INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE ISOLATED RAT HEART perfused by the Langendorff technique yields abnormally high coronary flow values and an increased rate of edema formation (17). Studies on blood-perfused heart-lung preparations are better suited for the examination of normal heart behavior (16). Several years ago we established a modified heart-lung preparation for measuring oxygen consumption and substrate uptake of the rat heart (9). In subsequent studies a rubber balloon, serving as a windkessel, was incorporated into the preparation (11), and the systemic circulation was substituted by an artificial shunt circuit (10).

In the present study further modifications of the preparation, which enable the coronary flow in the blood-perfused rat heart to be measured at variable left ventricular pre- and afterload, will be described. Furthermore, the improved preparation allows the systolic coronary flow to be distinguished from diastolic flow and permits the effect of coronary compliance on coronary circulation to be studied.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Heart-lung preparation. Nine male Wistar rats (body wt 366 ± 14 g; heart wt 0.923 ± 0.074 g) were anesthetized with urethan (1.2 g/kg body wt ip) and ventilated after tracheotomy with a Schuler ventilator (Hugo Sachs, Freiburg, Germany). A broad opening was made in the thorax with a parasternal cut, and the pericardium was removed. A schematic of the experimental setup is shown in Fig. 1.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Schematic of experimental setup. In heart-lung preparation systemic circulation was replaced by shunt circuit including resistor. Pulmonary flow (Fpulm) was measured with perivascular flow probe and aortic flow (Faorta) with cannulating flow probe, which was inserted into aortic root. Distal end of the Faortic probe was fitted to a 4-way stopcock. One outlet of stopcock was connected to the shunt circuit and the other to a rubber balloon serving as a windkessel. For measuring coronary perfusion pressure (Paorta), left ventricle was punctured with a 20-gauge Abbocath and pushed forward into aortic root. Left ventricle was punctured with another Abbocath for measurement of left ventricular pressure (PLV). Circulating blood volume was varied either by infusion of heparinized blood or by bloodletting.

Experimental protocol. After surgical preparation of the animals, the resistor was closed and the heart was allowed to stabilize for a period of 10 min at high left ventricular afterload. Thereafter, the resistance in the shunt circuit was varied randomly by opening and closing the resistor. The circulating blood volume was varied by either bloodletting or by infusion of heparinized blood (0.1-0.2 ml/min). Pressure and flow values were measured under steady-state conditions, i.e., at the earliest 5 min after the respective change in shunt resistance.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Representative original tracings of pressure and flow data recorded with totally closed (A), partially opened (B), and widely opened resistor (C). From top to bottom PLV amplitude, left ventricular end-diastolic pressure (EDP), Fpulm, Faorta, central aortic pressure (Paorta), and rate of rise in left ventricular pressure (dP/dt). Dotted horizontal line, zero flow in aortic root; dotted vertical lines, onset and end of systole and diastole, respectively. Area 1, right ventricular stroke volume, which is identical to left ventricular stroke volume under steady-state conditions. Area 2, left ventricular stroke volume minus that volume flowing into coronary arteries during systole. Difference between volume 1 and 2 corresponds to systolic coronary flow. Area 3 corresponds to diastolic coronary flow. Note that diastolic aortic flow is equal to systolic aortic flow if Starling resistor is closed (A). Retrograde diastolic aortic flow diminishes with decreasing shunt resistance (B and C).

Data analysis. Systole was defined as the fraction of time spent between the onset of left ventricular pressure rise, measured from the highly amplified pressure signal, and closure of the aortic valve, measured at zero flow of the descending branch of the aortic flow signal (Fig. 2A). Diastole was defined as the period between closure of the aortic valve and the next onset of left ventricular pressure rise. Systolic and diastolic coronary flow was evaluated from the pulmonary and aortic flow signal. Integrated pulmonary flow (area 1 in Fig. 2) represented right ventricular stroke volume, whereas integrated systolic aortic flow (area 2 in Fig. 2) represented left ventricular stroke volume minus that volume flowing into the coronary arteries during systole. Hence, under steady-state conditions, the difference between integrated pulmonary flow and integrated systolic aortic flow corresponded to that portion of left ventricular stroke volume flowing into the coronary arteries during systole. Integrated retrograde aortic flow during diastole (area 3 in Fig. 2) represented the volume stored in the windkessel during systole and flowing into the coronary arteries during diastole. The sum of systolic and diastolic coronary flow yielded total coronary flow, and aortocaval shunt flow was calculated from the difference between systolic and diastolic aortic flow. Coronary perfusion pressure was identical to central aortic pressure, which was evaluated from the tracings every 10 ms and integrated over systole, diastole, and the whole heart cycle. Under each condition, pressure and flow signals were evaluated from two individual beats so that each pressure and flow value represents the mean of two individual measurements.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Representative tracings of pressure and flow data before and during determination of zero Faorta. PLV amplitude, left ventricular EDP, Fpulm, Faorta, central Paorta, and left ventricular dP/dt, recorded before (A) and after closing 4-way stopcock (B). Faorta is zero when stopcock is closed. In this case, shunt circuit and windkessel are also functionally separated from preparation so that decay in diastolic Paorta depends on elastic properties of coronary arteries and on coronary resistance.

