INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Acute lung injury (ALI), an early stage of acute respiratory distress syndrome (ARDS), results from various causes including inflammatory damage, volutrauma, and oxygen toxicity associated with mechanical ventilation (1, 2). Its pathologic changes usually include edema, hemorrhage, hyaline membranes, infiltration of neutrophils, and desquamation and necrosis of bronchiole and alveolar epithelial cells, as well as with bronchiolar dilatation, alveolar atelectasis or overextension, and intraparenchyma pseudocysts (2). The search for effective respiratory therapy to prevent ALI and acute respiratory failure is of clinical importance (3). Among the new techniques and strategies on prevention and treatment of ALI/ ARDS, surfactant and inhaled nitric oxide (iNO) tend to be promising as supported by recent clinical studies (4). However, the effects of these therapies are limited, possibly because of either independent application per se or delayed intervention, allowing severe lung injury to develop. Surfactant dysfunction and deficiency is one of the mechanisms in ARDS (11), and it is associated with inadequate alveolar expansion, lung fluid clearance, vascular-to-alveolar permeability, and altered intrapulmonary blood flow, leading to ventilation-perfusion mismatching and hypoxic respiratory failure. Use of conventional mechanical ventilation (CMV) and high oxygen potentiates lung damage by shear-force stress and hyperoxic injury, which are often seen in immature neonates, resulting in interstitial emphysema, air leaking, and alveolar septal remodeling. These problems may also be found in children and adults with ALI and ARDS after a long term mechanical ventilation. Noninvasive ventilation such as pressure support ventilation (PSV) and continuous positive airway pressure (CPAP) aimed to minimize mechanical ventilation and/or hyperoxia induced ALI is applied for ALI/ARDS and hypoxic respiratory failure (12). In the present study, using an oleic-acid-induced ALI model, we investigated whether earlier treatment with a combined surfactant, iNO, and PSV would effectively prevent ALI and hypoxic respiratory failure.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surfactant

A porcine lung surfactant phospholipid extract (40 mg/ml) was prepared as reported elsewhere (16). Disaturated phosphatidylcholine (DSPC) and total phospholipids (TPL) in this preparation were determined according to the methods described by Mason and colleagues (17) and Bartlett (18), respectively. In this batch, DSPC was about 50% of TPL and surfactant protein B and C were less than 1% of TPL, as separated by Sephadex LH-20 liquid-gel chromatography and measured by amino acid analysis. Its surface properties were determined in a pulsating bubble surfactometer (see below).

Inhalation of Nitric Oxide

NO gas at 1,000 ppm (Shanghai BOC, Shanghai, China) was supplied to the inspiratory line of a ventilator circuit about 20 cm proximal to the endotracheal tubing connector (Y-piece). Control of NO gas flow and monitoring of NO and NO₂ concentration were achieved by a mass flow controller and an electrochemical NO/NO₂ monitor (16).

Animal Management

Protocols for animal care and experimental management were approved by the Children's Hospital Scientific Committee, Shanghai Medical University. Forty-seven healthy adult New Zealand White rabbits with a mean body weight of 2.8 ± 0.3 kg were used. The animals were sedated intramuscularly with diazepam (5 mg/ml) at 2 mg/kg, and anesthetized intravenously with pentobarbital sodium (10 mg/ml) at 20 mg/kg. Additional pentobarbital sodium was provided at 0.5 ml/ kg/h to maintain anesthesia. The animals were orally intubated with a cuffed endotracheal tubing (3.0 mm inner diameter) and breathed room air at an approximate rate of 50 breaths/min. An 18-G cannula was put into the right femoral artery for collection of blood samples and measurement of mean systemic arterial blood pressure (Psa) with a pressure transducer and a monitor (Department of Physiology, Shanghai Medical University, Shanghai). When the animals were stabilized after the initial operation, blood was taken for measurement of baseline values of Pa_O2, Pa_CO2, and pH and for determination of nitrite/nitrate and methemoglobin (MetHb) (see below). Baseline values for dynamic compliance (Cdyn) and resistance of the respiratory system (Rrs) were measured with a pneumotachograph GM 250 Navigator (Newport Medical Instruments, Newport Beach, CA) using an infant-type differential pressure transducer placed in the Y-piece. During the measurement, the animals were connected to a Wave ventilator E-200 (Newport Medical Instruments) set at synchronized intermittent mandatory ventilation (SIMV) mode, a fraction of inspired oxygen (FI_O2) of 0.21, frequency of 40 breaths/min, peak inspiratory pressure of 10 to 15 cm H₂O to provide a tidal volume (VT) at 8 to 10 ml/kg, and inspiratory-to-expiratory time ratio of 1:1.5. Cdyn was expressed as ml/cm H₂O/kg and Rrs as cm H₂O/L/s from an average of 10 consecutive breaths. After these measurements, animals were allowed to breathe room air again at approximately 50 breaths/min and VT and minute ventilation volume (E) were provided at 8 to 10 ml/kg and 0.4 to 0.5 L/min/kg, respectively, and recorded by the same monitoring system. Physiologic intrapulmonary shunting (S/T) was determined by measuring oxygen content in mixed venous blood (when FI_O2 was temporarily raised to 1.0), and its values were calculated using the standard shunt equation (19), and expressed as a percentage of the total pulmonary blood flow. The animals were then subjected to the experimental protocols.

