lAce Vol. 9. No.
3 647
March 19X7:647 54
Determinants of Infarct Size During Permanent Occlusion of a
Coronary Artery in the Closed Chest Dog
TETSUJI MIURA, MD,* DEREK M. YELLON, PHD. t DAVlD J. HEARSE, DSc,t
JAMES M. DOWNEY, PHD*
Mobile. Alabama and London , Eng/and
One hundred nine dogs subjected to experienced exper-
imental coronary occlusion were retrospectively exam-
ined in an attempt to identify factors influencing infarct
size. A coronary artery was occluded by an embolus.
The field of the occluded artery (zone at risk) was de-
termined by subsequent autoradiography of 141-cer-
ium-Iabeled microspheres that were injected into the left
ventricle 2 minutes after embolization. Fifty-eight dogs
were analyzed after 24 hours of embolization and 51
after 48 hours. Infarct size (assessed by tetrazolium
staining) was directly proportional to the size of the zone
at risk in both groups. The percent of the risk zone that
developed infarction was independent of risk zone size
in both the 24 and 48 hour groups. No differences were
seen between male and female animals or between the
dogs with left circumflex or left anterior descending ar-
tery occlusion. Infarct size was also independent of the
heart rate-systolic pressure product at the time of coro-
nary occlusion, and of the time of year.
The identification of interventions that limit myocardial in-
farct size is currently the subject of intense investigation
and controversy. The valid assessment of any intervention
in an animal model requires that the investigator be able to
predict how large the infarct would have been in the absence
of that intervention. To make such an accurate prediction
requires an understanding of the determinants of infarct size
in that model. In the dog it has been known for some time
that the size of the field of the occluded artery (the zone at
risk of infarction) is a primary determinant of infarct size
From the ' Department of Physiology. University of South Alabama .
Mob ile. Alabama and the t Heart Research Unit. The Rayne Institute .
S\. T homas ' Hospital . Lond on , England . Thi s work was supported in part
by Grant HL 2064 8 from the Nation al Institute s o f Health . Bethesda.
Maryland .
Manuscript received November 14. 1985: revised manuscript recei ved
Augu st 6. 1986. accepted September 2. 1986.
Addre ~s for reprints: James M . Down ey . PhD . MSB Room 3024.
Department of Physiology. College of Medicine. University of South Al-
abama . Mobile . Alabama. 36688.
(' 1987 by the American College of Cardiology
Eighty percent of the variability in the portion of the
risk zone that infarcted in this population could be ex-
plained by the level of collateral flow in the risk zone 2
minutes after embolization. The linear regression be-
tween the percent of the risk zone that developed in-
farction and collateral flow was the same in slope and
intercept between the 24 and 48 hour groups. The cor-
relation betweeninfarct sizeand collateral flow wasslightly
better when collateral flow was expressed as a percent
of flow in the nonischemic zone than when it was ex-
pressed in absolute terms. In conclusion, infarcts reach
their ultimate size in the first 24 hours of ischemia and
the size of the zone at risk and the level of blood flow
through native collateral channels are the major deter-
minants of infarct size after permanent occlusion of a
coronary artery.
(J Am Coli CardioI1987;9:647-54)
( I). The percent of the risk zone that proceeds to infarction
in untreated dogs varies widely, however, and the factors
that alter this percent remain unclear. The duration of oc-
clusion (2), the level of collateral flow (2-6 ), the size of
the risk zone (5.7), the branch that is occluded (8), the sex
of the dog (9- 11), the myocardial oxygen consumption at
occlusion (12) and even the time of year (13) have all been
proposed as determinants of the extent of necrosis within
the risk zone.
Over the past 5 years we have conducted a series of
experiments in which the effect of various interventions on
infarct size in the dog was examined ( 13- 17, and unpub-
lished data). ln these studies the coronary arteries were
occluded by introducingemboli into the coronary circulation
through a catheter and the bed of the occluded artery was
visualized by microsphere autoradiography. Because each
study included a control group of about 10 animals, we have
now accumulated a data base that includes a very large
number of control dogs. Over the years, emboli of various
sizes were employed and either the left anterior descending
0735-1097 /87/$3.50
648 MIURA ET AL.
DETERMINANTS OF INFARCT SIZE
or the left circumflex artery was embolized randomly. Dogs
of either sex were included and experiments were performed
throughout the calendar year. Finally, a wide range of heart
rate and systolic blood pressure products, which provide an
estimate of myocardial oxygen consumption, were ob-
served. Although all of the dogs had a permanent coronary
occlusion, some of the hearts were analyzed 24 hours after
embolization and others were analyzed at 48 hours. We can
now retrospectively examine this data base to find out which,
if any, of the proposed determinants actually correlated with
infarct size.
Methods
