AMERICAN JOURNAL. OF PHYSIOLOGY Vol. 228, No. 2, February 1975. PrinseQ in U.S.A,
Cardiopulmonary ramp,
baroreflexes:
and square-wave THOMAS Defartment
C. LLOYD, JR. of Physiology, Indiana
stimulation
University Medical
LLOYD, rr~~~~~ C., JR. Curdiopulmonary horeflexes: effects of staircase, ramf, and square-wave stimulation. Am. J. Physiol. 228(2) : 470-476. 1975.--Anesthetized dogs were pump perfused in a system that held systemic arterial pressure constant. Pulmonary arterial beds were pressurized, and the induced transient falls of systemic vascular resistance (SVR) were measured. An approximately linear relationship between percent fall of SVR and pulmonary artery pressure was obtained. Left atria and pulmonary venous beds were pressurized using square-pulse, staircase, and ramp wave forms, and changes in SVR were measured. The effect of varying the interval between stimuli was appraised. An approximately linear relationship was obtained for responses to square-wave left atria1 forcing pressure (Ha). With staircase forcing, the change of SVR per unit change of Ha depended upon prestimulus Pla. With ramp forcing, responses increased approximately linearly with respect to the logarithm of the rate of pressure rise. Responses to square-wave stimulation varied directly with respect to the interstimulus interval. This study sho\vs that left atrial-pulmonary vein baroreflexes are capable of producing substantial short-term falls of SVR, and that the response depends upon stimulus pressure, rate of pressure change, base-line pressure, and pressure history. left blood
atrium; pressure
pulmonary regulation
vessels;
systemic
vascular
effects of staircase,
resistance;
I T HAS BEEN SHOWN that stimulation of stretch receptors in the pulmonary artery, or left atrium and pulmonary veins may aflcct systemic vascular resistance. Distention of one or another of these chambers typically produces systemic vasodilation (7, 8, 16, 17, 23, 26, 28), although not always (2, 4, 10, 13-15). While the role of these cardiopulmonary barorcflexes in homeostasis remains uncertain, there is evidence that they provide a continuous sympathoinhibitory influence (19, 22). Aff ercnt nerve discharge patterns imply the existence of several types of cardiopulmonary receptors which respond to strain, or rate of strain, or both, and which may show different rates of adaptation (26-28). Some of the apparent differences among receptors may relate to their mode of attachment to the walls of the chamber in question (1, 27). Left atria1 type-B receptors are slowly adapting and their discharge frequency parallels the atria1 v-wave (25). The role and behavior of type-A receptors are less uniformly agreed upon, but they may be rapidly adapting and sensitive to rate of change of left atria1 wall tension (26). Pulmonary arterial baroreceptors resemble the systemic arterial baroreceptors in having responses that are prominently influenced by pulse pressures
Center,Indianapolis,
Xndiana 46202
(4). More recent studies have established the existence, if not always the role, of other receptors which utilize nonmyelinated afierent fibers (3, 20, 2 1). The neurophysiologic data have not been completely accompanied by equivalent hemodynamic data which would display such phenomena as rate sensitivity and adaptation. Nor have the systemic vascular elects of reflexes from the pulmonary arterial bed been contrasted in the same preparation with those from the pulmonary veins and left atrium. The present study was designed to further characterize the cardiopulmonary baroreflexes by answering the following questions: 1) how do the responses to square-wave forcing of pulmonary arterial pressure compare with responses while similarly forcing the pressure within the pulmonary veins and left atrium; 2) is there an influence of prestimulus left atria1 pressure upon the change in systemic vascular resistance produced per unit change in left atria1 pressure; 3) does the rate of rise of pressure influence the hemodynamic response to left atria1 distention; 4) for a given left atria1 square-wave pressure stimulus, is there an effect of the interstimulus interval duration upon response? This investigation was carried out with a canine preparation using cardiac bypass perfusion under the condition of constant systemic arterial pressure. Our procedure incorporated a method that provided for separate distention of the pulmonary arterial bed and of the pulmonary venous bed and left atrium. METHOD
Sixteen mongrel dogs, weighing from 13 to 20 kg, (average, 16 kg) were used for this study+ After anesthesia with 30 mg/kg iv pentobarbital, and 1 mg/kg im morphine sulfate, the dog was placed supine and pump ventilated with room air. The chest was opened through a longitudinal sternotomy. The left subclavian artery and left lower-lobe pulmonary artery were dissected free, and the was pericardium was opened. The right atria1 appendage cannulated with an 8-mm ID stainless steel cannula for the return of blood to a pump oxygenator. The subclavian artery was divided and the central end was cannulated for perfusion of the systemic circulation. The left lower-lobe artery was divided and a centrally directed cannula was installed to provide an input port for pressure forcing. A z-mm OD catheter was advanced centrally through the right internal thoracic artery for measurement of systemic arterial pressure. Heparin (100 mg) was given to prevent clotting. Ventricular fibrillation was then induced with an
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
CARDIOPULMONARY
K4ROREFLEXES
electric shock; lung ventilation was arrested, and external perfusion was begun. The lungs passively collapsed through a water seal to an end-expiratory pressure of 3-5 cmHzO. A left ventricular drainage cannula was placed through the apex and a left atria1 cannula was placed through its appendage. Both ventricles were clamped with a single large clamp (made from a 10-inch Rochester-Ochsner forceps) placed across the ventricles as close to the atrioventricular groove as possible. This clamp isolated the ventricles from the pulmonary artery and aorta as well as from the atria. The atria continued to beat vigorously during the remainder of the experiment. Catheters for acquisition of pressures were advanced into the left atrium and main pulmonary artery through the walls of the tubing attached to their cannulas. The tubing to these cannulas provided either for drainage from, or imposition of pressure on, the respective sides of the pulmonary circulation. To effect independent distention of the pulmonary arterial or pulmonary venous-left atria1 components, we inflated the segment with gas using 95 7~ OZ-5 % CO, with the wave form and pressures desired. Surface tension prevented capillary gas passage at the pressures used (18). While one compartment was pressurized, the other was allowed to drain freely through its cannula and tubing. When the left atrium was pressurized, diversion of coronary and bronchial venous blood into the pulmonary arteries caused pulmonary arterial pressure to rise transiently. Such increases were of less than 3 cmHz0 magnitude, so pressure remained nominally at zero. Similar small changes in atria1 pressure occurred during arterial pressurization, but on a few occasions large increases of pulmonary arterial pressure caused atria1 pressure to rise above 5 cmHz0. In the ultimate analysis, the marked sensitivity of the atria1 reflex led us to discard responses to pulmonary arterial pressurization if atria1 pressure rose above 5 cmHz0. The external perfusion system consisted of a bubble oxygenator (constructed locally) and two Manostat rotary tubing pumps which were connected in parallel. The speed of one pump was manually variable; the other pump was controlled electronically by a voltage proportional to the difference between a chosen set point and systemic arterial controlled pump adequately pressure. The electronically kept pressure constant during induced falls in systemic vascular resistance as large as 30 %. Manual control of the second pump during evoked responses permitted pressure to be held constant during larger resistance falls. Since both pumps ran simultaneously, the electronic controller provided the necessary fine adjustment when manually controlling the other pump, thereby freeing the operator from the need for critical performance. The oxygenator consisted of a vertical glass column of 50 cm length and 2 cm ID, into which blood flowed by gravity. Oxygen was bubbled into the base of this column through a l-cm OD Plexiglas tube that contained 4 rows of 24 holes. Hole diameter was 0.75 mm and the rows were spaced 5 mm apart. The upper one-third of the glass column was surrounded by a 10.5 x 2 7.5 cm cylindrical Plexiglas defpaming chamber that was half-filled with polypropylene gauze mesh. Blood from the defoaming chamber flowed by gravity into a Plexiglas reservoir from
471 which it was pumped. The priming volume of this system was about 500 ml, but we intentionally filled it with 1 liter to provide sufficient fluid for the large transient flow increases produced reflexly. The priming fluid was 6 Y& dextran in 0.9% NaCl solution (Abbott Dextran 75), to which was added additional amounts of a physiological salt solution (17) if necessary during the experiment. Added volumes ranged from 300 to 1,000 ml. Oxygen flow rate was never measured, but was set to provide a pH of reservoir blood between 7.3 and 7.5 U. Arterial blood Pea always exceeded 150 torr whenever measured by intermittent sampling, and it was typically above 300 torr. Reservoir blood temperature was maintained between 34 and 38°C by heating the dog and the oxygenator assembly with infrared lamps. Except where noted, all cannulas were glass, and the interconnecting tubing was Silastic or gum rubber. After use, all components were soaked overnight in Alconox or Haemosol solution. With the exception of the Plexiglas chambers, everything was then boiled for 2 h in 2 % NaHC03 solution, then rinsed and air-dried. We continuously recorded systemic arterial pressure, pulmonary arterial pressure (Ppa), and left atria1 pressure (Pla), using the mid-chest plane for zero reference. Systemic arterial inflow was monitored using an electromagnetic Aowmeter. Systemic vascular resistance (SVR) was defined as systemic arterial pressure divided by systemic inflow. In many experiments SVR was obtained electronically and continuously recorded (16). This was not done in all experiments because of technical problems. Systemic blood flow was initially set near 100 ml/kg per min by choice of the arterial pressure set point. If base-line resistance changed much during the experiment, set-point adjustments were made to maintain the flow near the original level. The various protocols were performed in random order. We tried to accomplish several in each experiment, but it was often impossible to perform all. We discontinued an experiment if responses to a given stimulus began to diminish; if large spontaneous variations of SVR appeared; or if the reservoir fluid volume began to decline rapidly, indicating either intravascular pooling or substantial transudative loss. The five subroutines were as follows: 1) In 13 preparations gas was admitted to the left atrium to cause a square-wave pressure change which lasted about 70 s. Between 3 and 13 (typically 6) such square pulses were used in each experiment. These ranged from 6 to 43 cmHz0 in magnitude, and occurred at a rate of 1 every 3-4 min. These were not true square-wave forcing pressures, in that the rise time was finite and the corners were somewhat rounded. Rise time was made as short as possible and dP/dt always exceeded 300 cmHaO/min. Between pulses, Pla was returned to an average base-line level of 5 cmHZO. This was constant for any given experiment, but varied among the groups from 0 to 12 cmHz0 because of variations in amount of drainage and configuration of the drainage tubing. 2) The pulmonary artery was similarly distended in but the range of pressure was seven preparations,
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
472 15-90 cmHxO. Base-line pressures were between 0 and -5 cmHz0. 3) In 11 dogs left atria1 pressure was increased from 0 to 40-45 cmHz0 in a staircase fashion (see Fig+ 4). Each step was of as uniform a size as possible in each run, but within the group they ranged between 6 and 10 cmHQ0. Pressure was restored abruptly to the prestimulus base line after obtaining the response at the highest pressure. These base-line pressures ranged from 0 to 6 cmHz0. Two such runs were made in four dogs, three runs were made in five dogs, and single runs were made in two dogs. to the left atrium to 4) In 13 dogs gas was admitted generate a ramp increase of pressure from a baseline Pla of 5 cmHz0 (range O-l 2). The rise was controlled manually, guided by the pressure record. Ramp slopes ranged from 7 to 400 cmHzO/min. From three to seven rates (typically five) were observed in each dog. Pressure was returned abruptly to base-line level upon reaching 40 cmHz0. In 10 dogs Pla was raised from base line to a pressure between 22 and 32 cmHz0 (average, 28), held steady for about 60 s, and then abruptly returned to base line for a ran,domly variable period between 5 and 100 s. At the end of that interval it was abruptly raised to the previous distending level for another -60 s. This sequence was continued until a number of interpulse intervals had been observed (from five to nine per animal). Results were analyzed in the following ways. As before (16, 17), we found that responses consisted of transient SVR falls which nearly completely resolved spontaneously in spite of maintained stimulus pressure. The steadystate changes were small and variable. We elected to consider only the transient maximum fall since this was easily quantifiable, was invariably present, and had been shown to behave in an unequivocally graded fashion in earlier experiments (I 7). The percentage fall of SVR from prestimulus level to the nadir during the response was calculated for each imposition of a forcing pressure. Results pertinent to the individual experiment subroutines were 1) Data obtained during single expressed differently. square-pulse forcing were reduced to graphs of percentage SVR fall versus the Pla or Ppa which existed during the pulse. 2) During staircase forcing, each pressure level was held until SVR had stabilized. Since the purpose of the was to display the influence of prestaircase sequence stimulus pressure on the subsequent response, the transient fall induced by each step was expressed as the percentage SVR fall per centimeter Hz0 step change of Pla, and each response so expressed was plotted as a function of the Pla that was present immediately preceding the step that produced it. 3) Ramp forcing caused responses which also had transient resistance minima. Each transient minimum was expressed as the percentage fall and these were then graphed as a function of the Pla ramp slope that caused ihem. $) The results of varying the interpulse recovery interval were displayed by plotting the percentage SVR fall observed transiently upon restoration of stimulus pressure against the interpulse time.
