PREVENTIVE EFFECT OF INHALED NITRIC OXIDE AND PENTOXIFYLLINE ON ISCHEMIA/REPERFUSION INJURY AFTER LUNG TRANSPLANTATION

Abstract

Background. 

The preventive effect of inhaled nitric oxide (NO) and pentoxifylline (PTX) administered during reperfusion has been demonstrated on experimental models of lung ischemia/reperfusion (I/R) injury but this strategy is not validated in clinical lung transplantation. The aim of this study was to assess retrospectively the protective effect of inhaled NO and PTX after lung transplantation.

Methods. 

Twenty-three consecutive patients who received inhaled NO (10 ppm) and PTX (NO-PTX group) at the time of reperfusion were compared retrospectively with (1) 23 consecutive patients transplanted just before the use of NO-PTX (control group 23); (2) 95 patients representing all the patients of the series who did not receive NO-PTX (control group 95), with respect to I/R injury related complications. In particular, the incidence of pulmonary reimplantation edema and early hemodynamic failure, the PaO₂/FIO₂ ratio as well as the duration of mechanical ventilation and the 2-month mortality rates were compared.

Results. 

Reimplantation edema was observed in 6/23 patients (26%) in the NO-PTX group vs. 13/23 patients (56%) in the control group 23 (P =0.035) and 48/95 patients (50%) in the control group 95 (P =0.035). The worst PaO₂/FIO₂ ratio during the first three postoperative days was 240±102 mmHg in the NO-PTX group vs. 162±88 mmHg (P =0.01) and 176±107 mmHg (P =0.01) in the control group 23 and the control group 95, respectively. The duration of mechanical ventilation was 2.1±2.4 days in the NO-PTX group vs. 7±9 days in the control group 23 (P =0.02) and 6±7 days in the control group 95 (P =0.01). The 2-month mortality rate was 4.3% in the NO-PTX group vs. 26% (P =0.04) and 21% (P =0.07) in the control group 23 and the control group 95, respectively.

Conclusions. 

The marked decrease in the incidence of allograft dysfunction compared with two historical control groups suggests that PTX and inhaled NO given before and throughout reperfusion are protective against I/R injury in the setting of clinical transplantation.

INTRODUCTION

Many complications may develop during the early postoperative period after lung transplantation accounting for the 10–20% perioperative mortality rate that is still observed in most centers. These complications are mainly related to infection, hemorrhage, acute rejection process, bronchial ischemia, or ischemia/reperfusion (I/R) injury ⁽¹⁾. In particular, the complications related to I/R injury are reimplantation edema and less frequently early hemodynamic failure ^(2–5), both of these being responsible for substantial morbidity or mortality. Many centers worldwide have been working on the prevention of lung I/R injury mainly by comparing in experimental models the effects of different preservation solutions. Recently, the beneficial effects of inhaled nitric oxide (NO), pentoxifylline (PTX), or a combination of the two administered prophylactically during reperfusion have been reported in experimental models ^(6–11). Positive results from these data have led some centers to preventively administer inhaled NO and/or PTX during lung reperfusion in clinical lung transplantation. Since 1998, the policy at our institution has been to use a combination of inhaled NO and PTX. To our knowledge, there has been no clinical data validating theses preventive strategies. Our purpose was to assess retrospectively the protective effect of inhaled NO and PTX on I/R injury after lung transplantation by comparing the early postoperative course in patients who received inhaled NO and PTX with that in patients who did not.

MATERIALS AND METHODS

Patients.

