Preserved Spontaneous Breathing Improves Cardiac Output during Partial Liquid Ventilation AXEL R. FRANZ, CHRISTINA MACK, JULIA REICHART, FRANK POHLANDT, and HELMUT D. HUMMLER Department of Pediatrics, Division of Neonatology and Pediatric Critical Care, University of Ulm, Ulm, Germany
The aim of this study was to examine whether preserved spontaneous breathing (SB) supported by proportional-assist ventilation (PAV) would improve cardiac output (CO) during partial liquid ventilation (PLV) in rabbits with and without lung disease if compared with time-cycled, volume-controlled ventilation (CV) combined with muscle paralysis (MP). PLV was initiated in 17 healthy rabbits and 17 surfactant-depleted rabbits using 12 to 15 ml/kg of perfluorodecaline. Both ventilatory modes, SBPAV and CVMP, were applied in random sequence using a crossover design. CO was measured by thermodilution. CO was significantly higher during SBPAV than during CVMP: 136 21 ml/kgmin (mean SD) versus 120 30 ml/kgmin (p 0.004) in healthy rabbits, and 147 19 ml/kgmin versus 111 13 ml/kgmin (p 0.0001) in surfactant-depleted rabbits, resulting in an improved oxygen delivery. This difference was mainly caused by a larger stroke volume during SBPAV, whereas there was little change in heart rate. In surfactant-depleted rabbits, SBPAV resulted in improved arterial blood pressure and arterial and mixed venous pH and in a higher PaO2 at the same level of PEEP and mean airway pressure. We conclude that during PLV, CO is higher during SBPAV than during CVMP, resulting in an improved oxygen delivery. In surfactantdepleted rabbits, improved CO, oxygen delivery, and arterial blood pressure resulted in higher pH, possibly reflecting improved tissue perfusion and oxygenation.
lator has been shown to improve gas exchange (8) at lower airway pressures (9) if compared with controlled mechanical ventilation without muscle paralysis. We have previously shown that PLV is feasible during spontaneous breathing (SB) supported by proportional-assist ventilation (PAV) (10), and that the benefits of PLV, i.e., improved oxygenation and pulmonary compliance, are preserved when PLV is applied during SBPAV in rabbits with meconium aspiration (11). Similarly, PLV was also successfully combined with synchronized intermittent mandatory ventilation in animals with healthy and injured lungs (12, 13). PAV, also known as elastic and resistive unloading, is a new mode of synchronized ventilation (14). It allows the patient to control not only the timing of each breath, but also inspiratory time, expiratory time, tidal volume, and minute ventilation. PAV may be a promising ventilatory mode to support spontaneous breathing during PLV as it allows to compensate for increased airway resistance caused by perfluorocarbon-filled airways, and, in synergism with PLV, for decreased lung compliance. The aim of this study was to examine the effect of SB PAV on hemodynamics in animals with and without lung disease undergoing PLV. We hypothesized that CO during PLV would be higher during SBPAV if compared with CVMP.
Keywords: Fluorocarbons; ventilation, mechanical; respiration, artificial; hemodynamics; cardiac output; paralysis
METHODS
Partial liquid ventilation (PLV), also known as perfluorocarbonassociated gas exchange, improves oxygenation and pulmonary compliance in several animal models of surfactant deficiency (1) and meconium aspiration syndrome (2), and also in infants, children, and adults with severe respiratory distress syndrome (3, 4). Most studies used PLV during time-cycled, pressure- or volume-controlled ventilation (CV) combined with muscle paralysis (MP). However, during mechanical gas ventilation, MP may result in a lower functional residual capacity (5), impaired oxygenation (5), decreased ventilation-perfusion match (6), and reduced pulmonary compliance (7), blood pressure, and cardiac output (CO) (6), if compared with preserved spontaneous breathing during mechanical ventilation. Furthermore, preserved spontaneous breathing in synchrony with the venti-
The experiments were approved by the animal care committee of Baden-Württemberg, Germany, and were performed according to current animal care guidelines. A detailed description of the methods is available in the online data supplement to this article.
Instrumentation and Experimental Procedures Thirty-eight New Zealand white rabbits weighing 3,587 246 g were anesthetized, intubated, and mechanically ventilated. A femoral arterial and a pulmonary arterial line were placed for blood sampling and for pressure and cardiac output monitoring. Twenty animals were surfactant-depleted by repeated saline lavage (15). Using a randomized crossover design (Figure 1), all animals underwent two experimental periods of 10 min duration each: one supported by SBPAV and one by CVMP. PLV was initiated by continuous intratracheal infusion of 12 to 15 ml/kg prewarmed perfluorodecaline using the ventilatory mode to which the animals had been first assigned. Additional perfluorodecaline was not administered throughout the study.
(Received in original form June 29, 2000 and in revised form March 15, 2001)
Ventilatory Technique
Presented in part at the Annual Meeting of the Society for Pediatric Research, Boston, MA, May 2000.
