400

Effect of Coronary Artery Occlusion onRegional Arterial and Venous O2 Saturation,

O2 Extraction, Blood Flow, and O2

Consumption in the Dog Heart

HARVEY R. WEISS

SUMMARY The effects of coronary artery ligation on regional arterial and venous O2 saturation,oxygen extraction, blood flow, and oxygen consumption were studied in occluded and unoccluded areasof the hearts of fourteen anesthetized open-chest dogs. In seven animals, a coronary artery wasoccluded for 10 minutes and in seven others a vessel was ligated for 2 hours. Microspectrophotometricobservations of small regional arteries and veins in quick-frozen hearts to determine regional O2extraction were combined with regional blood flow measurements with radioactive microspheres todetermine regional myocardial O2 consumption by the Fick principle. Flow was significantly lower inthe occluded compared to the control area at both times. The subendocardial: subepicardial flow ratiowas reversed at 2 hours in the occluded area. Oxygen extraction was greater in the occluded areas.Oxygen consumption was lower in the occluded area. At 2 hours, the subendocardial: subepicardialconsumption ratio was reversed in the occluded area, indicating a greater decrement in consumptionin the subendocardium. Measurements of arterial saturation indicate an increasing number of bloodvessels with O2 saturations below 80% with coronary artery occlusion. These vessels were found in alloccluded areas. This would indicate a marked heterogeneity of blood flow within the area of occlusion.Some vessels may have normal flow and others low or no flow. There was heterogeneity of flow andoxygen extraction which led to areas with relatively normal oxygen consumption in the core of theinfarct. CircRes 47: 400-407, 1980

MYOCARDIAL infarction leads to a reduction inblood flow and oxygen consumption and an increasein oxygen extraction in the affected region of theheart (Blair, 1969; Marshall et al., 1974; Obeid etal., 1972; Owen et al., 1970; Ramakrishna et al.,1975; Weiss and Lipp, 1979). The infarct whichproduces the ischemia lowers tissue Po2, causes aloss of cellular constituents, accumulation of anaer-obic end products, rupture of lysosomal mem-branes, and ultimately cell death and necrosis(Blair, 1969; Obeid et al., 1972; Owen et al., 1970;Apstein et al., 1977; Khuri et al., 1975; Kloner et al.,1977; Lipp and Weiss, 1978). Within the zone ofinfarction, myocardial blood flow appears to beaffected more adversely in the deeper, subendocar-dial region than in the more superficial subepicar-dial region of the left ventricular free wall (Beckeret al., 1973; Hoffman, 1978). The present study wasdesigned, in part, to determine for the first time theregional oxygen extraction and oxygen consumptionwithin an area of ischemia. Under control condi-tions, it has been found that oxygen extraction and

From the Department of Physiology and Biophysics, College of Med-icine and Dentistry of New Jersey, Rutgers Medical School, Piscataway,New Jersey.

This project was supported by Grant HL-21172 from the NationalHeart, Lung and Blood Institute.

Address for reprints: Dr. Harvey R. Weiss, CMDNJ-Rutgers MedicalSchool, Department of Physiology & Biophysics, P.O. Box 101, Piscata-way, New Jersey 08854.

Received October 25, 1979; accepted for publication April 8, 1980.

oxygen consumption are greater in the deeper, sub-endocardial region (Weiss and Sinha, 1978; Weisset al., 1978a).

During an infarction, blood flow may be quiteheterogeneous. There have been claims for the ex-istence of a border zone of reduced flow, althoughthis has been questioned (Hearse et al., 1977; Hirzelet al., 1977). Within the core of the infarct, thereare areas where, if flow is later restored, no returnof blood flow is found (Camilleri and Fabiani, 1977).This "no reflow" phenomenon, may lead to areas ofearly flow heterogeneity in the infarct. In the pres-ent report, I have studied this effect during is-chemia. Spatial heterogeneity of flow also has beenobserved in normal heart tissue (Marcus et al., 1975,Marcus et al., 1977).

To examine the relationship between regionaloxygen supply and demand in an ischemic area, aquantitative technique to measure regional myo-cardial O2 consumption was employed. Microspec-trophotometric observations of small regional ar-teries and veins in quick-frozen hearts were madeto determine regional O2 extraction and were com-bined with regional measurements of blood flowwith radioactive microspheres to determine re-gional myocardial O2 consumption (Sinha et al.,1975; Sinha et al., 1977; Weiss and Sinha, 1978;Weiss et al., 1978a). Thus for the first time, therelation between oxygen supply and consumptioncould be determined quantitatively in an area of

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

REGIONAL O2 CONSUMPTION IN INFARCTS/Weiss 401

infarction. The microspectrophotometric observa-tion of regional arterial O2 saturation was also em-ployed to give an estimate of regional flow hetero-geneity.

MethodsThis study was conducted on 14 adult mongrel

dogs of either sex ranging in weight between 16 and27 kg. The animals were anesthetized with sodiumpentobarbital, 30 mg/kg, iv, and this was supple-mented as required. The trachea was intubated andartificial ventilation was instituted with a HarvardPump. The FA(O was monitored and maintainedconstant by adjusting ventilation. A femoral arterywas catheterized and the catheter was advancedinto the aortic arch. This catheter was used torecord aortic blood pressure and heart rate and forarterial blood sampling.

