INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

When vascular endothelial cells are stimulated to increase their production of nitric oxide (NO), the NO rapidly diffuses into subjacent vascular smooth muscle cells, where it causes relaxation and vasodilation (1). As a result of the extremely high affinity of NO for hemoglobin, NO is rapidly inactivated in the bloodstream, with a half-life time estimated to be 111 to 130 ms (2). Accordingly, exogenously administered inhaled nitric oxide (iNO) causes pulmonary vasodilation in well aerated areas of the lung with no systemic hemodynamic side effects, unlike intravenously administered vasodilators (3).

Since 1991, both observational clinical and animal experimental data have suggested that iNO may be beneficial in patients with pulmonary hypertension and acute respiratory distress syndrome (ARDS) (4, 5). The potential therapeutic value of iNO includes improved oxygenation and lowered pulmonary arterial pressure, both of which might lead to a reduction in morbidity or mortality (6, 7). Unfortunately, the response to iNO on both gas exchange and pulmonary artery pressure is not uniform or predictable in any individual patient (8, 9). The reason(s) for these variable responses are unclear. Regardless, this heterogeneous response may be one reason why randomized, multicenter studies have not demonstrated any significant benefit so far from this novel therapy in patients with ARDS (10).

We recently reported that oxygenation in experimental acute lung is strongly affected by the postinjury perfusion pattern (11). When perfusion redistributes away from areas of edematous lung injury, gas exchange is only moderately abnormal. We hypothesized that in such cases, treatment with iNO would have little additional benefit on oxygenation. In contrast, when perfusion redistribution after ALI is interfered with, oxygenation is severely affected. In these cases, iNO could restore perfusion redistribution and improve oxygenation. The present study was designed to test this hypothesis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Preparation

These studies were approved by the Washington University School of Medicine's Animal Studies Committee. Fifteen adult mongrel dogs weighing 18 to 22 kg were anesthetized with pentobarbital sodium (40 mg/kg), intubated with a cuffed endotracheal tube, and ventilated (FI_O2 = 1.0) with a Harvard pump respirator (Harvard Apparatus Co., South Natick, MA) at a positive end-expiratory pressure (PEEP) of 5 cm H₂O and a tidal volume of 15 ml/kg in the supine position, with the respiratory rate adjusted to achieve a normal arterial Pa_CO2. Additional barbiturate was administered as necessary to eliminate spontaneous breathing.

Instrumentation was performed with animals in the supine position. After percutaneous insertion of two 8.5 Fr introducers (Baxter Healthcare Corp., Irvine, CA) in both femoral veins, a 7.5 Fr balloon-tipped pulmonary artery catheter (Baxter) and a 110-cm 7 Fr pig-tailed catheter (Cook, Bloomington, IN) were positioned in the pulmonary artery under fluoroscopic visualization. A 20-Ga artery catheter (Arrow International, Reading, PA) was placed into the right femoral artery via the Seldinger technique for blood sampling; a 6.0 Fr introducer (Cook, Bloomington, IN) was percutaneously inserted into the right external jugular vein for drug and radionuclide administration. Catheter patency was maintained by periodic infusion of heparinized saline (1 U/ml).

Cardiac output was measured in triplicate by the thermodilution technique with a cardiac output computer (American Edwards Laboratories, Irvine, CA). Pressure transducers (Baxter) were calibrated to the center of the lateral chest and connected to a Mennen model 742 monitor (Mennen, Clarence, NY) for pulmonary arterial, pulmonary wedge, and systemic arterial pressure recordings. Blood gases were analyzed using a Model 1305 blood gas analyzer (Instrumentation Laboratories, Lexington, MA).

