INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: lung transplantation; bronchial arteries; reperfusion injury; bronchoalveolar lavage

Current protocol for lung transplantation does not reinstate the systemic circulation of the graft through bronchial artery revascularization (BAR). Recently there has been increased concern that systemic bronchial circulation is relevant for survival of the transplanted lung and that reanastomosis of the bronchial vessels might alter ischemia-reperfusion injury posttransplant (1).

Physiologically, the systemic bronchial arteries receive 3 to 5% of cardiac output, which is low compared with pulmonal artery blood flow. Nevertheless, this small arterial flow is of significant importance for the vital airway defense, fluid balance, and metabolic functions of the lung (4). The bronchial arteries arise from intercostal arteries to follow the airways into the pulmonal tissue. Besides supplying the airways, they deliver nutritive supply to lymph nodes, act as the vasa vasorum to the pulmonary vessels, and finally form bronchopulmonary anastomoses along the alveoli (5). Bronchial blood flow increases manifold in response to lung injury, such as inflammation or pulmonary vascular obstruction (4, 6, 7). Lung injury induced by ischemia and reperfusion is characterized by increased permeability of lung microvessels, pulmonary hypertension, and diminished surfactant production (8). Attenuation of this lung injury by function-going bronchopulmonary anastomoses has been observed in acute and chronic obstruction of the pulmonary artery (12, 13).

Within this background, we tested the hypothesis that BAR in lung transplantation may reduce ischemia-reperfusion injury of the graft. The two different operational concepts (left lung allotransplantation with or without BAR) were compared in a standardized left lung allotransplantation model in beagle dogs, concerning their influence on enzyme activities in bronchoalveolar lavage fluid (BALF). Cell type-specific enzyme activities were taken as markers of ischemia of pulmonary endothelium (angiotensin-converting enzyme [ACE]), pneumocytes type I (carboxypeptidase M [CPM]), and pneumocytes type II (alkaline phosphatase [AP]) over a period of 5 h after transplantation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental Model

Twelve beagle dogs underwent left lung allotransplantation from 12 donor dogs. In six recipients, transplantation was carried out according to clinical standard without BAR (LTX group); in the other six recipients, transplantation was performed with bronchial arterial revascularization (BAR group). Sedation and induction of anesthesia were followed by intubation, intravenous neuroleptanalgesia, and volume-controlled ventilation at a fraction of inspired oxgen (FI_O2) of 100% and an additional positive end-expiratory pressure (PEEP) of 8 mm Hg after transplantation. All animals received human care in compliance with the European Convention on Animal Care throughout the experiment.

Surgical Procedure

After median sternotomy and systemic anticoagulation (300 IU heparin/kg), both donor lungs were preserved by flush perfusion with 4° C cold Euro-Collins solution using a catheter placed in the pulmonary artery. The thoracic organs were excised en bloc and the left lung was separated for transplantation. Grafts were stored inflated at 4° C until transplantation.

Recipients underwent a left posterolateral thoracotomy to excise the left lung and replace it by a donor graft. To prevent thromboembolic complications, 300 IU/kg heparin were instilled 15 min before reperfusion. After identical surgical procedures and a total ischemic time of 6 h (4 h ± 15 min cold ischemia, 2 h ± 15 min warm ischemia) in both groups, the grafts were reperfused. Consecutively in the BAR group, functional bronchial arteries were identified by bronchoarterial backflow (14) and revascularized by patch anastomosis onto the recipient's aorta within 30 ± 5 min after start of reperfusion.

Control Parameters

Monitoring of systemic and pulmonal pressures, cardiac output, pulmonary artery blood flow, blood gas analysis, and hemoglobin content was performed throughout the experiment. Blood samples were taken from the left atrium in donors before surgery (Control), in recipients before reperfusion, and from a catheter out of the left pulmonary vein in recipients during graft reperfusion, to obtain ACE activity in blood.

To perform bronchoalveolar lavage (BAL), the bronchial system was entered by a bronchoscope (Storz, Germany) under sight to reach wedge position, instilling and removing warm saline in four 20-ml aliquots to recover at least 50% of the instilled volume. The right middle lobe of donors (control values) and of the recipients was lavaged approximately 45 min before reperfusion (values before reperfusion). Two more BALs were performed at 2 and 4 h of reperfusion in adjacent segments of the lingula of the transplanted lung, corresponding to the lavaged segments of the right lung. Samples were centrifuged at 500 g for 10 min to sediment cellular material, and supernatants were frozen at 80° C. Lactate dehydrogenase (LDH) and AP were determined automatically by Synchron-LX-Systems (Beckman Instruments GmbH, Unterschleissheim, Germany). ACE activity was determined using a test kit (ACE-Color FS 116; MAST-Diagnostika, Reinfeld, Germany). CPM activity was determined fluorospectrophotometrically by the method of Tan and coworkers (15).