Under artificial ventilation, the chest was openedat the 5th interspace on the left side. The pericar-dium was opened and tied back. A catheter wasplaced in the left atrium and another was placed inthe left ventricular cavity through the apical dim-ple. The left ventricular catheter was used to recordpressure and the rate of pressure change, dp/dt. Acatheter was placed in the right external jugularvein and advanced at least 2 cm into the coronarysinus. An interval of at least 30 minutes was allowedfor the preparation to stabilize.

In the control period, heart rate, blood pressureand dp/dt were recorded on a Beckman R411 re-corder. Anaerobically obtained arterial and coro-nary sinus blood samples were used for analysis ofblood gases, pH (Instrumentation Laboratory,model 113) and hematocrit and hemoglobin concen-tration (Fisher Hemophotometer). The main trunkof the left anterior descending artery was then tiedoff. The tie was placed below at least one majorbranch to keep the size of the experimental infarctsmall. No animals used in these experiments hadvisible surface collaterals to the area supplied bythe left anterior descending artery. Ten minutesafter occlusion of the coronary artery, recordingsand blood samples were obtained. A dose of ap-proximately 1-2 million carbonized 85Sr-labeled mi-crospheres, 15 ± 3 jtim in diameter (3M Co.) wasinjected as a bolus of approximately 0.5 ml into theleft atrial catheter and then flushed with 5 ml ofsaline in seven of the experimental animals. A ref-erence sample method was used to obtain myocar-dial flow measurements (Buckberg et al., 1971). Thesample was obtained from the femoral artery cath-eter with a peristaltic pump set at 7 ml/min.

When radioactive microsphere injections hadbeen completed, the hearts were fibrillated to arrestblood flow during the freezing process. Immediately,with a large pair of shears, I cut the ventricles belowthe atrioventricular ring and dropped them intoliquid nitrogen-cooled liquid propane. It has beenshown that the time for fibrillation and freezing isso short that no changes occur in arterial or venous

oxygen saturation (Weiss and Sinha, 1978). Thefrozen hearts were stored at -70°C until analyzed.In the other seven dogs, the occlusion was contin-ued for 2 hours. Recordings and blood samples werethen obtained. A dose of 1-2 million carbonized85Sr-labeled microspheres was injected and a refer-ence blood sample was obtained. The hearts werethen removed as described.

Immediately adjacent duplicate transmural sam-ples of the left ventricular free wall were cut fromthe center of the ischemic area and also from anarea unaffected by ligation of the blood vessel.Samples were prepared for analysis of regional mi-crosphere distribution and for microspectrophoto-metric analysis as described previously (Weiss etal., 1978a). Briefly, 30-/tm thick frozen tissue sec-tions were cut on a cold rotary microtome in a-25°C cold box. Each section was then transferredto a precooled slide, covered with deoxygenatedsilicone oil, and rapidly transferred to the micro-spectrophotometer cold stage. Arteries and veins,20 to 150 /xm in diameter, were located in the regionsof interest and absorbances at 560, 523, and 506 nmwere obtained to give O2 saturation of the bloodcontained within the vessels. Only vessels seen intransverse sections were studied, so that the lightpath was only through blood. In all, 1257 vesselswere examined microspectrophotometrically. Thisconsisted of between seven and 10 arteries andseven and 10 veins in the subepicardial, middle, andsubendocardial regions of the occluded and nonoc-cluded regions. The adjacent tissue samples wereprepared for blood flow determination. Blood flowwas determined in ml/min per 100 g (Buckberg etal., 1971).

Regional oxygen extraction (ml O2/IOO ml blood)was calculated in the control area as the localarteriovenous difference multiplied by the arterialhemoglobin concentration times the maximal O2combining capacity of 1.36 ml 02/g hemoglobin.Oxygen extraction for the occluded region was ob-tained from the unoccluded regional arterial O2saturation less the occluded regional venous O2saturation. Using the Fick principle, the oxygenconsumption for the regions of interest was deter-mined as the product of O2 extraction and regionalblood flow (Weiss et al., 1978a). The regional ratioof O2 supply to consumption was determinedby dividing the local O2 supply by local O2 con-sumption, Cao2 x Q/Q X (Cao2 - Cvo2), whereCao2 and Cvo2 are arterial and venous O2 contentand Q is the blood flow. This reduces toSao2/(Sao2 - Svo2), where Sao2 and Svo2 are thepercent oxyhemoglobin in the arterial and venousblood, respectively.