Administration and Measurement of NO

A tank of NO source gas with a mixture of 800 ppm NO in nitrogen (Praxair, St. Louis, MO) was used to deliver inspired NO concentrations at 10 and 40 ppm (Figure 1). In order to administer the desired concentration of NO, the 800 ppm NO was introduced continuously into the inspiratory limb of the breathing circuit 90 cm proximal to the Y piece at a known flow rate using a flowmeter (Alltech digital flow check; Alltech, Deerfield, IL). For coarse adjustment, NO/N₂ flow was calculated from NO concentration of the source gas and inspiratory flow. For fine adjustment, the inspiratory concentration of NO was measured by sampling a portion of the inspiratory gas (300 ml/ min) from the inspiratory limb 20 cm proximal to the Y piece. The sampled gas was delivered to a Sievers Model NOA 280 (Sievers, Boulder, CO) rapid response (250 ms) chemiluminescent NO detector, which was calibrated according to the manufacturer's instructions (12). The NO concentration was digitized and recorded for calculation of the mean inspired NO concentration during administration of iNO. For concentrations of 10 and 40 ppm iNO, the fluctuation in inspired NO concentration caused by the used NO delivery system was in a range of less than ± 1 ppm NO. The fractional concentration of oxygen in inspired gas (FI_O2) was measured in the inspiratory limb 10 cm proximal to the Y piece with a Ventronics oxygen analyzer model 5524 (Ventronics, Temecula, CA).

View larger version (9K): [in this window] [in a new window]   Figure 1.   NO administration setup. Pure oxygen was delivered to the mechanical ventilator gas inlet (O2  ). From a tank of nitric oxide source gas (NO/NO2) with a mixture of 800 ppm nitric oxide in nitrogen, nitric oxide was introduced continuously into the inspiratory limb of the breathing circuit 90 cm proximal to the Y piece at a known flow rate. The inspiratory concentrations of nitric oxide and oxygen were measured in the inspiratory limb 20 and 10 cm proximal to the Y piece, respectively (see text for more details).

PET Techniques

Regional pulmonary blood flow (PBF) and regional lung water concentration (LWC) were measured with positron emission tomographic (PET) imaging. These measurements were obtained with a "Super-PETT" 3000 system built in-house. Design features, methods for calibration, corrections for activity decay, and corrections for photon attenuation have been described previously (13).

The animals were placed in the scanner with the most caudal tomographic slice about 1-2 cm below the level of the dome of the diaphragm. Data were recorded simultaneously from seven slices with a center-to-center separation of 1.05 cm and an in-plane full-width half-maximum spatial resolution of 0.85 cm. The image reconstruction resolution was set at 12 mm.

The methods used to measure PBF and LWC, including supporting validation studies, have also been described previously in detail (13, 17). In general, PET is used to measure the tissue concentration and distribution of a positron-emitting radionuclide, which in the present study was simply H₂¹⁵O. The activity data measured with PET, when combined with blood activity (used as a reference) and analyzed with an appropriate compartmental mathematical model, yield tomographic images representative of PBF and LWC.

Experimental Protocols

We studied 15 dogs, divided into three groups (Figure 2). In Group OA only (n = 5), the only intervention was 0.08 ml/kg oleic acid (OA) given into the central venous catheter 120 min prior to the administration of iNO. In Group E+OA (n = 5), 15 µg/kg of Escherichia coli endotoxin (Fisher Scientific, Pittsburgh, PA) was injected intravenously, and preceded the same dose of OA by 30 min. This dose of endotoxin is known to eliminate hypoxic pulmonary vasoconstriction without systemic hemodynamic changes (11). In the control group (n = 5), the only intervention was the administration of iNO. Otherwise, all animals were treated the same.

View larger version (14K): [in this window] [in a new window]   Figure 2.   Experimental timeline. Asterisks indicate timing of blood sample collections for eicosanoid analysis. First long upright arrow (E/P) indicates timing at which either endotoxin or placebo was given. Second long upright arrow (OA/P) indicates timing at which either oleic acid or placebo was given. Short upright arrows (PET-scan) always indicate timing of hemodynamic measurements and PET-scans. Short horizontal arrows (iNO) indicate the periods of 10 and 40 ppm inhaled nitric oxide administration.

