Dose-Response Curves of Inhaled Nitric Oxide with and without Intravenous Almitrine in Nitric Oxide-responding Patients with Acute Respiratory Distress Syndrome

Methods: Six critically ill patients (aged 44plus/minus 7 yr) were studied during early stage of their acute respiratory failure (Murray score: 2.6 plus/minus 0.1). All responded to 15 parts per million (ppm) of inhaled nitric oxide by an increase in Pa^O2of at least 40 mmHg at FI^O21. Hemodynamic and respiratory parameters were recorded continuously from pulmonary artery and systemic catheters. Inspiratory, expiratory, and mean intratracheal nitric oxide concentrations were monitored continuously using a fast response time chemiluminescence apparatus (NOX 4000, Seres, Aix-en-provence, France). On day 1, 6 inspiratory concentrations of nitric oxide were randomly administered: 0.15, 0.45, 1.5, 4.5, 15, and 45 ppm to determine the dose response of inhaled nitric oxide on Pa^O2, pulmonary shunt, mean pulmonary artery pressure, and pulmonary vascular resistance index. On day 2, a continuous intravenous infusion of almitrine at a dose of 16 micro gram *symbol* kg sup -1 *symbol* min sup -1 was administered and dose response to inhaled nitric oxide was repeated according to the same protocol as during day 1. A constant FI^O2of 0.85 was used throughout the study.

Results: Nitric oxide induced a dose-dependent increase in Pa^Osub 2 for inspiratory nitric oxide concentrations ranging between 0.15 and 1.5 ppm. Almitrine increased Pa^O2/FI^O2from 161 plus/minus 30 to 251 plus/minus 45 mmHg (P < 0.001) and pulmonary vascular resistance index from 455 plus/minus 185 to 527 plus/minus 176 dyn *symbol* s *symbol* cm sup -5 *symbol* m²(P < 0.05), and decreased pulmonary shunt (Q with dot^S/Q with dot^T) from 35 plus/minus 2 to 33 plus/minus 3%(P < 0.001). During almitrine combined with nitric oxide, a dose-dependent increase in Pa^O2was observed for inspiratory nitric oxide concentrations ranging between 0.15 and 1.5 ppm. Almitrine plus nitric oxide 1.5 ppm increased Pa^O2/FI^O2from 161 plus/minus 30 to 355 plus/minus 36 mmHg (P < 0.001), decreased (Q with dot^S/Q with dot^Tfrom 35 plus/minus 2 to 24 plus/minus 2%(P < 0.001), pulmonary vascular resistance index from 455 plus/minus 185 to 385 plus/minus 138 dyn *symbol* s *symbol* cm sup -5 *symbol* m²(P < 0.05), and mean pulmonary artery pressure from 31 plus/minus 4 to 28 plus/minus 4 mmHg (P < 0.001).

Conclusions: In 6 patients with early acute respiratory distress syndrome and highly responsive to inhaled nitric oxide, the administration of intravenous almitrine at a concentration of 16 micro gram *symbol* kg sup -1 *symbol* min sup -1 induced an additional increase in Pa^O2. Dose response of nitric oxide was not changed by the administration of almitrine and a plateau effect was observed at inspiratory nitric oxide concentrations of 1.5 ppm.

ACUTE respiratory distress syndrome (ARDS) is characterized by impaired gas exchange caused by maldistribution of ventilation perfusion ratios. Increased pulmonary artery pressure caused by vasoconstriction and mechanical obstruction of pulmonary vessels frequently is associated with impaired alveolar ventilation caused by pulmonary edema. In consolidated lung regions, hypoxic pulmonary vasoconstriction reduces lung perfusion to limit arterial hypoxemia. However, during the inflammatory process characterizing ARDS, many vasodilators are released by the activated pulmonary endothelium and might inhibit hypoxic pulmonary vasoconstriction and enhance arterial hypoxemia. During the past 20 years, many attempts have been made to reestablish normal ventilation perfusion relationships through pharmacologic manipulation of pulmonary circulation. When administered in experimental acute lung injury and ARDS, intravenous vasodilators invariably have resulted in worsening arterial oxygenation and systemic hypotension through a non-selective vasodilation of both systemic and pulmonary vessels. In the late 1980s, it was demonstrated that intravenous almitrine bismesylate, a selective vasoconstrictor of pulmonary vessels, could improve arterial oxygenation in some patients with ARDS. However, this beneficial effect, likely related to a predominant vasoconstriction of pulmonary vessels located in nonventilated lung areas, was inconsistent and associated with increased pulmonary arterial pressure. More recently, it was shown that the administration of low concentrations of inhaled nitric oxide to patients with ARDS resulted in a significant reduction in pulmonary artery pressure associated with a significant increase in arterial oxygenation. This beneficial effect, likely related to a selective vasodilation of pulmonary vessels located in ventilated lung areas and to a diversion of pulmonary blood flow away from nonventilated lung areas, has been shown to be enhanced in some patients by the concomitant intravenous administration of almitrine. However, Wysocki et al. found the effect inconsistent, particularly in patients who were unresponsive to nitric oxide alone with a significant increase in Pa sub O². Except for 2 patients whose mean intratracheal concentrations of nitric oxide were measured using a chemiluminescence apparatus, inspiratory nitric oxide concentrations administered to the patients were calculated and assumed to be in the range of 5–10 parts per million (ppm). .