Experiment Protocols

Oleic acid (0.89 mg/µl) (Cat. no. O1008; Sigma Chemical, St. Louis, MO) was first diluted with 10 volumes of bovine serum albumin (10 mg/ml) in saline. A total of 60 µl/kg body weight of this diluted oleic acid was slowly infused into the pulmonary circulation over 15 to 20 min, using a microtransfusion pump and a syringe connected to an infant feeding tube placed in the right atrium of the heart. Thirty-five animals were subjected to the oleic-acid infusion, and blood gas and lung function parameters were measured as explained above. Respiratory failure was defined as Cdyn decreased by > 30% from its baseline level (20), Pa_O2 < 60 mm Hg when breathing room air, and S/ T increased to > 20%, and this moment was regarded as representing treatment time 0. The animals were connected to the same ventilator, and FI_O2 was raised to 0.3 and kept constant throughout the rest of the experiment. A combined mode of PSV and CPAP set at 3 cm H₂O was then applied. PSV level was adjusted to provide VT at 8 to 10 ml/ kg, and E of 0.5 to 0.6 L/min/kg (approximately 1.5 to 2.0 L/min per animal) by regulating breath frequency through adjustment of trigger sensitivity for PSV. This adjustment aimed at achieving average values of blood pH, Pa_CO2, and Pa_O2 in ranges of 7.30 to 7.50, 40 to 50, and 60 to 80 mm Hg, respectively. The animals were then randomly allocated to four treatment groups, receiving: (1) PSV only (Control group, n = 10); (2) inhalation of 20 ppm NO (NO group, n = 8); (3) a bolus containing 100 mg of surfactant phospholipids/kg body weight (Surf group, n = 9); and (4) 100 mg surfactant phospholipids/kg and inhalation of 20 ppm NO (SNO group, n = 8). Surfactant was instilled into the animal's lungs via an endotracheal tube at 2.5 ml/kg while the animal's body was rotated. All the animals were subsequently ventilated with PSV at identical VT and E for another 12 h, or until early death. Values for arterial Pa_O2, Pa_CO2, pH, Cdyn, Rrs, VT, and E were measured each hour. Set PSV level (cm H₂O), measured respiratory rate (RR, times per min), peak insufflation pressure (PIP, cm H₂O), and mean airway pressure (, cm H₂O) during PSV were recorded each hour from monitoring indicator of the Wave ventilator. S/T was measured at 6 and 12 h of the treatment. During the experiment, intravenous lactated Ringer's solution (pH, 4.5 to 7.0) was infused at 5 to 8 ml/kg/h, and 2.5% bicarbonate sodium was given to overcome metabolic acidosis when necessary. At the end of the period of ventilation, animals were killed by an overdose of 5% pentobarbital sodium and the animal lungs were processed (see below). For animals that survived less than 12 h, measured values for physiologic parameters recorded before deterioration of heart rate and blood pressure were included in the final analysis, and their lungs were processed immediately after death.

