Several clinical situations require subjecting lung tissue to more or less prolonged periods of ischemia, with the consequent risk of acute lung injury after reperfusion. Most of the studies on how to preserve the lung tissue from the effects of ischemia-reperfusion (IR) are done in experimental or clinical models of lung transplantation with cold ischemia.1 There are clinical situations, however, in which the progressive cooling of the lung is not possible, nor is it possible to perfuse it with a preservation solution before interrupting blood circulation. Among these situations are lung resections with angioplasty of the pulmonary artery2–4 and lung lobe transplants from living donors. Less frequent, almost anecdotal situations are the ex situ pulmonary resections of central tumors with reimplantation of the viable pulmonary lobe or lobes.5 In these cases, part of the lung tissue suffers a more or less prolonged period of warm (normothermic) ischemia, while reperfusion edema and the need for prolonged post-op ventilation are frequent.6

In this context, the study of lung injury due to IR and potential modulating therapies requires experimental models in which ischemia is initiated without cooling or previous lung preservation. An experimental model reproducing this situation is lung autotransplantation in animals. The objective of this present study is to present a model of lung autotransplantation in pigs in order to study the earliest phases of the IR syndrome and analyze the hemodynamic, gasometric and biochemical changes that take place during this period.

The study was done with the approval of the Animal Experiment and Research Committee of the institution, following all times the European and Spanish guidelines for the manipulation and care of experimental animals.

In six large-white pigs, we carried out a procedure of orthotopic lung autotransplantation. Mean animal weight was 42.8kg. The intervention consisted of left pneumonectomy with cranial lobectomy ex situ, reimplantation of the caudal lobe and reperfusion for 30min. Mean procedure time was 289min (range: 232–325min) and the mean pulmonary ischemia time was 90min (range: 84–97min). At the beginning of the procedure and during the periods of ischemia and reperfusion, different parameters were analyzed to identify hemodynamic, blood gas and biochemical changes in the model.

Standard experimental models that study pulmonary IR are usually associated with the progressive cooling of the organ, perfusion with preservation solutions or extracorporeal circulation. Therefore, they are not ideal for the study of situations like the reconstruction of the pulmonary artery or lobe transplant from a living donor, in which the organ is suddenly subjected to normothermic or warm ischemia. It is necessary to develop experimental models in order to study therapies that reduce the risk of acute lung injury in these patients. The majority of the research on pulmonary warm ischemia is done in small rodents by means of vascular occlusion techniques11,12 and less frequently include surgical procedures similar to those carried out in humans. Pulmonary autotransplantation in large mammals largely meets this objective, as reported in dogs,13–17 sheep18 and pigs,19 but few of these studies thoroughly analyze the inflammatory mediators during normothermic IR.20

In this present study, we analyzed the hemodynamic, blood gas and biochemical changes during pulmonary autotransplantation procedures in pigs. Van Raemdonck et al. observed that the tolerance to warm ischemia in non-ventilated rabbit lungs was 1h.21 Yamazaki et al., in a canine model of pulmonary autotransplantation, set the limit of 120min as the maximum tolerable time of warm ischemia for the lung to be viable and for the animal to survive.22 The mean time of lung ischemia in our study, 90min, makes for a foreseeable development of a non-lethal lung injury with physiopathological manifestations that allow us to analyze the effect of potential treatments.

In experimental studies of normothermic IR, acute pulmonary injury involves increased pulmonary vascular resistance, pulmonary edema and deterioration of the gas exchange capacity. In the majority of these studies, the ischemia is maintained for 2–3h and these alterations are manifested after several hours of reperfusion.15,23,24 In our model, designed for the study of the early inflammatory response, the different hemodynamic variables were relatively stable throughout the procedure. We observed a significant increase in the pulmonary capillary pressure during the reperfusion and gasometric alterations after the start of the one-lung ventilation and 10min after initiating reperfusion. We consider that this clinical stability confers a greater reliability to the observed biochemical findings.

As for the biochemical mechanisms implicated in the IR injury of any organ, numerous studies signal the production of oxygen free radicals (OFR), the activation of polymorphonuclear leukocytes and the production of proinflammatory cytokines as important mediators in the inflammatory response.

During IR, OFR are generated, which cause cell lysis by lipid peroxidation of the free fatty acids of the membranes.25 The degree of lipid peroxidation, and indirectly the presence of OFR, can be measured by the presence of lipid peroxides (LPO) and tissue MDA. In the present model, a progressive increase of MDA and LPO is observed in the lung tissue, from the final phase of the ischemia until 30min after the reperfusion, which clearly indicates that the increase in tissue oxidative stress is a premature consequence of the IR. Similar results have been observed after normothermic ischemia, including in isolated lungs in rats26 and by using in situ occlusion of the pulmonary artery in dogs.27

In our experiment, we have observed a significant increase in tissue MPO during reperfusion, indicating recruitment and progressive activation of leucocytes in the lung tissue. Today it is known that the role of the neutrophils is important in the late phase of IR, but during the earlier phases the role of the macrophages and lymphocytes predominates.25 Eppinger et al., in a model of warm pulmonary IR in rats, demonstrated that the alterations observed during the first 30min of reperfusion were regardless of the leukocyte activation, whose effects began to become patent after 4h of reperfusion.28 Our data show that this leukocyte infiltration/activation is a phenomenon that begins a few minutes after the reperfusion of the organ.

