Cardiovascular Research 1992;26:791-797
79 1
Myocardial oxygen requirements during experimental cardiopulmonary resuscitation Roy V Ditchey, Yoichi Goto, and JoAnn Lindenfeld Objective: The aims were to determine myocardial oxygen requirements during cardiopulmonary resuscitation (CPR), and to test the hypothesis that endogenous catecholamines have a major effect on myocardial oxygen requirements in this setting. Methods: Myocardial oxygen consumption (MV02) was measured during 20 minutes of CPR in eight anaesthetised dogs. Coronary blood flow was maintained at prearrest levels using an external pump to provide a permissive level of oxygen delivery during ventricular fibrillation. Oxygen content was measured in arterial and coronary sinus blood samples under prearrest conditions and at 5 rnin intervals during CPR. Four dogs were given propranolol ( I mg.kg-’) following the 5 min measurements. Results: MVo2 averaged 108.7(SEM 12.8)%of the initial prearrest values after 5 min CPR (n=8). After 10 min CPR, MVo2 fell to 53.8(13.3)% of the initial prearrest values in the subset of animals given propranolol after the 5 rnin measurements (n=4), but remained at prearrest levels in untreated animals ( ~ ~ 0 . 0for5 an interactive effect between treatment and time). MVoz subsequently tended to decrease with time in untreated animals, but remained a high percentage of prearrest values throughout the 20 rnin period of CPR. Conclusions: These findings suggest that endogenous sympathetic stimulation of the fibrillating heart results in high myocardial oxygen requirements during CPR. Cardiovascular Research 1992;26:79 1-797
A
lthough available evidence suggests that closed chest cardiopulmonary resuscitation (CPR) cannot maintain coronary blood flow at prearrest the level of flow required to meet cardiac metabolic demands during CPR is unknown. In experimental preparations in which coronary flow is maintained with donor blood, myocardial oxygen consumption (MVor) typically is much lower during ventricular fibrillation than when the heart is beating and performing external work.” However, cardiovascular collapse is accompanied by intense activation of the sympathetic nervous system, and p adrenergic stimulation of the heart is known to increase myocardial oxygen requirements during ventricular fibrillation.” We hypothesised that endogenous neurohumoral sympathetic stimulation of the heart during CPR raises myocardial oxygen requirements to a high level during resuscitation from ventricular fibrillation. To test this hypothesis, we measured myocardial oxygen consumption before and during a 20 rnin period of CPR and observed the effects of pharmacological p adrenergic blockade. Because the low levels of coronary blood flow typically present during CPR might limit myocardial oxygen consumption by restricting oxygen delivery, we artificially maintained coronary flow at prearrest levels during CPR using an extracorporeal perfusion system.
Methods Studies were performed in eight mongrel dogs weighing an average of 27.6 kg (range 25.5 to 30.5 kg). The dogs were anaesthetised with morphine sulphate (2 mg.kg-’) and (Y chloralose (100 mg.kg-l), intubated, and mechanically
ventilated with a device that mixed room air with 100% oxygen by a venturi effect. The experiments were approved by the institutional animal care and use committee of the University of Vermont. Fluid filled catheters were inserted via a femoral artery and vein and positioned in the ascending aorta and right atrium, respectively. Catheter position was confirmed by appropriate pressure recordings, and by direct palpation. In addition, large diameter (approximately 4 mm internal diameter) plastic catheters with multiple side holes were inserted into the left external jugular vein via a side branch and into the descending thoracic aorta via the remaining femoral artery. A left lateral thoracotomy was performed, and the heart was suspended in a pericardial cradle. After injecting heparin, the coronary sinus was ligated near its insertion into the right atrium, and an additional large diameter catheter was passed through the chest wall, inserted through a small incision in the coronary sinus, and tied in place. This catheter was connected to the external jugular catheter via a rubber tube, diverting coronary sinus blood flow into the left external jugular vein (fig 1). The catheter in the coronary sinus had a side hole immediately proximal to the site of insertion to which a separate fluid filled catheter was attached and passed through the chest wall to allow measurement of coronary sinus pressure. The rubber tube comprising the extrathoracic portion of this drainage system had an internal diameter of approximately I cm, and was used to reduce resistance to blood flow (compared to the smaller diameter tubing required for intravascular insertion) and to provide a means of sampling coronary sinus blood by needle puncture. The large diameter catheter in the descending thoracic
Cardiology Unit, Department of Medicine, Medical Center Hospital of Vermont, Burlington, VT 05401, USA: R V Ditchey, Y Goto; Cardiology Division, University of Colorado Health Sciences Center, Denver, CO, USA: .ILindenfeld. Correspondence to Dr Ditchey.