For the CVMP periods, time-cycled, volume-controlled mechanical ventilation was applied with the settings: tidal volume (VT), 8 ml/kg; PEEP, 6 cm H2O; FIO2, 1.0; respiratory rate (RR), 30/min. The RR was adjusted to achieve a PaCO2 between 40 and 55 mm Hg. Thereafter, animals were paralyzed intravenously using 0.1 mg/kg of vecuronium. PAV was applied by servo-controlling the airway pressure (Paw) according to the flow signal, where the pressure increase from PEEP (6 cm H2O) applied per unit of inspiratory flow (Kr) determines the degree of resistive unloading and the pressure increase from PEEP applied per unit of inspired volume (Ke) determines the degree of elastic unloading. The applied Paw was the sum of the resistive and elastic components at any point in time during a spontaneous breathing cycle: Paw (Kr Flow) (Ke Volume) PEEP, with the exception that unloading was confined to inspiration. The gain settings for Kr and Ke
Supported by Grant No. Hu 793/1-2 from the Deutsche Forschungsgemeinschaft (German Research Foundation). Requests for reprints should be addressed to Helmut D. Hummler, M.D., University Children’s Hospital, Division of Neonatology and Pediatric Critical Care, Prittwitzstr. 43, 89075 Ulm, Germany. Correspondence should be addressed to Axel Franz, M.D., University Children’s Hospital, Division of Neonatology and Pediatric Critical Care, Prittwitzstr. 43, 89075 Ulm, Germany. E-mail:
[email protected] This article has an online data supplement, which is accessible from this issue’s table of contents online at www.atsjournals.org Am J Respir Crit Care Med Vol 164. pp 36–42, 2001 Internet address: www.atsjournals.org
37
Franz, Mack, Reichart, et al.: Cardiac Output during Partial Liquid Ventilation
Figure 1. Study design for both studies. Panel A for healthy rabbits, and Panel B for surfactant-depleted rabbits. A/C assist control ventilation; VT tidal volume, SBPAV spontaneous breathing supported by proportional-assist ventilation; CVMP timecycled, volume-controlled ventilation with muscle paralysis.
RESULTS
were set to compensate for the airway resistance of the intubated rabbit and to keep the PaCO2 at 40 to 55 mm Hg. The FIO2 was 1.0. At the end of each experimental period, arterial and mixed venous blood gas measurements were taken, and CO was measured by thermodilution.
Seventeen animals without lung disease and 17 surfactantdepleted animals completed the protocol (Figure 1). One of the healthy animals was excluded before randomization because it did not resume spontaneous breathing after instrumentation despite a PaCO2 of 60 to 70 mm Hg and discontinuation of ketamine infusion for 2 h. Of the 20 animals intended for surfactant depletion, two died during or immediately after the lavage procedure prior to randomization. In a third animal, we were unable to place the Swan-Ganz catheter into the pulmonary artery. In the remaining 17 surfactant-depleted animals, dynamic lung compliance during assist-control ventilation decreased from 0.95 0.42 ml/cm H2O/kg before lavage to 0.53 0.35 ml/cm H2O/kg after lavage, and the PaO2 decreased from 469 68 mm Hg to 241 115 mm Hg (FIO2 1.0). The dose of ketamine administered was similar during SBPAV and CVMP (51.3 12.1 mg/kg h versus 50.0 13.3 mg/kg h in healthy animals, and 63.8 15.8 mg/kg h versus 61.8 15.6 mg/kg h in surfactant-depleted animals). The ketamine dose was identical in 31 of the 34 animals. CO, SV, and oxygen delivery were higher during SBPAV
Data Analysis and Statistical Evaluation The primary outcome variable was CO. Secondary outcome variables were oxygen delivery and consumption, heart rate, stroke volume (SV), shunt fraction, mean arterial (ABP) and pulmonary arterial blood pressure, central venous pressure, arterial and mixed venous blood gases, mean airway pressure ( Paw ), minute ventilation, VT, and RR. The number of animals to be studied was determined by a sample size calculation (See online data supplement). The data are presented as mean standard deviation (over both treatment periods). All variables were analyzed using two-way ANOVA for repeated measurements (SAS version 8.0, SAS Institute, Cary, NC). Mean differences between treatments with corresponding 95% confidence intervals and two-sided p values after adjustment for period effects are reported. Additionally, the influence of SBPAV compared with CV on CO during gas ventilation prior to initiation of PLV and the effect of MP on PaO2 during PLV were examined in two subgroups of animals using a two-sided paired t test. The significance level was 0.05. No correction was performed for multiple comparisons of secondary outcome variables.