Experimental preparation. Data from 92 dogs were
reviewed to examine the effects of risk zone size, sex of
the animal, coronary artery occluded, seasonal changes and
heart rate-systolic pressure product on infarct size. The dogs,
weighing 12 to 33 kg, were anesthetized with sodium pen-
tobarbital, 30 mg/kg intravenously, and I million units of
penicillin were injected intramuscularly for infection pro-
phylaxis. Dogs were intubated and spontaneously breathed
room air. The carotid artery was exposed by a midline
cervical incision and a percutaneous catheter was placed in
the jugular vein. After intravenous injection of 5,000 units
of heparin, a special cannula (18) was inserted into the
carotid artery. The cannula tip was introduced into the left
coronary ostium and a plastic bead, 2.5 or 3 mm in diameter,
was flushed through the cannula and into the coronary artery
by 10 ml of saline solution. The cannula was then withdrawn
and 5 minutes later was advanced into the left ventricle
where 20 million microspheres, 15 p.m in diameter and
labeled with 141-cerium, were injected to mark the zone at
risk. The cannula was removed, the cervical wound was
closed and the dogs were returned to their kennel. All sur-
gery was performed under sterile conditions.
In our early experience with this model we observed that
thrombi often formed behind the embolus and occluded
proximal branches. This would make the measured risk zone
appear to be smaller than it actually was. We found that we
could prevent thrombus formation behind the embolus with
heparin. Therefore, 5,000 units of heparin in 1,000 ml of
saline solution were continuously infused at 1,000 ml/24 h
through the jugular vein catheter. Of the 92 dogs, 20 had
a catheter placed in the femoral artery and a reference blood
sample was withdrawn starting 15 seconds before injection
of microspheres for 2 minutes by a Harvard syringe pump.
Calculation of volume of zone at risk and of infarct.
All dogs were killed with a sodium pentobarbital overdose
either 24 (n = 48) or 48 (n = 44) hours after embolization.
The hearts were removed and weighed and the location of
the embolus was noted. Each heart was then frozen and
sliced transversely into 4 mm thick sections using an electric
lACC Vol 9. No.3
March 1987:647-54
meat slicer. The slices were incubated in triphenyltetrazo-
lium (I % wt/vol) in phosphate buffer (100 roM, pH 7.4) at
37°C for 20 minutes to visualize the infarct (19). The tissue
slices were placed on an acrylic sheet with a layer of plastic
food wrap placed over the slices followed by a sheet of X-
ray film (Kodak no screen NS-ZT). The whole package was
placed in a light-proof box and stored at - 20°C for 48 to
72 hours to expose the film. The films were then removed
and developed. After autoradiography the heart slices on
the acrylic sheet were taken to a cold room where a clear
acetate sheet was placed on the slices and the outlines of
the sections and the infarcts were traced. Autoradiograms
clearly revealed a perfusion defect that was designated the
zone at risk (14). The volume of the zone at risk and the
infarct was calculated by planimetry.
Measurement of collateral flow. Collateral blood flow
to the zone at risk had been measured in 20 of the 92 dogs.
Three transmural samples (about 0.5 to 1.0 g each) from
the zone at risk and three from the nonischemic region were
taken from each heart. Ischemic and nonischemic samples
were cut out from the same slice using the autoradiogram
as a guide. To avoid contamination with nonischemic tissue,
samples from the zone at risk were taken at least 5 mm
inside the border of the zone at risk. Each transmural sample
was divided into endocardial, midmyocardial and epicardial
samples of approximately equal size. Three samples from
each layer were combined, weighed and counted in a gamma
counter. Collateral blood flow was expressed as a percent
of the flow of the corresponding nonischemic layer.
Absolute blood flow was calculated using the equation
F = R x TIB,
where F = myocardial blood flow (rnl/rnin per g), R
withdrawal rate of reference blood sample, T = counts/g
in tissue sample and B = total counts in blood samples
(20). In addition, we prepared another 17 dogs (10 dogs
with 24 hour and 7 with 48 hour occlusions) in which col-
lateral blood flow was measured. The surgical procedure
was the same as in the other animals but they are not included
in the retrospective analysis.
Statistical analysis. All results are expressed as mean
± I SE unless otherwise indicated. Differences between two
groups were tested by Student's t test. Linear regressions
were obtained by the least squares method and the differ-
ences between regression lines were tested by analysis of
covariance.
These protocols were reviewed and approved to be hu-
mane by the Animal Use Committee at the University of
South Alabama. Workers in the animal care facility were
authorized to euthanize any animals showing undue signs
of discomfort or postoperative complications such as hem-
orrhage or infection. All procedures were conducted in strict
accordance with the guidelines of the American Physiolog-
ical Society on animal usage.
lACC Vol. 9. No.3 MIURA ET AL. 649
March 1987:647~ 54 DETERMINANTS OF INFARCT SIZE

