AMERICAN JOURNAL OF PHYSIOIBGY VoI. 230, No. 1, January 1976. Printed
in
U.S.A.
Intramyocardial distribution of blood flow in hemorrhagic shock in anesthetized dogs EDWIN L, CARLSON, SAMUEL Cardiovascular Research Institute San Francisco, California Ml43
L, SELINGER, and Department
subendocardial blood flow; myocardial edema; alpha-adrenergic blockade; mannitul; phenoxybenzamine; myocardial oxygen consumption; myocardial oxygen extraction
myocardialfunction and in causing irreversibility of hemorrhagic shock is not clear. Evidence that the heart is partly responsible for the irreversible changes includes the finding of subendocardial hemorrhages, necrosis and zonal contraction lesions (18, 19, 29>, release of myocardial enzymes (34), shift of left ventricular function curves to the right after prolonged hemorrhage (3, 8, 15, 16, 21, 34, 37, 39), and decreased cardiac response to sympathetic stimulation (14) and the demonstration that toxic substances released in shock can depress the myocardium and cause cardiac failure (25). These changes could be due TV myocardial ischemia since in shock there is marked depression of coronary blood flow (3, 11,13,16-18,21, 28,33,4(I) and increased myocardial oxygen extraction (18, 19, 21, 26, 33). Furthermore, hyperbaric oxygen during the hypotensive phase prevents the subendocardial hemorrhages and increases lactate extraction (32) and, finally, increasing the coronary blood flow by infusing dipyridamole (3) or mainPART
myocardial
PLAYED
JULIEN I. E. HOFFMAN of California,
taining or raising aortic or coronary arterial pressure (15, 37) will maintain or restore normal myocardial function, There are, however, many reasons to question the importance of myocardial deterioration in causing irreversibility. No one has shown that the subendocardial lesions are sufficient to cause severe myocardial dysfunction. The decrease of coronary blood flow does not of itself indicate myocardial ischemia since myocardial work and oxygen uptake are usually reduced in shock (11, 21, 28, 33) and coronary sinus oxygen tensions remain above the level at which ischemia is usually seen (26, 38). There is usually no biochemical evidence for anaerobic myocardial metabolism in shock (9,16, 21,27, 28, 40). Finally, in the dog irreversibility is noted when hemoconcentration occurs (23) and this suggests that extravascular pooling is a major cause of irreversibility. In keeping with this view, Rothe and Selkurt (36) observed that massive fluid infusions after the shed blood had been returned could prevent cardiovascular decay for many hours. Even the undoubted reductions of ventricular function tend to occur late and have not been proved to be the major cause of the irreversibility. One major difficulty in interpreting results that argue against myocardial ischemia in shock is that they do not allow for possible maldistribution of blood flow within the myocardium. Ischemia of some regions plus hyperemia of others would not only make total coronary flows hard to assessbut could also produce opposing changes in regional coronary venous oxygen tension and lactate concentrations that could sum to produce normal values, That maldistribution of flow might occur is suggested not only by subendocardial lesions but also by changes in the pattern of phasic coronary blood flow. Thus Entman et al. (13) and Granata et al. (17) have pointed out that during hypotension arterial blood pressure is stable but there is an increasing proportion of systolic blood flow in the left coronary artery; this change has in other studies been associated with subendocardial ischemia (6). Since maldistribution could be so important in drawing conclusions in this field and since there are no reported data on it in hemorrhagic shock, we decided to study the effect of hemorrhagic shock on the total coronary blood flow and its regional distribution. Because in shock there is sympathetic stimulation and eventual tissue edema, we pretreated some dogs with an alpha-adrenergic blocker, phenoxybenzamine (PBZ), and also in some infused mannitol late in hemorrhage.