Statistical Analysis

Data are given as mean ± SD. Data that passed the double-sided nonparametric U test with p < 0.05 were considered as statistically significant. Figures 1-4 are displayed as box-whisker plots, including median, 5th, 25th, 75th, and 95th percentiles.

View larger version (29K): [in this window] [in a new window]   Figure 1.   LDH activity in BALF was significantly higher in both groups at 2 h and 4 h of reperfusion compared with control values and values before reperfusion (RP) (#p < 0.01). Results are displayed as box-whisker plots, including median, 5th, 25th, 75th, and 95th percentiles.

View larger version (28K): [in this window] [in a new window]   Figure 2.   CPM activity as a marker for pneumocyte type I injury was significantly elevated in the BAR group and the LTX group during RP compared with control values (#p < 0.01). Results are displayed as box-whisker plots, including median, 5th, 25th, 75th, and 95th percentiles.

View larger version (23K): [in this window] [in a new window]   Figure 3.   AP activity in BALF after 2 h of RP was significantly elevated in the LTX group compared with the BAR group (#p < 0.05) and with control values (*p < 0.01). Results are displayed as box-whisker plots, including median, 5th, 25th, 75th, and 95th percentiles.

View larger version (25K): [in this window] [in a new window]   Figure 4.   Conventionally transplanted animals of the LTX group showed significantly higher ACE activity in BALF after 2 h of RP compared with the BAR group (*p < 0.05) and with control values (#2 h: p < 0.01). Results are displayed as box-whisker plots, including median, 5th, 25th, 75th, and 95th percentiles.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Table 1 compares the BAR and LTX groups with respect to pulmonary and systemic pressures as well as cardiac output, blood gas analysis, and pulmonal arterial flow of the transplanted lung during reperfusion. No statistically significant differences have been observed between the two groups.

The activity of LDH in BAL (Figure 1) rose significantly in both groups within 2 h of reperfusion compared with control values (control: 7.71 ± 8.3 IU/L; BAR: 60.14 ± 63.5 IU/L; LTX: 67.0 ± 69.1 IU/L); p < 0.01). Thereafter, LDH activity declined, but at 4 h of reperfusion was still significantly above control values (control: 7.71 ± 8.3 IU/L; BAR 49.14 ± 35.7 IU/L; LTX: 32.14 ± 34.3 IU/L; p < 0.01). CPM activity in BALF (Figure 2) increased significantly after 2 and 4 h of reperfusion in both groups compared with control values (p < 0.01). No statistically significant differences in LDH and CPM activity have been found between BAR and LTX.

After 2 h of reperfusion, AP activity in BAL (Figure 3) was significantly increased in the LTX group compared with the BAR group and control values (LTX: 68.17 ± 19.1 IU/L; BAR: 29.17 ± 16.9 IU/L; control: 14.00 ± 12.9 IU/L; p < 0.05). After 4 h of reperfusion, AP activities in LTX animals were still elevated, but decreased from the 2 h values. ACE activity in BAL (Figure 4) after 2 h of reperfusion showed a significant increase in the LTX group (1.39 ± 1.1 IU/L) compared with the BAR group (0.35 ± 0.6 IU/L; p < 0.05) as well as compared with control values (0.25 ± 0.6 IU/L; p < 0.01). ACE activity in pulmonary venous blood of both groups (Figure 5) was significantly higher at reperfusion compared with control values (p < 0.05). After reperfusion, ACE activity decreased continuously in both groups and was significantly lower than control values at 3, 4, and 5 h of reperfusion (p < 0.05), without any statistically significant difference between the two groups.

View larger version (42K): [in this window] [in a new window]   Figure 5.   ACE activity in pulmonary venous blood of the transplanted left lung (left PV) was significantly reduced after 180, 240, and 300 min of RP and significantly higher at start of RP compared with control values (*p < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchial arteries follow the airways as far as the bronchioles, forming a peribronchial and a submucosal plexus. They also supply the visceral pleura and the lymph nodes and act as vasa vasorum for pulmonary arteries and veins (4). Pump described bronchopulmonary connections between the bronchial microvasculature and the pulmonary microvessels in the alveolar capillary bed (5). As pulmonary hypertension and pulmonary edema both are severe side effects in the early period after lung transplantation, it cannot be excluded that the additional nutritive supply of pulmonary endothelium, as well as alveolar epithelium, by bronchial arteries might be of significant importance to limit lung tissue damage after transplantation.

To evaluate this hypothesis, we monitored lung tissue damage during the early period after transplantation, by analyzing marker enzyme activities in BAL supernatants in LTX animals compared with BAR animals. The two experimental groups were handled in the same manner regarding surgical procedures, duration of ischemia, and implantation time. The en bloc resection of thoracic organs during explantation, the atraumatic preparation, and the identification of bronchial arteries by bronchoarterial backflow after transplantation, ensured a viable and functional bronchial circulation after reanastomosis on the descending aorta in the BAR group (14). The standardization of experimental procedures resulted in stable and comparable systemic and pulmonary circulation data (Table 1).