Statistics. All values are expressed as means ± SD. Differences between hemodynamic variables measured at various shunt resistances were assessed by Kruskal-Wallis one-way analysis of variance on ranks. All calculations, as well as regression analysis, were performed on a personal computer using a commercial statistical program (Statgraphics).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Initially, we evaluated the elastance of the arterial tree in four pilot experiments on anesthetized rats with intact circulation. Aortic flow and central aortic pressure were measured under open-chest conditions. Sixteen beats showed an exponential drop in diastolic aortic pressure, and elastance estimated according to the method of Frank (5) was found to be 693 ± 149 mmHg/ml.

In further pilot experiments, the elastance of the heart-lung preparations was adjusted to the elastance of the arterial tree of rats with intact circulation. For this purpose, three different-sized rubber balloons were connected to the preparations. The smallest balloon had an unstretched volume of 0.55 ml, the middle sized 1.0 ml, and the largest 1.5 ml. In heart-lung preparations connected to the small balloon elastance was 1,055 ± 724 mmHg/ml, in preparations connected to the middle-sized balloon 726 ± 280 mmHg/ml, and in preparations connected to the large balloon 526 ± 147 mmHg/ml. This shows that only in preparations connected to the middle-sized balloon, the elastance was identical to that measured in the pilot experiments in rats with intact circulation. In the following only results obtained from preparations connected to the middle-sized balloon are presented.

Mean values and standard deviation of nine basic hemodynamic variables measured at different values of shunt resistance are summarized in Table 1. With falling shunt resistance, shunt flow and pulmonary flow increased, whereas all the other variables, including coronary flow, were independent of shunt resistance. Cardiac output, which corresponds to pulmonary flow, was low because in the present preparation systemic circulation was reduced to one artificial shunt circuit. This resulted in a fourfold increase in total peripheral resistance compared with the situation in situ, where numerous pathways are joined in parallel. Coronary flow per gram ventricular weight ranged from 2.36 to 18.55 ml · min¹ · g¹ (mean value 7.84 ± 4.11 ml · min¹ · g¹).

The present preparation enabled total coronary flow to be subdivided into systolic and diastolic coronary flow. Total coronary flow did not correlate to aortic pressure because diastolic coronary flow increased with rising aortic pressure, whereas systolic coronary flow simultaneously tended to decrease. Systolic and diastolic coronary flow affected each other, so that diastolic coronary flow decreased with rising systolic coronary flow (Fig. 4A). Diastolic coronary flow ranged from 1.65 to 7.44 ml · min¹ · g¹ (mean value 4.28 ± 1.54 ml · min¹ · g¹). Diastolic coronary flow diminished with rising diastolic coronary resistance (Fig. 4B).

View larger version (12K): [in this window] [in a new window]   Fig. 4.   A: interrelationship between diastolic and systolic coronary flow. At low systolic coronary flow diastolic coronary flow was high and vice versa. B: dependence of diastolic coronary flow on diastolic coronary resistance. Diastolic coronary flow decreased with rising diastolic coronary resistance. Data values for individual preparations were labeled with different symbols.

Systolic coronary flow ranged from 2.78 to 15.26 ml · min¹ · g¹ (mean value 3.56 ± 4.86 ml · min¹ · g¹). Systolic coronary flow increased with increasing compliance (Fig. 5A), whereas diastolic flow declined simultaneously. Consequently, the proportion of systolic to diastolic coronary flow increased with rising coronary compliance (Fig. 5B). Coronary compliance per gram ventricular weight was calculated between 1.7 × 10⁴ and 24.4 × 10⁴ ml · mmHg¹ · g¹ (mean value 6.6 × 10⁴ ± 6.0 × 10⁴ ml · mmHg¹ · g¹). It is worth mentioning that no significant effect of heart rate on the proportion of systolic to diastolic coronary flow could be detected.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Dependence of coronary flow data on coronary compliance. Systolic coronary flow (A), proportion of systolic to diastolic coronary flow (B), and coronary capacitance flow (C) all increased with rising coronary compliance. Symbols as in Fig. 4.