792
Ditchey, Goto, Lindenfeld
Ilf J
From femoral artery
Large diameter
' E
_
t2g 2% 0 0
u=
_
-_ -
0-
-1s-
-
_.
~~
-
_-
-__ ~
~
--
~
_.__
-
___
-
-_
-____ . _
-
Figure 2 An example of the phasic relationships between
%?t
jf
To LAD
To circumflex
ascending aortic pressure and pressure and flow within the extracorporeal coronary perjiusion sytem under prearrest conditions. Instantaneous jlow increased during diastole and decreased during systole.
tj/ From coronary sinus
Figure 1 Extracorporeal coronary perfusion system (see text for
explanation). LADdefi anterior descending coronary artery: circumflex=circumflexcoronary artery.
aorta was connected to a Masterflex model 7523-00 peristaltic pump, through which arterial blood was withdrawn and forced sequentially through a filter and a heating coil submerged in a water bath maintained at approximately 37°C. The distal end of this tube was passed through the chest wall and then bifurcated via a Y connector into two limbs, each of which was connected distally to an adapter suitable for insertion into a coronary artery (fig 1). The proximal (approximately 2-3 cm) segments of the left anterior descending and circumflex coronary arteries were freed from surrounding fat and connective tissue, following which lignocaine (1.5 mg.kg-') was given intravenously. After filling the extracorporeal perfusion system with blood, the left anterior descending segment was ligated proximally and clamped distally. One of the distal limbs of the perfusion system was clamped and the tip of the other was inserted into the left anterior descending coronary artery near the site of ligation and tied in place. This limb of the coronary perfusion system had a side hole immediately proximal to the site of insertion to which a separate fluid filled catheter was attached and passed through the chest wall to allow measurement of coronary arterial pressure. This catheter, the catheter attached to the side of the coronary sinus drainage system, and the ascending aortic and right atrial catheters were connected to Could P23XL pressure transducers with zero reference set at the mid-chest level. Coronary blood flow then was restored by pumping femoral arterial blood into the left anterior descending coronary artery at a rate sufficient to raise the mean coronary arterial-coronary sinus pressure difference to a level slightly above the simultaneously recorded mean aortic-right atrial pressure difference. After a brief recovery period, the second limb of
the arterial perfusion system was inserted into the circumflex coronary artery in a similar manner, following which the rate at which arterial blood was pumped into the coronary arterial system was increased until the mean coronary arterialcoronary sinus pressure difference was restored to an appropriate level. Successful cannulation of each coronary artery and restoration of flow required approximately 1-5 min coronary arterial occlusion per artery. In some cases, proximal branches were ligated if they could not be included in the portion of the artery perfused artificially. The resolution of reactive hyperaemia (ie, a decrease in the flow required to maintain the same pressure in the first few minutes following relief of coronary occlusion) was taken as evidence that circulatory control mechanisms were not disrupted by instrumentation. The portion of the coronary perfusion catheter system leading from the femoral artery to the Masterflex pump contained a small segment of rubber tubing that allowed sampling of arterial blood by direct needle puncture. In some cases, phasic coronary flow signals were recorded using a cannulating electromagnetic flow probe inserted into the coronary perfusion system and connected to a Carolina Medical Electronics model FM501 square wave electromagnetic flow meter. Zero reference for the flow signal was determined by simultaneously turning off the perfusion pump and clamping the tube containing the electromagnetic flow probe. The Masterflex pump was calibrated in vitro, and quantitative levels of phasic coronary blood flow were estimated on that basis. Recorded flow signals were used for the purpose of illustration only. The filter in the coronary perfusion system contained an air buffer (fig 1) that effectively removed pulsations due to the peristaltic pump and allowed instantaneous coronary flow to vary during the cardiac cycle in response to changes in downstream resistance (fig 2). Finally, a fluid filled catheter was inserted into the left atrium via the left atrial appendage, and a bipolar electrode was sewn to the surface of the left ventricular free wall, following which the left atrial catheter and the electrical leads from the epicardial electrode were exteriorised through small incisions in the chest wall. This catheter was connected
Myocardial oxygen consumption during cardiopulmonary resuscitation