The aim of this study performed selectively in critically ill patients with ARDS who responded to inhaled nitric oxide with a significant increase in arterial oxygenation was twofold: first, to assess whether the additive effect of inhaled nitric oxide on the almitrine-induced increase in arterial oxygenation was dose-dependent; second to assess whether inhaled nitric oxide could completely reverse almitrine-induced increase in pulmonary arterial pressure and pulmonary vascular resistance.

This is part of a comprehensive study about the combination of inhaled nitric oxide with intravenous almitrine in patients with severe ARDS. Three goals originally were defined: first, to assess the dose response of inhaled nitric oxide in combination with a fixed dose of almitrine (16 micro gram *symbol* kg sup -1 *symbol* min sup -1); second, to determine the proportion of patients previously known as responding to inhaled nitric oxide by an increase in Pa^O2who further improve their arterial oxygenation when combining intravenous almitrine (partial results of this second study have been published in abstract form ); third, to assess dose response of almitrine in combination with a fixed inspiratory concentration of inhaled nitric oxide (5 ppm). Methods and results of the Current study concern the first part of the comprehensive study.

During a 6-month period, 17 consecutive hypoxemic patients with ARDS diagnosed on or after admission to the Surgical Intensive Care Unit of La Pitie Hospital in Paris (Department of Anesthesiology) were prospectively screened early in the course of their respiratory disease after written informed consent was obtained from each patient's next of kin. The study was approved by the Comite Consultatif de Protection des Personnes dans la Recherche Biomedicale of La Pitie-Salpetriere Hospital and supported by l'Assistance Publique Hopitaux de Paris. Inclusion criteria were:(1) bilateral infiltrates on bedside chest radiographs;(2) Pa^O2less or equal to 200 mmHg using an FI^O2of 1.0 and zero end-expiratory pressure;(3) bilateral and extensive hyperdensities on a high-resolution spiral thoracic computed tomographic scan;(4) positive response to inhaled nitric oxide, defined as a decrease in mean pulmonary arterial pressure (MPAP) of at least 2 mmHg and an increase in Pa^Osub 2 (FI^O21) of at least 40 mmHg after nitric oxide inhalation at an inspiratory concentration of 15 ppm. These cutoffs for response to nitric oxide were set to select patients responding to nitric oxide by a decrease in MPAP and an increase in Pa^O2of a sufficient magnitude to allow the determination of dose-response curves. It is obvious that when the variation of the parameter studied (either Pa^Osub 2 or pulmonary artery pressure) is in the range of the error of measurement, it becomes difficult to accurately assess the dose response. Exclusion criteria were:(1) pulmonary edema of cardiac origin defined as a pulmonary capillary wedge pressure > 18 mmHg and a left ventricular ejection fraction < 50% as estimated by a bedside transesophageal echocardiogram;(2) circulatory shock defined as a systolic arterial pressure < 90 mmHg or dependence on exogenous catecholamines; and (3) cardiac dysrhythmias. These exclusion criteria were intended to exclude patients with cardiovascular instability in whom an accurate evaluation of the hemodynamic effects of inhaled nitric oxide and almitrine was difficult. Among the 17 patients initially screened for inclusion, 11 had to be excluded (no response to nitric oxide n = 5, circulatory shock n = 4, pulmonary edema of mixed origin n = 1, atrial fibrillation n = 1). Finally, 6 patients responding to inclusion and exclusion criteria could be included.

In each patient, the trachea was orally intubated with a Hi-Lo Jet number 8 Mallinckrodt tube (Mallinckrodt Inc, Argyle, NY) which incorporates 2 side ports, one ending at the distal tip of the endotracheal tube and a more proximal port ending 6 cm from the tip. These additional channels were used for continuous monitoring of tracheal pressure and tracheal concentrations of inhaled nitric oxide. After inclusion into the study, all patients were sedated and paralyzed with a continuous intravenous infusion of fentanyl 250 micro gram/hr, flunitrazepam 1 mg/hr and vecuronium 4 mg/hr and the lungs were ventilated using conventional mechanical ventilation (Cesar Ventilator, Taema, France). For each patient, tidal volume and respiratory rate were adjusted to maintain minute ventilation constant throughout the study. An inspiratory time of 30%, a positive end-expiratory pressure of 10 cmH²O and an FI^O2of 0.85 were maintained throughout the study period. To detect changes in FI^O2induced by inhalation of nitric oxide, FI^O2was monitored continuously using an Oxygen sub 2 analyzer (Seres 4000, Aix-en-Provence, France). All patients were monitored using a fiberoptic thermodilution pulmonary artery catheter (Oximetrix Opticath catheter, Abbot Critical Care System) and a radial or femoral arterial catheter.

To assess accurately the extension of pulmonary hyperdensities and, therefore, the severity of ARDS, each patient was taken to the Department of Radiology (Thoracic Division). Lung scanning was performed from the apex to the diaphragm using a Tomoscan SR 7000 (Philips, Eindhoven, The Netherlands). All images were observed and photographed at a window width of 1600 Hounsfield units (HU) and level of 700 HU. An intravenous injection of 80 ml contrast material was made in each patient to differentiate pleural fluid collection from consolidated lung parenchyma. Evaluation included thin section and spiral computed tomograms in all patients. The thin section computed tomographic examination consisted of a series of 1.5 mm thick sections with 20-mm intersection spacing selected by means of a thoracic scout view during a 25-s period of apnea, the paralyzed patient being disconnected from the ventilator (pulmonary volume equal to apneic functional residual capacity). For spiral computed tomography, contiguous axial sections 10-mm thick were reconstructed from the volumetric data obtained during a 15-s apnea. In each patient, a semi-quantitative assessment of parenchymal consolidation in zero end-expiratory pressure was performed according to a previously described technique. .