One of the consequences of the tissue increase of OFR is the activation of the HO/CO system. HO is a microsomal enzyme that catalyzes the limiting step of the degradation of the heme group, converting it into biliverdin, CO and Fe+++. There is growing evidence of the protective role of this enzyme and its metabolites against IR injury.29 The constitutive isoform, HO-2, is expressed in baseline conditions in numerous tissues, while the inducible isoform, HO-1, also called heat shock protein 32, rises in situations of stress such as IR. On the other hand, CO presents as a potent anti-inflammatory and anti-apoptotic molecule, and its effect seems to be mediated by the activation of the mitogen-activated protein kinase (MAPK) pathway.30 Along the lines of these studies, in our study we have observed an increase of tissue HO activity during reperfusion, which indirectly implies a local increase in CO, while the decrease in CO levels in blood could be due to its peripheral consumption upon reacting with OFR and NO. Some authors have described similar results in rats during hepatic31 and pulmonary30 warm IR.

In addition to the activation of the HO/CO system, it has been reported that the OFR stimulate the synthesis and release of different cytokines and chemokines by the tissue macrophages. This activation of the macrophages has been related with the early phase of the inflammatory response to the IR.25,32 IL-1β and TNF-α are proinflammatory cytokines whose implication in the first stages of the inflammatory response to IR is well-known, but recently a crucial role in the process has been attributed to MCP-1, a chemokine that regulates the migration and activation of monocytes and macrophages, although its protective or deleterious function has still not been well defined.32,33 In our experiment, these three molecules are significantly higher in the lung tissue during ischemia, maintaining high levels after 10 and 30min of reperfusion. The progressive increase in TNF-α stands out, which would back the results of other authors which indicate this cytokine as a crucial inflammatory mediator in the initial phase as well as the late phase of IR.34 Our data could be related with a study by Krishnadasan et al., performed in mice, in which an increase in IL-1β and TNF-α is observed in the lung tissue 60min after reperfusion after a period of 90min of warm ischemia.11 Nevertheless, in the same experimental model, MCP-1 is only detected after 4h of pulmonary reperfusion,35 a discrepancy that could be due to the differences in the experimental model or to the sensitivity of the measuring technique.

The alteration of the metabolism of NO has been widely studied in different IR models, but with frequently contradicting conclusions due to the polyvalent effect of NO.25 The NO produced by the constitutive isoforms of the nitric oxide synthase enzyme, NOSe and NOSn, would be responsible for the beneficial physiological effects such as the control of the vascular tone or the aggregation of neutrophils and platelets, while NO synthesized by the inducible isoform, NOSi, would be implicated in physiopathological processes such as IR syndrome.36 We have observed in our experiment a marked decrease in blood NO and less activity of NOSe and NOSn in the lung tissue, both during ischemia as well as during reperfusion. However, there was a significant increase in the NOSi activity. To what extent these changes and their possible therapeutic modification either positively or negatively affect IR injury has still not been clarified and will have to be the object of future studies.

In order to more thoroughly analyze the biomolecular profile of the inflammatory response in this model, we have studied the activation of the MAPK. Recent studies implicate the MAPK as intracellular mediators of the inflammatory response to IR, therefore the final components of this signaling pathway are transferred to the nucleus and activate transcription factors and the expression of certain genes. The MAPK include different families of enzymes and routes of activation, and those that stand out are MAPK p38, JNK and ERK, but the specific role of each is still debated.37 The first two are activated by proinflammatory cytokines like IL-1β and TNF-α, and also OFR.38 It has also been observed that MCP-1 is capable of activating the three signaling routes.39 In addition, in vitro studies (with pulmonary endothelial cells) and in vivo ones (warm pulmonary IR in rats) show that CO has an anti-apoptotic effect during IR and that this effect is mediated by MAPK p38.30 The significant increase in MAPK p38 activity that we detect in the ischemic lung suggests the activation of this intracellular signaling pathway in response to different mediators, such as IL-1β, TNF-α, MCP-1 or CO that are also found to be high at the end of the ischemia period. Likewise, the generation of OFR, evident by the lipid peroxidation products, could be involved in this activation of the MAPK pathway. Future research should clarify to what degree this pathway amplifies and perpetuates the inflammatory response or, on the contrary, plays a protective role against IR.

In summary, we present an experimental procedure of swine lung autotransplantation as a model for studying pulmonary IR syndrome in large mammals, and consequently in humans. We have reported in detail the hemodynamic, blood gas and biochemical changes that may serve as control parameters in the study of new modulating therapies for acute lung injury due to IR. The model, in addition to presenting a great similarity with surgical procedures such as lung angioplasty or lobe transplantation from a living donor, could serve as a stepping stone to the development of lung autotransplantation techniques that, in the future, could offer real hope for surgical treatment in patients with lung cancer and low respiratory functional reserve. The association with cutting-edge procedures, such as ex vivo lung perfusion40 could widen to new extremes the field of investigation of this model.

This study has been supported by funding from the Health-Care Research Fund (Fondo de Investigación Sanitaria - FIS PI070840 and FIS PI070481) and from the Spanish Society of Pulmonology and Thoracic Surgery (SEPAR 2006/121).

The authors declare no conflicts of interest.