In the 47 rabbits, eight received intravenously a sham volume of 0.9% NaCl instead of oleic acid and were subjected to the protocol. As these animals breathed freely and did not need PSV for maintenance of ventilation, CPAP only was applied, but the breathing volume was confirmed to be 8 to 10 ml/kg, and lung function was measured in the same manner as for the oleic acid-injured animals. These animals were killed when they completed all the measurements of physiologic parameters at baseline and treatment times 0, 1, 6 and 12 h. These animals were designated as normal control (Normal) and their lungs (four animals each) were processed identically as those of the oleic acid-exposed animals subjected to either histologic examination and morphometric analysis or to measurements of phospholipids and proteins in bronchoalveolar lavage (BAL) and wet-to-dry lung weight ratio (W/D). Another four of the 47 animals were killed after the baseline measurement of physiologic parameters, and their lungs were processed the same as those for BAL and W/D in the Normal group, and these animals were designated as the Normal₀ group.

Wet-To-Dry Lung Weight Measurements and Bronchoalveolar Lavage

Methods for BAL and W/D measurement have been reported elsewhere (16). Briefly, after inspection of pneumothorax, in each group, four animals were subjected to W/D measurement and right lung lavage, and the supernatant of the bronchoalveolar lavage fluid (BALF) after low speed centrifugation was used for chemical and biophysical analysis of surfactant phospholipids (see below).

Chemical Analysis of Bronchoalveolar Lavage Fluid

Processing of BALF was performed according to protocols reported elsewhere (16). Amounts of DSPC and TPL were determined according to Mason and colleagues (17) and Bartlett (18), respectively, and corrected by the total volume of BALF and body weight. Values for TPL are presented as mg/kg, and those for DSPC as percentage of TPL (DSPC/TPL). Total proteins (TP) in BALF were measured according to Lowry and colleagues (21), corrected by total volume of BALF and body weight and are presented as mg/kg. The DSPC and TP ratio was expressed as µg/mg.

Surface Tension Measurements

Measurements of surface tension of surfactant extracted from BALF were performed with a pulsating bubble surfactometer (PBS; Electronetics Corp., Buffalo, NY). After determination of TPL, aliquots of BAL fluid were extracted with chloroform/methanol (2:1, vol/vol), and the amount of TPL in the chloroform phase was weighed after evaporation of chloroform under nitrogen gas. It was resuspended in 0.9% NaCl to yield a final concentration of TPL at approximately 5 mg/ml. Values for minimum and maximum surface tension (min, max, respectively) were obtained at minimum and maximum bubble size (radius, 0.4 and 0.55 mm), respectively, corresponding to 50% change of the bubble surface area. Pulsating rate of PBS was set at 25 cycles/min and the sample chamber was kept at 37° C. Values of min and max at 5 min were averaged from four determinations for each sample. The batch of surfactant preparation used in this study was suspended in 0.9% NaCl at a phospholipid concentration of 5 mg/ml.

Measurements of Nitrite/Nitrate and Methemoglobin (MetHb)

In order to evaluate adverse effect of iNO, blood and urine samples representing baseline and treatment times 0, 6, and 12 h were taken for measurement of nitrite and nitrate using a modified Griess method as described by Shi and colleagues (22), and values were expressed as µmol/L. MetHb was determined according to the method described by Hegesh and colleagues (23), and expressed as percentage of total hemoglobin.

Histologic and Morphometric Examination of Lungs

Four or five animals of each group were used for histologic examination of the lungs. The lungs were perfused for 30 min via the pulmonary arteries with 4% formaldehyde at a pressure of 65 cm H₂O while both lungs were first inflated to an airway pressure of 30 cm H₂O for 1 min, and then deflated to 10 cm H₂O for the rest of time. Representative lung tissue blocks from all lung lobes were embedded in paraffin. Sections stained with hematoxylin-eosin were examined by light microscopy for evidence of lung injury, and the lung injury was scored for edema, neutrophil infiltration, hemorrhage, bronchiole epithelial desquamation, and hyaline membrane formation. A score scaled at 0 to 4 represents the severity: 0 for no or very minor, 1 for modest and limited, 2 for intermediate, 3 for widespread or prominent, and 4 for widespread and most prominent. Lung expansion was quantified using the point-counting method, and expressed as volume density (V_V) of aerated alveolar spaces, using total parenchyma as reference volume (24). Fifty fields of each lung section were examined from each animal (magnification: ×300), and field-to-field variability was determined by calculating the coefficient of variation of V_V (CV[V_V]), and a low value of CV[V_V] indicates homogeneity of alveolar aeration. These works were performed in a blinded manner so that investigators were not able to identify the treatment and outcome of the animals until completion of the measurement.