Using a repeated measure design, I employedfactorial analysis of variance to determine whetherdifferences existed between areas, occluded andnonoccluded and regionally with depth, for arterialand venous O2 saturations, oxygen extraction, bloodflow and oxygen consumption in the hearts after 10

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

402 CIRCULATION RESEARCH VOL. 47, No. 3, SEPTEMBER 1980

TABLE 1 Effect of Occlusion on Pressures, dp/dt, and Heart Rate

10-minute occlusiongroup (n = 7):

ControlOcclusion (10-min)

2-hour occlusiongroup (n = 7):

ControlOcclusion (10-min)Occlusion (2-hr)

123119

127117128

AP(mm Hg)

± 19/100 ±+ 25/98 ±

+ 20/99 ±± 18/93 ±± 22/100 ±

1922

141312

119112

121110114

LVP(mm Hg)

± 21/4 ±± 21/7 ±

± 20/1 +± 17/2 ±+ 14/2 ±

45

232

dp/dt(mm Hg/sec)

2191 ± 8811846 ± 815

2210 + 8121991 ± 6011852 ± 484

HR(beate/min)

177 + 29171 ± 26

165 ±38173 ± 20180 ±26

All values are expressed as mean ± SD. AP = aortic pressure; LVP = left ventricular pressure; dp/dt =positive dp/dt; HR = heart rate.

minutes and 2 hours of occlusion. The statisticalsignificance of the differences was determined bythe Student-Newman-Keuls procedure (Steel andTorrie, 1960). A value of P < 0.05 was accepted assignificant. The distribution of the data was testedfor normality using the modified Kolmogorov-Smir-nov statistic.

ResultsAt the onset of the occlusion, there were small

initial falls in blood pressure and dp/dt. By 10minutes postocclusion, however, pressures had re-turned to the control level (Table 1). The bloodpressure and heart rate remained unaltered for thenext 2 hours. Blood gasses and pH in arterial andcoronary sinus blood were not altered by occlusionwhen measured at 10 minutes or 2 hours postocclu-sion (Table 2).

Myocardial Blood FlowOcclusion significantly lowered regional coronary

blood flow (Fig. 1). In the area affected by ligationof the left anterior descending artery, mean trans-mural blood flow was 23.1 ± 28.6 (mean ± SD) ml/min per 100 g, and the flow in the unoccluded areawas 99.2 ± 38.0 ml/min per 100 g after 10 minutes.After 2 hours of occlusion, flow was 29.1 ± 36.4 ml/min per 100 g in the occluded area and 109.1 ± 42.7ml/min per 100 g in the unoccluded area. At 10minutes, there was no significant difference withdepth in either the occluded or unoccluded area.After 2 hours, the occluded area had a subendocar-

dial flow of 18.8 ± 38.3 ml/min per 100 g and asubepicardial flow of 42.9 ± 42.2 ml/min per 100 g.This difference was not significant due to the greatvariability of the flow values. In the unaffected areaat 2 hours, blood flow was significantly higher inthe subendocardial region compared to the subepi-cardial region. The subendocardial-to-subepicardialflow ratio was significantly lower in the occludedarea, 0.44 ± 0.52 compared to the control area, 1.24± 0.12, at 2 hours. There was no significant differ-ence in the ratio at 10 minutes postocclusion.

Arterial O2 Saturation, Venous O2Saturation, and O2 Extraction

In the area unaffected by ligation of the leftanterior descending artery, arterial saturation av-eraged 93.5 ± 3.3% in the animals occluded for tenminutes and 94.3 ± 2.6% in the dogs occluded for 2hours. There was no significant difference in arterialO2 saturation with vessel size in the unoccludedarea, nor was there a subepicardial vs. subendocar-dial difference. When the unaffected area in all 14animals were examined, only 3.9% of the vesselshad O2 saturations below 80% (Fig. 2). Occlusionsignificantly lowered arterial O2 saturation in theaffected area compared to the control area. In theaffected area, arterial O2 saturation was lowered to86.1 ± 5.2% after 10 minutes of occlusion and to 81.2± 6.4% in the area occluded for 2 hours. Thesevalues were significantly below control and differentfrom each other. There was some tendency forsubendocardial arteries to have a slightly lower

TABLE 2 Arterial and Coronary Sinus Blood Gases and pH

Arterial Coronary sinus

10-minute occlusion group:ControlOcclusion (10-min)

2-hour occlusion group:ControlOcclusion (10-min)Occlusion (2-hr)

Po2

(mm Hg)

70+ 1169 + 7

75 + 472 ±577 + 6

Pco2

(mm Hg)

37 ± 235 ± 2

36 ± 635 ± 633 ± 5

pH

7.41 ±7.42 ±

7.42 ±7.39 +7.44 ±

0.050.06

0.090.080.05

Po2

(mm Hg)

25 ±623 ±4

21 + 219 ±520+ 1

Pco2

(mm Hg)

49 ± 547 ± 6

48± 748 ± 746 ± 6

PH

7.35 ± 0.057.36 ± 0.05

7.35 ± 0.107.38 ± 0.097.39 ± 0.08

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

REGIONAL O2 CONSUMPTION IN INFARCTS/ Weiss 403

120

110

100

s 90

o

I 80Ei TO

o 60o

50

30

20

10

• Io NON-OCCLUDED REGION

• OCCLUDED REGION

EPI MID ENDOOCCLUSION-IOmin.

