CORONARY HAEMODYNAMICS AND MYOCARDIAL METABOLISM DURtNG CARDIOPULMONARY BYPASS J. P. RICHARDSON’ AND T. G.W. BAKER: University of Melbourne Department of Surgery, S t Vincent’s Hospital, Melbourne, Victoria, a n d Open Heart Surgery Unit, University of Melbourne Hospitals.

An experimental animal preparation is described which has allowed us to determine coronary haemodynamics and the overall myocardial metabolic rate of the disease-free, beating, normothermic heart during total cardiopulmonary bypass. We have confirmed that under these conditions the required coronary blood flow is about SOml/min/lOOg myocardium, which can be achieved at a low perfusion pressure (60 mm Hg or 8 kPa), and that overall myocardial metabolism is represented by an oxygen utilization of just over Zml/rnin/lOOg. We have also defined the associated dynamic parameters which characterize oxygen transport in the coronary circulation during bypass. By controlling coronary perfusion, analogous in a clinical setting to selective coronary perfusion, we have demonstrated a simple linear correlation between perfusion pressure and coronary flow, indicating a “fixed” coronary vascular resistance. Under these conditions the myocardial oxygen uptake rises towards a maximum of 4ml/rnin/lOOg with the expected associated changes in the parameters characterizing oxygen transport in the coronary circulation. These trends are postulated as being a reflection of alterations in regional myocardisl perfusion. By extrapolating to the clinical situation, these data have appeared to give valid information concerning the adequacy of coronary perfusion.

ALTHOUGH it is more than 20 years since the introduction of cardiopulmonary bypass techniques to facilitate cardiac surgery, firm and final agreement has not yet been achieved regarding the criteria for optimal conditions in the coronary circulation during bypass operations, nor for the intimately related problem of the best technique for preserving myocardial viability and function. It is true that the oxygen uptake of the nonworking left ventricle h?s been known fora long time (McKeever et alii, 19%). This datum and many others of similar ilk have allowed the development of a seemingly rational concept among perfusionists and cardiac surgeons for apparently satisfactory methods of protecting the myocardium during periods of whole body perfusion. Howevkr, it has 1 Senior Associate. 2 Cardiac Anaesthetist. Reprints. J. P Richardson, Albert St Medical Centre. 372 Albert Street, East Melbourne, Victoria. 3002, Australia

AUST. N.Z. J. SURG.VOL. 49 - No. 6, DECEMBER, 1979

become clear that systems which were believed theoretically to be adequate have in fact been associated with a significant and disturbingly frequent amount of myocardial damage during the perioperative period (Buckberg, 1977; Hultgren et alii, 1973, Neutze et alii, 1974; Balibreaetalii, 1975; lsom etalii, 1973; Barratt-Boyes et alii, 1976 At the present time, there is occurring around the world a great deal of experiment and clinical observation aimed at improving myocardial preservation during operation, chiefly by combining hypotherrnia with chemical arrest of myocardial electromechanical activity. Nevertheless, because of the conflicting nature of the data already available, it wasdeemed a worthwhile exercise to investigate a new coronary haemodynamics and myocardial metabolism in an experimental animal preparation during standard cardiopulmonary bypass in order to clarify current clinical concepts. 721


Since the heart isan aerobic organ, it reliesalmost exclusively on the oxidation of substrates for the generation of energy and, therefore in a steady state, determination of the heart's oxygen consumption provides a precise measure of this organ's total metabolism. The investigation reported here has sought to define coronary blood flow and myocardial oxygen uptake in a beating heart at 37" C which has been bypassed by a simple, standard, cardiopulmonary bypass circuit. The observations have been made both with autoperfusion of the coronary system from the aorta and also with selective coronary perfusion and aortic cross clamping.

MATERIAL AND METHODS The Experimental Preparation Adult mongrel dogs weighing between 15 and 35 kg were anaesthetized withan intravenous injection of sodium pentobarbital (30 mg/kg), and paralysed with gallamine triethiodide (2.5 mg/kg). Respiration was maintained with a Bird respirator through an endotracheal tube prior to the establishment of the extracorporeal circuit. Through a midline sternal splitting incision, the mediastinurn and, inevitably, both pleural cavities were opened. Cardiopulmonary bypass was established by placing large bore (32Ch-36Ch) catheters in both venae cavae through purse-stringed incisions in the right atrial wall. The venous blood was drained intoa Rygg-Kyvsgaard disposable bubble oxygenator, which was ventilated with a gas mixture of oxygen and either 2.5% or 5% carbon dioxide. Oxygenated blood was returned to the dog by a single roller pump, through a counter current heat exchanger intoa metalcannula (4 mm t o 5 5 mm internal diameter) fixed in one or other femoral artery. The experimental animal was perfused at a rate of 80-100 mI/kg/mi n . The extracorporeal circuit had been primed with a solution containing 1 litre of compound sodium lactate solution and % litre of 10°/~Rheomacrodex in normal saline. This degree of haemodilution gave haematocrit values in theanimals rangingfrom 15 to 35, with a median figure of 25. A third large bore (34Ch) multiholed drainage catheter was introduced through a purse-stringed stab incision in the right atrial wall, passed across the atrial cavity, and through the tricuspid valve into the right ventricle. Blood drained from this catheter was returned to the exygenator. The volume of this blood flow could be determined by a timed collection in a graduated burette which was incorporated in thisdrainage line. When thecaval snares had been secured around the cannulae and when the pulmonary artery had also been occluded by tightening a 722


