Circulatory and renal effects of β-adrenergic-receptor stimulation in pregnant sheep

Circulatory and renal effects of ,a-adrenergic-receptor stimulation in pregnant sheep G. Kleinman, B. Nuwayhid, R. Rudelstorfer,* A. Khoury, K. Tabsh, S. Murad, C. R. Brinkman III, and N. S. Assali Los Angeles, California The effects of {3·adrenergic-receptor stimulation with ritodrine on systemic and pulmonary hemodynamics and on renal handling of water and electrolytes were studied in unanesthetized, chronically instrumented pregnant sheep. Each animal was studied during control, ritodrine, and recovery periods, each lasting 60 minutes, with the use of three different modes of hydration. {3-Receptor stimulation produced a significant increase in heart rate and cardiac output arid a decrease in systemic vascular resistance. Pulmonary arterial and wedge pressures tended to increase. These circulatory effects were similar for the three types of hydration and they persisted after cessation of infusion. In terms of its renal effects, {3-receptor stimulation elicited a profound decrease in urine flow and in the excretions of sodium and potassium, irrespective of the mode of hydration. The antidiuresis and antinatriuresis were accompanied by no changes in plasma osmolality and sodium concentration, whereas plasma potassium levels decreased. All of these effects persisted for 60 minutes after the cessation of infusion. In the water-loaded experiments, the antidiuresis seemed to be related to increased antidiuretic hormone secretion; in the saline-loaded experiments, however, both the antidiuresis and antinatriuresis appeared to be related to increased renal reabsorption. The changes in renal hemodynamics seemed to have an insignificant role. The amount of fluid retained in the body was greater when ritodrine was infused with saline solution than with dextrose solution. These cardiovascular and renal studies suggest that a circulatory overload may be the major factor in the pathogenesis of pulmonary edema observed during {3~adrenergic-receptor stimulation: (AM. J. 0BSTET. GYNECOL. 149:865, 1984.)

/3-Adrenergic agonists have been widely used in obstetrics to arrest premature labor through their specific effects on the pregnant uterus. Agonists such as ritodrine, fenoterol, and terbutaline have been preferred because of their more selective action on the /3 2 -adrenergic receptors located in the myometrial effector system. Despite their claimed selectivity, however, these agents exert significant circulatory, renal, and metabolic effects depending on the dose and duration of treatment. There have been several reports of pregnant subjects who developed pulmonary edema while receiving /3agonists intravenously with isotonic saline or glucose solutions. 5 • 9 The incidence of this complication has been estimated to be about 5% of all patients treated with f3-agonists. 5 Although several contributing factors such as twin gestation, concomitant corticosteroid therapy, and fluid overload have been suggested, the preFrom the Department of Obstetrics and Gynecology, University of California (Los Angeles) School of Medicine. Supported by Grants HL-01755 and HL-13634 from the National Institutes of Health. Received for publication October 25, 1983; revised February 2, 1984; accepted March 9, 1984. Reprint requests: N. S. Assali, M.D., UCLA School of Medicine, Department of Obstetrics and Gynecology, Los Angeles, CA 90024. *Recipient of National institutes of Health Fogarty Grant and Fellow in Reproductive Physiology (from Austria).

cise mechanisms underlying the pulmonary edema associated with /3-adrenergic-receptor stimulation are, as yet, unclear. The present studies were designed to provide answers to the following questions: (1) What are the systemic and pulmonary hemodynamic effects of /3-adrenergic agonists when these are used by constant infusion in a manner similar to that used in human pregnancy? (2) What are the effects of /3-adrenergicreceptor stimulation on renal functions including water and electrolyte excretion and antidiuretic hormone disease? (3) What influence does the composition of the infusate used to administer the /3-agonist have on the circulatory and renal effects? (4) Could any of these factors be contributing to the pulmonary edema?

Material and methods Studies were carried out on l 0 chronically instrumented, unanesthetized pregnant ewes (100 to 135 days' gestation) weighing 40 to 50 kg. With the animals under local infiltration anesthesia (l% Xylocaine) supplemented with intramuscular diazepam, polyvinyl catheters were inserted into a femoral artery and vein and advanced into the descending aorta and the vena cava, respectively. These catheters served for monitoring arterial pressure and heart rate and for periodic arterial and venous blood samplings for de-

865

866

Kleinman et al.

