Improved oxygenation and increased cardiac output result from the interaction of spontaneous breathing and mechanical ventilation (1–3). Interestingly, these beneficial effects of spontaneous breathing were observed only during airway pressure release ventilation (APRV) but not during pressure support ventilation (1,3). APRV is a unique approach to ventilatory support because ventilation is achieved by a periodic pressure release from a preset continuous positive airway pressure (CPAP) level to a lower airway pressure (Paw), rather than by increasing Paw intermittently higher than CPAP as performed during more conventional forms of mechanical ventilation (4,5). APRV can provide full ventilatory support for patients who are not capable of spontaneous breathing and is then identical to pressure-controlled ventilation. APRV can also be used for partial ventilatory support with the unique possibility of unrestricted spontaneous breathing throughout the whole ventilatory cycle. Finally, if equal inspiratory and expiratory airway pressures are chosen, APRV is identical to CPAP. APRV was introduced for the treatment of patients with severe acute lung injury (4) and has been studied in experimental lung injury (1,2,5) as well as in clinical trials (3,6–8). Two studies have shown that lung collapse may occur rapidly during each expiration, followed by recruitment during the succeeding inspiration (9,10). These observations call for an analysis of the effects on respiratory mechanics and gas exchange of the short pressure releases during APRV with and without the application of positive end-expiratory pressure (PEEP).
Material and Methods
After approval of the local animal ethics committee, 15 healthy pigs (mixed Swedish country breed, weight 33 ± 5 kg) were anesthetized and mechanically ventilated. In nine pigs (Study group) baseline values were obtained 30 min after instrumentation, followed by induction of lung injury with oleic acid injected via a central venous catheter. After a stabilization period of 120 min measurements were repeated, then three ventilatory settings were sequentially applied for 1 h each in random order. Thereafter, mechanical ventilation was studied once more to assess any change in lung function over time (Fig. 1). Hemodynamic and ventilatory measurements were performed at the end of each study period; thereafter the pigs were paralyzed for a brief period to measure compliance of the respiratory system and the amount of mechanical ventilation during partial ventilatory support. Controlled mechanical ventilation with PEEP 20 cm H2O and an inspiratory pressure of 50 cm H2O followed each study period for 5 min to have a consistent volume history before each study mode.
In a second group of 6 pigs the stability of our lung injury model over time was evaluated during mechanical ventilation after induction of lung injury (Time Control group). Handling of animals and data collection were performed as in the Study group. This resulted in a total study period of approximately 8 h.
Forty mg azaperonum (Stresnil®, Janssen, Belgium) were given IM as premedication before transport from the farm. General anesthesia was induced with atropine (0.04 mg · kg−1), tiletamin/zolazepam (Zoletil®, Reading, Carros, France) (6 mg · kg−1), and xylazin (Rompun®, Bayer, Leverkusen, Germany) (2.2 mg · kg−1) IM, followed by a constant infusion of 400 mg · h−1 clomethiazole (Heminevrin®, Astra, Södertälje, Sweden), 100 μg · h−1 fentanyl, 200 μg · h−1 ketamine, and 6 mg · h−1 xylazin. The animals were endotracheally intubated with a cuffed tube and connected to the ventilator.
Before baseline measurements, 1000 mL Ringer’s acetate (Pharmacia AB, Stockholm, Sweden) at body temperature was infused; thereafter, fluid replacement was aimed at a constant hemoglobin value and a stable systemic arterial blood pressure. This resulted in an average infusion rate of 20–30 mL · h · kg−1 after induction of lung injury.
Mechanical ventilation was initiated using the BIPAP mode of the Evita 4 ventilator (Drägerwerke, Lübeck, Germany) with an Fio2 = 0.5, a respiratory rate of 20 breaths · min−1, and an inspiratory:expiratory ratio (I:E) of 1:2. The expiratory airway pressure (Pawlow) was set to 0 cm H2O, and the inspiratory airway pressure (Pawhigh) was chosen so that tidal volumes (Vt) were in the range of 10 mL · kg−1. Pawhigh was then increased until all spontaneous breathing efforts stopped as assessed from the simultaneously displayed curves of gas flow and airway pressure. During the course of the study Pawhigh was further adjusted if either spontaneous breathing efforts appeared during mechanical ventilation periods or if minute ventilation e increased >10% because of a change of respiratory mechanics.