snare around it, the drainage from the right side of the heart (right atrium and ventricle) was regarded as being the coronary sinus return. The left ventricle was "vented" by a small-bore catheter (28Ch) inserted into its cavity through a purse-stringed stab incision i n the apexof the ventricle. This drainage was also returned to the oxygenator venous reservoir via a burette by means of which the volume of blood appearing in this chamber during cardiopulmonary bypass could be measured. The purpose of these arrangements was, first, to prevent the development of intracavitary tension, and second, to collect the bronchial circulation return as well as coronary return reaching the left ventricular cavity by Thebesian channels. In practice it was found that flow through this vent did not amount to more than a few millilitres per minute, and therefore this volume was omitted from the calculation of total coronary flow. In some experiments, the ascending aorta was cross clamped and the coronary arteries were then perfused by cannulating the root of the aorta. This cannula was connected through a second roller pump with the arterial end of the oxygenator. This allowed the coronary arteries to be perfused at varying pressures and flows. Directly Observed Parameters Arterial blood pressure was recorded continuously on a direct writing record system utilizing a finecatheter, passedtoa central aortic position from a femoral artery, and connected to a Statham pressure transducer. An additional arterial pressure monitor was used in those experiments in which the coronary arteries were perfused from the isolated aortic root by a separate cannula. In these instances the aortic root pressure was monitored and was regarded as the coronary perfusion pressure. The electrocardiograph was recorded continuously from surface electrodes. Core temperature was monitored with an oesophageal thermistor probe. Serum electrolyte and acid-base status were maintained in a stable state after periodic assessment of serum chemistry, the timing of which was determined by the necessity to obtain blood samples to measure myocardial oxygen consumption. When any particular set of experimental conditions had been stable for at least 10 to 15 minutes, pairs of samples of blood were taken from the arterial line and from the coronary venous drainage line for blood gas and serum electrolyte estimations. At the same time coronary venous flow was measured. Generally two pairs of simultaneous arterial and coronary venous samples were obtained at each AUST. N.Z. J. SURG.VOL. 49 - No. 6, DECEMBER, 1979


determination, and coronary flow measurement was similarly made in duplicate. Blood gas analysis was made using a Radiometer Astrup triple electrode system, and a value for the oxygen tension, the carbon dioxide tension, and the pH of each sample wasdetermined. From thesedata haemoglobin oxygen saturation wascalculated with a Severinghaus (1966) blood gas calculator. Serum levels of sodium and potassium ions were determined by standard techniques, as were the haemoglobin concentration and the packed cell volume. At the conclusion of each experiment the heart was weighed. Derived Parameters The peripheral vascular resistance across thecoronary vascular bed was calculated as the ratio of the mean pressure drop to the total coronary blood flow and was expressed in conventional peripheral vascular resistance units as millimetres of mercury pressure per millilitre of flow per minute. In order to compare the results of one experiment with another the resistance unit was "standardized" by dividing the total myocardial mass into 100 g compartments each of which wasassumed to haveaflow which was proportional to its mass. Thus two resistancescould be calculated for any steady state of the coronary circultion, viz., total coronary resistance and resistance per 100 g of myocardial mass. In accordance with the Fick principle, the myocardial oxygen consumption was calculated as the product of the coronary blood flow and the coronary arteriovenous oxygen content difference. This value was also "standardized" by regarding the mass of myocardium to be divided into 1009 compartments and expressing the myocardial oxygen uptake as millilitres of oxygen per 100 g of myocardium per minute. Analysis of Data All sets of data have been subjected to standard statistical analyses (Fisher, 1970; Snedecor and Cochran, 1967).