to a fifth Gould P23XL pressure transducer for measurements of left atrial pressure. The pericardial cradle was released and the chest was closed in layers and drained of air, following which the dogs were secured in a supine position and ventilatory parameters were adjusted so that arterial pH was between 7.35 and 7.45, and arterial Pcoz was between 4.65 and 6.00 kPa. Sodium bicarbonate solution was given intravenously when necessary. In addition, normal saline was given in a volume sufficient to raise mean left atrial pressure to approximately 5 mm Hg. This was done both as an arbitrary replacement for the extracorporeal blood pool and insensible fluid losses during instrumentation, and to reduce variability in myocardial oxygen consumption during ventricular fibrillation due to differences in cardiac di~tension.’~ The rate at which blood was pumped into the coronary arterial system was adjusted a final time to restore the mean coronary arterial-coronary sinus pressure difference to a level slightly above the mean aortic-right atrial pressure difference present under these conditions, and then was held constant for the remainder of the study. Aortic, right atrial, coronary arterial, coronary sinus, and left atrial pressures and coronary flow signals were recorded under prearrest conditions (ie, prior to induction of ventricular fibrillation), and arterial and coronary sinus blood samples were obtained in heparinised glass syringes for determination of oxygen content. The arterial blood sample was also used for measurement of packed cell volume. In addition, a separate coronary arterial blood sample was obtained for determination of prearrest adrenaline and noradrenaline concentrations and placed in a tube containing EGTA and glutathione. After a brief interval, prearrest pressure measurements were repeated, and a second set of arterial and coronary sinus blood samples was obtained for determination of oxygen contents to assess reproducibility. Blood samples were stored on ice until the end of the study. In five dogs, heart rate at the time of the first set of prearrest measurements was calculated from an electrocardiogram or the phasic coronary flow signal (electrocardiograms and phasic pressures and flow signals were not recorded in the remaining three dogs). Ventricular fibrillation then was induced by applying a low voltage alternating current to the epicardial electrode. The current was discontinued, and anteroposterior chest compression was initiated using a pneumatic chest compression device (Michigan Instruments Life Aid model X1004 cardiopulmonary resuscitator). The chest was compressed at a rate of 60min-’ with a compression duration equal to approximately 50% of each cycle, and a compression force sufficient to produce between 2.0 and 2.5 inches of posterior sternal displacement, as estimated by a displacement transducer connected to the compression piston. The release phase of the compression cycle was prolonged by approximately 0.5 s every fifth compression to allow lung inflation to a peak inspiratory pressure of 20 cm of water by a synchronised pressure limited ventilator. Chest compression force was adjusted, if necessary, during the remainder of the study in order to maintain a relatively constant extent of sternal displacement. Higher levels of chest compression force and greater sternal displacement were avoided to minimise the possibility of dislodging the coronary perfusion system from the coronary arteries. CPR was continued for a total of 20 min, during which combined left anterior descending and circumflex coronary blood flow was maintained at the same level present during prearrest measurements. This provided a constant and relatively high
793
level of coronary blood flow during CPR, such that myocardial oxygen consumption could vary (ie, either increase or decrease) in response to metabolic requirements, rather than be limited by inadequate coronary blood flow and restricted oxygen delivery. Pressure and flow recordings were repeated 5 , 10, 15, and 20 min after the onset of ventricular fibrillation and CPR. Arterial and coronary sinus blood samples for determination of oxygen content and arterial samples for blood gas analysis, packed cell volume, and determination of plasma adrenaline and noradrenaline concentrations were obtained at each interval. Four dogs were studied without further intervention. In the remaining four dogs, propranolol (1 mg-kg-’) was injected through the left atrial catheter after recording pressures and obtaining arterial and coronary sinus blood samples at the 5 min CPR interval. Arterial and coronary sinus blood oxygen contents were measured using a Lex-02-Con oxygen content analyser. In most cases, the oxygen content in each sample was measured twice and the results averaged. The average difference between duplicate measurements in 67 samples