Systolic coronary inflow can be subdivided into one portion immediately flowing through the resistance vessels and another portion that is stored for the present and flows during the next diastole. The stored volume was calculated from coronary compliance multiplied by the difference between the end-systolic and the end-diastolic aortic pressure. The time constant by which intramyocardial blood volume changed was evaluated graphically from the semilogarithmic pressure-time plots and had a value of 0.28 ± 0.10 s. Capacitance flow resulted from the product of stored volume and heart rate and was on average 2.52 ± 2.98 ml · min¹ · g¹. Coronary capacitance flow depended linearly on coronary compliance (Fig. 5C).

The evaluation of coronary capacitance flow permitted systolic and diastolic intramyocardial flow values to be estimated. Systolic intramyocardial flow, calculated from the difference between systolic inflow and capacitance flow, was 1.04 ± 2.72 ml · min¹ · g¹, and diastolic intramyocardial flow, calculated from diastolic inflow plus capacitance flow, was 6.80 ± 2.68 ml · min¹ · g¹, i.e., intramyocardial systolic flow was only 13% of coronary flow and intramyocardial diastolic flow as much as 87%.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present heart-lung preparation, systolic and diastolic coronary flow was evaluated from the two flow signals: 1) pulmonary flow measured with a perivascular electromagnetic flow probe and 2) aortic flow measured with a cannulating electromagnetic flow probe. Electromagnetic flowmetry implies two sources of error: 1) calibration of the flow signal and 2) detection of zero flow. The cannulating flow probe was calibrated with blood in a model circuit. Multiple calibrations at flow rates between 10 and 80 ml/min yielded almost identical calibration factors. For calibration of the pulmonary flow, the perivascular flow probe was placed around the aortic arch of a rat mounted into a model circuit. However, the calibration of the perivascular flow probe in this way may not necessarily result in correct values, because the probe is not calibrated with the same vessel used in the experiment. Therefore, we checked the calibration of the pulmonary flow probe in additional experiments on rats, in which the pulmonary flow area was correlated to cardiac output, measured with the Fick method (12). Calibration factors of the pulmonary flow probe, evaluated either with the Fick method or with the model circuit, were identical.

Zero flow of the pulmonary flow probe was taken to be the diastolic pulmonary flow signal. Zero flow of the aortic flow probe was determined by closing the four-way stopcock, situated just behind the aortic flow probe. Aortic zero flow was additionally checked in 10 beats, measured with a totally closed resistor. Under this condition, the windkessel was still in operation so that systolic and diastolic aortic flow should be exactly identical in value but opposite in direction. In practice, with a closed resistor systolic aortic flow per beat was 30.6 ± 11.4 µl and diastolic flow 30.4 ± 11.0 µl.

Because systolic coronary flow was obtained from the difference between two flow signals, deficiencies possibly involved in electromagnetic flowmetry should particularly affect the evaluation of this variable. In Fig. 6A systolic flow values, evaluated from two individual beats under respective steady-state conditions, are plotted against the respective mean values. The small distance between the individual values and the line of unity indicates small dispersion. The same proves true for pulmonary flow, aortic flow, diastolic and total coronary flow, and even for coronary compliance (Fig. 6B). For all these variables, the average difference between individual values and the respective mean values was less than 6%. This small dispersion speaks in favor of a reproducible evaluation of the respective variable.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   A: systolic coronary flow per beat, evaluated from 2 respective pulmonary and aortic flow signals (ordinate) is plotted against corresponding mean value (abscissa). B: plot of coronary compliance values, calculated from 2 respective beats, against corresponding mean values. Small dispersion indicates accurate evaluation of both variables. Solid lines, line of unity; dashed lines, lines of displacement.

Coronary flow per gram heart weight, measured at various left ventricular pre- and afterload, was on average 7.84 ± 4.11 ml · min¹ · g¹. In most studies on the rat heart, coronary flow is determined with microspheres. The following values have been measured: 5.10 ± 0.84 (22) and 6.17 ± 0.43 ml · min¹ · g¹ (21) in the awake rat and 4.10 ± 2.24 (24) and 5.5 ± 0.6 ml · min¹ · g¹ (18) in rats anesthetized with pentobarbital.