terminations of respiratory gases and electrolytes. A jugular vein was exposed and two polyvinyl catheters were implanted; the smaller catheter served for infusion of the /3-mimetic as well as for dye injection to determine the cardiac output. The larger jugular catheter was used as a portal for introduction of a Swan-Ganz catheter into the pulmonary artery for monitoring of its pressure, as well as the pulmonary wedge pressure. The position of the pulmonary catheter was ascertained by attaching it to a pressure transducer and observing the wave form on a dynograph. Technical details of this type of chronic instrumentation have been published elsewhere. 4 For the renal studies, in addition to the above-described instrumentation, a Foley catheter was introduced into the bladder for accurately timed collection of urine flow. A recovery period lasting 4 to 5 days after operation was allowed before the animal was subjected to a study according to the following two experimental protocols. Cardiovascular experiments. This part of the study was designed to test the effects of intravenous infusion of ritodrine given with three different modes of hydration on: (1) systemic arterial, pulmonary arterial, and pulmonary wedge pressures, (2) heart rate and cardiac output, and (3) systemic and pulmonary vascular resistances. There were three different types of hydration. (1) Normal fluid maintenance tests consisted of giving the animal an intravenous infusion of normal saline solution at a rate of 125 ml/hr. The infusion started 1 hour before the testing period and continued for the duration of the experiments. Total fluid administered was approximately 500 mi. (2) Saline hydration tests consisted of giving the animal an initial priming load of I L of normal saline solution over 15 to 30 minutes, followed by a constant infusion of saline solution at a rate of 250 ml/hr for the duration of the experiment. (3) Dextrose hydration tests consisted of priming the animal with I L of 5% dextrose and water (over 15 to 30 minutes) followed by a constant infusion of the same solution at a rate of 250 ml/hr for the duration of the experiment. During each test, the animal stood quietly in its cage and was allowed 45 to 60 minutes to adjust to the testing laboratory. A control period lasting 1 hour was observed during which the animal received only the saline or dextrose infusion. A ritodrine treatment period lasting 1 hour then followed, during which ritodrine was added to the infusate at a dose of 5 p,g/kg/min, with the use of a constant-infusion pump. A recovery period lasting 1 hour was observed during which ritodrine was discontinued and the animal received only the dextrose or saline solution. During all these periods, systemic and pulmonary arterial pressures were recorded continuously; pulmonary wedge pressure and cardiac output were deter-

August 15, 1984 Am. J. Obstet. Gynecol.

mined at frequent intervals. Blood respiratory gases and hematocrit were assayed twice during each period. Mean pressures, heart rate, stroke volume, and cardiac output (indocyanine green dilution) were computed by methods previously described. 4 Systemic vascular resistance was calculated from the ratio of mean arterial pressure and cardiac output, and pulmonary vascular resistance was calculated from the ratio of mean pulmonary pressure minus wedge pressure and cardiac output. Renal studies. This series of experiments carried out on the same group of unanesthetized sheep was designed to test the effects of ritodrine and the type of hydration on: (1) renal plasma flow and glomerular filtration rate, (2) urine flow and osmolal and free water clearances, and (3) urinary excretion of sodium and potassium and their respective plasma levels. The same three different types of hydration and three experimental periods already described were carried out during each test. Renal clearances, including collections of urinary and plasma samples, were made every 30 minutes during all of the periods of study as described elsewhere. 6 Glomerular filtration rate was determined by the inulin clearance and the renal plasma flow was determined by the para-aminohippurate clearance. Free water and osmolal clearances were computed by the standard formula. 6 Excretion of sodium and potassium was computed from the product of urine flow, and the urinary concentrations and filtered load of these ions were calculated from the product of glomerular filtration rate and plasma concentration. The methodology for the determinations of all of these renal studies has been published previously. 6 Antid.iuretic hormone studies.* For studying the effects of {3-adrenergic-receptor stimulation on blood levels of antidiuretic hormone, a separate group of three animals was used. In each animal, physical and emotional stress was minimized by implantation of only one arterial and one venous catheter in the animal under local anesthesia and by allowing a long period (2 hours) of adaptation to the laboratory environments. Each animal received an infusion of 5% dextrose in water at a rate of 250 ml/hr throughout the three experimental periods as in the other studies; the same dosage of ritodrine in the other experiments was used. During the control ritodrine and recovery periods, plasma levels of arginine vasopressin were determined every 30 minutes. Arginine vasopressin levels were determined by a *We are indebted to Drs. D. Fisher and R. D. Leake for the arginine vasopressin radioimmunoassays which were carried out under United States Public Health Service Grant No. HD-06335 from the National Institute of Child Health and Human Development.

