Effects of Ventilation on Hemodynamics and Myocardial Blood Flow during Active Compression-Decompression Resuscitation in Pigs

This study was approved by our institutional animal investigation committee. The animals were managed in accordance with the guidelines of the American Physiological Society. Fourteen pigs (24–25 kg) received 4 mg/kg intramuscular azaperone 1 h before surgical preparation. Anesthesia was induced with 10 mg/kg intravenous metomidate. The pigs were stabilized in the dorsal recumbent position and an endotracheal tube was inserted. Ventilation was performed with a servo ventilator with 65% nitrous oxide in oxygen at 20 breaths/min and with a tidal volume adjusted to maintain Pa^CO2at 35 mmHg. Anesthesia was maintained using a continuous infusion of metomidate (0.5 mg *symbol* kg sup -1 *symbol* h sup -1) and two doses of 0.024 mg/kg buprenorphine at the beginning of surgical preparation and before induction of cardiac arrest, respectively. Ringer's solution (6 ml *symbol* kg sup -1 *symbol* h sup -1) was administered continuously. A lead II ECG was monitored. Fluid-filled catheters were advanced via femoral cutdowns into the abdominal aorta and right atrium. A pulmonary artery catheter was inserted via external jugular vein cut-down. A 7-French pigtail catheter was placed into the left ventricle for injection of radionuclide microspheres and iced saline. For the measurement of body temperature, a thermistor was placed into the aorta. Before cardiac arrest, 300 units/kg sodium heparin was administered intravenously to prevent intracardiac clot formation. All animals were autopsied to check correct positioning of the catheters, and to look for damage to the rib cage and internal organs.

Aortic and right atrial pressures were measured via fluid-filled catheters using pressure transducers that were calibrated to atmospheric pressure at the level of the right atrium. Pressure tracings were continuously recorded using two monitors and a data acquisition/control unit. Measurements were performed at intervals of 30 s before cardiac arrest and at intervals of 1 s during CPR. An electronic substraction circuit was used to record arteriovenous pressure differences (aortic minus right atrial pressure difference), which reflect coronary perfusion pressure. Cardiac output was measured by thermodilution technique (5 ml iced saline injected into the left ventricle) using a cardiac output computer. Blood gases were measured with a blood gas analyzer and corrected for body temperature. Expired minute volume was sampled between min 1 and 3 and between min 7 and 9, respectively, after beginning of CPR through a nonrebreathing valve into a gas-tight reservoir and measured afterward with a spirometer.

Before and 90 s and 5 min after epinephrine administration during CPR, myocardial blood flow was measured with radiolabeled microspheres according to Heymann et al. [15]as previously described. [2]At the end of the experiment, the entire heart was removed. The left ventricular free wall including the endocardial and epicardial tissue layers was weighed, homogenized, and radioactivity from tissue and blood was measured with a gamma scintillation spectrometer.

Before induction of cardiocirculatory arrest, hemodynamic parameters, and blood gases were measured. A 50-Hz, 60-V alternating current was applied via two subcutaneous needle electrodes to induce ventricular fibrillation. Cardiocirculatory arrest was defined as that point at which the aortic pulse pressure decreased to zero and the electrocardiogram showed ventricular fibrillation. Thirty seconds after induction of cardiocirculatory arrest, ACD was begun with a pneumatically driven automatic piston device (ACD Controller, Ambu International, Glostrup, Denmark). The chest compression rate was 80/min, and the duration of compression was 50% of the total cycle time. Unlike in humans in which a handheld compression/suction device provided with a silicone rubber suction cup (Ambu CardioPump, Glostrup, Denmark) is used for ACD, in the current study the compression pad was wired to the mid-sternum to ensure constant contact with the animal's chest wall and to prevent displacement of the pad during ACD. A velocity of compression of 7.48 inches/s was held constant. Based on experiments using the same animal model, the chest compression force was adjusted to produce 25% of sternal displacement of the pig's anteroposterior diameter. Active decompression was set to produce a sternal displacement of 10% greater than the resting anteroposterior diameter, for which a suction force of approximately 200 N was necessary. The decompression velocity was 7.48 inches/s.

During ACD with IPPV, mechanical ventilation with 100% of oxygen was performed at 20 breaths/min independently of chest compression at a tidal volume shown to result in a Pa^CO2of 35 mmHg before induction of cardiac arrest. In animals without IPPV, 100% oxygen was supplied by an oxygen reservoir bag connected to the endotracheal tube.