EPI MID ENDOOCCLUSION-2hr

FIGURE 1 Regional coronary blood flow in the subep-icardial (EPI), middle (MID), and subendocardial(ENDO) regions of the left ventricular free wall. Flowsin the occluded and unoccluded regions are depicted forthose animals occluded for 10 minutes and 2 hours(mean ± SD).

saturation, 81%, compared to subepicardial arteries,85%, but the difference was not significant (P =0.06). There was a large statistically significant in-crease in the number of vessels having saturationsbelow 80% both at 10 minutes and more so at 2hours postocclusion (Fig. 2).

Venous O2 saturation was significantly loweredby occlusion at 10 minutes, 34.4 ± 6.2% unoccludedvs. 25.4 ± 6.7% occluded, and also at two hours, 37.7± 5.7% vs. 25.3 ± 5.2%. There were no significantdifferences with depth in the occluded region withrespect to venous O2 saturation at either time. Theunoccluded area in the dogs occluded for 10 minuteshad a significant subepicardial-to-subendocardialvenous O2 saturation gradient, 36.2 ± 4.4% vs. 31.2± 4.3%. The difference was not significant in thecontrol area at 2 hours, 38.9 ± 6.0% vs. 35.4 ± 4.6%.The histograms of venous O2 saturation in variousregions of the occluded and nonoccluded area areshown in Fig. 3. Note that even after 2 hours ofocclusion some veins have high O2 saturations.

The amount of oxygen extracted from the arterialblood was significantly greater in the occluded areacompared to the nonoccluded area (Fig. 4). Approx-imately 15% more oxygen was removed from theblood in the occluded area compared to the nonoc-cluded area at 10 minutes and approximately 21%greater regional O2 extraction was found at 2 hours.

In the occluded area, no subepicardial-to-subendo-cardial gradient in O2 extraction was found eitherat 10 minutes or 2 hours. In the group occluded for10 minutes, subepicardial O2 extraction was signifi-cantly lower than subendocardial in the controlarea (13.1 ± 2.4 vs 14.3 ± 3.7 ml O2/100 ml blood).The difference was not statistically significant inthe control area of the 2-hour occlusion group.

Myocardial O2 ConsumptionOxygen consumption was significantly lower in

the occluded areas than in the nonoccluded areas(Fig. 5). The occluded area had only 29% of the O2consumption of the nonoccluded area. Time of oc-clusion did not affect the difference in consumption.In the occluded area, there was no subepicardial-to-subendocardial O2 consumption difference at eithertime. After 10 minutes of occlusion, the unoccludedarea exhibited no depth effect with regard to O2consumption. At 2 hours, however, subepicardial O2consumption was significantly below that of thesubendocardial region. The average oxygen con-sumption of the normal area was 12.6 ± 3.8 ml O2/min per 100 g at 10 minutes and 12.4 ± 8.7 at 2hours, whereas the average consumption in theoccluded area was 2.8 ± 4.6 02/min per 100 g at 10minutes and 3.8 ± 5.1 at 2 hours. The subendocar-dial-to-subepicardial O2 consumption ratio was sig-nificantly higher in the unoccluded area, 1.31 ±0.26, than in the occluded area, 0.45 ± 0.56, at 2hours. This reflected differences in the flow ratio.Differences in the O2 consumption ratio were notsignificant at 10 minutes. The data describing re-gional blood flow, arterial and venous O2 saturation,extraction and consumption showed a normal dis-tribution.

The relation of O2 supply to consumption as

60uj

= = 5 0

10 MIN OCCLUDED REGION 2 HBS OCCLUDED REGION

"0 JO 40 60 80 ICO 0 20 40 60 80 100

% ARTERIAL 0 , SATURATION

20 40 60 80 100

FIGURE 2 Histograms of arterial O2 saturations foundin small regional vessels of occluded and nonoccludedareas of the left ventricular myocardium. In panel A, allarterial O2 saturations found in the subepicardial, mid-dle, and subendocardial regions of normal myocardiumare displayed. In panel B, all arterial O2 saturations inthe occluded region of the seven dogs in which flow wasstopped for 10 minutes are shown. Panel C, shows allarterial O2 saturations in the occluded region of thedogs in which flow was stopped for 2 hours. Note thelower arterial O2 saturation with increasing length ofocclusion.

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

404 CIRCULATION RESEARCH VOL. 47, No. 3, SEPTEMBER 1980

30

20

10

0

40LLJ

^ 30UJ

cc§ 20oO 10

0

40

30

20

10

NONOCCLUDED REGION

EPI

n..MID

1 . .ENDO

rrrr20 40 60 80 100

10 MIN. OCCLUDED REGION

EPI

MID

ENDO

2 HRS. OCCLUDEDREGION

EPI

MID

ENDO

20 40 60 80 1000 20 40 60 80 100

% VENOUS 0* SATURATION

FIGURE 3 Histograms of venous O2 saturations found in small veins of the left ventricular subepicardium (EPI),middle (MID), and subendocardium (ENDO) in occluded and nonoccluded areas. The histogram of the nonoccludedregion is from data from 14 hearts, whereas the histogram of both the 10-minute and 2-hour occluded regions are fromseven dogs each. Note that, although the means are lower in the occluded areas, there are still many vessels with ahigh venous O2 saturation.

expressed by the 02 supply/O2 consumption ratio isgiven in Table 3. Whereas subendocardial ratioswere slightly lower than corresponding regionalsubepicardial ratios, these differences were not sig-nificant in the nonoccluded area. Occlusion, how-ever, significantly lowered the O2 supply :O2 con-sumption ratio both in the 10-minute and 2-hourgroups, in the three transmural regions studied.

