INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cyclooxygenase (COX) inhibitors (1, 2) and inhibitors of the synthesis or effect of nitric oxide (NO) (2) have been shown to enhance hypoxic pulmonary vasoconstriction (HPV), suggesting that the release of endogenous prostaglandins and NO inhibits HPV.

If the release of endothelium-derived vasodilating prostaglandins and NO inhibits HPV in lung injury, these compounds could contribute to arterial hypoxemia. In accordance with this hypothesis, treatment (6, 7) and pretreatment (8, 9) with COX inhibitors have been shown to improve gas exchange in experimental lung injury. Inhibitors of NO synthase (NOS) have also been shown to preserve gas exchange in models of acute lung injury (10, 11), possibly by restoring HPV (11). On the other hand, some studies show that NO could be protective in lung injury, by limiting the increase in microvascular permeability (12) and/or by reducing microvascular pressure (12, 14, 15). However, NOS inhibitors given after oleic acid-induced lung injury in dogs, did not affect gas exchange (16, 17).

In the study reported here, we investigated whether the combined blockade of COX by indomethacin and of NOS by N-nitro-L-arginine (L-NA) could produce additional effects on HPV, and hence convey better protection of gas exchange in oleic acid-induced lung injury than blockade of either COX or NOS alone. Agonistic (18, 19) as well as antagonistic (20, 21) interactions between the COX and NOS pathways have been reported. Moreover, inhibition of either one of these two enzymes was found to result in a maximum pressor response to hypoxia (1). Therefore, inhibition of one pathway when the other is already blocked could not further increase HPV.

In the present study, we evaluated pulmonary vascular tone by constructing plots of pulmonary artery pressure (Ppa) minus occluded Ppa (Ppao) divided by cardiac output (Q) ([Ppa  Ppao]/Q), to avoid flow-dependent changes in Ppa  Ppao (22) and in mediator release. We measured intrapulmonary shunt with the weakly soluble inert gas sulfur hexafluoride (SF₆) at constant Q to avoid flow-dependent changes in shunt during lung injury.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of the U.S. National Institutes of Health, and were approved by the Committee on the Care and Use of Animals in Research of the Brussels Free University School of Medicine.

Thirty-six mongrel dogs (16 to 45 kg) were anesthetized with pentobarbital sodium (25 mg/kg intravenously), paralyzed with pancuronium bromide (0.2 mg/kg intravenously), intubated, and ventilated (Elema 900 B Servo ventilator; Siemens, Solna, Sweden) with a tidal volume of 15 to 20 ml/kg (adjusted to maintain arterial PCO₂ between 30 and 40 mm Hg), a respiratory rate of 12 breaths/min, and an inspired O₂ fraction (FI_O2) of 0.4 during the preparation. Pentobarbital (2 mg/kg intravenously) and pancuronium (0.2 mg/kg intravenously) were given on a repeated, hourly basis to maintain anesthesia and paralysis. Femoral and pulmonary artery catheters were inserted for measurements of systemic and pulmonary hemodynamics and for sampling of arterial and mixed venous blood. A balloon catheter was advanced into the inferior vena cava through a right femoral venotomy. A large-bore cannula was inserted into the left femoral artery and vein to act as an arteriovenous bypass. A left jugular venous catheter was placed for fluid and drug administration. Thrombus formation along the catheters was prevented by heparin (100 U/kg intravenously).

Vascular pressures were recorded and measured at end expiration. Heart rate was determined from a continuously monitored electrocardiographic lead. Cardiac output was measured with the thermodilution technique. Arterial and mixed venous blood gas tensions were determined immediately after the drawing of samples with a tonometered automated analyzer (ABL2; Radiometer, Copenhagen, Denmark) and corrected for temperature. When excessive metabolic acidosis occurred, it was corrected with a slow infusion of sodium bicarbonate. Body temperature was kept constant with an electric heating pad.

Intrapulmonary shunt was measured with SF₆ as a tracer gas and through use of the standard shunt equation. Starting 20 min before the first set of measurements, SF₆ was administered in a continuous infusion (5 ml/min). Samples of 10 ml of arterial and mixed venous blood were simultaneously withdrawn, equilibrated with 35 ml N₂ in a heated bath for 45 min, and analyzed for SF₆ with an electron capture detector (5890A gas chromatograph; Hewlett-Packard, Palo Alto, CA) (23).

Experimental Protocol

Hypoxic pulmonary vasoconstriction. Three consecutive hypoxic tests were performed, yielding six plots of (Ppa  Ppao)/Q (Table 1). After ensuring steady-state conditions for 20 min at FI_O2 = 0.4 (stable heart rate, systemic arterial pressure, Ppa), a first five-point (Ppa  Ppao)/ Q plot was generated for all dogs with the bypass open (1 point), with the bypass closed (1 point), and after stepwise inflations of the inferior vena caval balloon (3 points). Each such plot was constructed in about 20 min. Systemic and pulmonary hemodynamics were determined at each point on the (Ppa  Ppao)/Q plots. Arterial and mixed venous blood gases were measured at the highest and at the lowest Q of each plot.