After instrumentation was completed, each experiment began with a background ("blank") PET scan. In addition, a transmission scan used to correct photon attenuation during emission scans and for the replacement of regions-of-interest for later image analysis (see below) was performed before and 90 min after OA injection. The following data were obtained at the time of each "dataset" collection: (1) a 15- to 18-s scan (used for the PBF measurement) obtained during a continuous infusion of about 60 mCi O-15 labeled water, (2) a 300-s scan obtained after equilibration of the O-15 water (for measurement of LWC and for the "apparent" blood-tissue partition coefficient for water used in the PBF calculation), and (3) pulmonary artery, pulmonary wedge, and systemic arterial pressures, cardiac output, and blood gas analysis. Blood was also drawn for eicosanoid analysis in Group OA only and in Group E+OA (see below).

After the first dataset was collected, either endotoxin or placebo was administered (Figure 2). Thirty minutes later, either OA or placebo was administered, followed 120 min later by a second dataset (no PET scan in control group). After the second dataset was completed, administration of iNO in a concentration of approximately 10 ppm iNO (range, 8 to 12 ppm iNO) was started. Thirty minutes later, a third dataset was obtained. Afterwards the concentration of iNO was changed to approximately 40 ppm (range, 32 to 49 ppm) and a fourth dataset was collected 30 min later. Sixty minutes after the end of iNO administration, the fifth and last dataset (no PET scan in control group) was obtained. After this, the dog was euthanized with additional pentobarbital followed by 20 ml saturated KCl.

Biochemical Analysis

Blood was drawn at baseline (Figure 2), just prior to giving OA (or placebo), 120 min after OA (or placebo), and at the end of the iNO administration for measurement of the stable metabolites of thromboxane (TxB₂) and prostacyclin (6-keto-prostaglandin F₁ alpha [PGF₁]) by an enzyme-immunoassay technique. At each time, a sample (for plasma) was obtained by drawing blood into a tube with EDTA (1 mg/ml) and indomethacin (5 µg/ml) added, and then centrifuged immediately at 5° C at 2,200 × g for 10 min. The plasma was removed and stored frozen at 30° C until assay.

Enzyme immunoassay of 6-keto-PGF₁ and TxB₂ was performed in 96-well microtiter plates precoated with 2 µg/well goat antirabbit immunoglobulin G (18). Before use, the plates were washed with 10² M phosphate buffer (pH, 7.4) containing 0.05% Tween 20 (wash buffer). The assay was performed in a total volume of 150 µl. In brief, 50 µl of acetylcholinesterase-conjugated eicosanoid tracer (Caymen, Ann Arbor, MI), 50 µl of antiserum directed against 6-keto-PGF₁ or TxB₂ (PerSeptive Biosystems, Framingham, MA), and 50 µl of a standard or sample in assay buffer were combined and incubated at 25° C for 18-20 h. After the plates were washed three times with wash buffer, Ellman's reagent (200 µl) was dispensed into each well. Absorbance was recorded at 412 nm in a microtiter plate spectrophotometer (BioTech, Winooski, VT) when the absorbance for the well containing the "0" standard (B_o) exceeded 0.200 absorbance units. Each sample was assayed in duplicate. A standard curve was generated for each assay. Sample eicosanoid concentrations were determined by comparison to a log-logit transformation of the standard curve. Eicosanoid concentrations were expressed as pg/ml blood.

Image Analysis

From each dog, the four contiguous tomographic slices with the most lung were analyzed from the seven slices reconstructed as part of each PET scan, encompassing most of the caudal lobes. Regions of interest from the right and left lungs were defined on each transmission scan, as previously described (21, 22).

The position of each region was kept in computer memory, and mean values for each region were obtained for all PET measurements performed. PBF was measured as ml/min/100 ml lung, and LWC as ml of water/100 ml lung. To normalize the regional PBF data for differences in cardiac output, PBF in each picture element (pixel) was expressed as a fraction of the total blood flow to the region.