All hearts were in sinus rhythm throughout. We noted that the range of values for heart rate varied slightly from one preparation to another and confirmed that there was a close linear correlation ( r = 0.73; P < 0.001) between heart rate and myocardial oxygen uptake. Estimations of myocardial oxygen uptake were therefore "standardized" to an adjucted heart rate of 125 in each preparation. Haem0dynarnics.- The aortic mean pressure varied widely from preparation to preparation, the range being 25 to 120 mm Hg (3.3 to 16.0 kPa). No special measures were taken to boost an unexpected low pressure other than increase the total. perfusion flow towards 100 ml/min/kg. In particular, no pharmacological interventions were made. Coronary arterial flow as measured from coronary venous return averaged 114 ml/min, indicating that the coronary blood flow was about 6% of the total perfused flow. The total peripheral resistance of the coronary circulation was about one-half of one millimetre of mercury pressure (67 Pa) for each millilitre of flow per minute. This value was remarkably constant over the range of observations. The haematocrit value is displayed in Table 1 to illustrate that there was a moderate toseveredegree of haemodilution in many of the smaller animals during the experiment. I t was considered possible that the smaller animals with greater haemodilution and thus less viscous perfusion fluid might exhibit a significant falling off in resistance to flow. Statistical correlation did not support this hypothesis ( r = 0.28; P 5 0.10). TABLE1 Auloperfused. Normal, Empty, Beating Heart at 37" C: Summary ol Parameters Measured

n = 22 Sample Mean Standard Error

A utoperfusion The first sets of analyses were made on data obtained from preparations on cardiopulmonary bypass at 37" with the heart beating in sinus rhythm and with the coronary tree being perfused in a normal fashion. These data are summarized in Tables 1 and 2. The heart weight in eleven preparations ranged from 350 g to 151 g, with a mean value of 253 g. DECEMBER,



Heart weight - g


Heart rate - beatsmin-'



61 (8.1)

6 (08)



0 51 (68)




Total myocardial oxygen uptake (heart rate 125) - mlmin-'



Arterial oxygen content - ml 100 ml-'



Aortic (i.e. perfusion) mean pressure - mm Hg (kPa) Total coronary blood flow

- ml m i t T

Total peripheral coronary resistance to flow - mm Hg (Pa)' Haematocrit - per cent


AUST. N.Z. J. SURG.VOL. 49 - No. 6,


Arterial to coronary sinus oxygen content difference - ml.lOO m1-I



Oxygen extraction ratio - per cent



Coronary sinus haemoglobin saturation - per cent


2 .o

43 6 (5.8)

2.1 (0.3)

Coronary sinus partial pressure oxygen - mm Hg (kPa)



CORONARY HAEMODYNAMICS AND MYOCARDIAL METABOLISM TABLE2 Autoperlused. Normal, Empty, Beating Heart at 37" C: Coronary Haemodynamics and Myocardial Metabolism "Standardized" Values per 100 g Myocardium n = 22 Sample Mean Standard Error Aortic mean pressure - mm Hg (kPa)

61 (8 1 )

6 (08)



Coronary resistance to flow - mm Hg (Pa)'

1.20 (160)


Myocardial oxygen uptake (heart rate 125) - rnl.rnirl



Coronary flow - ml.min-'

These data are "standardized" (in so far as this is necessary) in Table 2, which shows three parameters of interest (coronary perfusate flow, coronary vascular resistance to flow, and myocardial oxygen uptake at a heart rate of 125) expressed asvaluesfor each 100 g of myocardium perfused. The mean coronary flow was 47 ml/min/100 g of myocardium but in these autoperfused coronarycirculation preparations, where the perfusion (central aortic) pressure varied from a low of 25 mm Hg (3 kPa) to a high of 120 mm Hg (16 kPa), there was a moderately good linear correlation ( r = 0.564; P -z 0.01) between pressure and flow. This is further demonstrated by a very constant resistance to flow through the coronary bed, 1.20 3 s.e. 0.08 mm Hg. min. ml-' (1602 11'),atall pressuresand flows. A significant correlation between coron-ary flow and coronary vascular resistance under the described conditionscould not be demonstrated ( r = 0.07). These results in respect of coronary flow are almost identical with those obtained by Hottenrott and Buckberg (1974),viz.,42+12 ml/min/lOOg, with a similar experimental preparation, though these workers were at pains to hold the perfusion pressure at 100 mm Hg (13.3 kPa) and consequently they computed a higher vascular resistance. However, these workers' "total coronary flow" excluded the flow in the left anterior descending coronary artery from, their calculations. It is difficult, therefore, to make exact comparisons. Myocardial oxygen uptake.- The mean myocardial oxygen uptake of the animals' hearts, which weighed on average 253 g. was 6.1 ml/min during autoperfusion. The myocardial oxygen uptake o? 2.16 ml/min/100 g myocardium was believed to represent the basal or resting energy utilization necessary to sustain the viability of the myocardium plus the energy required to maintain active contrac-, tion of the myocardium at zero load. The mean oxygen content of the perfusate as it left the oxygenator was almost 13 m1/100 ml, and the 724