The same procedure was repeated at FI_O2 = 0.1. Each FI_O2 was applied for at least 6 min to allow stabilization before measurements.

The same sequence of (Ppa  Ppao)/Q plots was repeated during two further hypoxic challenges, each beginning at 30 min after the start of either an infusion of solvent (NaCl 0.9%), or an infusion of indomethacin alone (1 mg/kg intravenous bolus followed by 1 mg/kg continuous infusion), or an infusion of L-NA alone (5 mg/kg intravenous bolus followed by 5 mg/kg continuous infusion), or an infusion of indomethacin alone followed by L-NA, or an infusion of L-NA alone followed by indomethacin (Table 1). Dogs were randomly allocated to one of five study groups. Infusion rates were adjusted so that the entire amount of drug was administered during the experiment. Indomethacin and L-NA (Sigma) were dissolved in 100 ml 0.9 % NaCl just before the experiments.

Oleic acid. After the drug infusion procedures the dogs were considered as pretreated with solvent, indomethacin, L-NA, or the combination of indomethacin and L-NA. FI_O2 was maintained at 0.4, and a last (Ppa  Ppao)/Q plot was constructed 90 min after a slow injection of 0.06 ml/kg oleic acid in the right atrium. Pa_O2 and intrapulmonary shunt were determined at FI_O2 = 0.4 during the construction of the first (baseline), fifth (after drug administration), and seventh (after oleic acid) (Ppa  Ppao)/Q plots.

Data Analysis

Results are expressed as mean ± SEM. Body surface area (m²) was calculated as 0.112 × weight (kg)^2/3. The individual (Ppa  Ppao)/Q plots were essentially linear. The correlation coefficients of all individual plots were > 0.8, except in four plots that were deleted from the analysis. A least-squares regression analysis was used to obtain the slopes and zero-flow pressure intercepts of the individual (Ppa  Ppao)/Q plots presented in Table 4. Values of Ppa  Ppao interpolated from the regression analyses for individual dogs were averaged at 0.5 L/min · m² intervals of Q from 2 to 4 L/min · m² to obtain the composite (Ppa  Ppao)/Q plots shown in the figures. A two-factor analysis of variance (ANOVA) for multiple measurements was used to assess the effects of hypoxia, oleic acid, and drugs on hemodynamics and blood gases. When the F ratio of the ANOVA reached a value of p < 0.05, modified t statistics were used to determine the means that differed significantly (24).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Data from the group given indomethacin followed by L-NA, and from the group given L-NA followed by indomethacin, are shown in Tables 2 and 3. Hemodynamic and blood gas variables before drug or solvent administration did not differ among the study groups. Reduction of Q significantly (p < 0.01) decreased mixed venous PO₂, mean systemic arterial pressure, Ppa, Ppao, and right atrial pressure in all study groups, as previously described (6, 17).

HPV

Before administration of drugs, hypoxia nonsignificantly affected slopes and pressure intercepts of the (Ppa  Ppao)/Q relationships (Table 4 and Figure 1), as was expected, since both "responders" (dogs with a spontaneous pulmonary vasoconstrictive response to hypoxia) and "nonresponders" were included in these analyses. Hypoxia reduced Pa_O2 and mixed venous PO₂ (p < 0.01), and produced variable increases in Q, heart rate, mean systemic arterial pressure, and Ppa (Tables 2 and 3), which were significant in some study groups but not in others, illustrating the wide variability in the hemodynamic responses to hypoxia (4, 8).

View larger version (35K): [in this window] [in a new window]   Figure 1.   Composite (Ppa  Ppao)/Q relationships after indomethacin alone (upper panel, left), L-NA alone (upper panel, right), indomethacin followed by L-NA (lower panel, left), and L-NA followed by indomethacin (lower panel, right) (*p < 0.05; **p < 0.01 for changes in slopes and/or in zero-flow pressure intercepts).

Solvent. Solvent had no significant effect on pulmonary hemodynamics or blood gases, and did not affect the (Ppa  Ppao)/Q relationships with hyperoxia or hypoxia (data not shown). Consequently, the hypoxic response at standardized Q from 2 to 4 L/min · m² was reproducible over three consecutive hypoxic challenges (Figure 2).

View larger version (34K): [in this window] [in a new window]   Figure 2.   Hypoxic response (mm Hg) (i.e., the increase in Ppa  Ppao when FIO2 is reduced from 0.4 to 0.1 at a standardized Q of 3 L/min · m2. Similar results are obtained for standardized Q from 2 to 4 L/min · m2 (*p < 0.05 compared with the previous column).