To evaluate the relationship of PBF to anatomic position within a region, the x- and y-coordinates for each pixel, along with the respective fractional PBF values for each pixel, were recorded. The pixel data were then sorted, first by their y-coordinate. Next, within each value for y, the data were sorted again by their x-coordinate. The result was a listing of the pixels by location, beginning in the most ventral-medial portion of the region and ending with the most dorsal-lateral portion of the region. Each region contained ~ 400 to 500 pixels. Arbitrarily, the data were divided into 20 "bins" stacked vertically in the ventral-dorsal direction, so each bin contained ~ 20 to 25 pixels, which could then be averaged. By keeping the number of bins per region and the number of tomographic slices per dog constant, bin values could be averaged across dogs, allowing comparisons between experimental groups.

To quantify perfusion redistribution, we determined the differences in fractional PBF between the baseline and subsequent PET data sets, and summed the difference in those bins in which PBF decreased between the different time points (21, 22).

Statistical Analysis

Data are presented as means ± SD. Statistical significance was determined by a repeated-measures analysis-of-variance using the General Linear Models Procedure of the Statistical Analysis System (SAS). Post-hoc testing with Tukey's honestly significant difference (HSD) test was limited to comparisons of baseline data, to changes from baseline data among the three experimental groups, and to differences among the groups at the same time point. We accepted p < 0.05 as indicating statistical significance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline Data

In general, the three groups were similar at baseline. In Group E+OA, lung water concentration (55 ± 8 ml/100 ml lung) was slightly but statistically significantly higher at baseline than in the control group and Group OA only (45 ± 13 and 45 ± 14 ml/100 ml lung). However, there was no other statistically significant difference between the three groups for the various systemic and pulmonary hemodynamic, blood gas, pulmonary blood flow, or cardiac output variables (Table 1 and Figures 3 and 4). Therefore, the higher baseline lung water concentration in the E+OA group probably reflects a modest degree of atelectasis in this group.

View larger version (21K): [in this window] [in a new window]   Figure 3.   Ratio of partial pressure of oxygen in arterial blood (PaO2) and fraction of inspired oxygen (FIO2) in the three experimental groups. Baseline data were obtained just prior to giving endotoxin or placebo. Pre-OA data were obtained just prior to giving oleic acid or placebo. After injury 0 ppm iNO data were obtained 120 min after oleic acid or placebo without administration of inhaled nitric oxide (iNO). The following data were collected 30 min after administration of 10 or 40 ppm iNO was started. 0 ppm iNO data were obtained 60 min after the end of iNO administration. Compared with the other two experimental groups or baseline, PaO2 was statistically significantly lower (crosses) in Group OA only and in Group E+OA at the after injury time point. Although oxygenation improved significantly (asterisks) in Group E+OA during iNO administration, iNO had no beneficial effect on oxygenation in the control group and in Group OA only.

View larger version (24K): [in this window] [in a new window]   Figure 4.   Mean arterial blood pressure (mABP) and mean pulmonary artery pressure (mPAP) in the three experimental groups. Baseline data were obtained just prior to giving endotoxin or placebo. Pre-OA data were obtained just prior to giving oleic acid or placebo. After injury 0 ppm iNO data were obtained 120 min after oleic acid or placebo without administration of inhaled nitric oxide (iNO). The following data were collected 30 min after administration of 10 or 40 ppm iNO was started. 0 ppm iNO data were obtained 60 min after the end of iNO administration. In both Group OA only and Group E+OA, values of mean arterial blood pressure at the After injury time point and during administration of iNO were statistically lower (crosses) than at baseline or compared with the control group. During administration of iNO, mean arterial blood pressure did not significantly change in all three experimental groups. At the after injury and the 10 ppm iNO time points, mean pulmonary artery pressure in Group OA only was statistically significantly higher (asterisks) than at baseline or compared with the other two experimental groups. During administration of iNO, mean pulmonary artery pressure fell clearly but not significantly in Group OA only, whereas in Group E+OA, mean pulmonary artery pressure increased significantly compared with baseline or the control group (open circles).

