with a substantial increase in alveolar-capillary barrier permeability. These findings indicate that the alveolar epithelium appears to be more resistant than the endothelium to injury in the setting of lung ischemia and reperfusion during transplantation. In addition, an intact functional epithelium is critically important for the resolution of reperfusion edema. Intact alveolar fluid clearance was correlated with less histologic injury, rapid resolution of hypoxemia, more rapid resolution of radiographic infiltrates, and a trend toward a shorter duration of mechanical ventilation and a shorter ICU stay. Measurement of alveolar fluid clearance may be useful clinically to assess alveolar epithelial transport function and predict outcome in patients with reperfusion lung injury after lung transplantation.