From 1988 to December 1999, 129 patients underwent lung transplantation in our center. Since January 1998, all recipients (n=23) were given inhaled NO and PTX during graft reperfusion on a systematic basis. We compared retrospectively the early postoperative course of these patients (NO-PTX group) with that of patients who did not receive NO or PTX (control groups) with respect to I/R injury related complications. In particular, the incidence of pulmonary reimplantation response, early hemodynamic failure as well as the PaO₂/Fraction of inspired O₂ (FIO₂) ratio during the first 3 postoperative days, the duration of mechanical ventilation and the two-month mortality rates were compared. We excluded from analysis all recipients transplanted for severe pulmonary hypertension because these patients experience perioperative complications mimicking those related to I/R injury. All the patients excluded from analysis were from control groups. Thus, the composition of the control groups were as follows: (1) 23 consecutive patients (control group 23) transplanted just before the use of NO-PTX PTX (from August 1994 to January 1998 and after exclusion of 2 patients with severe preoperative pulmonary hypertension) regardless of the preservation solution that had been used; (2) 95 patients (control group 95) who represent all the patients of our series who did not receive NO-PTX (after exclusion of 11 patients who were transplanted for severe pulmonary hypertension). Furthermore, we analyzed 5 periods of our transplantation experience by grouping the patients of our series (with the exception of the 11 patients who had a severe preoperative pulmonary hypertension) in 5 cohorts of consecutive patients. Each cohort is composed of 23 to 24 patients, the last cohort representing the NO-PTX group. The incidence of reimplantation edema, early hemodynamic failure, and 2-month mortality was compared in the five different groups.

The characteristics of the patients composing the NO-PTX group and the two control groups are described in Table 1. The type of procedure was a single lung transplantation in 102 patients and a bilateral lung transplantation in 16 patients.

Table 1: Sociodemographic and medical characteristics of patients

Transplantation procedure.

Donor selection was based on widely accepted guidelines ⁽¹²⁾. Donor and recipient matching was based on ABO compatibility and cytomegalovirus status. Particular attention was not taken to match donor and recipient lung size. In our series, the preservation of the graft was achieved with either topical cooling (in the early experience) or with flush perfusion of the pulmonary artery. In the latter technique, several preservation solutions of either intracellular or extracellular composition have been used from the onset of our transplantation program. The type of preservation and the composition of the preservation solutions in the NO-PTX group and in the two control groups are given in Table 2. The flush perfusion technique consisted of a hypothermic flush perfusion of pulmonary artery using the preservation solution cooled to 4°C which was delivered into the main pulmonary artery for approximately 4 min to a total volume of 50 ml/kg. After this procedure, the graft was excised and stored in a plastic bag filled with saline at 4°C. In the NO-PTX group, the preservation solution used during the harvesting of the graft was in all cases that developed by the Papworth hospital in Cambridge which is of extracellular type ⁽¹³⁾.

Table 2: Techniques of preservation

Single lung transplantations were performed using the classical procedure ⁽¹⁴⁾ and the bilateral lung transplantation technique was a bilateral sequential operation in all cases ⁽¹⁵⁾. In the case of single transplant procedure, the side of transplant was chosen according to the results of lung perfusion scan, computerized tomography of the thorax, and history of previous thoracic surgery. All procedures were carried out without the use of cardiopulmonary bypass. In the NO-PTX group, PTX (Torental, Hoeschst Laboratories, Paris, France) was given at the dose of 400 mg which was administered just before reperfusion and perfused over a 30-min period. Inhaled NO was given at a concentration of 10 ppm from a tank containing NO in nitrogen at the concentration of 225 ppm (CFPO, Meudon la Forêt, France) and was delivered by continuous injection at the Y piece of the breathing circuit. In France, the perioperative use of inhaled NO has been approved by the national regulatory authorities in the setting of cardiothoracic surgery including lung and cardiac transplantation as part of a compassionate release program (Autorisation Temporaire d’Utilisation). Approximation of NO flow rate necessary to achieve the desired NO concentration was performed according to Wysocki et al. ⁽¹⁶⁾. For a given inspiratory time expressed as a fraction of total time (Ti%), NO concentration within inspired gas mixture (Ci) was given by the following formula: Ci=(Cb.V_NO.Ti%)/Vtot where Cb is the NO concentration in the tank, V_NO the NO flow rate delivered continuously using a calibrated nitrogen flow-meter, and Vtot the sum of V_NO.Ti% and inspired minute ventilation delivered by the ventilator. The use of this formula in the determination of the desired NO concentration was validated in the first five patients with an online chemiluminescence nitrogen oxide analyzer (Eco-Physics, Pratteln, Switzerland) ⁽¹⁷⁾. Inhaled NO administration was begun just before reperfusion and maintained for 8 hr after reperfusion. The mean graft ischemic times in the three groups were not significantly different (Table 1). Postoperative initial immunosuppression consisted of cyclosporine, azathioprine and corticosteroids. The dosage of cyclosporine was adjusted to produce a trough level of 200–300 ng/ml as measured by radioimmunoassay of whole-blood samples. Azathioprine was given in a daily dose of 2–2.5 mg/kg.

