- Email: [email protected]
Hemodynamic and fluid responses to furosemide infusion in the ovine fetus
Rodgers et al.
17.
18. 19.
20.
21.
and classification of enzymes. New York: Academic Press, 1991. Quantin B, Murphy G, Breathnach R. Pump-1 cDNA codes for a protein with characteristics similar to those of classical collagenase family members. Biochemistry 1989; 28:5327-34. Woessner jFj, Taplin CJ. Purification and properties of a small latent matrix metalloproteinase of the rat uterus. j Bioi Chern 1988;263:16918-25. Rentrop MB, Knapp B, Winter H, Schweizer j. Aminoalkylsilane-treated glass slides as support for in situ hybridization of keratin cDNAs to frozen tissue sections under varying fixation and pretreatment conditions. Histochemj 1986;18:271-6. McDonnell SM, Navre M, Coffey Rj, Matrisian LM. Expression and localization of the matrix metalloproteinase PUMP-1 (MMP-7) in human gastric and colon carcinomas. Mol Carcin 1991;4:527-33. Muller D, Quantin B, Gesnel Me, Millon CR, Abecassisj,
January 1993 Am J Obstet Gynecol
22.
23.
24.
25.
Breathnach R. The collagenase gene family in humans consists of at least four members. Biochem j 1988;253: 187-92. Melton D, Krieg P, Rebagliati M, Maniatis T, Zinn K, Green M. Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promotor. Nucl Acids Res 1984; 12:7035-56. Cox KH, DeLeon DM, Angerer LM, Angerer Re. Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetric RNA probes. Dev Bioi 1984;101: 485-502. Angerer LM, Stoler MH, Angerer RC. In: Valentino KL, Eberwine jH, Barchas jD, eds. In situ hybridization with RNA probes: an annotated recipe. New York: Oxford University Press, 1987:42-70. Cathala G, Savouret j-F, Mendez B, et al. A method for isolation of intact, translationally active ribonucleic acid. DNA 1983;2:329-35.
Hemodynamic and fluid responses to furosemide infusion in the ovine fetus Thomas F. Kelly, MD, Thomas R. Moore, MD, and Robert A. Brace, PhD
San Diego, California OBJECTIVES: The direct effects of furosemide infusion have not been systematically examined in the fetus. Our objective was to explore the hemodynamic and urinary responses to a 4-hour infusion of furosemide into fetal sheep. STUDY DESIGN: Furosemide (0, 1, 5, or 10 mg/hr) was infused into the fetal inferior vena cava for 4 hours in 15 chronically catheterized near-term sheep. RESULTS: Relative to vehicle infusion, furosemide produced dose-dependent Increases in fetal arterial pressure (analysis of variance, p < 0.05 when comparing groups), fetal heart rate (p < 0.0001), urine flow (p < 0.0001), and urine osmolality, sodium, and chloride, concentrations (p < 0.0001). Concomitantly, there were dose-dependent decreases in fetoplacental blood volume, fetal plasma chloride (p < 0.0001) and fetal venous pressure (p < 0.05). The changes in urine flow rate and fetal arterial pressure were positively correlated (r = 0.46, P < 0.01), suggesting that the diuresis was mediated in part by fetal arterial pressure. The four-quadrant amniotic fluid index increased during the furosemide infusions but not during vehicle infusions (p = 0.022). CONCLUSIONS: Furosemide infuSion caused a marked dose-dependent diuresis in the ovine fetus that appears to be partly mediated by increases in vascular pressure. Although amniotic fluid volume increases and fetal blood volume decreases, the reduction in blood volume was small compared with the urine volume excreted. (AM J OBSTEr GVNECOL 1993;168:260-8.)
