Cardiovascular Research, 1976, 10, 81-90.
Distribution of coronary collateral flow in acute myocardial ischaemic injury: effect of propranolol' R O B E R T A . K L O N E R , K E I T H A . R E I M E R ~and , ROBERT B. J E N N I N G S ~
From the Department of Pathology, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, Illlinois 60611, USA
The local distribution of coronary collateral flow was mapped using 8 (I tracer microspheres following circumflex coronary occlusions in dogs. Overall flow to the ischaemic posterior papillary muscle and subjacent myocardium decreased to 2 I of anterior free wall flow. Collateral flow was non-uniform, however, and was lower in the subendocardium (14%)than in the subepicardium (27 %). Brief temporary occlusions with and without propranolol therapy showed that collateral flow was less and the inner/outer wall ratio was unaltered in treated animals. Although propranolol reduces infarct size after coronary occlusions, this effect appears not to be related to increases in collateral flow.
AUTHORS' SYNOPSIS
This work was supported in part by grant HE 08729, H L I8833 and contract NHLI-72-2984 from the National Institutes of Health (Heart and Lung Institute) and by a grant from the Chicago and Illinois Heart Association. R.A.K. was a predoctoral fellow (National Institutes of Health Grant G M 00131) during the time of this study. 'Present address: Department of Pathology, Box 3712, Duke University Medical Center. Durham, N C 27710, USA.
flow during circumflex occlusion and to compare it with the known distribution and timing of development of myocardial necrosis. Propranolol significantly reduces the amount of necrosis produced by either temporary or permanent coronary occlusion and 24 h of survival (Reimer et al, 1973; Rasmussen et al, 1974). Its mechanism of action is unknown. One hypothesis is that it improves myocardial 0, supply, either in absolute terms or relative to oxygen demand. In support of this hypothesis, Becker et a1 ( I97 1) using I5 p radioactive microspheres and 10 g samples of heart showed that propranolol increased the inner/outer wall ratio of flow to ischaemic myocardium. However, they did not measure the effect of propranolol on absolute myocardial collateral flow to different regions of the myocardium. The second major purpose of our study, therefore, was to assess the effect of propranolol on both the absolute rate and local distribution of myocardial blood flow following circumflex coronary occlusion.
Downloaded from by guest on June 5, 2016
When the circumflex coronary artery in dogs is occluded for less than 20min, myocardial cell necrosis does not occur. Longer occlusions cause cell death (Jennings et at, 1969). As the duration of occlusion is increased, necrosis occurs first in the subendocardial and then in the midepicardial and finally in the subepicardial zones. Previous studies have shown that blood flow in ischaemic myocardium is reduced to a greater extent in the subendocardium than the subepicardium (Becker et a / , 1971; Moir, 1972; Becker et al, 1973; Wusten et al, 1974). Moreover, it seems likely that the distribution of myocardial necrosis is related to local distribution of collateral flow. The first purpose of this study was quantitatively to map the distribution of myocardial blood
02 Kloner, Reimer, and Jennings Materials and methods Mongrel dogs of either sex were fasted overnight and were anaesthetized with sodium pentobarbital (30 mg/kg). The dogs were intubated and ventilated with an Harvard respirator pump (Model 1063). Lead I1 of the electrocardiogram was monitored continuously throughout the experiment on a Grass (Model 5P1) polygraph. Arterial blood pressure was recorded from a brachial artery catheter attached to a Statham strain gauge. Catheters were also placed in the femoral artery and vein for reference blood flow collection and thioflavin-S injection, respectively (see below). After opening the left fourth interspace, the lung was retracted and the pericardium was opened. A polyethylene tube was placed in the left atrial appendage and tied into place with a purse string suture. The circumflex branch of the left coronary artery was isolated 1 to 2cm from the aorta. Temporary coronary occlusions were made with a snare and permanent occlusions with ligatures.
Modified from Hoffman, Rudolph. and Heymann, Department of Pediatrics, University of California Medical School at San Francisco.
