Donald, David E.

Determinations of pressure, blood flows, and calculated pulmonary vascular resistance were made prior to and during acute individual occlusion of the right and left main branches of the pulmonary artery in random sequence in eight anesthetized, closed-chest dogs witn chronic atrial septal defects. Similar observations were made with occlusion of only the left pulmonary artery in three additional dogs. Obstruction of the pulmonary artery to one lung was immediately followed by a rise in systolic pressure in the right ventricle and pulmonary artery proximal to the obstruction, from an average normal value of 33.1 mm. Hg to an average value of 40.9 mm. Hg (range 32 to 47). This change was not associated with demonstrable systematic effects on right atrial or systemic arterial pressure or on the systemic blood flow. As would be expected, the average calculated ratio of pressure to flow (vascular resistance) across the unoccluded lung was elevated during the obstruction. The resistance value during the occlusion, however, was less than the estimated pressure/flow ratio for this same lung, based on the measurements made prior to the occlusion. This apparent effective decrease in vascular resistance of the unoccluded lung is probably related to dilation of the pulmonary vessels as a result of an increase in intravascular pressure caused by the sudden removal of approximately half of the pulmonary vasculature from the circulation. In spite of the elevation in right ventricular systolic pressure, occlusion of the left pulmonary artery was associated with a slight increase in pulmonary blood flow and a definite increase in the magnitude of the left-to-right shunt across the atrial septal defect, whereas no systematic effects were obtained during obstruction of the right branch in association with the same approximate levels of right ventricular systolic pressure. These data are regarded as further evidence that the preferential shunting of blood from the right lung in the presence of an atrial septal defect of the secundum type is primarily due to the anatomical relation between the opening of these vessels into the left atrium and the position of the defect in the atrial septum. Furthermore, they emphasize the importance of dynamic factors, related to the manner, site, and direction in which the pulmonary venous blood drains into the left atrium, as important determinants of the left-to-right shunt via such defects.

Systemic and pulmonary blood flows and pressures were determined in anesthetized closed-chest dogs with chronic atrial septal defects and in dogs with intact cardiac septa. In the former animals, graded obstruction of the main pulmonary artery associated with right ventricular systolic pressures of more than 80 mm. Hg was produced without evidence of failure of the right side of the heart. A considerable decrease in the magnitude of the left-to-right shunts and an increase in the size of the right-to-left shunts occurred; these changes were associated with only minor changes in filling pressure of the right ventricle, and systematic changes in systemic arterial pressure did not appear. Analysis of right atrial pressure pulses suggested the development of an increased force of contraction of the right atrium and no evidence of tricuspid regurgitation under these circumstances. In contrast, partial obstruction of the main pulmonary artery in the dogs with intact cardiac septa, so as to increase the right ventricular systolic pressure to more than 60 mm. ITg, was associated with pronounced increases in mean right atrial pressure and striking decreases in cardiac output and systemic arterial pressure. The pressure recordings from the right atrium and indicator-dilution curves demonstrated the development of significant tricuspid regurgitation. The deterioration of right ventricular function in dogs with intact cardiac septa appears to be related to overdistention of the right ventricle and possibly also to a disproportionate reduction in coronary blood flow to the myocardium caused by a decrease in systemic arterial pressure and the concomitant increase in resistance to coronary venous drainage from the right side of the heart due to the increased right atrial and right ventricular pressures. It is concluded that under the conditions of these studies, an atrial septal defect provides a considerable increase in resistance to the heinodynamic effects associated with right ventricular hypertension caused by acute partial obstruction of the main pulmonary artery. The possible duration of this beneficial effect beyond the 33 to 160 minutes of observation in these studies has not been explored.

Indicator-dilution curves were recorded simultaneously from the left atrium and the femoral artery during injections of indocyanine green into the left ventricle in normal dogs and in dogs with chronic mitral regurgitation. A study was made-without the use of thoracotomy-of the effect of the following factors on the amount of early appearing indicator detected in the left atrium: (a) the timing of the injection of indicator in relation to the cardiac cycle, (b) the position of the injecting catheter in the left ventricle, and (c) the position of the sampling catheter in the left atrium. It was found that the time and site of the injection into the left ventricle had little effect on the amount of early appearing dye detected in the left atrium and that, although the position of the sampling catheter in the left atrium was important, reproducible results were attained when the sampling site was restricted to a region in close proximity to the mitral valve. A close correlation was demonstrated between the volume of regurgitation determined independently on a hydraulic basis and the volume estimated from the fraction of early appearing indicator detected in the left atrium. It is concluded that further work with this upstream-sampling technique is justified in order to study its possibilities as a practical method for detecting and quantitating mitral regurgitation in human beings.