Identification of early complications linked to I/R injury.

Pulmonary reimplantation edema was diagnosed in the presence of the following criteria: (1) diffuse alveolar infiltrate involving the lung allograft and developing within the first three days following lung transplantation; (2) PaO2/FIO₂ ratio <200 mmHg; (3) no evidence of bacterial infection or rejection. Gas exchange during the early postoperative period was assessed by retrieving in the case records of each patient both the best and the worst PaO₂/FIO₂ ratio recorded within the first 3 days after lung transplantation. In the NO-PTX group, gas exchange was assessed beyond the 8th hour after reperfusion, that is after cessation of NO administration.

Early hemodynamic failure was defined ⁽²⁾ as the association of the following criteria: (1) occurrence within the first 3 postoperative days of severe systemic hypotension: systolic pressure <80 mmHg without vasopressive drugs and despite fluid administration; (2) requirement of i.v. vasoactive drugs (dopamine>10 μg/kg/min or epinephrine when dopamine dosage up to 20 μg/kg/min was ineffective). The duration of mechanical ventilation (in days) and the 2-month mortality rate were also recorded.

Statistical analysis.

We compared the early postoperative course of the NO-PTX group with that of each of the two control groups. Results are expressed as the mean with SD for quantitative variables. Comparisons of continuous and categorical variables were made using the Student’s t tests and χ² tests, respectively. All P values are two tailed and P <0.05 are considered to indicate statistical significance. Survival rate up to 60 days was estimated using Kaplan-Meier method. Log-rank test assessed survival rate difference between NO-PTX group and each of the two control groups.

RESULTS

The main results are described in Table 3 and in Table 4. Reimplantation edema was observed in 6/23 patients (26%) in the NO-PTX group vs. 13/23 patients (56%) in the control group 23 (P =0.035) and 48/95 patients (50%) in the control group 95 (P =0.035). The worst PaO₂/FIO₂ ratio during the first three postoperative days was 240±102 mmHg in the NO-PTX group vs. 162±88 mmHg (P =0.01) and 176±107 mmHg (P =0.01) in the control group 23 and the control group 95, respectively. Similarly, concerning the best PaO₂/FIO₂ ratio, a statistical difference was observed between NO-PTX group (415±100 mmHg) and each of the control groups [(278±109 mmHg (P =0.001) and 304±118 mmHg (P =0.001) in control group 23 and in control group 95, respectively]. The duration of mechanical ventilation was 2.1±2.4 days in the NO-PTX group vs. 7±9 days in the control group 23 (P =0.02) and 6±7 days in the control group 95 (P =0.01). The 2-month mortality rate was 4.3% in the NO-PTX group vs. 26% (P =0.04) and 21% (P =0.07) in the control group 23 and the control group 95, respectively. Despite a tendency to a lower incidence of early hemodynamic failure in the NO-PTX group (13%) than in the control group 23 (26%) (P =0.25) and in the control group 95 (24%) (P =0.25), statistical significance was not achieved.

Table 3: Comparison between the NO-PTX group and the 23 patients transplanted before NO-PTX treatment (control group 23)

Table 4: Comparison between the NO-PTX group and the 95 patients transplanted before NO-PTX treatment (control group 95)

The 2-month survival curve according to Kaplan-Meier is depicted in Figure 1. The log-rank test identified a survival benefit in the group that received NO-PTX compared with control group 23 (P =0.04) but no statistical difference was detected between NO-PTX group and control group 95 (P =0.06). In four of the six patients who died in the control group 23, and in 11 of 20 patients who died in the control group 95, death was directly related to allograft dysfunction. The only death in the NO-PTX group was ascribed to a severe reimplantation edema associated with early hemodynamic failure. Despite aggressive ICU supportive care, the patient died at day 2 postoperatively from multiorgan failure.