was less than 2% of the mean value. The plasma from each blood sample drawn for measurement of catecholamine concentrations was separated by centrifugation and frozen at -70°C. Plasma adrenaline and noradrenaline concentrations in each sample were determined at a later date using high pressure liquid chromatography. Pressures, flows, and electrocardiograms were recorded with a Hewlett Packard model 7848A eight channel recorder. Mean pressures were averaged over at least two respiratory cycles for prearrest measurements and over 20 chest compression cycles for each set of CPR measurements. Myocardial oxygen consumption in the distribution supplied by the artificial perfusion system was calculated as the product of coronary blood flow and the measured coronary arterial-coronary sinus oxygen content difference. Because coronary blood flow was held contant during each study, the flow term in the calculation was eliminated by expressing MVoz as a percentage of the first prearrest value. This was done to avoid possible errors in flow calibration and also served to normalise for baseline differences in MVo2 between dogs. Data analysis Differences between prearrest and CPR values for the study variables of interest were tested for statistical significance by repeated measures analysis of variance. The variables analysed were MVoz, plasma adrenaline and noradrenaline concentrations, and arterial pH and PCOZ.In the case of MVoz, two separate analyses were performed. First, to test the hypothesis that MVoz during CPR is a high percentage of prearrest values, an analysis of variance was restricted to data obtained under prearrest conditions and after 5 min of CPR when conditions were comparable in all eight dogs (ie, prior to propranolol administration). Second, to assess the effect of p adrenergic blockade, an analysis of variance was performed considering the control and propranolol treated dogs as separate subsets for all six prearrest and CPR conditions. For all other variables, the data are reported separately to show the general comparability of the two study groups, but statistical analyses for the variables of interest were performed using only combined data. Multiple comparison testing of the differences between measurements made under the first prearrest condition and each subsequent condition, if indicated, was performed using Dunnett’s test. Data are reported as the mean(SEM). A p value ~ 0 . 0 5was considered significant.
794
Ditchey, Goto,
Lindenfeld
Table I Oxygen contents and consumption under prearrest conditions and after 5 min cardiopulmonary resuscitution (CPR). Vrrlries are means(SEM), n=8
Prearrest 1 Prearrest 2 CPR 5
Arterial olvgen content (ml.df')
Coronarv sinus oxygen content (m/.df')'
Arteritrl-coronary sinus oxyRen content difference
M.vocardio1 oxygen consumption (% of Prrorrest 1 )
14.9( I .O) 1S.O(O.9) 16.3(0.7)
6.7(0.6) 6.8(0.6) 7.7( I .O)
R.l(O.7) 8.2(0.6) 8.7( 1 . I )
I00 102.7(2.6) I08.7( 12.8)
Myocardial oxygen consumption after 5 min of CPR did not differ significantly from prearrest values. Prearrest 1 and 2=the first and second sets of measurements made under prearrest conditions; CPR S=measurements made after approximately 5 rnin CPR.
Table I1 Oxygen contents and consumptions in control and propranolol groups. Values are means(SEM) Arteriul oxygen content (ml.dC')
Coronary sinus oxygen content (ml.df').
Prearrest 1 Prearrest 2 CPR 5 CPR 10 CPR IS CPR 20
13.4(0.S) 13.7(0.4) 14.9(0.5) 14.9(0.8) lS.l(O.9) 15.3(0.9)
6.4(0.4) 6.4(0.6) 7 . 3 I .2) 8.0(I .4) 9 . 3I .S) 9.9( I .6)
Prearrest I Prearrest 2 CPR 5 CPR 10 CPR IS CPR 20
l6.3(1.7) 16.3(1.5) 17.7(0.9) 17.1(0.9)
7.l( I .2) 7.l(l.2) 7.8i1.9j 12.2( 1.3) I I .8(0.9) 1 I .6(0.9)
16.5(0.8) 16.2( 1 .O)
Control group
(lid)
Arterial-cornnor?.sinus oxygen content d$ference
Myocurdiul u.ry,gen consumption (9% of Preurrest 1 )
7.0(0.4) 7.3(0.4) 7.4( I . I ) 6.9( 1.1) S.6(0.8) S.4(0.8)
100
(Ai.dr I I
Propranolol group (n=4J Y.2( 1.2) 9.2(0.9) 9.9( 1.7) 4.9( 1.1) 4.7(0.5) 4.6(0.8)
104.8(2.6) 80. I ( 12.2) 77.N 10.4) 100
100.6(4.6) 10Y.2(17.9) 53.8(13.3) 5 1 X(6.6) 49..5(5.1)
CPR=cardiopulmonary resuscitation; control group=animals not given propranolol during CPR; propranolol group=animals given propranolol I mg.kg-' through a left atrial catheter after the CPR 5 measurements; CPR 10. 15, and 20=measurements made after approximately 10, IS. and 20 min CPR, respectively; other abbreviations are defined in table 1. Propranolol administration significantly decreased myocardial oxygen consumption during CPR (p
79 1