Statistics

Data are presented as means and standard deviation (SD). Survival rate was examined with chi-square and Fisher's exact tests. Continuous parametric data were subjected to analysis of variance (ANOVA) followed by the Student-Newman-Keuls post-hoc test for between-group differences, and by Wilcoxon's signed-rank test for within-group differences. For the lung injury score, a Kruskal-Wallis test was used to detect differences across the groups, followed by the Wilcoxon-Mann-Whitney test for differences between the two groups. A p value  0.05 was regarded as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General Conditions of the Animals

Animal body weight was similar across the groups. Infusion of oleic acid induced, in all the animals, respiratory failure in 60 ± 24 min, and no significant difference was found among the four groups. In these animals, heart rate and Psa at baseline were 244 ± 24/min and 73.2 ± 10.7 mm Hg, and at treatment time 0, maintained at 230 ± 27/min and 60.8 ± 12.1 mm Hg, respectively (p < 0.01 compared with the corresponding values at baseline). At 6 h, these parameters were 237 ± 44/ min and 61.2 ± 12.9 mm Hg, respectively. Bradycardia and hypotension were found in some animals at different period of treatment. Survived animal numbers at different experimental time are shown in Figure 1A. One animal in the Control group, three each in the NO and Surf groups, and five in the SNO group, survived 12 h (Control versus SNO, p < 0.05). The early death of the animals was mainly due to cardiac depression in association with severe hypoxemia, and acidosis. Concentration of iNO in the NO and SNO groups was kept at 20 ppm with a variation less than 1.0 ppm, and NO₂ was lower than 1.5 ppm throughout the treatment period.

View larger version (21K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 1.   (A) Survived animal numbers. (B) PaO2/FIO2. (C ) Dynamic compliance (Cdyn). (D) Resistance of respiratory system (Rrs) at different experimental times. Labels and group definitions: open circles = Control; closed circles = nitric oxide (NO); open triangles = surfactant (Surf ); closed triangles = surfactant and nitric oxide (SNO); open squares = normal animals without oleic acid injury (Normal). The baseline is the moment before induction of lung injury by oleic acid. Labels and ticks are means and SD. Statistical analysis: SNO versus Control at 12 h, p < 0.05 for survival number. Values of PaO2/FIO2, Cdyn, and Rrs at treatment time 0 h versus baseline, p < 0.01 (within-group) for Control, NO, Surf, and SNO groups; at 2 h, NO versus Control, p < 0.01, and NO versus Surf and SNO, p < 0.05 for PaO2/FIO2; at 6 h, Control, NO, Surf, and SNO versus Normal, p < 0.01 for PaO2/FIO2 and Cdyn; at 6 h, NO and SNO versus Control, p < 0.05 for PaO2/FIO2; at 12 h, Control, NO, Surf, and SNO versus Normal, p < 0.01 for PaO2/FIO2; at 12 h, NO, Surf, and SNO versus Control, p < 0.05 for PaO2/FIO2, and Cdyn.

