Resuscitation, 20 (1990) 57-66 Elsevier Scientific Publishers Ireland Ltd.
57
Factors influencing variable outcomes after ventricular fibrillation cardiac arrest of 15 minutes in dogs*p*** Mark Angelos ****, Harvey Reich and Peter Safar International Resuscitation Research Center (IRRC), Department of Anesthesiology and Critical Care Medicine, Center for Emergency Medicine and Presbyterian-University Hospital University of Pittsburgh, PA (U.S.A.) (Received December 8th, 1989; Revision received April 3rd. 1990, Accepted May 17th, 1990)
Animal experiments with cardiac arrest and cardiopulmonary resuscitation (CPR) despite controlled insult and postinsult life support, have yielded variable individual outcomes. This report concerns 10 dog experiments with a standardized model of VF cardiac arrest with no flow for 10 min followed by CPR basic life support (BLS) from VF 10 to 15 min and then CPR advanced life support (ALS) with epinephrine at 15 min. Defibrillating countershocks began at 17 min, for restoration of spontaneous circulation. After controlled ventilation to 20 h and intensive care to % h, outcome was evaluated using the overall performance category (OPC) 1 (normal) (n5) vs. OPC 2-4 (impaired) (n5) (P < 0.001). We searched for correlations between normal vs. impaired outcome in various prearrest, arrest and postarrest factors that are suspected to influence postarrest neurologic deficit. Prearrest variables were similar in the normal and impaired groups. Resuscitation variables were similar in both. Coronary perfusion pressure during CPR-ALS was higher in the normal outcome group (P = 0.03). Among postarrest variables, postarrest reperfusion pressure pattern (initial hypertensive bout), blood glucose, cardiac output, Hct, pHa, Pao, and Pace, were the same. Our data support the importance of maximizing coronary perfusion pressure not only for restoration of heart beat but also as a possible predictor of improved cerebral outcome. Cardiopulmonary resuscitation - Cardiac arrest Coronary perfusion pressure - Hypothermia
INTRODUCTION
external Studies of cardiac arrest (no flow) reversed by standard cardiopulmonary resuscitation (CPR) basic life support (BLS) and advanced life support (ALS), have shown that with the same duration of ischemic insult and the same detailed protocols of resuscitation and prolonged life support, there can be variable neurologic outcome in both animals [l-6] and patients [6-g]. In con*Supported by a 1986/87 fellowship grant from the Emergency Medicine Foundation and the American College of Emergency Physicians, ** and the Asmund S. Laerdal Foundation. ***Presented at the Second International Conference on Emergency Medicine, Brisbane, Australia, October 24-28.1988, Address all correspondence and reprint requests to: Mark Angelos, M.D., Cox Institute, Department of Emergency Medicine, Wright State University, 3525 Southern Boulevard, Kettering, OH 45429, U.S.A. Printed and Published in Ireland
58
trolled cardiac arrest-CPR animal studies, many variables have been shown to affect neurologic outcome. Other variables have been suspected as factors and also been controlled [lO,l 11. In addition to the duration of the primary insult of temporary complete brain ischemia, there is the variable low flow state of CPR and numerous known or unknown factors post-CPR, which may contribute to secondary brain damage and the multiorgan systems postresuscitation syndrome [ 12,131. While factors affecting restoration of spontaneous heartbeat have been examined extensively, factors important for the final neurologic outcome of cardiac arrest survivors are relatively unexplored. Our objective was to determine significant differences between dogs with normal neurologic outcome and dogs with impaired neurologic outcome following cardiac arrest. We retrospectively determined which of the factors known or suspected to influence outcome may have differed between groups. We used our standardized model of cardiac arrest, resuscitation and postarrest intensive care [3], which yielded complete neurologic recovery in five dogs and impaired neurologic recovery in five other dogs. MATERIALS AND METHODS
This study was approved by our institution’s Animal Care and Use Committee. Ten healthy male coonhounds of 22.5 [18-291 kg body weight and 8.5 [6-111 months of age were fasted overnight with free access to water. We used our dog ventricular fibrillation (VF) cardiac arrest outcome model that uses CPR ALS for restoration of spontaneous circulation, controlled intermittent positive pressure ventilation (IPPV) for 20 h, and intensive care to 96 h [1,5,10,11]. All experiments were performed within 3 months by the same team. Briefly, anesthesia was induced with ketamine 10 mg/kg i.m. and maintained with N,O:O, 50:50% plus halothane 0.25-1.0% (adjusted for arterial pressure control) delivered via endotracheal tube and IPPV. Hydration was with i.v. Ringers solution, 4 ml/kg per h. Tidal volumes were 15 ml/kg with the ventilatory frequency (I) adjusted to maintain end tidal and arterial PCO, to 30-35 mmHg. Positive end expiratory pressure (PEEP) of 5 cmH,O was used. The electrocardiogram (ECG) and electroencephalogram (EEG) were monitored noninvasively. Stomach and bladder were catheterized. Sterile femoral cutdowns were used for the necessary cannulations. Continuously monitored were: heart rate (HR), ECG, mean arterial pressure (MAP), central venous pressure (CVP), pulmonary artery (core) temperature (Tpa), end tidal CO,, EEG and F,o,. Intermittently monitored were: pulmonary artery and occlusion pressure (Ppa, Ppao), cardiac output (CO), hematocrit (Hct), blood glucose, Pao,, Pace,, pHa, arterial base excess (BE), urine flow and pupillary size and reactivity to light. Coronary perfusion pressure during CPR was calculated as intrathoracic aortic pressure minus right atria1 pressure, both during thoracic diastole. All pressures were determined using electromechanical pressure transducers. Variables controlled before and for 24 h after arrest were: MAP at 110 f 15 mmHg (with norepinephrine or trimethaphan); CVP and Ppao at 3-15 mmHg (with i.v. fluids); Tpa at 37.0 + 2.OmC (with external heating/cooling); Pao, at > 100 mmHg with F,o, and PEEP; Pace, at 30-35 mmHg (by adjusting f); pHa at 7.2-7.5 and BE at
59