Table 1. Summary of Hemodynamic and Infarct Size Data for 92 Dogs

24 Hour group 4X Hour group

Value SE Value SE
Systolic pressure (mm Hg) IX'l II 1M 6
Heart rate (beats/min) 166 5 166 7
Rate-pressure product (mm Hg/rnin)» 26.06'l 1.170 27.22'l 1.202
Heart weight (gl 27'l X 222 12
Infarct (g) IO.X o.v 1'l.7 2.7
Risk zone (g) 16.4 U 25.6 3.1
Percent of ventricle infarcted \. 'l 0.3 'l.'l U
Percent of ventricle at risk 4.'l 0.4 12.3 1.6
Infarct/risk zone (0/<: l 6X.O 3.2 76.4 2.6

*Heart rate-systolic blood pressure product.

Results
Means and standard errors of hemodynamic and infarct
size data from 92 dogs are summarized in Table I . Coronary
blood flow data in 37 dogs in which flow data were available
are presented in Table 2.
Relation between risk zone size and infarct size. Be-
cause we used two different sizes of emboli and the size of
the hearts ranged from 91 to 360 g, the risk zone varied
from 4 to 124 g (1.5% to 42.3%) of the ventricular mass.
Figure I shows that the absolute size of the infarct after 48
and 24 hours of coronary occlusion correlated closely with
the absolute size of the zone at risk (Y = 0.766X - 0.236,
r = 0.919 for the 48 hour studies and Y = 0.563X +
1.541, r = 0.778 for the 24 hour studies). These findings
indicate that the size of the infarct was proportional to the
size of the zone at risk. Note that eight dogs in the 48 hour
group having a risk zone larger than 35 g were excluded
from Figure I for clarity. They were also excluded from
the F test analysis because no dogs in the 24 hour group
had values in this range and comparison with an extrapolated
part of a regression line is not valid. Those eight data points
can be seen in Figure 2, however. The slope of the 48 hour
regression line was significantly steeper than that for the 24
hour group (p < 0.01), suggesting that a greater percent of
the risk zone necrosed in the 48 hour studies.
The X axis intercepts in both groups approximated zero,
suggesting that the extent of necrosis of the risk zone was
independent of the risk zone size. This was also evident
when the infarct and risk zone were represented as a percent
of the ventricular mass rather than as an absolute mass (Y
= 0.751X - 0.273, r= 0.926 for the 48 hour group and
Y = 0.541X + 0.703, r = 0.792 for the 24 hour group).
Again the regression lines projected to the origin. When the
percent of the risk zone with infarction was plotted against
either absolute risk zone size (Fig. 2A) or risk zone as
percent of the ventricular mass (Fig. 2B), there was no
correlation. That confirmed that the size of the risk zone
had no influence on the extent to which it necrosed.
Infarct size in male versus female dogs. In the 24 hour
group, the percent of the risk zone with infarction was 62.1
± 4.4o/t: in male dogs (n = 24) and 73.0 ± 5.1 o/c in female