Lung Function Measurements

Tidal volume was kept at 8 to 10 ml/kg during the whole experimental peroid with no significant difference across the groups. Values for Pa_O2/FI_O2, Cdyn, and Rrs at baseline level and during the treatment period are shown in Figures 1B- 1D. Values for S/T are shown in Figure 2. The respiratory failure was evidenced by marked decrease of Pa_O2/FI_O2 and Cdyn (Figures 1B and 1C), and increase of Rrs and S/T (Figures 1D and 2). Infusion of oleic acid resulted in a decrease of mean Pa_O2 from 94 to 48 mm Hg (p < 0.01) when the animals breathed room air. This corresponds to a reduction of mean Pa_O2/FI_O2 from 445 to 235 mm Hg. Also, mean values for Cdyn were decreased from 1.59 ml/cm H₂O/kg at baseline to 0.85 ml/cm H₂O/kg at treatment time 0 h (p < 0.01), respectively, and Rrs were increased from 38 to 45 cm H₂O/L/s (p < 0.01), respectively. During treatment period, Cdyn and Rrs were variable and did not reach the normal level (Figures 1C and 1D). E was also increased from 0.46 to 0.58 L/min/kg (p < 0.01), whereas the mean breath rate was increased from 48 to 90 breaths/min (p < 0.01) before PSV was applied. At the beginning of the randomized treatment, PSV was applied higher in the Surf and SNO groups than in the Control and NO groups (p < 0.05), adjusted to maintain Pa_CO2 within a range of 40 to 50 mm Hg and pH at 7.20 to 7.40. Relatively low level of mean PSV was applied to the NO group (10.3 ± 3.4 cm H₂O) compared with the Surf and SNO groups (14.3 ± 2.6 and 14.3 ± 3.8 cm H₂O, respectively, p < 0.05) at 6 h of the treatment, but at 12 h there were no differences between the three treated groups (about 12 cm H₂O). Measured PIP levels were about 7 cm H₂O higher than the corresponding PSV, but levels were kept at 6 to 8 cm H₂O during the treatment. Mean values of RR and E were kept at 65 breaths/min and 0.6 L/ min/kg, respectively, among the four treatment groups, which enabled adequate blood gas exchange and survival. Values for S/T were 9.4% at baseline and increased to 27% at the treatment time 0 h while the animals breathed spontaneously, but there was no significant difference across the groups receiving oleic acid (Figure 2). During the treatment, significant reduction of S/T was found in the NO, Surf, and SNO groups at 6 and 12 h. In the Control group, there was persistent deterioration of lung function and early death under the concurrent ventilation settings applied. In the NO, Surf and SNO groups, there were modest improvements in Pa_O2/FI_O2, Cdyn, and Rrs (Figures 1B-1D). In the Normal group, these parameters did not deteriorate during in the whole experimental period, and PIP, , RR, and E were lower than in those exposed to oleic acid injury and various treatments.

View larger version (31K): [in this window] [in a new window]   Figure 2.   Intrapulmonary shunting at baseline, treatment time 0, 6, and 12 h. The baseline is before induction of lung injury by oleic acid. Group definitions: C = Control; NO = nitric oxide; Surf = surfactant; SNO = surfactant and nitric oxide; Normal = normal lungs not injured by oleic acid. Bars and ticks are means and SD. Statistical analysis: Values for Control, NO, Surf, and SNO groups at treatment time 0 h versus baseline p < 0.05 (within-group); at 6 and 12 h, NO, Surf, SNO, and Normal versus Control, p < 0.01, NO, Surf, and SNO versus Normal, p < 0.05. For animal numbers at each time point refer to Figure 1A.

Chemical Analysis of BALF

Values for TPL, DSPC/TPL, TP, and DSPC/TP in BALF are shown in Table 1. TPL and DSPC/TPL in BALF were significantly higher in the Surf and SNO groups than in the Control and NO groups. TP in BALF was significantly lower in the NO, Surf, and SNO groups than in the Control group, the SNO group being the lowest one. DSPC/TP values in the Surf and SNO groups were also significantly higher than that in the Control and NO groups, and even matching that of the Normal group (for the SNO group).

Surface Tension Measurements

When the exogenous surfactant suspension was measured in PBS, mean values of min and max reached 1 and 30 mN/m at 1 min, and were maintained at 2 and 30 mN/m at 5 min, respectively. Values for min and max of phospholipids from BALF after 5 min of pulsation are shown in Table 1. In the Control and NO groups, min and max remained > 15 and 35 mN/m, respectively, throughout the period of pulsating, whereas in the Surf and SNO groups, mean values of min and max were significantly lower than those of the Control and NO groups, but didn't reach that of the Normal and Normal₀ groups in which even lower surface tension values were observed.