0 f 7 mEq/kg (with f and NaHCO, i.v.); blood glucose prearrest at 90-180 mg/dl (by withholding i.v. glucose); and urine flow at > 1 ml/kg per h (with i.v. fluids and furosemide 0.25-1.0 mg/kg i.v. if needed). Insult and resuscitation Two preinsult baseline measurements were performed in sequence within 30 min of the insult, during a steady state of light anesthesia with N,O:O, 50:50% plus halothane < 0.5% and IPPV. Prior to the insult, N,O and halothane were discontinued. IPPV with 0, 100% was given for 1 min followed by room air for 4 min to lighten anesthesia in a standardized manner, while the dogs were paralyzed with pancuronium. In pilot experiments, the same anesthesia without paralysis did not cause awakening. External heating and i.v. infusion were stopped. VF was then induced with an external transthoracic electric shock of 60 V AC, which was repeated if necessary. Simultaneously IPPV was stopped. VF resulted in an isoelectric EEG within lo-20 s. VF was allowed to continue for 10 min of no flow. This was followed by CPR-BLS from VF 10 to 15 min (low flow) and then CPR-ALS beginning at VF 15 min. The first defibrillating countershock was at VF 17 min. Attempts at restoration of spontaneous circulation were standardized and carried out until success or VF time 40 min. At VF 10 min, all dogs received IPPV with Fro, 0.21, tidal volumes of 20 ml/kg, f = 12 and mechanical external chest compressions at a rate of 6O/min by a Michigan Instruments Thumper. CPR was performed with the dog taped supine into a V-shaped trough, in a 10’ right oblique position, to apply chest compressions (of 100 lbs) over the left margin of the sternum. This we have previously found to provide better MAP during CPR than mid-sternal compressions. Following CPR-BLS (simulating bystander CPR), CPR-ALS was started at VF 15 min by changing F,o, from 0.21 to 1.O and administering epinephrine 0.02 mg/kg and NaHCO, 1 mEq/kg through the central venous line. External chest compressions were continued as before. At VF 17 min (after 2 min of CPR with epinephrine) the first external defibrillating countershock of 100 J was administered. If necessary, two additional countershocks of 200 J and 300 J were immediately given. If unsuccessful, the same dose of epinephrine was repeated at VF 18 min followed by 3 countershocks of 300 J each. If still in VF, lidocaine 1 mg/kg i.v. was administered, again followed by 3 countershocks of 300 J. Additional NaHCO, i.v. was administered to keep BE at + 7 mEp/kg based on arterial blood gases. Immediately following defibrillation a hypertensive bout with MAP B 140 mmHg for about 5 min was achieved with a titrated i.v. infusion ofasrepinephrine. Prolonged life support After restoration of spontaneous circulation and a hypertensive bout, MAP and other variables were controlled at prearrest levels. Cardiac output > 60%‘of prearrest control values was maintained with i.v. fluids and dobutamine l-15 &kg per min by titrated infusion. The dogs remained immobilized with pancuronmm and sedated with fentanyl 5 pg/kg i.v. as needed., Pancuronium and fentanyl were discontinued at 20 h. IPPV was with 0, 100% for the first 2 h and then with N,O:O, 50:50010for analgesia from 2-20 h. Respiratory care included tracheal suctioning, intermittent deep lung inflations and frequent turning. Tetracycline 50 mg was given
i.v. every 8 h, for 2 days, for infection prophylaxis. Normothermia was maintained. At 20 h, paralysis was reversed with neostigmine 0.5 mg plus atropine 0.5 mg i.v. and the dogs were extubated and given 0, by face mask to keep Pao, > 80 mmHg. When the dogs were hemodynamically stable, had normal blood gas values and were conscious, all lines were removed and they were transferred to a crib or floor mat. Supplemental 0,, i.v. fluids and urine output monitoring were continued until the dog could eat and drink and move about on his own. All dogs received continuous 24 h monitoring and nursing care until 96 h. Evaluation The 10 experiments reported here all followed protocol with survival to 96 h. Those excluded using our standard exclusion criteria [ 10,l l] were reported elsewhere [3]. The 10 dogs of this study were separated into a normal outcome group (n5) and an impaired outcome group (n5). The former had complete neurologic recovery, whereas the latter had varying degrees of neurologic deficit, as reflected in our previously documented ‘overall performance categories’ (OPC) [ 10,111. Briefly, OPC 1 = normal; 2 = moderate disability; 3 = severe disability; 4 = coma or vegetative state; and 5 = brain death or death. In this report all 5 dogs of the “normal outcome group” had final 96 h OPC 1; whereas all 5 dogs of the “impaired outcome group” had OPC 2 (n3), OPC 3 (nl), or OPC 4 (nl). Differences in neurologic outcome between groups were confirmed using neurologic deficit (ND) scores (ND 0% = normal, 100% = brain death) [lO,l 11. ND scores were 3 & 6 in the normal group and 39 f 12 in the impaired group (P< 0.01). Retrospective data analysis comparing the two outcome groups was done on the following variables which may influence outcome: (1) prearrest baseline, hemodynamic variables, blood gases, Tpa, blood gluocse and Hct; (2) arrest-resuscitation variables such as MAP and coronary perfusion pressures during BLS and ALS, drug and countershock requirements and time to restoration of spontaneous circulation; and (3) postarrest Tpa, MAP, C.O., Hct, and blood glucose. Neurologic difference between the normal and impaired group was documented by the Wilcox Rank Sum test. Variables (Tables I-III) were calculated as mean values f standard deviation. Statistical comparisons between groups were done using analysis of variance. Differences with P < 0.05 were considered significant. RESULTS
Prearrest baseline variables (Table I) were the same between normal and impaired group except for pHa values, which were statistically higher in the impaired group. All pH prearrest values were within the normal range in both groups. Arrest-resuscitation variables (Table II) showed no statistically significant difference between normal and impaired groups except for coronary perfusion pressure (Pig. l), which was higher during CPR-ALS in the normal outcome group. Restoration of spontaneous circulation was achieved within 2-4 min of ALS attempts in both groups. CPR-ALS produced coronary perfusion pressures before successful defibrillation of 47 f 8 mmHg in the normal outcome group vs. 34 f 8 mmHg in the impaired outcome group (P< 0.05).
61 Table I.
Prearrest baseline variables.
Heart rate MAP, mmHg CVP, mmHg Ppaop, mmHg CI, l/min Tpa Blood glucose Pao,, mmHg Pace,, mmHg pHa BD, mEq/l Hct, %
Normal outcome group (n=S)
Impaired outcome group (n=S)
105
115 102 8 11 3.4 36.7 148 218 35 7.41 + 1.4 36
105 9 10 3.1 36.2 133 245 31 7.36 +4.9 40
f 1s i:8 *3 24 f 0.5 f: 0.6 r?: 32 f 27 23 f: 0.02 f 2.2 +4
f 20 f 16 +4 *4 + 1.0 + 0.9 f 25 f 47 +2 + 0.04* + 3.3 *4
‘P = 0.04.
While MAP during BLS and ALS was not significantly different between the two outcome groups, aortic diastolic pressure rose more in the normal outcome group with the change from BLS to ALS. In the normal outcome group aortic diastolic pressure increased from 39 f 16 mmHg during BLS to 63 + 16 mmHg during ALS Table II.