Systolic and diastolic arterial pressures (SAP and DAP, respectively), and systolic and diastolic pulmonary arterial pressures (SPAP and DPAP, respectively) were measured simultaneously using the arterial cannula and the fiberoptic pulmonary artery catheter connected to two calibrated pressure transducers (91 DPT-308 Mallinckrodt) positioned at the midaxillary line. Systemic and pulmonary arterial pressures, electrocardiogram, tracheal pressure measured through the distal port of the endotracheal tube, gas flow, and tidal volume (VT) measured using a heated and calibrated Hans Rudolph pneumotachograph, were recorded simultaneously and continuously on a Gould ES 1000 recorder (Gould Instruments, Cleveland, OH) at a paper speed of 1 mm/s throughout the entire study period.

In each phase (see experimental protocol), when a steady-state level was obtained--defined as a leveling of the pulmonary arterial pressure--SAP, DAP, SPAP, DPAP, pulmonary capillary wedge pressure, right atrial pressure, VT, tracheal pressure, and gas flow were recorded at a paper speed of 50 mm/s. Mean arterial pressure was calculated as 1/3 SAP + 2/3 DAP. Mean pulmonary artery pressure was measured by planimetry as the mean of four measurements performed at end-expiration. SAP, DAP, SPAP, DPAP, pulmonary capillary wedge pressure, and right atrial pressure were also measured at end-expiration. Cardiac output was measured using the thermodilution technique and a bedside computer allowing the recording of each thermodilution curve (Oximetrix 3 SO²/CO Computer). Four serial injections of 10 ml 5% dextrose solution at room temperature performed at random during the respiratory cycle were used to avoid errors related to the use of cold thermodilution injectate and to average the cyclic variations in cardiac output related to continuous positive pressure ventilation. Heart rate was measured from the recorded electrocardiogram. Systemic and pulmonary arterial blood samples were simultaneously withdrawn within 1 min after the measurements of cardiac output (after discarding an initial 10-ml heparin contaminated aliquot). Arterial pH, Pa^O2, PvO², and PaCO²were measured using an IL BGE blood gas analyzer (Instrumentation Laboratory, Paris, France). Hemoglobin concentration, methemoglobin concentration, arterial and mixed venous oxygen saturations (SaO²and SvO²) were measured using a calibrated OSM3 hemoximeter (Radiometer Copenhagen, Neuilly-Plaisance, France). Arterial and mixed venous blood samples that showed hemoglobin concentrations differing by more than 0.1 g/100 ml were considered diluted and the highest hemoglobin concentration was used to calculate oxygen contents. Standard formulas were used to calculate cardiac index, pulmonary vascular resistance index (PVRI), systemic vascular resistance index, right and left stroke work indices, true pulmonary shunt (Q with dot^S/Q with dot^T), arteriovenous oxygen difference, oxygen delivery, oxygen extraction ratio, and oxygen consumption.*

In four patients, respiratory volume-pressure curves were measured using a 1–1 syringe (Model Series 5540, Hans Rudolph, Inc., Kansas City) as follows: the endotracheal tube was disconnected from the ventilator to allow functional residual capacity to be reached; then, 100-ml increments of Oxygen²were given with a 1.5-s pause at the end of each injection. Simultaneous recording of tracheal pressure measured through the distal port of the endotracheal tube was performed. Airway pressures on inflation and deflation were recorded on the Gould ES 1000 recorder at a paper speed of 10 mm/s. Each P-V curve was constructed allowing determination of opening pressure, static respiratory compliance calculated as the slope of the curve between 500–1000 ml and quasi-static respiratory compliance, obtained by dividing the VT by the corresponding airway pressure. Opening pressure could be clearly identified in only one patient. A positive end-expiratory pressure of 10 cmH²O was systematically applied to all patients.

Nitric oxide was released from three different tanks of nitrogen that had nitric oxide concentrations of 25, 900, and 2,000 ppm measured using chemiluminescence (Air Liquide, Meudon, France). Nitric oxide was delivered within the inspiratory limb of the ventilator just after the Fisher-Paykel humidifier via a nitrogen flowmeter delivering flows in the range of 10–1500 ml/min (CFPO, Meudon, France). Depending on the concentration of inhaled nitric oxide delivered to the patient, the flow of nitric oxide coming from the tank represented 30–450 ml/min of nitrogen. With the aid of the calibrated and heated pneumotachograph (Model Series 3500B, Hans Rudolph, Inc., Kansas City) attached to the proximal end of the endotracheal tube, VT was reduced to compensate exactly for the added volume of nitrogen and nitric oxide coming from the tank. Thus, VT and minute ventilation delivered to the patients were kept constant for all concentrations of inhaled nitric oxide.