TABLE 1. HEMODYNAMICAL PARAMETERS Healthy Animals (n 17 ) Raw Data* SBPAV Cardiac output, ml/kgmin Stroke volume, ml/kg Heart Rate, beats/min Shunt fraction Mean arterial blood pressure, mm Hg Mean pulmonary arterial pressure, mm Hg Central venous pressure, mm Hg Oxygen delivery, ml/kg min Oxygen consumption, ml/kg min Systemic vascular resistance, dyne s/cm5/kg Estimated PVR, dyne s/cm5/kg
136 21 0.65 0.12 213 18 0.21 0.08 52.4 13.2 16.6 2.1 7.3 1.5 20.6 4.0 5.6 1.1 26.8 8.4 10.0 2.1
CVMP
Surfactant-depleted Animals (n 17)
2-way RM-ANOVA Treatment Effect†
120 30 16 [5; 26] 0.59 0.14 0.06 [0.02; 0.11] 207 25 6 [0; 12] 0.13 0.05 0.07 [0.03; 0.11] 49.4 14.9 2.7 [0.5; 6.0] 15.5 1.7 1.2 [0.4; 1.9] 7.6 1.5 0.4 [0.6; 0.11] 18.6 4.9 1.8 [0.4; 3.3] 5.3 0.9 0.3 [0.2; 0.8] 28.2 8.9 1.4 [3.8; 1.0] 10.7 1.9 0.6 [1.6; 0.3]
Raw Data*
p Value‡
SBPAV
0.007 0.012 0.04 0.0009 0.10 0.004 0.008 0.018 0.25 0.23 0.19
147 19 0.66 0.12 226 21 0.24 0.07 57.4 8.3 16.3 1.5 6.6 1.5 22.7 3.5 5.1 0.9 27.7 4.2 9.0 1.5
CVMP
2-way RM-ANOVA Treatment Effect†
111 13 37 [31; 43] 0.52 0.09 0.14 [0.11; 0.17] 216 26 10 [3; 17] 0.28 0.06 0.04 [0.08; 0.01] 47.6 6.6 9.8 [7.0; 12.7] 16.1 1.2 0.2 [0.2; 0.7] 7.6 1.5 1.0 [1.3; 0.6] 16.7 2.2 6.1 [5.2; 7.0] 4.7 0.9 0.4 [0.3; 1.2] 29.1 5.7 1.5 [3.4; 0.5] 11.7 1.8 2.8 [3.4; 2.2]
p Value‡ 0.0001 0.0001 0.01 0.09 0.0001 0.32 0.0001 0.0001 0.23 0.13 0.0001
Definition of abbreviations: CVMP time-cycled, volume-controlled ventilation with muscle paralysis; PVR pulmonary vascular resistance (the PVR was estimated presuming that the left atrial pressure was zero mm Hg); SBPAV spontaneous breathing supported by proportional assist ventilation. * Raw data are presented as mean SD (over both periods). † The treatment effect is presented as the mean difference [(SBPAV) – (CVMP)] adjusted for period effects [95% confidence interval]. ‡ Two-sided p value by 2-way RM-ANOVA.
38
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
VOL 164
2001
was significantly higher during SBPAV. Mean central venous pressure (measured against ambient pressure) was lower during SBPAV in both healthy and surfactant-depleted rabbits. In healthy rabbits, mean pulmonary arterial pressure was higher during SBPAV, whereas in surfactant-depleted rabbits, mean pulmonary arterial pressure was similar during both ventilatory modes (Table 1). In surfactant-depleted rabbits, pulmonary vascular resistance was significantly lower during SBPAV. Systemic vascular resistance was similar during both modes of ventilatory support (Table 1). In surfactant-depleted animals, SBPAV resulted in an improved PaO2 and an improved pH (Table 2), whereas in healthy animals, the PaO2 was lower during SBPAV. Similarly, there was a trend towards a lower shunt fraction during SBPAV in surfactant-depleted rabbits, whereas the shunt fraction was higher during SBPAV in healthy rabbits (Table 1). The complete results of arterial and mixed venous blood gas analyses are shown in Table 2. SBPAV was associated with higher RR, lower VT, and higher minute ventilation (Table 3). The Paw was higher during SBPAV in healthy animals and was similar during both ventilatory modes in surfactant-depleted animals. A subgroup analysis in surfactant-depleted animals that had been randomized to receive CVMP first showed that MP resulted in a reduction of PaO2 during CV despite liquidfilled lungs and despite similar VT, RR, minute ventilation, and Paw (Table 4). In animals that had been randomized to receive SBPAV first (Figure 1), SBPAV also resulted in higher CO if compared with CV during mechanical gas ventilation, i.e., before initiation of PLV. (Details are shown in Table E1 in the online data supplement).
Figure 2. Relative changes in cardiac output (CO) after change of ventilatory mode. Relative change was calculated as [(2nd CO – 1st CO)/ CO during CVMP]. Upper panel: healthy rabbits. Lower panel: surfactant-depleted rabbits. SBPAV spontaneous breathing supported by proportional assist ventilation; CVMP time-cycled, volume-controlled ventilation with muscle paralysis. Closed bars represent individual animals; open bars represent the mean change for each group of animals and each sequence of ventilatory modes.
