Abstract
A randomized, controlled clinical trial was performed with patients with acute respiratory distress syndrome (ARDS) to compare the effect of conventional therapy or inhaled nitric oxide (iNO) on oxygenation. Patients were randomized to either conventional therapy or conventional therapy plus iNO for 72 h. We tested the following hypotheses: (1) that iNO would improve oxygenation during the 72 h after randomization, as compared with conventional therapy; and (2) that iNO would increase the likelihood that patients would improve to the extent that the Fi O2 could be decreased by ⩾ 0.15 within 72 h after randomization. There were two major findings. First, That iNO as compared with conventional therapy increased PaO2 /Fi O2 at 1 h, 12 h, and possibly 24 h. Beyond 24 h, the two groups had an equivalent improvement in PaO2 /Fi O2 . Second, that patients treated with iNO therapy were no more likely to improve so that they could be managed with a persistent decrease in Fi O2 ⩾ 0.15 during the 72 h following randomization (11 of 20 patients with iNO versus 9 of 20 patients with conventional therapy, p = 0.55). In patients with severe ARDS, our results indicate that iNO does not lead to a sustained improvement in oxygenation as compared with conventional therapy.
Short-term exposure to inhaled nitric (iNO) can decrease pulmonary vascular resistance and improve oxygenation in patients with the acute respiratory distress syndrome (ARDS) (1-13). Most reported studies of this have, however, lacked control groups, and have primarily evaluated the physiologic effects of iNO for 10 to 160 min. When studies have provided information about chronic therapy, they have generally highlighted the physiologic consequences of stopping iNO for a few minutes, rather than the effects of iNO compared with baseline or the response over time. Consequently, little information exists about the long-term efficacy of iNO on oxygenation in ARDS patients as compared with a control group.
We undertook a randomized, controlled clinical trial to answer two major questions: (1) Does iNO therapy improve oxygenation as compared with conventional therapy during the 72 h after patient randomization? (2) Does iNO increase the likelihood that patients will improve sufficiently to be managed with a persistent decrease in Fi O2 ⩾ 0.15 within the 72 h following randomization? We also addressed two secondary issues. Since most studies have emphasized the short-term response to iNO, we investigated whether the acute change in PaO2 /Fi O2 after 1 h of iNO correlates with the response after 72 h. We also looked for specific clinical features that would predict improved oxygenation over 72 h.
From January 1994 to June 1996, 40 patients with ARDS from the intensive care units (ICUs) at the University of Utah Hospital were entered into a randomized, controlled clinical trial to compare conventional therapy with iNO on oxygenation during the 72 h after randomization. The protocol was approved by the University of Utah Institutional Review Board. Informed consent was obtained from each patient's family. The inclusion and exclusion criteria are shown in Table 1. All patients meeting these criteria were eligible for entry.
| Inclusion criteria | ||
| A. ARDS | ||
| (1) Acute respiratory failure requiring mechanical ventilation. | ||
| (2) Bilateral pulmonary radiographic infiltrates. | ||
| (3) PaO2 /Fi O2 < 150. | ||
| (4) No clinical evidence for left atrial hypertension or heart failure as the primary disorder or, if a pulmonary artery catheter was present, a pulmonary arterial occlusion pressure ⩽ 18 mm Hg. | ||
| B. Patient receiving immediately before randomization Fi O2 ⩾ 0.8 for at least 12 h or Fi O2 ⩾ 0.65 for at least 24 h. | ||
| Exclusion criteria | ||
| A. Pregnancy | ||
| B. Patients not expected to survive hospitalization because of underlying disease, such as active malignancy. | ||
Patients from the Medical Intensive Care Unit (MICU), Surgical Intensive Care Unit (SICU), and Intermountain Burn Center (Burn Center) were eligible for inclusion. Patients were randomized within each ICU, using balanced blocks of 14 patients. After randomization, patients received either conventional therapy or conventional therapy plus 5 to 20 ppm iNO for the next 72 h. The primary endpoint for efficacy was improvement in oxygenation within 72 h following randomization. This and other study definitions are provided in Table 2. We selected this endpoint because of our impression that it would represent significant clinical improvement. During the 72 h after randomization, Fi O2 was adjusted to maintain PaO2 between 55 and 65 mm Hg, or SaO2 between 88% and 92%.
| Improvement in oxygenation = a decrease in Fi O2 ⩾ 0.15 that began within 72 h after randomization and persisted for at least 24 h. |
| ARDS reversal = patient managed for ⩾ 24 h on an Fi O2 < 0.5. |
| Survival = patient discharged alive from hospital. |
| Decrease in Fi O2 over the 72 h after randomization = the difference in Fi O2 between times 0 h and 72 h. |
If oxygenation was not improved after 72 h of randomized treatment, patients assigned to conventional therapy began treatment with iNO. Patients randomized to iNO therapy who did not meet the definition for oxygenation improvement within 72 h were slowly tapered off iNO over several days. Patients receiving conventional therapy could be crossed over to iNO2 earlier if they met predefined criteria for clinical deterioration.
NO was delivered through a Siemens Servo 900 C or Servo 300 ventilator (Siemens, Danvers, MA), using a bleed-in adapter placed in the inspiratory limb of the circuit, just distal to the humidifier. The volume of tubing from the site of NO delivery to the patient Y-connector was 537 ml at atmospheric pressure and 597 ml when exposed to a pressure of 20 cm H2O. NO was supplied from tanks containing 2,200 ppm, with the balance gas being nitrogen (A-L Compressed Gases, Salt Lake City, UT). For pediatric patients, tanks containing 800 ppm NO were used. A high-purity, corrosion-resistant regulator with a very low flowmeter (25 to 200 ml/min) was used to regulate the flow rate. Inspired gas was sampled 185 cm downstream from the bleed-in port and 15 cm proximal to the patient Y-connector. NO and NO2 concentrations were measured with an electrochemical sensor (Exidyne Instrumentation Technologies, Exton, PA). The exhaled gas was scavenged with a Nitric Oxide Scavenger–Vacuum Safety Interface (Boehringer Laboratories Inc., Norristown, PA). Methemoglobin concentrations were measured in all blood gas analyses.
The iNO concentration was controlled by the patient's attending physician (all coinvestigators). The suggested protocol was to begin with 5 ppm. During the first 24 h, the concentration was generally increased by 5 ppm at 6-h intervals, so that a concentration response at 5, 10, 15, and 20 ppm was obtained. After the first 24 h, the concentration was adjusted according to the patient's condition and response to the various NO concentrations. iNO therapy was continuous. During the 72 h after randomization, the mode of mechanical ventilation was left unchanged.