Arrest-resuscitation
variables. Normal outcome group (n = 5)
Epinephrlne requirement mg/kg NaHCO, requirement mEq/kg Countershocks required (#) Time aortic diastolic pressure < 50 mmHg, min Time coronary perfusion pressure < 30 mmHg, min VF time, min MAP during BLS, mmHg MAP during ALS, mmHg Systolic pressure during BLS. mmHg Systolic pressure during ALS, mmHg Coronary Perf. Press. during BLS. mmHg Coronary Perf. Press. during ALS, mmHg ‘P = 0.03.
Impaired outcome group (n = 5)
0.02 f 0.003
0.04 + 0.05
1
1
2
+2
3
*3
14.3 f 3.6
16.8 f 1.4
14.1 17.3 55 81
+ f f f
14.5 18.3 48 70
f f f f
87
f 12
73
*7
116
f 20
112
26
+ 10
24
zt8
*47
*8
34
+8
2.6 0.5 15 I5
2.6 1.8 11 10
f 20
62
Coronary Perfusion Pressure (CorPP) during CfPR 60
El
Normal Outcome Group
cl
Impaired Outcome Group
CPR-I 1 min
s
*
T
CPR-BLS _ 5 mfn
CPR-ALS 7 min *P = 0.03
CPR Time
Fig. 1. Coronary perfusion pressures (mean and S.D.), during CPR-BLS and CPR-ALS. *CorPP during CPR-ALS significantly higher in normal outcome group (P< 0.03).
(P < 0.05); while in the impaired outcome group, that increase was from 36 + 4 to 49 f 7 mmHg (P< 0.05). Postarrest variables such as Tpa, hypertensive bout (MAP), blood glucose, CO. and Hct showed no statistically significant difference between groups. There was, however, a trend toward lower values in the normal outcome group for Tpa at 1 h, blood glucose, Hct at 2 h; and a slightly higher hypertensive bout (Table III). Table 11. Postresuscitation variables. Normal outcome group (n=S) Tpa (1 h) MAP max. in first 5 min, mmHg MAP 30 min, mmHg MAP 60 min, mmHg Blood glucose (2 h) mg/dl CI (1 h), Vmin CI (4 h). l/min CI (12 h), l/min Hct (Zh), olo
36.3 rt 0.06
Impaired outcome group (n=5) 36.9 k 1.0
226 130 124
f 38 + 17 f 18
205 115 124
f 22 f 16 f 20
166 4.07 2.88 3.64 42
f 27 f 1.57 f 0.46 f 1.04 +7
187 4.76 3.3 3.f6 46i
t60 f 1.39 * 0.90 f 0.56 rs
63 DISCUSSION
We evaluated retrospectively factors that might influence neurologic outcome after a standardized VF cardiac arrest no flow of 10 min followed by CPR-BLS of 5 min and CPR-ALS, i.e. a clinically relevant scenario. Data during attempts at restoring spontaneous circulation with CPR-ALS starting at VF 15 min were also compared. Among the many variables examined (Tables I-III) there was a significantly higher coronary perfusion pressure during ALS in the normal outcome group. Coronary perfusion pressure generated during CPR is well established as an important factor in achieving successful defibrillation and restoration of spontaneous heart beat with hypertension or normotension [14-191. Myocardial blood flow during CPR has been show to correlate well with measured coronary perfusion pressure [20-221. Downey et al. [23] observed in dogs that a coronary perfusion pressure of 28 mmHg was required to maintain subendocardial blood flow in the fibrillating heart. Kern et al. [24] reported a positive effect on cardiovascular resuscitability and 24 h survival after cardiac arrest and CPR in dogs. Survivors had a mean coronary perfusion pressure of 29 mmHg vs. 11 mmHg in non-survivors. In our study, dogs had coronary perfusion pressures of 26-55 mmHg (mean 40 f 11 mmHg) during CPR-ALS and were resuscitated. This study is the first to demonstrate improved neurologic outcome at 96 h with higher coronary perfusion pressures. Coronary perfusion pressure is probably reflective of cerebral perfusion pressure which correlates well with cerebral blood flow, There is considerable support from laboratory studies by us and others that improved cerebral perfusion pressure during CPR leads to improved cerebral outcome [2,6,25,26]. Although we did not measure cerebral perfusion pressure in this study, we found in another study a strong correlation between coronary and cerebral perfusion pressures [ 181. Both systolic and mean arterial pressures were higher in the normal outcome group, suggesting again the importance of perfusion pressures early in reperfusion (Table II). These differences were not statistically different perhaps due to the small sample size, but may be important clinical indicators of perfusion. Improved coronary perfusion pressure during CPR and ALS no doubt is only one of several factors influencing ultimate neurologic outcome. Using a controlled model, important pre- and post-arrest variables were controlled, in order to better identify important differences during resuscitation. During resuscitation, the normal outcome group tended to require less epinephrine, have a shorter VF time and less total ischemia time measured as the total time coronary perfusion pressure was less than 30 mmHg. During ALS when coronary perfusion pressure was determined, all animals had received one dose of epinephrine in compliance with the protocol. The trend towards higher total epinephrine requirements in the impaired outcome group is a reflection of a slightly longer resuscitation time. Ischemia, including both global (no flow) and partial (low flow) is an important determinant of outcome [5,6]. We tried to quantitate the total ischemia time (both global and partial) as the time coronary perfusion pressure was less than 30 mmHg. This included time during arrest, resuscitation and post-defibrillation if the dog was hypotensive. These times, as well as total VF times, were very similar between groups. These differences,
64
although small, may have been important factors. However, due to the small numbers in the study they were not statistically significant. Post-arrest hypoperfusion may be a significant factor in impaired neurologic function. In order to limit variability in post-resuscitation cerebral &hernia, mean arterial pressure was closely regulated during the first 20 h. Previous studies have demonstrated improved cerebral outcome following a brief hypertensive surge after defibrillation [26,27]. In accordance with this previous work, a hypertensive surge lasting approximately 5 min was generated immediately after defibrillation by using a titrated norepinephrine infusion. Peak mean arterial pressure during the hypertensive surge was similar in both groups and exceeded 200 mmHg. There are many variables that might influence outcome in addition to the insult itself. For this reason control of pre-and post-arrest variables is imperative. Essential requirements for animal models used in cerebral resuscitation have been previously reviewed [lo]. This study further demonstrates the good outcome that can be achieved after a prolonged cardiac arrest when high coronary perfusion pressures are generated. After a mean VF time of 17.3 min, including 10 min without CPR or ventilation, five dogs had normal neurologic function at 96 h. Five other dogs with lower coronary perfusion pressure, but still higher pressures than those recommended for successful defibrillation, had return of spontaneous circulation and 96 h survival with impaired neurologic outcome. Attaining high coronary perfusion pressure during cardiac arrest would appear to be more important than the techniques used to attain them, i.e. high dose epinephrine, vest CPR, cardiopulmonary bypass, etc. CONCLUSION
Using a standardized cardiac arrest model with external closed-chest CPR and controlled post-resuscitation care, survivors with normal neurologic outcome had higher coronary perfusion pressure than survivors with impaired neurologic outcome. Improved coronary perfusion pressure probably reflects cerebral perfusion pressure during early cardiac arrest reperfusion. Coronary reperfusion pressure may be an important indicator not only of successful cardiac defibrillation and early survival, but also of neurologic recovery. ACKNOWLEDGEMENTS
S. William Stezoski, Henry Alexander and Frank Houghton helped with the experiments. Lisa Cohn helped with editing. Gale Foster helped with preparation of the manuscript. REFERENCES 1 2
P. Vaagenes, R. Cantadore, P. Safar et al., Amelioration of brain damage?.by lidoflazine after prolonged ventricular fibrillation cardiac arrest in dogs, Crit. Care Med., 12 (l!jSS) 846-855. N. Bircher and P. Safar, Cerebral preservation during cardiopulmonary resuscitation, Crit. Care Med., 13 (1985) 185-190.