DISCUSSION
(Table 1 and Figure 2), and the difference between both ventilatory modes was larger in surfactant-depleted animals. The difference in CO was mainly caused by a larger SV during SBPAV, whereas heart rate was only slightly higher (Table 1). In healthy animals, there was a trend towards a higher ABP during SBPAV. In surfactant-depleted animals, ABP
During mechanical gas ventilation, preserved spontaneous respiratory efforts are associated with increased CO and increased arterial blood pressure in animals and humans (6, 16). The present study comparing SBPAV with CVMP shows that these benefits are preserved during PLV. Moreover, our
TABLE 2. BLOOD GAS ANALYSES Healthy Animals (n 17) Raw Data*
Arterial pH PaO2, mm Hg PaCO2, mm Hg BE, mmol/L§ SaO2, %¶ Mixed venous pH PvO2, mm Hg PvCO2, mm Hg SvO2, %¶
Surfactant-depleted Animals (n 17) 2-way RM-ANOVA
Raw Data*
2-way RM-ANOVA
SBPAV
CVMP
Treatment Effect†
p Value‡
SBPAV
CVMP
Treatment Effect†
p Value‡
7.29 0.04 277 112 53.6 9.0 1.4 2.9 99.3 0.9
7.30 0.05 384 52 51.4 6.3 1.3 2.8 99.9 0.0
0.01 [0.04; 0.01] 109 [163; 56] 2.3 [3.3; 7.9] 0.1 [0.9; 0.7] 0.6 [1.1; 0.1]
0.32 0.0006 0.40 0.77 0.019
7.32 0.04 260 98 46.0 4.9 2.5 3.6 99.4 0.6
7.30 0.04 160 51 46.6 4.1 3.1 3.1 98.4 1.2
0.01 [0.00; 0.03] 100 [46; 154] 0.5 [3.0; 1.9] 0.65 [0.4; 1.8] 1.0 [0.3; 1.6]
0.03 0.0013 0.66 0.23 0.007
7.25 0.05 53.7 8.1 64.3 8.7 74.0 10.9
7.25 0.05 53.8 9.5 62.5 7.4 73.6 11.8
0.00 [0.03; 0.02] 0.2 [3.2; 2.8] 1.7 [4.1; 7.6] 0.2 [2.9; 3.4]
0.84 0.88 0.54 0.89
7.28 0.04 57.8 8.4 53.7 4.0 79.1 7.3
7.26 0.04 49.5 4.9 54.2 4.7 71.2 7.2
0.02 [0.01; 0.03] 8.3 [3.2; 13.5] 0.4 [3.1; 2.3] 7.9 [3.7; 12.1]
0.0001 0.004 0.75 0.0013
Definition of abbreviations: CVMP time-cycled, volume-controlled ventilation with muscle paralysis; SBPAV spontaneous breathing supported by proportional assist ventilation. * Raw data are presented as mean SD (over both periods). † The treatment effect is presented as the mean difference [(SBPAV) – (CVMP)] adjusted for period effects [95% confidence interval]. ‡ Two-sided p value by 2-way RM-ANOVA. § Calculated value. ¶ Calculated value, assuming 100% human adult hemoglobin and a P 50 of 26.7 mm Hg.
39
Franz, Mack, Reichart, et al.: Cardiac Output during Partial Liquid Ventilation TABLE 3. PARAMETERS OF VENTILATION Healthy Animals (n 17) Raw Data*
Respiratory rate, breaths/min Tidal volume, ml/kg Minute ventilation, ml/kg min Mean airway pressure, cm H2O Mean esophageal pressure, cm H2O
Surfactant-depleted Animals (n 17)
2-way RM-ANOVA
Raw Data*
2-way RM-ANOVA
SBPAV
CVMP
Treatment Effect†
p Value
SBPAV
CVMP
Treatment Effect†
p Value
79 12 3.9 0.5 313 70 9.0 0.7 8.1 1.5
19 3 8.7 0.4 173 35 8.1 0.8 8.2 1.2
60 [53; 66] 4.8 [5.1; 4.6] 140 [105; 175] 0.8 [0.5; 1.2] 0.1 [0.6; 0.5]
0.0001 0.0001 0.0001 0.0001 0.82
104 17 4.1 0.6 426 58 10.3 1.0 7.6 1.8
48 11 6.3 0.5 309 86 10.4 1.1 8.8 1.7
56 [44; 68] 2.2 [2.5; 2.0] 115 [71; 160] 0.12 [0.6; 0.4] 1.3 [1.9; 0.6]
0.0001 0.0001 0.0001 0.61 0.0007
Definition of abbreviations: CVMP time-cycled, volume-controlled ventilation with muscle paralysis; SBPAV spontaneous breathing supported by proportional assist ventilation. * Raw data are presented as mean SD (over both periods). † The treatment effect is presented as the mean difference [(SBPAV) – (CVMP)] adjusted for period effects [95% confidence interval]. ‡ Two-sided p-value by 2-way RM-ANOVA.