Table 2. Summary of Regional Coronary Blood Flow in the 37 Dogs in Which Collateral Flow
Was Measured

24 Hour group 48 Hour group

Value SE Value SE
Normal zone flow*
Subendocardial 2.00 0.15 2.05 O.I?
Midmyocardial I.XO 0.10 I. 'l3 0.15
Subepicardial UX O.O'l I.M 0.15
Collateral flow*
Subendocardial 0.16 0.05 O.O'l 0.04
Midmyocardial 0.33 0.10 0.16 0.04
Subepicardial 0.72 0.10 0.45 0.12
Collateral flow* (% normal LOne)
Subendocardial X.OO 2.10 3.70 tOO
Midmyocardial 16.70 3.60 6.40 1.60
Subepicardial 27.50 6 10 20.XO 4.50
*Flows expressed in millileters per minute per gram.
650 MIURA ET AL. lACC Vol. 9, No.3
DETERMINANTS OF INFARCT SIZE March 1987:647· 54

3lI
0 ,,48 hr /

" 0/ /
/

25 /

" ,,"/ "/ /

0
"
/ < ;
20 0
"
/" " Figure 1. Relation between the size of the zone at risk
" "" 24hr
/

W / 0 (g) and infarct size (g) after either 24 hours (circles)

N
....... I /"''' 0 or 48 hours (triangles) of coronary occlusion. Eight
UJ
I-
0"
0 0 points having risk zones greater than 35 g in the 48
U 0 hour group were omitted from the figure to allow an
a:1 0"
4:
LL " expanded horizontal axis, Those points are included in
Z Figures 2A and 2B,
.......

10 15 20 21 3lI 315

REGION AT RISK
dogs (n = 18), In the 48 hour group, the percent of risk Effect of left anterior descending versus circumflex
zone infarcting was 82.8 ± 3.4% in male dogs (n = 10) artery occlusion on infarct size. The embolus was found
and 77,5 ± 4.4% in female dogs (n = 6). In both 24 and in the left circumflex artery in 40 dogs (25 in the 24 hour
48 hour groups, there were no significant differences in this protocol and 15 in the 48 hour protocol) and in the left
percent between male and female dogs. anterior descending artery in 22 dogs (13 in the 24 hour and

0
o .0 0" 0

0 & "
W
I-
"6,,00 L"" '\ " " " "
.6. 4 AO
U
a:
4:
0
0,f~",,44 "
7 o 00 0 A 0"
LL
Z
....... "8 0
0.6..6. A
"
0 00
W 00
Z
0
N "
" 0
,,0
,fo '" 0

::.::::
UJ
....... 0 " 0

a::
0
"
<t4i!

10 20 3D 40 !ill 60 70 60 !II UII 110 120 I3lI 140

A Figure 2. Relation between the percent of risk zone

REGION AT RISK with infarction and the size of zone at risk (g), A, Risk
zone in grams (abscissa); B, risk zone size as a percent
100 o <P of ventricular mass, Circles, Hearts analyzed 24 hours
0 A
o
iDA
A A A A
after occlusion; triangles, hearts analyzed at 48 hours,
90 A
A
"0 00 " A A A A A
0 " A A A 0
WlIO 0 A " A
I-
u 00 " A Al'A
a: 70 o ~O A ~ A
A
AA A
4:
LL OA oA 0 0
Z 60 0 o 0 0
.......
o 0
A
W50 A
0 0 A
Z A
0 A 0'0 0
N ~o
0 A
:::.::: 0
UJ3lI
.......
a: 20
"
0
il"1
10