65 3
4 5
6 I 8 9 10
11
12 13 14 15 16 17 18 19
20
21 22
23 24 25
M. Angelos, H. Reich, P. Safar et al., Cardiac resuscitability and neurologic recovery after clinically relevant cardiac arrest of 20 min with cardiopulmonary bypass vs. cardiopulmonary resuscitation in dogs, Anesthesiology, 67 (1987) A147. E. Cerchiari, E. Klain. P. Safar et al., Cardiovascular post-resuscitation syndrome (PRS) after ventricular fibrillation cardiac arrest (VFCA) in dogs, Crit. Care Med., 16 (1988) 445 (abstract). P. Safar, N.S. Abramson, M. Angelos et al., Emergency cardiopulmonary bypass for resuscitation from prolonged cardiac arrest, Am. J. Emerg. Med., 8 (1990) 55-67. P. Safar, Resuscitation from clinical death: Pathophysiologic limits and therapeutic potentials, Crit. Care Med., 16 (1988) 923-941. R.G. Myerburg, C.A. Conde, R.J. Sung et al., Clinical elcctrophysiologic and hemodynamic profile of patients resuscitated from prehospital cardiac arrest, Am. J. Med., 68 (1980) 568-576. E. Bass, Cardiopulmonary arrest pathophysiology and neurologic complications, Ann. lnt. Med., 103 (1985) 920-927. N.S. Abramson, P. Safar, K.M. Detre et al. and the BRCT I Study Group, Randomized clinical study of thiopental in comatose survivors of cardiac arrest, N. Engl. J. Med., 314 (1986) 397-403. P. Safar, S.E. Gisvold, P. Vaagenes et al., Long term animal models for the study of global brain ischemia. In: Protection of Tissues Against Hypoxia. Editors: A. Wauquier, M. Borgers and W.K. Amery, Elsevier, Amsterdam, 1982, pp. 14-170. P. Safar, F. Sterz, Y. Leonov et al., Systematic development of cerebral resuscitation after cardiac arrest. Three promising treatments: Cardiopulmonary bypass, hypertensive henodilution and mild hypothermia. In: Causes and Mechanisms of Secondary Brain Damage. Editors: A. Baethmann et al., Springer Verlag, Heidelberg, in press. V.A. Negovsky, A.M. Gurvitch and E.S. Zolotokryhna, Postresuscitation Disease, Elsevier, Amsterdam, 1983. P. Safar, Effects of the postresuscitation syndrome on cerebral recovery from cardiac arrest, Crit. Care Med., 13 (1985) 932-935. G. Crile and D.H. Dolley, Experimental research into resuscitation of dogs killed by anesthetics and asphyxia, J. Exp. Med., 8 (1906) 713-738. C.J. Wiggers, The physiological bases for cardiac resuscitation from ventricular fibrillation. Method for serial defibrillation, Am. Heart J., 20(1940)413. L.C. Harris, B. Kirimli and P. Safar, Ventilation-cardiac compression rates and ratios in cardiopulmonary resuscitation, Anesthesiology, 28 (1967) 806. J.S. Redding, Drug therapy during cardiac arrest. In: Advances in Cardiopulmonary Resuscitation. Editors: P. Safar and J. Elam, Springer-Verlag, New York, 1977, pp. 113-l 17. N. Bircher and P. Safar, Comparison of standard and ‘new’ closed-chest CPR and open-chest CPR in dogs, Crit. Care Med., 9 (1981) 384. J.T. Niemann, J.M. Criley and J.P. Rosborough et al., Predictive indices of successful cardiac resuscitation after prolonged arrest and experimental cardiopulmonary resuscitation, Ann. Emerg. Med., 14 (1985) 521-528. S.H. Ralston, W.D. Voorhees and C.F. Babbs, Intrapulmonary epinephrine during prolonged cardiopulmonary resuscitation: Improved regional blood flow and resuscitation in dogs, Am. J. Emerg. Med., 13 (1984) 79-86. H.R. Halperin, J.E. Tsitlik, A.D. Guerci et al., Determinance of blood flow to vital organs during cardiopulmonary resuscitation in dogs, Circulation, 73 (1986) 539-550. A.R. Michael, A.D. Guerci, R.C. Koehler et al., Mechanisms by which epinephrine augments cerebral and myocardial perfusion during cardiopulmonary resuscitation in dogs, Circulation, 69 (1984) 822-835. J.M. Downey, R.W. Chagrasulis and V. Hemphill, Quantitative study of intramyocardial compression in the fibrillating heart, Am. J. Physiol., 237 (1979) H191-H196. K.B. Kern, G.A. Ewy W.D. Voorhees et al., Myocardial perfusion pressure: A predictor of 24 h survival during prolonged cardiac arrest, Resuscitation, 16 (1988) 241-250. K-A. Hossmann, Resuscitation potentials after prolonged global cerebral ischemia in cats, Crit. Care Med., 16 (1988) 964-971.
26
27
K.-A. Hossmann, Hemodynamics of postischemic reperfusion of the brain. In: Protection of the Brain from Ischemia. Editors: P.R. Weinstein and AI. Fayden, Vol. 3, Williams and Wilkins, San Francisco, in press. P. Safar, S.W. Stezoski and E.M. Nemoto, Amelioration of brain damage after 12 min cardiac arrest in dogs, Arch. Neurol., 33 (1976) 91-95.