data suggest that the higher CO during SBPAV results in improved oxygen delivery, whereas oxygen consumption remains unchanged. It appears to be unlikely that factors other than the ventilatory modalities contributed significantly to the observed difference in CO since possible confounding factors have been excluded carefully: First, the dose of ketamine administered as continuous intravenous infusion was identical during both SBPAV and CVMP. Second, vecuronium was chosen for pharmacologic muscle paralysis during CVMP because this drug has been reported to have no cardiovascular side effects (17). Third, PaCO2 was carefully matched between ventilatory modes because PaCO2 may influence autonomic nervous output with an acute increase of PaCO2 resulting in higher heart rate, CO, ABP, myocardial contractility, and plasma epinephrine (18, 19). In this study, improved CO during SBPAV was predominantly caused by increased SV. Increased SV can result from improved contractility, reduced afterload, or improved diastolic filling (as a result of venous return and/or ventricular compliance). Because contractility is probably unaffected by the two modes of ventilation, the improvement of CO during SBPAV is most likely caused by the effect of the pressure changes transmitted to the heart chambers and the intrathoracic and intra-abdominal vessels, which may improve or impair ventricular filling and/or afterload. Spontaneous inspiration is associated with decreased pleural/intrathoracic pressure and results in decreased right atrial
pressure (measured against ambient pressure), increased blood flow in the vena cava (i.e., venous return to the right atrium), and, consecutively, in increased right ventricular enddiastolic volume, increased right ventricular SV, and pulmonary arterial blood flow (20–22). Although vena cava flow decreases again during spontaneous expiration, mean vena cava flow is higher during spontaneous respiration than during breathholding (23). In contrast, increased intrathoracic pressure, as during Valsalva maneuvers or mechanical inspiration, with or without PEEP, is known to reduce venous return and to reduce right ventricular end diastolic volume, and this reduction in venous return increases with increasing intrathoracic pressure and increasing tidal volume (16, 21, 22). Our observation that preserved spontaneous respiration was not only associated with negative esophageal pressure deflections (Figure 3), but also with a lower central venous pressure (as measured against ambient pressure), and a higher CO, if compared with CVMP, is in agreement with these studies. Furthermore, the difference of CO between the two ventilatory modes was higher in surfactant-depleted animals, which showed more respiratory effort indicated by more negative esophageal pressure deflections than in animals with healthy lungs. On the other hand, an increased Paw is known to reduce venous return, and Paw was higher during SBPAV in the healthy rabbits and was the same during both ventilatory modes in surfactant-depleted rabbits, suggesting that Paw could not have been the main factor determining venous re-
TABLE 4. SUBGROUP ANALYSIS OF THE EFFECT OF PARALYSIS ON GAS EXCHANGE DURING VOLUME-CONTROLLED PARTIAL LIQUID VENTILATION* Surfactant-depleted Animals (n 8)
PaO2 PaCO2 Mean airway pressure Minute ventilation Respiratory rate Tidal volume Cardiac output
mmHg mmHg cmH2O ml/kg min breaths/min ml/kg ml/kg min
CV
CVMP
Treatment Effect†
p Value‡
241 67 47.5 7.4 10.5 0.62 277 66 42 9 6.2 0.3 124 27
160 49 46.8 2.8 10.5 1.0 313 106 48 14 6.3 0.4 110 11
81 [18; 143] 0.7 [5.2;6.7] 0.0 [0.6;0.6] 36 [103;31] 6 [16;3] 0.1 [0.3;0] 14 [3;32]
0.02 0.78 0.98 0.24 0.16 0.10 0.10
Definition of abbreviations: CV time-cycled; volume-controlled ventilation. * Subgroup analysis in animals that had been randomized to receive controlled ventilation (CV) with muscle paralysis (MP) first (see Figure 1). After filling the lungs with 12 ml/kg perfluorodecaline, all animals first received CV and then CVMP. Values are mean SD. † Treatment effect is presented as mean difference [(CV) (CVMP)] with corresponding 95% confidence interval. ‡ Two-sided p value by paired t test.
40
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
Figure 3. Characteristic traces of airflow, airway (Paw), esophageal (Pes), pulmonary arterial (PAP), arterial (ABP), and central venous (CVP) pressures during spontaneous breathing supported by proportional-assist ventilation (left panel), and time-cycled, volume-controlled ventilation with muscle paralysis (right panel) in a surfactantdepleted animal. Note the marked negative esophageal pressure deflections below baseline during SBPAV.
turn in our study animals. The sum forces on the right atrium affecting venous return (resulting from transmitted Paw and from contractions of the respiratory muscles) is probably more accurately described by the mean esophageal pressure. Mean esophageal pressure was indeed lower during SBPAV in the surfactant-depleted rabbits (despite similar Paw), and it was similar between the ventilatory modes in the healthy animals (despite a higher Paw during SBPAV) (Table 3). The validity of this observation, however, is limited by the uncertainty to what extent esophageal pressure reflects juxta-cardial pressure in this model. Apart from a possible improvement of venous return, SBPAV may also have beneficial effects on right ventricular afterload as suggested by our findings of a lower estimated pulmonary vascular resistance and a lower central venous, i.e., right atrial pressure during SBPAV. The higher pulmonary vascular resistance during CVMP, i.e., the ventilatory mode with higher VT, is in agreement with previous studies, which showed that increasing inflation of the lung above functional residual capacity by increasing VT or PEEP will increase pulmonary vascular resistance (24, 25). However, the validity of our observation is limited, because we were unable to measure left atrial pressure or pulmonary artery wedge pressure. Because of the size of the rabbits, the balloon of the Swan-Ganz catheter could not be positioned far enough into a pulmonary artery branch. Inflating the balloon caused circulatory failure and thereby prevented meaningful measurements. We estimated pulmonary vascular resistance presuming that left atrial pressures have been similar during both ventilatory modes (Table 1). However, since SB may result in increased left atrial pressures by impairing left ventricular filling and emptying (26), we rather may have underestimated the difference in pulmonary vascular resistance between SBPAV and CVMP. The effects of spontaneous and mechanical inspiration on the left ventricle are almost exactly opposite to the effects on the right ventricle and apparently are inconsistent with our findings: Whereas mechanical inspiration improves left ven-
VOL 164
2001