mean arterial to coronarysinusdifference in oxygen content was 4.8 m1/100 ml. There was a remarkably constant oxygen extraction ratio of 37% (i.e., the proportion of oxygen utilized by the myocardium to that supplied in the coronary arterial blood) exhibited by the myocardium in this preparation. In a similar and expected fashion, the partial pressure of oxygen in the coronary sinus effluent was very nearly constant at 43 mm Hg (5.7 kPa), and this corresponded to a relatively high mean haemoglobin saturation of the coronary sinus blood of 72%. All of these data referable to the overall metabolic pattern of the myocardium are summarized in Tables 1 and 2. The values for oxygen consumption of the bypassed, beating, unloaded left ventricle at 37OC, do not differ significantly from those found by Hottenrott and Buckberg (1974) - viz., 3.2 5 0.6 ml/min/l00 g - in a verysimilaranimal preparation, but using a radionuclide microsphere method for measuring total and regional coronary blood flow. Controlled Coronary Perfusion Haem0dynamics.- When the volume and pressure of coronary perfusion were controlled, the pressure was increased in stepwise fashion from 40 to 200 mm Hg (5.3 to26.7 kPa) by increasing the rate of coronary perfusion, and it was found that there was a very close linear correlation (r = 0.898) between pressure and flow throughout the range. The statistical data are set out in Table 3, and the regression line of coronary flow on perfusion pressure is depicted in Figure 1. The close correlation between pressure and flow is further demonstrated by the fact that the total coronary vascular resistance to flow was remarkably constant, resistance per 100 g of perfused myocardium being 1.70+s.e.0.068 mm Hg.rnin.m1-'(226+9 TABLE3 Normal, Empty, Beating Heart ar 3 7 ° C Haemodynarnrcs of Controlled Coronary Perfusion n = 30 Correlation of coronary flow ( Y ) against perfusion pressure ( X )

r = 0 898 p c 0.001

Regression coefficient ( b )of coronary flow on perfusion pressure

b = 0.68 t s d 0.06

Regression equation of coronary flow (Y ml.min-'.IOO g-') on perfusion pressure ( X mm Hg).l mm Hg = 133 Pa

Y = 0.68X - 7.04 + s d 1444


Mean coronary vascular resistance per 100 g 1.70 (226) perfused myocardium - mm Hg (Pa) ' +,s.e. 0.068(9.0) (s.d. = standard deviation : s.e. = standard error)



No. 6, DECEMBER, 1979










.-'c100E . -

E 8 l


.IOOnl-' / ///

/ . .


/'/ '-'/'








1 60

1 100

ml I


I 800

.roog-' 1 1

Ppf- mm Hg

.. A

FIGURE1:Thisdepicts the regression lineof coronarybloodflow (Qca) on coronary perfusion mean pressure (Ppf) in the perfused, beating, normothermic. empty heart. The dashed lines are drawn two standard deviations from the regression line and the experimental data points are depicted. The equation of the regression line and related statistics are set out in Table 3. (1 mm H g = 133 Pa)') over a range of pressures from 40 to 200 mm Hg (5.3 to 26.7 kPa) andflowsfrom 20 to 160 ml/min. Significant correlation between coronary flow and coronary vascular resistance again could not be demonstrated ( r = 0.01). However, under these conditions of controlled coronary perfusion the higher mean flow (+ Sooh) compared with the autoperfusion situation was achieved by a disproportionate increase in perfusion pressure (+ 108%), indicating a 31% rise in total coronary vascular resistance with very high flows, and this rise in resistance was significant at the 1% level. The statistics are set out in Table 4. Myocardial Oxygen Uptake.- When coronary perfusion is controlled, the principal determinant of coronary flow is the perfusion pressure. Somewhat similarly, the myocardial oxygen uptake during controlled perfusion is, to a significant degree, determined by the coronary flow. The data in relation to myocardial oxygen uptake and to the arteriovenous


, ,






50 100 150 - m I. mi,-I. ioog-'

FIGURE 2: This composite depicts the regression lines of myocardial oxygen uptake (OrnyO2) in the lower panel, and of arterial to coronary sinus oxygen content difference ( h a - c s 0 2 ) in the upper panel, on coronary blood flow (Qca) in the perfused, beating, normotherrnic, empty heart. The dashed lines are drawn two standard deviations from the regression lines and the experimental data pointsare depicted.Theequations of the regression lines and related statistics are set out in Table 5.

difference in oxygen content in the coronary circulation are summarized in Table 5 and depicted graphically in Figure 2. There is only moderately good linear correlation between coronary flow and oxygen uptake ( r = +0.5765), although it is highlysignificant. A greater number of data points, particularlyat the extreme ranges, could quite conceivably yield a better, possibly asymptotic, regression line. The analysis further reveals quite a close negative linear correlation between coronary flow and arteriovenous difference in oxygen content ( r = -0.7573), as would be expected, and here again more data concerning the correlation at very high and very low flows might significantly alter the shape fo the regression line in those regions.