Key words: Fetal, urinary output, amniotic fluid index, blood volume, sheep
From the Division of Perinatal Medicine, Department of ReproductIVe Medicine, University of California, San Diego. Supported by National Institutes ofHealth grant HD 23724 from the National Institute of Child Health and Human Development. Received for publication January 30, 1992; reVISed July 9, 1992; accepted July 16, 1992. Reprint requests: Robert A. Brace, PhD, Department of Reproductive Medicine, University of California, San Diego, LaJolla, CA 920930802. 6/1/41092
260
Furosemide is a diuretic that has been used in a variety of clinical situations during pregnancy, including evaluation of fetal urine production and human fetal intravascular transfusions. 1. 2 Although this drug has been studied in various adult settings, there are little data regarding the fetal responses. Prior studies reported that a bolus injection of furosemide into the ovine fetus produced increases in fetal urine produc0002-9378/93 $1.00
+ 0.20
Volume 168 Number I, Part 1
Fetal sheep responses to furosemide infusion
261
tion.'· 4 However, the mechanism of furosemide-induced diuresis, particularly the role of cardiovascular responses, has not been elucidated completely. Further, it is not known whether a continuous infusion of furosemide will produce a sustained diuresis in the fetus and a resultant expansion of amniotic fluid volume. Thus our study was designed to determine the fetal cardiovascular and urinary responses to a continuous infusion of furosemide.
Table I. Fetal and amniotic control values
Material and methods Animal preparation and maintenance. The protocol used in this study was approved by our institution's animal subjects committee. Fifteen time-dated, nearterm pregnant sheep were used. Their source, preparation, and maintenance have been previously described.' Gestational age at the time of surgery was 125 ± 1 (mean ± SE) days. In brief, the ewe was anesthetized with 1 gm intravenous thiopental sodium, and anesthesia was maintained with 1% to 2% halothane in oxygen. The maternal abdomen was prepared and draped, and a laparotomy was performed. The uterus was exposed and opened, and the fetal hind limbs were extracted. Vascular catheters were placed in the fetal inferior vena cava and descending aorta via the femoral vessels. The fetal urinary bladder was catheterized via a suprapubic incision. Additional catheters were attached to the fetal limbs for accessing the amniotic fluid. The hysterotomy and laparotomy incisions were closed, and the maternal inferior vena cava and descending aorta were catheterized. All catheters were tunneled subcutaneously and exited through a small incision in the left flank. The animals were maintained postoperatively for 5 days on prophylactic antibiotics. 5 All vascular catheters were flushed daily with heparin sodium 1000 Vlml (Elkins-Sinn, Cherry Hill, N.J.) to maintain patency. Experimentation began on postoperative day 5. At the time of experimentation the animals were given furosemide or control infusions randomly, and most had more than one dose given. No animal received the same dose twice. At least 48 hours elapsed between experiments on each animal. Thirty-two vehicle (normal saline solution, Abbott Laboratories, North Chicago) or furosemide infusions were performed in 15 animals. Experimental protocol. A I-hour control period was followed by a four-hour infusion of furosemide into the fetal vena cava at doses of 0, 1, 5, or to mgihr. These doses were chosen because of previous literature showing significant urinary effects in the neonate at 1 mg/kg.6 The infusion was maintained at 0.12 mVmin (Harvard Pump, Harvard Apparatus, South Natic, Mass.). Mter the infusion there was a I-hour recovery period. Experimental measures. Fetal arterial pressure,
Values are mean ± SEM; n = 32 for all, except n = 24 for amniotic fluid index.
