INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: cardiac arrest; ventricular fibrillations; asphyxia; myocardial function

Experimental and clinical investigations now support the notion that deaths after initially successful resuscitation from ventricular tachycardia or ventricular fibrillation (VF), at least in part, are caused by postresuscitation impairment in myocardial function (1). This contrasts with patients who present with extremes of bradycardia or asystole in whom there is more often a primary respiratory or neural cause for cardiac arrest (7). Asphyxial causes of cardiac arrest, resulting from airway obstruction or failure of ventilation, account for the vast majority of instances of cardiac arrest in pediatric victims (8). Suboptimal exchange of oxygen and carbon dioxide is associated with progressive hypoxemia, hypercarbia, bradycardia, hypotension, and loss of consciousness before pulselessness. The onset of asphyxial cardiac arrest is therefore more gradual in contrast to the pulselessness, loss of consciousness, and apnea after the sudden onset of VF, which is the predominant cause of cardiac arrest in older adults (9).

In the present study, we sought to compare postresuscitation myocardial function and duration of survival in animals in which cardiac arrest was induced by VF and animals in which cardiac arrest followed asphyxiation caused by apnea. Additional studies were performed in which the duration of asphyxial cardiac arrest was adjusted to correspond to the duration of cardiac arrest in the VF group and in which the animals received a comparable number of electrical shocks.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals received humane care following approval of the Institute of Critical Care Medicine's Animal Use and Care Committee, and in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health. Our Institute is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care.

Animal Preparation

Male Sprague-Dawley breeder rats, weighing between 450 and 550 g, were fasted overnight except for free access to water. Anesthesia was initiated by intraperitoneal injection of 45 mg/kg pentobarbital sodium and supplemented with additional hourly doses of 10 mg/kg. Animals were placed on a surgical board in a supine position with extremities immobilized. The trachea was orally intubated with a 14-gauge cannula (Abbocath-T; Abbott Hospital Products, North Chicago, IL) mounted on a blunt needle by techniques previously described (13).

From the surgically exposed right carotid artery, an 18-gauge polyethylene catheter (Intramedic PE50; Becton Dickinson, Sparks, MD) was advanced into the left ventricle for measurement of left ventricular (LV) pressure and both rate of pressure rise at LV pressure of 40 mm Hg (dP/dt₄₀) and rate of pressure decrease (negative dP/dt). An 18-gauge polyethylene catheter was advanced through the left external jugular vein into the right atrium for injection of thermal tracer. A 3-French pediatric radial arterial catheter (C-PUM-301J; Cook Critical Care, Ellettsville, IN) was advanced from the right external jugular vein into the right atrium. Two additional Intramedic PE50 catheters were advanced through the surgically exposed left femoral artery and vein into the abdominal aorta and into the inferior vena cava for measurements of arterial pressures and for arterial and venous blood sampling. Cardiac output was measured with a thermocouple microprobe (number 9030-12-34; Columbus Instruments, Columbus, OH), which was advanced through the right femoral artery into the thoracic aorta. Blood temperature was monitored with this sensor and maintained between 36.5° C and 37.5° C with the aid of infrared heating lamps. End-tidal PCO₂ (EtPCO₂) was measured with a side-stream infrared CO₂ analyzer (End-Tid IL 200; Instrumentation Laboratory Inc., Lexington, MA). For animals randomized to dysrhythmic cardiac arrest, a precurved guidewire, supplied with the pediatric radial artery catheter, was advanced into the right ventricle until an endocardial electrocardiogram (ECG) was documented.