Indomethacin alone. Indomethacin did not affect the (Ppa  Ppao)/Q relationship in hyperoxia, but increased its slope in hypoxia (Table 4 and Figure 1), and increased the hypoxic response (Figures 1 and 2). Under conditions of a constant infusion of this drug, the hypoxic response remained comparable during the subsequent hypoxic test (Table 4, Figures 1 and 2). At FI_O2 = 0.4, with the arteriovenous bypass open, indomethacin had no effect on Q, heart rate, Ppao, or right atrial pressure, but increased systemic arterial pressure from 108 ± 5 mm Hg to 126 ± 5 mm Hg (p < 0.05).

L-NA alone. L-NA did not affect the (Ppa  Ppao)/Q relationship in hyperoxia, but increased its slope and pressure intercept in hypoxia (Table 4 and Figure 1), and increased the hypoxic response, which remained stable during the subsequent hypoxic test (Table 4, Figures 1 and 2). At FI_O2 = 0.4, with the arteriovenous bypass open, L-NA did not affect heart rate, decreased Q from 5.0 ± 0.5 L/min · m² to 3.5 ± 0.2 L/min · m², and increased systemic arterial pressure from 104 ± 5 mm Hg to 149 ± 9 mm Hg, Ppa from 15 ± 1 mm Hg to 18 ± 2 mm Hg, Ppao from 5 ± 1 mm Hg to 9 ± 1 mm Hg, and right atrial pressure from 3 ± 1 mm Hg to 6 ± 1 mm Hg (all p < 0.05).

Indomethacin and L-NA in combination. Combination of the two inhibitors did not affect the hyperoxic (Ppa  Ppao)/ Q relationship, but increased its slope and pressure intercept in hypoxia (Table 4 and Figure 1) and resulted in a supplementary increase in HPV over that with each drug given alone, whatever the order of administration (Figures 1 and 2). At FI_O2 = 0.4, with the arteriovenous bypass open, the combination decreased Q and increased systemic arterial pressure, Ppa, Ppao, and right atrial pressure as compared with the data recorded before drug administration (Tables 2 and 3). Under the same experimental conditions, adding indomethacin to L-NA increased systemic arterial pressure, Ppa, Ppao, and right atrial pressure over the respective values with L-NA alone (Table 3), which was not the case when L-NA was added to indomethacin (Table 2).

Oleic Acid-Induced Lung Injury

After administration of oleic acid, data from both groups pretreated with the combination of indomethacin and L-NA were comparable, and were pooled (Table 5). After administration of solvent, oleic acid had no significant effect on the (Ppa  Ppao)/Q relationship (Figure 3). However, oleic acid significantly affected the (Ppa  Ppao)/Q relationship after indomethacin alone, after L-NA alone, and after the combination of indomethacin and L-NA, by increasing both slopes and pressure intercepts (all p < 0.02, Figure 3). Pa_O2 and the SF₆-based value for intrapulmonary shunt, measured at an identical Q of about 3.0 L/min · m², were comparable in all study groups and remained so at baseline and after drug or solvent administration before treatment with oleic acid (data not shown), indicating no degradation in gas exchange during the HPV protocol. Oleic acid decreased Pa_O2 and increased the SF₆-based shunt value in all study groups (p < 0.01), but this alteration in gas exchange was less pronounced after indomethacin and indomethacin combined with L-NA than after solvent or L-NA alone (Table 6).

View larger version (28K): [in this window] [in a new window]   Figure 3.   Composite (Ppa  Ppao)/Q relationships before and after oleic acid in dogs pretreated with solvent (upper panel, left), indomethacin (upper panel, right), L-NA (lower panel, left), and the combination of indomethacin and L-NA (lower panel, right) (*p < 0.02 for changes in both slopes and zero-flow pressure intercepts).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The new findings in this study are that a combination of indomethacin and L-NA amplified HPV over its value with each drug given alone, and that pretreatment with indomethacin, but not with L-NA, attenuated the gas-exchange degradation in oleic acid-induced lung injury. The results also confirm that indomethacin and L-NA have no effect on hyperoxic pulmonary vascular tone (1, 2, 4, 25, 26), and that both compounds separately enhance HPV (1). Pentobarbital anesthesia slightly inhibits HPV (27) and can modify the pulmonary vascular response to compounds affecting the COX pathway (28), although possibly not to compounds affecting the NO pathway (29). We therefore limit our conclusions to intact pentobarbital-anesthetized dogs.