CARLSON, EDWIN L., SAMUEL L. SELINGERJOSEPH UTLEY, AND JULIEN II. E. HOFFMAN. Intramyocardial distribution of blood flow in hemorrhagic shock in anesthetized dogs. Am. J. Physiol. 230(l): 41-49. 1976. -In 34 anesthetized, open-chest dogs aortic blood pressure was kept at 35-40 mmHg for 3 h to determine if maldistribution of coronary blood flow (CBF) could contribute to the irreversibility of hemorrhagic shock. Six dogs were pretreated with phenoxybenzamine (PBZ) and 11 dogs (3 with PBZ) received hypertonic mannitol infusions in late hemorrhage. Changes of heart rate, cardiac output, and peripheral resistance were similar to those described by others. In untreated dogs total and left ventricular CBF fell, as did coronary vascular resistance. However, minimal coronary resistance after transient &hernia rose progressively and the ratio of subendocardial:subepicardial flow fell, as did the percentage of diastolic coronary flow. Mannitol infusion returned CBF and steady-state and minimal postischemic coronary resistance to control values and also returned to normal the increased myocardial water content found in late hemorrhage. Phenoxybenzamine delayed but did not prevent the rise of coronary vascular resistance or decreased subendocardial flow. These studies suggest that there may be subendocardial ischemia, possibly due to myocardial edema, in hemorrhagic shock.
THE
JOSEPH UTLEY, AND of Pediatrics, University
BY IMPAIRED
ischemia
41
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
42 MATERIALS
CARLSON AND
%lETHODS
Thirty-four mongrel dogs weighing 25-35 kg were used. Twenty dogs were anesthetized with pentobarbital sodium (30 mglkg iv) and the rest received a mixture of ar-chloralose (50 mg/kg) and urethan (250 mgkg) in propylene glycol intravenously. We noted no difference between results obtained with these two anesthetics. The dogs were divided into two groups. Twenty-eight dogs were classified as untreated and six dogs were given 5 mg PBZ/kg iv l-2 h before bleeding; these six had all been given pentobarbital sodium. Before the PBZ was given to these six dogs they were given intravenous dosesof 0. 05-l mg methoxamine until a moderate rise of arterial blood pressure was produced; when the same dose was given 1 h after injection of PBZ there was no response, thus indicating adequate alpha-adrenergic blockade. After the dogs were anesthetized and intubated they were ventilated with room air by a volume respirator. Pressures in the left ventricle and supravalvar aorta were measured by end-hole catheters inserted through the right femoral and brachial arteries, except for eight untreated dogs in which ventricular pressure w‘as measured by a needle through the ventricular wall, In these eight untreated dogs, one artery was used for a catheter with its tip in the supravalvar aorta for later infu .sion of mannitol. A large polyvinyl cannula was placed in the left femoral artery for bleeding into a reservoir and a catheter in the left brachial artery was used to obtain reference samples during the microsphere injection (6, 10). One femoral vein was used for intravenous injection of drugs; the other was cannulated to allow return of shed blood from the reservoir at appropriate times. The left thorax was opened between the fourth and fifth ribs, the pericardium incised, and the heart supported in a pericardial cradle. Catheter positions were checked by palpation. Two end-hole catheters were inserted into the left atrium through its appendage, one to measure pressure and the other to inject microspheres. The coronary sinus was cannulated to obtain blood samples for oxygen contents (measured by Lex-O&on, Lexington Instrument Co.), to measure coronary sinus pressure and initially to check for shunting of microspheres (2). The cir cumflex coronary artery was cleared of fat and an electromagnetic flow transducer was placed around it and connected to the flowmeter (Omnicraft, Inc., Oakland, Calif.). Distal to the flow transducer a ligature was placed on the coronary artery to obtain occlusion zeros and to test reactive hyperemia (6); care was taken that no vessels were between the occluder and the transducer. An electrocardiogram was also recorded. Left ventricular and aortic pressures were measured with Micron strain gauges; left atria1 and coronary sinus pressures were measured with Statham P23Db strain gauges. All tracings were recorded on a Beckman Dynograph pen recorder. Total and regional myocardial blood flows were measured by injection of L-2 million 15pm-diameter spheres labeled with 1251and g-pm-diameter spheres labeled with 141Ce 85Sr and 46Sc (3M Company, St. Paul, Minn.). ThHre is little difference in the ability of spheres
ET
AL.