Experimental Procedures

Anesthetized animals were mechanically ventilated beginning 15 min before inducing apnea or VF, using a high-frequency jet ventilator of our own design as previously described (13). Tidal volume was established at 0.65 ml/100 g animal weight and respiratory frequency at 100/min. For an initial interval of 5 min of mechanical ventilation, the fraction of inspired oxygen (FI_O2) was established at 1.0, and baseline blood gases were measured. The interval was to obtain baseline values that would be comparable to those obtained after resuscitation when animals were also returned to an FI_O2 of 1.0. After baseline measurements were completed, the FI_O2 was decreased to 0.21. Four groups of animals were investigated. The first group represented apnea without airway obstruction. Mechanical ventilation was discontinued and apnea was induced after a right atrial injection of 0.4 mg/kg pancuronium bromide (Organon Inc., West Orange, NJ). This group was designated "NMB." In the second group, the procedure was identical except that the endotracheal tube was occluded with a plastic male Luer lock adapter (Baxter Healthcare Corp., Deerfield, IL). This group was designated "NMB-O." Asphyxia was confirmed when there was no evidence of endotracheal gas flow, on the capnogram over an interval of 4 min. In the setting of asphyxia, cardiac arrest was defined as a mean aortic pressure decline to less than 30 mm Hg. Initial management was restoration of mechanical ventilation with an FI_O2 of 1.0. Failing to restore aortic pressure of greater than 30 mm Hg within 30 s, precordial compression was begun in addition to continued mechanical ventilation using a pneumatically driven mechanical chest compressor. Precordial compression was maintained at a rate of 200/min and synchronized to provide a compression/ventilation ratio of 2:1 with equal compression-relaxation intervals as previously described (13). Depth of compression was adjusted to decrease the anteroposterior diameter of the chest by 30% and subsequently adjusted to secure a coronary perfusion pressure (CPP)  20 mm Hg. Our rationale for selecting different methods for producing asphyxia was to individually investigate whether there were any differences in postresuscitation myocardial function after respiratory arrest induced by apnea alone or apnea/asphyxia associated with airway obstruction.

The third group represented cardiac arrest resulting from VF. VF was induced with delivery of an alternating current of 2 to 6 mA to the right ventricular endocardium as previously described (3, 6, 13). Four minutes after onset of VF, precordial compression and mechanical ventilation were started and continued for 2 min. Defibrillation was then attempted with a 2-J DC shock delivered between the anterior chest and conductive foil in contact with the skin of the posterior thorax. Failing to reverse VF, precordial compression was resumed for an additional interval of 30 s, followed by a second sequence of three shocks. A third sequence of defibrillatory shocks was attempted if VF persisted after an additional 30 s of precordial compression. In the fourth group, apnea was induced to produce cardiac arrest in an identical fashion to that of the NMB group. Cardiac arrest was maintained for an interval comparable to that of the VF group. Because investigations in our laboratory had demonstrated that electrical shocks of themselves adversely affect postresuscitation myocardial function (6), approximately the same number and energy of electrical shocks were delivered over comparable intervals to these animals (NMB-E) during resuscitation even though these animals maintained a regular rhythm.

Resuscitation was defined as the reestablishment of a mean aortic pressure of 60 mm Hg or greater for an interval exceeding 5 min. Hemodynamic measurements and mechanical ventilation with an FI_O2 of 1.0 were continued for a total of 4 h. The endotracheal tube and catheters were then removed. After recovery from anesthesia, animals were returned to their cages, and physical activity and survival were recorded at 24-h intervals for a total of 72 h.

After 72 h, the animals were euthanized with an intraperitoneal injection of 100 mg/kg of pentobarbital. Autopsy was performed on all animals with gross inspection of thoracic and abdominal organs to identify potential adverse effects of the resuscitative interventions.

Measurements

Blood was sampled in amounts of 0.4 ml from the aorta and the inferior vena cava for laboratory measurement, and an equal quantity of blood was transfused into the left femoral vein from an anesthetized donor rat of the same colony as previously described (13). Aortic and central venous blood pH, PCO₂, and PO₂, and lactate were measured with a Stat Profile Ultra (Nova Biomedical Co, Waltham, MA). ECG limb leads II and aVR were recorded with the aid of needle electrodes. Cardiac output was measured with the thermodilution technique in which a bolus of 200 µl of saline at a temperature of between 10 and 15° C was injected into the right atrium. Cardiac index was computed with an adaptation of a commercially available data acquisition system and software (National Instruments, Austin, TX). Aortic, LV, and right atrial pressures, ECG, and PET_CO2 were recorded on a PC-based data acquisition system using CODAS software (DATAQ Instruments, Akron, OH) at a sampling rate of 300/s as previously described (3, 6, 13). CPP was calculated as the difference between (decompression) diastolic aortic pressure and time-coincident right atrial pressure as previously described.