In 13 canine hearts, 158 disappearance curves for 133 Xe and antipyrine-1 125 I, given by intra-arterial slug injection, were recorded at a wide range of perfusion rates. Flow rates (ml/100 g/min) calculated from these curves by a variety of methods were compared with measured flow rates (F a ) per weight of perfused tissue. Perfusion of isolated, supported hearts and of anterior descending coronary arteries in open-chest dogs provided similar data. The semilogarithmic slope of curves from apex or whole heart decreased with time, particularly at high flow rates. There was a small, consistent difference in shape between antipyrine and xenon curves, suggesting that radioactivity in fat contributed somewhat to this tailing. Estimation of flow rate from the steepest semilog slope yielded an average value of 1.1 F a for all rates; estimation from slope at 30% of peak radioactivity gave 0.9F a . The curves were closely described by a two-exponential equation which gave flow estimates of 0.95F a when collimation limited the observations to the heart apex, and lower values when the whole heart was observed. Peak height/area methods gave values of approximately 0.75F a in spite of various compensations for the impossibility of recording the curve until radioactivity = 0.

The effect of exercise on the vasoconstrictive response to sympathetic stimulation was studied by observing the changes in perfusion pressure in dogs' hindlimbs perfused at constant flow. During steady-state simulated exercise, a time-dependent increase occurred between 10 and 30 minutes in the magnitude of response to lumbar trunk stimulation or carotid sinus hypotension. This was present (although to a reduced Degrees) in drug-induced vasodilatation and in exercising limbs perfused at constant pressure. The time-dependent change was not due to (1) repetition of stimulus, (2) artifact from muscle electrodes, (3) circulating vasoactive substances, (4) skin circulation, or (5) pH, P CO 2 , or temperature of the perfusing blood. When norepinephrine was used as the vasoconstrictive stimulus, no time-dependent change was noted. After stabilization, the frequency-response curve of exercising muscle exceeded that at rest at all blood concentrations of norepinephrine and at all save the lowest frequencies of stimulation. It was concluded that vascular smooth muscle can develop greater tension from greater initial length. However, during exercise some factor which operates between the sympathetic nerve ending and the receptor site on vascular smooth muscle initially tends to depress this enhanced ability to develop tension.

A field study was made of the effect of β-adrenergic receptor blockade (propranolol; Kö 592) on the racing performance of six normal greyhounds and three with chronic extrinsic cardiac denervation. Records were made of the time to race a 5/16-mile course before and after blockade; the heart rate was recorded by telemetry. After the administration of the β-adrenergic receptor-blocking agent to the normal greyhound, racing time was slightly increased, and maximal heart rate was slighdy decreased. After blockade in the cardiac denervated greyhounds, racing time was prolonged, cardiac acceleration was severely limited, and the animals finished running in a state of collapse. The data indicated diat die cardiostimulant action of both sympathetic nerves and circulating catecholamines was necessary for maximal performance. After blockade of one or the other of these mechanisms, racing performance was slightly reduced from normal. Withdrawal of both adjuvants severely limited the performance of maximal exercise.

In dogs studied without thoracotomy, under morphine-pentobarbital anesthesia and with an average heart rate of 71 beats/minute, atrial systole contributes to ventricular function to a small but significant degree. Stroke volume increased by 7 to 8% when atrial systole was timed to occur 0.1 second prior to closure of the atrioventricular valves, and decreased a similar magnitude when the contribution of atrial systole to ventricular filling was rendered ineffective by the use of an A-V delay of zero so that atrial emptying into the ventricles was prevented by the closed A-V valves. Increases or decreases in right atrial pressure of several centimeters of water induced by infusion of 6% dextran and by hemorrhage, respectively, were not associated with demonstrably significant alterations in these effects.

The distances between the aortic valve and sites in the aorta from which no retrograde flow of injected indicator could be detected in the left ventricle were determined, without thoracotomy, in two normal dogs and in ten dogs with chronic experimental aortic regurgitation. It was found that these distances attained maximal values when the aortic injections were timed to occur just at the end of systole or onset of diastole. Distances downstream from the origin of the left subclavian artery from which no retrograde flow occurred to this vessel were determined similarly in these same animals and, in confirmation of Warner and Toronto, similar relationships to the timing of the injections were obtained. These distances down the aorta, designated as "disappearance distances", were converted to "disappearance volumes" by measuring the diameters of the silhouette of successive aortic segments from aortograms made after contrast medium was injected into the left ventricle of these animals. The regurgitant flow through the aortic valve per heartbeat was estimated on the basis of the amount of immediately appearing indicator detected in the left ventricle following injection of indicator approximately 1 cm downstream from the aortic valve. Correlation coefficients of 0.88 and 0.77 were obtained between these stroke-regurgitation values and the disappearance volumes from the aortic valve and subclavian artery, respectively. The standard errors of the estimate of the regurgitant flow per beat from these regression equations were ± 2.7 and ± 2.9 ml/beat, respectively. The correlation coefficients between stroke-regurgitation values and the disappearance distances from the valve and subclavian artery were 0.83 and 0.81, respectively, and the respective standard errors of the estimate of stroke regurgitation from these regression equations were ± 3.2 and ± 3.3 ml/beat. It is concluded that determination of the maximal distances downstream in the aorta of sites from which retrograde flow of injected indicator can be detected in the left ventricle or subclavian artery is a valid method for estimating the severity of experimental aortic regurgitation in dogs. Since determinations of the disappearance distance from the aortic valve can be carried out readily in conjunction with measurement of the left ventricular regurgitation fraction, it is suggested that advantage be taken of the increase in reliability which accrues from the combined use of these two independent methods for the estimation of the degree of aortic regurgitation.