TAELE4 Normal. Empty Beating Heart af 37°C Coronary Haemodynamics Autoperfusion v Controlled Perfusion Controlled Perfusion ( n = 30) Mean Standard Error

Autoperfusion ( n = 22)' Mean Standard Error Mean perfusion pressure - mrn Hg (kPa) Coronary blood flow - ml rnin 100 g-' Coronary vascular resistance/100 g - rnrn Hg (Pa) min rnl-'

61 (8 1 )

6 (08)

127 ( 1 6 9 )

25 (3 3 )

47 7


78 8


1 70 (226)

0 068 (9 0)

1 2 0 (160)

0 0 8 (11) tt=274 PCOOl

AUST. N.Z. J. SURG. VOL. 49 - No. 6, DECEMBER, 1979






The Perfused, Beating. Normothermic Empty Heart v Coronary Blood Flow per 100 g Myocardium ( X ) - ml, Linear Corre1atio.n and Regresston Statistics n = 30

Y‘ = +my02


3 ca-csm

Y3 = OER

Y‘ = scs

r P b Sb

+0.5765 co.001

-0 7573 c 0 001

-0.621 5 co.001

t0.4362 C0.05

t0.0166 4 38 x 10-3

-0.0439 7.2 x 10-3

-0.2759 6.70 x 1 T 2

to 1931 7.53 x 10-2

Y= sy.x

1.42 + 0.0166X 0.77

7.77 - 0.0439X 111

52.65 - 0.2759X 9 71


Y ’ - 3my02 ~2 - Oca-cs02 Y3 - OER Y4 - s c s YS - PcsO2



55.94 + 0.1931X 11.67

YS = Pcso2


< 0.001

+O 2722 3.82 x 1O P 26.33 + 0.2722): 5 93

-Myocardial oxygen consumption per 100 g myocardium - rnl -Arterial to coronary sinus oxygen content difference per 100 rnl - rnl -Oxygen extraction ratio - per cent per cent -Coronary sinus haemoglobin saturation -Coronary sinus partial pressure of oxygen - mm Hg (1 rnrn Hg = 133 Pa)


- significance: b - regression coefficient: Sb - standard deviatton of regression coefficient, Y= - regression equation: Sy.x - standard deviation from regression line

* * r - correlation coefficient: P

When the oxygen extraction rate (OER), the volume of oxygen extracted by the myocardium expressed as a percentage of the total coronary arterial oxygen content, was compared with the coronary flow, reasonably good negative correlation (r = -0.6798) was demonstrated, and this was reflected









Qa-m I . m i n-I. ioog-’ FIGURE 3: This composite depicts the regression lines of the oxygen extraction ratio (OER) in the lowest panel, of the haemoglobin saturation i n the coronary sinus blood (Scs) in the middle panel, and of the partial pressure of oxygen in the coronary sinus b!ood ( p c s 0 2 ) in the uppermost panel, on coronary blood flow (Oca) in the perfused. beating, normothermic. empty heart. The dashed lines aredrawn twostandard deviationsfrom the regression line and most experimental data points are depicted. One or two data points in each panel could not be shown because the ordinate axis has been truncated as a graphic economy. The equations of the regression lines and related statistics are set out in Table 5.


by significant positive correlation of coronary sinus haemoglobin saturation (r=0.4362) and partial pressure of oxygen (r = 0.8026) with coronary flow. The statistical data are set out in Table 5, and the regression lines are depicted in Figure 3. DISCUSSION

Haemodynamics It is well known and universally accepted that in a working heart in an intact animal there is indeed a very marked degree of “autoregulation” achieved largely through a variable coronary vascular resistance to flow which, in turn, is closely related to and apparently dependent o n local metabolic myocardial demands. The appropriate alterations to coronary vascular resistance are mediated through neurohumoral (catecholamines), hypoxic (adenosine),. acid/base disturbances, and other related mechanisms (Gorlin, 1971). Exactly analogous provi-sions cardiopulmonary bypass were postulated b y Shaw et alii (1962) in experiments involving only the circumflex coronary artery in dogs, but these mechanisms were believed to operate only in the ”normal” range of perfusion mean pressures, from 70-140 mm Hg (9.3 to 18.7 kPa). In ourobservations. which involved total coronary blood flow, we were not able to confirm this often quoted dictum, and coronary flow seemed very largely determined by perfusion pressure through the whole range of low, “normal”, and high perfusion pressures. It may be relevant t o the potentiality for variations of coronary vascular resistance while the heart is empty, beating, and at 37”, that we did observe a 30% higher resistance when the coronary flow was very considerably augmented (+ 60%) during a period of controlled coronary perfusion. But. if it is possible to extrapolate from our observations to the AUST. N.Z. J. SURG. VOL. 49 - No. 6,DECEMBER, 1979


clinical situation, it would be dangerous to assume that if too high a perfusion pressure obtains, the coronary resistance will also be high and thus tend to protect the myocardium from the dangers of high pressure, high flow perfusion (Balibrea et alii 1975; Brown, 1969). On the other hand, in our autoperfusing preparation a mean perfusion pressure of only 60 mm H g (8 kPa) seemed adequate, on a prior; grounds at least, since it allowed a myocardial blood flow (47 mI/lOOg/min) which isabout one-half of the mean flow in the basal state of the working human myocardium (Gorlin, 1971), and studies of oxygen availability and uptake (vide infra) tend to confirm the adequacy of these pressures and flows in the bypassed, non-working heart.