9 pm and 15 pm in diameter to measure regional flows within the myocardium (41), but we always injected the 15-pm spheres in the control state. The microspheres were suspended in saline to which Tween 80 had been added to a final concentration of 0.5% to reduce aggregation. The vial with the‘microspheres was ultrasonicated to break up clumps and placed on a vibrator just before injection to ensure uniform suspension. Then the microspheres were flushed in with E-20 ml of warmed saline over 15-25 s. After the injection the vial and connecting tubing were counted for residual radioactivity; usually none was found. Arterial reference samples were withdrawn at a steady rate of 9-11 ml/min for 1.5 min. At the end of the experiment, the dog was killed by injection of pentubarbital sodium. The heart was cut into left ventricular free wall, septum, and right ventricular free wall, which were cleared of fat, large vessels, and valves and then weighed. The tissues were placed in 10% Formalin for about 1 wk, then reweighed and cut into layers. The left ventricular free wall was divided inti endocardial, middle, and epicardial layers of about equal mass; the septum was divided into right and left halves; and the right ventricle was left as one layer. The individual sections were then cut into small pieces, placed in vials, and their gamma emission was counted for 2-5 min in a multichannel pulse-height analyzer with variable regions of interest (Nuclear-Chicago Corp.). The counts per minute of the tissues and reference blood samples were put on computer cards that were then run on an IBM 360 computer for determination of flows, flows per gram, and cardiac output (2, 6, 10). The percentage of total counts of each nuclide in the region of interest was related to the height of tissue in the vial and corrections for the efficiency of counting blood and tissue were included in the computer program. Flow to any cardiac region was computed as C;F,IC,, where C, and C, are counts per minute in the cardiac region and the arterial reference sample, respectively, and F, is the rate of withdrawal of the reference sample (6, 10). For any region of the heart the flows in component pieces were added and that total divided by the total weight of those pieces gave the average flow per gram for that region. Cardiac output was computed as Ci l F,/C,, where Ci is the total counts per minuk of the spheres injected into the left atrium (2). In three dogs the percentage of the total cardiac microspheres appearing in the coronary sinus was measured by withdrawal of a second reference sample from the coronary sinus (2). The percentage of microspheres not trapped was less than 1% of those in the myocardium in control and shock states. We found considerable variability in the sensitivity of the electromagnetic flow transducer, probably due to changes in hematocrit, changes in the boundary between the transducer and the blood within the vessel, and perhaps variable asymmetry of flow. Therefore we related the mean deflection in the coronary flow tracing to the left ventricular coronary flow measured by microspheres to derive a calibration factor for the flow tracing. From this it was possible to determine stroke coronary flow, the proportion of flow in systole and diastole, and the coronary vascular resistance. For each micro-
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
REGIONAL
CORONARY
BLOOD
FLOW
IN
HEMORRHAGIC
43
SHOCK
sphere injection there was a new calibration factir. We believe that this procedure is more accurate than terminal calibration of the flowmeter in acute studies because of the demonstrated accuracy of coronary flows measured by microspheres (41), Coronary vascular resistance calculations were made before and after lo-15 s of coronary occlusion in order to evaluate steady-state and postischemic hyperemic resistances. To determine if the postischemic vasodilator response was maximal, the coronary artery was occluded on some occasions for 45 s during the control state. In none of the tests did peak flow after occlusion for 45 s differ from that obtained after 10-15 s. Three different coronary vascular resistances were measured, both in the steady state and during maximal reactive hyperemia. Mean coronary resistance was derived by dividing the mean pressure difference between aortic and coronary sinus pressures by coronary flow in milliliters per minute per 100 g. Second, late diastolic coronary vascular resistance was measured with flows taken from the phasic flow tracing just before the rise in left ventricular pressure and the corresponding values in the pressure tracings (17). Because these pressures and flows are single points and thus subject to