Effect of iNO on Normal Healthy Dogs

Before iNO administration, there was no change in any measured variable, indicating a stable experimental preparation. During administration of 10 and 40 ppm iNO, we did not observe statistically significant changes in any measured variable, including systemic blood pressure, pulmonary artery pressure, cardiac output, PBF, and Pa_O2/FI_O2 (Tables 1 and 2 and Figures 3-5).

View larger version (14K): [in this window] [in a new window]   Figure 5.   Average ventral-dorsal distribution of fractional pulmonary blood flow (PBF) in the Control group. Bin numbers represent equal collections of picture elements on multiple positron emission tomography images of PBF, with lower bin numbers situated in ventral regions and higher bin numbers situated in the dorsal regions of the lungs. Each symbol represents the mean value for all dogs in the Control group. In the Control group, administration of iNO had no effect at all on PBF.

Effect of iNO after OA Alone

In Group OA only, 120 min after oleic acid infusion, the pulmonary arterial pressure increased significantly (from 16 ± 2 at baseline to 25 ± 4 mm Hg), systemic blood pressure fell significantly (from 131 ± 23 at baseline to 108 ± 11 mm Hg), oxygenation deteriorated significantly (from 567 ± 32 at baseline to 437 ± 67 mm Hg), cardiac output also fell (from 2.2 ± 0.4 at baseline to 1.7 ± 0.4 L/min), LWC increased significantly by 38%, and a significant amount of perfusion redistribution was detected when compared with baseline (Tables 1 and 2 and Figures 3 and 4). These observations were associated with an increase in both TxB₂ levels (from 62 ± 107 at baseline to 117 ± 217 pg/ml) and 6-keto-PGF₁ levels (64 ± 13 at baseline to 125 ± 26 pg/ml) (Table 3).

During the administration of iNO, mean arterial blood pressure did not change in this group (Figure 4), whereas mean pulmonary artery pressure decreased modestly but not statistically significantly in this group (from 25 ± 4 at 0 ppm iNO to 23 ± 5 mm Hg at 40 ppm iNO). Concentrations of TxB₂ and 6-keto-PGF₁ remained elevated during administration of iNO (Table 3). Furthermore, Pa_O2/FI_O2 increased only marginally (from 437 ± 67 at 0 ppm iNO to 467 ± 41 at 10 ppm iNO, and 471 ± 52 mm Hg at 40 ppm iNO) during administration of iNO (Figure 3). These changes in Pa_O2/FI_O2 were statistically significant. Ventilation with iNO at either concentration had no effect on the distribution of PBF in this group (Figure 6).

View larger version (19K): [in this window] [in a new window]   Figure 6.   Average ventral-dorsal distribution of fractional pulmonary blood flow (PBF) in Group OA only. Bin numbers represent equal collections of picture elements on multiple positron emission tomography images of PBF, with lower bin numbers situated in ventral regions and higher bin numbers situated in the dorsal regions of the lungs. Each symbol represents the mean value for all dogs in Group OA only. There is a significant reduction of relative PBF ("perfusion redistribution") in the dorsal lung regions (the injured regions that develop the greatest increase in lung water) in Group OA only. Administration of 10 and 40 ppm inhaled NO (iNO) had no effect on the distribution of PBF in this group.

Effect of iNO after Endotoxin + OA

Thirty minutes after endotoxin, 6-keto-PGF₁ levels increased significantly when compared with baseline, whereas the increase in TxB₂ levels did not reach statistical significance in the analysis of variance. Also at 30 min, cardiac output decreased statistically significantly (from 2.9 ± 0.7 at baseline to 2.3 ± 0.7 L/min) (Tables 1 and 3). However, there were no further changes in the remaining measured variables 30 min after endotoxin.