Discussion

When a vessel supplying a portion of the leftventricular free wall is ligated, there is an immedi-ate fall in flow in the affected area. This leads to arapid fall in local contractility, dp/dt, work, andelectrical activity in this area (Marshall et al., 1974;Ramakrishna et al., 1975; Khuri et al., 1975). Venous02 saturation falls in the veins draining the area(Blair, 1969; Marshall et al., 1974; Obeid et al., 1972;Owen et al., 1970; Ramakrishna et al., 1975; Weissand Lipp, 1979). Venous CO2 tension also is in-creased, and there are changes in lactate, pyruvate,glucose, potassium, phosphate, and pH (Blair, 1969;Marshall et al., 1974; Obeid et al., 1972; Owen et al.,1970; Weiss and Lipp, 1979). Within the tissue,there is a fall in tissue Po2 and a rise in tissue Pco2(Khuri et al., 1975; Lipp and Weiss, 1978). Theseeffects and the fall in blood flow seem more severein the deeper, subendocardial region of the affectedarea (Khuri et al., 1975; Lipp and Weiss, 1978;Becker et al., 1973; Hoffman, 1978).

1 bl

ood

EOO

E

00

1

20

18

16

14

12

10

8

6

4

2

ft

-

-

•

O1

10

i

1

I i I0

1 ?

0 NON-OCCLUDED REGION

• OCCLUDED REGION

I01

EPI MID ENDO EPI MID ENDOOCCLUSION-IOmin. OCCLUSION- 2hr.

FIGURE 4 Regional myocardial oxygen extraction(Cao2 —CvoJ in the subepicardial, middle, and suben-docardial regions of the left ventricular free wall. Oxy-gen extraction in the occluded and unoccluded regionsis shown for dogs in which the left anterior descendingartery was occluded for 10 minutes and 2 hours (mean± SD).

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

REGIONAL O2 CONSUMPTION IN INFARCTS/ Weiss 405

20

18

16

H |4o1I 12

10

|oo

z 6O(S>

4

2

0

O NON-OCCLUDED REGION

• OCCLUDED REGION

I i

EPI MID ENDO EPI MID ENDOOCCLUSION-10min. OCCLUSION - 2 hr.

FIGURE 5 Regional myocardial oxygen consumption(Vo-z) in the subepicardial, middle, and subendocardialregions of the left ventricular free wall. Oxygen con-sumption in the occluded and unoccluded regions isdepicted for animals in which the left anterior descend-ing artery was occluded for 10 minutes and 2 hours(mean ± SD).

Although occlusion of the left anterior descend-ing artery lowered blood flow in this area, there wasno subepicardial-to-subendocardial gradient ineither the occluded or nonoccluded area after 10minutes of ligation (Fig. 1). In dogs with the arteryoccluded for 2 hours, there was a significant reversalof the ENDO/EPI flow ratio, which was 1.24 ± 0.12in the nonoccluded area and 0.44 ± 0.52 in theoccluded area. In a normal heart, microspheres ofthis size tend to show a slightly higher subendocar-dial than subepicardial flow (Gross and Winbury,1973). Occlusion dramatically reverses this finding(Becker et al., 1973; Hoffman, 1978). I then studiedthe differences in flow distribution within the areaof occlusion through examination of the arterial O2saturation.

In the control area, less than 4% of the arterieswere found to have saturations of less than 80%.The low O2 saturations were unrelated to vessel sizeor depth within the heart. This has been reportedpreviously (Weiss and Sinha, 1978). In the occludedarea, arterial O2 saturation was lower than in theunaffected area at 10 minutes and even lower at 2

hours. This is clear evidence for spatial heteroge-neity of blood flow. The arterial wall provides asignificant barrier to diffusion of oxygen into themyocardium. It has been shown that the oxygensaturation of large arteries is the same as small onesin the heart (Weiss and Sinha, 1978), indicating noloss of oxygen. Further, in fibrillating hearts andskeletal muscle, no loss of oxygen was seen in arter-ies greater than 20 /tin in diameter for at least 2minutes (Weiss and Sinha, 1978). Since, after occlu-sion for at least 10 minutes, some myocardial arter-ies had quite low O2 saturations, this indicates thatblood flow in these vessels must have been veryslow or completely stopped. The evidence indicatesthat O2 saturation would fall only in myocardialarteries after considerable flow reduction. This lowflow state occurs in only some arteries, since othershave normal O2 saturations. The percentage of ar-teries with low O2 saturation, and therefore lowflows, is higher at 2 hours than at 10 minutes. Thealternate explanation for the low arterial O2 satu-ration, i.e., a reduction in the external wall diffusionbarrier, appears unlikely especially after only 10minutes of ischemia. Changes in the partial pressureof gases in the extracellular fluid could well affectarterial O2 saturation as well. Large microregionaldifferences in the partial pressure of O2 and CO2could only occur, however, if there were large vesselby vessel variations in flow.