Figure 1: Two-month cumulative survival assessed by Kaplan-Meier method. The NO-PTX group was compared with each of the two control groups using the log-rank test. In the NO-PTX group, survival was better than that of the control group 23 (P =0.04) whereas statistical significance was not achieved between NO-PTX group and control group 95 (P =0.06).

Figure 2 describes the 2-month mortality rate and the incidence of reimplantation edema and early hemodynamic failure in the five cohorts of patients. It appears that over time, we did not observe a lower 2-month mortality rate and a lower incidence of reimplantation edema or early hemodynamic failure until we used NO and PTX on a systematic basis (last cohort 01/1998–01/2000).

Figure 2: Two-month mortality, incidence of reimplantation edema and early hemodynamic failure in 5 cohorts of consecutive patients (23 to 24 patients in each cohort) representing all the patients of the series from 1988 to 1999. The dates of the cohorts are given on the x axis. The last cohort on the right represents the NO-PTX group. Over time, we did not observe a lower 2-month mortality and a lower incidence of edema or early hemodynamic failure before the use of NO-PTX.

DISCUSSION

The group of patients who were given preventively inhaled NO and intravenous PTX had a lower incidence of reimplantation edema and a shorter duration of mechanical ventilation than the control groups. The 2-month survival was also greater in the NO-PTX group than in one of the control groups. The NO-PTX group tended to have a lower incidence of early hemodynamic failure than the control groups. The preventive effect of the combination of NO and PTX was however not observed in all patients since one of them died as a direct result of I/R injury.

Overall mortality after lung transplantation is still notable. According to the International Registry, the 1-year actuarial survival is approximately 70%, most of the deaths occurring within 30 days after transplant ⁽¹⁸⁾. The main cause of death during the first postoperative month is ascribed to nonspecific graft failure ^(1,18,19). Early graft failure after lung transplantation, presumably related to I/R injury has been given various names such as reimplantation edema, reperfusion edema, primary graft failure, or allograft dysfunction. This complication refers to the occurrence of a noncardiogenic pulmonary edema with gas exchange impairment, associated in the most severe cases with hemodynamic failure ⁽²⁾. The morbidity related to early graft failure is thus significant. Histologically, diffuse alveolar damage is the hallmark of early graft failure ⁽¹⁾. Ischemia/reperfusion injury is now believed to be the predominant mechanism underlying this complication.

The potential impact of I/R injury on outcome of lung transplantation has led many centers worldwide to work extensively on its prevention, notably by work in experimental models. These models include mainly lung transplantation in rabbits, dogs, rats, or pigs and also several animal models of isolated perfused lungs. From such models, it has been shown that I/R injury is mediated by an influx of polymorphonuclear neutrophils within the lung. Their activation leads to the release of cytotoxic mediators including reactive oxygen species. Extensive experimental research aimed at comparing the effect of different preservation solutions during harvesting has been carried out. Although a beneficial effect of some solutions (in general extracellular type solutions) has been demonstrated, no clinical data have permitted validation of this approach. An other experimental approach focused on the effect of several agents administered at the time of graft reperfusion that aimed at reducing the neutrophil-dependent lung injury. Pentoxifylline, a methyl xanthine derivative has been shown to reduce neutrophil-dependent lung injury in several animal models ^(20,21) by decreasing neutrophil accumulation. There was thus a rationale for a potential beneficial effect of PTX on I/R injury. Using a model of rat-isolated perfused lung, Regnier and colleagues have shown that PTX given before reperfusion was able to reduce the increase in lung microvascular permeability after I/R, compared with placebo ⁽⁷⁾. This effect was associated with a decrease in lung sequestration of neutrophils. The beneficial effect of PTX on I/R injury was confirmed by the same group in a model of left lung allotransplantation in pigs. The animals that were given PTX before and throughout reperfusion had better gas exchange and less neutrophils sequestration within the lung than in the control group ⁽⁶⁾.