Volume 149 Number 8

radioimmunoassay method used by Weitzman et aU Sham experiments. Three animals were used to test the effects of the instrumentation and the saline hy· dration without ritodrine on the cardiovascular and renal functions. Each animal was studied during the same periods and under the same experimental conditions as in the "saline hydration series" except no ritodrine was added. Cardiovascular and renal parameters were monitored as already described. Statistical analyses. To minimize individual variations, each animal served as its own control for either the circulatory or the renal studies and was used for more than one test. Experimental protocols were alternated with an interval of 3 to 5 days between subsequent tests. The means of the various tests performed on the same animal were used to compute the response to ritodrine as well as to the type of hydration. Statistical evaluation was performed under the direction of Lee Youkeles (Department of Biomathematics), and the BMDP Biomedical Statistical software package was used. Differences between control period and /3-adrenergic-receptor stimulation periods were analyzed with program P3D, and a paired t test was used. Differences between hydration protocols were assessed with the P7D program with the use of an analysis of the vanance. Values of p < 0.05 were considered significant. Results

Effects of the experimental procedure on the animal. The surgical procedures and repeated tests were well tolerated by the animals. The average control values for arterial pressure, cardiac output, heart rate, arterial blood gases, and hematocrit during the control period were within the ranges of values previously reported for similar chronic instrumentation and were not significantly different from those of the sham operation animals. In this latter group, arterial pressure, cardiac output, urine flow, and osmolar and free water clearance were not significantly different during the three periods of study (Table 1). Circulatory effects CHANGES IN SYSTEMIC HEMODYNAMICS. The effects of ritodrine infusion with the three different types of hydration on the systemic arterial pressure, heart rate, stroke volume, cardiac output, and systemic vascular resistance are presented in Fig. 1. Mean arterial pressure decreased slightly during ritodrine infusion (not significant) and the changes were the same irrespective of the type of hydration. Heart rate increased by about 50% and stroke volume decreased by about 20% during /3-stimulation and the changes were similar in the three different types of hydration; heart rate remained elevated during most of the recovery period. Cardiac output progressively

Effects of .a-adrenergic-receptor stimulation

867

Table I. Data obtained from sham experiments (values represent mean ± l SE)

Parameter Mean arterial pressure (mm Hg) Cardiac output (L/min) Urine flow (mll min) Osmolal clearance (ml/min) Free water clearance

Recovery 100 ± 2

102 ± 3

104 ± 3

5.3 ± 0.4

5.4 ± 0.6

5.7 ± 0.9

6.7 ± 1.2

7.6 ± 1.4

6.7 ± 1.1

6.7 ± 1.9

6.0 ± 1.5

6.1 ± 1.4

0.1 ± 3.7

0.6 ± 1.8

0.4 ± 1.0

increased, reaching about 40% above control values during ritodrine infusion and remained elevated during the recovery period. Systemic vascular resistance decreased by about 70% during ritodrine infusion and remained below control during recovery; the changes were similar in the three types of hydration (Fig. 1). CHANGES IN PULMONARY HEMODYNAMICS. Fig. 2 presents the data on the effects of ritodrine on mean pulmonary arterial and pulmonary wedge pressures and on pulmonary vascular resistance. In contrast to the effects on the systemic circulation, /3-receptor stimulation with three different modes of hydration consistently increased the pulmonary artery and pulmonary wedge pressures even though the increment was of borderline significance (p > 0.05); pulmonary vascular resistance decreased by only 15% (p > 0.05) as compared to the systemic vascular resistance which decreased by 70% (Fig. 2). Here again, the type of hydration had no influence on the pulmonary hemodynamic changes produced by /3-adrenergic-receptor stimulation. Renal effects URINE FLOW. In Fig. 3 are shown the values for the urine flow in the three different modes of hydration and during the three experimental periods. The figures depict the values of 30-minute clearance periods. In the series of tests during which the animal received a "maintenance" amount of saline solution (see Material and Methods), control urine flow was relatively low (average, 1 ml/min); it decreased to an average of 0.5 mllmin during the first 30 ~inutes of the ritodrine infusion period and to 0.3 ml/min during the second collection period. Urine flow remained below control values during recovery (Fig. 3). The changes in the other renal parameters in this series of experiments were not significantly different from those of the "saline hydration" series; therefore, they will not be discussed further for the sake of brevity. In the series of hydration tests with isotonic saline

868

Kleinman et al.

August 15, 1984 Am. J. Obstet. Gynecol.

I RITOORINE I RECOVERY ~-----d---.=. .. ,.. -~------~------:.:;

CONTROL MEAN ARTERIAL PRESSURE (mmHol

120l 80

i

40

I

I

0

MEAN PULMONARY ARTERY PRESSURE (mmHg)

I

I HEART RATE (#/mini

150l 100

CONTROL

30

0.9% NoCI Hydration ... ---• 5% D/W Hydration

50

1 ± SEM

0 CARDIAC OUTPUT liter /min

7. ~

l

(mmHg/liter. min- 1 )

I I I

0

I

12

PULMONARY WEDGE PRESSURE (mmHg)

I

. . r---=--=---=-~-- ~P=====-

---...-41"1---~-

2.5

~~i~~;:~CE ~~l

1

l----~o:.. -... .... ~'

~-----f1·-·

-.,.___

~--~

--~-

~

8 ''----,--+--,------iCL---...----, 30

60

90

120

150

180

TIME (min)

PULMONARY VASCULAR RESISTANCE (mmHg/ liter· min- 1 )

I

I

I

--;-_:::.:i=:.-.=.-;~ ~-~~=~--=~

I I

I

I __..,_1-.....!.