The following data were collected for each patient: demographic information, acute physiology and chronic health evaluation-III (APACHE III) score (14) at ICU admission and randomization, presence or absence of sepsis (15), conditions associated with ARDS, lung injury score (LIS) (16), cardiopulmonary physiologic measurements, and outcome. Acute organ dysfunction was identified for the following systems: cardiovascular (17), gastrointestinal (17), neurologic (18), coagulation (18), hepatic, and renal. Acute hepatic failure was defined as a total bilirubin > 8 mg/dl, alanine aminotransferase (ALT) > twice normal, or encephalopathy. Acute renal failure required a serum creatinine (Cr) > 3 mg/dl, increase in serum Cr ⩾ 1.5 mg/dl, 24-h urine output < 600 ml or < 1 ml/kg/h, or hemodialysis or ultrafiltration (excluding patients receiving dialysis before admission). A nonpulmonary multiple-organ dysfunction score was calculated by assigning one point for each organ dysfunction. To evaluate any effect on bleeding, we noted acute hemorrhage requiring blood transfusion.
The following cardiopulmonary measurements were obtained: PaO2 /Fi O2 , positive end expiratory pressure (PEEP), mean systemic arterial pressure, central venous pressure (CVP), mean pulmonary arterial pressure (Ppa), pulmonary arterial occlusion pressure Ppao, pulmonary vascular resistance (PVR), systemic vascular resistance, and thermodilution cardiac output (CO) averaged from at least three measurements. Quasi-static respiratory system compliance (Cqs) was obtained by dividing the Vt by the plateau airway pressure minus PEEP. In patients receiving iNO, the Fi O2 values used in calculations were reduced by 1% to adjust for the dilution by the nitrogen in the iNO gas. We recorded PEEP, PaO2 , Fi O2 , and PaO2 /Fi O2 at 12-h intervals during the 72 h after randomization, using the values closest to the respective time points.
We selected a sample size of 40 subjects because it would allow us to detect a difference of between 35 and 40% in the frequency of a persistent decrease in Fi O2 ⩾ 0.15, assuming a two-tailed test and α = 0.05 and β = 0.80.
Data were analyzed with Student's t test for paired and unpaired data, repeated measures analysis of variance (ANOVA), linear regression, and contingency table analysis with Fisher's exact test. Values are presented as mean ± SEM. Sample size estimates were calculated according to standard approaches (19).
The families of 41 patients with severe ARDS were asked to have their affected family member participate in the study. One family refused participation; 40 patients were enrolled. The patient characteristics at randomization are shown in Table 3. Twenty-two patients were enrolled from the MICU, 13 from the SICU, and five from the Burn Center. The patients in the two groups had similar characteristics at randomization (Table 4).
| Age/Sex | Underlying Disease | Etiology or Risk Factor | Lung Injury Score | APACHE III Score | Acute Organ Dysfunction | ARDS Duration (d ) | Fi O2 | PEEP (cm H2O) | PaO2 (mm Hg) | Outcome | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Randomized to conventional therapy | ||||||||||||||||||||
| 38 M | Sepsis, pneumonia | 3.25 | 40 | 1 | 1.00 | 10 | 57 | Died | ||||||||||||
| 32 F | HIV | Sepsis | 3.75 | 76 | CV | 6 | 1.00 | 19 | 53 | Died | ||||||||||
| 79 F | Sepsis | 3.25 | 74 | 6 | 1.00 | 7.5 | 43 | Died | ||||||||||||
| 45 F | Liver | Sepsis, pneumonia | 3.00 | 130 | CV, N, C | 2 | 1.00 | 5 | 53 | Survived | ||||||||||
| 39 F | Liver | Sepsis, pneumonia | 4.00 | 58 | C, GI | 9 | 0.90 | 18 | 65 | Survived | ||||||||||
| 35 M | Sepsis, pneumonia | 3.75 | 35 | CV, C | 16 | 1.00 | 15 | 55 | Survived | |||||||||||
| 31 M | Pancreatitis | 3.75 | 133 | CV, R, GI | 0.5 | 1.00 | 23 | 49 | Died | |||||||||||
| 2 M | Burn | 3.75 | — | 2 | 0.70 | 17 | 75 | Survived | ||||||||||||
| 57 F | Smoke inhalation | 4.00 | 56 | CV | 16 | 1.00 | 23 | 78 | Died | |||||||||||
| 35 M | Aspiration | 4.00 | 50 | CV, R, GI | 2 | 1.00 | 15 | 73 | Survived | |||||||||||
| 31 F | Sepsis, pneumonia | 4.00 | 40 | 6 | 1.00 | 17 | 62 | Died | ||||||||||||
| 31 M | Aspiration | 3.75 | 69 | CV, C | 1 | 1.00 | 20 | 34 | Survived | |||||||||||
| 27 M | Trauma | 3.50 | 98 | 0.5 | 0.90 | 22 | 62 | Survived | ||||||||||||
| 69 M | Hypertransfusion | 3.50 | 100 | CV, R, C, GI | 1 | 0.75 | 16 | 66 | Survived | |||||||||||
| 60 F | Pancreatitis | 3.75 | 103 | CV, H, C, GI | 1 | 1.00 | 15 | 56 | Died | |||||||||||
| 46 M | Sepsis | 3.75 | 68 | C | 6 | 1.00 | 20 | 62 | Survived | |||||||||||
| 57 M | Heart transplant | ?Amiodarone | 3.75 | 71 | 1 | 0.90 | 17 | 55 | Died | |||||||||||
| 24 F | Aspiration | 3.50 | 73 | CV, H, C | 0.5 | 1.00 | 18 | 56 | Survived | |||||||||||
| 2 M | Burn | 3.75 | — | C | 1 | 1.00 | 12 | 58 | Survived | |||||||||||
| 79 F | S/P bypass | 3.50 | 54 | CV | 1 | 1.00 | 12 | 45 | Died | |||||||||||
| Randomized to iNO therapy | ||||||||||||||||||||
| 48 F | Sepsis | 3.00 | 65 | GI | 2 | 0.80 | 5 | 66 | Survived | |||||||||||
| 43 M | Liver | Aspiration, pancreatitis | 4.00 | 55 | GI | 9 | 1.00 | 16 | 66 | Survived | ||||||||||
| 45 F | Sepsis | 3.75 | 76 | CV | 0.5 | 0.75 | 12 | 46 | Died | |||||||||||