NO appears to be a key modulator in I/R injury. Endothelial cell dysfunction in I/R injury is manifested by a loss of NO-dependent vasodilatation ⁽²²⁾ and NO has also been shown to be protective in models of cardiac and mesenteric I/R injury ^(23–25). In pulmonary I/R injury, despite the potential harmful effect related to peroxynitrite formation, the beneficial effect of exogenous NO has been shown in various animal models both in isolated perfused lungs and in experimental lung transplantation ^(8,10). The effect of NO may be mediated by the following mechanisms: inhibition of xanthine oxydase thus limiting the generation of superoxyde anion, prevention of neutrophil adherence and activation, and protection against the neutrophil-derived reactive oxygen species ^(22,26). Given the protective effect of both NO and PTX on I/R lung injury, a combined effect of both treatments was investigated on a rat model of lung transplantation. This combined treatment with PTX and NO resulted in a synergistic enhanced protection against I/R injury compared with NO or PTX alone ⁽¹¹⁾.

In clinical transplantation, NO has been used successfully in the treatment of allograft dysfunction ⁽²⁶⁾. However, few clinical data on the use of NO or PTX in the prevention of I/R injury are available despite very promising experimental results. In a recently presented abstract, a beneficial effect of inhaled NO given prophylactically at reperfusion time in 18 consecutive lung transplantation recipients was reported but the study was not comparative ⁽²⁷⁾. Our results suggest that the combination of NO and PTX given at the time of reperfusion is able to reduce the incidence of I/R injury-related complications. We hypothesize that the reduction in the 2-month mortality rate is secondary to this effect. Considering the separate effects of PTX and NO, we could have chosen to administer these agents separately but the experimental data cited above led us to combine PTX and NO to obtain a synergistic effect. To our knowledge, this is the first study suggesting the potential benefit of NO and/or PTX in the prevention of I/R injury in clinical lung transplantation.

Obvious criticisms of this work need to be mentioned. The design of the study is retrospective and thus, bias can not be excluded. The early mortality and the incidence of reimplantation edema are high in the control groups. The overall two-month mortality rate before the use of NO-PTX was 21%. This rate is clearly higher than that reported by several groups but is consistent with what is commonly observed in France. In the French Registry 1998 (Etablissement Français des Greffes) which take into account all the transplant centers, the 1-month mortality after lung transplantation varies from 20 to 25% according to the era of the procedure (unpublished data). In the International Registry (which is not exhaustive), no specific data about the 2-month mortality are given but the latter may be approximately estimated to range between 15 and 22% according to the era ⁽¹⁸⁾. In a recent study by Khan et al. from the Cleveland Clinic Foundation, the two-month mortality rate was approximately 20%⁽²⁸⁾. More than half of the control patients developed reimplantation edema that represents a high incidence of this complication. However, this rate, which is consistent with that previously described by our group in the first 40 patients of our series ⁽²⁹⁾, may be explained by a rather liberal definition of pulmonary reimplantation response compared with other groups ⁽¹⁹⁾. Recently, Khan et al., using a definition of pulmonary reimplantation response that was similar to ours, reported a 57% incidence of this complication ⁽²⁸⁾. Our data may also reflect a learning curve in our team. The surgeons who performed the transplantation procedures before the use of NO-PTX were the same as those who operated on during the NO-PTX period. However, to ensure that we were not facing a learning effect, we compared the NO-PTX group not only with the 95 patients who did not receive NO-PTX but also with 23 consecutive patients who were transplanted just before the use of NO-PTX. These latter patients (control group 23) were transplanted from August 1994 to December 1997, i.e., late after the onset of our experience. Whatever the control groups, a significant difference in PaO₂/FIO₂ ratio, incidence of edema, and duration of ventilation was found in favor of the NO-PTX group suggesting that our results were due to a true effect of NO-PTX administration rather than to a “time” effect. The analysis of the five cohorts representing all the patients of our series also indicate that a learning curve was not observed in our center at least for the three studied parameters. A prospective randomized comparative study comparing the early outcome with and without PTX-NO would obviously have been more appropriate. Given the scarcity of lung donors in France, such a study would have been difficult to undergo in a single center.

In conclusion, the marked decrease in the incidence of allograft dysfunction in the NO-PTX group compared with two historical control groups suggests that PTX and inhaled NO given before and throughout reperfusion are protective against I/R injury in the setting of clinical transplantation. Thus, the promising data obtained with these treatments in experimental models seem to be confirmed. We hypothesize that the beneficial effect of NO-PTX is ascribed to a dual effect on graft vasculature and neutrophil sequestration.