T.
C. LLOYD,
JR.
RESULTS
The general form of responses seen upon square-wave forcing of Pla resembled those reported earlier (16, 17) The pressure control system generally he1.d systemic arterial pressure within 5 cmHz0. However, in the largest responses (50 % SVR fall or greater) of the largest dogs, the required increases in flow exceeded pumping capacity, and on those few occasions pressure briefly fell as much as 20 cmHz0. At such times systemic flow approached 4 liters/min. In each of the 13 experiments in which we sought the relationship between the magnitude of a square-wave pulse of Pla and the resulting fall of SVR, responses were related to input pressure magnitudes. The data have been combined in Fig. 1. Not apparent in the figure is the tendency for the response curve of most individual experiments to level off beyond atria1 pressures of about 35 cmHz0. The effect of this behavior on the figure is to cause a wide variation among responses above that The stimuli and responses were significantly pressure. correlated. The linear regression equation for all data is = 1.55X 3.56 (r = 0.79, 12 = 87). The x-intercept Y is at 2.3 cmHz0. Since a linear relation was most apparent only for Pla less than 35 cmHz0, a regression analysis of that data subset was done, yielding the equation y = 2.09x - 13.7 (r = 0.81, 12 = 68). In this case the x-intercept pressure is 6.4 cmHg0. Square-wave forcing of Ppa caused changes of SVR resembling those seen when forcing Pla, i.e., a transient minimum which returned to control resistance within 3060 s. The fractional change of resistance at the nadir was related to stimulus pressure magnitude. The data from all seven exneriments are displayed in Fig. 2. The linear l
0
0
0 0
0
0
o”
8, 0 000:
i
o*
8
0 0
0
0
00
0
0
0
0
0
0
8
0 0
0 0
0
0
0
0”:
0
Q 0
O0 0 0 00
1, Percentage fall of systemic vascular resistance (SVR) vs. atria1 pressure (Pla) during square-pulse forcing of Pla in 13 dogs.
FIG.
left
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
CARDIOPULMONARY
473
BAROREFLEXES
60 0 0
50 0 0
Ir40 > CT) -- 3o u w
0 0 0 eQ 0
o"
0
0 0
o
0
0
320
a
IO
0
O0
0
0
0
a
0 0
0 OO
I
IO
I
20
0
IOQ
30
pm 2. Percentage puhnonary arterial Ppa in 7 dogs. FIG.
I
1
1
I
I
40
50
60
70
80
cm
I
90
Hz0
fall of systemic vascular resistance pressure (Ppa) during square-pulse
(SVR) forcing
vs. of
regression equation for these is y = 0.5~ - 2.8 (r = 0.71, n = 39), and th c x-intercept is 5.2 cmHz0. These responses differed from those obtained by forcing Pla in that they were smaller for any given stimulus and there was no consistent tendency for responses to level off within the range of Ppa used. Increase of Pla in a staircase fashion caused transient SVR falls with each step. Typically, SVR returned to the control level after each transient fall, but occasionally there was a trend to a progressively lower plateau SVR following each step. These trends were manifestations of the small and inconsistent steady-state changes seen in these experiments and described previously (17). However, we did not assume a fixed base-line SVR for evaluation of the response, but based the fractional resistance change at each response on the resistance present immediately before each step. In all of the 11 experiments the fractional fall of SVR per centimeter Hz0 rise of Pla was related to the Pla from which each step rose. Responses were small at a low prestep Pla, rose to a maximum near Pla of 15 cmHz0, and then fell as Pla was further raised. The large variation among experiments militated against Instead presenting our findings with a graph of all data. we have grouped the data into subunits of G.5 cmHz0 prestep Pla and determined the average response obtained in each unit. (Because of the small group size, all results obtained between Pla of 35 and 45 cmHz0 have been collected into a single unit.) The unit averages are displayed in Fig. 3. Each individual experiment had a graph similar in shape to that of the figure. Recorded data from one experiment are given in Fig. 4. When using ramp wave forms to distend the left atrium, we found that transient SVR minima were again produced. The SVR fall tended to develop more slowly at the lowest pressure rise rates, and at these rates the resistance often returned toward control level while Pla was still increasing. On the other hand, at the highest rise rates, the greatest SVR fall often did not appear until after Pla was returned to base line. Since Pla was manually controlled, exact linearity of rise was difficult to attain. The ramp slope
FIG, 3. Percentage fall of systemic vascular resistance (SVR) per unit change of left atria1 pressure (Pla) vs. prestimulus Pla during staircase forcing of Pla. Data from 11 dogs have been grouped, and the group averages (IEJZ,) have been graphed.
6
fml/min)
d-2000 -0 Pea
(cm
H20)
-A00 -0 PI0
km
H20) f-7
ii I FIG.
corded fashion.
L-20 -0
4, Systemic arterial inflow (0) - and aortic pressure (Pao) rewhile forcing left atria1 pressure (Pla) to rise in a staircase Dog 789.
used for graphic analysis was that which was present over most of the ramp, or, when appropriate, was that slope present over the portion of the ramp from the first significant pressure rise to the point at which the SVR minimum occurred. Although subject to quantitative errors, the data are adequate to show that the SVR response is influenced by the rate of pressure rise. In all 13 experiments there was a direct relationship between rate of pressure rise and the vast ular response. Again because of the large variation in absolute response magnitudes, and in number of data points obtained in each experiment, presentation of all data is undesirable here. We have illustrated the effect of pressure rise rate by giving all the data from the six experiments that had the greatest range of slopes (Fig. 5). Figure 6 shows a typical experiment. In 6 of the 10 experiments performed to see the effect of interpulse duration on response, there was a significant sustained fall of resistance (from 5 to 15 %) while Pla was elevated, whereas in 4, SVR returned to control level after the initial transient fall. When Pla was restored to base
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
474
T. 70
Pla
(cm
c.