Inspiratory, expiratory, and mean concentrations of nitric oxide and NO²were measured continuously using a fast response time chemiluminescence apparatus (NOX 4000 Seres, Aix-en-provence, France). Intratracheal gas was sampled using continuous aspiration through the proximal side port of the endotracheal tube, i.e., 162 cm from the site of nitric oxide administration. The NOX 4000 is a chemiluminescence apparatus specifically designed for medical use. When using an aspiration flow rate of 150 ml/min, the response time is approximately 30 s and only mean concentrations of nitric oxide can be measured accurately. When an aspiration flow rate of 1000 ml/min is selected, the response time is approximately 200 ms and inspiratory and expiratory nitric oxide concentrations can be accurately measured (Figure 1). We tested the in vitro accuracy of the NOX 4000 for measuring periodic fluctuations of nitric oxide concentrations. The NOX 4000 was connected to 3 reference nitric oxide tanks having a nitric oxide concentration of 700 parts per billion, 8,000 parts per billion, and 26 ppm via a solenoid valve that was closed and opened during various periods to mimic mechanical ventilation. As shown in Figure 2, fluctuations of nitric oxide concentrations were adequately measured by the NOX 4000 with a precision of 5%.

During the study, inspiratory and expiratory nitric oxide concentrations were measured continuously and recorded by setting the aspiration flow rate of the NOX 4000 at 1000 ml/min. In addition, in steady-state conditions, mean intratracheal nitric oxide concentrations were measured by setting the aspiration flow rate of the NOX 4000 at 150 ml/min. When the aspiration flow rate was changed, tidal volume was modified accordingly to achieve a constant minute ventilation and a stable nitric oxide concentration. To increase the precision of measurements, two different operating ranges of measurement were used according to the concentrations of nitric oxide administered to the patient: an operating range of 0–5 ppm was selected for inspiratory concentrations of 0.15, 0.45, 1.5, and 4.5 ppm, and an operating range of 0–100 ppm was selected for concentrations of 15 and 45 ppm. When an operating range of 0–5 ppm was selected, calibration was performed using a tank of nitric oxide having a reference concentration of 0.945 ppm (CFPO, Air Liquide); when an operating range 0–100 ppm was selected, calibration was performed using a tank of nitric oxide having a reference concentration of 22.8 ppm (CFPO, Air Liquide). Nitrogen oxides were calibrated using the same reference tanks according to the manufacturer's instructions. The oxygen analyzer of the NOX 4000 was used for continuous monitoring of oxygen concentration to ensure that FI^O2was maintained constant during nitric oxide inhalation, whatever the concentration administered.

The study was performed during a 2-day period in each patient. On day 1, the protocol consisted of 5 consecutive phases, at the end of which hemodynamic and respiratory parameters were measured.

Phase 1: Positive End-expiratory Pressure without Nitric Oxide (Control 1). Baseline measurements were made after 1 hr of steady-state conventional mechanical ventilation using the following ventilatory settings: FI^O20.85, positive end-expiratory pressure = 10 cm H sub 2 O, inspiratory time = 30%, respiratory frequency = 18 plus/minus 1.6 bpm, VT = 734 plus/minus 51 ml.

Phase 2: Positive End-expiratory Pressure 10 cmH²O with Nitric Oxide at a Fixed Inspiratory Concentration of 15 ppm (nitric oxide 15 ppm). Using the same ventilatory settings as during phase 1, inhaled nitric oxide at an inspiratory tracheal concentration of 15 ppm--as determined by the NOX 4000--was added into the inspiratory limb of the ventilator. To maintain a constant minute ventilation, the VT delivered by the ventilator was slightly reduced. Simultaneously, the FI^O2of the gas delivered by the ventilator was slightly increased to maintain the FI^O2at 0.85. After 30 min at steady state, hemodynamic and respiratory measurements were performed.

Phase 3: Positive End-expiratory Pressure 10 cmH²O without Nitric Oxide (Control 2). Using the same ventilatory settings as in phase 1 (VT and FI^O2included), nitric oxide was stopped and after 30 min at steady state, hemodynamic and respiratory parameters were measured.

Phase 4: Positive End-expiratory Pressure 10 cmH²O with Nitric Oxide at Increasing Inspiratory Concentrations (Dose-response Curve). Using the same ventilatory settings as in phase 1, increasing inspiratory tracheal concentrations of nitric oxide were administered in a random order according to a logarithmic scale: 0.15, 0.45, 1.5, 4.5, 15, and 45 ppm. Because the concentration of 45 ppm was associated with a long-lasting increase in blood methemoglobin concentration, which interfered with the calculation of venous and arterial Oxygen²contents and pulmonary shunt, it was not included in the randomization and was always administered as the last concentration. For each inspiratory tracheal concentration of nitric oxide, expiratory and mean intratracheal concentrations of nitric oxide were measured and recorded. In addition, VT and FI^O2were adjusted to maintain a constant minute ventilation and an FI^O2of 0.85 as assessed using the pneumotachograph and the oxygen analyzer.

Phase 5: Positive End-expiratory Pressure 10 cmH²O without Nitric Oxide (Control 3). Using the same ventilatory settings as in phase 1 (VT and FI^O2included), nitric oxide 45 ppm was stopped and after 60 min at steady state, hemodynamic and respiratory parameters were measured.