--- _____]_ ___ T __ _

8

~~--~

4

l--=¥.:--r--Ti---r--1 I I

0

I I

I

I

I - - 0 . 9 % NoCI Maintenance I - - - 0.9 % NoCI Hydration I ... ---... 5 % D/W Hydration I I± SEM I

: -----

~r-------o.l,',,, : ___.-I-------1 .

0

10

12

sol 60

RITODRINE I RECOVERY

I

r-- I

0

· '

I 40

!---

.,..._ ___. 0.9% NoCI Maintenance

0

STROKE VOLUME (ml)

20

I

I

I

:

I

I

I

I I

I I I

:1

0

30

60

90

120

150

180

TIME(min)

Fig. 1. Mean arterial pressure, heart rate, stroke volume, cardiac output, and systemic vascular resistance during control, ritodrine, and recovery periods. Figures represent mean ± 1 SE and are for the three different types of hydration (n = 10 experiments).

Fig. 2. Values for mean pulmonary arterial pressure, pulmonary wedge pressure, and pulmonary vascular resistance during control, ritodrine, and recovery periods. Figures represent mean ± 1 SE and are for the three different types of hydration (n = eight experiments).

solution, control urine flow averaged about 4.5 ml/min; it decreased by an average of 50% during the first ritodrine period and by 70% during the second period (p < 0.01); urine flow remained significantly below control values for the 60 minutes of the recovery period (Fig. 3). Maximal water diuresis was obtained with the infusion of 5% dextrose in water during which control urine flow reached 8 to 10 ml/min. A profound antidiuresis occurred during ,8-receptor stimulation with the urine flow reaching an average of 3 ml/min during the first ritodrine period and 2 ml/min during the second period (p < 0.001); during the recovery period, urine flow remained far below control levels (Fig. 3). FREE WATER AND OSMOLAL CLEARANCES. In Fig. 4 are presented the data on the changes in urine flow as a function of its two components, the free water and osmolal clearances for the glucose and the saline hydration experiments. During the control period of the saline infusion series, the free water clearance (representing the nonobligated urinary fraction under antidiuretic hormone control) was negative, which indicates that in these experiments very little, if any, solute-free water was excreted in the urine. The free water clearance did

not change significantly during ritodrine infusion with saline solution and remained at the same levels during the recovery period (Fig. 4). The osmolal clearance, on the other hand, which represents the obligatory urine excreted with solute, was positive and very high during the control period; it decreased by about 50% during the first 30 minutes of ritodrine administration and by 75% during the second 30 minutes; it remained significantly below control values during the recovery period (Fig. 4). In the series of experiments with 5% dextrose hydration, control free water clearance and osmolal clearance were positive and very high and the urine was hypotonic. Stimulation of the ,8-adrenergic-receptor system produced a profound fall in both clearances and the urine became hypertonic; the osmolal clearance decreased by about 50% during the first ritodrine period and by 60% during the second period; it remained below control values after cessation of ,8stimulation. The free water clearance decreased by 90% during the first ritodrine period and became frankly negative during the second period, indicating that all of the free water had been reabsorbed; the free water clearance remained negative for 60 minutes after cessation of the infusion (Fig. 4).

Effects of ,8-adrenergic-receptor stimulation

Volume 149 Number 8

869

Ritodrine HC I 51'g/kg/min

I 10

I.....

T± ISE

:

---- 0.9% NaCI Maintenance - - 0.9% NaCI Hydration 5% D/W Hydration

· · . . .j

ml/min

i·.

I·.. I

5

30

\.

60

90

120

150

180

MINUTES

Fig. 3. Values for mean urine flow during control, ritodrine, and recovery periods for the three different types of hydration. Figures represent mean± 1 SE (n = 14 experiments).

Table II. Data on renal handling of sodium in the saline and dextrose hydration experiments

Period

Saline Control First ritodrine period Second ritodrine period Dextrose Control First ritodrine period Second ritodrine period

Filtered load (j.iliql min)

Excreted (j.iliqlmin)

Reabsorbed (j.iliqlmin)