After 5 min of CPR, all animals received epinephrine 0.2 mg/kg in a 10-ml volume of normal saline into the right atrium over a period of 5 s. Hemodynamic measurements and acquisition of aortic and mixed venous blood samples were performed before and 90 s and 5 min after epinephrine administration. After acquiring the last blood sample during CPR, direct current shocks were applied. Initially, up to three countershocks (3 J/kg) were administered. In cases of persistent ventricular fibrillation, epinephrine was administered at the same dose as the previous one, and after a further 90 s of CPR, up to three more countershocks (5 J/kg) were administered. The same protocol (without defibrillation) was used if asystole or pulseless electrical activity occurred. Successful resuscitation was defined as the presence of coordinated electrical activity together with a systolic blood pressure above 80 mmHg for a duration of at least 5 min, during which no further resuscitative measures were applied.

(^[35])Values are expressed as median as well as 25th and 75th percentiles. For within-group comparisons, the Wilcoxon signed rank test was used. For comparisons between groups, the Mann-Whitney U test was used. Correlations between single parameters were analyzed by regression analysis. Statistical significance was considered at P < 0.05.

There were no differences in baseline hemodynamics, myocardial blood flow, blood gases, or minute ventilation between animals in the ACD-IPPV or ACD groups before induction of ventricular fibrillation (Table 1, Table 2, Table 3). Whereas in ACD-IPPV-treated animals, spontaneous circulation could be restored within 20 s after removal of the last blood sample, in animals subjected to ACD alone, restoration of spontaneous circulation could not be achieved. Autopsy revealed no damage to the thoracic cage or the internal organs in either group.

Mean arterial pressure (median as well as 25th and 75th percentiles) was greater in the ACD-IPPV group compared with the ACD group before (44, 40–51 and 33, 26–37 mmHg; P < 0.05) as well as 90 s (73, 66–80 and 50, 48–57 mmHg; P < 0.01) and 5 min after epinephrine administration (53, 40–63 and 40, 33–43 mmHg)(P < 0.05). Similarly, mean central venous pressure was greater in ACD-IPPV-treated animals in comparison with ACD-treated animals before (22, 20–31 and 14, 10–18 mmHg; P < 0.05) and 90 s (27, 25–34 and 18, 15–21 mmHg; P < 0.01) and 5 min after administration of epinephrine (25, 21–33 and 14, 11–17 mmHg; P < 0.01). In both groups, mean arterial pressure and mean central venous pressure increased 90 s after epinephrine administration as compared with point in time before epinephrine administration (P < 0.05). Mean coronary perfusion pressure was not significantly different between both groups before (24, 7–31 and 19, 16–21 mmHg) as well as 90 s (49, 28–53 and 33, 29–36 mmHg) and 5 min after epinephrine administration (28, 7–42 and 23, 20–30 mmHg). In both groups, coronary perfusion pressure increased 90 s after epinephrine administration as compared with mechanical measures alone (P < 0.05). Cardiac index was not significantly different between both groups before (27, 15–33 and 26, 20–28 ml *symbol* min sup -1 *symbol* kg sup -1) as well as 90 s (19, 17–22 and 17, 14–24 ml *symbol* min sup -1 *symbol* kg sup -1) and 5 min (20, 14–24 and 20, 17–23 ml *symbol* min sup -1 *symbol* kg sup -1) after epinephrine administration. Ninety seconds after epinephrine administration, cardiac index was decreased as compared with preepinephrine levels. The decrease in cardiac index was significant only in the ACD group (Table 1).

Myocardial blood flow was not different between both groups before (30, 19–43 and 25, 18–34 ml *symbol* min sup -1 *symbol* 100 g sup -1) and 90 s after epinephrine administration (55, 39–68 and 39, 20–60 ml *symbol* min sup -1 *symbol* 100 g sup -1). However, 5 min after epinephrine administration, myocardial blood flow was greater in animals subjected to ACD-IPPV (33, 25–39 ml *symbol* min sup -1 *symbol* 100 g sup -1) as compared with animals subjected to ACD alone (15, 10–27 ml *symbol* min sup -1 *symbol* 100 g sup -1; P < 0.05). Only in ACD-IPPV-treated animals did myocardial blood flow increase after epinephrine administration as compared to mechanical measures alone (P < 0.05)(Table 2). In the ACD-IPPV group there was a significant correlation between coronary perfusion pressure and myocardial blood flow (r = 0.68, P < 0.01), but no such correlation was found in animals treated with ACD alone (r = 0.36, ns). Five minutes after epinephrine administration, seven animals in the ACD-IPPV group and two animals in the ACD group had myocardial blood flow values > 20 ml *symbol* min sup -1 *symbol* 100 g sup -1 (P < 0.05, Fisher's exact test).