Measurements of Nitrite/Nitrate and Methemoglobin

Values of nitrite/nitrate in serum and urine and blood MetHb at baseline were 87 ± 27 µmol/L, 500 ± 250 µmol/L, and 1.0 ± 0.6%, respectively, and tended to increase during inhalation of NO in the NO and SNO groups, but the highest levels of these parameters were 166 ± 25 µmol/L, 705 ± 230 µmol/L, and 1.4 ± 1.2%, respectively, at 12 h of treatment, with no significant difference across the groups.

Wet-to-Dry Lung Weight Ratio

Values of W/D in various groups of experimental animals are shown in Table 2. Low values for W/D were observed in the three treated groups, and the lowest level was found in animals in the SNO group. However, these values were still higher than that of the Normal and Normal₀ groups.

Histological and Morphometric Findings

Results from morphometric analysis for alveolar expansion are shown in Table 2. In the NO, Surf, and SNO groups, aeration of alveoli was significantly improved compared with that in the Control group, as reflected by increased V_V, but values for CV (V_V) in these groups were similar. As shown by lung injury score in Table 3, there were prominent edema, infiltration of neutrophils, and intrapulmonary hemorrhage in the lungs of animals in the Control and NO groups. In the Surf and SNO groups, there was moderately improved aeration of alveoli, but significantly alleviated edema, hemorrhage and bronchiole and alveolar epithelial desquamation in the SNO group. Photomicrographs representative of lung aeration and injury patterns of each group are shown in Figure 3.

View larger version (154K): [in this window] [in a new window]   View larger version (130K): [in this window] [in a new window]   View larger version (106K): [in this window] [in a new window]   View larger version (101K): [in this window] [in a new window]   View larger version (102K): [in this window] [in a new window]   Figure 3.   Photomicrographs of histologic findings in the animal lungs. All the lungs were fixed by pulmonary arterial perfusion at deflation pressure of 10 cm H2O for 30 min. (A) Control (PSV only), there are atelectasis, edema, hemorrhage, neutrophil infiltration, and desquamation and necrosis of bronchiole epithelial cells. (B) Nitric oxide (NO), there are prominent edema and neutrophil infiltration, but aeration was improved. (C ) Surfactant (Surf ), neutrophil infiltration, and desquamation of bronchiole epithelial cells are still visible, but edema was markedly attenuated. (D) Surfactant and iNO (SNO), only moderate neutrophil infiltration was present. (E ) Normal lungs without oleic acid injury (Normal). Hematoxylin-eosin stain; original magnification ×60.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Oleic acid-induced edema and lung injury has been well established as a pathologic model for studies of mechanisms as well as various therapies in ALI and ARDS (25). Attempts made by Putensen and colleagues (26) have demonstrated that CPAP (10 cm H₂O, FI_O2, 0.3) and iNO improved ventilation-perfusion in canine ALI because of vasoconstriction induced by inhibition of endogenous NO production. The current study was designed to evaluate preventive effects of combined surfactant and iNO with PSV. Our main finding was that such a combined treatment resulted in improved ventilation efficiency, gas exchange and survival rate, and attenuated lung injury. The treatment for NO, Surf, and SNO groups was implemented at a relatively early stage of ALI, and it resulted in different responses that are related to ventilation style and surfactant.

PSV is a flow-cycled, assisted ventilation mode, initiated by spontaneous breathing and ended as gas flow, that is reduced to less than 30% of the peak flow, or a preset peak insufflation pressure is attained. It is usually used in conjunction with synchronized intermittent mandatory ventilation or CPAP to compensate minute ventilation volume and reduce work of breathing. Several recent clinical studies have reported that PSV is effective in improving gas exchange and ventilation-perfusion distributions in patients with pulmonary edema or ARDS (12- 14). We thought that PSV, when used together with CPAP, should maintain adequate ventilation for early ALI in this animal model, so that effects of surfactant and/or iNO therapy and their interactions could be evaluated. The data revealed that use of PSV only didn't show substantial effects in improving blood oxygenation and preventing lung injury, although VT and Pa_CO2 were kept to the targeted ranges. In contrast, a combined surfactant and iNO with PSV did exert such effects as evidenced by improved Pa_O2/FI_O2, Cdyn, S/T, DSPC/TP, W/D, lung morphology, and injury scores, and prolonged survival time, suggesting that preventive effects of PSV alone at early ALI in this model is very limited unless other therapies are combined.