In pilot experiments on two preparations the infusion of 40 µg adenosine per minute resulted in an increase in coronary flow of 20%. Our relatively high coronary flow values and the small adenosine-induced increase in coronary flow indicate that in the present experiments the coronary bed was widely dilated, which might be attributed to the abnormal high resistance in the shunt circuit.

This heart-lung preparation, however, does not allow for a distinction between left and right ventricular perfusion. In the rat heart, right ventricular flow per gram muscle is about 65% of the left ventricular flow (22), although the right ventricle generates pressures that are only about 20% of those of the left ventricle. However, based on the ratio of right to left ventricular weight it can be estimated that only 15% of total coronary flow contributes to the blood supply of the right ventricle and as much as 85% to the supply of the left ventricle. Coronary flow values, measured in the present preparation, therefore must be referred predominantly to left ventricular flow.

The proportion of systolic to diastolic coronary flow depends on the elastic properties of the arterial tree. The elastance of the reduced arterial tree of the heart-lung preparations was, with a value of 726 ± 280 mmHg/ml, very similar to that measured in rats with intact circulation, which was 693 ± 149 mmHg/ml. In a study by Westerhof et al. (23), the capacitance of the arterial system of the rat is given as 2 × 10⁶ g¹ · cm⁴ · s². Assuming this value is a misprint and should correctly be 2 × 10⁶ g¹ · cm⁴ · s², conversion from capacitance into elastance yields 376 mmHg/ml. In experiments on isolated rat hearts, Kühn and Brachfeld (14) set the elastance of their windkessels to 300 and 675 mmHg/ml, respectively, i.e., their larger value is very similar to the elastance calculated in our pilot experiments on rats with intact circulation.

In the present heart-lung preparation coronary compliance represented combined values of right and left ventricular vessels of different size and was on average 0.066 ± 0.060 ml · mmHg¹ · 100 g¹. Many studies on coronary compliance were performed on the dog heart, and compliance of extramural arteries was found to be in the range of 0.001 to 0.002 ml · mmHg¹ · 100 g¹ (for literature review see Ref. 6). Most of the coronary capacitance, however, resides in the small intramural vessels, so that calculated values of intramyocardial compliance are far greater than the epicardial compliance (1, 19). The compliance of small intramural coronary vessels of the dog heart depends on distending pressure and vascular smooth muscle tone and has been evaluated to be in the range of 0.01 to 0.07 ml · mmHg¹ · 100 g¹ (2, 15, 20). The relatively high values of coronary compliance calculated from the present experiments corroborates the presumption that in the rat coronary capacitance is primarily located in small intramural vessels.

Left ventricular contraction causes systolic inflow into the coronary arteries distending the capacitance vessels. The time constant by which intramyocardial blood volume changed was in the order of 0.3 s. Lee et al. (15) measured time constants between 55 ms when tone was present and 105 ms with maximal dilatation. On the other hand, time constants up to 3.2 s have been reported (7, 20). Not only, however, do distending forces act on intramural vessels. Because ventricular contraction squeezes blood from the capacitance vessels, distending and compressing forces act simultaneously (3, 4, 20). Depending on whether the distending or the compressing forces predominate, systolic coronary flow will become antegrade or retrograde (3, 8, 13, 20).

The rate of change of vascular diameter causes a capacitance flow. In the present experiments, capacitance flow depended on coronary compliance (Fig. 5C). At a high coronary compliance, capacitance flow was high and much of the systolic inflow was stored. The stored volume was available during the next diastole, in which diastolic inflow was low. Consequently, the proportion of systolic to diastolic inflow was high at high coronary compliance (Fig. 5B). On the other hand, at low coronary compliance, systolic inflow and the stored volume were low, so that diastolic inflow predominated and hence the proportion of systolic to diastolic flow was low.

The essential improvement of the present heart-lung preparation is that it permits separation of systolic from diastolic coronary flow and it allows the effect of coronary compliance on coronary circulation in the beating heart to be studied. It could be shown that the proportion of systolic to diastolic coronary flow varies considerably. Systolic as well as diastolic coronary flow depends on coronary compliance because the quantity of blood stored in the coronary arteries during systole and flowing out during the next diastole is limited by the compliance of the coronary arteries.