On day 2, hemodynamic and respiratory effects of the association almitrine-nitric oxide were studied. The protocol was made of five consecutive phases, at the end of which hemodynamic and respiratory parameters were measured.

Phases 1 and 2 were exactly similar to phases 1 and 2 performed at day 1.

Phase 3: Positive End-expiratory Pressure 10 cmH²O with Almitrine Alone (Almitrine¹). Using the same ventilatory settings as in phase 1 (VT and FI^O2included), a continuous intravenous infusion of almitrine at a concentration of 16 micro gram *symbol* kg sup -1 *symbol* min sup -1 was administered using an electric pump. After a steady state of 1 hour had been obtained, hemodynamic and respiratory parameters were measured.

Phase 4: Positive End-expiratory Pressure 10 cmH²O with Almitrine and Nitric Oxide at Increasing Inspiratory Concentrations (Dose-Response Curve). Using the same ventilatory settings as in phase 3, increasing inspiratory concentrations of nitric oxide were added to the continuous infusion of almitrine exactly like during phase 4 performed at day 1. FI^O2and minute ventilation were adjusted in a similar way.

Phase 5: Positive End-expiratory Pressure 10 cmH²O with Almitrine Alone (Almitrine²). Using the same ventilatory settings as in phase 1 (VT and FI^O2included), nitric oxide 45 ppm was stopped and almitrine was maintained at a concentration of 16 micro gram *symbol* kg sup -1 *symbol* min sup -1. After 60 min at steady state, hemodynamic and respiratory parameters were measured.

Effects of nitric oxide in the presence or absence of almitrine on hemodynamic and respiratory parameters were analyzed by two-way analysis of variance for two within factors, i.e., factor "almitrine (absence or presence)" and factor "dose of nitric oxide." We decided a priori that when a significant dose-effect relationship was found for nitric oxide by the analysis of variance, the values measured at each level of nitric oxide in the absence or in the presence of almitrine were to be compared with those obtained at 15 ppm nitric oxide in the absence of almitrine. Contrast analysis was used for these comparisons. The significance level was fixed at 5% but owing to the nature of the analysis of variance we used the criterion of Huynh and Feld rather than the classical F value. Calculations were made using BMDP software (UCLA at Los Angeles). All values are expressed as mean plus/minus SEM.

Among the six men enrolled in the study, three were admitted to the Surgical Intensive Care Unit after multiple trauma and three were admitted with postoperative complications after major surgery. All were studied at the early phase of the respiratory disease (first 5 days). As shown in Table 1and Table 2, all patients had severe ARDS characterized by arterial hypoxemia, increased Q with dot^S/Q with dot^T, pulmonary artery hypertension, reduced respiratory compliance, and consolidation of lung parenchyma extended to at least 50% of total lung volume. In three patients, ARDS was caused by lung infection but none were in septic shock.

Inspiratory, expiratory, and mean intratracheal concentrations of nitric oxide were found remarkably stable in each patient and very reproducible from one patient to another. Table 3shows that inspiratory intratracheal nitric oxide concentrations were 1.5 times greater than mean intratracheal nitric oxide concentrations. Expiratory concentrations of nitric oxide progressively increased with mean nitric oxide concentrations. For an inspiratory nitric oxide concentration of 0.15 ppm, nitric oxide could be measured in expired gas only in three patients. For an inspiratory nitric oxide concentration of 0.45 ppm, nitric oxide could be measured in expired gas in 5 patients. From inspiratory nitric oxide concentrations of 1.5 ppm, nitric oxide could be measured in expired gas of all patients.

As shown in Table 4, nitric oxide 15 ppm (inspiratory concentration) induced a 20% reduction in MPAP (P < 0.001), a 30% decrease in PVRI (P < 0.05), a 15% decrease in RWSI (P < 0.05), a 15% reduction in Q with dot^S/Q with dot^T(P < 0.001), a 25% increase in SVO²(P < 0.001), and a 60% increase in Pa^O2/FI^O2(P < 0.001). For MPAP (P < 0.001), PVRI (P < 0.05), right stroke work index (P < 0.001), Q with dot^S/Q with dot^T(P < 0.001), SVO²(P < 0.001), and Pa^O2/FI^O2(P < 0.001) a significant dose-dependent effect was found in the range 0.15–4.5 ppm. All other hemodynamic and respiratory parameters remained unchanged. Inhaled nitric oxide-induced changes in hemodynamic and respiratory parameters were quite reproducible in each patient from day 1 to day 2.

As shown in T4-5, almitrine 16 micro gram *symbol* kg sup -1 *symbol* min sup -1 induced a 20% increase in MPAP (P < 0.001), a 30% increase in PVRI (P < 0.05) a 13% increase in right stroke work index, a 12% reduction in Q with dot^S/Q with dot^T(P < 0.001), a 60% increase in Pa^O2/FI^O2(P < 0.001), and a 15% increase in SVO²(P < 0.001). Almitrine-induced increase in arterial oxygenation was quantitatively similar to inhaled nitric oxide-induced improvement in arterial oxygenation. All other hemodynamic and respiratory parameters remained unchanged after almitrine administration.