some variability, we also calculated mean diastolic coronary vascular resistance as mean diastolic coronary flow divided into the mean pressure difference between the aortic and coronary sinus pressures in diastole. These latter two resistances should more accurately depict the resistance in the coronary bed when extravascular forces are minimal. The amount of myocardial water was determined in two dogs during late hemorrhage and in two others during late hemorrhage after mannitol infusions. In these, at the end of the experiment, full-thickness samples of left ventricular free wall were taken. The samples, weighing about I g, were placed in weighed vials, reweighed, and dried in an oven at 56*C for 24 h. They were then placed in a desiccator and weighed daily until their weights were stable+ The area between the aortic and left ventricular pressures in diastole (DPTI) and the area under the left ventricular pressure curve in systole (SPTI) were measured by planimetry (6). The DPTI multiplied by heart rate is an index of the pressure available for producing subendocardial flow in diastole and is proportional to subendocardial flow if resistance is constant. Therefore, DPTI is also an index of potential subendocardial oxygen supply, which can be quantitated by multiplying DTPI by arterial oxygen content (5), Experimental protocol. Once the preparation had been completed, control measurements were made of blood gases and pH, hematocrits, and arterial and coronary sinus oxygen contents. Occlusion zeros were obtained for the flow tracing and the hyperemic flow response and when control flows had returned the first batch of microspheres was injected. Then the dogs were bled over 30 min to a mean arterial pressure of 35-40 mmHg with a pressure reservoir that allowed the dog to maintain a constant pressure and automatically transfuse itself late in hemorrhage. After the dog had first stabilized at the low arterial pressure, the second set of
microspheres was injected; this was termed early hemorrhage (EH). The microsphere injection was preceded by occlusion zeros for the flow tracing and the hyperemic flow response and measurement of blood gases and pH, hematocrit, and oxygen contents. The dog was then not interfered with until just before autotransfusion, as indicated by.a change in the phasic flow tracing to show late systolic backflow (17). Occlusion zeros and postischemic hyperemic responses were obtained, blood samples were taken, and microspheres were injected. This stage was termed late hemorrhage (LH). Autotransfusion began soon after this time, and when 25% of the shed blood had been taken up the rest was slowly infused. When blood pressure, heart rate, and coronary flow were stable, blood samples were drawn and microspheres injected; this phase was termed the posttransfusion state (PT). The dog was then followed until definite signs of cardiovascular decay (CD) developed-that is, progressive decrease of blood pressure and coronary flow. Because of the hypotension at this time, microspheres were injected without obtaining blood samples and occlusion zeros. All dogs were kept at an arterial blood pH between 7,2 and 7.3 by varying respiratory rate and giving sodium bicarbonate. Because there were five states and only four different nuclides, every untreated dog was given the control injection and the three other sets of microspheres were given at three randomly chosen states out of the remaining four states of hemorrhage. Dogs given PBZ were not studied to the posttransfusion stage. Eight untreated and three PBZ-treated dogs were allowed to progress to late hemorrhage, appropriate measurements being made. Then 20% mannitol was infused into the root of the aorta at 5 ml/min for 15-20 min. The tubing to the reservoir was open so that aortic pressure did not rise. After the infusion another set of measurements was made; the dogs were then killed and the hearts analyzed as above. RESULTS
Times, volumes, and hematocrits. Table 1 shows that the time taken to reach EH was similar in untreated and PBZ-treated dogs, but the time to LH was signifi1. Times, volumes bled, and hematocrits at different stages of shock TABLE
Control
EH
LH ----
Time, min Untreated Untreated + mannitol Phenoxybenzamine
50*5.3(13) 48*2.3(7)3 59?2.9(6)$
Volume bled, ml/kg Untreated Phenoxybenzamine
38*22.2(20) 24?2.9(6)*
43*2.1(23) 27~2.7(5)*
Hematmrit, % Untreated Phenoxybenzamine
43*1.6(12) 362 1.2(6)*
39*2.2(16) 32k 1.6(6)t
115*8.0(19) 124? 13.6(7)$ 1732 18.4(6)*
Values are mean 2 1. SE, with numbers of observations in parentheses. All comparisons are by unpaired-# test with corresponding values for untreated doga: l P < 0.05; t 0.05 < P < 0.10; $ 0.10
in
U.S.A.