alii (1972), Hottenrott and Buckberg (1974)and Hottenrott et alii (1974). The heart that was particularly at risk was that which was fibrillated during bypass or that in which there was marked left ventricular hypertrophy. In these situation the subendocardial myocardium was most particularlyat risk, and Estes et a h (1966) had demonstrated the anatomical differences of the intramyocardial vascular supply which could invoked to explain the vulnerability of the subendocardium.

Myocardial metabolism.- Since the heart is an aerobic organ, that is to say, since it relies almost exclusively on the oxidation of substrates for energy, it can develop only a small oxygen debt. Therefore, in a steady state, determination of the heart’s oxygen consumption provides a precise measurement of this organ’s total metabolism, and an understanding of the categories involved in myocardial oxygen uptake of the empty, beating, perfused heart at 37°C will provide important insights to the cardiac surgeon and to clinical cardiopulmonary bypass perfusionists and technicians.

If it Is accepted that our evidence suggests that a myocardial flowof about05 ml/g/min issufficientto maintain myocardial viability and contraction against zero load at 37”C, and this belief is supported by the findings of Hottenrott and Buckberg (1974), then one might suspect that a coronary flow around 300 ml/min in a non-hypertrophied, beating, empty human heart used by lsom etalii(1973) might be very greatly in excess of that which is necessary at 30°-34”C, though the practice of the New York University Hospital group was in line with the often quoted statistic of 1 mI/g/min for perfusing a heart under “basal” (but also working) conditions at 37” C. This would then throw open to doubt many of the implications of this otherwise very thoughtful and careful clinical study. On the other hand, “high” total coronary blood flow of only about 150 ml/min at a “high” mean perfusion pressure of 80 mm Hg (10.7 kPa) in, presumably, grossly hypertrophied hearts of patients having aortic valve replacement, as reported by Barratt-Boyes et alii (1976) might be expected to represent underperfusion, even at 30’ C, and it was therefore not surprising to learn that metabolic studies, reported in this paper, suggested inadequate myocardial protection.

It has been known for many years that the total metabolism of the arrested, quiescent heart is only a small fraction of that of the working organ. Whereas the oxygen consumption of the beating heart, contracting against a load, ranges from 8 to 15 ml/min/lOO g of ventricle, the oxygen consumption of the arrested heart or quiescent papillary muscle amounts to only 1.O to 1.5 ml/min100 g of ventricle, and this quantity of oxygen is required for the normal metabolic processes in cardiac cells not associated with contraction (Sonnenblick and Skelton, 1971). This “basal” requirement is slightly less than the “basal” value which we observed (2.16 ml/min/100 g myocardium), and the small difference could quite well be attributed to the energy required to maintain active contraction of the myocardium, 125 timedmin, against zero load at 37°C.

In respect of regional coronary blood flow, our observations can provide no useful information, which is not to say that regional difference in myocardial flow may not be of considerable importance in the non-working, beating and perfused heart. Sullivan etalii (1967) and Liedtkeetalii(1970) have demonstrated that with coronary atherosclerosis myocardial flow is non-uniform, and that the degree of non-uniformity is dependent not only on the degree of local arterial obstruction but also on the compensating collateral vessels. The importance of regional differences in myocardial blood flow during cardiopulmonary bypass was pinpointed in several pertinent article by Buckberg et

To summarize at this point, during cardiopulmonary bypass at normal temperature, the empty, beating and non-diseased heart can adequately be perfused: (i) with a mean perfusion pressure of about 60 mm Hg (8 kPa) (below normaI”basa1” levels); (ii) with a total coronary blood flow of about 50 ml/min/100 g (about 5O0/o of “basal” working levels); (iii) resulting in a myocardial oxygen consumption of about 2 ml/min/100 g (about 10-15% of “basal” working energy requirements); and (iv) resulting in an arteriovenous difference in oxygen content in the coronary circulation of slightly less than 5 ml Oz/lOO ml flow, (less than half of ”basal” working requirements); and that when theseconditions prevail there