After lung injury was established, mechanical ventilation as described above was always chosen as a first and last ventilatory mode to assess the degree of lung injury induced by oleic acid injection, and to evaluate any change in lung function over time. In between, the following three ventilatory settings with different pressure release levels were compared using the BIPAP mode and an Fio2 of 0.5.
- 1) APRV with pressure release to ambient airway pressure (APRVZEEP).
- The time intervals for Pawhigh and Pawlow were set to 2 and 1 s, respectively. Thus, the rate of mechanically delivered breaths was 20 · min−1 with an I:E ratio of 2:1. Pawhigh matched the mean airway pressure (Pawmean) of mechanical ventilation whereas Pawlow was set to 0 cm H2O to obtain the greatest possible inspiratory-expiratory pressure difference and thus the greatest possible mechanical ventilatory support for a given Pawhigh and respiratory rate.
- 2) APRV with pressure release to 50% of the CPAP level (APRVPEEP).
- The time intervals of Pawhigh and Pawlow as well as the pressure level of Pawhigh were identical as during APRVZEEP. Pawlow was decreased to the value that was 50% of Pawhigh to provide less ventilatory support compared with APRVZEEP. If Pawhigh was an uneven number, Pawlow was increased to the next integer (e.g., if Pawhigh was 11 cmH2O, Pawlow was set to 6 cmH2O).
- 3) CPAP.
- The respiratory rate, I:E ratio and Pawhigh were identical as during APRVZEEP, but Pawlow was chosen identical to Pawhigh. Thus, the ventilator generated no pressure swings, so these ventilator settings were actually identical to CPAP.
Oleic acid (Apoteksbolaget, Göterborg, Sweden) 0.1 mL/kg suspended in 20 mL isotonic saline was slowly (over 20 min) injected via the central venous catheter. If the Sao2 decreased to less than 85% during the injection, no further oleic acid was given. During injection blood pressure was stabilized with titrated doses of adrenaline.
For measurements of Vt and Paw, the readings of the ventilator were used. The amount of spontaneous breathing during APRV was estimated as e − mechanical ventilation, with the amount of mechanical ventilation being determined during complete muscle paralysis. Compliance (Crs) of the total respiratory system was estimated as Vt/Pawhigh during complete paralysis, with Pawhigh and Pawlow set to match the values during mechanical ventilation. Because mechanical ventilation and APRV resulted in a nearly instantaneous increase and decrease of Paw, mean airway pressure (Pawmean) was estimated as (Pawhigh · tinsp + Pawendexpiration · texp) · (tinsp + texp)−1, where tinsp and texp denote the time intervals of Pawhigh and Pawlow, respectively.
For pressure measurement and arterial blood sampling, an 18-gauge catheter was inserted into the left carotid artery together with a thermistor-tipped fiberoptic catheter (Pulsiocath 4F FT PV 2024; Pulsion Medical System, Munich, Germany) that was advanced into the descending aorta for measurements of cardiac output (t) and intrathoracic blood volume (ITBV). A Swan-Ganz catheter and an 18-gauge catheter were introduced into the right external jugular vein. The position of catheters was confirmed by pressure tracing.
Systemic, pulmonary arterial and central venous pressures were displayed on a bedside monitor together with the oxyhemoglobin saturation (Sao2) (Series 7010, Tram; Marquette Electronics Inc.) and recorded with reference to atmospheric pressure at midthoracic level at end-expiration. t and ITBV were calculated automatically (Pulsion COLD Z-021; Pulsion Medical System) after injecting 8–10 mL of 1 mg/mL indocyanine green (ICG) (Pulsion Medical System) mixed in sterile water (temperature ca. 5°C–7°C). randomly within the respiratory cycle (11,12). The mean of triplicate measurements was calculated and used for statistical evaluation. Oxygen delivery (O2), oxygen consumption (O2), systemic vascular resistance, and pulmonary vascular resistance were calculated with standard equations.