Intramyocardial distribution of blood flow in hemorrhagic shock in anesthetized dogs EDWIN L, CARLSON, SAMUEL Cardiovascular Research Institute San Francisco, California Ml43
L, SELINGER, and Department
subendocardial blood flow; myocardial edema; alpha-adrenergic blockade; mannitul; phenoxybenzamine; myocardial oxygen consumption; myocardial oxygen extraction
myocardialfunction and in causing irreversibility of hemorrhagic shock is not clear. Evidence that the heart is partly responsible for the irreversible changes includes the finding of subendocardial hemorrhages, necrosis and zonal contraction lesions (18, 19, 29>, release of myocardial enzymes (34), shift of left ventricular function curves to the right after prolonged hemorrhage (3, 8, 15, 16, 21, 34, 37, 39), and decreased cardiac response to sympathetic stimulation (14) and the demonstration that toxic substances released in shock can depress the myocardium and cause cardiac failure (25). These changes could be due TV myocardial ischemia since in shock there is marked depression of coronary blood flow (3, 11,13,16-18,21, 28,33,4(I) and increased myocardial oxygen extraction (18, 19, 21, 26, 33). Furthermore, hyperbaric oxygen during the hypotensive phase prevents the subendocardial hemorrhages and increases lactate extraction (32) and, finally, increasing the coronary blood flow by infusing dipyridamole (3) or mainPART
myocardial
PLAYED
JULIEN I. E. HOFFMAN of California,
taining or raising aortic or coronary arterial pressure (15, 37) will maintain or restore normal myocardial function, There are, however, many reasons to question the importance of myocardial deterioration in causing irreversibility. No one has shown that the subendocardial lesions are sufficient to cause severe myocardial dysfunction. The decrease of coronary blood flow does not of itself indicate myocardial ischemia since myocardial work and oxygen uptake are usually reduced in shock (11, 21, 28, 33) and coronary sinus oxygen tensions remain above the level at which ischemia is usually seen (26, 38). There is usually no biochemical evidence for anaerobic myocardial metabolism in shock (9,16, 21,27, 28, 40). Finally, in the dog irreversibility is noted when hemoconcentration occurs (23) and this suggests that extravascular pooling is a major cause of irreversibility. In keeping with this view, Rothe and Selkurt (36) observed that massive fluid infusions after the shed blood had been returned could prevent cardiovascular decay for many hours. Even the undoubted reductions of ventricular function tend to occur late and have not been proved to be the major cause of the irreversibility. One major difficulty in interpreting results that argue against myocardial ischemia in shock is that they do not allow for possible maldistribution of blood flow within the myocardium. Ischemia of some regions plus hyperemia of others would not only make total coronary flows hard to assessbut could also produce opposing changes in regional coronary venous oxygen tension and lactate concentrations that could sum to produce normal values, That maldistribution of flow might occur is suggested not only by subendocardial lesions but also by changes in the pattern of phasic coronary blood flow. Thus Entman et al. (13) and Granata et al. (17) have pointed out that during hypotension arterial blood pressure is stable but there is an increasing proportion of systolic blood flow in the left coronary artery; this change has in other studies been associated with subendocardial ischemia (6). Since maldistribution could be so important in drawing conclusions in this field and since there are no reported data on it in hemorrhagic shock, we decided to study the effect of hemorrhagic shock on the total coronary blood flow and its regional distribution. Because in shock there is sympathetic stimulation and eventual tissue edema, we pretreated some dogs with an alpha-adrenergic blocker, phenoxybenzamine (PBZ), and also in some infused mannitol late in hemorrhage.