| 19 M | Sepsis | 3.75 | 98 | CV, C | 0.5 | 1.00 | 16 | 53 | Survived | |||||||||||
| 26 F | Sepsis, pneumonia | 3.25 | 43 | 1 | 0.90 | 10 | 67 | Survived | ||||||||||||
| 17 F | Aspiration | 4.00 | 62 | C | 4 | 0.80 | 15 | 68 | Survived | |||||||||||
| 32 M | Liver | Aspiration | 3.50 | 76 | CV, H, C | 4 | 1.00 | 17 | 49 | Died | ||||||||||
| 18 M | Sepsis, pneumonia | 4.00 | 67 | 6 | 1.00 | 16 | 33 | Died | ||||||||||||
| 47 M | Sepsis, pneumonia | 4.00 | 57 | CV, C | 5 | 0.85 | 15 | 57 | Died | |||||||||||
| 33 M | Sepsis | 3.50 | 65 | CV, R, C | 3 | 0.80 | 14 | 77 | Died | |||||||||||
| 55 F | Sepsis | 3.50 | 63 | CV, H, C | 25 | 0.75 | 10 | 61 | Died | |||||||||||
| 21 M | 75 d S/P BMT | Sepsis, pneumonia | 3.75 | 118 | 3 | 1.00 | 15 | 60 | Died | |||||||||||
| 24 F | Smoke inhalation | 4.00 | 49 | H, C | 4 | 0.80 | 20 | 68 | Died | |||||||||||
| 23 M | SLE | Sepsis, pneumonia | 3.75 | 57 | R, C | 9 | 1.00 | 16 | 53 | Died | ||||||||||
| 38 M | Sepsis | 3.75 | 131 | CV, R | 0.5 | 1.00 | 15 | 38 | Died | |||||||||||
| 44 M | Sepsis | 3.75 | 52 | CV, C | 2 | 0.90 | 16 | 72 | Survived | |||||||||||
| 1 M | Burn | 4.00 | — | CV | 16 | 0.98 | 16 | 66 | Survived | |||||||||||
| 34 M | Liver | Sepsis, pancreatitis | 4.00 | 73 | CV, C, GI | 3 | 1.00 | 20 | 71 | Survived | ||||||||||
| 30 F | Trauma | 4.00 | 90 | N | 18 | 1.00 | 20 | 46 | Died | |||||||||||
| 18 M | Trauma | 3.75 | 57 | N, C | 1 | 0.90 | 20 | 55 | Survived | |||||||||||
| Conventional (n = 20) | Nitric Oxide (n = 20) | p Value | ||||
|---|---|---|---|---|---|---|
| Age, yr | 41 ± 5 | 31 ± 3 | 0.08 | |||
| Male/female | 12/8 | 13/7 | 0.75 | |||
| Sepsis | 8/20 | 13/20 | 0.13 | |||
| Quasi-static respiratory compliance | ||||||
| ⩽ 19 ml/cm H2O† | 7/19 | 9/17 | 0.50 | |||
| Fi O2 | 0.96 ± 0.02 | 0.91 ± 0.02 | 0.12 | |||
| PEEP, cm H2O | 16 ± 1 | 15 ± 1 | 0.52 | |||
| PaO2 , mm Hg | 58 ± 2 | 59 ± 3 | 0.83 | |||
| PaO2 /Fi O2 , mm Hg | 62 ± 4 | 64 ± 4 | 0.65 | |||
| SaO2 , % | 83 ± 2 | 85 ± 1 | 0.61 | |||
| PaCO2 , mm Hg | 44 ± 3 | 42 ± 3 | 0.57 | |||
| Tidal volume, ml/kg | 7.4 ± 1 | 7.7 ± 1 | 0.81 | |||
| Lung injury score | 3.66 ± 0.06 | 3.75 ± 0.06 | 0.32 | |||
| ARDS duration before | ||||||
| randomization, d | 4 ± 1 | 6 ± 1 | 0.31 | |||
| APACHE III score at | ||||||
| randomization | 74 ± 7 | 71 ± 5 | 0.77 | |||
| Nonpulmonary multiple-organ- | ||||||
| system dysfunction score | 1.55 ± 0.3 | 1.60 ± 0.2 | 0.89 | |||
The average iNO concentrations during the 72 h after randomization are shown in Figure 1. NO2 measurements never exceeded 1 ppm. iNO did not increase methemoglobin, which was measured in all blood gas analyses. The measured NO concentrations in the inspired gas corresponded with those predicted on the basis of the concentration of NO in the gas tank and its flow rate as a percentage of total inspiratory flow.
Fig. 1. Mean concentrations of inspired NO during the 72 h after randomization. Values are mean ± SEM (n = 16).
[More] [Minimize]Conventional therapy. Twenty patients were randomized to conventional therapy. Two patients died within the 72 h after randomization. One patient met a criterion for clinical deterioration, and at 36 h after randomization was crossed over to iNO. In a protocol violation, one patient who had not improved was crossed over to iNO after 48 h instead of 72 h. Sixteen patients received conventional therapy for the full 72 h after randomization.
iNO. Twenty patients were randomized to iNO. Four patients died within 72 h after randomization. Sixteen patients received iNO for the full 72 h after randomization.
In patients randomized to iNO, 1 h of therapy significantly increased PaO2 /Fi O2 (baseline: 64 ± 4, versus 87 ± 6; p = 0.0004, n = 20). Criteria used to define a short-term oxygenation response to iNO include an increase in PaO2 /Fi O2 ⩾ 20% (11, 13, 20) or an absolute increase ⩾ 10 mm Hg (9, 10, 21). Sixty percent of patients randomized to iNO met one or both criteria after 1 h. Although patients randomized to conventional therapy did not routinely have arterial blood gas samples drawn at 1 h after randomization, NO therapy probably increased PaO2 / Fi O2 as compared with conventional therapy, since in the conventional therapy group the PaO2 /Fi O2 at 12 h was not significantly different from the baseline value.
We compared PEEP, PaO2 , Fi O2 , and PaO2 /Fi O2 in the two groups throughout the 72 h after randomization (Figures 2 and 3). The PEEP level averaged 16 ± 1 cm H2O, and did not change or differ between the two groups during the 72 h. Both treatments significantly reduced Fi O2 (p = 0.0001) (Figure 2A) and increased PaO2 /Fi O2 (p = 0.0005) (Figure 3) over the 72 h. When analyzed over the entire 72 h, there was no significant treatment effect on the course of PaO2 , Fi O2 , or PaO2 /Fi O2 . Separate comparisons at each time point indicated that patients randomized to iNO had a higher PaO2 at 12 h and a lower Fi O2 at 12 and 24 h after randomization (Figure 2A and B). As compared with conventional therapy, iNO significantly increased PaO2 /Fi O2 at 12 h (p < 0.01) and tended to increase it at 24 h (p < 0.06) (Figure 3). Beyond 24 h, the two groups had equivalent values for PaO2 /Fi O2 (Figure 3).