LLOYD,
JR.
H20)
0
60 0
*1* I
50
0
________/
SW
0 0
(arbitrary
units)
-4
1
L -0
/
i40
0 /YOY(/
--
0
0 0
0
z 30 s 0
20
0
FIG.
0
repetitive pressure between
0
0 / 0 f4
IO
0
/
0
5
I 0
0
IO
20 rise
Pla
40 rate,
a0 cm
FIG. 5. Percentage fall of systemic vascular rate of rise of left atria! pressure (Ha). Graphs ments are shown.
Pla
(cm
160 H20/min
320
resistance (SVR) vs. of 6 separate experi-
l-l201 -40
A SVR
(arbitrary
-2 L
4
1 min
6. Systemic vascular ramp increases of left atria1 periment. Dog 796. FIG.
20
-0
resistance (SVR) recorded during pressure (Pla) at 4 rates in a single ex-
30
40 interval,
FIG. 7. Percentage fall of systemic interval allowed between square-wave Data from 5 dogs are shown.
50 set
vascular stimuli
60
70
resistance of fixed
80
(SVR) vs. magnitude.
in those six experiments with a sustained resistance depression, resistance returned rapidly to the control level during the first 4-6 s of the interpulse interval. In any event, the resistance upon which each percentage fall was based
line
was that resistance present immediately prior to restoration of the stimulus Pla. In all 10 experiments we found that responses became larger as the recovery interval was increased from 6 to 100 s. Recovery intervals of 40-80 s were required before responses became as large as the one seen when the atria1 pressure was first increased. The recovery graphs from five experiments, chosen on the basis of having a wide range of response magnitudes, are illustrated in Fig. 7, and data recorded during a single experiment are shown in Fig. 8. DISCUSSION
-0
units)
IO
8. Systemic vascular resistance (SVR) recorded during application of a fixed square-pulse increase of left atria1 (Pla) when different recovery intervals have been allowed stimuli. Dog 800.
Our previous work had shown that simultaneous pressurization of the left atrium and the entire pulmonary vascular system caused a large transient, and smaI1 steadystate, systemic vasodilation ( 16, 17). The present study showed that responses similar to those obtained in the combined system attended pressurization of either the pulmonary arterial or vein-atria1 compartments independently. The pulmonary arterial and vein-atria1 baroreflex responses differ most in their sensitivity: that of the atrium and veins being three- to fourfold greater than that of the pulmonary arteries. In comparison with an earlier study (17), the sensitivity of the vein-atria1 component acting alone and unaffected by systemic baroreflexes is not significantly different from that of the combined system acting in the presence of systemic baroreflexes. Because of its greater sensitivity, we have focused attention on the vein-atria1 baroreflex in the present study to show that those responses depend upon preexisting pressure, rate of stimulus pressure rise, and the interval between stimuli+ Since not all observers agree that atria1 distention causes systemic vasodilation, it is of interest to compare reports. Doutheil and Kramer (7) showed a reflex fall of SVR related to magnitude of rise of left atria1 pressure. Although extensive data were not presented, it appears that responses reached a maximum in 20-30 s, and that the interval of stimulation was also of this duration. Edis et al. (8) obtained a fall of SVR with either of two techniques: by simultaneously inflating three balloons placed in pulmonary vein-atria1 junctions, or by distending a pouch of the left atrium. Evoked responses tended to be well maintained during the course of a 2-min period of distention, although there was apparently a transient minimum achieved initially. With both methods, response magnitude
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
CARDIOPULMONARY
BAROREFLEXES
was determined by amount of distention. Carswell et al. (2) using smaller balloons than Edis et al. also distended three of the a triovenous junctions. Apparen tlY these were not d istended simultaneously, thus creating wh at may be an i mpor tant di flerence in rate and magnitude of stimulation in comparison with other studies. Transient falls of SVR lasting approximately 22 s occurred in one-half of their experiments. However, those workers concentrated their interest on results obtained after 2 min of distention and discarded the importance of the transient fall. At 2 min the SVR was not significantly different from control. Furnival et al. (10) did not find that SVR fell when small balloons were inflated in the vein-atria1 junctions (in the manner of Carswell et al.), but they too concentrated on late efiects. Linden (15) has used the data of Carswell et al. and Furnival and co-workers along with his own to conclude that atria1 distention does not significantly influence SVR. In comparison, I consistently find that atrial pressurization causes a transient fall of SVR that reaches its nadir between 20 and 30 s, and which is thereafter variably, although poorly, maintained for stimulus periods as long as 5 min. Thus while there has been some inconsistency in results, there has been an even greater inconsistency in methods of stimulation and time of observation. It seems that chances for seeing reflex vasodilation are greatest when larger portions of the veins and atrium are distended quickly and when one looks for early effects. In my experiments, and in those of Doutheil and Kramer (7), a further complication may be caused by the inclusion of stretch of ventricular receptors at or near the mitral valve ring, since evidence suggests that ventricular stretch may induce systemic hypotension (7, 20). If systemic hy ‘po tension is a result of atria1 distention, what receptors may be responsible? There are several similarities between the SVR changes we found and the electrical activity of myelinated atria1 aRerent fibers. Thus both firing rate and SVR response are proportional to extent of distention (12, 24, 25), electrical activity and SVR response both depend upon prestimulus pressure (25, 27), and the firing rate and the change of SVR are both related to the logarithm of the rate of distention (11). may Although the observed patterns of reflex vasodilation seem to have equivalen ts in activity of type-B receptors, it may be incorrect to speculate that they were responsible for the changes we induced. Oberg and Thor& (21) reported that electrical stimulation of cat myelinatcd cardiac afferent fibers caused a small vasoconstriction and cardiac acceleration (similar to the responses descri bed by Furnival and co-workers (10)). On the other hand, stimulation of nonmyelinated afferents caused prominent hypotension and bradycardia (similar to the results of Edis et al., Doutheil and Kramer, and myself). 6berg and Thor& ascribed only a minor role to myelinated afferents in the cardiac nerve, but in contrast, held that rcpowcrful circulatory responses can be elicited from cardiac receptors firing in nonmedulated aflerents.” They later went on to show that such a rcflcx can attend brief (
Cardiopulmonary ramp,
baroreflexes:
and square-wave THOMAS Defartment
C. LLOYD, JR. of Physiology, Indiana
stimulation
University Medical