I I I I I I I I I
0 10 15 20 25 3lI 315 40 6 50

B %VENTRICULAR MASS AT RISK

JACC Vol. 9, No.3 MIURA ET AL. 651
March 1987:647 54 DETERMINANTS OF INFARCT SIZE

100
0
90 0 0
0 IP '&
LU A 0
I-
A .A +. A 0

U 80 lJA A 1111 0 0
0
A • 0
A
0:
«
A
AA
A
8 • A 0

LL
Z
>--i
70
o.
A
A
0 0
0
0

~o
""t 0
6
•
A
0 0

A
60 A 0 Q)
0
A 0 0
W 00
0
Figure 3. Percent of the risk zone with infarction plot- Z 50 0
0
0 0
ted against the day of the year that the experiment was A A

performed. No correlation could be found. Symbols as

N 00

0
A
• 0

in Figure 2. 0
0

~ 10

DAY OF THE YEAR

9 in the 48 hour group). There was no significant difference
in percent of risk zone with infarction between the dogs
with the circumflex or left anterior descending artery oc-
cluded (64.6 ± 3.7% versus 72.6 ± 7.2% in the 24 hour
group, 79.2 ± 4.2% versus 77.8 ± 5.1% in the 48 hour
group).
We also found no difference in the collateral flow to the
two regions in the subgroup with collateral flow data. Al-
though the mean collateral flow to the circumflex occlusion
beds did exceed that to the anterior descending occlusion
beds (39.0% of flow to the normal region [n = 121 versus
24.9% [n = 15]), the values had a wide range (3.5 to 49%
and 4.3 to 88%, respectively) and the difference was not
significant.
Relation between the time of the year and infarct size.
Figure 3 shows the percent of the risk zone with infarction
plotted against the day of the year for the entire data set.
The time of year had no discernible effect on the size of
the resulting infarcts.
Relation between rate-pressure product and infarct
size. We were able to calculate heart rate-systolic pressure
product at the time of occlusion for 76 of the animals. The
correlation coefficient between rate-pressure product and
percent of risk zone with infarction was 0.05 (p = NS).
Relation between collateral blood flow and infarct size.
Figure 4 illustrates the relation between infarct size and
collateral blood flow 2 minutes after coronary embolization.
Infarct size is plotted as a percent of the zone at risk and
collateral flow is expressed as flow to the subepicardial one-
third of the ventricular wall and is expressed as a percent
of that flow to the nonischemic regions. Percent necrosis
and subepicardial blood flow were highly correlated both in

tOO
• C c
CJ 90 .. c
....~ c o

.
LLJ
Figure 4. Relation between collateral blood flow t; 80- c c
00

to the subepicardium of the risk zone 2 minutes IT 0

0
.. ..
after coronary occlusion and percent of the risk ~ 70 oCc&
o
zone with infarction. Circles, 24 hour experi- Z
1---1
60
O ....OA
o
o •
ments; squares, 48 hour experiments from the o
W o
present study. The regression lines are shown for
~50
these groups (Y = 86.4 - 0.9X for 24 hour N c
group, Y = 92.1 - I.IX for 48 hour group) ~ 40
which are not significantly different (p > 0.05). If)
1---130
Closed circles and closed triangles represent ex- 0:
periments by Reimer and Jennings (2); closed tri- 1.L20
CJ o
angles represent animals experiencing 96 hours o
~ 10
of occlusion; closed circles represent those with
reperfusion at 4 hours.

COLLATERAL FLOW 1% NORMAL)