There is considerable evidence for spatial inho-mogeneity of flow within a region of myocardium,e.g., subendocardium (Marcus et al., 1975; Marcuset al., 1977). These investigators have found thattissue samples weighing about 1 g may have verydifferent flows even in the same myocardial region.Steenbergen et al. (1977), using NADH fluorescencephotography, demonstrated on the surface of is-chemic and hypoxic saline-perfused rat hearts, dis-crete heterogeneous anoxic zones. The measure-ments of arterial O2 saturation indicate, for the firsttime, that these differences in flow exist within anarea of infarction. Furthermore, this inhomogeneity

TABLE 3 02 Supply/O2 Consumption Ratio[Calculated as Sao2/(Sao.,, —Svo.J

Occluded regionEPI

MID

ENDO

Non-occluded regionEPI

MID

ENDO

Postocclusion

10 min

1.42*±0.19

1.36±0.12

1.37±0.14

1.65±0.14

1.63±0.21

1.50±0.16

2 h r

1.39±0.12

1.40±0.15

1.36±0.04

1.72±0.20

1.73±0.26

1.60±0.11

All values are expressed as mean ± SD.

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

406 CIRCULATION RESEARCH VOL. 47, No. 3, SEPTEMBER 1980

exists on a vessel-by-vessel basis and, with time,more of the vessels appear to have low flow. Neitherthe heterogeneity demonstrated by these investi-gators nor the possibility of vessel by vessel flowdifferences suggested in the present report shouldaffect regional blood flow measurements with radio-active microspheres. Flow distribution is normaland samples obtained should reflect regional aver-ages (Steel and Torrie, 1960). We also have esti-mated that the number of microspheres in the lowflow samples is greater than 500, which shouldmaintain reasonable accuracy (Buckberg et al.,1971).

While I investigated flow and arterial O2 satura-tion in the obvious center of an area of low flow, itwas possible that some small areas of normal flowwere present (Hirzel et al., 1977). These areas ofnormal flow could explain the many vessels with ahigh arterial O2 saturation. There are many reportsof collateral vessels and arterial anastomosisthrough which blood can be supplied to the areanormally supplied by the ligated vessel (Berne andRubio, 1969; Cohen, 1978; Eliska and Eliskova,1970). This appears to be especially probable indogs. Distribution of collateral flow within the areaof infarction also could be inhomogeneous due toslight differences in path length, pressure differ-ences, vessel sizes, regional work, or regional bulg-ing. It appears that the number of arteries with lowO2 saturation is increased in the group studied 2hours after occlusion. This may be related to edema,tissue swelling, or other changes that block arterialflow (Kloner et al., 1977). this may also be relatedto the "no-reflow" phenomenon (Camilleri and Fa-biani, 1977), in which, when reinstituted after aperiod of time, blood flow does not return uniformlyto reperfuse all of the blocked vessels.

In heart, tissue Po2, and venous and capillary O2saturation, are lower in the subendocardial regioncompared to the subepicardial region in controlanimals (Weiss and Sinha, 1978; Weiss et al., 1978b;Holtz et al., 1977). In the nonoccluded area in thepresent study, venous O2 saturation is lower at 10minutes postocclusion in the subendocardium butnot at 2 hours. Some of the differences between thepresent findings at 2 hours in the control area andother studies with normal hearts may be due to theeffect of the infarct. While pressure appears unal-tered (Table 1) the level of circulating catacholam-ines may be higher during infarction. The unaf-fected region may be doing slightly more work tocompensate for the nonworking area.

In the occluded area, venous O2 saturation wassignificantly lower than that found in the controlarea, and no regional differences were observed.Investigators measuring the O2 saturation throughcatheterization of a small vein draining the in-farcted area have also found a sharp fall in venoussaturation in infarcts (Blair, 1969; Marshall et al.,1974; Obeid et al., 1972; Owen et al., 1970; Weissand Lipp, 1979). From the histograms shown infigure 3, it is clear that while the average regional

venous saturation is lower in the affected area, thedistribution of saturation is just shifted to slightlylower values. The degree of heterogeneity seen inthe infarct is large. There still are many veins inthe infarcted area with high O2 saturations. Thiscould be due to nonhomogeneous extraction of ox-ygen, or nonhomogeneous blood flow distribution.Some of the heterogeneity also could be due tonormal areas within the area of infarct, collateralflow distribution and venous anastomosis (Berneand Rubio, 1969; Cohen, 1978; Eliska and Eliskova,1970).