Myocardial oxygen requirements during experimental cardiopulmonary resuscitation Roy V Ditchey, Yoichi Goto, and JoAnn Lindenfeld Objective: The aims were to determine myocardial oxygen requirements during cardiopulmonary resuscitation (CPR), and to test the hypothesis that endogenous catecholamines have a major effect on myocardial oxygen requirements in this setting. Methods: Myocardial oxygen consumption (MV02) was measured during 20 minutes of CPR in eight anaesthetised dogs. Coronary blood flow was maintained at prearrest levels using an external pump to provide a permissive level of oxygen delivery during ventricular fibrillation. Oxygen content was measured in arterial and coronary sinus blood samples under prearrest conditions and at 5 rnin intervals during CPR. Four dogs were given propranolol ( I mg.kg-’) following the 5 min measurements. Results: MVo2 averaged 108.7(SEM 12.8)%of the initial prearrest values after 5 min CPR (n=8). After 10 min CPR, MVo2 fell to 53.8(13.3)% of the initial prearrest values in the subset of animals given propranolol after the 5 rnin measurements (n=4), but remained at prearrest levels in untreated animals ( ~ ~ 0 . 0for5 an interactive effect between treatment and time). MVoz subsequently tended to decrease with time in untreated animals, but remained a high percentage of prearrest values throughout the 20 rnin period of CPR. Conclusions: These findings suggest that endogenous sympathetic stimulation of the fibrillating heart results in high myocardial oxygen requirements during CPR. Cardiovascular Research 1992;26:79 1-797
A
lthough available evidence suggests that closed chest cardiopulmonary resuscitation (CPR) cannot maintain coronary blood flow at prearrest the level of flow required to meet cardiac metabolic demands during CPR is unknown. In experimental preparations in which coronary flow is maintained with donor blood, myocardial oxygen consumption (MVor) typically is much lower during ventricular fibrillation than when the heart is beating and performing external work.” However, cardiovascular collapse is accompanied by intense activation of the sympathetic nervous system, and p adrenergic stimulation of the heart is known to increase myocardial oxygen requirements during ventricular fibrillation.” We hypothesised that endogenous neurohumoral sympathetic stimulation of the heart during CPR raises myocardial oxygen requirements to a high level during resuscitation from ventricular fibrillation. To test this hypothesis, we measured myocardial oxygen consumption before and during a 20 rnin period of CPR and observed the effects of pharmacological p adrenergic blockade. Because the low levels of coronary blood flow typically present during CPR might limit myocardial oxygen consumption by restricting oxygen delivery, we artificially maintained coronary flow at prearrest levels during CPR using an extracorporeal perfusion system.
Methods Studies were performed in eight mongrel dogs weighing an average of 27.6 kg (range 25.5 to 30.5 kg). The dogs were anaesthetised with morphine sulphate (2 mg.kg-’) and (Y chloralose (100 mg.kg-l), intubated, and mechanically
ventilated with a device that mixed room air with 100% oxygen by a venturi effect. The experiments were approved by the institutional animal care and use committee of the University of Vermont. Fluid filled catheters were inserted via a femoral artery and vein and positioned in the ascending aorta and right atrium, respectively. Catheter position was confirmed by appropriate pressure recordings, and by direct palpation. In addition, large diameter (approximately 4 mm internal diameter) plastic catheters with multiple side holes were inserted into the left external jugular vein via a side branch and into the descending thoracic aorta via the remaining femoral artery. A left lateral thoracotomy was performed, and the heart was suspended in a pericardial cradle. After injecting heparin, the coronary sinus was ligated near its insertion into the right atrium, and an additional large diameter catheter was passed through the chest wall, inserted through a small incision in the coronary sinus, and tied in place. This catheter was connected to the external jugular catheter via a rubber tube, diverting coronary sinus blood flow into the left external jugular vein (fig 1). The catheter in the coronary sinus had a side hole immediately proximal to the site of insertion to which a separate fluid filled catheter was attached and passed through the chest wall to allow measurement of coronary sinus pressure. The rubber tube comprising the extrathoracic portion of this drainage system had an internal diameter of approximately I cm, and was used to reduce resistance to blood flow (compared to the smaller diameter tubing required for intravascular insertion) and to provide a means of sampling coronary sinus blood by needle puncture. The large diameter catheter in the descending thoracic
Cardiology Unit, Department of Medicine, Medical Center Hospital of Vermont, Burlington, VT 05401, USA: R V Ditchey, Y Goto; Cardiology Division, University of Colorado Health Sciences Center, Denver, CO, USA: .ILindenfeld. Correspondence to Dr Ditchey.