652 MIURA ET AL.
DETERMINANTS OF INFARCT SIZE
the 48 hour group (Y = 92.1 - 1.1X, r = -0.91) and
in the 24 hour group (Y = 86.4 - 0.9X, r = -0.86).
Neither the slopes nor the intercepts were significantly dif-
ferent between 24 hour and 48 hour groups.
The correlation was also strong between percent necrosis
and transmural collateral flow (Y = 91.8 - 2.5X, r =
-0.95, for the 48 hour group and Y = 86.0 - 1.8X, r =
- 0.83, for the 24 hour group). Subendocardial blood flow
also correlated with percent necrosis (r = - 0.78 and - 0.74
for 48 hour and 24 hour groups, respectively); however, the
correlation was not as tight as that for subepicardial or
transmural flow. When the collateral blood flow was ex-
pressed in absolute values in milliliters per minute per gram,
the correlations tended to be slightly poorer (r = - 0.94
and -0.77 for subepicardial flow, r = -0.92 and -0.65
for transmural flow, r = -0.69 and -0.56 for subendo-
cardial flow in the 48 hour and the 24 hour group, respec-
tively).
Discussion
The present study indicates that infarct size in this model
is directly proportional to the size of the risk zone with the
percent of the risk zone developing infarction independent
of risk zone size over a wide range. However, the percent
of the risk zone with infarction in any given dog was vari-
able. Approximately 80% of that variability could be ac-
counted for by the differences in collateral blood flow among
the animals. Neither the sex of the animal, the coronary
artery that was embolized, heart rate-systolic pressure prod-
uct at the time of coronary occlusion, nor the season of the
year exerted a discernible effect on the size of the resulting
infarcts. Because, except for the analysis for duration of
occlusion, only a univariate analysis was conducted, it is
possible that some minor correlations were obscured by the
variability introduced by the wide range in collateral flow.
However, collateral flow data were unavailable for most of
the data base; therefore, a multivariate analysis was not
feasible.
Influence of risk zone size. Several studies (5-7) using
conscious dogs and barium gelatin angiography to delineate
the risk zone found the percent of risk zone with infarction
to be strongly dependent on risk zone size. Those studies
indicate that no infarction would occur when the risk region
was less than 20% of left ventricular mass. In the present
study the zero infarct intercepts (X intercepts in Fig. I and
2) were indistinguishable from zero. Furthermore, 71 of the
dogs had a risk zone involving less than 15% of the ven-
tricular mass, well below the 20% threshold. Yet a large
percent of the risk zone developed infarction in virtually all
of those animals.
We have no clear explanation for the discrepant results.
Recently, Jugdutt (21) compared the infarct size of anes-
thetized dogs with that of conscious dogs; the risk zones
lACC Vol. 9. NO.3
March 1987:647 54
were determined by barium gelatin angiography in both
groups. Although the anesthetized dogs had a larger infarct
than the conscious dogs (56 versus 33% of risk zone), both
the conscious and the anesthetized groups demonstrated a
zero infarct intercept of about 20%. A similar finding was
reported by Reimer et al. (6) in a cooperative study of
unconscious dogs in which the risk zone had been delineated
by postmortem dye injection. Those findings suggest that
anesthesia is not the reason for the absence of a high zero
infarct intercept in the present study.
The method used to estimate the size of the risk zone
might also account for the difference. A high zero infarct
intercept could result from a systematic overestimation of
the zone at risk by certain methods. Accordingly most, but
not all (6), studies reporting a high zero infarct intercept
have used the barium angiographic method (5-7,21). It has
been proposed that the" anatomic" risk zone that is marked
by barium angiography or double dye infusion postmortem
is different from the "physiologic" risk zone that is marked
by microsphere autoradiography or dye injection in vivo,
because the former delineates the anatomic territory of the
occluded branch whereas the latter only marks the zone with
an actual flow deficit under coronary occlusion (22). Al-
though this differentiation seems obvious, no study has ac-
tually compared the anatomic and physiologic risk zone size
using all of these techniques in the same heart. However,
some data are available. Nakamura et al. (23) marked the
risk zone with radiomicrospheres injected into the coronary
artery before it was ligated; because this technique sharply