Myocardial function was assessed from measurements of LV pressure. The rate of LV pressure increase was measured at a LV pressure of 40 mm Hg (dP/dt₄₀) by analog differentiation. It represents an index of isovolemic contractility. The maximal rate of LV pressure decline ( dP/dt) was measured as an estimate of myocardial relaxation.

Neurologic deficit was scored at 24, 48, and 72 h after resuscitation on a scale of 0 (fully alert and active) to 500 (nonreactive with apnea) as described by Hendrickx and coworkers (9).

Statistical Analyses

Measurements were reported as mean ± SD. Baseline measurements between groups were compared by analysis of variance (ANOVA). Time-based measurements within groups were compared by ANOVA repeated measurements. A p value of less than 0.05 was regarded as significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Thirty-four animals were investigated. Two animals randomized to the NMB group were excluded from analysis and rerandomized because retrospective review indicated that the mean aortic pressure did not decline to less than 30 mm Hg such that the animals failed to fulfill the criteria of asphyxial cardiac arrest.

Baseline hemodynamic measurements did not differ significantly among the four groups. There were also no significant differences in weight, baseline blood gases, or baseline arterial blood lactate among the four groups (Table 1).

Asphyxial cardiac arrest occurred approximately 170 s after injection of the neuromuscular blocking agent. The durations of asphyxial cardiac arrest in the NMB group and NMB-O group were 103 and 105 s, respectively. Ventricular fibrillation appeared in only one animal among the asphyxial groups and that animal was successfully defibrillated. Excepting this animal, there was a consistent and progressive slowing of the supraventricular rhythm from baseline levels of 371 ± 23 to 86 ± 23. In animals in which VF was induced, an average of seven electrical shocks were required before restoration of spontaneous circulation (ROSC) (Table 2). The duration of cardiopulmonary resuscitation (CPR), and specifically ventilation and precordial compression before ROSC, averaged 174 s. In the asphyxial group, ventilation and precordial compression were followed by ROSC within an average of 56 to 77 s (Table 2). All animals in the asphyxial group were resuscitated. No significant differences in postresuscitation myocardial function between NMB and NMB-O, the two initial asphyxial groups, were demonstrated based on cardiac index, dP/dt₄₀ (Figure 1), negative dP/dt, and LV end-diastolic pressure (LVEDP) (Figure 2).

View larger version (21K): [in this window] [in a new window]   Figure 1.   There were no significant differences in either cardiac index or dP/dt40 between two asphyxial groups (NMB and NMB-O). Numbers in parentheses indicate the number of animals. Closed circles: NMB; open circles: NMB-O.

View larger version (18K): [in this window] [in a new window]   Figure 2.   There were no significant differences in either  dP/dt40 or LVEDP between two asphyxial groups (NMB and NMB-O). Numbers in parentheses indicate the number of animals. Closed circles: NMB; open circles: NMB-O.

Because the experimental models of asphyxial cardiac arrest (NMB and NMB-O) did not take into account the longer durations of cardiac arrest observed for the VF model or the effects of electrical shocks, we were prompted to control these variables. We therefore extended our studies to include an additional model of asphyxial cardiac arrest. Durations of cardiac arrest in this model were adjusted to correspond to the average duration established in the VF model. In addition, we elected to deliver electrical shocks with energies and in numbers that corresponded to the VF model, even though the rhythm remained sinus bradycardia. In this additional group (NMB-E), cardiac index and dP/dt₄₀ were compared with the VF group as shown in Figure 3. Postresuscitation myocardial function based on the cardiac index and dP/dt₄₀ was significantly less impaired in the NMB-E group. Lesser impairment in myocardial relaxation, as measured by disproportionately greater negative dP/dt and significantly lower LVEDP were also documented in this modified asphyxial model (Figure 4). However, neither neurologic deficits measured at 24, 48, and 72 h in survivors (Table 3) nor the duration of survival among the four groups were significantly different (Table 2).