CARLSON, EDWIN L., SAMUEL L. SELINGERJOSEPH UTLEY, AND JULIEN II. E. HOFFMAN. Intramyocardial distribution of blood flow in hemorrhagic shock in anesthetized dogs. Am. J. Physiol. 230(l): 41-49. 1976. -In 34 anesthetized, open-chest dogs aortic blood pressure was kept at 35-40 mmHg for 3 h to determine if maldistribution of coronary blood flow (CBF) could contribute to the irreversibility of hemorrhagic shock. Six dogs were pretreated with phenoxybenzamine (PBZ) and 11 dogs (3 with PBZ) received hypertonic mannitol infusions in late hemorrhage. Changes of heart rate, cardiac output, and peripheral resistance were similar to those described by others. In untreated dogs total and left ventricular CBF fell, as did coronary vascular resistance. However, minimal coronary resistance after transient &hernia rose progressively and the ratio of subendocardial:subepicardial flow fell, as did the percentage of diastolic coronary flow. Mannitol infusion returned CBF and steady-state and minimal postischemic coronary resistance to control values and also returned to normal the increased myocardial water content found in late hemorrhage. Phenoxybenzamine delayed but did not prevent the rise of coronary vascular resistance or decreased subendocardial flow. These studies suggest that there may be subendocardial ischemia, possibly due to myocardial edema, in hemorrhagic shock.
THE
JOSEPH UTLEY, AND of Pediatrics, University
BY IMPAIRED
ischemia
41
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
42 MATERIALS
CARLSON AND
%lETHODS
Thirty-four mongrel dogs weighing 25-35 kg were used. Twenty dogs were anesthetized with pentobarbital sodium (30 mglkg iv) and the rest received a mixture of ar-chloralose (50 mg/kg) and urethan (250 mgkg) in propylene glycol intravenously. We noted no difference between results obtained with these two anesthetics. The dogs were divided into two groups. Twenty-eight dogs were classified as untreated and six dogs were given 5 mg PBZ/kg iv l-2 h before bleeding; these six had all been given pentobarbital sodium. Before the PBZ was given to these six dogs they were given intravenous dosesof 0. 05-l mg methoxamine until a moderate rise of arterial blood pressure was produced; when the same dose was given 1 h after injection of PBZ there was no response, thus indicating adequate alpha-adrenergic blockade. After the dogs were anesthetized and intubated they were ventilated with room air by a volume respirator. Pressures in the left ventricle and supravalvar aorta were measured by end-hole catheters inserted through the right femoral and brachial arteries, except for eight untreated dogs in which ventricular pressure w‘as measured by a needle through the ventricular wall, In these eight untreated dogs, one artery was used for a catheter with its tip in the supravalvar aorta for later infu .sion of mannitol. A large polyvinyl cannula was placed in the left femoral artery for bleeding into a reservoir and a catheter in the left brachial artery was used to obtain reference samples during the microsphere injection (6, 10). One femoral vein was used for intravenous injection of drugs; the other was cannulated to allow return of shed blood from the reservoir at appropriate times. The left thorax was opened between the fourth and fifth ribs, the pericardium incised, and the heart supported in a pericardial cradle. Catheter positions were checked by palpation. Two end-hole catheters were inserted into the left atrium through its appendage, one to measure pressure and the other to inject microspheres. The coronary sinus was cannulated to obtain blood samples for oxygen contents (measured by Lex-O&on, Lexington Instrument Co.), to measure coronary sinus pressure and initially to check for shunting of microspheres (2). The cir cumflex coronary artery was cleared of fat and an electromagnetic flow transducer was placed around it and connected to the flowmeter (Omnicraft, Inc., Oakland, Calif.). Distal to the flow transducer a ligature was placed on the coronary artery to obtain occlusion zeros and to test reactive hyperemia (6); care was taken that no vessels were between the occluder and the transducer. An electrocardiogram was also recorded. Left ventricular and aortic pressures were measured with Micron strain gauges; left atria1 and coronary sinus pressures were measured with Statham P23Db strain gauges. All tracings were recorded on a Beckman Dynograph pen recorder. Total and regional myocardial blood flows were measured by injection of L-2 million 15pm-diameter spheres labeled with 1251and g-pm-diameter spheres labeled with 141Ce 85Sr and 46Sc (3M Company, St. Paul, Minn.). ThHre is little difference in the ability of spheres
ET
AL.