792
Ditchey, Goto, Lindenfeld
Ilf J
From femoral artery
Large diameter
' E
_
t2g 2% 0 0
u=
_
-_ -
0-
-1s-
-
_.
~~
-
_-
-__ ~
~
--
~
_.__
-
___
-
-_
-____ . _
-
Figure 2 An example of the phasic relationships between
%?t
jf
To LAD
To circumflex
ascending aortic pressure and pressure and flow within the extracorporeal coronary perjiusion sytem under prearrest conditions. Instantaneous jlow increased during diastole and decreased during systole.
tj/ From coronary sinus
Figure 1 Extracorporeal coronary perfusion system (see text for
explanation). LADdefi anterior descending coronary artery: circumflex=circumflexcoronary artery.
aorta was connected to a Masterflex model 7523-00 peristaltic pump, through which arterial blood was withdrawn and forced sequentially through a filter and a heating coil submerged in a water bath maintained at approximately 37°C. The distal end of this tube was passed through the chest wall and then bifurcated via a Y connector into two limbs, each of which was connected distally to an adapter suitable for insertion into a coronary artery (fig 1). The proximal (approximately 2-3 cm) segments of the left anterior descending and circumflex coronary arteries were freed from surrounding fat and connective tissue, following which lignocaine (1.5 mg.kg-') was given intravenously. After filling the extracorporeal perfusion system with blood, the left anterior descending segment was ligated proximally and clamped distally. One of the distal limbs of the perfusion system was clamped and the tip of the other was inserted into the left anterior descending coronary artery near the site of ligation and tied in place. This limb of the coronary perfusion system had a side hole immediately proximal to the site of insertion to which a separate fluid filled catheter was attached and passed through the chest wall to allow measurement of coronary arterial pressure. This catheter, the catheter attached to the side of the coronary sinus drainage system, and the ascending aortic and right atrial catheters were connected to Could P23XL pressure transducers with zero reference set at the mid-chest level. Coronary blood flow then was restored by pumping femoral arterial blood into the left anterior descending coronary artery at a rate sufficient to raise the mean coronary arterial-coronary sinus pressure difference to a level slightly above the simultaneously recorded mean aortic-right atrial pressure difference. After a brief recovery period, the second limb of
the arterial perfusion system was inserted into the circumflex coronary artery in a similar manner, following which the rate at which arterial blood was pumped into the coronary arterial system was increased until the mean coronary arterialcoronary sinus pressure difference was restored to an appropriate level. Successful cannulation of each coronary artery and restoration of flow required approximately 1-5 min coronary arterial occlusion per artery. In some cases, proximal branches were ligated if they could not be included in the portion of the artery perfused artificially. The resolution of reactive hyperaemia (ie, a decrease in the flow required to maintain the same pressure in the first few minutes following relief of coronary occlusion) was taken as evidence that circulatory control mechanisms were not disrupted by instrumentation. The portion of the coronary perfusion catheter system leading from the femoral artery to the Masterflex pump contained a small segment of rubber tubing that allowed sampling of arterial blood by direct needle puncture. In some cases, phasic coronary flow signals were recorded using a cannulating electromagnetic flow probe inserted into the coronary perfusion system and connected to a Carolina Medical Electronics model FM501 square wave electromagnetic flow meter. Zero reference for the flow signal was determined by simultaneously turning off the perfusion pump and clamping the tube containing the electromagnetic flow probe. The Masterflex pump was calibrated in vitro, and quantitative levels of phasic coronary blood flow were estimated on that basis. Recorded flow signals were used for the purpose of illustration only. The filter in the coronary perfusion system contained an air buffer (fig 1) that effectively removed pulsations due to the peristaltic pump and allowed instantaneous coronary flow to vary during the cardiac cycle in response to changes in downstream resistance (fig 2). Finally, a fluid filled catheter was inserted into the left atrium via the left atrial appendage, and a bipolar electrode was sewn to the surface of the left ventricular free wall, following which the left atrial catheter and the electrical leads from the epicardial electrode were exteriorised through small incisions in the chest wall. This catheter was connected