The fall in venous O2 saturations indicated anincrease in the amount of oxygen extracted by thearea supplied by the ligated vessel. In the presentstudy, I used arterial O2 saturations outside the areato compute the O2 extraction (i.e., normal arterialO2 saturation). O2 extraction is normally computedas the O2 content of the blood entering a region lessthe O2 content of the blood leaving it. To choose tocompute A-VO2 differences from arterial O2 satu-ration and content measurements from vesselswithin the ischemic area would be inappropriatesince some O2 is already lost to this region.

Increases in oxygen extraction provide a secondmechanism to maintain regional metabolic rate inthe affected region, in addition to blood flow in-creases. In normal hearts, oxygen extraction isgreater in the subendocardium compared to thesubepicardium (Weiss and Sinha, 1978; Weiss et al.,1978a). In the present study, this is also true at 10minutes after ligation in the control area. No othersignificant differences were found. In the infarct,extraction may be at its limit in many vessels. Thetotal average O2 extraction may not be maximum ifsome of the tissue is not in serious difficulty. Thereis sufficient evidence to claim a degree of inhomo-geneity within the infarct. The inhomogeneity offlow, O2 saturation, O2 extraction, and O2 consump-tion should not affect the accuracy of regional av-erages. As long as a sufficient sample size is ob-tained, statistical differences with a normal distri-bution can be studied.

There is evidence in many tissues that the he-matocrit of small vessels may be lower than that inthe large ones (Klitzman and Duling, 1979). This,however, may not be true in heart (Myers andHonig, 1964). Even if small cardiac vessels had alow hematocrit compared to large ones, this wouldnot affect computed arterial-venous O2 saturationdifference or extraction, since small and large ves-sels have the same saturation (Weiss and Sinha,1978). Further, no regional differences in hematocrithave been found in heart (Myers and Honig, 1964).

In the area affected by ligation, measurementsindicated a sharp reduction in work (Apstein et al.,1977; Pashkow et al., 1977). I and others report amuch reduced oxygen consumption in the area offlow reduction (Blair, 1969; Marshall et al., 1974;Obeid et al., 1972; Owen et al., 1970; Ramakrishnaet al., 1975; Weiss and Lipp, 1979). No regionaldifferences with depth were found except at 2 hours

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

REGIONAL O2 CONSUMPTION IN INFARCTS/Wms 407

when subendocardial consumption was higher thansubepicardial in the control region. The ENDO:EPIconsumption ratio at 2 hours was also significantlydifferent between the control areas and areas withreduced flow, indicating that the subendocardialregion was more adversely affected. In many typesof stress, this region is more severely affected (Hoff-man, 1978).

The ratio of oxygen supply to oxygen consump-tion is lower in the occluded than in the controlarea. No significant differences with depth into theleft ventricular free wall were observed. The ratioof supply to consumption in the infarcted area onlyranges from 1.3 to 1.4. Under other circumstances,the ratio may be lower (Daniell, 1973; Scott, 1961).Some of the needs of the tissue are being suppliedanaerobically, and portions of the tissue are dying.Due to inhomogeneity in the tissue, this ratio ishigher in some regions than others. It is clear fromthis study that the tissue affected by the ligation ofthe left anterior descending coronary artery sufferssupply vs. consumption imbalances in an inhomo-geneous manner. This within-region variation inblood supply and oxygen extraction in the core ofthe developing infarct may imply the possibility ofpreservation of some portions of the endangeredarea. This could be accomplished by increasingoxygen delivery through those vessels that are stillfunctional in the infarct.

ReferencesApstein CS, Deckelbaum L, Mueller M, Hagopian L, Hood WB

Jr. (1977) Graded global ischemia and reperfusion. Cardiacfunction and lactate metabolism. Circ 55: 864-872

Becker LC, Ferreira R, Thomas M (1973) Mapping of leftventricular blood flow with radioactive microspheres in exper-imental coronary artery occlusion. Cardiovasc Res 7: 391-400

Berne RM, Rubio R (1969) Acute coronary occlusion: Earlychanges that induce coronary dilatation and the developmentof collateral circulation. Am J Cardiol 24: 776-781

Blair E (1969) Significance of sampling sites in the study ofmyocardial metabolism in regional ischemia. J Thorac Car-diovasc Surg 58: 271-278

Buckberg GD, Luck JC, Payne DB, Hoffman JIE, Archie JP,Fixler DE (1971) Some sources of error in measuring regionalblood flow with radioactive microspheres. J Appl Physiol 31:498-504

Camilleri JP, Fabiani JN (1977) No-reflow phenomenon andacute myocardial ischemia. The need for further investigation.Biomedicine 26: 353-356

Cohen MV (1978) Quantitation of collateral and ischemic flowswith microspheres and diffusible indicator. Am J Physiol 234:H487-H495

Daniell HB (1973) Coronary flow alterations on myocardialcontractility, oxygen extraction, and oxygen consumption. AmJ Physiol 225: 1020-1025

Eliska O, Eliskova M (1970) Arterial and venous circulation inischemic foci in the myocardium. Angiology 7: 65-76

Gross GJ, Winbury MM (1973) Beta adrenergic blockade onintramyocardial distribution of coronary blood flow. J Phar-macol Exp Ther 187: 451-464