tricular SV (26), spontaneous inspiration results in decreased left ventricular SV (27, 28). Furthermore, spontaneous inspiration impairs left ventricular filling (22, 27, 28), whereas mechanical inspiration improves left ventricular filling (26). These beneficial effects of increased intrathoracic pressure on left ventricular performance result in an immediate increase in stroke volume and is reflected in a transient increase of ABP at the time of mechanical inspiration (Figure 3). Our finding of an increased CO during SBPAV suggests, that in our animals, the benefit from improved venous return to the right heart and a possible decrease in right ventricular afterload during SBPAV exceeds the potential benefit of CV and higher intrathoracic pressure on the left heart. This is in agreement with previous studies showing that, when cardiac function was normal, CO was primarily dependent on venous return and not on left ventricular afterload (29). Our findings probably reflect the hemodynamic effects of the two ventilatory modes on healthy hearts, and may therefore not apply to subjects in cardiac failure, because in a condition with left ventricular failure, CO may be more dependent on changes in left ventricular loading than on changes in venous return (29). On the basis of the mechanisms outlined above, we speculate that the increases of CO and SV during SBPAV probably do not reflect an effect specific for PAV but rather the effect of spontaneous respiratory activity. Consequently, this effect may also be seen during unsupported SB or during other modes of assisted ventilation. Subgroup analyses suggest that the beneficial effect of SBPAV on CO and SV is also present during mechanical gas ventilation (Table E1 in the ONLINE DATA SUPPLEMENT). Because the study was not designed to examine this effect during gas ventilation, the sequence of the ventilatory modes was not randomized, and the Paw and the PaCO2 were not well matched in animals without lung disease. During mechanical gas ventilation, preserved spontaneous breathing is not only associated with an improved CO, but also with increased ventilation-perfusion matching (6) and improved oxygenation (5). This may at least in part be caused by preserved diaphragmatic muscle tone that, in synergism with PEEP, maintains functional residual capacity, improving ventilation-perfusion matching and oxygenation. Furthermore, spontaneous breathing may improve ventilation-perfusion match and oxygenation by preferentially ventilating dependent, i.e., well-perfused regions of the lung, whereas controlled ventilation in paralyzed subjects is preferentially directed to non-dependent regions (30, 31). Because PLV itself improves oxygenation and ventilation-perfusion matching by redirecting blood flow to nondependent regions (32), and by recruiting collapsed alveoli, it may thereby reduce the beneficial effect of spontaneous breathing. However, this study has shown, that even during PLV, SBPAV results in improved oxygenation and a trend towards a reduced shunt fraction in surfactant-depleted animals. Furthermore, subgroup analyses of surfactant-depleted animals with liquid-filled lungs comparing CV with CVMP showed that MP during PLV results in impaired gas exchange (Table 4). We did not measure diaphragmatic muscle activity by electromyography, but air flow and airway and esophageal pressure tracings show that there was diaphragmatic muscle activity during CV that was eliminated by MP (Figure 4). Although limited by the fact that the sequence of CV and CVMP was not randomized, these data seem to support the importance of diaphragmatic muscle tone for oxygenation despite “liquid PEEP”. We speculate that diaphragmatic activity caused changes in local transpulmonary pressures with consecutive changes of FRC and ventilation distribution. These
Franz, Mack, Reichart, et al.: Cardiac Output during Partial Liquid Ventilation
Figure 4. Characteristic traces of airflow, esophageal (Pes), and airway (Paw) pressures during time-cycled, volume-controlled ventilation (CV, left panel), and controlled ventilation with muscle paralysis (CVMP, right panel) in a surfactant-depleted animal. Note the fluctuations of Pes and airflow during the expiratory phase of the mechanical respiratory cycle during CV, which were eliminated by MP, suggesting diaphragmatic muscle activity during CV.
changes in transpulmonary pressures between ventilatory modes may have been enhanced by liquid filling of the lungs and by supine positioning because of the inherent effects of liquid filling (33) and supine positioning (34) on the vertical gradients of local pleural and transpulmonary pressures. Additionally, it cannot be excluded that side effects of Vecuronium on the pulmonary vasculature caused either by changes in lung volume or by other unknown mechanisms could have influenced ventilation perfusion matching. In contrast, in healthy animals, functional residual capacity and thereby oxygenation appear to be sufficiently maintained by PEEP 6 cm H2O and infrequent breaths with VT of about 8 ml/kg, whereas the markedly lower VT adopted by the healthy animals during SBPAV was associated with impaired oxygenation and intrapulmonary shunting. The low VT may have resulted in increased ventilation perfusion heterogeneity and/or diffusion limitation as described previously in animals without lung disease (35). Because blood oxygen saturation was calculated and not measured by co-oximetry, all data presented in this study based on oxygen saturation, including oxygen delivery and shunt fraction, should be interpreted cautiously. However, since PCO2 and pH were well matched between treatment modalities, and changes in 2,3DPG concentrations are unlikely during the short experimental period, any bias caused by calculation of oxygen saturation must have occurred similarly in both ventilatory modalities. The differences found between both ventilatory modalities therefore most likely reflect true differences. The adjustment of ventilator settings is a problem during PAV since the relationships of elastic and resistive loads to VT and flow are, at least in part, nonlinear (36–38). Furthermore, measurements of lung mechanics may be difficult and highly variable in spontaneously breathing subjects (39), and inertial loads, which are not routinely considered in these measurements, may become relevant with partially liquid-filled lungs during PLV. Therefore, we adjusted the degree of unloading according to arterial blood gases and the breathing pattern of the animals (described in detail in the Online-Only Repository), which may have resulted in a nonuniform degree of unloading if calculated as a proportion of elastic and resistive loads. Patients supported by PAV maintain PaCO2 over a range of degrees of unloading by increasing or decreasing their own
41
respiratory effort (40). If CO depends on the degree of spontaneous respiratory effort, one may speculate that keeping the degree of unloading at the lower tolerable limit would maximize the beneficial effect of SBPAV on CO, but may result in decreased patient comfort. Another limitation of this study is that because of its design it did not allow to study definite end points such as mortality. Nevertheless, the improvement in CO, ABP, PaO2 and oxygen delivery demonstrated in surfactant-depleted animals may be important in the critically ill patient with acute lung injury: preservation of spontaneous respiratory activity may allow to maintain CO and ABP at lower inotropic support and volume requirements and to maintain oxygenation at lower FIO2 and oxygen toxicity. Although the small improvement of arterial and mixed venous pH at the same level of PaCO2 found during SBPAV in surfactant-depleted animals is not clinically relevant in this animal model, it may indicate improved tissue perfusion and oxygenation. However, whether our results can be extrapolated to a condition of oxygen supply limitation still requires further studies. In summary, we have shown that preserved spontaneous breathing supported by PAV during PLV results in improved CO and SV, and may also improve oxygen delivery in comparison with CVMP. This effect is more pronounced in surfactant-depleted rabbits. As during gas ventilation, preserved spontaneous breathing during PLV may also improve oxygenation and reduce shunt fraction in surfactant-depleted rabbits. Acknowledgment : The writers gratefully appreciate the support by N. Claure, University of Miami, FL, USA; B. Jilge and B. Kuhnt, Animal Research Center, University of Ulm, Germany; and M. Kron, Department of Biometry and Medical Documentation, University of Ulm, Germany.