Experimental design Dogs were assigned randomly to three groups. I Double occlusions in untreated dogs In eight dogs, spheres were successfully injected during each of two 7-min circumflex coronary occlusions separated by a 30-min period of reflow. A different nuclide was used for each injection which was begun at 5 min after occlusion. Ten min before the second occlusion, normal saline (0.8 ml/kg) was given intravenously as a control for the propranolol treatment group. Reference blood flows were collected from the femoral artery catheter during each sphere injection as described above. During the second occlusion, a catheter was placed briefly in the circumflex artery distal to the tie to measure peripheral coronary pressure (PCP). At the end of the second 7-min ligation, thioflavin-S (4% in normal saline, 1 .O ml/kg), a fluorescent dye which binds to vascular endothelium, was injected intravenously and 10 s later the heart was excised. 2 Double occlusions before and after propranolol Nine experiments were performed exactly as described for group 1, except that 10min before the second occlusion 5 mg/kg of propranolol in 0.5 ml/kg saline was given, via the femoral artery, followed by an additional 0.3 ml/kg normal saline flush. This dose of propranolol results in significant reduction in infarct size (Reimer et al, 1973); and, although it is a high dose with both @-blockingand membrane-stabilizing actions, it has never resulted in severe cardiac decompensation, ie, hypotension, in our experiments.
3 Single occlusions after propranolol To determine whether the first occlusion and sphere injection altered the flow values obtained during the second occlusion after propranolol treatment in group 2, 11 dogs were pretreated with propranolol and were then subjected to a single occlusion and sphere injection.
Downloaded from by guest on June 5, 2016
Microspheres Carbonized microspheres with a diameter of 8 i1 p were obtained from the 3M Company, Minneapolis, Minnesota. One batch was impregnated with strontium 85 (17.6 mCi/g) and a second batch with cerium 141 (27.8 mCi/g). One mCi of each was suspended in 20ml of normal saline plus a few drops of Tween 80 to prevent clumping (stock solutions). The radioactivity/sphere was calculated by numerically counting spheres ( >400/sample) on a transparent grid viewed by light microscopy and then by counting the gamma activity in a Packard Autogamma counter (Model 5986). Before each experiment, the stock solutions were suspended uniformly by shaking in an Eberbach mechanical agitator. A volume containing approximately 800 OOO to 1 100 OOO spheres was placed into a specially designed injecting vial' which was agitated continuously for 30 min, ultrasonicated for 1 to 2min just before injection and was agitated manually until the moment of injection. In preliminary experiments, no clumping of spheres was observed by microscopic examination when this protocol was followed. To measure myocardial blood flow, the spheres were injected into the left atrium over a 20-s period by flushing the injecting vial with 7 to 10 ml of normal saline (at 37°C). The atrial catheter was flushed with an additional 5 ml of warmed saline. Beginning 5 s before injection and continuing for
exactly 2min, a reference blood sample was withdrawn from the femoral artery catheter at 10 ml/min using an Harvard constant withdrawal pump (Model 600-000). Preliminary experiments demonstrated that most of the spheresentering the reference catheter were collected during the first 30-s interval and that the 2-min period was adequate to collect 99.9% of spheres which could be obtained with longer collecting periods.) The flow rate was calculated from the weight of the blood collected and the time of collection.
83 CollateralJlow in acute myocardial ischaemia Ten dogs died from ventricular fibrillation during the first occlusion and six dogs survived the first occlusion and sphere injection but died at release or during the second occlusion. These were eliminated from the study.
Mapping blood flow Each excised heart was weighed and the left ventricle (including septum) was dissected free and frozen to make further dissection easier. The frozen left ventricle and septum then were subdivided into 13 sections and numbered as shown in Fig. 1 . The block containing the posterior papillary muscle (PP) was cut into longitudinal strips and each slice was photographed under ultra-violet (UV) light so that the distribution of thioflavin-S fluorescence could be evaluated. Sections then were split into inner and outer halves, and the inner and outer halves of each slice of the PP were further subdivided into fluorescent and nonfluorescent portions. Each portion of tissue was weighed, and the 85Sr and/or lrlCe activities were determined by pulse height analysis using the Packard autogamma counter.
MBF= C m
RBF Y
Cr where MBF = myocardial blood flow (ml/rnin.g-l), RBF = reference blood flow (ml/rnin), C , = counts/1000 s/g of heart, and C r = total counts in the reference flow sample/ 1000 s. Cardiac output (CO) also was calculated as
co=-RBF cr
total counts/l000 s injected into the dog.
In addition to absolute flows, the ratio of inner and outer blood flow (I/O) was calculated.