The square-wave electromagnetic flowmeter was used to assess the upstream sampling dye-dilution technic for quantitating aortic regurgitation by simultaneous determinations of the regurgitant fraction by flowmeter, as given by the ratio of regurgitant volume to total stroke volume (RV/TSV) and the identical ratio of regurgitant flow to total forward flow (Q R /Q F ), and the regurgitant fraction by dye curve, as given by the ratio of the areas of the immediately appearing portion of the left ventricular curve to the area of the femoral artery dye curve (LV A /FA A ), after injection of indicator approximately 1 cm downstream to the aortic valve. A good correlation (r = 0.91) was found between these two independent methods which can be represented by the regression equation Q R /Q F = 0.82 (LV A /FA A ). This correlation indicates that the concept that the regurgitant fraction of indicator (LV A /FA A ) is proportionate to the regurgitant fraction of blood (RV/TSV) is valid. The observed systematic difference between flow-meter and dye-curve regurgitant fractions may be due to the fact that time-average rather than volume-average concentrations of the indicator are obtained by sampling of blood through eatheters. Two different and independent calibration technics were used to establish a calibration factor for converting recorded flowmeter deflections to volumetric units for measurement of rates and volumes of blood flow in the ascending aorta. The first technic involved simultaneous recording of dye-dilution curves and velocity pulses in the ascending aorta by the flowmeter before the production of aortic regurgitation. The second involved bidirectional perfusion of the isolated aorta at known rates of blood flow with simultaneous flowmeter recordings at the termination of each experiment. The data obtained validated the use of the flowmeter for measurements of total stroke volume and regurgitant stroke volume. The close correspondence between flowmeter recordings of the regurgitant flow and hydraulic estimates of regurgitant flow, based on the size of the defect in the aortic valve measured at necropsy and continuous records of the pressure gradient across the valve, also lent strong support to the validity of these measurements. The relationship Q R = Q S (Q R /Q F )/(1 - Q R /Q F ) can be used to estimate regurgitant flow from values of the regurgitant fraction (Q R /Q F ) and the systemic flow (Q S ). The mathematical relationship of the regurgitant fraction to regurgitant flow, determined by the dilution-curve technic, plus the variability obtained between simultaneous flowmeter and dilution-curve estimates of regurgitation, indicate that the error in the estimation of regurgitant flow by the dye-dilution technic is least for small-to-moderate degrees of regurgitation and increases progressively with the severe degrees of regurgitation associated with regurgitant fractions of greater than 0.6. On the basis of 231 simultaneous determinations by the indicator-dilution and flowmeter technics carried out in nine dogs, and assuming that the flowmeter values were uniformly correct, it was estimated that the standard deviation of the errors in measurement of regurgitant flow by the upstream sampling dye-dilution method in these experiments was 0.15 L per minute for regurgitant fractions below 0.6. and that this variability increased so that the standard deviation was 1.3 L per minute at regurgitant fraction levels of 0.9 to 1.0. The results of determinations carried out over a range of heart rates from 39 to 216 beats per minute indicate that the indicator-dilution method is applicable over a wide range of heart rates. The observed changes in aortic regurgitation with variation in heart rate correlated closely with the changes estimated by hydraulic equations based on the hydrodynamics of steady flow through an orifice. The increase in regurgitant stroke volume with increased duration of diastole at slow heart rates was disproportionately greater than the increase in total stroke volume, so that the effective forward (systemic) flow fell to extremely low levels at slow heart rates, particularly in dogs with large defects in the aortic valve. The applicability of the upstream sampling indicator-dilution technic to the detection and estimation of the severity of aortic regurgitation in human beings merits further study.

In dogs studied without thoracotomy the detection of immediately appearing indicator in the left ventricle following its injection approximately 1 cm downstream from the aortic valve can be used as a reliable index of the presence or absence of aortic regurgitation. With this type of injection and under the conditions of these experiments, the ratio of the area encompassed by the immediately appearing portion of the dilution curve recorded from the left ventricle to the area encompassed by the primary portion of the curve recorded at the femoral artery bears a close positive correlation (correlation coefficient: 0.9) to the severity of aortic regurgitation as estimated by back perfusion of the valve at necropsy. Within limitations, the position of the tip of the sampling catheter in the left ventricle in dogs with aortic regurgitation is not an important determinant of the amount of immediately appearing indicator detected in this chamber. Furthermore, although the amount of indicator detected in the left ventricle following injections of very short duration is time-dependent, if the duration of the injection is extended to cover the full systolic or diastolic phase of one heartbeat or continues over one or more heartbeats, the exact timing of this injection in relation to the cardiac cycle is not an important determinant of this variable. It is concluded that under these conditions the positive correlation established between the amount of immediately appearing dye detected in the left ventricle (expressed as the regurgitant fraction) and the severity of aortic regurgitation determined by back perfusion at necropsy can be used as a valid means of estimating the severity of aoritic regurgitation in dogs. The applicability of this indicator-dilution method to the study of clinical aortic reguragitation merits further study.