Reabsorbed

13,140 ± 10 11,680 ± 12 12,410 ± 8

650 ± 15 220 ± 7 90 ± 6

12,490 ± 8 11,460 ± 9 12,320 ± 6

95.0 98.1 99.3

13,140 ± 10 11,680 ± 12 12,410 ± 8

150 ± 7 75 ± 6 40 ± 4

12,990 ± 8 11,605 ± 7 12,370 ± 6

98.8 99.4 99.8

RENAL PLASMA FLOW AND GLOMERULAR FILTRATION RATE. Fig. 5 shows the effects of ,8-receptor stimulation on renal plasma flow and glomerular filtration rate during the three experimental periods (the data represent the saline and glucose hydration tests combined because the difference between the two series was not significant). Renal plasma flow decreased significantly (p < 0.01) during ritodrine infusion and returned to control values during the recovery period; the fall in renal plasma flow occurred at a time when the cardiac output was rising (see Fig. 1). Calculation of the fraction of cardiac output that goes to the kidney (ratio of renal plasma flow to cardiac output) showed a decrease by more than 70% during ,8-receptor stimulation; the mechanisms of this decrease are not known. Glomerular filtration decreased by about 30% (p < 0.05) during the first period of ritodrine administration but recovered during the second period (the difference between control and the latter values was not significant); during the recovery period, glomerular filtration rate was close to control values. (Fig. 5). Be-

(%)

cause of the greater fall in renal plasma flow than in glomerular filtration rate, filtration fraction tended to increase during ,8-receptor stimulation (Fig. 5). EXCRETIONS OF SODIUM AND POTASSIUM. In Fig. 6 are presented the data on urinary sodium and potassium excretions for both the saline and glucose hydration experiments. In the control period of the saline hydration series, urinary sodium excretion was high, averaging about 620 ~-tEq/min. It decreased to about 220 ~-tEq/min during the first period of ritodrine infusion and to < 100 ~-tEq/min during the second period (p < 0.01); sodium excretion remained significantly below control values during the recovery period (Fig. 6). In the dextrose hydration series, control sodium excretion averaged 160 ~-tEq/min. It decreased by about 50% during the first period of ritodrine infusion and by about 80% during the second period; it remained below control values during recovery (Fig. 6). In both the saline and dextrose hydration experiments, the decrease in sodium excretion was totally related to a rise in the amount reabsorbed without any

870

Kleinman et al.

12

August 15, 1984 Am. J. Obstet. Gynecol.

.9 N NaCI Hydration

600

8 c E

....

c ~·E 400

a..:::::

4

UE

E

200 o+-~~r-~~~~~~~~--~r-­

-2

12

I

CJ

5% D/W Hydration I

(0.34} (0.24}

(0.18}

COSM

~ CH20

120

- - Urine Flow

c

I I

Filtration Fraction

8

c c·-

E

=E =>'-

.......

-=E

E

80 40

O+-~~L-~~~~M-~~~~r-~rr--­

-2

I

30

60:

t

90

120:

150

30

180

significant correlation with the amount filtered (Table II). ,8-Adrenergic-receptor stimulation with ritodrine also significantly decreased the excretion of potassium regardless of the type of hydration (p < 0.01); potassium excretion remained below control values during the recovery period (Fig. 6). The decrease in sodium excretion during saline infusion was accompanied by no changes in plasma sodium and osmolality; plasma potassium, however, decreased and continued to fall during the recovery period. With dextrose infusion, the plasma sodium concentration decreased by an average of 10 mEq/L and potassium also decreased. Blood pH, Po 2 , and Pco2 were not significantly affected by ,8-stimulation; hematocrit increased during ,8-stimulation, probably because of a contracting action on the spleen; plasma albumin did not change (Table III). FLUID RETENTION. In Fig. 7 are presented the data on the amount of fluid retained (intake- output) during ritodrine infusion with isotonic saline or dextrose solution. A significantly greater amount of fluid was retained in the body when ritodrine was infused with isotonic saline solution. Effects on arginine vasopressin levels. Fig. 8 shows the

90

120:

I

150

ISO min

t

Ritodrine HCI

Ritodrine HCI

molal (COSM) and free water (CH,P) clearances during control, ritodrine, and recovery periods, for the saline and dextrose hydration experiments (for more details, see Material and methods and Results) (n = 8 experiments).

I

t

t

Fig. 4. Values for urine flow superimposed on those for os-

60:

Fig. 5. Effects of ritodrine infusion on renal plasma flow (CPAH) and glomerular filtration rate (inulin clearance). Note

that ,8-stimulation decreased renal plasma flow more than glomerular filtration rate. Consequently, filtration fraction (in numbers) tended to increase. All renal hemodynamic parameters returned to control values during the recovery period (n = 10 experiments).

changes in plasma levels of arginine vasopressin produced by ,8-receptor stimulation during dextrose hydration. During the control period, arginine vasopressin values were stable at low levels, which indicates the efficiency of water loading with the dextrose infusion. ,8-Receptor stimulation with ritodrine progressively increased arginine vasopressin plasma levels with marked individual variations; plasma arginine vasopressin remained above control values during the recovery period (Fig. 8). Comment

The present data obtained from unanesthetized pregnant sheep indicate that activation of the ,8-adrenergic system by pharmacologic agents that are used to arrest premature labor elicit significant central hemodynamic, renal, and water and electrolyte effects which may be considered as contributing factors to the pathogenesis of pulmonary edema. Systemic and pulmonary hemodynamics. It is clear from this series of experiments that stimulation of the ,8-adrenergic receptors by ritodrine, which is presumed

Effects of /3-adrenergic-receptor stimulation

Volume 149 Number 8

1200

871

,-*P<0.05----,

c

.E 'CJ

.......