LLOYD, rr~~~~~ C., JR. Curdiopulmonary horeflexes: effects of staircase, ramf, and square-wave stimulation. Am. J. Physiol. 228(2) : 470-476. 1975.--Anesthetized dogs were pump perfused in a system that held systemic arterial pressure constant. Pulmonary arterial beds were pressurized, and the induced transient falls of systemic vascular resistance (SVR) were measured. An approximately linear relationship between percent fall of SVR and pulmonary artery pressure was obtained. Left atria and pulmonary venous beds were pressurized using square-pulse, staircase, and ramp wave forms, and changes in SVR were measured. The effect of varying the interval between stimuli was appraised. An approximately linear relationship was obtained for responses to square-wave left atria1 forcing pressure (Ha). With staircase forcing, the change of SVR per unit change of Ha depended upon prestimulus Pla. With ramp forcing, responses increased approximately linearly with respect to the logarithm of the rate of pressure rise. Responses to square-wave stimulation varied directly with respect to the interstimulus interval. This study sho\vs that left atrial-pulmonary vein baroreflexes are capable of producing substantial short-term falls of SVR, and that the response depends upon stimulus pressure, rate of pressure change, base-line pressure, and pressure history. left blood
atrium; pressure
pulmonary regulation
vessels;
systemic
vascular
effects of staircase,
resistance;
I T HAS BEEN SHOWN that stimulation of stretch receptors in the pulmonary artery, or left atrium and pulmonary veins may aflcct systemic vascular resistance. Distention of one or another of these chambers typically produces systemic vasodilation (7, 8, 16, 17, 23, 26, 28), although not always (2, 4, 10, 13-15). While the role of these cardiopulmonary barorcflexes in homeostasis remains uncertain, there is evidence that they provide a continuous sympathoinhibitory influence (19, 22). Aff ercnt nerve discharge patterns imply the existence of several types of cardiopulmonary receptors which respond to strain, or rate of strain, or both, and which may show different rates of adaptation (26-28). Some of the apparent differences among receptors may relate to their mode of attachment to the walls of the chamber in question (1, 27). Left atria1 type-B receptors are slowly adapting and their discharge frequency parallels the atria1 v-wave (25). The role and behavior of type-A receptors are less uniformly agreed upon, but they may be rapidly adapting and sensitive to rate of change of left atria1 wall tension (26). Pulmonary arterial baroreceptors resemble the systemic arterial baroreceptors in having responses that are prominently influenced by pulse pressures
Center,Indianapolis,
Xndiana 46202
(4). More recent studies have established the existence, if not always the role, of other receptors which utilize nonmyelinated afierent fibers (3, 20, 2 1). The neurophysiologic data have not been completely accompanied by equivalent hemodynamic data which would display such phenomena as rate sensitivity and adaptation. Nor have the systemic vascular elects of reflexes from the pulmonary arterial bed been contrasted in the same preparation with those from the pulmonary veins and left atrium. The present study was designed to further characterize the cardiopulmonary baroreflexes by answering the following questions: 1) how do the responses to square-wave forcing of pulmonary arterial pressure compare with responses while similarly forcing the pressure within the pulmonary veins and left atrium; 2) is there an influence of prestimulus left atria1 pressure upon the change in systemic vascular resistance produced per unit change in left atria1 pressure; 3) does the rate of rise of pressure influence the hemodynamic response to left atria1 distention; 4) for a given left atria1 square-wave pressure stimulus, is there an effect of the interstimulus interval duration upon response? This investigation was carried out with a canine preparation using cardiac bypass perfusion under the condition of constant systemic arterial pressure. Our procedure incorporated a method that provided for separate distention of the pulmonary arterial bed and of the pulmonary venous bed and left atrium. METHOD
Sixteen mongrel dogs, weighing from 13 to 20 kg, (average, 16 kg) were used for this study+ After anesthesia with 30 mg/kg iv pentobarbital, and 1 mg/kg im morphine sulfate, the dog was placed supine and pump ventilated with room air. The chest was opened through a longitudinal sternotomy. The left subclavian artery and left lower-lobe pulmonary artery were dissected free, and the was pericardium was opened. The right atria1 appendage cannulated with an 8-mm ID stainless steel cannula for the return of blood to a pump oxygenator. The subclavian artery was divided and the central end was cannulated for perfusion of the systemic circulation. The left lower-lobe artery was divided and a centrally directed cannula was installed to provide an input port for pressure forcing. A z-mm OD catheter was advanced centrally through the right internal thoracic artery for measurement of systemic arterial pressure. Heparin (100 mg) was given to prevent clotting. Ventricular fibrillation was then induced with an
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
CARDIOPULMONARY
K4ROREFLEXES
electric shock; lung ventilation was arrested, and external perfusion was begun. The lungs passively collapsed through a water seal to an end-expiratory pressure of 3-5 cmHzO. A left ventricular drainage cannula was placed through the apex and a left atria1 cannula was placed through its appendage. Both ventricles were clamped with a single large clamp (made from a 10-inch Rochester-Ochsner forceps) placed across the ventricles as close to the atrioventricular groove as possible. This clamp isolated the ventricles from the pulmonary artery and aorta as well as from the atria. The atria continued to beat vigorously during the remainder of the experiment. Catheters for acquisition of pressures were advanced into the left atrium and main pulmonary artery through the walls of the tubing attached to their cannulas. The tubing to these cannulas provided either for drainage from, or imposition of pressure on, the respective sides of the pulmonary circulation. To effect independent distention of the pulmonary arterial or pulmonary venous-left atria1 components, we inflated the segment with gas using 95 7~ OZ-5 % CO, with the wave form and pressures desired. Surface tension prevented capillary gas passage at the pressures used (18). While one compartment was pressurized, the other was allowed to drain freely through its cannula and tubing. When the left atrium was pressurized, diversion of coronary and bronchial venous blood into the pulmonary arteries caused pulmonary arterial pressure to rise transiently. Such increases were of less than 3 cmHz0 magnitude, so pressure remained nominally at zero. Similar small changes in atria1 pressure occurred during arterial pressurization, but on a few occasions large increases of pulmonary arterial pressure caused atria1 pressure to rise above 5 cmHz0. In the ultimate analysis, the marked sensitivity of the atria1 reflex led us to discard responses to pulmonary arterial pressurization if atria1 pressure rose above 5 cmHz0. The external perfusion system consisted of a bubble oxygenator (constructed locally) and two Manostat rotary tubing pumps which were connected in parallel. The speed of one pump was manually variable; the other pump was controlled electronically by a voltage proportional to the difference between a chosen set point and systemic arterial controlled pump adequately pressure. The electronically kept pressure constant during induced falls in systemic vascular resistance as large as 30 %. Manual control of the second pump during evoked responses permitted pressure to be held constant during larger resistance falls. Since both pumps ran simultaneously, the electronic controller provided the necessary fine adjustment when manually controlling the other pump, thereby freeing the operator from the need for critical performance. The oxygenator consisted of a vertical glass column of 50 cm length and 2 cm ID, into which blood flowed by gravity. Oxygen was bubbled into the base of this column through a l-cm OD Plexiglas tube that contained 4 rows of 24 holes. Hole diameter was 0.75 mm and the rows were spaced 5 mm apart. The upper one-third of the glass column was surrounded by a 10.5 x 2 7.5 cm cylindrical Plexiglas defpaming chamber that was half-filled with polypropylene gauze mesh. Blood from the defoaming chamber flowed by gravity into a Plexiglas reservoir from