57
Factors influencing variable outcomes after ventricular fibrillation cardiac arrest of 15 minutes in dogs*p*** Mark Angelos ****, Harvey Reich and Peter Safar International Resuscitation Research Center (IRRC), Department of Anesthesiology and Critical Care Medicine, Center for Emergency Medicine and Presbyterian-University Hospital University of Pittsburgh, PA (U.S.A.) (Received December 8th, 1989; Revision received April 3rd. 1990, Accepted May 17th, 1990)
Animal experiments with cardiac arrest and cardiopulmonary resuscitation (CPR) despite controlled insult and postinsult life support, have yielded variable individual outcomes. This report concerns 10 dog experiments with a standardized model of VF cardiac arrest with no flow for 10 min followed by CPR basic life support (BLS) from VF 10 to 15 min and then CPR advanced life support (ALS) with epinephrine at 15 min. Defibrillating countershocks began at 17 min, for restoration of spontaneous circulation. After controlled ventilation to 20 h and intensive care to % h, outcome was evaluated using the overall performance category (OPC) 1 (normal) (n5) vs. OPC 2-4 (impaired) (n5) (P < 0.001). We searched for correlations between normal vs. impaired outcome in various prearrest, arrest and postarrest factors that are suspected to influence postarrest neurologic deficit. Prearrest variables were similar in the normal and impaired groups. Resuscitation variables were similar in both. Coronary perfusion pressure during CPR-ALS was higher in the normal outcome group (P = 0.03). Among postarrest variables, postarrest reperfusion pressure pattern (initial hypertensive bout), blood glucose, cardiac output, Hct, pHa, Pao, and Pace, were the same. Our data support the importance of maximizing coronary perfusion pressure not only for restoration of heart beat but also as a possible predictor of improved cerebral outcome. Cardiopulmonary resuscitation - Cardiac arrest Coronary perfusion pressure - Hypothermia
INTRODUCTION
external Studies of cardiac arrest (no flow) reversed by standard cardiopulmonary resuscitation (CPR) basic life support (BLS) and advanced life support (ALS), have shown that with the same duration of ischemic insult and the same detailed protocols of resuscitation and prolonged life support, there can be variable neurologic outcome in both animals [l-6] and patients [6-g]. In con*Supported by a 1986/87 fellowship grant from the Emergency Medicine Foundation and the American College of Emergency Physicians, ** and the Asmund S. Laerdal Foundation. ***Presented at the Second International Conference on Emergency Medicine, Brisbane, Australia, October 24-28.1988, Address all correspondence and reprint requests to: Mark Angelos, M.D., Cox Institute, Department of Emergency Medicine, Wright State University, 3525 Southern Boulevard, Kettering, OH 45429, U.S.A. Printed and Published in Ireland
58
trolled cardiac arrest-CPR animal studies, many variables have been shown to affect neurologic outcome. Other variables have been suspected as factors and also been controlled [lO,l 11. In addition to the duration of the primary insult of temporary complete brain ischemia, there is the variable low flow state of CPR and numerous known or unknown factors post-CPR, which may contribute to secondary brain damage and the multiorgan systems postresuscitation syndrome [ 12,131. While factors affecting restoration of spontaneous heartbeat have been examined extensively, factors important for the final neurologic outcome of cardiac arrest survivors are relatively unexplored. Our objective was to determine significant differences between dogs with normal neurologic outcome and dogs with impaired neurologic outcome following cardiac arrest. We retrospectively determined which of the factors known or suspected to influence outcome may have differed between groups. We used our standardized model of cardiac arrest, resuscitation and postarrest intensive care [3], which yielded complete neurologic recovery in five dogs and impaired neurologic recovery in five other dogs. MATERIALS AND METHODS
This study was approved by our institution’s Animal Care and Use Committee. Ten healthy male coonhounds of 22.5 [18-291 kg body weight and 8.5 [6-111 months of age were fasted overnight with free access to water. We used our dog ventricular fibrillation (VF) cardiac arrest outcome model that uses CPR ALS for restoration of spontaneous circulation, controlled intermittent positive pressure ventilation (IPPV) for 20 h, and intensive care to 96 h [1,5,10,11]. All experiments were performed within 3 months by the same team. Briefly, anesthesia was induced with ketamine 10 mg/kg i.m. and maintained with N,O:O, 50:50% plus halothane 0.25-1.0% (adjusted for arterial pressure control) delivered via endotracheal tube and IPPV. Hydration was with i.v. Ringers solution, 4 ml/kg per h. Tidal volumes were 15 ml/kg with the ventilatory frequency (I) adjusted to maintain end tidal and arterial PCO, to 30-35 mmHg. Positive end expiratory pressure (PEEP) of 5 cmH,O was used. The electrocardiogram (ECG) and electroencephalogram (EEG) were monitored noninvasively. Stomach and bladder were catheterized. Sterile femoral cutdowns were used for the necessary cannulations. Continuously monitored were: heart rate (HR), ECG, mean arterial pressure (MAP), central venous pressure (CVP), pulmonary artery (core) temperature (Tpa), end tidal CO,, EEG and F,o,. Intermittently monitored were: pulmonary artery and occlusion pressure (Ppa, Ppao), cardiac output (CO), hematocrit (Hct), blood glucose, Pao,, Pace,, pHa, arterial base excess (BE), urine flow and pupillary size and reactivity to light. Coronary perfusion pressure during CPR was calculated as intrathoracic aortic pressure minus right atria1 pressure, both during thoracic diastole. All pressures were determined using electromechanical pressure transducers. Variables controlled before and for 24 h after arrest were: MAP at 110 f 15 mmHg (with norepinephrine or trimethaphan); CVP and Ppao at 3-15 mmHg (with i.v. fluids); Tpa at 37.0 + 2.OmC (with external heating/cooling); Pao, at > 100 mmHg with F,o, and PEEP; Pace, at 30-35 mmHg (by adjusting f); pHa at 7.2-7.5 and BE at
59