References 1. Leach CL, Holm B, Morin FCr, Fuhrman BP, Papo MC, Steinhorn D, Hernan LJ. Partial liquid ventilation in premature lambs with respiratory distress syndrome: efficacy and compatibility with exogenous surfactant. J Pediatr 1995;126:412–420. 2. Foust R III, Tran NN, Cox C, Miller TFJ, Greenspan JS, Wolfson MR, Shaffer TH. Liquid assisted ventilation: an alternative ventilatory strategy for acute meconium aspiration injury. Pediatr Pulmonol 1996; 21:316–322. 3. Leach CL, Greenspan JS, Rubenstein SD, Shaffer TH, Wolfson MR, Jackson JC, DeLemos R, Fuhrman BP. Partial liquid ventilation with perflubron in premature infants with severe respiratory distress syndrome. The LiquiVent Study Group. N Engl J Med 1996;335:761–767. 4. Hirschl RB, Conrad S, Kaiser R, Zwischenberger JB, Bartlett RH, Booth F, Cardenas V. Partial liquid ventilation in adult patients with ARDS: a multicenter phase I-II trial. Adult PLV Study Group. Ann Surg 1998;228:692–700. 5. Miller J, Law AB, Parker RA, Sundell H, Silberberg AR, Cotton RB. Effects of morphine and pancuronium on lung volume and oxygenation in premature infants with hyaline membrane disease. J Pediatr 1994;125:97–103. 6. 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;159:1241–1248. 7. Bhutani VK, Abbasi S, Sivieri EM. Continuous skeletal muscle paralysis: effect on neonatal pulmonary mechanics. Pediatrics 1988;81:419-422. 8. Cleary JP, Bernstein G, Mannino FL, Heldt GP. Improved oxygenation during synchronized intermittent mandatory ventilation in neonates with respiratory distress syndrome: a randomized, crossover study. J Pediatr 1995;126:407–411. 9. Hummler H, Gerhardt T, Gonzalez A, Claure N, Everett R, Bancalari E. Influence of different methods of synchronized mechanical ventilation on ventilation, gas exchange, patient effort, and blood pressure fluctuations in premature neonates. Pediatr Pulmonol 1996;22:305–313. 10. Hummler HD, Schulze A, Pohlandt F, Thome U. Dynamics of breathing during partial liquid ventilation in spontaneously breathing rabbits supported by elastic and resistive unloading. Pediatr Res 2000;47:392–397. 11. Hummler HD, Thome U, Schulze A, Schnabel R, Pohlandt F, Franz
42
12.
13.
14. 15.
16.
17.
18.
19.
20. 21.
22.
23. 24.
25.