Results Coronary flow distribution in non-ischaemic myocardium The normal variation in regional blood flow distribution to the left ventricle as measured by the microsphere technique was assessed in five sham-operated dogs (no coronary occlusion). In
‘ Modified from Hoffman, Rudolf,
and Heyrnann, Department of Pediatrics, University of California Medical School at San Francisco.
1 Maps of the left ventricle showing mean valuesfor myocardial blood flow in untreated (A) and propranolol-treated (B) dogs. The left ventricle is opened with the endocardial surface face up. I t is cut into 13 tissue blocks, bused on longitudinal cuts on either side of the papillary muscles and on horizontal cuts at the base of the posterior papillary muscle (PP) and at the junction of the lower and middle one-third of the PP and the anterior papillary muscle (AP). Each section (except no. 13) is divided into inner and outer halves. A : Mean values for myocardial blood flow in 17 dogs with untreated circumflex artery occlusion (first occlusions of all dogs in groups 2 and 3). Values are expressed as innerlouter flow (rnl/min.g-’). B: Mean values for myocardial blood flow in 20 dogs treated with propranolol before circumflex artery occlusion (second occlusions of group 3 and single occlusions of group 4). Propranolol reduced the absolute flow in both non-ischaemic and ischaemic myocardium but had no significant effect on the 1 / 0 ratios. FIG.
these dogs, the average flow rate to all sections of the left ventricular wall was similar and flow to section no. 1, ie, the section containing the posterior papillary muscle (1.1 1 k0.34 ml/ minegl), was virtually identical to flow in the anterior free wall (AFW=sections 8 and 9
Downloaded from by guest on June 5, 2016
Calculations of flow The blood flow to each portion of heart was calculated’ by ratio and proportion as
84 Kloner, Reimer, and Jennings
therefore was chosen as the non-ischaemic control muscle within each heart. Flow in ischaemic myocardium Following circumflex coronary occlusion, flow to the section of the left ventricle containing the PP (representing the central zone of most severe ischaemia) was reduced to 0.20+0.04 ml/min.g-' (21 % of AFW) during the first occlusion in the 17 dogs of groups 1 and 2. The reduction in flow, however, always was greater in the inner half of the left ventricular wall than in the outer half and corresponded to 14+_4%and 27+5';;: of the flow to the AFW, respectively (Table 1 ; Fig. 2). Flows obtained during a repeat untreated occlusion (group 1) were similar to those observed during the first occlusion. Maximum reductions in flow were found not only in the PP but also in the area between the PP and septum (section 2). The sections peripheral to the areas of maximum ischaemia (sections 3, 4, and 5) showed more moderate decreases in absolute flow and 1/0 ratio (Fig. I A).
Effect of propranolol on coronary flow distribution Propranolol significantly reduced flow to the non-ischaemic areas (group 2). The reduction in
TABLE
1
Local blood jlo w in ischaemic and non-ischaemic myocardium in untreated and propranolol-treated dogs Mean flow in ischaemic PP (section no. I )
Croup I ( t i = & 1st occlusion (untreated) 2nd occlusion (untreated)
P Croup 2 ( n = 9) 1st occlusion (untreated) 2nd occlusion (propranolol)
P Group 3 ( i f = 1 1 ) 1st occlusion (propranolol) P (group 3 YS pooled 1st occlusion of groups 1 and 2)
*
Standard error of the mean.