Pa^O2was higher before as well as 90 s and 5 min after epinephrine administration in the ACD-IPPV group as compared with the ACD group. Mixed venous P^O2was not different between both groups at any point in time. Ninety seconds after epinephrine administration, Pa^O2increased in ACD-IPPV-treated, but not in ACD-treated animals in comparison with preepinephrine levels, whereas mixed venous P^O2did not change in any group between points in time before and 90 s after epinephrine administration. Pa^CO2and mixed venous P^CO2were less before as well as 90 s and 5 min after epinephrine administration in pigs subjected to ACD-IPPV in comparison with ACD treated pigs. While 90 s after epinephrine administration Pa^CO2and mixed venous P^CO2remained unchanged in the ACD-IPPV group, in the ACD group, Pa^CO2and mixed venous P^CO2increased in comparison with preepinephrine levels. Arterial and mixed venous pH values were higher before as well as 90 s and 5 min after epinephrine administration in ACD-IPPV-treated animals in comparison with animals treated with ACD alone. Ninety seconds after epinephrine administration, arterial pH decreased in the ACD group, but not in the ACD-IPPV group whereas mixed venous pH decreased in both groups in comparison with preepinephrine levels. The absolute value of the arterial-mixed venous Pa^CO2gradient was greater before as well as 90 s and 5 min after administration of epinephrine in animals treated with ACD-IPPV in comparison with ACD-treated animals. The arterial-mixed venous P^CO2gradient increased in the ACD-IPPV group and decreased in the ACD group 90 s after epinephrine administration in comparison with preepinephrine levels (Table 3).

In the ACD-IPPV group, minute volume was 6.5, 6.2–6.8, and 6.1, 5.8–6.4 l/min after 1 and 7 min of CPR, respectively. In the ACD group, minute volume was 4.2, 2.3–4.6, and 1.6, 0.8–2.3 l/min after 1 and 7 min of CPR, respectively (Figure 1). Calculated tidal volume was 325, 310–339, and 305, 288–322 ml after 1 and 7 min of ACD-IPPV (minute volume divided by respiratory rate: 20/min) and was 52, 28–57, and 19, 329 ml after 1 and 7 min of ACD alone (minute volume divided by chest compression rate: 80/min).

In comparison with standard CPR, ACD has been shown to improve hemodynamics, vital organ blood flow, and immediate resuscitation success in animal as well as in human studies. Myocardial blood flow, which is one of the most important factors for restarting the arrested heart, was significantly greater when ACD was compared with standard CPR in pigs during cardiac arrest. [2,3].

Because ACD increases negative intrathoracic pressure, in addition to augmenting venous return and myocardial blood flow, ACD is able to increase minute ventilation independently of IPPV in comparison to other CPR techniques. The ventilatory aspects of ACD have raised interest in studies investigating ventilation generated by ACD versus other CPR techniques and the potential role of ACD as a substitute for rescue breathing during basic life support. [12–14].

Even with standard CPR, the necessity for ventilation during basic life support has been challenged recently by animal investigations. In dogs, it was observed that an arterial oxygen saturation greater than 90% was maintained for the first 5 min of ventricular fibrillation with chest compressions alone [16]and with concern to outcome, in swine, chest compressions alone appeared to be as effective as chest compressions combined with ventilation. [17]In contrast, in another study in pigs, greater hypoxia and hypercarbic acidosis as well as a lower rate of return of spontaneous circulation was found in animals in which the lungs were not ventilated compared with animals [18]in which the lungs were ventilated.

In the current study, in pigs receiving ACD without IPPV, otherwise ideal conditions including tracheal intubation, 100% oxygen, and a very short period of ventricular fibrillation were provided. However, 10 min after starting CPR, only in the ACD-IPPV group, myocardial blood flow was greater than the critical threshold of 20 ml *symbol* min sup -l *symbol* 100 g sup -1.

Recently, it has been shown that during ACD coronary perfusion is present during both compression and active decompression and is increased in comparison to that during standard CPR. [2]In the current study, coronary perfusion pressure was not significantly different in pigs subjected to ACD-IPPV as compared with pigs subjected to ACD alone and thus, may not have been the major determinant for increased myocardial blood flow with ACD-IPPV at the end of the resuscitation phase.