BAL was performed to assess lung injury in a cell-specific but nondetrimental manner. The determination of acellular components in BALF cannot be precisely standardized, and no satisfactory marker for the dilution factor during lavage has been found so far (16). A dilution factor such as albumin or urea has not been shown to be useful because of the damaged blood-air barrier and pulmonary edema after transplantation. Subjects of equal weight were used to perform BAL in corresponding compartments of the lung, which supports standardization.

Enzyme activities in the acellular component of BALF, which are commonly found inside cells, like AP and LDH, are considered to be sensitive markers for disruption of cellular integrity (17). Although determination of common LDH activity in blood samples of patients is rather unspecific, LDH in BAL is a valuable parameter to survey lung injury (17). CPM activity in BALF has been observed to be a characteristic marker for pneumocyte type I injury (18). Dragovic and coworkers observed, by immunoprecipitation studies with antisera, that all carboxypeptidase activity inhibited by 2-mercaptomethyl-3-guanidinoethylthiopropanoic acid (MGTA) in BAL is due to the presence of authentic CPM (19).

AP activity in BAL is mainly produced in the lung itself and was taken as a specific marker for pneumocytes type II damage (17, 20, 21). ACE, a dipeptidylcarboxypeptidase, located membrane bound on the luminal side of pulmonary endothelium, and increased concentrations in BAL are a marker for pulmonary endothelial cell injury (22, 23). Ischemia-reperfusion injury of the lung has been reported to accelerate shedding of ACE from pulmonary endothelium (23), which provoked us to carry out measurements in BAL and pulmonary venous blood.

Looking at the results from this aspect, reperfusion after preservation and ischemia induces a release of unspecific as well as specific enzymes from pulmonary tissue, which has been observed by analysis of BAL during reperfusion. A significant augmentation of marker enzymes for lung ischemia (LDH) and pneumocytes type I injury (CPM) was found in BAL of both groups during reperfusion (p < 0.01; Figures 1 and 2). Although operational trauma to the lung was higher in the BAR group owing to the patch anastomosis, LDH and CPM activities did not differ significantly between the two groups.

However, if reperfusion through the pulmonal arteries was assisted by BAR, a specific effect on ACE and AP activities in BAL was found. Both enzyme activitiesAP as the marker for pneumocytes type II and ACE as the marker for pulmonary vasculaturewere significantly lower in the BAR group than in the LTX group after 2 h of reperfusion (Figures 3 and 4). Both cell types are especially altered during ischemia and reperfusion of the lung (24). The acute damage of type II pneumocytes found in the LTX group corresponds to the morphologic findings of Fehrenbach and coworkers, which demonstrated the sensitivity of these cells to ischemia (10).

ACE activity in left pulmonary venous blood peaked at the start of reperfusion and decreased significantly in both observed groups during reperfusion compared with control values (Figure 5). Loss of serum ACE activity has already been observed by Bedrossian and coworkers in patients with adult respiratory distress syndrome (25). We presume that ACE activity in serum rose because of release into the pulmonal vascular system and fell after the start of reperfusion owing to dilution in the peripheral blood. In contrast to the findings in BAL, there were no significant differences of serum ACE activity during reperfusion between the two groups, which might be due to the overall dilution in the animals. The significantly lower ACE activity in BAL after 2 h of reperfusion in the BAR group might be a direct result of clearing by bronchial arterial perfusion of the pulmonary endothelium. Nevertheless, ACE in lung tissue can also be produced by pulmonary macrophages (22). If this was true, a lower degree of macrophage activation after BAR could as well result in minor onset rejections after transplantation, as the macrophages seem to play a key role in the initial phase of tissue injury and in the repair phase (26).

All these results may imply some clinical relevancy, as a higher degree of ischemia-reperfusion injury appears to make lung tissue more predisposed for acute periods of rejection in the follow-up of transplantation (27, 28). Rejection episodes are a major risk factor for a later development of chronic transplant failure (29). In recent years, evidence has emerged that chronic airway ischemia or hypoxia may contribute to the development of bronchiolitis obliterans syndrome (11, 30). Discussion still persists whether BAR diminishes ischemia- reperfusion injury and ultimately reduces the development of a bronchiolitis obliterans syndrome (1, 31). Maclean and coworkers recently demonstrated that inhibition of ACE from Day 1 after transplantation could limit airway obliteration in a rat model of bronchiolitis obliterans (32).

In conclusion, BAR in lung transplantation has a positive impact on BAL concentrations of injury marker enzymes for pneumocytes type II and pulmonary endothelium in the early postischemic period after lung transplantation. This result could favor the hypothesis that the bronchial circulation could compensate pulmonal arterial perfusion deficit by using bronchopulmonary anastomoses (4, 13). Whether BAR possibly affects long-term function of the transplanted lung and delays onset of bronchiolitis obliterans syndrome by limiting ischemia- reperfusion injury of the graft needs further study.