Inner
Outer
(m1lmin.g ')
(ml/min.g-')
Mean /low in non-ischaemic anterior left ventricular wall(sections no. 8 and Y) Inner
I10
Outer
(ml/min.g-') (m1lniin.g')
I10
0.14~k0.04' 0.27$0.05 0. I6 1 0 . 0 6 0.3110.07 NS NS
0.4750.07 0.4510.08 NS
0.95 10.10 0.95 i 0.10 NS
0.97k0.10 NS
0.99 1 0.05 0.96 i 0.06 NS
0. I7 1 0 . 0 6 0.08 t 0.03 ~0.025
0.2310.061 0.13i0.04
0.63 0.10 0.48i0.10 NS
1.1510.19
0.9110.16
1.1710.17 0.91 t0.17
0.98 0.03 I .03 -1 0.05
0.06 t 0.01
0.14-0.03
0.13
0.88'0.09
0.8710.09
1.03 i 0.04
NS
NS
NS
NS
NS
-0.025
0.57
NS
s.0.025
1.0010.12
0.025
+
NS
Downloaded from by guest on June 5, 2016
pooled) ( I .0820.36 ml/min.g-l). Flow to the inner and outer halves of the myocardium also were similar, so that the mean 1/0 ratios in the PP and AFW were 1.04+_0.04and 1.10+0.04, respectively. In two sham-operated dogs which received two sequential injections, one with 14Te and one with the results with the second isotope were similar to those obtained with the first. In the 17 dogs of groups 1 and 2, the blood flows and 1/0 ratios in the AFW during the first episode of circumflex coronary occlusion were similar to those measured in the sham group. Also, measurements of AFW flows and 1/0 ratios during two sequential untreated occlusions (group 2) were reproducible in each dog as well as within the group as a whole (Table 1). These results indicate that induction of ischaemia in the posterolateral wall does not produce any marked alterations in coronary aterial flow to the non-ischaemic anterior free wall. Moreover, since the AFW arterial flows are virtually identical to those noted in the PFW of the sham-operated animals, the AFW flows can serve as reference flow rates for estimating reductions in the PFW flow as percent of control flow. The AFW of the left ventricle never is cyanotic following circumflex occlusion, and this area (sections 8 and 9 pooled; see Fig. 1)
85 Collateral flow in acute myocardial ischaemia 1 i r occl-unfreored lrr and 2nd occl unrreared 2nd occl-propronolol Nanirchocmlc myocordum Noniichoemlc mvocard8um
01.
'
.?Id hr occl occl
coronary occlusion (group 3). Results from these dogs (Table 1) were similar to those observed following propranolol in the doubleocclusion dogs.
(jLl I IIrf0 2I 2
occl
occl
occl
occl
occl
occl
PP flow produced by circumflex occlusion after propranolol was also significantly greater than the reduction produced by an identical occlusion in the same animal before propranolol (Table 1, Figs. 1B and 2). The 1 / 0 ratios in both ischaemic and non-ischaemic areas were not significantly changed by propranolol, indicating that there was no relative increase in subendocardial flow. When the flow to the ischaemic PP was expressed as a percentage of the flow to the non-ischaemic AFW, propranolol had no significant effect. In order to ascertain that the first occlusion itself did not produce the depression in arterial flow observed with propranolol, flow was measured with a single sphere injection in 1 1 dogs after propranolol treatment and during circumflex
Other indices of collateral flow and ischaemic injury Thiojlavin-S distribution
The dye, thioflavin-S (TS), when injected intravenously 10s before excision of the heart, allows visualization (by fluorescence under ultraviolet light) of areas which were perfused versus areas receiving little or no flow (Kloner et al, 1975). Sections of myocardium from the AFW and interventricular septum always demonstrated homogeneous yellow-green thioflavin-S fluorescence and thus served as an internal control for adequacy of the injection procedure. Conversely, sections of the PP of hearts with circumflex coronary occlusion always showed non-fluorescence involving the subendocardial third of the myocardium with patchy extensions into the mid- and occasionally the subepicardial myocardium. The percentage of each PP which was non-fluorescent ("/,NF) was quantitated from photographs and was 54% in the untreated group (group 2) and 63% in propranololtreated dogs (Table 3). Thus, propranolol-
Downloaded from by guest on June 5, 2016
2 Mean flow in non-ischaemic anterior free wall (sections 8 and 9 pooled) and ischaemic P P (section I ) myocardium during two untreated paired coronary occlusions in eight dogs and during a paired untreated coronary occlusion and treafed coronary occlusion in nine dogs. (Bars indicate plus or minus one SEM.) In dogs with two untreated occlusions neither anterior free wall ( A F W )nor PPPow changed significantly between the first and second occlusions. I n dogs treated with propranolol before the second occlusion, flow decreased significantly in both ischaemic and non-ischaemic myocardium during the propranolol-treated occlusion. I =mean flow (ml/min.g-') in the inner half of myocardium, O=mean flow (mI/mirrg-l) in the outer halfof myocardium. The 110 ratio was less in ischaemic myocardium than in non-ischaemic myocardium in all dogs. occl= occlusion. FIG.
