Reprinted from THE AMERICAS JOURNU. OF P~rs~o~ocu Vol. 182, No. I, July, 1955 Printed in U.S.A. Effects of Varying the Output of a Mechanical Left Ventricle on the Circulation in the Dog1 EDWARD D. FREIS, HERBERT P. BROIDA, CHARLES A. HUFNAGEL AND JOHN C. ROSE From the Cardioaascdar Research Laboratory, Georgefom Unioersity Hospital, the Departments 0.i Medicine and Surgery, Georgetown L'niaersity .Wtnol qf' Medicine, and the Heat and Power Dikion, Nn~ional Btweau of Standards, U'asltin@on, D. C. T HE EFFECTS of varying filling pressures and peripheral resistances upon the out- put of the isolated heart have been known since the Starling `heart-lung prepara- tion' was devised (I). In such experiments an intact heart and lung were used, the venous return and peripheral resistance being varied under controlled conditions. The reverse ex- periment, that of using an intact peripheral circulation and an artificial left ventricular pump, which can be varied at will over a wide range of outputs and rates, has not been at- tempted until recently. An experimental heart pump permitting such studies has been de- veloped (z), and the present report is concerned with the effects of variations in rate and output of this pump upon systemic arterial, venous, and pulmonary arterial pressures and pulmo- nary blood flow. MATERIALS AND METHODS All experiments were carried out on healthy mongrel dogs weighing 16-25 kg. Each animal was anesthetized with sodium pentobarbital 25 mg/kg administered intravenously and maintained with endotracheal oxygen. A detailed description of the apparatus, the surgical technique for replacing the left ventricle with the artificial pump, and the methods of recording have been provided in a previous communication (2). A pump output of KUX~~~OO ml/min. was used at the time that the extracorporeal circulation was sub- stituted. The systemic arterial pressure pulses were then observed for the presence of `impure' pulse waves indicating a continued output from the left ventricle as well as from the pump (2). If such `contaminated' waves were observed the drainage tube in the left auricle was adjusted until the pressure pulses became pure indicating that the bypass had become complete. In manyinstancescontamination of the systemic arterial pulse waves with left ventricular blood could not he Received for publication December 14, 1954. 1 Supported in part by a research grant from the American Heart Association. completely eliminated, especially at high outputs. How- ever, these were easily recognized by simple inspection of the tracings and were eliminated from the reported data. Although substitution of a pump for the left ventricle was carried out in 36 dogs (21 with the pump described) a major portion of the data reported here was obtained in five animals where complete satisfactory experiments were obtained. RESULTS Effects of Variations in Stroke Volume and Pump Rate With Constant Minute Outputs. If the minute output was constant proportionate variations in stroke volume and heart rate did not alter the mean arterial pressure. As the rate was decreased and the stroke volume in- creased proportionately to keep the minute output constant, there was an increase in pulse pressure due to both a rise in systolic pressure and a decrease in diastolic pressure (fig. I). However, mean systemic arterial pressure, venous and pulmonary arterial pressures and return flow remained unchanged. Right Ventricular Rate. Since only the left ventricle was being bypassed in these experi- ments right ventricular function remained in- tact. There was no tendency for the right ventricle to assume the same rate as the pump. There also were no consistent changes in right ventricular rate when the output of the left ventricular pump was changed. Marked in- creases in the left ventricular output from IOOO to more than 3000 ml/min. were not accom- panied by an increase in the rate of the right ventricle. Thus, the changes in right ventricu- lar output described below are due almost entirely to changes in stroke volume, rather than rate. The mean right ventricular rate for the en- tire group of animals was 136 (range 75-180) beats/min. However, in the great majority of I92 FREIS, BROIDA, HUFNAGEL AND ROSE experiments the rates of the right ventricle varied between 120 and 160 beats,/min. In any given animal the extreme range of change in rates during the experiment averaged 25 (range 10-45) beats/min. The changes ob- served, as stated above, however, followed no consistent pattern in relation to changes in pump output (table I). That this pattern of right ventricular re- sponse was