AUST. N.Z. J. SURG.VOL. 49 - No. 6, DECEMBER, 1979



is (v) a constant oxygen extraction ratio of 37% (about one-half of the “basal” working heart ratio); (vi) a constant coronary sinus haemoglobin saturation of i’2°/~,(three times the “basal” working level); and (vii) a constant coronary sinus partial pressure of oxygen of 43 mm Hg (5.7 kPa), (about twice the “basal” working value). These seven parameters would appear to give a reasonably accurate indication of adequate, safe perfusion of the bypassed beating heart, and, taken together, would help to show whether the myocardium was being underperfused (hypoxic) or overperfused (traumatized). When, in our experiments, higher perfusion pressures were employed, very high coronary blood flows resulted - and this alone causes myocardial trauma (Brown, 1969), and there is an increase in myocardial oxygen uptake, long since noted (Gregg, 1963; Kahler et alii, 1963; Sarnoff et a h , 1963), but still difficult to explain satisfactorily apart from invoking better subendocardial perfusion. The regression line of this increase in oxygen utilization plotted against coronary flow is likely to be asymptotic and remain below 4 ml Oz/min/100 g myocardium.


latter two parameters were noted to be normal (i.e. high) before bypass. In contrast, in the meticulous study reported by Barratt-Boyes et alii (1976) reference is made to “high pressure and high flow” coronary perfusion of a group of patients undergoing aortic valve replacement, and therefore with, almost certainly, a large ventricular muscle mass. Reference has already been made earlier in this paper to the probably inadequate coronary flow in these patients. This is borne out by the fact that the myocardial oxygen uptake was of the order of 9 ml/rnin, even at “29.5” C”, suggesting a ventricular mass around 400450 g, requiring a total coronary flow of about 200225 ml/min (compared to 150 ml/min which they received), and this inadequate flow predictably resulted in a high arteriovenous oxygen content difference in the coronary circuit of between 6.5 and 7.5 ml Oz/lOO ml flow. The relative inadequacy of these myocardial perfusions was confirmed by the authors on the basis of the extended myocardial metabolic studies which they performed on these patients. CONCLUSION

Along with this basic datum about oxygen uptake at high coronary flows there is found the related, and therefore expected, tailing off of the arteriovenous oxygen difference which, from the shape of its regression line would appear to reach an absolute minimum just below 2 ml Oz/min/l00 ml blood flow. In like fashion, at very high coronary flows, the oxygen extraction ratio appears likely to have an asymptote of about lo%, the haemoglobin saturation of the coronary sinus effluent rises to a “high” of about 85%, and the partial pressure of oxygen in the coronary sinus blood reaches a plateau at about 70 mm Hg (9.3 kPa). In the light of these “experimental” facts it is instructive to look critically at two of theverycareful clinical studies reported in the surgical literature. lsom et alii, (1973) reported on a series of patients. submitted tocardiopulmonary bypass with a modest degree of hypothermia (30’34” C ) , in whom the average myocardial oxygen uptake was 6.7 ml/rnin, indicating a mean myocardial mass of about 300 g, and we have already suggested that coronary perfusion flows of around 300 ml/min in heartsof thissize was likely to be excessive. This doubtless was the reason that no difference could be detected in Oxygen uptake between hearts perfused above and below 300 ml/min, that the oxygen extraction ratio was a low 27%. and that the arteriovenous oxygen content difference in the coronary circuit was a low 3.2 ml Oz/lOO rnl flow, although, as expected, these


The purpose of this investigation was, as initially indicated, to characterize the haemodynamics and the overall pattern of cardiac metabolism in the bypassed, beating, normothermic heart, in order to clarify current clinical concepts concerning optimal methods of preserving myocardial viability and function during cardiopulmonary bypass operations. We believe these aspirations have been achieved. I t has been possible to define with reasonable precision desirable coronary blood flows, the perfusion pressure necessary to achieve these, the myocardial oxygen requirements under these conditions (and hence the total cardiac metabolism), and various parameters reflecting the oxygen extraction rate which can be determined from coronary sinus blood samples and which indicate whether or not the myocardium is being under perfused or overperfused. Finally, the data have been used to extrapolate to the clinical situation as described in two papers in the recent surgical literature relating to myocardial perfusion during bypass operations. In each case, it would have been possible from our data to predict that, in one case, the myocardium was being overperfused and, in the other case, that the reverse obtained. The authors of these articles had come to similar conclusions on the basis of very elaborate myocardial metabolism studies which now appear to be unnecessary in deciding whether a particular perfusion is adequate or not. AUST.N.Z. J. SURG.VOL. 49 - No. 6, DECEMBER, 1979