471 which it was pumped. The priming volume of this system was about 500 ml, but we intentionally filled it with 1 liter to provide sufficient fluid for the large transient flow increases produced reflexly. The priming fluid was 6 Y& dextran in 0.9% NaCl solution (Abbott Dextran 75), to which was added additional amounts of a physiological salt solution (17) if necessary during the experiment. Added volumes ranged from 300 to 1,000 ml. Oxygen flow rate was never measured, but was set to provide a pH of reservoir blood between 7.3 and 7.5 U. Arterial blood Pea always exceeded 150 torr whenever measured by intermittent sampling, and it was typically above 300 torr. Reservoir blood temperature was maintained between 34 and 38°C by heating the dog and the oxygenator assembly with infrared lamps. Except where noted, all cannulas were glass, and the interconnecting tubing was Silastic or gum rubber. After use, all components were soaked overnight in Alconox or Haemosol solution. With the exception of the Plexiglas chambers, everything was then boiled for 2 h in 2 % NaHC03 solution, then rinsed and air-dried. We continuously recorded systemic arterial pressure, pulmonary arterial pressure (Ppa), and left atria1 pressure (Pla), using the mid-chest plane for zero reference. Systemic arterial inflow was monitored using an electromagnetic Aowmeter. Systemic vascular resistance (SVR) was defined as systemic arterial pressure divided by systemic inflow. In many experiments SVR was obtained electronically and continuously recorded (16). This was not done in all experiments because of technical problems. Systemic blood flow was initially set near 100 ml/kg per min by choice of the arterial pressure set point. If base-line resistance changed much during the experiment, set-point adjustments were made to maintain the flow near the original level. The various protocols were performed in random order. We tried to accomplish several in each experiment, but it was often impossible to perform all. We discontinued an experiment if responses to a given stimulus began to diminish; if large spontaneous variations of SVR appeared; or if the reservoir fluid volume began to decline rapidly, indicating either intravascular pooling or substantial transudative loss. The five subroutines were as follows: 1) In 13 preparations gas was admitted to the left atrium to cause a square-wave pressure change which lasted about 70 s. Between 3 and 13 (typically 6) such square pulses were used in each experiment. These ranged from 6 to 43 cmHz0 in magnitude, and occurred at a rate of 1 every 3-4 min. These were not true square-wave forcing pressures, in that the rise time was finite and the corners were somewhat rounded. Rise time was made as short as possible and dP/dt always exceeded 300 cmHaO/min. Between pulses, Pla was returned to an average base-line level of 5 cmHZO. This was constant for any given experiment, but varied among the groups from 0 to 12 cmHz0 because of variations in amount of drainage and configuration of the drainage tubing. 2) The pulmonary artery was similarly distended in but the range of pressure was seven preparations,
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
472 15-90 cmHxO. Base-line pressures were between 0 and -5 cmHz0. 3) In 11 dogs left atria1 pressure was increased from 0 to 40-45 cmHz0 in a staircase fashion (see Fig+ 4). Each step was of as uniform a size as possible in each run, but within the group they ranged between 6 and 10 cmHQ0. Pressure was restored abruptly to the prestimulus base line after obtaining the response at the highest pressure. These base-line pressures ranged from 0 to 6 cmHz0. Two such runs were made in four dogs, three runs were made in five dogs, and single runs were made in two dogs. to the left atrium to 4) In 13 dogs gas was admitted generate a ramp increase of pressure from a baseline Pla of 5 cmHz0 (range O-l 2). The rise was controlled manually, guided by the pressure record. Ramp slopes ranged from 7 to 400 cmHzO/min. From three to seven rates (typically five) were observed in each dog. Pressure was returned abruptly to base-line level upon reaching 40 cmHz0. In 10 dogs Pla was raised from base line to a pressure between 22 and 32 cmHz0 (average, 28), held steady for about 60 s, and then abruptly returned to base line for a ran,domly variable period between 5 and 100 s. At the end of that interval it was abruptly raised to the previous distending level for another -60 s. This sequence was continued until a number of interpulse intervals had been observed (from five to nine per animal). Results were analyzed in the following ways. As before (16, 17), we found that responses consisted of transient SVR falls which nearly completely resolved spontaneously in spite of maintained stimulus pressure. The steadystate changes were small and variable. We elected to consider only the transient maximum fall since this was easily quantifiable, was invariably present, and had been shown to behave in an unequivocally graded fashion in earlier experiments (I 7). The percentage fall of SVR from prestimulus level to the nadir during the response was calculated for each imposition of a forcing pressure. Results pertinent to the individual experiment subroutines were 1) Data obtained during single expressed differently. square-pulse forcing were reduced to graphs of percentage SVR fall versus the Pla or Ppa which existed during the pulse. 2) During staircase forcing, each pressure level was held until SVR had stabilized. Since the purpose of the was to display the influence of prestaircase sequence stimulus pressure on the subsequent response, the transient fall induced by each step was expressed as the percentage SVR fall per centimeter Hz0 step change of Pla, and each response so expressed was plotted as a function of the Pla that was present immediately preceding the step that produced it. 3) Ramp forcing caused responses which also had transient resistance minima. Each transient minimum was expressed as the percentage fall and these were then graphed as a function of the Pla ramp slope that caused ihem. $) The results of varying the interpulse recovery interval were displayed by plotting the percentage SVR fall observed transiently upon restoration of stimulus pressure against the interpulse time.