One hundred twenty minutes after OA, TxB₂ levels remained elevated compared with Group OA only. Even more dramatic were the changes in 6-keto-PGF₁ levels, which increased nearly three and a half times the 30-min level and seven times the baseline value (Table 3). These changes are similar to those previously reported (11).

Equally dramatic were the changes in hemodynamic and oxygenation variables after OA: there was no significant increase in pulmonary arterial pressure in this group (Figure 4); perfusion redistribution was largely eliminated (Table 3 and Figure 7), and oxygenation fell from 558 ± 70 at baseline to 119 ± 53 mm Hg after OA (Figure 3). Compared with baseline, the changes in cardiac output, systemic blood pressure, and the percent increase in LWC (35%) were nearly the same as in Group OA only (Tables 1 and 2). These changes too are similar to those previously reported (11).

View larger version (21K): [in this window] [in a new window]   Figure 7.   Average ventral-dorsal distribution of fractional pulmonary blood flow (PBF) in Group E+OA. Bin numbers represent equal collections of picture elements on multiple positron emission tomography images of PBF, with lower bin numbers situated in ventral regions and higher bin numbers situated in the dorsal regions of the lungs. Each symbol represents the mean value for all dogs in the Group E+OA. In contrast to Group OA only, the relative pattern of pulmonary blood flow is essentially unchanged after injury in Group E+OA. In this group, administration of iNO restored redistribution of PBF from the dorsal edematous lung regions to a level similar to that in Group OA only.

In Group E+OA, administration of iNO was associated with a slight but not statistically significant decrease in mean arterial blood pressure (Figure 4). Interestingly, mean pulmonary artery pressure did not fall in Group E+OA during ventilation with iNO. As in Group OA only, eicosanoid concentrations were not significantly affected by iNO administration (Table 3).

In contrast to Group OA only, PBF changed statistically significantly because of iNO administration. In Group E+OA, the administration of iNO restored the redistribution of PBF from the dorsal edematous lung regions to a level similar to that in Group OA only (Table 2 and Figure 7). These changes in PBF distribution were accompanied by a very significant improvement in oxygenation and were reversed when iNO was discontinued (Figure 3).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The oleic acid model is a well characterized and predictable model of acute lung injury (23). Within 90 to 120 min of administration, animals develop significant pulmonary edema. In general, animals show a significant reduction of relative pulmonary blood flow to these edematous lung regions ("perfusion redistribution") associated with mild-moderate pulmonary hypertension. Unlike some other models of lung injury in other species (24, 25), these hemodynamic changes continue to develop during the first 8 h after injury and are sustained for at least 28 h (26). Thus, the OA-induced model of ALI has been widely studied to gain a better understanding of the pathophysiologic and morphologic changes that take place in ARDS. This model has also been frequently used to investigate the effects of inhaled nitric oxide in ALI (27, 28).

Using the noninvasive quantitative nuclear medicine imaging technique of positron emission tomography, we found no effect of iNO on the distribution of pulmonary perfusion and on oxygenation in normal healthy animals (Figures 3 and 5). In normal animals, the absence of changes in PBF are not unexpected (29, 30). More importantly, we also observed only minor changes in oxygenation associated with a modest fall in pulmonary arterial pressure in the injured animals of Group OA only during the administration of iNO (Figures 3 and 4). These findings are in agreement with the results of others (27, 28). In an OA-induced model of ALI in pigs, Shah and coworkers (28) observed a significant dose-dependent decrease in pulmonary arterial pressure without changes in Pa_O2 during NO inhalation. Although they attributed the lack of effect to the "severity" of the lung injury model, we believe the answer lies in the effect of injury on the pre-iNO perfusion pattern (Figure 6). After OA injury in Group OA only (prior to the administration of iNO), PBF spontaneously redistributed away from the dependent injured lung regions to the nondependent noninjured lung regions, presumably because of an intact hypoxic pulmonary vasoconstriction (HPV) response. The administration of iNO had no additional effect on this injury response to regional PBF. Accordingly, there was no significant effect on oxygenation.