9 pm and 15 pm in diameter to measure regional flows within the myocardium (41), but we always injected the 15-pm spheres in the control state. The microspheres were suspended in saline to which Tween 80 had been added to a final concentration of 0.5% to reduce aggregation. The vial with the‘microspheres was ultrasonicated to break up clumps and placed on a vibrator just before injection to ensure uniform suspension. Then the microspheres were flushed in with E-20 ml of warmed saline over 15-25 s. After the injection the vial and connecting tubing were counted for residual radioactivity; usually none was found. Arterial reference samples were withdrawn at a steady rate of 9-11 ml/min for 1.5 min. At the end of the experiment, the dog was killed by injection of pentubarbital sodium. The heart was cut into left ventricular free wall, septum, and right ventricular free wall, which were cleared of fat, large vessels, and valves and then weighed. The tissues were placed in 10% Formalin for about 1 wk, then reweighed and cut into layers. The left ventricular free wall was divided inti endocardial, middle, and epicardial layers of about equal mass; the septum was divided into right and left halves; and the right ventricle was left as one layer. The individual sections were then cut into small pieces, placed in vials, and their gamma emission was counted for 2-5 min in a multichannel pulse-height analyzer with variable regions of interest (Nuclear-Chicago Corp.). The counts per minute of the tissues and reference blood samples were put on computer cards that were then run on an IBM 360 computer for determination of flows, flows per gram, and cardiac output (2, 6, 10). The percentage of total counts of each nuclide in the region of interest was related to the height of tissue in the vial and corrections for the efficiency of counting blood and tissue were included in the computer program. Flow to any cardiac region was computed as C;F,IC,, where C, and C, are counts per minute in the cardiac region and the arterial reference sample, respectively, and F, is the rate of withdrawal of the reference sample (6, 10). For any region of the heart the flows in component pieces were added and that total divided by the total weight of those pieces gave the average flow per gram for that region. Cardiac output was computed as Ci l F,/C,, where Ci is the total counts per minuk of the spheres injected into the left atrium (2). In three dogs the percentage of the total cardiac microspheres appearing in the coronary sinus was measured by withdrawal of a second reference sample from the coronary sinus (2). The percentage of microspheres not trapped was less than 1% of those in the myocardium in control and shock states. We found considerable variability in the sensitivity of the electromagnetic flow transducer, probably due to changes in hematocrit, changes in the boundary between the transducer and the blood within the vessel, and perhaps variable asymmetry of flow. Therefore we related the mean deflection in the coronary flow tracing to the left ventricular coronary flow measured by microspheres to derive a calibration factor for the flow tracing. From this it was possible to determine stroke coronary flow, the proportion of flow in systole and diastole, and the coronary vascular resistance. For each micro-
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
REGIONAL
CORONARY
BLOOD
FLOW
IN
HEMORRHAGIC
43
SHOCK
sphere injection there was a new calibration factir. We believe that this procedure is more accurate than terminal calibration of the flowmeter in acute studies because of the demonstrated accuracy of coronary flows measured by microspheres (41), Coronary vascular resistance calculations were made before and after lo-15 s of coronary occlusion in order to evaluate steady-state and postischemic hyperemic resistances. To determine if the postischemic vasodilator response was maximal, the coronary artery was occluded on some occasions for 45 s during the control state. In none of the tests did peak flow after occlusion for 45 s differ from that obtained after 10-15 s. Three different coronary vascular resistances were measured, both in the steady state and during maximal reactive hyperemia. Mean coronary resistance was derived by dividing the mean pressure difference between aortic and coronary sinus pressures by coronary flow in milliliters per minute per 100 g. Second, late diastolic coronary vascular resistance was measured with flows taken from the phasic flow tracing just before the rise in left ventricular pressure and the corresponding values in the pressure tracings (17). Because these pressures and flows are single points and thus subject