T.
C. LLOYD,
JR.
RESULTS
The general form of responses seen upon square-wave forcing of Pla resembled those reported earlier (16, 17) The pressure control system generally he1.d systemic arterial pressure within 5 cmHz0. However, in the largest responses (50 % SVR fall or greater) of the largest dogs, the required increases in flow exceeded pumping capacity, and on those few occasions pressure briefly fell as much as 20 cmHz0. At such times systemic flow approached 4 liters/min. In each of the 13 experiments in which we sought the relationship between the magnitude of a square-wave pulse of Pla and the resulting fall of SVR, responses were related to input pressure magnitudes. The data have been combined in Fig. 1. Not apparent in the figure is the tendency for the response curve of most individual experiments to level off beyond atria1 pressures of about 35 cmHz0. The effect of this behavior on the figure is to cause a wide variation among responses above that The stimuli and responses were significantly pressure. correlated. The linear regression equation for all data is = 1.55X 3.56 (r = 0.79, 12 = 87). The x-intercept Y is at 2.3 cmHz0. Since a linear relation was most apparent only for Pla less than 35 cmHz0, a regression analysis of that data subset was done, yielding the equation y = 2.09x - 13.7 (r = 0.81, 12 = 68). In this case the x-intercept pressure is 6.4 cmHg0. Square-wave forcing of Ppa caused changes of SVR resembling those seen when forcing Pla, i.e., a transient minimum which returned to control resistance within 3060 s. The fractional change of resistance at the nadir was related to stimulus pressure magnitude. The data from all seven exneriments are displayed in Fig. 2. The linear l
0
0
0 0
0
0
o”
8, 0 000:
i
o*
8
0 0
0
0
00
0
0
0
0
0
0
8
0 0
0 0
0
0
0
0”:
0
Q 0
O0 0 0 00
1, Percentage fall of systemic vascular resistance (SVR) vs. atria1 pressure (Pla) during square-pulse forcing of Pla in 13 dogs.
FIG.
left
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
CARDIOPULMONARY
473
BAROREFLEXES
60 0 0
50 0 0
Ir40 > CT) -- 3o u w
0 0 0 eQ 0
o"
0
0 0
o
0
0
320
a
IO
0
O0
0
0
0
a
0 0
0 OO
I
IO
I
20
0
IOQ
30
pm 2. Percentage puhnonary arterial Ppa in 7 dogs. FIG.
I
1
1
I
I
40
50
60
70
80
cm
I
90
Hz0
fall of systemic vascular resistance pressure (Ppa) during square-pulse
(SVR) forcing
vs. of
regression equation for these is y = 0.5~ - 2.8 (r = 0.71, n = 39), and th c x-intercept is 5.2 cmHz0. These responses differed from those obtained by forcing Pla in that they were smaller for any given stimulus and there was no consistent tendency for responses to level off within the range of Ppa used. Increase of Pla in a staircase fashion caused transient SVR falls with each step. Typically, SVR returned to the control level after each transient fall, but occasionally there was a trend to a progressively lower plateau SVR following each step. These trends were manifestations of the small and inconsistent steady-state changes seen in these experiments and described previously (17). However, we did not assume a fixed base-line SVR for evaluation of the response, but based the fractional resistance change at each response on the resistance present immediately before each step. In all of the 11 experiments the fractional fall of SVR per centimeter Hz0 rise of Pla was related to the Pla from which each step rose. Responses were small at a low prestep Pla, rose to a maximum near Pla of 15 cmHz0, and then fell as Pla was further raised. The large variation among experiments militated against Instead presenting our findings with a graph of all data. we have grouped the data into subunits of G.5 cmHz0 prestep Pla and determined the average response obtained in each unit. (Because of the small group size, all results obtained between Pla of 35 and 45 cmHz0 have been collected into a single unit.) The unit averages are displayed in Fig. 3. Each individual experiment had a graph similar in shape to that of the figure. Recorded data from one experiment are given in Fig. 4. When using ramp wave forms to distend the left atrium, we found that transient SVR minima were again produced. The SVR fall tended to develop more slowly at the lowest pressure rise rates, and at these rates the resistance often returned toward control level while Pla was still increasing. On the other hand, at the highest rise rates, the greatest SVR fall often did not appear until after Pla was returned to base line. Since Pla was manually controlled, exact linearity of rise was difficult to attain. The ramp slope
FIG, 3. Percentage fall of systemic vascular resistance (SVR) per unit change of left atria1 pressure (Pla) vs. prestimulus Pla during staircase forcing of Pla. Data from 11 dogs have been grouped, and the group averages (IEJZ,) have been graphed.
6
fml/min)
d-2000 -0 Pea
(cm
H20)
-A00 -0 PI0
km
H20) f-7
ii I FIG.
corded fashion.
L-20 -0
4, Systemic arterial inflow (0) - and aortic pressure (Pao) rewhile forcing left atria1 pressure (Pla) to rise in a staircase Dog 789.
used for graphic analysis was that which was present over most of the ramp, or, when appropriate, was that slope present over the portion of the ramp from the first significant pressure rise to the point at which the SVR minimum occurred. Although subject to quantitative errors, the data are adequate to show that the SVR response is influenced by the rate of pressure rise. In all 13 experiments there was a direct relationship between rate of pressure rise and the vast ular response. Again because of the large variation in absolute response magnitudes, and in number of data points obtained in each experiment, presentation of all data is undesirable here. We have illustrated the effect of pressure rise rate by giving all the data from the six experiments that had the greatest range of slopes (Fig. 5). Figure 6 shows a typical experiment. In 6 of the 10 experiments performed to see the effect of interpulse duration on response, there was a significant sustained fall of resistance (from 5 to 15 %) while Pla was elevated, whereas in 4, SVR returned to control level after the initial transient fall. When Pla was restored to base
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
474
T. 70
Pla
(cm
c.