Gas Exchange and Ventilation Perfusion Relationship
Arterial and mixed venous blood gas samples were analyzed with ABL 300 and OSM 3 Hemoximeter (Radiometer, Copenhagen, Denmark).
Determination of the ventilation-perfusion distribution (a/) was done with the multiple inert gas elimination technique (13) in the Study group during all measurements except for the second period of mechanical ventilation. In the Time Control group, the multiple inert gas elimination technique was performed in 5 animals (one animal was excluded because of technical problems) at baseline, 2 and 6 h after induction of lung injury. The standard deviations of the logarithmic distribution of perfusion (Logsd) and ventilation (Logsd) were calculated as measures of the dispersion (mismatch) of the blood flow and ventilation distribution, respectively.
All data are presented as mean ± sd if not stated otherwise. P < 0.05 was chosen as the level of significance. Not all data were normally distributed as tested with the Shapiro-Wilk’s-W-Test, therefore a Friedman ANOVA was used to analyze differences between the ventilatory settings, followed by Wilcoxon’s signed rank test, if significant differences were detected. Calculations were performed with the software package Statistica® on a personal computer.
At baseline, the Study group and Control group had similar characteristics except for an increased heart rate in the Control group. Oleic acid injection affected hemodynamics, oxygenation and respiratory mechanics significantly and caused a comparable degree of lung injury in both groups as shown in Tables 1 and 2.
Ventilatory Variables and Hemodynamics
Spontaneous breathing accounted for 8.1 ± 1.8 L/min during CPAP, 5.5 ± 1.9 L/min (66.6 ± 13.9% of ˙Ve) during APRVPEEP and 4.6 ± 2.0 L/min (56.3 ± 14.9% of ˙Ve) during APRVZEEP (P < 0.001 for differences between groups). ˙Ve (8 L/min) and respiratory rate of mechanical and spontaneous breaths combined (26 breaths/min), in contrast, were almost identical during CPAP and both APRV patterns. Thus, the decrease of Paw during APRVPEEP from 12 ± 4 cm H2O to 7 ± 2 cm H2O (Fig. 1) allowed the lungs to expire passively 135 ± 50 mL per breath whereas the further decrease of Pawlow from 7 + 2 cm H2O to 0 cm H2O during APRVZEEP increased Vt only by additional 36 mL to 171 ± 49 mL. Crs tended to be higher during CPAP compared with both APRV patterns (19 ± 4 versus 17 ± 4 mL/cm H2O;P = 0.17), but the application of PEEP had no influence on Crs. Hypercapnia (Paco2 60 mm Hg) developed during CPAP and both APRV patterns, whereas a Paco2 between 35–45 mm Hg was necessary to completely override spontaneous breathing efforts during mechanical ventilation.
Hemodynamic variables such as ˙Qt, systemic and pulmonary arterial pressure as well as ˙DO2 were similar during CPAP and both APRV modes.
In the Study and Control groups, oleic acid injection increased shunt and ˙Va/˙Q-mismatch, as indicated by Logsd˙Q (Tables 2 and 4), whereas perfusion of poorly ventilated lung areas was minor before and after induction of lung injury (Fig. 2).
Shunt and ˙Va/˙Q-mismatch were lower during CPAP-breathing than APRV (Table 3). Thus, Pao2 was significantly increased (P < 0.01) with CPAP compared with both APRV patterns and the application of PEEP = 7 cm H2O had only minor effects on gas exchange as compared with APRVZEEP (Fig. 3). However, no close correlation was found between the amount of spontaneous breathing and venous admixture, shunt, or Logsd˙Q. Alveolar ventilation decreased by approximately 40% when ventilation changed from mechanical ventilation to CPAP or APRV and dead space ventilation (% of ˙Ve) increased significantly. In contrast, no difference in dead space was noted, whether or not spontaneous breathing was supported by cyclic airway pressure release.
Stability of the Lung Injury Model
In the Study group, lung injury tended to improve over time. Compared with the measurements performed 2 h after induction of lung injury, pulmonary vascular resistance was significantly decreased (469 ± 154 versus 426 ± 69 dyn · s · cm−5) and mixed venous oxygen saturation was significantly increased (52.6 ± 9.3 versus 62.8 ± 7.4%) 4 h later. Furthermore, oxygenation was slightly better (P = 0.086) after 6 than 2 h (Pao2 122 ± 34 versus 99 ± 39 mm Hg, P = 0.086) and Pawhigh could be reduced by 2 cm H2O. In the Control group, in contrast, lung injury was more or less stable during the 6-h observation period (Table 2).