Fig. 2. Course of Fi O2 (A) and PaO2 (B) throughout the 72 h after randomization. In both groups, Fi O2 decreased significantly over the 72 h (p < 0.0001). Over the entire 72 h, the two groups did not differ significantly. PaO2 remained constant over the 72 h. Data are shown for patients who completed 72 h of therapy (n = 16 in each group). Values are mean ± SEM. §p < 0.05 versus conventional therapy at time point.
[More] [Minimize]
Fig. 3. Course of PaO2 /Fi O2 throughout the 72 h after randomization. In both groups, PaO2 /Fi O2 increased significantly over the 72 h (p < 0.0005). Over the entire 72 h, the two groups did not differ significantly. Data are shown for patients who completed 72 h of therapy (n = 16 in each group). Values are mean ± SEM. +p < 0.01 versus conventional therapy at time point.
[More] [Minimize]The primary endpoint for analysis was the frequency with which patients improved sufficiently to be managed with a persistent decrease in Fi O2 ⩾ 0.15 during the 72 h following randomization. iNO therapy did not significantly increase the frequency of improved oxygenation (iNO: 11 of 20 [55%], versus conventional therapy: nine of 20 [45%], p = 0.55). If we exclude the six patients who died during the 72 h after randomization, iNO did not significantly augment the likelihood of persistent improvement in oxygenation (10 of 16 patients [63%] with iNO versus nine of 18 patients with conventional therapy [50%], p = 0.51).
Since our definition of persistent improvement in oxygenation was arbitrary, we tested whether using other criteria would alter the results, such as a persistent decrease in Fi O2 ⩾ 0.1, ⩾ 0.2, or ⩾ 0.25 that began within 72 h. A persistent decrease in Fi O2 ⩾ 0.1 occurred in 10 of 20 patients (50%) given conventional therapy and 12 of 20 (60%) given iNO (p = 0.54). Six of 20 patients (30%) given conventional therapy had a reduction in Fi O2 ⩾ 0.2, as compared with nine of 20 patients (45%) receiving iNO (p = 0.51). Five of 20 patients (25%) given conventional therapy achieved a reduction in Fi O2 ⩾ 0.25, compared with seven of 20 patients (35%) given iNO (p = 0.51).
If a patient improved to the point of being manageable with a persistent decrease in Fi O2 ⩾ 0.15 within 72 h after randomization, this development predicted ARDS reversal. Seventeen of the 20 patients who had this improvement had reversal of ARDS. In contrast, only six of 20 patients who did not improve within this time interval had ARDS reversal (p = 0.0006). This relationship held for patients randomized to either conventional or iNO therapy (p < 0.03 for each group). Regardless of therapy, survival also correlated with a persistent decrease in Fi O2 ⩾ 0.15 within 72 h after randomization. Fourteen of 20 patients who met this criterion survived. In contrast, only six patients survived among the 20 who did not improve over the 72 h (p < 0.02).
Using linear regression, with the decrease in Fi O2 over 72 h after randomization as the dependent variable, we looked for clinical features associated with improvement in Fi O2 . In patients randomized to conventional therapy, we were unable to identify predictive features. The best predictor in patients randomized to iNO was Cqs at randomization, using the compliance ranges in the LIS system (16). Among patients randomized to iNO, those with Cqs > 19 ml/cm H2O had a greater decrease in Fi O2 over 72 h than those with a Cqs ⩽ 19 ml/cm H2O (0.33 ± 0.05 [n = 6] versus 0.04 ± 0.06 [n = 7], p = 0.003). In patients randomized to iNO, those with ARDS for ⩽ 3 d before randomization had a larger decrease in Fi O2 (ARDS ⩽ 3 d: 0.28 ± 0.06, versus ARDS > 3 d: 0.06 ± 0.08, n = 8 in each group, p < 0.04).
Among patients randomized to conventional therapy, the ARDS duration or Cqs did not affect the 72 h response (data not shown). Since iNO was more effective in patients with Cqs > 19 ml/cm H2O or with ARDS ⩽ 3 d before randomization, we compared the therapies in these subgroups. In patients with Cqs > 19 ml/cm H2O, conventional therapy decreased Fi O2 over 72 h by 0.22 ± 0.09 (n = 8) versus 0.33 ± 0.06 (n = 6) with iNO (p = 0.33). In patients with ARDS for ⩽ 3 d, conventional therapy reduced Fi O2 during 72 h by 0.22 ± 0.08 (n = 10), compared with 0.28 ± 0.06 (n = 8) with iNO (p = 0.56).
A host of clinical features failed to influence the decrease in Fi O2 produced after 72 h of conventional or iNO therapy. These variables included age, gender, etiology, APACHE III score, LIS, or the level of Ppa, PVR, PEEP, Fi O2 , PaO2 , or PaO2 /Fi O2 at randomization. Some authors have speculated that septic patients have a diminished response to iNO (13, 22). In patients randomized to iNO in our study, sepsis did not influence the acute increase in PaO2 /Fi O2 after 1 h or the decrease in Fi O2 after 72 h (0.21 ± 0.06 with sepsis [n = 9] versus 0.13 ± 0.10 without sepsis [n = 7], p = 0.48).
The acute change in baseline PaO2 /Fi O2 produced after 1 h of iNO did not correlate with the change observed after 72 h of therapy (Figure 4). Criteria used to define a short-term oxygenation response to iNO include an increase in PaO2 /Fi O2 ⩾ 20% (11, 13, 20) or an absolute increase ⩾ 10 mm Hg (9, 10, 21). Based on these two criteria, an acute response to iNO at 1 h or lack of such a response did not predict whether a patient would have a persistent decrease in Fi O2 ⩾ 0.15 within 72 h or ARDS reversal (p < 0.67 for both outcomes). The decrease in Fi O2 after 72 h was 0.12 ± 0.08 (mean ± SEM) in acute responders to iNO versus 0.24 ± 0.06 in nonresponders (p = 0.31). Additionally, the increase in PaO2 /Fi O2 after 72 h of iNO was 26 ± 10 in acute responders versus 40 ± 10 in nonresponders (p = 0.36). The use of other published criteria for an acute oxygenation response, such as an increase in PaO2 /Fi O2 ⩾ 30% or an absolute increase ⩾ 40 or 50 mm Hg, did not improve the predictive value (8, 22-24).