0 f 7 mEq/kg (with f and NaHCO, i.v.); blood glucose prearrest at 90-180 mg/dl (by withholding i.v. glucose); and urine flow at > 1 ml/kg per h (with i.v. fluids and furosemide 0.25-1.0 mg/kg i.v. if needed). Insult and resuscitation Two preinsult baseline measurements were performed in sequence within 30 min of the insult, during a steady state of light anesthesia with N,O:O, 50:50% plus halothane < 0.5% and IPPV. Prior to the insult, N,O and halothane were discontinued. IPPV with 0, 100% was given for 1 min followed by room air for 4 min to lighten anesthesia in a standardized manner, while the dogs were paralyzed with pancuronium. In pilot experiments, the same anesthesia without paralysis did not cause awakening. External heating and i.v. infusion were stopped. VF was then induced with an external transthoracic electric shock of 60 V AC, which was repeated if necessary. Simultaneously IPPV was stopped. VF resulted in an isoelectric EEG within lo-20 s. VF was allowed to continue for 10 min of no flow. This was followed by CPR-BLS from VF 10 to 15 min (low flow) and then CPR-ALS beginning at VF 15 min. The first defibrillating countershock was at VF 17 min. Attempts at restoration of spontaneous circulation were standardized and carried out until success or VF time 40 min. At VF 10 min, all dogs received IPPV with Fro, 0.21, tidal volumes of 20 ml/kg, f = 12 and mechanical external chest compressions at a rate of 6O/min by a Michigan Instruments Thumper. CPR was performed with the dog taped supine into a V-shaped trough, in a 10’ right oblique position, to apply chest compressions (of 100 lbs) over the left margin of the sternum. This we have previously found to provide better MAP during CPR than mid-sternal compressions. Following CPR-BLS (simulating bystander CPR), CPR-ALS was started at VF 15 min by changing F,o, from 0.21 to 1.O and administering epinephrine 0.02 mg/kg and NaHCO, 1 mEq/kg through the central venous line. External chest compressions were continued as before. At VF 17 min (after 2 min of CPR with epinephrine) the first external defibrillating countershock of 100 J was administered. If necessary, two additional countershocks of 200 J and 300 J were immediately given. If unsuccessful, the same dose of epinephrine was repeated at VF 18 min followed by 3 countershocks of 300 J each. If still in VF, lidocaine 1 mg/kg i.v. was administered, again followed by 3 countershocks of 300 J. Additional NaHCO, i.v. was administered to keep BE at + 7 mEp/kg based on arterial blood gases. Immediately following defibrillation a hypertensive bout with MAP B 140 mmHg for about 5 min was achieved with a titrated i.v. infusion ofasrepinephrine. Prolonged life support After restoration of spontaneous circulation and a hypertensive bout, MAP and other variables were controlled at prearrest levels. Cardiac output > 60%‘of prearrest control values was maintained with i.v. fluids and dobutamine l-15 &kg per min by titrated infusion. The dogs remained immobilized with pancuronmm and sedated with fentanyl 5 pg/kg i.v. as needed., Pancuronium and fentanyl were discontinued at 20 h. IPPV was with 0, 100% for the first 2 h and then with N,O:O, 50:50010for analgesia from 2-20 h. Respiratory care included tracheal suctioning, intermittent deep lung inflations and frequent turning. Tetracycline 50 mg was given
i.v. every 8 h, for 2 days, for infection prophylaxis. Normothermia was maintained. At 20 h, paralysis was reversed with neostigmine 0.5 mg plus atropine 0.5 mg i.v. and the dogs were extubated and given 0, by face mask to keep Pao, > 80 mmHg. When the dogs were hemodynamically stable, had normal blood gas values and were conscious, all lines were removed and they were transferred to a crib or floor mat. Supplemental 0,, i.v. fluids and urine output monitoring were continued until the dog could eat and drink and move about on his own. All dogs received continuous 24 h monitoring and nursing care until 96 h. Evaluation The 10 experiments reported here all followed protocol with survival to 96 h. Those excluded using our standard exclusion criteria [ 10,l l] were reported elsewhere [3]. The 10 dogs of this study were separated into a normal outcome group (n5) and an impaired outcome group (n5). The former had complete neurologic recovery, whereas the latter had varying degrees of neurologic deficit, as reflected in our previously documented ‘overall performance categories’ (OPC) [ 10,111. Briefly, OPC 1 = normal; 2 = moderate disability; 3 = severe disability; 4 = coma or vegetative state; and 5 = brain death or death. In this report all 5 dogs of the “normal outcome group” had final 96 h OPC 1; whereas all 5 dogs of the “impaired outcome group” had OPC 2 (n3), OPC 3 (nl), or OPC 4 (nl). Differences in neurologic outcome between groups were confirmed using neurologic deficit (ND) scores (ND 0% = normal, 100% = brain death) [lO,l 11. ND scores were 3 & 6 in the normal group and 39 f 12 in the impaired group (P< 0.01). Retrospective data analysis comparing the two outcome groups was done on the following variables which may influence outcome: (1) prearrest baseline, hemodynamic variables, blood gases, Tpa, blood gluocse and Hct; (2) arrest-resuscitation variables such as MAP and coronary perfusion pressures during BLS and ALS, drug and countershock requirements and time to restoration of spontaneous circulation; and (3) postarrest Tpa, MAP, C.O., Hct, and blood glucose. Neurologic difference between the normal and impaired group was documented by the Wilcox Rank Sum test. Variables (Tables I-III) were calculated as mean values f standard deviation. Statistical comparisons between groups were done using analysis of variance. Differences with P < 0.05 were considered significant. RESULTS
Prearrest baseline variables (Table I) were the same between normal and impaired group except for pHa values, which were statistically higher in the impaired group. All pH prearrest values were within the normal range in both groups. Arrest-resuscitation variables (Table II) showed no statistically significant difference between normal and impaired groups except for coronary perfusion pressure (Pig. l), which was higher during CPR-ALS in the normal outcome group. Restoration of spontaneous circulation was achieved within 2-4 min of ALS attempts in both groups. CPR-ALS produced coronary perfusion pressures before successful defibrillation of 47 f 8 mmHg in the normal outcome group vs. 34 f 8 mmHg in the impaired outcome group (P< 0.05).
61 Table I.
Prearrest baseline variables.
Heart rate MAP, mmHg CVP, mmHg Ppaop, mmHg CI, l/min Tpa Blood glucose Pao,, mmHg Pace,, mmHg pHa BD, mEq/l Hct, %
Normal outcome group (n=S)
Impaired outcome group (n=S)
105
115 102 8 11 3.4 36.7 148 218 35 7.41 + 1.4 36
105 9 10 3.1 36.2 133 245 31 7.36 +4.9 40
f 1s i:8 *3 24 f 0.5 f: 0.6 r?: 32 f 27 23 f: 0.02 f 2.2 +4
f 20 f 16 +4 *4 + 1.0 + 0.9 f 25 f 47 +2 + 0.04* + 3.3 *4
‘P = 0.04.
While MAP during BLS and ALS was not significantly different between the two outcome groups, aortic diastolic pressure rose more in the normal outcome group with the change from BLS to ALS. In the normal outcome group aortic diastolic pressure increased from 39 f 16 mmHg during BLS to 63 + 16 mmHg during ALS Table II.