delineates the field of the artery, their risk zone estimations
should have been anatomic. In that study, the zero infarct
intercept of the risk zone was near zero. Furthermore, post-
mortem dye infusion and microsphere autoradiography in a
single heart yield the same risk zone size in our experience
(15). The differentiation into anatomic versus physiologic
risk zone is hardly a clear one and certainly does little to
explain these discrepant results. Whatever the explanation,
risk zone size does not influence the extent of necrosis in
this model.
Influence of the site of occlusion on infarct size. Becker
et al. (5) reported a clear difference in the extent of necrosis
between left anterior descending versus circumflex artery
occlusions with their conscious model incorporating barium
angiography. No such differences were apparent in our study,
however, even when collateral flow was taken into account.
Thus, in our model both arteries seem to be equally vul-
nerable to infarction. Again we have no explanation for the
discrepant results.
Relation between collateral flow and infarct size. The
very close correlation between percent necrosis of the risk
zone and collateral blood flow measured 2 minutes after
coronary occlusion indicates that blood flow through native
collateral channels is the major determinant of the extent of
infarction within the risk zone. On this point, our results
lACC Vol. 9, No.3
March 1987:647 54
corroborate the work of Reimer and Jennings (2); they also
described a close linear relation between percent necrosis
of the risk zone and epicardial flow in their open chest
model. In that model, the circumflex branch was either
occluded for 6 hours and reperfused for 4 days or simply
ligated for 4 days. Although their method of producing
coronary occlusion was quite different from ours, their
regression line for percent infarction versus epicardial col-
lateral blood flow was virtually identical to those presented
here. Data from their study are plotted along with ours in
Figure 4. This relation should be very useful because it
seems to be remarkably reproducible.
Relation between the duration of ischemia and infarct
size. Analysis of the data in Figure I indicates that the
infarcts were slightly larger in the 48 hour experiments than
in the 24 hour studies. This suggests that the infarcts had
not reached their ultimate size by 24 hours. This difference
was not supported, however, by the collateral blood flow
plots in Figure 4. In those studies the regression lines were
not different in either slope or intercept; apparently several
more dogs with rich collateral circulation were fortuitously
included in the 48 hour protocols.
Influence of cardiac metabolism on infarct size. When
the transmural collateral flow was expressed as a percent of
normal zone flow, it gave a slightly better correlation coef-
ficient than when it was expressed as absolute flow. That
could be the result of an influence of oxygen demand at the
time of occlusion. As pointed out by Nienaber et al. (12),
collateral flow expressed as a percent of normal flow should
be a good approximation of supply/demand ratio in the
ischemic area and may explain the slightly better fit. Another
explanation may be that technical shortcomings in our ref-
erence blood sample make our absolute flow measurements
less precise than those normalized against normal zone flow.
We were unable to find any correlation between infarct
size and rate-pressure product. The recent study by Reimer
et al. (6) also reported that the rate-pressure product seemed
to be of influence in the conscious dog model but not in the
anesthetized model. Pentobarbital may increase oxygen de-
mand to such a high level that it no longer influences myo-
cardial necrosis. Another possibility is that the rate-pressure
product value measured at the time of coronary occlusion
is simply not an accurate enough estimate of the myocardial
oxygen demand during the entire infarction process. Which-
ever is the case, the rate-pressure product at the time of
coronary occlusion was not found to be a predictor of infarct
size in our anesthetized model.
Conclusions. Myocardial infarcts reach their ultimate
size in the first 24 hours, and the size of the infarct is directly
proportional to the size of the zone at risk. The percent of
the zone at risk that develops infarction is independent of
the size of the risk zone but is closely related to the level
of collateral blood flow shortly after coronary occlusion.
Neither the sex of the animal, the season nor the location