Fig. 4. The change from baseline PaO2 /Fi O2 (P/F ratio) after 1 h of inhaled NO (iNO) does not correlate significantly with the change after 72 h of iNO. The relationship between the change in P/F ratio after 1 h and 72 h was y = 34 − 0.003 x; r2 = 0.000007; p = 0.99 (n = 16).
[More] [Minimize]In patients randomized to iNO, the increase in baseline PaO2 /Fi O2 after 1 h was similar in selected subgroups, such as patients with Cqs ⩽ 19 ml/cm H2O or > 19 ml/cm H2O, ARDS ⩽ 3 d or > 3 d before randomization, absence or presence of sepsis, or meeting or failing to meet our definition of improved oxygenation.
Since iNO may increase bleeding time (25) and inhibit platelet aggregation (26), we specifically evaluated whether iNO would affect bleeding during the 72 h after randomization. Among the 40 patients in our study one patient receiving iNO had a bleeding episode requiring blood transfusion. While receiving iNO, another patient developed an acute myocardial infarction due to occlusion of his left anterior descending artery, received intracoronary urokinase and intravenous heparin, and suffered an intracranial hemorrhage within the following 12 h.
Inhaling NO for short periods can improve arterial oxygenation in ARDS patients (1-13). The increase in PaO2 /Fi O2 noted after inhalation of NO for 10 to 160 min has created enthusiasm for its potential long-term use. In extrapolating from short-term results to long-term efficacy in ARDS, investigators usually assume that ongoing iNO therapy will produce an improvement in oxygenation as compared with conventional therapy. In the present study we tested this key assumption. Additionally, we sought to determine whether improved oxygenation would translate into a clinically relevant endpoint (i.e., the ability to safely decrease Fi O2 ).
Inhaling NO for 1 h significantly increased PaO2 /Fi O2 over the patient's baseline values. This improvement coincides with the expected response based on other reports. Our study extends previous work by following the effect of iNO over 72 h and comparing the response with patients randomized to conventional therapy. Patients treated with iNO had a significantly higher PaO2 /Fi O2 at 12 h (p < 0.01) and possibly at 24 h (p < 0.06) than did the control group (Figure 3). However, after this transient improvement, iNO produced no oxygenation benefit as compared with conventional therapy (Figure 3).
We also assessed the ability of the two therapies to improve oxygenation to the point at which a patient could be safely managed with a persistent decrease in Fi O2 ⩾ 0.15. Although the selection of this endpoint was arbitrary, its occurrence correlated with ARDS reversal (p = 0.0006) and survival (p < 0.02). The two treatments did not significantly differ in the frequency with which patients achieved this endpoint. If we used other definitions for a persistent decrease in Fi O2 , such as a decrease ⩾ 0.1, ⩾ 0.2, or ⩾ 0.25, iNO did not significantly increase the occurrence of these endpoints.
Our results indicate that iNO does not produce a sustained improvement in oxygenation as compared with conventional therapy. It is unknown whether our findings can be generalized. Several features of our study should be discussed.
Our entry criteria identified patients with severe ARDS who were not responding to standard therapy. All patients meeting these criteria were eligible for entry. We entered 40 of the 41 patients offered admission to the study. Our patients were heterogeneous in age, risk factors, and ARDS duration before randomization. Our patients had severe ARDS as indicated by their PaO2 /Fi O2 , Fi O2 , and LIS (Tables 3 and 4). Whether our results apply to patients with milder ARDS is unknown.
The major etiologies in our study—sepsis, aspiration of gastric contents, and trauma—are similar to those in other studies of iNO in ARDS (1, 3, 9-11, 27). Twenty-one of the 40 patients in our study had sepsis as the primary risk factor associated with ARDS. Investigators have speculated that patients with sepsis or septic shock may be less responsive to iNO (13, 22). Among patients randomized to iNO, in our study, the presence of sepsis or septic shock did not affect the acute increase in PaO2 /Fi O2 or the decrease in Fi O2 over 72 h. Thus, we found no evidence that septic patients were less responsive to iNO than nonseptic patients. However, we cannot rule out the possibility that there exist subgroups, based on the cause of ARDS, that may be more or less responsive to iNO.
ARDS duration before randomization in our study ranged from 0.5 to 25 d, with a mean of 5 ± 1 d (mean ± SEM) and a median of 3.5 d. The ARDS duration before randomization in our study generally corresponds with or is shorter than in other published iNO studies (3, 9, 11, 27-29).
Most patients randomized to iNO therapy had a concentration–response curve generated during the first 24 h, receiving 5, 10, 15, and 20 ppm for 6 h each. Other investigators have commonly used iNO concentrations within this range or higher in ARDS patients (1, 3, 9, 11-13, 30). The optimal iNO concentration in ARDS patients is unknown. This arises, in part, because the correct endpoint for dosing is uncertain: effect on pulmonary vascular resistance, oxygenation, oxidant damage, or other variables. We also used a delivery system that tends to produce small boluses of iNO, although the large volume between the site of NO administration and the patient, and the high respiratory rate (usually > 20/min) may minimize this effect (31-33). Overall, we used a standard approach to dosing and a commonly employed dosage range; whether other dosing approaches or methods of delivery would have produced different results is unknown.
Although the study was unblinded, several steps were taken to ensure that patients were managed in a similar fashion during the 72-h period after randomization. Patients were separately randomized in each ICU. The rationale for decreasing Fi O2 was the same in each unit. The objective was to use a Fi O2 that produced a PaO2 between 55 and 65 mm Hg or an SaO2 in the 88 to 92% range. During the 72 h period, the ventilator mode was not changed. Additionally, the definition for improved oxygenation required that a decrease in Fi O2 ⩾ 0.15 persist for at least 24 h. This requirement focused attention on persistent rather than transient improvement. Two pieces of data provide evidence that patients were managed in a similar fashion. First, the PEEP level was similar in both groups and remained unchanged during the 72 h. Second, the two groups had a similar PaO2 during the 72 h (Figure 2B).