During APRV, mechanical ventilatory support achieved by decreasing Paw periodically below the CPAP level (which causes a decrease of Pawmean) was associated with increased shunt as compared with CPAP breathing, even though a short pressure release time of only one second was used. In addition, application of PEEP 7 cm H2O had only minor effects on gas exchange and did not change compliance compared to APRV without PEEP. No difference was found for dead space ventilation between CPAP and APRV. Therefore, the cyclic release of airway pressure should be confined to the minimum necessary to avoid respiratory muscle fatigue. However, compared with mechanical ventilation all modes with spontaneous breathing, APRV and CPAP, increased ˙Qt without any measurable effect on oxygen consumption.
Pawmean has an important influence on oxygenation and is usually closely related to mean alveolar pressure (14). It has even been suggested that oxygenation is simply a function of Pawmean, irrespective of the ventilatory mode used, including CPAP (15). Because of the concept of APRV (4), mechanical ventilatory support is achieved by periodically decreasing Paw below the CPAP level and thereby decreasing Pawmean. Thus, the increased shunt and ventilation perfusion mismatch, indicated by an increased Logsd˙Q, during APRV compared with CPAP may in part be explained by differences in Pawmean. Other groups, however, reported that oxygenation was similar (16) or even better (17) with CPAP than mechanical ventilation, when the end-expiratory airway pressure was equal, and when Pawmean, thus, should have been decreased during CPAP. In our study, Pawmean was equal or even less during the spontaneous breathing modes than during mechanical ventilation, but still oxygenation was slightly or significantly improved with spontaneous breathing. Thus, oxygenation is not solely dependent on Pawmean.
The end-expiratory lung volume is another important factor that influences gas exchange and end-expiratory lung volume is dependent on the PEEP level. Thus, the periodic pressure release lower than CPAP during APRV may have facilitated atelectasis formation. Lung collapse occurs rapidly within the first second of an expiration, even when higher PEEP levels were used (9,10). Thus, the rather small increase of Vt ( 36 mL), when Pawlow was decreased from 7 (APRVPEEP) to 0 cm H2O (APRVZEEP), indicates that parts of the lung were most likely flooded or collapsed at end-expiration during APRVPEEP, so that the additional decrease of Paw to ambient pressure during APRVZEEP resulted in little further exhalation. This may also explain why Crs was unaffected whether or not PEEP was applied during APRV. However, during mechanical ventilation and APRVZEEP no PEEP was used, but the alveolo-arterial oxygen difference was significantly less (P < 0.05) during APRVZEEP (163 ± 44 mm Hg) compared with mechanical ventilation (208 ± 37 mm Hg). Therefore, differences of the end-expiratory airway pressure do not fully explain differences of gas exchange observed with CPAP, APRV and mechanical ventilation in the present study.
In experimental lung injury (1,2) and acute respiratory distress syndrome patients (3,18), spontaneous breathing reduces shunt when combined with APRV. The mechanisms by which this is achieved are not fully elucidated. Diaphragmatic contractions reduce the size of atelectasis in dependent lung areas during general anesthesia (19), and in subjects lying supine, the dorsally located parts of the diaphragm move more during spontaneous breathing than during mechanical ventilation (20). Spontaneous breathing may therefore reopen atelectasis that is preferentially located in dependent areas and could thereby improve gas exchange.
Compared with mechanical ventilation, CPAP and APRV increased ˙Qt in the present study, but the magnitude of this increase was independent from the amount of spontaneous breathing. A better cardiac performance with CPAP than mechanical ventilation has previously been reported (16,17,21), and may be explained by improved venous return of blood to the heart during CPAP, because intrathoracic pressure decreases during spontaneous inspirations. However, conflicting results have also been published (15,22). This discrepancy may be explained by differences of the arterial carbon dioxide tension among the ventilatory modes compared. An increase of the Paco2 is associated with a decrease of the systemic vascular resistance and an increase of the ˙Qt (23,24), and may thus explain the higher ˙Qt during CPAP and APRV compared with mechanical ventilation in the present study. However, Putensen et al. (1–3) showed that ˙Qt increased in the presence of spontaneous breathing with APRV, even though Paco2 was unchanged or even less after animals or patients started to breath spontaneously. Thus, the higher ˙Qt during CPAP and both APRV patterns than during mechanical ventilation may be explained by both hypercapnia and spontaneous breathing per se.