E

w

--

:::l..

:::l

> 200

800

0. :::l

0

z

0

::>

I Q) ~

.E 0.9%NaCI

•

I

c

5%0/W

"C

±SE

I.J_

400

:::l

300 c E

' t:r w

200

>

5% D/W Hydration

.9NNaCI

Hydration

:::1...

100

~

::>

30

60

I

t

90

1201

150

180

t

Ritodrine HCI

Fig. 6. Values for sodium (UNa· V) and potassium (UK· V) excretions during the control, ritodrine, and recovery periods for the saline (open bars) and dextrose (black bars) experiments. Note the profound antinatriuresis that followed /3-stimulation and persisted for the entire recovery period (n = 10 experiments).

Fig. 7. Amount of fluid retained by the animals for the saline and dextrose hydration experiments calculated as the difference between intake and output during the entire experiments (180 minutes). Note the marked difference in the degree of body fluid expansion during the experiments with saline infusion (n = 10 experiments). Table III. Average values for blood respiratory gases, hematocrit, and plasma albumin during the three periods of study Parameter

pH to be more selective for the myometrium, elicits significant cardiac effects as reflected by the marked increase in the heart rate and cardiac output. The increase in cardiac output is mediated more through the chronotropic than the inotropic action on the myocardium as reflected by the changes in heart rate and stroke volume. The slight fall in the systemic arterial pressure and vascular resistance suggests that some vascular beds (probably in the skeletal muscle) have receptors which respond to /3 2 -stimulation. These findings are in agreement with those of other reports.3· 4 The fact that the cardiac effects persisted after the interruption of the infusion indicates that the {3-adrenergic receptors continue to be activated for some time after the cessation of infusion. The present data also show that the mode of hydration, whether saline or glucose, dose not alter the manner by which the cardiovascular receptors respond to /3 2 -adrenergic stimulation. In regard to the lungs, the present series of experiments show that the driving pressure in the pulmonary vascular bed tended to increase during {3-receptor stimulation despite a slight fall in the pulmonary vascular resistance. These findings in the sheep are in

Po2 Pco2

Hematocrit Albumin

J

Control

7.51 95.0 33.0 24.5 2.7

± 0.01

± 3.6 ± 2.1 ± 0.7 ± 0.1

Ritodrine

7.49 98.8 31.3 30.0 2.8

± 0.02 ± 2.2

± 2.7 ± 0.8

± 0.1

Recovery

7.50 99.6 31.1 26.3 2.65

0.02 2.8 2.5 0.7 ± 0.1 ± ± ± ±

agreement with those of others in dogs and monkeys. 3 They suggest that, during activation of the {3-adrenergic receptors, the capacity of the pulmonary vascular bed does not expand sufficiently to accommodate the increment in volume flow and, therefore, the driving pressure tends to increase. Erdmann et al. 8 have demonstrated in unanesthetized nonpregnant sheep that an increase in vascular pressure, no matter how small, may have a significant effect on lung fluid balance. Therefore, the changes observed by us in the pulmonary circulation of pregnant sheep may become a contributing factor to the circulatory overload elicited by {3-adrenergic activation as discussed further below. Renal function and water and electrolyte balance. The role of the {3-adrenergic system in the control of renal functions has been a controversial subject. Levy et al. 2 and Klein et al.' 0 observed in dog:_s and rats that isoproterenol produced a profound antidiuresis which was independent of antidiuretic hormone. In their subsequent studies in human subjects, however,

872 Kleinman et al.

August 15, 1984 Am.

Ritodrine· HCI

4

e

...... :::>

2:. 11.. > 2


:2: (J)