Hearse DJ, Opie LH, Katzeff IE, Lubbe WF, Van Der Werff TJ,Peisach M, Boulle G (1977) Characterization of the "borderzone" in acute regional ischemia in the dog. Am J Cardiol 40:716-726

Hirzel HO, Sonnenblick EH, Kirk ES (1977) Absence of a lateralborder zone of intermediate creatine phosphokinase depletion

surrounding a central infarct 24 hours after acute coronaryocclusion in the dog. Circ Res 41: 673-683

Hoffman JIE (1978) Determinants and prediction of transmuralmyocardial perfusion. Circulation 58: 381-391

Holtz J, Grunewald WA, Manz R, Restorff WV, Bassenge E(1977) Intracapillary hemoglobin oxygen saturation and oxy-gen consumption in different layers of the left ventricularmyocardium. Pfluegers Arch 370: 253-258

Khuri SF, Flaherty JT, O'Riordan JB, Pitt B, Brawley RK,Donahoo JS, Gott VL (1975) Changes in intramyocardial STsegment voltage and gas tension with regional myocardialischemia in the dog. Circ Res 37: 455-463

Klitzman B, Duling BR (1979) Microvascular hematocrit andred cell flow in resting and contracting striated muscle. Am JPhysiol 237: H481-H490

Kloner RA, Fishbein MC, Maclean D, Braunwald E, Maroko PR(1977) Effect of hyaluronidase during the early phase of acutemyocardial ischemia: An ultrastructural and morphometricanalysis. Am J Cardiol 40: 43-49

Lipp JA, Weiss HR (1978) Blood flow and relative tissue oxy-genation of normal and partially ischaemic myocardium: Ef-fect of CO2. Clin Exp Pharmacol Physiol 5: 567-577

Marcus ML, Kerber RE, Ehrhardt J, Abboud FM (1975) Threedimensional geometry of acutely ischemic myocardium. Circ52: 254-263

Marcus ML, Kerber RE, Ehrhardt JC, Falsetti HL, Davis DM,Abboud FM (1977) Spatial and temporal heterogeneity of leftventricular perfusion in awake dogs. Am Heart J 94: 748-754

Marshall RJ, Parratt JR, Ledingham IMcA (1974) Changes inblood flow and oxygen consumption in normal and ischemicregions of the myocardium following acute coronary arteryligation. Cardiovasc Res 8: 204-215

Myers WW, Honig CR (1964) Number and distribution of cap-illaries as determinants of myocardial oxygen tension. Am JPhysiol 207: 653-660

Obeid A, Smulyan H, Gilbert R, Eich RH (1972) Regionalmetabolic changes in the myocardium following coronary ar-tery ligation in dogs. Am Heart J 83: 189-196

Owen P, Thomas M, Young V, Opie L (1970) Comparisonbetween metabolic changes in local venous and coronary sinusblood after acute experimental coronary arterial occlusion.Am J Cardiol 25: 562-570

Pashkow F, Hollan R, Brooks H (1977) Early changes in con-tractility and coronary blood flow in normal areas of theischemic porcine heart. Am Heart J 93: 349-357

Ramakrishna NC, Sobel BE, Shell WE, Covell JW (1975) Al-tered myocardial oxygen consumption after coronary occlusionin anesthetized dogs. Am Heart J 89: 333-337

Scott JC (1961) Myocardial coefficient of oxygen utilization. CircRes 9: 906-910

Sinha AK, Neubauer JA, Lipp JA, Weiss HR (1975) Oxygensaturation determination in frozen blood. Microvasc Res 10:312-321

Sinha AK, Neubauer JA, Lipp JA, Weiss HR (1977) Bloodoxygen saturation determination in frozen tissue. MicrovascRes 14: 133-144

Steel RGD, Torrie JH (1960) Principles and procedures of sta-tistics. New York, McGraw-Hill

Steenbergen C, Deleeuw G, Barlow C, Chance B, Williamson JR(1977) Heterogeneity of the hypoxic state in perfused ratheart. Circ Res 41: 606-615

Weiss HR, Lipp JA (1979) Graded flow reductions and O2consumption of small partially ischemic region of dog leftventricle. Arch Intern Pharmacodyn Ther 237: 128-138

Weiss HR, Sinha AK (1978) Regional oxygen saturation of smallarteries and veins in the canine myocardium. Circ Res 42:119-126

Weiss HR, Neubauer JA, Lipp JA, Sinha AK (1978a) Quantita-tive determination of regional oxygen consumption in the dogheart. Circ Res 42: 394-401

Weiss HR, Bailin JJ, Levy PJ (1978b) Effect of mild hypovole-mia and fluid replacement on regional oxygenation, flow andsmall vessel blood content in the rabbit heart. Circ Shock 5:365-374

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from

H R Weissextraction, blood flow, and O2 consumption in the dog heart.

Effect of coronary artery occlusion on regional arterial and venous O2 saturation, O2

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1980 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.47.3.4001980;47:400-407Circ Res. 

http://circres.ahajournals.org/content/47/3/400.citationWorld Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on July 12, 2018http://circres.ahajournals.org/

Dow

nloaded from