ACKNOWLEDGEMENTS This work was made possible only by unstinted access to the facilities of the Animal ResearchTheatres of the Experimental Medical and Surgical Department of St Vincent’s Hospital, Melbourne, and *the Clinical Biochemistry and Haematology Departments of the same institution undertook all of the biochemical and haematology measurements. Two of the Perfusion Technicians from the Open Heart Surgery Unit of the University of Melbourne Hospitals, Messrs K. Mason and P. Ewer, supervised the cardiopulmonary bypass circuits. Ms C. A. Richardson typed the manuscript. REFERENCES BALIBREA, J. L. BULLON,A,. DE LA FUENTE,A,. ALCARON, A. D. FARINAS. J., CoLLANTES, P.. GIL, M., GOMBAU, M.. MORALES, R., and SANCHEZ. F. (1975), Thorax, 30: 371. BARRATT-BOYES. B. G., HARRIS,E. A., KENYON.A. M., LINDOP, C. A. and SEELYE, E. R. (1976). J. thorac. cardiovasc. Surg., 72: 133. BROWN,A. H. (1969), J. thorac. cardiovasc. Surg.. 58: 655. BUCKBERG. G. D. TOWERS, 6.. PAGLIA, D. E.. MULDER,D. G. and MALONEY,J . V. (1972), J. fhorac. cardiovasc. Surg. 64: 669. BUCKBERG, G. D. (1977), Ann. thorac. Surg., 24: 379. ESTES, E. H., ENTMAN. M. L., DIXON, H. 8. and HAECKEL. D. 8. (1966). Amer. Heart J.. 71: 58.


FISHER,R. A. (1970) Statistical Methods for Research Workers, 14th edition, Oliver & Boyd: Edinburgh. GORLIN,R. (1971) Brit. Heart J., Suppl. 33: 9. GREGG.D. E. (1963), Circulat. Res.. 13: 497. HOTTENROTT, C. and BUCKBERG, G. D. (1974), J. thorac. cardiovasc. Surg.. 68: 626. HOTTENROTT. C., MALONEY, J. V. and BUCKBERG. G. D. (1974) J. thorac. cardiovasc. Surg., 68: 634. HULTGREN, H. N., MIYAGAWA. M..BUCH.W. and ANGELL,W. W. (1973). Amer. Heart J., 85: 167. . ISOM. 0.W., KUTIN,N. D.. FALK,E. A. and SPENCER, F. C. (1973) J. thorac. cardiovasc. Surg., 66: 705. KAHLER,R. L.. BRAUNWALD. E.. KELMINSON.L. L. KEDES,L.. CHIDSEY, C. A. and SEGAL, S. (1963).Circu/at. Res.. 13: 501. LIEDTKE. A. J., LA RAIA,P. J., BRKENHAGEN, D., KEMP. H. G. and GORLIN.A. (1970), Clin. Res., 18: 318. MCKEEVER. W. P., GREGG, D. E. and CANNEY, P. C. (1958), Circulat. Res., 6: 612. NEUTZE.J. M., DRAKELEV. M. J., BARRATT-BYES, B. G. and HUBBERT. K. (1974). Amer. Heart J., 88: 425. SARNOFF, S. J., GILMORE.J. P., SKINNER,N. S ,WALLACE, A. G. and MITCHELL, J. H. (1963), Circulat. Res. 13: 514. SEVERINGHAUS. J . W. (1966), J. appl. Physio.. 21: 1108. SHAW. R. F., MOSHER. P., ROSS, J., JOSEPH, J. I. and LEE,A. S. J. (1962), J. thorac. cardiovasc. Surg.. 44: 608. SNEDECOR, G. W., COCHRAN, W. G. (1967), Statistical Methods, 6th edition, Iowa State University Press: Arnes. Iowa. SONNENBLICK, E. H. and SKELTON,C. L. (1971). Mod. Conc. cardiov. Dis., 40: 9. SULLIVAN,J. M.. TAYLOR, W J.. ELLIOTT,W. C. and GORLIN,R. (1967), J. clin. Invest., 46: 1402.

ISOLATED LIVER PERFUSION: THE CHOICE OF ANAESTHETIC B. R. DOBBS, J. N. BAXTER, B. C. GALLAND AND D. LEE Department of Surgery, University of Otago Medical School, Dunedin, New Zealand.

Perfused livers isolated from rats under halothane anaesthesia produced greater amounts of bile, released smaller amounts of aspartate aminotransferase, and had a much greater ability to maintain a constant concentration of glucose in perfusates than those obtained with ether or pentobarbitone. Little or no effect was shown on the ability of the liver to synthesize urea and to retain potassium within the organ. It appears, therefore, that halothane is the anaesthetic of choice when removing the liver from the laboratory rat.

PENTOBARBITONE and ether are the two anaesthetic agents in common use in the course of experimental procedures on small animals. When pentobarbitone is used for prolonged procedures, periodic Pulses of the drug are necessary with the Reprints. Dr D . Lee. Department of Surgery, University of Otago, School, P 0 Box 913, Dunedin. New Zealand



- No. 6, DECEMBER, 1979

result that the anaesthesia is not smooth and overdosing is not uncommon. In addition, commercial preparations of pentobarbitone contain ethanol, which can have undesirable metabolic effects. Prolonged use of ether results in room contamination which is hazardous (due to itsexplosive Properties) and leads to poor operator performance. In 729

Coronary haemodynamics and myocardial metabolism during cardiopulmonary bypass.

749KB Sizes 0 Downloads 0 Views