As shown in F2-5, Figure 3, Figure 4, Figure 5, Figure 6, Figure 7, the combination of a constant infusion of almitrine at a concentration of 16 micro gram *symbol* kg sup -1 *symbol* min sup -1 with inspiratory increasing concentrations of inhaled nitric oxide in the range 0.15–45 ppm induced a dose-dependent increase in Pa^O2(P < 0.001) and a dose-dependent reduction in MPAP, PVRI, right stroke work index, and Q with dot^S/Q with dot^T(P < 0.01). A plateau effect was observed for inspiratory intratracheal nitric oxide concentrations between 1.5 and 4.5 ppm. When inhaled nitric oxide was combined with almitrine, the dose-response curve of Pa^O2/FI^O2was shifted upward (P < 0.05). As shown by the absence of interaction between the factors almitrine and nitric oxide, the profile of the dose-response curve was not significantly affected by the presence of almitrine. The difference between Pa^O2/FI^O2measured with almitrine + nitric oxide and with nitric oxide alone did not appear to be dose-dependent, suggesting that the effect of each drug was additive and not synergistic. In contrast, the combination of almitrine to nitric oxide significantly affected the dose-response curve of MPAP. The profile of nitric oxide-dose response on MPAP was shifted rightward when nitric oxide was combined with almitrine (P < 0.05, for the interaction between the factors "almitrine" and "nitric oxide"). In patients 2, 4, 5, and 6 the association almitrine-nitric oxide was continued as an integral part of the clinical care. In these 4 patients, in whom the lungs were ventilated using FI^O2ranging from 0.8 to 1 prior to almitrine-nitric oxide, FI^O2could be reduced to between 0.35–0.5. Almitrine-nitric oxide was administered during 4, 2, 3, and 6 days, respectively.

As shown in T3-5, MetHb and NO²significantly increased at inspiratory nitric oxide concentrations of 15 ppm. For an inspiratory nitric oxide concentration of 45 ppm, a mean intratracheal NO²concentration of 1.2 ppm was observed.

This study shows that a combination of almitrine and inhaled nitric oxide can markedly increase Pa^O2/FI^O2in hypoxemic patients with ARDS who respond to nitric oxide with an increased Pa^Osub 2 of > 40 mmHg. This beneficial effect is dose-dependent for inspiratory nitric oxide concentrations ranging between 0.15 and 1.5 ppm. The maximum effect of the combination on arterial oxygenation is better than what can be obtained with nitric oxide or almitrine alone. To facilitate the understanding of the results presented, the discussion will be divided into three parts: discussion of the cardiorespiratory effects of the combination almitrine-nitric oxide, discussion of clinical monitoring and measurement of inhaled nitric oxide, and discussion of the dose-response of nitric oxide with and without almitrine.

It is well known that nitric oxide and almitrine, despite their opposite effects on pulmonary vessels, can reduce intrapulmonary shunt and improve arterial oxygenation in patients with ARDS. Nitric oxide administered by inhalation, has a vasodilating action that is doubly selective: it selectively dilates preconstricted pulmonary vessels and only pulmonary arteries and veins perfusing ventilated lung areas. Therefore, in patients with ARDS, hypoxic pulmonary vasoconstriction, which is a critical mechanism for limiting perfusion of consolidated lung and avoiding a further increase in Q with dot^S/Q with dot^T, is preserved in nonventilated lung areas. Almitrine administered intravenously constricts pulmonary arteries by a mechanism which mimics and competes with hypoxic pulmonary vasoconstriction. When hypoxic pulmonary vasoconstriction is already present, almitrine does not further constrict pulmonary arteries. When hypoxic pulmonary vasoconstriction is absent, almitrine reestablishes a pulmonary arterial constriction similar to hypoxic pulmonary vasoconstriction. For reasons not yet elucidated, almitrine-induced constricting effect is limited to pulmonary arteries and not pulmonary veins and systemic vessels. In addition, the constrictor effect is dose-dependent: at concentrations less than 5 micro gram *symbol* kg sup -1 *symbol* min sup -1 almitrine potentiates hypoxic pulmonary vasoconstriction whereas at concentrations greater than 5 micro gram *symbol* kg sup -1 *symbol* min sup -1, it constricts pulmonary arteries independently of any hypoxic challenge and might even inhibit hypoxic pulmonary vasoconstriction. In our patients, and according to previous studies, large doses of almitrine were used and, consequently, almitrine induced a marked and significant increase in pulmonary arterial pressure and PVRI. In patients with high permeability type pulmonary edema, such an increase may increase the amount of fluid traversing the alveolo-capillary barrier and contribute to an additional deterioration in gas exchange. However, this deleterious effect is mainly dependent on the hydrostatic capillary pressure, which in turn, depends on pulmonary venous pressure. Because almitrine selectively constricts pulmonary arteries, it is likely that pulmonary venous pressure and hydrostatic capillary pressure do not increase, thereby limiting the deleterious effect of almitrine-induced pulmonary hypertension on lung parenchyma. However, almitrine-induced pulmonary hypertension represents an additional increase in right ventricular afterload which might precipitate acute right ventricular failure. Therefore, it appears reasonable to limit its administration to patients with a preserved right ventricular function and with an MPAP not exceeding 30 mmHg. The current study demonstrates that the addition of inhaled nitric oxide to almitrine totally reverses almitrine-induced pulmonary vasoconstriction if inspiratory concentrations of nitric oxide above 1.5 ppm are used. At these concentrations, the toxicity of inhaled nitric oxide appears minimal. .