Arrest-resuscitation
variables. Normal outcome group (n = 5)
Epinephrlne requirement mg/kg NaHCO, requirement mEq/kg Countershocks required (#) Time aortic diastolic pressure < 50 mmHg, min Time coronary perfusion pressure < 30 mmHg, min VF time, min MAP during BLS, mmHg MAP during ALS, mmHg Systolic pressure during BLS. mmHg Systolic pressure during ALS, mmHg Coronary Perf. Press. during BLS. mmHg Coronary Perf. Press. during ALS, mmHg ‘P = 0.03.
Impaired outcome group (n = 5)
0.02 f 0.003
0.04 + 0.05
1
1
2
+2
3
*3
14.3 f 3.6
16.8 f 1.4
14.1 17.3 55 81
+ f f f
14.5 18.3 48 70
f f f f
87
f 12
73
*7
116
f 20
112
26
+ 10
24
zt8
*47
*8
34
+8
2.6 0.5 15 I5
2.6 1.8 11 10
f 20
62
Coronary Perfusion Pressure (CorPP) during CfPR 60
El
Normal Outcome Group
cl
Impaired Outcome Group
CPR-I 1 min
s
*
T
CPR-BLS _ 5 mfn
CPR-ALS 7 min *P = 0.03
CPR Time
Fig. 1. Coronary perfusion pressures (mean and S.D.), during CPR-BLS and CPR-ALS. *CorPP during CPR-ALS significantly higher in normal outcome group (P< 0.03).
(P < 0.05); while in the impaired outcome group, that increase was from 36 + 4 to 49 f 7 mmHg (P< 0.05). Postarrest variables such as Tpa, hypertensive bout (MAP), blood glucose, CO. and Hct showed no statistically significant difference between groups. There was, however, a trend toward lower values in the normal outcome group for Tpa at 1 h, blood glucose, Hct at 2 h; and a slightly higher hypertensive bout (Table III). Table 11. Postresuscitation variables. Normal outcome group (n=S) Tpa (1 h) MAP max. in first 5 min, mmHg MAP 30 min, mmHg MAP 60 min, mmHg Blood glucose (2 h) mg/dl CI (1 h), Vmin CI (4 h). l/min CI (12 h), l/min Hct (Zh), olo
36.3 rt 0.06
Impaired outcome group (n=5) 36.9 k 1.0
226 130 124
f 38 + 17 f 18
205 115 124
f 22 f 16 f 20
166 4.07 2.88 3.64 42
f 27 f 1.57 f 0.46 f 1.04 +7
187 4.76 3.3 3.f6 46i
t60 f 1.39 * 0.90 f 0.56 rs
63 DISCUSSION
We evaluated retrospectively factors that might influence neurologic outcome after a standardized VF cardiac arrest no flow of 10 min followed by CPR-BLS of 5 min and CPR-ALS, i.e. a clinically relevant scenario. Data during attempts at restoring spontaneous circulation with CPR-ALS starting at VF 15 min were also compared. Among the many variables examined (Tables I-III) there was a significantly higher coronary perfusion pressure during ALS in the normal outcome group. Coronary perfusion pressure generated during CPR is well established as an important factor in achieving successful defibrillation and restoration of spontaneous heart beat with hypertension or normotension [14-191. Myocardial blood flow during CPR has been show to correlate well with measured coronary perfusion pressure [20-221. Downey et al. [23] observed in dogs that a coronary perfusion pressure of 28 mmHg was required to maintain subendocardial blood flow in the fibrillating heart. Kern et al. [24] reported a positive effect on cardiovascular resuscitability and 24 h survival after cardiac arrest and CPR in dogs. Survivors had a mean coronary perfusion pressure of 29 mmHg vs. 11 mmHg in non-survivors. In our study, dogs had coronary perfusion pressures of 26-55 mmHg (mean 40 f 11 mmHg) during CPR-ALS and were resuscitated. This study is the first to demonstrate improved neurologic outcome at 96 h with higher coronary perfusion pressures. Coronary perfusion pressure is probably reflective of cerebral perfusion pressure which correlates well with cerebral blood flow, There is considerable support from laboratory studies by us and others that improved cerebral perfusion pressure during CPR leads to improved cerebral outcome [2,6,25,26]. Although we did not measure cerebral perfusion pressure in this study, we found in another study a strong correlation between coronary and cerebral perfusion pressures [ 181. Both systolic and mean arterial pressures were higher in the normal outcome group, suggesting again the importance of perfusion pressures early in reperfusion (Table II). These differences were not statistically different perhaps due to the small sample size, but may be important clinical indicators of perfusion. Improved coronary perfusion pressure during CPR and ALS no doubt is only one of several factors influencing ultimate neurologic outcome. Using a controlled model, important pre- and post-arrest variables were controlled, in order to better identify important differences during resuscitation. During resuscitation, the normal outcome group tended to require less epinephrine, have a shorter VF time and less total ischemia time measured as the total time coronary perfusion pressure was less than 30 mmHg. During ALS when coronary perfusion pressure was determined, all animals had received one dose of epinephrine in compliance with the protocol. The trend towards higher total epinephrine requirements in the impaired outcome group is a reflection of a slightly longer resuscitation time. Ischemia, including both global (no flow) and partial (low flow) is an important determinant of outcome [5,6]. We tried to quantitate the total ischemia time (both global and partial) as the time coronary perfusion pressure was less than 30 mmHg. This included time during arrest, resuscitation and post-defibrillation if the dog was hypotensive. These times, as well as total VF times, were very similar between groups. These differences,
64
although small, may have been important factors. However, due to the small numbers in the study they were not statistically significant. Post-arrest hypoperfusion may be a significant factor in impaired neurologic function. In order to limit variability in post-resuscitation cerebral &hernia, mean arterial pressure was closely regulated during the first 20 h. Previous studies have demonstrated improved cerebral outcome following a brief hypertensive surge after defibrillation [26,27]. In accordance with this previous work, a hypertensive surge lasting approximately 5 min was generated immediately after defibrillation by using a titrated norepinephrine infusion. Peak mean arterial pressure during the hypertensive surge was similar in both groups and exceeded 200 mmHg. There are many variables that might influence outcome in addition to the insult itself. For this reason control of pre-and post-arrest variables is imperative. Essential requirements for animal models used in cerebral resuscitation have been previously reviewed [lo]. This study further demonstrates the good outcome that can be achieved after a prolonged cardiac arrest when high coronary perfusion pressures are generated. After a mean VF time of 17.3 min, including 10 min without CPR or ventilation, five dogs had normal neurologic function at 96 h. Five other dogs with lower coronary perfusion pressure, but still higher pressures than those recommended for successful defibrillation, had return of spontaneous circulation and 96 h survival with impaired neurologic outcome. Attaining high coronary perfusion pressure during cardiac arrest would appear to be more important than the techniques used to attain them, i.e. high dose epinephrine, vest CPR, cardiopulmonary bypass, etc. CONCLUSION
Using a standardized cardiac arrest model with external closed-chest CPR and controlled post-resuscitation care, survivors with normal neurologic outcome had higher coronary perfusion pressure than survivors with impaired neurologic outcome. Improved coronary perfusion pressure probably reflects cerebral perfusion pressure during early cardiac arrest reperfusion. Coronary reperfusion pressure may be an important indicator not only of successful cardiac defibrillation and early survival, but also of neurologic recovery. ACKNOWLEDGEMENTS
S. William Stezoski, Henry Alexander and Frank Houghton helped with the experiments. Lisa Cohn helped with editing. Gale Foster helped with preparation of the manuscript. REFERENCES 1 2
P. Vaagenes, R. Cantadore, P. Safar et al., Amelioration of brain damage?.by lidoflazine after prolonged ventricular fibrillation cardiac arrest in dogs, Crit. Care Med., 12 (l!jSS) 846-855. N. Bircher and P. Safar, Cerebral preservation during cardiopulmonary resuscitation, Crit. Care Med., 13 (1985) 185-190.