not due to the presence of moder- ator reflexes is suggested by the fact that blocking doses of hexamethonium failed to alter the relative constancy of right ventricular rate. However, the right ventricle remained capable of increasing its rate during these ex- periments since if epinephrine was injected a significant tachycardia resulted. Mean Arterial Pressure. At the lowest out- put of I ljmin. the mean arterial pressure ranged between $3 and 64 mm Hg. As the out- puts were increased there was a rise in arterial pressure. At the highest output of 3.5 l/min. the range of mean arterial pressure varied be- tween IOO and 123 mm Hg. For a given change in output the increments in pressure usually appeared to be greater in the lower ranges of outputs than in the higher ranges. As the out- puts were decreased the descending limb of the pressure output curve tended to approximate a mirror image of the ascending limb (fig. 2). Pulmonary Arterial Pressure. The pressure in the pulmonary artery also rose with increases in output. The ranges of mean pulmonary arterial pressures were 11-20 mm Hg at out- puts of 1.0 l/min. and 24-36 mm Hg at outputs ARTERI 150 100 50 RATE- 100 per min. 50 per min. STROKE VOL.- 1st~. 30cc. MEAN PRESSURE-74 mmH 74 mm. Hg. RETURN-1.58L. I 1.58 L. per mia FIG. 1. Cuttings of systemic arterial pressure pulses in a pump dog showing that at a constant pump output of 1500 cc/min. alterations in rate alter only the pulse pressure. The mean arterial pressure and venous and pulmonary arterial pressures as well as total left auricular return flow remain the same. TABLE I. KELATIONSHIP OF THE RATE OF THE IN- TACT RIGHT VENTRICLE TO STEPWISE ALTERATIONS IN MINDTE `LEFT PUMP' OUTPUT ~- 1000 1250 ISW 1750 2000 2250 PjOO 3000 3250 3500 3000 2250 2000 1750 1500 1250 1000 Av. R.V. rate Range R.V rates Right Ventricular Rate, strokes/min. Dog 21 Dog 26 Dog 35 Dog 36 150 140 I44 140 132 150 160 160 150 150 90 90 ii," 90 132 II0 120 140 20-q 0'1 140 32-14. 155 50-16~ ,8 120 90 `4-120 `Left pump' rate IOO strokes/min. in all experiments. of 2.5 l/min. In the two animals in which the outputs were increased to 3.5 l/min. the pul- monary arterial pressures in both cases rose to 31 mm Hg. When pressure changes were plotted against output changes on a semi-logarithmic scale there was a striking parallelism of pulmonary and systemic arterial pressure curves in the individual dog (fig. 2). In all five dogs the slope and configuration of the pulmonary arterial pressure curve closely patterned the systemic arterial curve for each animal. In addition, the position of the pulmonary and systemic pres- sure-output curves for each dog in relation to the curves for the other dogs was the same in both the pulmonary and systemic graphs. Since, as discussed below, there was storage of blood in the dogs' vascular system when the pump outputs were increased, it seemed pos- sible that the parallelism between pulmonary and systemic pressure-output curves in indi- vidual dogs was due to proportionate increases in blood volume in the pulmonary and systemic arteries. If this were the case, however, the blood storage curve should parallel the pres- sure curves. Blood volume changes were deter- VARYING VENTRICLE OUTPUT AND CIRCULATION I93 MEAN 100 SYSTEMIC ARTERIAL 60 PRESSURE (MM HG) 60 30 MEAN PULMONARY ARTERIAL IO + STEPWISE OUTPUT INCREASES --3 STEPWISE OUTPUT DECREASES I I LEGEND o 20 0 21 * 26 0 35 . 36 I I I I I I I I I I I I I II I III III III III I.0 1.4 I.6 2.2 2.6 3.0 3.4 3.4 3.0 2.6 2.2 I.6 1.4 I.1 "LEFT PUMP" OUTPUT (LITERS) FIG. 2. Chart showing the logarithmic plot in 5 dogs, systemic arterial pressure and mean pulmonary arterial pressure against pump output as the outputs are increased and then decreased in steps. See text for discussion. mined in dogs 35 and 36. As the outputs were increased there was no close parallelism be- tween the curves of blood storage and pressure 0%. 3). Venous Pressure. The responses of venous pressure to changes in pump output were small (fig. 4). At the lowest pump outputs the venous pressures varied between 20 and 55 mm Hz0 and at the highest outputs between 50 and 60 mm HzO. It must be emphasized that' when the heart begins to fail the venous pressures become much higher than those reported here (3). Although the venous pressure did not parallel the systemic arterial pressure as closely in the individual dogs as did the pul- monary arterial pressure, the directional trends were the same. It should be pointed out that the venous pressure was difficult to measure accurately because of the small magnitude of the changes and also because of