The combination of nitric oxide with almitrine increased arterial oxygenation in the six patients enrolled in the present study. All of them were nitric oxide responders in terms of arterial oxygenation and pulmonary arterial pressure and they showed a significant additional increase in Pa^O2when nitric oxide and almitrine were combined. Only five patients responded to almitrine by a significant increase in arterial oxygenation whereas one patient did not show any change in Pa^O2. As demonstrated recently, the combination nitric oxide-almitrine increases in only 50–60% of patients with ARDS. The reasons for this variable response are not yet known. Wysocki et al. suggested that patients who do not respond to inhaled nitric oxide, also do not respond to the combination of almitrine-nitric oxide. This is the reason why nitric oxide responders only were enrolled in the current study. As a matter of fact, the 6 patients who fulfilled the inclusion criteria all responded to the combination almitrine-nitric oxide by an increase in Pa sub O²of at least 10% of the Pa^O2obtained with nitric oxide alone. However, the small number of patients included in the current study precludes any definitive conclusion on the predictive factors determining the response to the association almitrine-nitric oxide.

Mechanisms by which almitrine increases inhaled nitric oxide-induced improvement in arterial oxygenation are not precisely known and can only be speculated upon. Because high doses of almitrine were used, it is highly likely that almitrine-induced pulmonary vasoconstriction involved the entire pulmonary arterial bed. Because in five patients, almitrine-induced pulmonary artery hypertension was associated with a significant decrease in intrapulmonary shunt, it is highly likely that the constricting effect predominated in non-ventilated lung areas, and that pulmonary blood flow was diverted toward ventilated lung areas. When inhaled nitric oxide was added, the vasoconstricting effect of almitrine was reversed selectively in pulmonary arteries perfusing ventilated alveolar spaces, thus likely contributing to enhance the redistribution of pulmonary blood flow toward well ventilated lung regions. As a consequence, pulmonary shunt further decreased and Pa^Osub 2 significantly increased. For nitric oxide concentrations greater than 1.5 ppm, PVRI and right stroke work index were similar during nitric oxide alone and almitrine-nitric oxide, although pulmonary shunt and arterial oxygenation were significantly different. In other words, very low doses of nitric oxide were sufficient to reverse the almitrine-induced vasoconstricting effect on pulmonary arteries.

Continuous and accurate monitoring of nitric oxide tracheal concentrations is critical for a safe administration of inhaled nitric oxide in patients with ARDS. Bedside chemiluminescence is generally considered as the reference method. However, chemiluminescence devices that are now widely used in the clinical setting were all designed for industrial use. Because they were aimed at measuring stable concentrations of nitric oxide and NO²in dry atmosphere, most of them have a 30–60-s response time although, intrinsically, chemiluminescence has a response time less than 100 ms. In fact, most devices used at the bedside measure mean intratracheal concentrations of nitric oxide and do not provide inspiratory and expiratory concentrations. Nevertheless, inspiratory nitric oxide concentration, which is the "effective" dose delivered to pulmonary vessels via alveolar spaces, should be monitored. In the current study, a chemiluminescence apparatus specifically designed for medical use was used. Because it has a response time of 200 ms, inspiratory and expiratory intratracheal concentrations were measured with a precision of 5%. Based on recordings of inspiratory, expiratory, and mean intratracheal concentrations of nitric oxide obtained in the six patients of the current study, the following statements can be made concerning uptake and distribution of inhaled nitric oxide. In paralyzed patients with volume-cycled mechanical ventilation and fixed ventilatory settings, inspiratory, expiratory, and mean intratracheal concentrations are remarkably stable in a given patient and from one patient to another. As shown in T3-5, inspiratory tracheal concentrations are 1.5-fold greater than mean concentrations. Expiratory concentrations increase with inspiratory concentrations suggesting that, at least in patients with ARDS, the inspiratory amount of nitric oxide delivered to the tracheobronchial tree is far from being totally taken up by the lungs. However, the difference between inspiratory and expiratory tracheal concentrations increases with increasing inspiratory concentrations suggesting that the higher the nitric oxide concentration, the greater amount of nitric oxide taken up by the lung.

Another important point should be emphasized. A mean nitric oxide concentration measured within the inspiratory limb of the ventilator using a chemiluminescence device with a slow response time is representative of "true" inspiratory tracheal concentrations only when nitric oxide is uniformly distributed within the tubing. This is not true when the gas is administered into the inspiratory limb, before the Y piece either continuously or discontinuously. In the current study, nitric oxide was introduced into the initial part of the inspiratory limb, so it is highly likely that nitric oxide concentrations were not homogeneously distributed within inspiratory tubing. Therefore, nitric oxide concentrations were monitored within the tracheobronchial tree using a fast response chemiluminescence device allowing an accurate determination of inspiratory tracheal nitric oxide concentrations.