65 3
4 5
6 I 8 9 10
11
12 13 14 15 16 17 18 19
20
21 22
23 24 25
M. Angelos, H. Reich, P. Safar et al., Cardiac resuscitability and neurologic recovery after clinically relevant cardiac arrest of 20 min with cardiopulmonary bypass vs. cardiopulmonary resuscitation in dogs, Anesthesiology, 67 (1987) A147. E. Cerchiari, E. Klain. P. Safar et al., Cardiovascular post-resuscitation syndrome (PRS) after ventricular fibrillation cardiac arrest (VFCA) in dogs, Crit. Care Med., 16 (1988) 445 (abstract). P. Safar, N.S. Abramson, M. Angelos et al., Emergency cardiopulmonary bypass for resuscitation from prolonged cardiac arrest, Am. J. Emerg. Med., 8 (1990) 55-67. P. Safar, Resuscitation from clinical death: Pathophysiologic limits and therapeutic potentials, Crit. Care Med., 16 (1988) 923-941. R.G. Myerburg, C.A. Conde, R.J. Sung et al., Clinical elcctrophysiologic and hemodynamic profile of patients resuscitated from prehospital cardiac arrest, Am. J. Med., 68 (1980) 568-576. E. Bass, Cardiopulmonary arrest pathophysiology and neurologic complications, Ann. lnt. Med., 103 (1985) 920-927. N.S. Abramson, P. Safar, K.M. Detre et al. and the BRCT I Study Group, Randomized clinical study of thiopental in comatose survivors of cardiac arrest, N. Engl. J. Med., 314 (1986) 397-403. P. Safar, S.E. Gisvold, P. Vaagenes et al., Long term animal models for the study of global brain ischemia. In: Protection of Tissues Against Hypoxia. Editors: A. Wauquier, M. Borgers and W.K. Amery, Elsevier, Amsterdam, 1982, pp. 14-170. P. Safar, F. Sterz, Y. Leonov et al., Systematic development of cerebral resuscitation after cardiac arrest. Three promising treatments: Cardiopulmonary bypass, hypertensive henodilution and mild hypothermia. In: Causes and Mechanisms of Secondary Brain Damage. Editors: A. Baethmann et al., Springer Verlag, Heidelberg, in press. V.A. Negovsky, A.M. Gurvitch and E.S. Zolotokryhna, Postresuscitation Disease, Elsevier, Amsterdam, 1983. P. Safar, Effects of the postresuscitation syndrome on cerebral recovery from cardiac arrest, Crit. Care Med., 13 (1985) 932-935. G. Crile and D.H. Dolley, Experimental research into resuscitation of dogs killed by anesthetics and asphyxia, J. Exp. Med., 8 (1906) 713-738. C.J. Wiggers, The physiological bases for cardiac resuscitation from ventricular fibrillation. Method for serial defibrillation, Am. Heart J., 20(1940)413. L.C. Harris, B. Kirimli and P. Safar, Ventilation-cardiac compression rates and ratios in cardiopulmonary resuscitation, Anesthesiology, 28 (1967) 806. J.S. Redding, Drug therapy during cardiac arrest. In: Advances in Cardiopulmonary Resuscitation. Editors: P. Safar and J. Elam, Springer-Verlag, New York, 1977, pp. 113-l 17. N. Bircher and P. Safar, Comparison of standard and ‘new’ closed-chest CPR and open-chest CPR in dogs, Crit. Care Med., 9 (1981) 384. J.T. Niemann, J.M. Criley and J.P. Rosborough et al., Predictive indices of successful cardiac resuscitation after prolonged arrest and experimental cardiopulmonary resuscitation, Ann. Emerg. Med., 14 (1985) 521-528. S.H. Ralston, W.D. Voorhees and C.F. Babbs, Intrapulmonary epinephrine during prolonged cardiopulmonary resuscitation: Improved regional blood flow and resuscitation in dogs, Am. J. Emerg. Med., 13 (1984) 79-86. H.R. Halperin, J.E. Tsitlik, A.D. Guerci et al., Determinance of blood flow to vital organs during cardiopulmonary resuscitation in dogs, Circulation, 73 (1986) 539-550. A.R. Michael, A.D. Guerci, R.C. Koehler et al., Mechanisms by which epinephrine augments cerebral and myocardial perfusion during cardiopulmonary resuscitation in dogs, Circulation, 69 (1984) 822-835. J.M. Downey, R.W. Chagrasulis and V. Hemphill, Quantitative study of intramyocardial compression in the fibrillating heart, Am. J. Physiol., 237 (1979) H191-H196. K.B. Kern, G.A. Ewy W.D. Voorhees et al., Myocardial perfusion pressure: A predictor of 24 h survival during prolonged cardiac arrest, Resuscitation, 16 (1988) 241-250. K-A. Hossmann, Resuscitation potentials after prolonged global cerebral ischemia in cats, Crit. Care Med., 16 (1988) 964-971.
26
27
K.-A. Hossmann, Hemodynamics of postischemic reperfusion of the brain. In: Protection of the Brain from Ischemia. Editors: P.R. Weinstein and AI. Fayden, Vol. 3, Williams and Wilkins, San Francisco, in press. P. Safar, S.W. Stezoski and E.M. Nemoto, Amelioration of brain damage after 12 min cardiac arrest in dogs, Arch. Neurol., 33 (1976) 91-95.