Myocardial oxygen consumption during cardiopulmonary resuscitation
to a fifth Gould P23XL pressure transducer for measurements of left atrial pressure. The pericardial cradle was released and the chest was closed in layers and drained of air, following which the dogs were secured in a supine position and ventilatory parameters were adjusted so that arterial pH was between 7.35 and 7.45, and arterial Pcoz was between 4.65 and 6.00 kPa. Sodium bicarbonate solution was given intravenously when necessary. In addition, normal saline was given in a volume sufficient to raise mean left atrial pressure to approximately 5 mm Hg. This was done both as an arbitrary replacement for the extracorporeal blood pool and insensible fluid losses during instrumentation, and to reduce variability in myocardial oxygen consumption during ventricular fibrillation due to differences in cardiac di~tension.’~ The rate at which blood was pumped into the coronary arterial system was adjusted a final time to restore the mean coronary arterial-coronary sinus pressure difference to a level slightly above the mean aortic-right atrial pressure difference present under these conditions, and then was held constant for the remainder of the study. Aortic, right atrial, coronary arterial, coronary sinus, and left atrial pressures and coronary flow signals were recorded under prearrest conditions (ie, prior to induction of ventricular fibrillation), and arterial and coronary sinus blood samples were obtained in heparinised glass syringes for determination of oxygen content. The arterial blood sample was also used for measurement of packed cell volume. In addition, a separate coronary arterial blood sample was obtained for determination of prearrest adrenaline and noradrenaline concentrations and placed in a tube containing EGTA and glutathione. After a brief interval, prearrest pressure measurements were repeated, and a second set of arterial and coronary sinus blood samples was obtained for determination of oxygen contents to assess reproducibility. Blood samples were stored on ice until the end of the study. In five dogs, heart rate at the time of the first set of prearrest measurements was calculated from an electrocardiogram or the phasic coronary flow signal (electrocardiograms and phasic pressures and flow signals were not recorded in the remaining three dogs). Ventricular fibrillation then was induced by applying a low voltage alternating current to the epicardial electrode. The current was discontinued, and anteroposterior chest compression was initiated using a pneumatic chest compression device (Michigan Instruments Life Aid model X1004 cardiopulmonary resuscitator). The chest was compressed at a rate of 60min-’ with a compression duration equal to approximately 50% of each cycle, and a compression force sufficient to produce between 2.0 and 2.5 inches of posterior sternal displacement, as estimated by a displacement transducer connected to the compression piston. The release phase of the compression cycle was prolonged by approximately 0.5 s every fifth compression to allow lung inflation to a peak inspiratory pressure of 20 cm of water by a synchronised pressure limited ventilator. Chest compression force was adjusted, if necessary, during the remainder of the study in order to maintain a relatively constant extent of sternal displacement. Higher levels of chest compression force and greater sternal displacement were avoided to minimise the possibility of dislodging the coronary perfusion system from the coronary arteries. CPR was continued for a total of 20 min, during which combined left anterior descending and circumflex coronary blood flow was maintained at the same level present during prearrest measurements. This provided a constant and relatively high
793
level of coronary blood flow during CPR, such that myocardial oxygen consumption could vary (ie, either increase or decrease) in response to metabolic requirements, rather than be limited by inadequate coronary blood flow and restricted oxygen delivery. Pressure and flow recordings were repeated 5 , 10, 15, and 20 min after the onset of ventricular fibrillation and CPR. Arterial and coronary sinus blood samples for determination of oxygen content and arterial samples for blood gas analysis, packed cell volume, and determination of plasma adrenaline and noradrenaline concentrations were obtained at each interval. Four dogs were studied without further intervention. In the remaining four dogs, propranolol (1 mg-kg-’) was injected through the left atrial catheter after recording pressures and obtaining arterial and coronary sinus blood samples at the 5 min CPR interval. Arterial and coronary sinus blood oxygen contents