The current study shows that dose-response curves of inhaled nitric oxide were not different during nitric oxide alone and almitrine-nitric oxide except for MPAP. Although a large dose of almitrine induced an additional increase in PVRI in each patient, similar concentrations of inhaled nitric oxide were required to decrease PVRI to the value obtained with nitric oxide alone. It is also clear that the optimum therapeutic effects of inhaled nitric oxide on pulmonary vascular resistance and arterial oxygenation were obtained with very low doses of nitric oxide, less than 4.5 ppm. The dose-response curve necessary to reverse almitrine-induced increase in pulmonary artery pressure together with basal pulmonary hypertension was different from the dose-response curve necessary to reduce basal pulmonary hypertension. In some patients, almitrine may increase cardiac index, which, in turn generates an additional increase in pulmonary artery pressure. Although the almitrine-induced increase in cardiac index has been previously reported, its mechanism remain to be elucidated. Inhaled nitric oxide having no influence on the flow component of pulmonary arterial hypertension, the combination of nitric oxide with almitrine will leave these patients with an increased MPAP. Almitrine-induced increase in cardiac output was observed in four patients of the current study and explains why, as a mean, MPAP remained greater during the combination nitric oxide-almitrine than during nitric oxide alone. Confirming previous studies, the optimum inspiratory nitric oxide concentration in terms of pulmonary vascular effects was found less than 4.5 ppm in patients with ARDS in the absence of circulatory shock. However, it has been shown that larger concentrations must be administered to patients treated with extracorporeal membrane oxygenation or in septic shock. Higher concentrations are also necessary in experimental animals and dose-response curves have been shown to be in the range 5–150 ppm in different experimental models of acute pulmonary hypertension. Finally, dose-response curves are probably species-dependent and dependent on the pathophysiology of pulmonary hypertension. In addition, they could also be influenced by the mode of administration of inhaled nitric oxide. Continuous administration of nitric oxide within the inspiratory limb of the ventilator, which is the most common mode of administration in Europe, does not deliver a constant concentration of nitric oxide during inspiration. As shown in F1-5, there is no inspiratory plateau but a peak inspiratory concentration rendering the inspiratory waveform sloping. This is likely related to the absence of uniform mixing of nitric oxide in the tidal volume. Because inspiratory flow is intermittent whereas nitric oxide is continuously delivered within the inspiratory limb, each tidal volume "pushes" into the tracheobronchial tree a bolus of nitric oxide incompletely mixed by the inspiratory flow. This peak inspiratory concentration only reaches pulmonary vessels during a fraction of inspiration and can be measured only using a fast response chemiluminescence apparatus. In contrast, it can be assumed that when nitric oxide is mixed with nitrogen and oxygen before being added to the ventilator, the inspiratory concentration is more constant throughout inspiration, the ventilator serving as a mixing chamber. However, during this method of administration that is widely used in North America, measurements of nitric oxide concentrations within the inspiratory limb have never been performed using a fast-response chemiluminescence apparatus to verify that nitric oxide is uniformly mixed in the inspiratory gas. The same inspiratory concentration administered either throughout the entire inspiration or during only a fraction of inspiration could have different hemodynamic effects. It can therefore be assumed that if pulmonary vessels are exposed to the same inspiratory nitric oxide concentration during a longer time, then hemodynamic effects should be more pronounced and of a longer duration. As a consequence, the dose-response curve could be displaced to the left: less inspiratory nitric oxide concentration would be required to obtain the same plateau effect. In fact, whatever is the method of nitric oxide administration, inspiratory nitric oxide concentrations less than 5 ppm are required to obtain the therapeutic effect. The very low inhaled nitric oxide concentration necessary to improve gas exchange and reduce pulmonary hypertension in patients with ARDS treated by conventional mechanical ventilation and without septic shock should reduce the risk of acute toxicity.

Reversible peripheral neuropathy has been reported after prolonged administration of almitrine to patients with chronic obstructive pulmonary disease. Although no other serious side effects have been described, the lowest dose of almitrine should be administered to critically ill patients. The minimum dose at which almitrine improves arterial oxygenation is not known. However, clinical experience suggests that doses of almitrine as small as 2 micro gram *symbol* kg sup -1 *symbol* min sup -1 might be associated with a significant increase in Pa sub O²in patients with ARDS. Although all previous animal and human studies of almitrine alone and almitrine combined with nitric oxide were performed using doses of almitrine around 15 micro gram *symbol* kg sup -1 *symbol* min sup -1, it is highly likely that smaller dose of almitrine would also result in some improvement in arterial oxygenation. Further studies are required to determine the optimum dose of almitrine that should be administered to patients with ARDS.

In conclusion, in patients in the early stage of ARDS and highly responsive to inhaled nitric oxide, the administration of a continuous intravenous infusion of 16 micro gram *symbol* kg sup -1 *symbol* min sup -1 of almitrine bismesylate may result in an additional improvement in arterial oxygenation. The maximum reduction in pulmonary shunt is obtained at inspiratory nitric oxide concentrations between 1.5 and 4.5 ppm. At these concentrations, the almitrine-induced pulmonary artery vasoconstriction is entirely reversed. However, in some patients, pulmonary artery pressure remains slightly increased because of an almitrine-induced increase in cardiac output. In the most hypoxemic patients, the marked increase in Pa^O2after the administration of nitric oxide and almitrine allows reducing FI^O2to less than 0.6, a level considered to be the threshold for oxygen toxicity.

The authors thank Dr. L. Bodin, Dr. P. Kalfon, and Dr. P. Poete, for their participation to the study; the nurses of the Surgical Intensive Care Unit, for their cooperation; and V. Connan, for preparing the manuscript.

* Vender J: Maximum utilization of pulmonary artery catheter monitoring. ASA Annual Refresher Course Lectures, San Francisco, 1985, 143 pp 1–6).