Mechanism of the effects by surfactant, iNO, and PSV probably relies on a synergy of these therapies for recruitment of gas exchange units at distal alveolar compartments, alteration of vascular-to-alveolar permeability, and relaxation of intrapulmonary vasculature in the injured lungs (14, 16, 26). In an oleic acid-induced pig ALI model, Shah and colleagues (27) found that iNO siginificantly alleviated pulmonary vascular resistance and intrapulmonary shunting in a dose-dependent manner. In our present study in the NO group, iNO at 20 ppm markedly improved Pa_O2 whereas PSV was kept at relatively low level (10 cm H₂O) during the first 6 h of treatment, and the effect was less prominent in the following 6 h. The lungs in this group showed only modest improvement in aeration and W/D but prominent edema, neutrophil recruitment, hemorrhage, and desquamation of small airways. In contrast, animals in the Surf and the SNO groups had moderate but consistent improvement of lung function while lung injuries were ameliorated. In all the three treated groups, low levels of , S/T, and W/D were associated with improved lung function and relatively low score of lung injury, especially in the SNO group, which is similar to our previous findings in oleic-acid-injured animals that had the synergistic therapeutic effects of surfactant and iNO as CMV at higher airway pressure and FI_O2 was needed (16), confirming the advantage of the current treatment modality for ALI.

Severe damages of the lungs may be associated with impairment of surfactant system secondary to the oleic acid injury. The significantly elevated surface tension of isolated phospholipids from BALF in both Control and NO groups may account for the respiratory distress and ventilation insufficiency, and is in accordance with the findings of Ito and colleagues (28), in which CMV with a higher VT resulted in more pronounced lung function impairment than with a lower VT because of a degradation of endogenous surfactant. As in this study the PIP to maintain VT was relatively low for the PSV group, and VT and CPAP levels were kept constant at 8 to 10 ml/kg and 3 cm H₂O, respectively, throughout the treatment period, the difference of endogenous surfactant degradation across the groups caused by tidal ventilation should be minimized. We found that the endogenous surfactant content in the control animals was comparable to that in the normal animals, but the ex vivo surface activities were significantly inferior, indicating that the functional abnormality of the endogenous surfactant in this model was related to the initial injury caused by oleic acid and subsequent development of proteinaceous edema in the lungs, leading to respiratory distress. In the present study, about 150% more TPL in BALF was found in the Surf and the SNO groups than in the Control and the NO groups, corresponding to 20% of the exogenous surfactant dose (about 10 mg/kg compared with 50 mg/kg for one side lung) from the alveolar compartment. Likewise, DSPC in TPL in the SNO group was also significantly higher than that in the Control group. These might account for the 50% reduction of TP and relatively lower surface tension of TPL in BAL, improved alveolar aeration, attenuated edema, and alleviated bronchiole and alveolar damages, along with prolonged survival time in the animals receiving surfactant, especially in the SNO group. Thus, it appears that the effects exerted by surfactant in the Surf and the SNO groups were mainly on alleviation of the injury. Although physiologic, biophysical, and histopathologic data didn't show significant differences between the Surf and the SNO groups, it is possible that exogenous surfactant facilitated NO diffusion and altered ventilation-perfusion mismatching in the injured lungs. A long-term experiment is required to verify sustained effects of this combination, and optimal exogenous surfactant dose-related efficacy and metabolism, in the injured lungs.

We conclude that a combined use of surfactant and iNO during PSV for experimental ALI and respiratory failure is relevant in improving gas exchange and lung mechanics, and ameliorating injuries. iNO with PSV reduced intrapulmonary shunting and improved blood oxygenation, but they had very limited impact on preventing or alleviating lung damage. Exogenous surfactant with PSV protected alveolar epithelial cells from injury and improved intrapulmonary shunting. A combined surfactant, iNO, and PSV exerted further effects on alveolar expansion, lung fluid adsorption, and prolonged survival time. Advantages of this combination need to be corroborated in even longer term experiment and clinical investigation.