Haemodynamic parameters The decreased coronary flow observed following propranolol was associated with decreases in heart rate, cardiac output, systolic blood pressure, and cardiac effort index (Table 2). Conversely, these parameters all were similar during each of two sequential occlusions in untreated dogs. There was a significant linear correlation between heart rate and flow to non-ischaemic zones in shams (r=0.98, P t 0 . 0 0 5 ) , and untreated dogs with occlusions (r=0.54, P
Distribution of coronary collateral flow in acute myocardial ischaemic injury: effect of propranolol' R O B E R T A . K L O N E R , K E I T H A . R E I M E R ~and , ROBERT B. J E N N I N G S ~
From the Department of Pathology, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, Illlinois 60611, USA
The local distribution of coronary collateral flow was mapped using 8 (I tracer microspheres following circumflex coronary occlusions in dogs. Overall flow to the ischaemic posterior papillary muscle and subjacent myocardium decreased to 2 I of anterior free wall flow. Collateral flow was non-uniform, however, and was lower in the subendocardium (14%)than in the subepicardium (27 %). Brief temporary occlusions with and without propranolol therapy showed that collateral flow was less and the inner/outer wall ratio was unaltered in treated animals. Although propranolol reduces infarct size after coronary occlusions, this effect appears not to be related to increases in collateral flow.
AUTHORS' SYNOPSIS
This work was supported in part by grant HE 08729, H L I8833 and contract NHLI-72-2984 from the National Institutes of Health (Heart and Lung Institute) and by a grant from the Chicago and Illinois Heart Association. R.A.K. was a predoctoral fellow (National Institutes of Health Grant G M 00131) during the time of this study. 'Present address: Department of Pathology, Box 3712, Duke University Medical Center. Durham, N C 27710, USA.
flow during circumflex occlusion and to compare it with the known distribution and timing of development of myocardial necrosis. Propranolol significantly reduces the amount of necrosis produced by either temporary or permanent coronary occlusion and 24 h of survival (Reimer et al, 1973; Rasmussen et al, 1974). Its mechanism of action is unknown. One hypothesis is that it improves myocardial 0, supply, either in absolute terms or relative to oxygen demand. In support of this hypothesis, Becker et a1 ( I97 1) using I5 p radioactive microspheres and 10 g samples of heart showed that propranolol increased the inner/outer wall ratio of flow to ischaemic myocardium. However, they did not measure the effect of propranolol on absolute myocardial collateral flow to different regions of the myocardium. The second major purpose of our study, therefore, was to assess the effect of propranolol on both the absolute rate and local distribution of myocardial blood flow following circumflex coronary occlusion.
Downloaded from by guest on June 5, 2016
When the circumflex coronary artery in dogs is occluded for less than 20min, myocardial cell necrosis does not occur. Longer occlusions cause cell death (Jennings et at, 1969). As the duration of occlusion is increased, necrosis occurs first in the subendocardial and then in the midepicardial and finally in the subepicardial zones. Previous studies have shown that blood flow in ischaemic myocardium is reduced to a greater extent in the subendocardium than the subepicardium (Becker et a / , 1971; Moir, 1972; Becker et al, 1973; Wusten et al, 1974). Moreover, it seems likely that the distribution of myocardial necrosis is related to local distribution of collateral flow. The first purpose of this study was quantitatively to map the distribution of myocardial blood
02 Kloner, Reimer, and Jennings Materials and methods Mongrel dogs of either sex were fasted overnight and were anaesthetized with sodium pentobarbital (30 mg/kg). The dogs were intubated and ventilated with an Harvard respirator pump (Model 1063). Lead I1 of the electrocardiogram was monitored continuously throughout the experiment on a Grass (Model 5P1) polygraph. Arterial blood pressure was recorded from a brachial artery catheter attached to a Statham strain gauge. Catheters were also placed in the femoral artery and vein for reference blood flow collection and thioflavin-S injection, respectively (see below). After opening the left fourth interspace, the lung was retracted and the pericardium was opened. A polyethylene tube was placed in the left atrial appendage and tied into place with a purse string suture. The circumflex branch of the left coronary artery was isolated 1 to 2cm from the aorta. Temporary coronary occlusions were made with a snare and permanent occlusions with ligatures.
Modified from Hoffman, Rudolph. and Heymann, Department of Pediatrics, University of California Medical School at San Francisco.