Because we observed in a previous study that a low dose of endotoxin, which is itself largely devoid of significant pulmonary effects, affects the development of perfusion redistribution in the OA lung injury model, we also investigated the effect of iNO in animals pretreated with low-dose endotoxin (11). In contrast to Group OA only, pretreatment with low-dose endotoxin blocked HPV because of increased plasma levels of prostacyclin. This resulted in a much greater mismatch of ventilation/perfusion after OA injury in Group E+OA. Thus, oxygenation was much worse than in Group OA only. In all animals of Group E+OA, the administration of iNO restored the redistribution of PBF to a similar level as in Group OA only (Figure 7), along with a dramatic improvement in oxygenation (Figure 3) because of partial reversal of ventilation/perfusion mismatch (Figure 7). The variable magnitude of the improvement in oxygenation depended on how well iNO restored perfusion redistribution and on the starting value for Pa_O2 prior to the administration of iNO. Because the levels of eicosanoids did not change during administration of iNO, the effects of iNO on the regional pulmonary blood flow were not due to an indirect effect on these eicosanoid levels.

If low-dose endotoxin "paralyzes" hypoxic pulmonary vasoconstriction, how is it possible that the administration of another vasodilator (iNO) caused (presumably) yet more vasodilation in relatively nonedematous lung regions, leading to the observed changes in perfusion pattern? The explanation may lie in the locus of effect of different vasodilators. For instance, prostacyclin is well known to cause precapillary pulmonary vasodilation (31). Thus, low-dose endotoxin, by causing dramatic increases in circulating prostacyclin, paralyzes the hypoxic pulmonary vasoconstrictor response, which is also known to be, primarily, a precapillary vasoconstrictive phenomenon (32, 33). In contrast, both oleic acid injury itself (21) and endotoxin (11) have been shown to increase pulmonary tissue and circulating levels of thromboxane, respectively. In contrast to the site of action of prostacyclin, thromboxane has a relatively greater effect on postcapillary pulmonary veins in dogs (34). Recently, Hillier and coworkers (35) provided convincing evidence that iNO primarily causes pulmonary venous vasodilation. Assuming a similar phenomenon occurs in the OA model, especially after endotoxin, iNO, by causing a reduction in pulmonary venous hypertension (caused, presumably, by thromboxane), will reduce the vascular resistance in areas of lung ventilated with iNO, even without causing any further pulmonary arterial vasodilation. The result is perfusion redistribution to the areas of lung ventilated with iNO.

The response to NO inhalation is not consistent in patients with ARDS (9). Treggiari-Venzi and colleagues (9) reported an increase in Pa_O2 in only 54% of patients with ARDS. Likewise, Mantkelow and colleagues (8) and Dellinger and coworkers (10) reported improvements in Pa_O2 in only 58 and 60% of patients, respectively. Not surprisingly then, the first randomized, multicenter, double-blinded studies of iNO did not demonstrate a significant benefit in patients with ARDS (10). Although the reason for this variability in patients is still not known, our data provide a plausible explanation.

The main finding of this study is that the response to iNO in ALI depends heavily on the pre-iNO perfusion pattern. Whether these experimental findings apply to patients with ARDS depends on several factors. First of all, the balance and importance of the vasodilatators NO and prostacyclin might be different in humans. Furthermore, most patients with ARDS receive various drugs, which can have an effect on pulmonary vascular tone and likewise on the distribution of regional PBF. In addition, timing may be important. We studied the effects of iNO during a very early stage of ALI when vasoconstriction, by one or more mechanisms, and not intravascular obstruction or mechanical compression, is responsible for virtually the entire amount of perfusion redistribution. At a later stage of experimental ALI or in patients with ARDS, intravascular obstruction or mechanical compression are certainly more important, and the effects of iNO may be altered. Nevertheless, if a similar phenomenon exists in humans, it may help explain the frequently reported variable response to iNO in ARDS.