Since 32 patients completed the entire 72-h study period, it is reasonable to ask what likely effect a larger sample size would have on our observations. We compared the effect of the treatments on PaO2 /Fi O2 at each 12-h interval. On the basis of our results, we estimated the sample sizes needed to demonstrate statistical significance at each 12-h time point. We found that at 36 h a sample of 66 to 80 patients would be needed to confirm a difference of 16 mm Hg in PaO2 /Fi O2 ; at 48 h a sample of 222 to 266 patients would be required for a difference of 8 mm Hg; at 60 h a sample of 2,510 to 3,012 patients would be needed to confirm a difference of 3 mm Hg; and at 72 h a sample of more than 264,000 would be needed to confirm a difference of 0.2 mm Hg. The doubtful clinical significance of these small differences in PaO2 /Fi O2 is a separate issue. Thus, a larger sample size would not change the fundamental observation that iNO, as compared with conventional therapy, transiently improves PaO2 /Fi O2 , but the comparative benefit is not sustained.
Our sample size of 40 subjects allowed us to detect a difference of 35 to 40% in the frequency of a persistent decrease in Fi O2 . Although iNO did not produce a statistically significant difference, it did, depending on the criteria used, increase the likelihood of achieving these endpoints by 10 to 15%. To detect differences of 10% or 15% in the rate of a persistent decrease in Fi O2 (assuming α = 0.05 and β = 0.80), one would need sample sizes of at least 782 or 342 patients, respectively.
ARDS duration and Cqs. In patients randomized to conventional therapy, we were unable to detect clinical features that influenced the decrease in Fi O2 over the 72 h after randomization. Factors thought to potentially influence the acute response to iNO, such as the level of pulmonary vascular resistance or pulmonary arterial pressure, did not correlate with the 72-h response to iNO. Two clinical characteristics, however, were associated with the ability of iNO to decrease Fi O2 over the 72 h period: ARDS ⩽ 3 d before treatment, and Cqs > 19 ml/cm H2O. When we compared conventional and iNO in these two subgroups, iNO did not significantly improve the clinical outcome.
Predictive ability of the acute response to iNO. Unfortunately, the acute increase in PaO2 /Fi O2 with iNO does not predict the 72-h response. The acute improvement in oxygenation with iNO was similar in patients who had very different longer-term oxygenation responses. Labeling patients as responders or nonresponders to iNO did not predict whether a patient would have persistently improved oxygenation. The lack of correlation between the acute and the long-term improvement suggests that the processes responsible for these effects are different. Alternatively, the processes may be similar, but sustaining the acute response may be controlled by factors associated with ARDS duration and the derangement in respiratory compliance.
Although we felt that improvement to the point at which a patient could be managed with a persistent decrease in Fi O2 ⩾ 0.15 within 72 h after randomization would be beneficial, we were surprised to find that it dramatically predicted ARDS reversal (p = 0.0006) and survival (p < 0.02). These results imply that a small improvement in respiratory status, as indicated by a persistent decrease in Fi O2 , is a very hopeful prognostic sign in ARDS patients. This suggestion will need to be confirmed in a larger group of ARDS patients.
During the 72 h after randomization, two patients receiving iNO had bleeding episodes. One patient had an intracranial hemorrhage while receiving iNO plus intravenous anticoagulants (heparin and urokinase). Since iNO can decrease platelet aggregation (26), our experience raises the concern that combining iNO with intravenous anticoagulants may increase the bleeding risk. Many patients treated with iNO received subcutaneous heparin; we identified no additional bleeding risk in these patients.
Our study demonstrates that conventional therapy is effective in improving oxygenation in patients with severe ARDS. We anticipated that only 10 to 15% of the patients randomized to conventional therapy in our study would improve to the point at which the Fi O2 could be reduced by ⩾ 0.15 within 72 h. Instead, 45% of patients randomized to conventional therapy achieved this improvement. Interestingly, little published information exists on the course of Fi O2 or PaO2 /Fi O2 over time in ARDS patients.
Our data indicate that inhaled NO acutely increases PaO2 / Fi O2 , with the response reaching a plateau after 12 to 24 h (Figure 3). At the time of final review of this manuscript, Troncy and colleagues reported on 30 patients with ARDS who were randomized to iNO or conventional therapy (34). The brief format of their research letter makes detailed comparison with our study difficult. However, they noted that iNO increased PaO2 /Fi O2 during the first day of therapy as compared with conventional therapy, but that this benefit disappeared by the second day. This finding is consistent with our observations. The reason why iNO no longer produces an additive advantage after approximately the first 24 h is unknown. Conceivably, the PaO2 /Fi O2 curves may converge because the same mechanisms account for improvement with both therapies; iNO may only bring them into play earlier. Alternatively, the two therapies may improve PaO2 /Fi O2 by distinct mechanisms. Stopping iNO in patients undergoing chronic therapy acutely increases pulmonary arterial pressure and decreases PaO2 (1, 3, 9). These changes raise the possibility that chronic iNO therapy may inhibit other vasodilator mechanisms or may activate counterregulatory mechanisms that limit the ability of iNO to further improve the matching of ventilation and perfusion. If this occurs, it could explain why the increase in PaO2 / Fi O2 with iNO reaches a plateau.
Whether the failure of iNO as compared with conventional therapy to persistently improve oxygenation translates into a lack of effect on ARDS reversal and survival is unknown. Our study focused on oxygenation, but iNO may produce other theoretical benefits through effects on lung leak (7, 35, 36), oxidant injury (35, 37, 38), right ventricular function (5, 30), adhesion molecules (39), or inflammatory mediators (39). It is unknown whether these experimental and short-term effects produce long-term clinical benefits. Alternatively, NO may produce harmful effects, such as increasing pulmonary vascular permeability (40-42), inactivation of surfactant (43, 44), damage to alveolar Type II cells (45), promotion of mutagenicity (46, 47), or enhancement of oxidant injury by generating peroxynitrite or nitrogen dioxide (48-50).
Experimentally, NO has the power to produce both beneficial and harmful effects depending on the setting in which it is generated (35, 37, 38, 40-42, 48, 50). Consequently, the use of such a reactive compound in acute lung injury is beset with uncertainty. Despite the theoretical harms or benefits that might arise from iNO therapy in ARDS, its potential to improve oxygenation serves as its primary rationale. We have shown in patients with severe ARDS that iNO does not produce a sustained improvement in oxygenation as compared with conventional therapy. We feel that the transient increase in PaO2 /Fi O2 produced by iNO does not justify its routine use in ARDS. Further studies are warranted to determine whether subgroups of patients might benefit from this therapy.
The authors wish to express their appreciation to the respiratory therapists, housestaff, and nurses for their support and for the professional and humane care they provided to the patients in the study. Without their assistance, this study could not have been performed.