LLOYD,
JR.
H20)
0
60 0
*1* I
50
0
________/
SW
0 0
(arbitrary
units)
-4
1
L -0
/
i40
0 /YOY(/
--
0
0 0
0
z 30 s 0
20
0
FIG.
0
repetitive pressure between
0
0 / 0 f4
IO
0
/
0
5
I 0
0
IO
20 rise
Pla
40 rate,
a0 cm
FIG. 5. Percentage fall of systemic vascular rate of rise of left atria! pressure (Ha). Graphs ments are shown.
Pla
(cm
160 H20/min
320
resistance (SVR) vs. of 6 separate experi-
l-l201 -40
A SVR
(arbitrary
-2 L
4
1 min
6. Systemic vascular ramp increases of left atria1 periment. Dog 796. FIG.
20
-0
resistance (SVR) recorded during pressure (Pla) at 4 rates in a single ex-
30
40 interval,
FIG. 7. Percentage fall of systemic interval allowed between square-wave Data from 5 dogs are shown.
50 set
vascular stimuli
60
70
resistance of fixed
80
(SVR) vs. magnitude.
in those six experiments with a sustained resistance depression, resistance returned rapidly to the control level during the first 4-6 s of the interpulse interval. In any event, the resistance upon which each percentage fall was based
line
was that resistance present immediately prior to restoration of the stimulus Pla. In all 10 experiments we found that responses became larger as the recovery interval was increased from 6 to 100 s. Recovery intervals of 40-80 s were required before responses became as large as the one seen when the atria1 pressure was first increased. The recovery graphs from five experiments, chosen on the basis of having a wide range of response magnitudes, are illustrated in Fig. 7, and data recorded during a single experiment are shown in Fig. 8. DISCUSSION
-0
units)
IO
8. Systemic vascular resistance (SVR) recorded during application of a fixed square-pulse increase of left atria1 (Pla) when different recovery intervals have been allowed stimuli. Dog 800.
Our previous work had shown that simultaneous pressurization of the left atrium and the entire pulmonary vascular system caused a large transient, and smaI1 steadystate, systemic vasodilation ( 16, 17). The present study showed that responses similar to those obtained in the combined system attended pressurization of either the pulmonary arterial or vein-atria1 compartments independently. The pulmonary arterial and vein-atria1 baroreflex responses differ most in their sensitivity: that of the atrium and veins being three- to fourfold greater than that of the pulmonary arteries. In comparison with an earlier study (17), the sensitivity of the vein-atria1 component acting alone and unaffected by systemic baroreflexes is not significantly different from that of the combined system acting in the presence of systemic baroreflexes. Because of its greater sensitivity, we have focused attention on the vein-atria1 baroreflex in the present study to show that those responses depend upon preexisting pressure, rate of stimulus pressure rise, and the interval between stimuli+ Since not all observers agree that atria1 distention causes systemic vasodilation, it is of interest to compare reports. Doutheil and Kramer (7) showed a reflex fall of SVR related to magnitude of rise of left atria1 pressure. Although extensive data were not presented, it appears that responses reached a maximum in 20-30 s, and that the interval of stimulation was also of this duration. Edis et al. (8) obtained a fall of SVR with either of two techniques: by simultaneously inflating three balloons placed in pulmonary vein-atria1 junctions, or by distending a pouch of the left atrium. Evoked responses tended to be well maintained during the course of a 2-min period of distention, although there was apparently a transient minimum achieved initially. With both methods, response magnitude
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 13, 2019.
CARDIOPULMONARY
BAROREFLEXES
was determined by amount of distention. Carswell et al. (2) using smaller balloons than Edis et al. also distended three of the a triovenous junctions. Apparen tlY these were not d istended simultaneously, thus creating wh at may be an i mpor tant di flerence in rate and magnitude of stimulation in comparison with other studies. Transient falls of SVR lasting approximately 22 s occurred in one-half of their experiments. However, those workers concentrated their interest on results obtained after 2 min of distention and discarded the importance of the transient fall. At 2 min the SVR was not significantly different from control. Furnival et al. (10) did not find that SVR fell when small balloons were inflated in the vein-atria1 junctions (in the manner of Carswell et al.), but they too concentrated on late efiects. Linden (15) has used the data of Carswell et al. and Furnival and co-workers along with his own to conclude that atria1 distention does not significantly influence SVR. In comparison, I consistently find that atrial pressurization causes a transient fall of SVR that reaches its nadir between 20 and 30 s, and which is thereafter variably, although poorly, maintained for stimulus periods as long as 5 min. Thus while there has been some inconsistency in results, there has been an even greater inconsistency in methods of stimulation and time of observation. It seems that chances for seeing reflex vasodilation are greatest when larger portions of the veins and atrium are distended quickly and when one looks for early effects. In my experiments, and in those of Doutheil and Kramer (7), a further complication may be caused by the inclusion of stretch of ventricular receptors at or near the mitral valve ring, since evidence suggests that ventricular stretch may induce systemic hypotension (7, 20). If systemic hy ‘po tension is a result of atria1 distention, what receptors may be responsible? There are several similarities between the SVR changes we found and the electrical activity of myelinated atria1 aRerent fibers. Thus both firing rate and SVR response are proportional to extent of distention (12, 24, 25), electrical activity and SVR response both depend upon prestimulus pressure (25, 27), and the firing rate and the change of SVR are both related to the logarithm of the rate of distention (11). may Although the observed patterns of reflex vasodilation seem to have equivalen ts in activity of type-B receptors, it may be incorrect to speculate that they were responsible for the changes we induced. Oberg and Thor& (21) reported that electrical stimulation of cat myelinatcd cardiac afferent fibers caused a small vasoconstriction and cardiac acceleration (similar to the responses descri bed by Furnival and co-workers (10)). On the other hand, stimulation of nonmyelinated afferents caused prominent hypotension and bradycardia (similar to the results of Edis et al., Doutheil and Kramer, and myself). 6berg and Thor& ascribed only a minor role to myelinated afferents in the cardiac nerve, but in contrast, held that rcpowcrful circulatory responses can be elicited from cardiac receptors firing in nonmedulated aflerents.” They later went on to show that such a rcflcx can attend brief (