fluctuations produced by the positive pressure respirator. Possibly under more ideal conditions a closer parallelism with arterial pressure might have been observed. Pulmonary Blood Flow. When the pump outputs increased the animal took up blood from the reservoir and when they decreased the reverse occurred with bleeding of the animal into the reservoir. Thus, in six experi- ments (in 4 dogs) in which the pump output was increased by I liter usually in steps of 250 cc there was a loss from the reservoir varying between 250 and 475 cc (average 345 cc) (table 2). Similarly, in the same experiments, when the outputs were decreased to the original level there was a gain in reservoir volume vary- ing between 50-500 cc (average 300 cc). The greatest change in reservoir level occurred usually within the first IO seconds after the pump output had been altered. After 30 sec- onds the reservoir level became constant indi- cating that the pulmonary blood flow was now equal to the output of the pump. At a later 194 FREIS, BROIDA, HUFKAGEL BND ROSE B 1 E-0 : L -PRESSURE E L 1 - - - BLOOD STORIGE `LEFT PUMP" OUTPUT (LITERS, I I I I I / I I I / I `O 1.0 1.4 1.8 2.2 2.6 3.0 A I FIG. 3. Chart constructed to compare alterations in mean pulmonary arterial pressure and blood storage in 2 dogs as pump outputs are increased in steps. The MPA pressures are plotted logarithmically. Blood storage is plotted on a similar scale for purposes of comparison. See text for discussion. stage in each experiment one of the indications of the onset of circulatory failure was a failing reservoir level even when the output was main- tained at a constant value. DISCUSSION Mean arterial pressure was uninfluenced by changes in rate and stroke volume so long as the minute output was kept constant. This effect has been surmised by others (4) but the evidence has not been conclusive due to the fact that in the intact animal changes in rate or stroke volume probably result in changes in minute output . Thus, Wiggers describes experi- ments in which cutting the vagi to produce tachycardia or infusing blood or saline to in- crease stroke volume resulted in an increase in mean arterial pressure (4). As a corollary, vagal stimulation or hemorrhage reduced mean pres- sure. In all of these experiments it was recog- nized that the minute output changed. In the present experiments rate and stroke volume could be varied without changing minute out- put. By this means it has been demonstrated conclusively that mean pressures and flows in the circulation are dependent upon minute output and not upon stroke volume or heart rate, per se. When the stroke volume was increased and the rate proportionately decreased with con- stant minute output there was a slight increase in systolic and decrease in diastolic pressure. These changes can be explained readily by the well known fact that the pressure pulse varies with cardiac ejection and arterial drainage (5). When cardiac ejection increases systolic pres- sure rises and when cardiac rate decreases arterial drainage is more complete resulting in a reduction of diastolic pressure. Despite marked increases in pump output with which the intact right ventricle kept pace `there were no significant changes in right ventricular rate. It is obvious, therefore, that the increase in right ventricular output was ac- complished entirely by an increase in right ventricular stroke volume rather than by rate changes. Thus, in agreement with Ballin and FIG. 4. Chart showing the logarithmic plot, in 1 dogs, of pressure against minute pump output as the outputs are in- creased and then decreased in steps. See text for discussion. -LEFT PUMP" OUTPUT UTERS, VARYING VENTRICLE OUTPUT AKD CIKCULATIOK 19.5 Katz (6), there was no evidence for the opera- tion of a Bainbridge type of reflex in these ex- periments. These results are also similar to those obtained by Patterson, Piper and Star- ling in the isolated heart lung (7). There was a crude, non-linear relationship between changes in right ventricular stroke volume and venous pressure. In contrast to the absence of change in mean arterial pressure with proportionate changes in rate and stroke volume there was a significant increase in mean pressure when the minute out- put was elevated. The fact that the increments in pressure usually were less at higher outputs indicated that the peripheral resistance de- creased as the outputs were increased. The shape of the pressure output curve cannot be explained by variations in blood storage since the curve of blood storage during increasing outputs failed to parallel the pressure curve (fig. 3). The reason