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE AR. Spontaneous breathing during partial liquid ventilation in animals with meconium aspiration. Pediatr Res 2001;49:572–580. Bendel-Stenzel EM, Mrozek JD, Bing DR, Meyers PA, Connett JE, Mammel MC. Dynamics of spontaneous breathing during patient-triggered partial liquid ventilation. Pediatr Pulmonol 1998;26:319–325. Bendel-Stenzel EM, Bing DR, Meyers PA, Connett JE, Mammel MC. Synchronized gas and partial liquid ventilation in lung-injured animals: improved gas exchange with decreased effort. Pediatr Pulmonol 1999;27:242–250. Younes M. Proportional assist ventilation, a new approach to ventilatory support. Theory. Am Rev Respir Dis 1992;145:114–120. Lachmann B, Robertson B, Vogel J. In vivo lung lavage as an experimental model of the respiratory distress syndrome. Acta Anaesthesiol Scand 1980;24:231–236. Downs JB, Douglas ME, Sanfelippo PM, Stanford W, Hodges MR. Ventilatory pattern, intrapleural pressure, and cardiac output. Anesth Anal 1977;56:88–96. Naguib M, Samarkandi AH, Bakhamees HS, Magboul MA, el-Bakry AK. Histamine-release haemodynamic changes produced by rocuronium, vecuronium, mivacurium, atracurium and tubocurarine. Br J Anaesth 1995;75:588–592. Carvalho CR, Barbas CS, Medeiros DM, Magaldi RB, Lorenzi Filho G, Kairalla RA, Deheinzelin D, Munhoz C, Kaufmann M, Ferreira M, Takagaki TY, Amato MB. Temporal hemodynamic effects of permissive hypercapnia associated with ideal PEEP in ARDS. Am J Respir Crit Care Med 1997;156:1458–1466. Rasmussen JP, Dauchot PJ, DePalma RG, Sorensen B, Regula G, Anton AH, Gravenstein JS. Cardiac function and hypercarbia. Arch Surg 1978;113:1196–1200. Brecher GA, Hubay CAN. Pulmonary blood flow and venous return during spontaneous respiration. Circ Res 1955;3:210–214. Hoffman JIE, Guz A, Charlier AA, Wilcken DEL. Stroke volume in conscious dogs; effect of respiration, posture, and vascular occlusion. J Appl Physiol 1965;20:865–877. Santamore WP, Heckman JL, Bove AA. Right and left ventricular pressure-volume response to respiratory maneuvers. J Appl Physiol 1984; 57:1520–1527. Wexler L, Bergel DH, Gabe IT, Makin GS, Mills CJ. Velocity of blood flow in normal human venae cavae. Circ Res 1968;23:349–359. Whittenberger JL, McGregor M, Berglund E, Borst HG. Influence of state of inflation of the lung on pulmonary vascular resistance. J Appl Physiol 1960;15:878–882. Cheifetz IM, Craig DM, Quick G, McGovern JJ, Cannon ML, Ungerleider RM, Smith PK, Meliones JN. Increasing tidal volumes and pulmonary overdistention adversely affect pulmonary vascular mechanics and cardiac output in a pediatric swine model. Crit Care Med 1998; 26:710–716.
VOL 164
2001
26. Robotham JL, Cherry D, Mitzner W, Rabson JL, Lixfeld W, Bromberger-Barnea B. A re-evaluation of the hemodynamic consequences of intermittent positive pressure ventilation. Crit Care Med 1983;11: 783–793. 27. Robotham JL, Mintzner W. A model of the effects of respiration on left ventricular performance. J Appl Physiol 1979;46:411–418. 28. Peters J, Fraser C, Stuart RS, Baumgartner W, Robotham JL. Negative intrathoracic pressure decreases independently left ventricular filling and emptying. Am J Physiol 1989;257:H120–131. 29. Pouleur H, Covell JW, Ross JJ. Effects of nitroprusside on venous return and central blood volume in the absence and presence of acute heart failure. Circulation 1980;61:328–337. 30. Froese AB, Bryan AC. Effects of anesthesia and paralysis on diaphragmatic mechanics in man. Anesthesiology 1974;41:242–255. 31. Puybasset L, Cluzel P, Chao N, Slutsky AS, Coriat P, Rouby JJ. A computed tomography scan assessment of regional lung volume in acute lung injury. The CT Scan ARDS Study Group. Am J Respir Crit Care Med 1998;158:1644–1655. 32. Doctor A, Ibla JC, Grenier BM, Zurakowski D, Ferretti ML, Thompson JE, Lillehei CW, Arnold JH. Pulmonary blood flow distribution during partial liquid ventilation. J Appl Physiol 1998;84:1540–1550. 33. Goldner M, Shapiro R, Amato M, Nahum A, Adams A, Barbas C, Marini JJ. Effect of incremental perfluorocarbon instillation on lung volumes (FRC) and regional pleural pressure (Ppl) [abstract]. Am J Respir Crit Care Med 1997;155:A87. 34. Yang QH, Kaplowitz MR, Lai-Fook SJ. Regional variations in lung expansion in rabbits: prone versus supine positions. J Appl Physiol 1989; 67:1371–1376. 35. Hernan LJ, Fuhrman BP, Kaiser RE, Penfil S, Foley C, Papo MC, Leach CL. Perfluorocarbon-associated gas exchange in normal and acidinjured large sheep. Crit Care Med 1996;24:475–481. 36. Kano S, Lanteri CJ, Duncan AW, Sly PD. Influence of nonlinearities on estimates of respiratory mechanics using multilinear regression analysis. J Appl Physiol 1994;77:1185–1197. 37. Guttmann J, Eberhard L, Fabry B, Bertschmann W, Wolff G. Continuous calculation of intratracheal pressure in tracheally intubated patients. Anesthesiology 1993;79:503–513. 38. Guttmann J, Kessler V, Mols G, Hentschel R, Haberthur C, Geiger K. Continuous calculation of intratracheal pressure in the presence of pediatric endotracheal tubes. Crit Care Med 2000;28:1018–1026. 39. Gonzalez A, Tortorolo L, Gerhardt T, Rojas M, Everett R, Bancalari E. Intrasubject variability of repeated pulmonary function measurements in preterm ventilated infants. Pediatr Pulmonol 1996;21:35–41. 40. Marantz S, Patrick W, Webster K, Roberts D, Oppenheimer L, Younes M. Response of ventilator-dependent patients to different levels of proportional assist. J Appl Physiol 1996;80:397–403.