MIURA ET AL. 653
DETERMINANTS OF INFARCT SIZE
of the coronary occlusion discernibly influence the size of
the infarct.
References
I. Lowe lE, Reimer KA, Jennings RB. Experimental infarct size as a
function of the amount of myocardium at risk. Am 1 Pathol 1978;90:
363-79.
2. Reimer KA, Jennings RB. The wavefront phenomenon of myocardial
ischemic cell death. 11. Transmural progression of necrosis within the
framework of ischemic bed size (myocardium at risk) and collateral
flow. Lab Invest 1979;40:633-44.
3. Bishop SP, White FC', Bloor CM. Regional myocardial blood flow
during acute myocardial infarction in the conscious dog. Circ Res
1976;38:429- 38.
4. Rivas F, Cobb FR, Bache Rl, Greenfield le. Relationship between
blood flow to ischemic regions and extent of myocardial infarction.
Serial measurement of blood flow to ischemic regions in dogs. Circ
Res 1976;38:439-47.
5. lugdutt BI, Hutchins GM, Bulkley BH, Becker Le. Myocardial in-
farction in the conscious dog: three-dimensional mapping of infarct,
collateral flow and region at risk. Circulation 1979;60: 1141-50.
6. Reimer KA, Jennings RB, Cobb FR, et al. Animal models for pro-
tecting ischemic myocardium: results of the NHLBI cooperative study.
Comparison of unconscious and conscious dog models. Circ Res 1985;56:
651-65.
7. Koyanagi S, Eastham CL, Harrison DG. Marcus ML. Transmural
variation in the relationship between myocardial infarct size and risk
area. Am 1 Physiol I982;242:H867-74.
8. Becker LC, Schuster EH, lugdutt BI, Hutchins GM, Bulkley BH.
Relationship between myocardial infarct size and occluded bed size
in dogs: difference between left anterior descending and circumflex
coronary artery occlusions. Circulation 1983;67:549-57.
9. Romson JL, Hook BG, Rigot YH, Schork MA, Swanson DP, Lucchesi
BR. The effect of ibuprofen on accumulation of indium-III-labeled
platelets and leukocytes in experimental myocardial infarction. Cir-
culation 1982;66: 1002-11.
10. Romson Jl., Hook BG, Kunkel SL, Abrams GD, Schork MA, Luc-
chesi BR. Reduction of the extent of ischemic myocardial injury by
neutrophil depletion in the dog. Circulation 1983;67:1016-23.
II. lolly SR, Kane Wl, Bailie MB, Abrams GD, Lucchesi BR. Canine
myocardial reperfusion injury. Its reduction by the combined admin-
istration of superoxide dismutase and catalase. Circ Res 1984;54:
277 85.
12. Nienaber C, Gottwik M, Winkler B, Shaper W. The relationship
between the perfusion deficit, infarct size and time after experimental
coronary artery occlusion. Basic Res Cardiol 1983;78:210-26.
13. Yellon OM, Hearse 01, Maxwell MP, Chambers DE, Downey 1M.
Sustained limitation of myocardial necrosis 24 hours after coronary
artery occlusion: verapamil infusion in dogs with small myocardial
infarcts. Am 1 Cardiol 1983;51: 1409- 13.
14. Chambers DE, Yellon DM, Hearse Dl , Downey 1M. Infarct size or
delay? Flurbiprofen limits apparent infarct size after 6 hours of coro-
nary occlusion in the dog, but not after 24 hours of occlusion. Am 1
CardioI1983;51:884-90.
15. Yoshida S, Downey 1M, Chambers DE, Hearse or, Yellon DM.
Nifedipine limits infarct size for 24 hours in closed chest coronary
emboli zed dogs. Basic Res Cardiol 1985;80:76-87.
16. Yoshida S, Downey JM, Yellon OM, Miura T, Iimura O. Diltiazem
reduced infarct size but not ventricular arrhythmias in 48 hour coronary
embolized dogs. Can 1 Cardiol 1985;1:346-53.
17. Akizuki S, Yoshida S, Chambers DE, et al. Infarct size reduction by
the xanthine oxidase inhibitor, allopurinol, in closed-chest dog. Car-
diovasc Res 1985;19:686-92.
654 MIURA ET AL.
DETERMINANTS OF INFARCT SIZE
18. Chagrasulis RW, Downey JM. Selective coronary embolization in
closed chest dogs. A~ J Physiol 1977;233:335-7.
19. Nachlas MM, Shnitka TK. Macroscopic identification of early myo-
cardial infarction by alterations in dehydrogenase activity. Am J Pathol
1963:42:379-405.
20. Heyman MA, Payne BD, Hoffman JIE, Rudolph AM. Blood flow
measurements with radionuclide-labelled particles. Prog Cardiovasc
Dis 1977;20:55-79.
21. Jugdutt BI. Different relations between infarct size and occluded bed
lACC Vol. 9, No.3
March 1987:647 54
size in barbiturate-anesthetized versus conscious dogs. J Am Coli
Cardiol 1985;6:1035-46.
22. Reimer KA, Jennings RB. Can we really quantitate myocardial cell
injury? In: Hearse OJ, Yellen OM, eds. Therapeutic Approaches to
Myocardial Infarct Size Limitation. New York: Raven Press, 1984:
163-84.
23. Nakamura M, Tomoike H, Sakai K, Ootsubo H, Kikuchi Y. Linear
relationship between perfusion area and infarct size. Basic Res Cardiol
1981;76:438-42.