Supported by grant IP50 HL50153 from the National Heart Lung and Blood Institute.
| 1. | Rossaint R., Falke K. J., López F., Slama K., Pison U., Zapol W. M.Inhaled nitric oxide for the adult respiratory distress syndrome. N. Engl. J. Med.3281993399405 |
| 2. | Gerlach H., Rossaint R., Pappert D., Falke K. J.Time-course and dose-response of nitric oxide inhalation for systemic oxygenation and pulmonary hypertension in patients with adult respiratory distress syndrome. Eur. J. Clin. Invest.231993499502 |
| 3. | Bigatello L. M., Hurford W. E., Kacmarek R. M., Roberts J. D., Zapol W. M.Prolonged inhalation of low concentrations of nitric oxide in patients with severe adult respiratory distress syndrome: effects on pulmonary hemodynamics and oxygenation. Anesthesiology801994761770 |
| 4. | Puybasset L., Rouby J.-J., Mourgeon E., Cluzel P., Souhil Z., Law-Koune J.-D., Stewart T., Devilliers C., Lu Q., Roche S., Kalfon P., Vicaut E., Viars P.Factors influencing cardiopulmonary effects of inhaled nitric oxide in acute respiratory failure. Am. J. Respir. Crit. Care Med.1521995318328 |
| 5. | Fierobe L., Brunet F., Dhainaut J.-F., Monchi M., Belghith M., Mira J.-P., Dall'ava-Santucci J., Dinh-Xuan A. T.Effect of inhaled nitric oxide on right ventricular function in adult respiratory distress syndrome. Am. J. Respir. Crit. Care Med.151199514141419 |
| 6. | Young J. D., Brampton W. J., Knighton J. D., Finfer S. R.Inhaled nitric oxide in acute respiratory failure in adults. Br. J. Anaesthesia731994499502 |
| 7. | Benzing A., Geiger K.Inhaled nitric oxide lowers pulmonary capillary pressure and changes longitudinal distribution of pulmonary vascular resistance in patients with acute lung injury. Acta Anaesthesiol. Scand.381994640645 |
| 8. | Puybasset L., Rouby J.-J., Mourgeon E., Stewart T. E., Cluzel P., Arthaud M., Poète P., Bodin L., Korinek A. M., Viars P.Inhaled nitric oxide in acute respiratory failure: dose-response curves. Intens. Care Med.201994319327 |
| 9. | Rossaint R., Gerlach H., Schmidt-Ruhnke H., Pappert D., Lewandowski K., Steudel W., Falke K.Efficacy of inhaled nitric oxide in patients with severe ARDS. Chest107199511071115 |
| 10. | Walmrath D., Schneider T., Schermuly R., Olschewski H., Grimminger F., Seeger W.Direct comparison of inhaled nitric oxide and aerosolized prostacyclin in acute respiratory distress syndrome. Am. J. Respir. Crit. Care Med.1531996991996 |
| 11. | McIntyre R. D., Moore F. A., Moore E. E., Piedalue F., Haenel J. S., Fullerton D. A.Inhaled nitric oxide variably improves oxygenation and pulmonary hypertension in patients with acute respiratory distress syndrome. J. Trauma391995418425 |
| 12. | Day R. W., Guarı́n M., Lynch J. M., Vernon D. D., Dean J. M.Inhaled nitric oxide in children with severe lung disease: results of acute and prolonged therapy with two concentrations. Crit. Care Med.241996215221 |
| 13. | Krafft P., Fridrich P., Fitzgerald R. D., Koc D., Steltzer H.Effectiveness of nitric oxide inhalation in septic ARDS. Chest1091996486493 |
| 14. | Knaus W. A., Wagner D. P., Draper E. A., Zimmerman J. E., Bergner M., Bastos P. G., Sirio C. A., Murphy D. J., Lotring T., Damiano A., Harrell F. E.The APACHE III prognostic system: risk prediction of hospital mortality for critically ill hospitalized adults. Chest100199116191636 |
| 15. | Bone R. D., Balk R. A., Cerra F. B., Dellinger R. P., Fein A. M., Knaus W. A., Schein R. M. H., Sibbald W. J.American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest101199216441655 |
| 16. | Murray J. F., Matthay M. A., Luce J. M., Flick M. R.An expanded definition of the adult respiratory distress syndrome. Am. Rev. Respir. Dis.1381988720723 |
| 17. | Dorinsky P. M., Gadek J. E.Multiple organ failure. Clin. Chest Med.111990581591 |
| 18. | Abraham E., Wunderink R., Silverman H., Perl T. M., Nasraway S., Levy H., Bone R., Wenzel R. P., Balk R., Allred R., Pennington J. E., Wherry J. C.Efficacy and safety of monoclonal antibody to human tumor necrosis factor α in patients with sepsis syndrome: a randomized, controlled, double-blind, multicenter clinical trial. J.A.M.A.2731995934941 |
| 19. | Brown B. W., Jr., and M. Hollander. 1997. Statistics: A Biomedical Introduction. John Wiley & Sons, New York. 119–121, 176–177. |
| 20. | Mira J. P., Monchi M., Brunet F., Fierobe L., Dhainaut J. F., Dinh-Xuan A. T.Lack of efficacy of inhaled nitric oxide in ARDS (letter). Intensive Care Med.201994532 |
| 21. | Wysocki M., Delclaux C., Roupie E., Langeron O., Liu N., Herman B., Lemaire F., Brochard L.Additive effect on gas exchange of inhaled nitric oxide and intravenous almitrine bismesylate in the adult respiratory distress syndrome. Intensive Care Med.201994254259 |
| 22. | Puybasset L., Stewart T., Rouby J.-J., Cluzel P., Mourgeon E., Belin M.-F., Arthaud M., Landault C., Viars P.Inhaled nitric oxide reverses the increase in pulmonary vascular resistance induced by permissive hypercapnia in patients with acute respiratory distress syndrome. Anesthesiology80199412541267 |
| 23. | Lu Q., Mourgeon E., Law-Koune J. D., Roche S., Vézinet C., Abdennour L., Vicaut E., Puybasset L., Diaby M., Coriat P., Rouby J.-J.Dose-response curves of inhaled nitric oxide with and without intravenous almitrine in nitric oxide-responding patients with acute respiratory distress syndrome. Anesthesiology831995929943 |