Fetal arterial pressure (mm Hg) Fetal venous pressure (mm Hg) Fetal heart rate (beats/min) Fetal urine flow rate (mVmin) Fetal arterial pH Fetal arterial Pco2 (mm Hg) Fetal arterial P02 (mm Hg) Fetal hematocrit (%) Fetal plasma protein concentration (gm/dl) Amniotic fluid index (mm)
47.5 2.9 167 0.62 7.31 53.8 20.5 34.2 3.8 78
± ± ± ± ± ± ±
± ±
±
0.8 0.4 3 0.05 0.01 0.5 0.4 0.7 0.1 3
venous pressure, amniotic fluid pressure, heart rate, and urine flow rate were continuously recorded on a polygraph recorder (Gould, Cleveland), and 30-second averages were stored with an on-line computer (Hewlett-Packard, Eybens, France) as prevously described. 7 • 8 Fetal urine was continuously returned to the amniotic compartment. Vascular pressures were referenced to the standard zero pressure reference for the fetus by continuously subtracting amniotic fluid pressure with the computer. Samples of fetal arterial blood (1.5 ml), maternal arterial blood (1.0 ml), fetal urine (0.5 ml), and amniotic fluid (1.0 ml) were taken at 30 and 60 minutes into the control period and every 60 minutes thereafter. Fetal blood was replaced immediately with an equal volume of heparinized maternal blood collected before the study. Fetal blood was used for the measurement of pH, Pco 2 , P0 2 at 39S C (model 1306 pHlblood gas analyzer, Instrumentation Laboratory, Lexington, Mass.), hematocrit in triplicate,9 plasma osmolality by freezing point depression (Advanced Instruments Inc., Needham Heights, Mass.), protein concentration (TS meter, American Optical Scientific Instruments, Buffalo, N.Y.), and Na+, K+ and CI- concentrations (model 5 + 5 electrolyte analyzer, Nova, Waltham, Mass.). The changes in blood volume relative to control were derived by using the inverse of the hematocrit as previously described. 9 Osmolality and electrolytes were measured on maternal blood, fetal urine, and amniotic fluid. A four-quadrant amniotic fluid index was performed by using real-time ultrasonography (3.5 MHz linear-array scanner, Siemens, Germany) in 24 of the 32 experiments. lo The amniotic fluid index measurement was not always successful because of an occasional uncooperative animal. Animals were killed with a pentobarbital sodium euthanasia solution (100 mg/kg) after completion of the experiments. Fetuses were towel dried and weighed. Assuming a 3% fetal weight gain per day, II weights at the time of experimentation were calculated for those animals not killed immediately after the last experiment. Statistical analysis. Experiments were divided into
262
Kelly, Moore, and Brace
January 1993 Am J Obstet Gynecol
110
D
.009
.............. ............. ....... ........
100
. ....... , ..... .
).)
'I-l
o
110
100hllo;.,i.....~
110
100
o
-1
1
2
3
4
5
TIME (hours) Fig. 1. Fetal arterial pressure responses to different infusions of furosemide (A, vehicle; B, 1; C, 5, and D, 10 mglhr; n = S for each dose) over 4 hours. Data expressed as mean ± SEM percent change from values averaged during I-hour preinfusion period. Shaded area, 95% Confidence interval of control values. Vertical dotted lines, Beginning and end of 4-hour infusion.
Table II. Fetal plasma, urine, amniotic fluid and maternal plasma control values Osmolality (mOsm/kg) Fetal plasma Maternal plasma Amniotic fluid Fetal urine
302 301 273 141
± 1 ± 1 ± 7 ± 7
Sodium
(mEq/L)
13S.9 144.3 115.0 32.2
± 0.3 0.3 ± 2.7 ± 3.2 ±
Potassium (mEq/L)
4.4 ± 0.1
4.3 ± 0.0
7.1 ± 0.6 7.0 ± O.S
Chloride (mEq/L)
10S.7 114.0 9S.5 30.7
± 0.4 ± 0.3 ± 2.4 ± 3.1
Values are mean ± SEM; n = 32.