Oxygen Cost of Spontaneous Breathing
Spontaneous breathing associated with CPAP and APRV did not increase ˙VO2 compared with mechanical ventilation. This is in accordance with numerous previous reports (CPAP (16,17,21,22), APRV (1–3,8). Apparently, the oxygen cost of spontaneous breathing is so small compared with the total body oxygen consumption that measurements of ˙VO2 have a poor signal-to-noise ratio to detect changes of the metabolic activity of respiratory muscles.
Stability of the Lung Injury Model
In pigs, oleic acid-induced lung injury is stable for several hours (25–27). The finding that lung injury improved somewhat over time in our Study group was therefore unexpected, and no such change was observed in the Control group. Thus, it seems reasonable to believe that the use of CPAP and/or APRV had a favorable influence on the degree of lung injury. Lefevre et al. (27) demonstrated that variability of Vt and respiratory rate improves oxygenation during mechanical ventilation. Such variability or other aspects of spontaneous breathing may explain this finding. However, because we did not evaluate the degree of lung injury before and after each APRV pattern and CPAP separately, we cannot draw definite conclusions in this regard.
In conclusion, in our acute lung injury model, oxygenation was better with CPAP breathing compared with partial ventilatory support with APRV. Because of the concept of APRV, airway pressure is decreased below the CPAP level when ventilatory support is initiated during APRV. This may promote lung collapse, even during short expiratory time intervals. In addition, Pawmean, which is one important determinant of oxygenation, decreases when mechanical ventilatory support increases during APRV. Thus, mechanical ventilatory support during APRV should be confined to the necessary minimum needed to compensate for the increased workload of breathing associated with acute respiratory failure, so that respiratory muscle fatigue is prevented. Nevertheless, compared with controlled-mechanical ventilation, both CPAP and APRV increased ˙Qt and tended to increase ˙DO2 at an equal or lower Pawmean.
We thank Ms. Eva-Maria Hedin, Ms. Anne Abrahamson, and Ms. Agneta Roneus for skillful technical help.
1. Putensen C, Rasanen J, Lopez FA, Downs JB. Effect of interfacing between spontaneous breathing and mechanical cycles on the ventilation-perfusion distribution in canine lung injury. Anesthesiology 1994; 81: 921–30.
2. Putensen C, Rasanen J, Lopez FA. Ventilation-perfusion distributions during mechanical ventilation with superimposed spontaneous breathing in canine lung injury. Am J Respir Crit Care Med 1994; 150: 101–8.
3. Putensen C, Mutz NJ, Putensen-Himmer G, Zinserling J. Spontaneous breathing during ventilatory support improves ventilation-perfusion distributions in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med 1999; 159: 1241–8.
4. Downs JB, Stock MC. Airway pressure release ventilation: a new concept in ventilatory support. Crit Care Clin 1987; 15: 459–61.
5. Stock MC, Downs JB, Frolicher DA. Airway pressure release ventilation. Crit Care Med 1987; 15: 462–6.
6. Cane RD, Peruzzi WT, Shapiro BA. Airway pressure release ventilation in severe acute respiratory failure. Chest 1991; 100: 460–3.
7. Rasanen J, Cane RD, Downs JB, et al. Airway pressure release ventilation during acute lung injury: a prospective multicenter trial. Crit Care Med 1991; 19: 1234–41.
8. Sydow M, Burchardi H, Ephraim E, et al. Long-term effects of two different ventilatory modes on oxygenation in acute lung injury: comparison of airway pressure release ventilation and volume-controlled inverse ratio ventilation. Am J Respir Crit Care Med 1994; 149: 1550–6.