30

60

90

120

150

180

TIME IN MINUTES

Fig. 8. Plasma levels of arginine vasopressin (AVP) during the

control, ritodrine, and recovery periods. Note the marked rise in these levels during ,13-stimulation and the persistence of elevated levels during the recovery period (n = lO experiments). these investigators 11 concluded that the antidiuresis of ,8-receptor stimulation was mediated through antidiuretic hormone release, together with changes in renal hemodynamics. Schrier et a!. 1 were able to abolish the antidiuretic effects of ,13-adrenergic-receptor stimulation with ablation of the hypothalamic-neurohypophyseal tract; they concluded that the antidiuresis is strictly dependent on antidiuretic hormone release. Recent in vivo and in vitro studies, however, have indicated that ,8-adrenergic-receptor stimulation exerts direct action on renal tubular handling of water and sodium. 12 - 14 The data obtained from our experiments on unanesthetized pregnant sheep show clearly that activation of the ,8-adrenergic receptors with ritodrine profoundly affects the renal handling of water and electrolytes through different mechanisms. In the series of experiments with maximum water diuresis (5% dextrose hydration), ,8-receptor stimulation not only strikingly reduced the urine flow but also changed the free water clearance from strongly positive to negative. Since the changes in glomerular filtration rate were minor compared to those of urine flow, this antidiuresis must have been related to increased water reabsorption. The data on antidiuretic hormone secretion support this hypothesis and indicate that ,8-receptor activation does, indeed, stimulate antidiuretic hormone release. The persistence of some antidiuretic action and elevated arginine vasopressin levels following cessation of the ritodrine infusion adds evidence to the circulatory data which showed persistent ,8-receptor stimulation following cessation of the infusion.

J. Obstet. Gynecol.

In the saline-hydrated experiments, the free water clearance was negative before, during, and after ,8-receptor stimulation; this suggests that, in these experiments, very little if any nonobligated water under antidiuretic hormone control existed in the urine. However, under these experimental conditions, ritodrine infusion still elicited profound oliguria largely related to a marked decrease in the excretion of the osmotically obligated water. Here again, the antidiuresis had no relation to the changes in renal hemodynamics; therefore, factors other than antidiuretic hormone must have played a role in the iso-osmotic reabsorption offluid in the kidney. The results of our experiments further show that, irrespective of the mode of hydrating the animal, ,8-adrenergic-receptor stimulation produces a profound fall in the excretion of sodium and potassium. The decrease in sodium excretion is totally related to an enhanced tubular reabsorption. This is in agreement with various in vivo and in vitro studies performed by other authors in different animal speciesY- 14 The question arises as to whether the antinatriuresis of ,8-stimulation is related to a direct effect on the renal tubular cells or is mediated through stimulation of the renin-angiotensin-aldosterone system or both. There is ample evidence in the literature which indicates that the ,8-adrenergic system plays an important role in renin secretion in health and disease. 15 • 16 These studies, which were carried out in different animal species including man, have shown that ,13-adrenergicreceptor stimulation increases renin blood levels within 15 to 20 minutes after the onset of the stimulus. Although in our series of experiments we did not measure the renin blood levels, it is difficult to rule out the possibility that the renin-angiotensin-aldosterone system might have played a role in the enhanced sodium reabsorption observed during ,13-receptor stimulation. Interpretation of the role of this system is, however, complicated by the behavior of potassium in the body compartments and its handling by the kidney. Our data show that ,8-receptor stimulation with ritodrine produced a profound fall in potassium excretion which was not reflected by a rise in the plasma levels of this cation. In fact, plasma potassium fell during and continued to fall after ritodrine infusion, probably because of movement of this ion into the intracellular compartment. At any rate, these changes in the intracellular and extracellular potassium levels affect the stimulation of aldosterone secretion and make the interpretation of the role of the renin-angiotensin-aldosterone system in the water and electrolyte reabsorption during ,8-stimulation extremely difficult. The decrease in the renal plasma flow, in the face of

Volume 149 Number 8

increasing cardiac output and falling systemic vascular resistance during /3-stimulation, is intriguing; it reflects an active vasoconstriction in the renal vascular bed, probably in the afferent arterioles. It is most likely that this renal vasoconstriction that occurred during /3-receptor stimulation is related to a rise in the reninangiotensin levels since the extreme sensitivity of this vascular bed to these substances has been well known for many years. Implications of our data in pulmonary edema. Although no animal developed signs or symptoms of pulmonary edema in any of our experiments, our results on systemic, pulmonary, and renal effects of /3-receptor stimulation permit the formulation of a conceptual mechanism of this complication. Pulmonary edema may be related to either hemodynamic factors generated by a circulatory overload or to toxic factors altering endothelial permeability and thereby permitting the diffusion of colloids and water into the extravascular compartment of the lungs. The available evidence collected by us (to be published) and by others seems to rule out the possibility that, in the dosage used clinically to arrest premature labor, the /3-mimetics have pulmonary toxic effects which may lead to a toxic type of pulmonary edema. On the contrary, the available information overwhelmingly points toward circulatory overload as being the primary factor. It is worthwhile, therefore, to dissect the various components of this circulatory overload which, in a given case, may combine in various proportions to produce pulmonary edema. The first component to be considered is the gestational increase in the cardiac output and blood volume. It is well known that these two parameters increase progressively during pregnancy, reaching a peak of 40% to 45% over nonpregnant levels at about 30 weeks. Although the maternal organism adjusts to this hypervolemic, hyperdynamic state without any increase in the systemic arterial and pulmonary arterial pressures, nevertheless and for the purpose of this discussion, it should be considered as an added load on the cardiovascular system which reaches a maximum at a time when the /3-mimetics are to be used, The second component to be discussed is related to the central hemodynamic effects of {3-adrenergic-receptor stimulation. As demonstrated in the present series of experiments and by others, even the most selective {3 2-adrenergic agonists such as ritodrine increase the cardiac output by more than 48%. From the hemodynamic point of view, this increase should be considered as an added circulatory load particularly on the pulmonary vascular bed as evidenced by the increase in the pulmonary arterial and pulmonary wedge pressures. The obvious inability of the pulmonary circula-