Experimental design Dogs were assigned randomly to three groups. I Double occlusions in untreated dogs In eight dogs, spheres were successfully injected during each of two 7-min circumflex coronary occlusions separated by a 30-min period of reflow. A different nuclide was used for each injection which was begun at 5 min after occlusion. Ten min before the second occlusion, normal saline (0.8 ml/kg) was given intravenously as a control for the propranolol treatment group. Reference blood flows were collected from the femoral artery catheter during each sphere injection as described above. During the second occlusion, a catheter was placed briefly in the circumflex artery distal to the tie to measure peripheral coronary pressure (PCP). At the end of the second 7-min ligation, thioflavin-S (4% in normal saline, 1 .O ml/kg), a fluorescent dye which binds to vascular endothelium, was injected intravenously and 10 s later the heart was excised. 2 Double occlusions before and after propranolol Nine experiments were performed exactly as described for group 1, except that 10min before the second occlusion 5 mg/kg of propranolol in 0.5 ml/kg saline was given, via the femoral artery, followed by an additional 0.3 ml/kg normal saline flush. This dose of propranolol results in significant reduction in infarct size (Reimer et al, 1973); and, although it is a high dose with both @-blockingand membrane-stabilizing actions, it has never resulted in severe cardiac decompensation, ie, hypotension, in our experiments.
3 Single occlusions after propranolol To determine whether the first occlusion and sphere injection altered the flow values obtained during the second occlusion after propranolol treatment in group 2, 11 dogs were pretreated with propranolol and were then subjected to a single occlusion and sphere injection.
Downloaded from by guest on June 5, 2016
Microspheres Carbonized microspheres with a diameter of 8 i1 p were obtained from the 3M Company, Minneapolis, Minnesota. One batch was impregnated with strontium 85 (17.6 mCi/g) and a second batch with cerium 141 (27.8 mCi/g). One mCi of each was suspended in 20ml of normal saline plus a few drops of Tween 80 to prevent clumping (stock solutions). The radioactivity/sphere was calculated by numerically counting spheres ( >400/sample) on a transparent grid viewed by light microscopy and then by counting the gamma activity in a Packard Autogamma counter (Model 5986). Before each experiment, the stock solutions were suspended uniformly by shaking in an Eberbach mechanical agitator. A volume containing approximately 800 OOO to 1 100 OOO spheres was placed into a specially designed injecting vial' which was agitated continuously for 30 min, ultrasonicated for 1 to 2min just before injection and was agitated manually until the moment of injection. In preliminary experiments, no clumping of spheres was observed by microscopic examination when this protocol was followed. To measure myocardial blood flow, the spheres were injected into the left atrium over a 20-s period by flushing the injecting vial with 7 to 10 ml of normal saline (at 37°C). The atrial catheter was flushed with an additional 5 ml of warmed saline. Beginning 5 s before injection and continuing for
exactly 2min, a reference blood sample was withdrawn from the femoral artery catheter at 10 ml/min using an Harvard constant withdrawal pump (Model 600-000). Preliminary experiments demonstrated that most of the spheresentering the reference catheter were collected during the first 30-s interval and that the 2-min period was adequate to collect 99.9% of spheres which could be obtained with longer collecting periods.) The flow rate was calculated from the weight of the blood collected and the time of collection.
83 CollateralJlow in acute myocardial ischaemia Ten dogs died from ventricular fibrillation during the first occlusion and six dogs survived the first occlusion and sphere injection but died at release or during the second occlusion. These were eliminated from the study.
Mapping blood flow Each excised heart was weighed and the left ventricle (including septum) was dissected free and frozen to make further dissection easier. The frozen left ventricle and septum then were subdivided into 13 sections and numbered as shown in Fig. 1 . The block containing the posterior papillary muscle (PP) was cut into longitudinal strips and each slice was photographed under ultra-violet (UV) light so that the distribution of thioflavin-S fluorescence could be evaluated. Sections then were split into inner and outer halves, and the inner and outer halves of each slice of the PP were further subdivided into fluorescent and nonfluorescent portions. Each portion of tissue was weighed, and the 85Sr and/or lrlCe activities were determined by pulse height analysis using the Packard autogamma counter.
MBF= C m
RBF Y
Cr where MBF = myocardial blood flow (ml/rnin.g-l), RBF = reference blood flow (ml/rnin), C , = counts/1000 s/g of heart, and C r = total counts in the reference flow sample/ 1000 s. Cardiac output (CO) also was calculated as
co=-RBF cr
total counts/l000 s injected into the dog.
In addition to absolute flows, the ratio of inner and outer blood flow (I/O) was calculated.
Results Coronary flow distribution in non-ischaemic myocardium The normal variation in regional blood flow distribution to the left ventricle as measured by the microsphere technique was assessed in five sham-operated dogs (no coronary occlusion). In
‘ Modified from Hoffman, Rudolf,
and Heyrnann, Department of Pediatrics, University of California Medical School at San Francisco.