HPV

Existing data confirm that inhibitors of COX (1, 2) and NOS (2), given separately, augment HPV. We wondered whether combined inhibition of these two enzymes could result in an additional increase in HPV. In a previous study in our laboratory, we examined the stimulus-response curve of HPV by measuring Ppa  Ppao at constant Q and at different values of FI_O2 in intact anesthetized dogs (1). The maximal hypoxic response occurred at FI_O2 = 0.1, and was comparable in responders without treatment, in responders treated with aspirin, and in nonresponders treated with aspirin (1). These data suggested that the magnitude of HPV was maximal after COX inhibition. The results of the present study clearly indicate that an additional increase in HPV can be obtained by coinhibition of the COX and NO pathways, suggesting that both pathways are largely independent in the pulmonary vascular wall.

Our data are in keeping with those of Sprague and colleagues, which were derived from anesthetized rabbits with unilateral alveolar hypoxia (2). Indomethacin and L-NA methyl ester (L-NAME) reduced the percentage of pulmonary blood flow to the hypoxic lung, and when the two inhibitors were combined, an additional diversion of blood flow away from the hypoxic lung was observed (2). Russell and coworkers showed that meclofenamate and L-NAME separately increased Ppa in isolated lungs from chronically hypoxic rats perfused at constant flow (25). When the two compounds were coadministered, the increase in Ppa was more substantial, suggesting attenuation of chronic hypoxic pulmonary hypertension by both COX products and NO (25).

Although the difference was not significant, the amplification of the hypoxic response in our study appeared more pronounced when indomethacin was added to L-NA than when L-NA was added to indomethacin. We believe that this observation results from the large variability in the response to hypoxia and to drugs, rather than from different drug effects according to the order of drug administration. Dogs present with a wide range of pulmonary vasopressor responses to hypoxia, this response usually being weak or mild (1, 4, 8). The enhancement of HPV after COX inhibition (1) and after NOS inhibition (4) is also highly variable, in responders as well as in nonresponders to hypoxia. In the present study we also observed wide variability in the response to hypoxia and to both inhibitors in the various study subgroups (Figures 1 and 2).

The foregoing results suggest that neither COX products with vasodilatory activity nor NO contribute to the low basal pulmonary vascular tone, but that they are released together and independently during acute hypoxia and strongly inhibit HPV, accounting for the modest hypoxic pulmonary vasopressor response usually observed in dogs.

Oleic Acid-Induced Lung Injury

Pretreatment with indomethacin limited the degradation in gas exchange in oleic acid-induced lung injury. Several studies have shown that pretreatment or treatment with COX inhibitors improves gas exchange in experimental lung injury (6). We previously observed that HPV is altered after administration of oleic acid, that it can be restored by indomethacin, and that pretreatment with indomethacin or aspirin preserves gas exchange, suggesting that the release of a vasodilating prostaglandin supports blood flow to injured lung regions in oleic acid-induced lung injury (8). In the present study we could not exclude the possibility that this protective effect came from inhibition of the formation of thromboxane A₂ (TXA₂), a compound that has been shown to increase pulmonary microvascular permeability (30). If this mechanism did occur, we would also expect a reduction in Ppa, since TXA₂ is a pulmonary vasoconstrictor (30, 31). However, because indomethacin increased Ppa  Ppao, our interpretation is that indomethacin blocked prostacyclin release after oleic acid administration, which preserved HPV and limited the deterioration in gas exchange (8).

Pretreatment with L-NA increased pulmonary vascular tone after administration of oleic acid, as has also been observed when NOS is blocked after the development of lung injury (16, 17), suggesting that NO release occurs in this model. However, contrary to our hypothesis, L-NA alone did not better protect gas exchange than did solvent, and the combination of L-NA with indomethacin provided no additional benefit with regard to gas exchange over that with indomethacin alone, indicating that inhibition of NOS before the development of oleic acid-induced lung injury does not preserve gas exchange. We have previously observed that L-NA given after oleic acid also does not affect gas exchange (17). Putensen and coworkers, using the multiple inert-gas elimination technique, showed that N-monomethyl-L-arginine did not affect the ventilation-perfusion inequality and gas exchange after oleic acid- induced lung injury in dogs (16). Taken together, these data suggest that endogenous NO is released from both ventilated and nonventilated areas of the lung, with the result that NOS inhibitors do not direct pulmonary blood flow to well-oxygenated areas. Stimuli other than hypoxia, such as cytokines, can account for NO release in lung injury (32). Alternatively, we cannot exclude the possibility that inhibition of NOS suppressed a protective effect of NO, offsetting a beneficial effect of improved effectiveness of HPV on gas exchange. NO could limit fluid leakage in lung injury by preserving the integrity of the microvascular barrier (12) and/or by reducing microvascular hydrostatic pressure (12). In the present study, Ppao, which indirectly reflects pulmonary postcapillary pressure, was similar in the various groups of animals studied (Table 5), which does not suggest an NO-induced reduction in microvascular pressure during lung injury.

In summary, these results show that although a vasodilating prostaglandin and NO both inhibit HPV in dogs with normal lungs, these compounds have different effects on gas exchange in oleic acid-induced lung injury.