Effects of {:!-adrenergic-receptor stimulation

873

tion to accommodate this extra load without increasing the pressure creates the tendency to augmented fluid diffusion to the pulmonary extravascular space as demonstrated by Erdmann et aJ.B The third component of circulatory overload is related to the amount of fluid reabsorbed by the kidney both isosmotically with sodium and as free water under antidiuretic hormone control. As shown in the present studies, this amount can be considerable, depending on the duration of infusion and the nature of the infusate. Obviously, because of the antinatriuretic properties of /3-agonists, the expansion of body fluids is greater when the /3-mimetics are administered with isotonic saline solutions than with dextrose solutions. When the overall effects of the three components of circulatory overload are added, it becomes clear that, during the administration of {3-mimetics to arrest premature labor, the stage is set for the development of pulmonary edema. In this regard, two important factors should be kept in mind. The first is that water and electrolytes are retained significantly in the extracellular space for the duration of the /3-receptor stimulation. The second is that interruption of the infusion does not mean total cessation of the stimulus. As our circulatory and renal studies show, the effects of infusion of /3-mimetics persist for 60 minutes or longer, depending on the duration of the treatment. REFERENCES 1. Schrier RW, Lieberman R, Ufferman RC. Mechanism of antidiuretic effect of beta-adrenergic stimulation. J Clin Invest 1972;51:97. 2. Levy J, GrinblatJ, Kleeman CR. Effects of isoproterenol on water diuresis in rats with congenital diabetes insipidus. AmJ Physiol1971;221:1728. 3. HauthJC, Hawkins GD, Kuehl TJ, Pierson WP. Ritodrine hydrochloride infusion in pregnant baboons. I. Biophysical effects. AM J 0BSTET GYNECOL 1983; 146:916. 4. Nuwayhid B, Cabalum MT, Lieb SM, eta!. Hemodynamic effects of isoxsuprine and terbutaline in pregnant and nonpregnant sheep. AMj 0BSTET GYNECOL 1980; 137:25. 5. Katz M, Robertson PA, Creasy RK. Cardiovascular complications associated with terbutaline treatment for preterm labor. AM J 0BSTET GYNECOL 1981;139:605. 6. Assali NS, Dignam WJ, Longo L. Renal function in human pregnancy. Ill. Effects of antidiuretic hormone (ADH) on renal hemodynamics and water and electrolyte excretion near term and postpartum. J Clin Endocrinol Metab 1960;20:581. 7. Weitzman RW, Glatz TH, Fisher DA. The effect of hemorrhage and hypertonic saline upon plasma oxytocin and arginine vasopressin in conscious dogs. Endocrinology 1978; 103:2154. 8. Erdmann AJ III, Vaughan TRJr, Brinham KL, Woolverton WC, Staub NC. Effect of increased vascular pressure on lung fluid balance in unanesthetized sheep. Cir Res 1975;37:271. 9. Philipsen T, Eriksen PS, Lynggard F. Pulmonary edema following ritodrine-saline infusion. Obstet Gynecol 1981; 58:304. 10. Klein LA, Lieberman B, Larks M, Kleeman CR. Interrelated effects of antidiuretic hormone and adrenergic

874

Kleinman et al.

August 15, 1984 Am.

drugs on water metabolism. Am J Physiol 1971;221: 1657. II. Levy J, Coburn J, Kleeman CR. Mechanism of the antidiuretic effect of beta-adrenergic stimulation in man. Arch Intern Med 1976;136:25. 12. Blendis LM, Auld RM, Alexander EA, Levinsky NG. Effect of beta and alpha-adrenergic stimulation on proximal sodium reabsorption in dogs. Clin Sci 1973;43:569. 13. Bellow-Reuss E. Effects of catecholamines on fluid reabsorption by the isolated proximal convoluted tubule. Am J Physiol I980;238:F347.

J. Obstet. Gynecol.

14. DiBona GF. Neurogenic regulation of renal tubular sodium reabsorption. Am J Physiol 1977;233:F73. 15. Oates HF, Stoker LM, Monaghan JC, Stokes GS. The beta-adrenoreceptor controlling renin release. Arch Int Pharmacodyn 1978;234:205. 16. Winer N, Chockshi DS, Yoon MS, Freedman AD. Adrenergic receptor mediation of renin secretion. J Clin Endocrinol 1969;29: 1168.