After 10 min of CPR, there were marked differences in Pa^Osub 2 and Pa^CO2as well as in mixed venous P^CO2levels between animals with and without IPPV. Myocardial hypercarbic acidosis was reported to reduce cardiac resuscitability independent of coronary perfusion pressure in a porcine model of cardiac arrest. [19]However, because we did not measure intramyocardial carbon dioxide we cannot conclude to which extent mixed venous P^CO2represented intramyocardial carbon dioxide in the current study and whether mixed venous and intramyocardial carbon dioxide, respectively, affected myocardial blood flow.

During hypoxemia in lambs, metabolic acidemia attenuated the hemodynamic response to epinephrine. [20]In our study, mean arterial and coronary perfusion pressures were significantly increased in the ACD-IPPV as well as in the ACD group 90 s after administration of epinephrine. Thus, a diminished hemodynamic response to epinephrine administration seems unlikely to be the underlying cause for a diminished myocardial blood flow in animals without IPPV.

The lower arterial-mixed venous P^CO2and pH gradients in the ACD group, which were also observed in a study in pigs whose lungs were not ventilated during CPR [13]indicate that an increase in Pa^CO2did not lead to a proportional increase in organ venous P^COsub 2. As a possible explanation for this phenomenon, it is suggested that under the conditions of CPR without IPPV the amount of carbon dioxide transported in arterial blood may be small in comparison to the total amount of carbon dioxide in organ venous blood and may not significantly contribute to mixed venous carbon dioxide content. In our study, cardiac index was equal in animals treated with ACD-IPPV and ACD and thus the arterial-mixed venous P^CO2gradient did not reflect cardiac output.

In comparison with other animal studies on the ventilatory effects of ACD, in our study minute ventilation and the capability to maintain Pa^O2and Pa^CO2in physiologic limits was in the lower range. In tracheally intubated, not mechanically ventilated supine beagles weighing 10–15 kg, a minute volume of 4.9 l/min was achieved in 1-min trials of ACD. [12]Another investigation in tracheally intubated mongrel dogs (13–22 kg) placed in the lateral position showed that after a previous phase of 9 min of different CPR techniques in combination with IPPV, a minute ventilation of 13 l/min was generated in 3-min trials of ACD without IPPV. [14]In pigs receiving ACD without IPPV over 10 min after 6 min of ventricular fibrillation, minute ventilation was greater, and hypoxia and hypercarbic acidosis was less as compared to animals receiving standard CPR. [13]Differences in age, animal species (i.e., regarding thorax configuration), and the methods used may account for some of the different results in the literature and in the current study with respect to the amount of minute ventilation achieved. In our study, in animals treated with ACD without IPPV, minute ventilation and blood gases were acceptable at the beginning of ACD but deteriorated markedly with ongoing resuscitation. However, even at the beginning of ACD, as in other studies on the ventilatory effects of ACD, at 80 cycles/min the calculated tidal volume was low and the ratio between dead space and tidal volume must have been high. Further studies are needed to investigate the possible mechanisms how ACD improves gas exchange during CPR.

There are several potential limitations of this study. Because compliance was not measured, it cannot be concluded whether alterations in the compliance of the thorax and lung, respectively, may have contributed to the deterioration of ventilation after a longer period of ACD compared with ACD and IPPV. Because we did not investigate long-term survival or neurologic outcome, we cannot conclude whether IPPV would have influenced these variables in comparison to omission of IPPV. Because of radiation protection regulations we were not able to measure myocardial blood flow before induction of ventricular fibrillation and to control this variable for intergroup differences at this time. However, myocardial blood flow was not different before and 90 s after epinephrine administration and hence the observed difference at 5 min after epinephrine is unlikely to be caused by preexisting differences. Finally, by using 100% of oxygen before induction of a 30-s period of ventricular fibrillation and by choosing an epinephrine dosage of 0.2 mg/kg, our model was designed to improve oxygenation, hemodynamics, and resuscitation success rather than to simulate real clinical conditions.

In conclusion, the results of this study demonstrate that gas exchange was critically low during the late phase of ACD without IPPV. While myocardial blood flow was well preserved after 10 min of ACD combined with IPPV and administration of epinephrine, under the same conditions but without IPPV myocardial blood flow was impaired and was less than the critical threshold necessary for restoration of spontaneous circulation.