1 Maps of the left ventricle showing mean valuesfor myocardial blood flow in untreated (A) and propranolol-treated (B) dogs. The left ventricle is opened with the endocardial surface face up. I t is cut into 13 tissue blocks, bused on longitudinal cuts on either side of the papillary muscles and on horizontal cuts at the base of the posterior papillary muscle (PP) and at the junction of the lower and middle one-third of the PP and the anterior papillary muscle (AP). Each section (except no. 13) is divided into inner and outer halves. A : Mean values for myocardial blood flow in 17 dogs with untreated circumflex artery occlusion (first occlusions of all dogs in groups 2 and 3). Values are expressed as innerlouter flow (rnl/min.g-’). B: Mean values for myocardial blood flow in 20 dogs treated with propranolol before circumflex artery occlusion (second occlusions of group 3 and single occlusions of group 4). Propranolol reduced the absolute flow in both non-ischaemic and ischaemic myocardium but had no significant effect on the 1 / 0 ratios. FIG.
these dogs, the average flow rate to all sections of the left ventricular wall was similar and flow to section no. 1, ie, the section containing the posterior papillary muscle (1.1 1 k0.34 ml/ minegl), was virtually identical to flow in the anterior free wall (AFW=sections 8 and 9
Downloaded from by guest on June 5, 2016
Calculations of flow The blood flow to each portion of heart was calculated’ by ratio and proportion as
84 Kloner, Reimer, and Jennings
therefore was chosen as the non-ischaemic control muscle within each heart. Flow in ischaemic myocardium Following circumflex coronary occlusion, flow to the section of the left ventricle containing the PP (representing the central zone of most severe ischaemia) was reduced to 0.20+0.04 ml/min.g-' (21 % of AFW) during the first occlusion in the 17 dogs of groups 1 and 2. The reduction in flow, however, always was greater in the inner half of the left ventricular wall than in the outer half and corresponded to 14+_4%and 27+5';;: of the flow to the AFW, respectively (Table 1 ; Fig. 2). Flows obtained during a repeat untreated occlusion (group 1) were similar to those observed during the first occlusion. Maximum reductions in flow were found not only in the PP but also in the area between the PP and septum (section 2). The sections peripheral to the areas of maximum ischaemia (sections 3, 4, and 5) showed more moderate decreases in absolute flow and 1/0 ratio (Fig. I A).
Effect of propranolol on coronary flow distribution Propranolol significantly reduced flow to the non-ischaemic areas (group 2). The reduction in
TABLE
1
Local blood jlo w in ischaemic and non-ischaemic myocardium in untreated and propranolol-treated dogs Mean flow in ischaemic PP (section no. I )
Croup I ( t i = & 1st occlusion (untreated) 2nd occlusion (untreated)
P Croup 2 ( n = 9) 1st occlusion (untreated) 2nd occlusion (propranolol)
P Group 3 ( i f = 1 1 ) 1st occlusion (propranolol) P (group 3 YS pooled 1st occlusion of groups 1 and 2)
*
Standard error of the mean.