to some variability, we also calculated mean diastolic coronary vascular resistance as mean diastolic coronary flow divided into the mean pressure difference between the aortic and coronary sinus pressures in diastole. These latter two resistances should more accurately depict the resistance in the coronary bed when extravascular forces are minimal. The amount of myocardial water was determined in two dogs during late hemorrhage and in two others during late hemorrhage after mannitol infusions. In these, at the end of the experiment, full-thickness samples of left ventricular free wall were taken. The samples, weighing about I g, were placed in weighed vials, reweighed, and dried in an oven at 56*C for 24 h. They were then placed in a desiccator and weighed daily until their weights were stable+ The area between the aortic and left ventricular pressures in diastole (DPTI) and the area under the left ventricular pressure curve in systole (SPTI) were measured by planimetry (6). The DPTI multiplied by heart rate is an index of the pressure available for producing subendocardial flow in diastole and is proportional to subendocardial flow if resistance is constant. Therefore, DPTI is also an index of potential subendocardial oxygen supply, which can be quantitated by multiplying DTPI by arterial oxygen content (5), Experimental protocol. Once the preparation had been completed, control measurements were made of blood gases and pH, hematocrits, and arterial and coronary sinus oxygen contents. Occlusion zeros were obtained for the flow tracing and the hyperemic flow response and when control flows had returned the first batch of microspheres was injected. Then the dogs were bled over 30 min to a mean arterial pressure of 35-40 mmHg with a pressure reservoir that allowed the dog to maintain a constant pressure and automatically transfuse itself late in hemorrhage. After the dog had first stabilized at the low arterial pressure, the second set of
microspheres was injected; this was termed early hemorrhage (EH). The microsphere injection was preceded by occlusion zeros for the flow tracing and the hyperemic flow response and measurement of blood gases and pH, hematocrit, and oxygen contents. The dog was then not interfered with until just before autotransfusion, as indicated by.a change in the phasic flow tracing to show late systolic backflow (17). Occlusion zeros and postischemic hyperemic responses were obtained, blood samples were taken, and microspheres were injected. This stage was termed late hemorrhage (LH). Autotransfusion began soon after this time, and when 25% of the shed blood had been taken up the rest was slowly infused. When blood pressure, heart rate, and coronary flow were stable, blood samples were drawn and microspheres injected; this phase was termed the posttransfusion state (PT). The dog was then followed until definite signs of cardiovascular decay (CD) developed-that is, progressive decrease of blood pressure and coronary flow. Because of the hypotension at this time, microspheres were injected without obtaining blood samples and occlusion zeros. All dogs were kept at an arterial blood pH between 7,2 and 7.3 by varying respiratory rate and giving sodium bicarbonate. Because there were five states and only four different nuclides, every untreated dog was given the control injection and the three other sets of microspheres were given at three randomly chosen states out of the remaining four states of hemorrhage. Dogs given PBZ were not studied to the posttransfusion stage. Eight untreated and three PBZ-treated dogs were allowed to progress to late hemorrhage, appropriate measurements being made. Then 20% mannitol was infused into the root of the aorta at 5 ml/min for 15-20 min. The tubing to the reservoir was open so that aortic pressure did not rise. After the infusion another set of measurements was made; the dogs were then killed and the hearts analyzed as above. RESULTS
Times, volumes, and hematocrits. Table 1 shows that the time taken to reach EH was similar in untreated and PBZ-treated dogs, but the time to LH was signifi1. Times, volumes bled, and hematocrits at different stages of shock TABLE
Control
EH
LH ----
Time, min Untreated Untreated + mannitol Phenoxybenzamine
50*5.3(13) 48*2.3(7)3 59?2.9(6)$
Volume bled, ml/kg Untreated Phenoxybenzamine
38*22.2(20) 24?2.9(6)*
43*2.1(23) 27~2.7(5)*
Hematmrit, % Untreated Phenoxybenzamine
43*1.6(12) 362 1.2(6)*
39*2.2(16) 32k 1.6(6)t
115*8.0(19) 124? 13.6(7)$ 1732 18.4(6)*
Values are mean 2 1. SE, with numbers of observations in parentheses. All comparisons are by unpaired-# test with corresponding values for untreated doga: l P < 0.05; t 0.05 < P < 0.10; $ 0.10