four groups, with each group receiving either 1,5, or 10 mglhr infusions of furosemide or vehicle. Data are expressed as mean ± SEM, and are plotted as either a change from control or as a percentage of control to reduce interanimal variability. Control values were taken as the mean value during the 60-minute preinfusion period. These values were analyzed for statistical differences among the four treatment groups by using a three-way analysis of variance, with time, treatment, and animal being the three factors. Changes with time in each treatment group were analyzed with a two-way analysis of variance with repeated measures. Assessment
of significant differences in the responses to the four infusions were derived from the interaction term of a three-way analysis of variance. Statistical significance was taken as p ::s; 0.05. To determine whether a significant relationship existed between variables, linear regression was used on the average values from each animal during the 4-hour infusion period. Results
The mean fetal age at experimentation was 134 ± 1 days; mean fetal weight was 3.03 ± 0.10 kg. Furosemide was infused at 0.35 ± 0.02 (n = 8), 1.83 ± 0.11
Fetal sheep responses to furosemide infusion
Volume 168 Number I, Part 1
263
: p < 0.05
1
a
-1
-1
o
2
1
3
4
5
TIME (hours) Fig. 2. Fetal venous pressure changes in response to either vehicle and I mglhr (n = 16, dots) or 5 and 10 mglhr (n = 16, sqw.res) infusions from 0 to 4 hours. Data expressed as mean ± SEM changes from values during I-hour preinfusion period. VertICal dotted lmes, Beginning and end of infusion; shaded area, 95% confidence interval of control values.
(n = 8), or 3.37 ± 0.32 (n = 8) mglkglkr in the 1, 5, and 10 mglhr groups, respectively. Control values for the maternal and fetal variables during the I-hour preinfusion period are listed in Tables I and II. There were no significant differences during the control period among the four groups except for a small difference in fetal plasma sodium in which the vehicle group was significantly lower than the three groups that received furosemide (137.6 ± 0.4 vs 138.9 ± 0.6 mEqlL, respectively p = 0.03). Fetal cardiovascular effects. Fetal arterial pressure increased significantly during the 5 and 10 mg/hr infusions (p < 0.05 when comparing groups, Fig. 1). During the 10 mglhr infusion, the peak in arterial pressure occurred earlier and was more sustained (p < 0.009) than the peaks associated with the I or 5 mglhr infusions. Fetal venous pressure was unchanged during the vehicle or 1 mglhr furosemide infusions but decreased significantly during the 5 and 10 mglhr infusions (p < 0.05, Fig. 2), with a reduction approaching 1.0 mID Hg during the last 2 hours of the infusion. There were significant differences in the fetal heart rate responses to vehicle versus furosemide infusion. Fetal heart rate fell significantly during the vehicle infusion from a control mean of 167 ± 15 to 141 ± 5 beats/min (p < 0.00001) by the end of the infusion. In the groups receiving furosemide there was a significant
dose-dependent effect (p < 0.0001), with mean heart rate changing by - 5, - 3, and + 8 beats/min during the last 2 hours of the 1, 5, and 10 mg/hr infusions, respectively. Fetal blood volume effects. Marked dose-related decreases in fetal blood volume were significant in all groups receiving furosemide compared with vehicle infusions (p < 0.00001, Fig. 3 ). Blood volume decreased by 3.6% ± 0.9%, 7.3% ± 1.7%, and 8.9% ± 1.3% in the 1, 5, and 10 mglhr groups, respectively, representing an average blood volume loss of 12, 23, and 31 mI.12 Concomitantly, fetal plasma protein concentration increased significantly in all groups receiving furosemide (p < 0.00001) by 7.1% ± 1.3%, 11.0% ± 2.2%, and 15.7% ± 2.6% at 4 hours into the infusion of 1,5, and 10 mglhr of furosemide. Fetal urine flow effects. Urine flow rose significantly above control in all groups receiving furosemide (p < 0.00001, Fig. 4). Overall, the average increases in urine flow in the 1, 5, and 10 mg/hr groups during the 4-hour infusion period were 0.6 ± 0.2, 0.9 ± 0.2, and 1.1 ± 0.2 mllmin, respectively, above control, corresponding to excess excretion volumes of 141, 221, and 252 ml. During the 1 and 5 mglhr infusions the peak flow occurred at 120 minutes into the infusion, whereas the peak occurred at 60 minutes during the 10 mg/hr infusion. The urine flow change exhibited a significant linear
264
Kelly, Moore, and Brace
January 1993 Am J Obstet Gynecol
105
>-I
95
+I
90
0 1-1
c::
0
u
85
0
105
.....