for the observed pressure output curve, therefore, is not apparent at this time and requires further study. Equally difficult to explain is the unusual parallelism observed between the pulmonary and systemic pressure output curves. The most striking observation was that the shape and position in relation to the other animals of the pulmonary pressure-output curve for each dog was generally the same as the shape and po- sition of the systemic pressure-output curve for that dog. Again the shape of these curves did not correlate with the blood storage curve. This observation suggests the possibility of an integrative mechanism whereby vasomotor tone is regulated simultaneously in the pulmo- nary and systemic circulations. However, no such integrative mechanism has been demon- ,strated or even suggested previously. As the pump outputs were increased there was a brief period in which the pump output exceeded the pulmonary venous drainage into the reservoir. It is primarily during this period that blood is stored. After this initial adjust- ment period the pulmonary venous drainage approaches the pump output and eventually an ,equilibrium is reached. The equilibrium is as- sociated with an increase in intravascular pressures and a decrease in reservoir level as compared to the equilibrium existing prior to the increase of output. The rise in all of the pressures suggests that the excess blood is TABLE 2. GAINS AND LOSSES m BLOOD VOLUME WITH I000 CC/MIX. INCREMENTS AND DECREMEKTS IN `LEFT PUMP' OC'TPVT Dog h-0. 31 31 3.5 34 AV. Dog No. 3' 31 35 Increasing Outpilt Output Altera&m, Blood Vol. Gain, cc/min cc 1000-2000 475 ropz-2'00 3 2.5 1000-2000 4oo 2coo-3000 3oo 1000~2000 2.50 2000-3000 325 34s Demwsin~ Outpzds Output Alteration, Blood Vol. Loss, cc/min cc 2000-1000 450 2100-1050 400 3000-2000 22j 2000-1000 500 3000-2000 50 2000-1000 `SO 3oo stored in both the greater and lesser circula- tions. That the storage is not primarily in the large veins is suggested by the fact that, when the right ventricle is functioning effectively as in these experiments, the venous pressure may not show much change. This is in contrast to the dog with a poorly functioning right ven- tricle, however, who as will be shown in a later communication, exhibits marked elevations of venous pressure (3). An exactly opposite se- quence of events occurs when the pump output is decreased. SUMMARY AND COXCLUSIOA-S A diaphragm type pump with independent controls of rate and stroke volume was substi- tuted for the left ventricle in dogs and the following observations made: mean arterial pressure was a function of minute output rather than of stroke volume or ventricular rate, per se. Increase in stroke volume with maintenance of a constant minute output by proportional reduction of pump rate produced an increase of pulse pressure without change in mean pres- sure. Increase in pump output produced paral- lel increases in the output of the intact right ventricle. The increase in right ventricular out- put was accomplished almost entirely by an increase in stroke volume rather than in right 196 FREIS, BROIDA, HUFNAGEL AND ROSE ventricular rate. With stepwise increases and then decreases in pump output the mean sys- temic and pulmonary arterial pressures usually exhibited parabolic pressure output curves characterized generally by larger pressure increments and decrements at the lower out- puts. The pulmonary pressure output curve tended to parallel the systemic curve in indi- vidual dogs. The reasons for these character- istics of the pressure-output curves are not clear at this time. Increase in pump output in these experiments produced blood storage. When the right ventricle was functioning efficiently there were negligible changes in venous pressure, suggesting that the excess blood was not stored primarily in the large veins. This technique appears to provide a method for exploring the responses of the cardiovascu- lar system to controlled changes in left ven- tricular output. I. 2. 3. 4. 5. 6. 7. REFERENCES KNOWLTON, F. P. AND E. H. STARLIXG. J. Physiol. 44: 206, 1912. ROSE, J. C., E. D. FREIS, C. A. HUFNAGEL, H. P. BROIDA, J. F. GILLESPIE AND P. J. RABILE. J. Appl. PhySiOl. 7: 580, 1955. ROSE, J. C., E. D. FREE, HUFNAGEL, C. A. AND H. P. BROIDA. In preparation. WIGGERS, C. J. Physiology in Hezlth and Disease. 5th Ed. Philadelphia: Lea, 1949, p. 606. HAMILTON, W. F. .4rlerial Pulse. In: Medical Physics, edited by 0. GLASSER, Chicago: Yr. Bk. Pub., 19, vol. I, p. 7. BALLIN, I. R. AND L. N. KATZ. Am. J. Physiol. 135:202,1941. PATTERSON, S. W., H. PIPER AND E. H. STARLING. J. Pkysiol. 48: 465, 1914.