were measured using a Lex-02-Con oxygen content analyser. In most cases, the oxygen content in each sample was measured twice and the results averaged. The average difference between duplicate measurements in 67 samples was less than 2% of the mean value. The plasma from each blood sample drawn for measurement of catecholamine concentrations was separated by centrifugation and frozen at -70°C. Plasma adrenaline and noradrenaline concentrations in each sample were determined at a later date using high pressure liquid chromatography. Pressures, flows, and electrocardiograms were recorded with a Hewlett Packard model 7848A eight channel recorder. Mean pressures were averaged over at least two respiratory cycles for prearrest measurements and over 20 chest compression cycles for each set of CPR measurements. Myocardial oxygen consumption in the distribution supplied by the artificial perfusion system was calculated as the product of coronary blood flow and the measured coronary arterial-coronary sinus oxygen content difference. Because coronary blood flow was held contant during each study, the flow term in the calculation was eliminated by expressing MVoz as a percentage of the first prearrest value. This was done to avoid possible errors in flow calibration and also served to normalise for baseline differences in MVo2 between dogs. Data analysis Differences between prearrest and CPR values for the study variables of interest were tested for statistical significance by repeated measures analysis of variance. The variables analysed were MVoz, plasma adrenaline and noradrenaline concentrations, and arterial pH and PCOZ.In the case of MVoz, two separate analyses were performed. First, to test the hypothesis that MVoz during CPR is a high percentage of prearrest values, an analysis of variance was restricted to data obtained under prearrest conditions and after 5 min of CPR when conditions were comparable in all eight dogs (ie, prior to propranolol administration). Second, to assess the effect of p adrenergic blockade, an analysis of variance was performed considering the control and propranolol treated dogs as separate subsets for all six prearrest and CPR conditions. For all other variables, the data are reported separately to show the general comparability of the two study groups, but statistical analyses for the variables of interest were performed using only combined data. Multiple comparison testing of the differences between measurements made under the first prearrest condition and each subsequent condition, if indicated, was performed using Dunnett’s test. Data are reported as the mean(SEM). A p value ~ 0 . 0 5was considered significant.
794
Ditchey, Goto,
Lindenfeld
Table I Oxygen contents and consumption under prearrest conditions and after 5 min cardiopulmonary resuscitution (CPR). Vrrlries are means(SEM), n=8
Prearrest 1 Prearrest 2 CPR 5
Arterial olvgen content (ml.df')
Coronarv sinus oxygen content (m/.df')'
Arteritrl-coronary sinus oxyRen content difference
M.vocardio1 oxygen consumption (% of Prrorrest 1 )
14.9( I .O) 1S.O(O.9) 16.3(0.7)
6.7(0.6) 6.8(0.6) 7.7( I .O)
R.l(O.7) 8.2(0.6) 8.7( 1 . I )
I00 102.7(2.6) I08.7( 12.8)
Myocardial oxygen consumption after 5 min of CPR did not differ significantly from prearrest values. Prearrest 1 and 2=the first and second sets of measurements made under prearrest conditions; CPR S=measurements made after approximately 5 rnin CPR.
Table I1 Oxygen contents and consumptions in control and propranolol groups. Values are means(SEM) Arteriul oxygen content (ml.dC')
Coronary sinus oxygen content (ml.df').
Prearrest 1 Prearrest 2 CPR 5 CPR 10 CPR IS CPR 20
13.4(0.S) 13.7(0.4) 14.9(0.5) 14.9(0.8) lS.l(O.9) 15.3(0.9)
6.4(0.4) 6.4(0.6) 7 . 3 I .2) 8.0(I .4) 9 . 3I .S) 9.9( I .6)
Prearrest I Prearrest 2 CPR 5 CPR 10 CPR IS CPR 20
l6.3(1.7) 16.3(1.5) 17.7(0.9) 17.1(0.9)
7.l( I .2) 7.l(l.2) 7.8i1.9j 12.2( 1.3) I I .8(0.9) 1 I .6(0.9)
16.5(0.8) 16.2( 1 .O)
Control group
(lid)
Arterial-cornnor?.sinus oxygen content d$ference
Myocurdiul u.ry,gen consumption (9% of Preurrest 1 )
7.0(0.4) 7.3(0.4) 7.4( I . I ) 6.9( 1.1) S.6(0.8) S.4(0.8)
100
(Ai.dr I I
Propranolol group (n=4J Y.2( 1.2) 9.2(0.9) 9.9( 1.7) 4.9( 1.1) 4.7(0.5) 4.6(0.8)
104.8(2.6) 80. I ( 12.2) 77.N 10.4) 100
100.6(4.6) 10Y.2(17.9) 53.8(13.3) 5 1 X(6.6) 49..5(5.1)
CPR=cardiopulmonary resuscitation; control group=animals not given propranolol during CPR; propranolol group=animals given propranolol I mg.kg-' through a left atrial catheter after the CPR 5 measurements; CPR 10. 15, and 20=measurements made after approximately 10, IS. and 20 min CPR, respectively; other abbreviations are defined in table 1. Propranolol administration significantly decreased myocardial oxygen consumption during CPR (p