View larger version (22K): [in this window] [in a new window]   Figure 3.   Significantly greater cardiac index and dP/dt40 were observed after resuscitation from asphyxial cardiac arrest (NMB-E) than that from dysrhythmic cardiac arrest (VF). Numbers in parentheses indicate the number of animals. Open triangles: VF; open squares: NMB-E. *p < 0.05, p < 0.01 versus VF.

View larger version (23K): [in this window] [in a new window]   Figure 4.   Significantly greater -dP/dt40 and lower LVEDP were observed after resuscitation from asphyxial cardiac arrest (NMB-E) than that from dysrhythmic cardiac arrest (VF). Numbers in parentheses indicate the number of animals. Open triangles: VF; open squares: NMB-E. *p < 0.05, p < 0.01 versus VF.

At autopsy, minor lung contusions were observed in five of 40 animals. However, there was no evidence of other adverse effects of the resuscitative interventions on thoracic and abdominal organs.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the United States, approximately 350,000 victims annually have fatal cardiac arrest before admission to the hospital. Although the initial success of CPR on these victims is 40%, a majority die within 72 h, primarily owing to recurrent VF associated with impaired myocardial function (1, 14). In most communities, CPR for victims of cardiac arrest yields a functional survival rate of only 5%.

The clinical course that follows successful resuscitation after dysrhythmic cardiac arrest in human victims has been documented by two recent multicenter studies. Approximately 60% of victims died within the initial 72 h after successful cardiac resuscitation. Refractory hypotension and recurrent VF were identified as predominant causes of postresuscitation death (1, 2). In the Brain Resuscitation Clinical Trial I, approximately 60% of successfully resuscitated patients died during the first 10 d. Arterial hypotension and recurrent VF were again identified as the primary events (15). We have implicated reversible myocardial dysfunction after initially successful resuscitation to explain, at least in part, "postresuscitation myocardial dysfunction" (3).

In the present study, no significant differences in cardiac function, neurologic responsiveness, or duration of survival were observed between asphyxial cardiac arrest with and without airway obstruction. However, the studies were performed in anesthetized animals after neuromuscular blockade in which the effects of airway obstruction were minimal in the absence of a struggle to maintain spontaneous ventilation. Because the duration of untreated cardiac arrest, and therefore the duration of no flow, is the single best predictor of the severity of postresuscitation myocardial dysfunction, we recognized the importance of controlling this variable (3, 4). Because the severity of postresuscitation myocardial dysfunction increased in close relationship to the total energy of the electrical shocks delivered before ROSC (6), a comparable number of electrical shocks was administered to the VF and NMB-E groups. We therefore conclude that postresuscitation myocardial systolic and diastolic function of the asphyxial group (NMB-E) in which both the duration of cardiac arrest and the number of electrical shocks were adjusted, continued to be less than that observed after dysrhythmic cardiac arrest caused by VF.

In contrast to the "sudden death" of dysrhythmic cardiac arrest, conscious victims of asphyxial cardiac arrest typically manifest respiratory distress. As in our experimental model, bradycardia typically advances to pulseless electrical activity rather than VF (16, 17). Bradycardia and hypotension are also the rule in hibernating animals, and these physiologic adaptations serve to minimize systemic and myocardial oxygen requirements. During the no flow or minimal flow state of clinical cardiac arrest, bradycardia would mitigate the detriment of oxygen deficits and resultant ischemic injury to the myocardium. This is in contrast to VF in which disproportionately greater myocardial oxygen requirements account for early and marked global myocardial ischemia (18, 19). These differences in metabolic demands are likely to explain, at least in part, the greater severity of postresuscitation myocardial ischemic injury when cardiac arrest is caused by VF (20).

Our studies have focused on adverse effects on the myocardium. However, it is well documented that asphyxia results in early cerebral ischemic injury (11). Therefore, we anticipated that survival in animals after asphyxial cardiac arrest was more contingent on neurologic outcomes. Although the present study does not secure that possibility, we do not exclude the potential importance of combined myocardial and neurologic ischemic injury as a determinant of outcomes.