Inner
Outer
(m1lmin.g ')
(ml/min.g-')
Mean /low in non-ischaemic anterior left ventricular wall(sections no. 8 and Y) Inner
I10
Outer
(ml/min.g-') (m1lniin.g')
I10
0.14~k0.04' 0.27$0.05 0. I6 1 0 . 0 6 0.3110.07 NS NS
0.4750.07 0.4510.08 NS
0.95 10.10 0.95 i 0.10 NS
0.97k0.10 NS
0.99 1 0.05 0.96 i 0.06 NS
0. I7 1 0 . 0 6 0.08 t 0.03 ~0.025
0.2310.061 0.13i0.04
0.63 0.10 0.48i0.10 NS
1.1510.19
0.9110.16
1.1710.17 0.91 t0.17
0.98 0.03 I .03 -1 0.05
0.06 t 0.01
0.14-0.03
0.13
0.88'0.09
0.8710.09
1.03 i 0.04
NS
NS
NS
NS
NS
-0.025
0.57
NS
s.0.025
1.0010.12
0.025
+
NS
Downloaded from by guest on June 5, 2016
pooled) ( I .0820.36 ml/min.g-l). Flow to the inner and outer halves of the myocardium also were similar, so that the mean 1/0 ratios in the PP and AFW were 1.04+_0.04and 1.10+0.04, respectively. In two sham-operated dogs which received two sequential injections, one with 14Te and one with the results with the second isotope were similar to those obtained with the first. In the 17 dogs of groups 1 and 2, the blood flows and 1/0 ratios in the AFW during the first episode of circumflex coronary occlusion were similar to those measured in the sham group. Also, measurements of AFW flows and 1/0 ratios during two sequential untreated occlusions (group 2) were reproducible in each dog as well as within the group as a whole (Table 1). These results indicate that induction of ischaemia in the posterolateral wall does not produce any marked alterations in coronary aterial flow to the non-ischaemic anterior free wall. Moreover, since the AFW arterial flows are virtually identical to those noted in the PFW of the sham-operated animals, the AFW flows can serve as reference flow rates for estimating reductions in the PFW flow as percent of control flow. The AFW of the left ventricle never is cyanotic following circumflex occlusion, and this area (sections 8 and 9 pooled; see Fig. 1)
85 Collateral flow in acute myocardial ischaemia 1 i r occl-unfreored lrr and 2nd occl unrreared 2nd occl-propronolol Nanirchocmlc myocordum Noniichoemlc mvocard8um
01.
'
.?Id hr occl occl
coronary occlusion (group 3). Results from these dogs (Table 1) were similar to those observed following propranolol in the doubleocclusion dogs.
(jLl I IIrf0 2I 2
occl
occl
occl
occl
occl
occl
PP flow produced by circumflex occlusion after propranolol was also significantly greater than the reduction produced by an identical occlusion in the same animal before propranolol (Table 1, Figs. 1B and 2). The 1 / 0 ratios in both ischaemic and non-ischaemic areas were not significantly changed by propranolol, indicating that there was no relative increase in subendocardial flow. When the flow to the ischaemic PP was expressed as a percentage of the flow to the non-ischaemic AFW, propranolol had no significant effect. In order to ascertain that the first occlusion itself did not produce the depression in arterial flow observed with propranolol, flow was measured with a single sphere injection in 1 1 dogs after propranolol treatment and during circumflex
Other indices of collateral flow and ischaemic injury Thiojlavin-S distribution
The dye, thioflavin-S (TS), when injected intravenously 10s before excision of the heart, allows visualization (by fluorescence under ultraviolet light) of areas which were perfused versus areas receiving little or no flow (Kloner et al, 1975). Sections of myocardium from the AFW and interventricular septum always demonstrated homogeneous yellow-green thioflavin-S fluorescence and thus served as an internal control for adequacy of the injection procedure. Conversely, sections of the PP of hearts with circumflex coronary occlusion always showed non-fluorescence involving the subendocardial third of the myocardium with patchy extensions into the mid- and occasionally the subepicardial myocardium. The percentage of each PP which was non-fluorescent ("/,NF) was quantitated from photographs and was 54% in the untreated group (group 2) and 63% in propranololtreated dogs (Table 3). Thus, propranolol-
Downloaded from by guest on June 5, 2016
2 Mean flow in non-ischaemic anterior free wall (sections 8 and 9 pooled) and ischaemic P P (section I ) myocardium during two untreated paired coronary occlusions in eight dogs and during a paired untreated coronary occlusion and treafed coronary occlusion in nine dogs. (Bars indicate plus or minus one SEM.) In dogs with two untreated occlusions neither anterior free wall ( A F W )nor PPPow changed significantly between the first and second occlusions. I n dogs treated with propranolol before the second occlusion, flow decreased significantly in both ischaemic and non-ischaemic myocardium during the propranolol-treated occlusion. I =mean flow (ml/min.g-') in the inner half of myocardium, O=mean flow (mI/mirrg-l) in the outer halfof myocardium. The 110 ratio was less in ischaemic myocardium than in non-ischaemic myocardium in all dogs. occl= occlusion. FIG.
Haemodynamic parameters The decreased coronary flow observed following propranolol was associated with decreases in heart rate, cardiac output, systolic blood pressure, and cardiac effort index (Table 2). Conversely, these parameters all were similar during each of two sequential occlusions in untreated dogs. There was a significant linear correlation between heart rate and flow to non-ischaemic zones in shams (r=0.98, P t 0 . 0 0 5 ) , and untreated dogs with occlusions (r=0.54, P