9. Neumann P, Berglund JE, Mondejar EF, et al. Dynamics of lung collapse and recruitment during prolonged breathing in porcine lung injury. J Appl Physiol 1998; 85: 1533–43.
10. Neumann P, Berglund JE, Mondéjar EF, et al. Effect of different pressure levels on the dynamics of lung collapse and recruitment in oleic-acid-induced lung injury. Am J Respir Crit Care Med 1998; 158: 1636–43.
11. Lichtwarck-Aschoff M, Beale R, Pfeiffer UJ. Central venous pressure, pulmonary artery occlusion pressure, intrathoracic blood volume, and right ventricular end-diastolic volume as indicators of cardiac preload. J Crit Care 1996; 11: 180–8.
12. Lichtwarck-Aschoff M, Zeravik J, Pfeiffer UJ. Intrathoracic blood volume accurately reflects circulatory volume status in critically ill patients with mechanical ventilation. Intensive Care Med 1992; 18: 142–7.
13. Wagner PD, Saltzman HA, West JB. Measurement of continuous distributions of ventilation-perfusion ratios: theory. J Appl Physiol 1974; 36: 588–99.
14. Marini JJ, Ravenscraft SA. Mean airway pressure: Physiologic determinants and clinical importance - Part 2: clinical implications. Crit Care Med 1992; 20: 1604–16.
15. Gattinoni L, Marcolin R, Caspani ML, et al. Constant mean airway pressure with different patterns of positive pressure breathing during the adult respiratory distress syndrome. Bull Eur Physiopathol Respir 1985; 21: 275–9.
16. Simonneau G, Lemaire F, Harf A, et al. A comparative study of the cardiorespiratory effects of continuous positive airway pressure breathing and continuous positive pressure ventilation in acute respiratory failure. Intensive Care Med 1982; 8: 61–67.
17. Shah DM, Newell JC, Dutton RE, Powers SR. Continuous positive airway pressure versus positive end-expiratory pressure in respiratory distress syndrome. J Thorac Cardiovasc Surg 1977; 74: 557–62.
18. Hormann C, Baum M, Putensen C, et al. Effects of spontaneous breathing with BIPAP on pulmonary gas exchange in patients with ARDS. Acta Anaesthesiol Scand Suppl 1997; 111: 152–5.
19. Hedenstierna G, Tokics L, Lundquist H, et al. Phrenic nerve stimulation during halothane anesthesia: effects on atelectasis. Anesthesiology 1994; 80: 751–60.
20. Froese AB, Bryan AC. Effects of anesthesia and paralysis on diaphragmatic mechanics in man. Anesthesiology 1974; 41: 242–55.
21. Rasanen J, Downs JB, Stock MC. Cardiovascular effects of conventional positive pressure ventilation and airway pressure release ventilation. Chest 1988; 93: 911–5.
22. Scott A, Hill AEG, Chakrabarti MK, Carruthers B. A comparison of the cardiorespiratory effects of continuous positive airway pressure breathing and continuous positive pressure ventilation in dogs. Br J Anaesth 1978; 50: 331–8.
23. Buhre W, Weyland A, Grüne F, et al. Influence of arterial carbon dioxide tension on systemic vascular resistance in patients undergoing cardiopulmonary bypass. Acta Anaesthesiol Scand 1998; 42: 167–71.
24. Kazmaier S, Weyland A, Buhre W, et al. Effects of respiratory alkalosis and acidosis on myocardial blood flow and metabolism in patients with coronary artery disease. Anesthesiology 1998; 89: 831–7.
25. Sum-Ping ST, Symreng T, Jebson P, Kamal GD. Stable and reproducible porcine model of acute lung injury induced by oleic acid. Crit Care Med 1991; 19: 405–8.
26. Grotjohan HP, van der Heijde RM, Jansen JR, et al. A stable model of respiratory distress by small injections of oleic acid in pigs. Intensive Care Med 1996; 22: 336–44.
© 2001 International Anesthesia Research Society
27. Lefevre GR, Kowalski SE, Girling LG, et al. Improved arterial oxygenation after oleic acid lung injury in the pig using a computer-controlled mechanical ventilator. Am J Respir Crit Care Med 1996; 154: 1567–72.