| 24. | Gerlach H., Pappert D., Lewandowski K., Rossaint R., Falke K. J.Long-term inhalation with evaluated low doses of nitric oxide for selective improvement of oxygenation in patients with adult respiratory distress syndrome. Intensive Care Med.191993443449 |
| 25. | Högman M., Frostell C., Arnberg H., Hedenstierna G.Bleeding time prolongation and NO inhalation. Lancet341199316641665 |
| 26. | Samama C. M., Diaby M., Fellahi J.-L., Mdhafar A., Eyraud D., Arock M., Guillosson J.-J., Coriat P., Rouby J.-J.Inhibition of platelet aggregation by inhaled nitric oxide in patients with acute respiratory distress syndrome. Anesthesiology8319955665 |
| 27. | Levy B., Bollaert P.-E., Bauer P., Nace L., Audibert G., Larcan A.Therapeutic optimization including inhaled nitric oxide in adult respiratory distress syndrome in a polyvalent intensive care unit. J. Trauma381995370374 |
| 28. | Benzing A., Loop T., Mols G., Geiger K.Effect of inhaled nitric oxide on venous admixture depends on cardiac output in patients with acute lung injury and acute respiratory distress syndrome. Acta Anaesthesiol. Scand.401996466474 |
| 29. | Zwissler B., Kemming G., Habler O., Kleen M., Merkel M., Haller M., Briegel J., Welte M., Peter K.Inhaled prostacyclin (PGI2) versus inhaled nitric oxide in adult respiratory distress syndrome. Am. J. Respir. Crit. Care Med.154199616711677 |
| 30. | Rossaint R., Slama K., Steudel W., Gerlach H., Pappert D., Veit S., Falke K.Effects of inhaled nitric oxide on right ventricular function in severe acute respiratory distress syndrome. Intensive Care Med.211995197203 |
| 31. | Westfelt U. N., Lundin S., Stenqvist O.Nitric oxide administration after the ventilator: evaluation of mixing conditions. Acta Anesthesiol. Scand.411997266273 |
| 32. | Young J. D., Dyar O. J.Delivery and monitoring of inhaled nitric oxide. Intensive Care Med.2219967786 |
| 33. | Imanaka H., Hess D., Kirmse M., Bigatello L. M., Kacmarek R. M., Steudel W., Hurford W. E.Inaccuracies of nitric oxide delivery systems during adult mechanical ventilation. Anesthesiology861997676688 |
| 34. | Troncy E., Collet J.-P., Shapiro S., Guimond J.-G., Blair L., Charbonneau M., Blaise G.Should we treat acute respiratory distress syndrome with inhaled nitric oxide? Lancet3501997111112 |
| 35. | Poss W. B., Timmons O. D., Farrukh I. S., Hoidal J. R., Michael J. R.Inhaled nitric oxide prevents the increase in pulmonary vascular permeability caused by hydrogen peroxide. J. Appl. Physiol.791995886891 |
| 36. | Benzing A., Bräutigam P., Geiger K., Loop T., Beyer U., Moser E.Inhaled nitric oxide reduces pulmonary transvascular albumin flux in patients with acute lung injury. Anesthesiology83199511531161 |
| 37. | Kavanagh B. P., Mouchawar A., Goldsmith J., Pearl R. G.Effects of inhaled NO and inhibition of endogenous NO synthesis in oxidant-induced acute lung injury. J. Appl. Physiol.76199413241329 |
| 38. | Chang, J., N. V. Rao, B. A. Markewitz, J. R. Hoidal, and J. R. Michael. 1996. Nitric oxide donor prevents hydrogen peroxide-mediated endothelial cell injury. Am. J. Physiol. 270(Lung Cell. Mol. Physiol. 14): L931–L940. |
| 39. | Chollet-Martin S., Gatecel C., Kermarrec N., Gougerot-Pocidalo M.-A., Payen D. M.Alveolar neutrophil functions and cytokine levels in patients with the adult respiratory distress syndrome during nitric oxide inhalation. Am. J. Respir. Crit. Care Med.1531996985990 |
| 40. | Berisha H. I., Pakbaz H., Absood A., Said S. I.Nitric oxide as a mediator of oxidant lung injury due to paraquat. Proc. Natl. Acad. Sci. U.S.A.91199474457449 |
| 41. | Mulligan M. B., Hevel J. M., Marletta M. A., Ward P. A.Tissue injury caused by deposition of immune complexes is L-arginine dependent. Proc. Natl. Acad. Sci. U.S.A.88199163386342 |
| 42. | Mulligan M. S., Warren J. S., Smith C. W., Anderson D. C., Yeh C. G., Rudolph A. R., Ward P. A.Lung injury after deposition of IgA immune complexes: requirements for CD18 and L-arginine. J. Immunol.148199230863092 |
| 43. | Haddad, I. Y., J. Crow, Y. Yoazu, J. S. Beckman, and S. Matalon. 1994. Concurrent generation of nitric oxide and superoxide damages surfactant protein A (SP-A). Am. J. Physiol. 267(Lung Cell Mol. Physiol. 11):L242–L249. |
| 44. | Haddad, I. Y., H. Ischiropoulos, B. A. Holm, J. S. Beckman, J. R. Baker, and S. Matalon. 1993. Mechanisms of peroxynitrite induced injury to pulmonary surfactant. Am. J. Physiol. 265(Lung Cell. Mol. Physiol. 9):L555–L564. |
| 45. | Hu, P., H. Ischiropoulos, J. S. Beckman, and S. Matalon. 1994. Peroxynitrite inhibition of oxygen consumption and ion transport in alveolar type II cells. Am. J. Physiol. 266(Lung Cell. Mol. Physiol. 10):L628–L634. |
| 46. | Arroyo P. L., Hatch-Pigott V., Mower H. F., Cooney R. V.Mutagenicity of nitric oxide and its inhibition by antioxidants. Mutation Res.2811992193202 |
| 47. | Isomura K.Induction of mutations and chromosome aberrations in lung cells following in vivo exposure of rats to NO. Mutation Res.1361984119125 |
| 48. | Gaston B., Drazen J. M., Loscalzo J., Stamler J. S.The biology of nitrogen oxides in the airways. Am. J. Respir. Crit. Care Med.491994538551 |
| 49. | Matalon S., Beckman J. S., Duffey M., Freeman B. A.Oxidant inhibition of epithelial active sodium transport. Free Radic. Biol. Med.61989557564 |
| 50. | Rubbo H., Radi R., Trujillo M., Telleri R., Kalyanaraman B., Barnes S., Kirk M., Freeman B. A.Nitric oxide regulation of superoxide and peroxynitrite-dependent lipid peroxidation. J. Biol. Chem.26919942606626075 |