.:P
D
100
'" ............ " ........ .01 ..........
...T
'I' ---- ----- ____ .1
T
.i.
T
:P
c
d(>
100
< 0.00001
'!' .....01"'.~"" .... ~ ... Jo:"'....... '" 10:10.......':.. 10.10.. . . . .1"Jo: 10.• • •:~..r"'Io.Io. ••• ~
< 0.00001
rz:I
~
95
~
90
~
§ 0
~
I:Q
85 105
~ rz:I ~
100 95 105 100 95
: P < 0.003
B
~
. .-
............I.;.:··:.:.··cc,.·~ ... ,.,..oj>·"'·" ...... "'... .,..,.,. ..'. . ,."'...... ·,.'Iw,..j.:... •• ...
...-
rr ::1~;~. ~. A
:
H
-1
0
~
..
:tt:~m''T.::h;",,*:*..t*:"".'_ _j";- - "..: .
.w.'' l
•
t+ . . . . ·!"'...... ~ .......
1
2
3
4
5
TIME (hours) Fig. 3. Fetal blood volume responses to infusion of vehicle (A, n = 8) or furosemide infusions (B, 1; C, 5; D, 10 mglhr; n = 8 for each group) from 0 to 4 hours. Data expressed as mean ± SEM percent of mean value during I-hour preinfusion period. Shaded area, 95% Confidence interval of control values; vertical dotted lines, beginning and end of 4-hour infusion.
correlation with the fetal arterial pressure change in all animals during the furosemide infusions (r = 0.46, P < 0.01). Fetal urine composition effects. The diuresis was associated with dose-dependent increases in urinary osmolality to 58 ± II, 78 ± 12, and 103 ± 12 mOsm/kg above control in the groups receiving furosemide at 1, 5, and 10 mglhr (p < 0.00001). Free water clearance during the vehicle infusion rose significantly in all groups receiving furosemide (Fig. 5). The peak free water clearance was 51% ± 27%, 139% ± 38%, and 66% ± 37% above control in the 1, 5, and 10 mglhr groups, respectively. Urinary sodium and chloride concentrations increased significantly during the diuresis (p < 0.00001). The sodium excretion rate increased in all furosemide groups in a dose-dependent fashion (98 ± 32, 150 ± 23, and 224 ± 33 ILEqlmin). The pattern of urinary chloride excretion was similar.
Fetal blood and plasma effects. There were no significant differences in the fetal plasma osmolality responses to vehicle or furosemide infusion. The urinary losses of electrolytes were reflected in significant decreases in fetal plasma chloride when comparing groups (p < 0.00001). At the 0, 1, 5, and 10 mglhr doses, changes averaged +0.5% ± 0.5%, -1.4% ± 0.5%, -3.0% ± 0.4%, and -3.7% ± 0.4%, respectively. Fetal plasma potassium did not change significantly. Fetal plasma sodium remained unchanged during the vehicle and 1 and 5 mglhr infusions but fell during the 10 mglhr infusion by 1.0% ± 0.2% (p = 0.002). Fetal pH and P0 2 did not change during vehicle or furosemide infusions. Fetal Pco2 increased significantly during the 5 and 10 mglhr infusions when compared with the vehicle (p < 0.003), with peaks of 2.1 ± 0.6 and 2.6 ± 0.9 mm Hg occurring respectively in the postinfusion period. Amniotic fluid effects. The amniotic fluid index was
Volume 168 Number I , Part 1
Fetal sheep responses to furosemide infusion
265
2
:p
D
< 0.00001
1
c
1
< 0.00001
a hw.......~··· · ·················· ·· ·············.· .. ..... : . ... ... .. .. .• . B
1
1
~
A
a : •••• -1
< 0.0001
t.....:•••••:.....~.....l.·,..'.:·· ::
a
:
1
2
3
4
NS
5
TIME (hours) Fig. 4. Fetal urine flow rate responses to infusion of vehicle (A, n = 8) or furosemide infusions (B, 1; C, 5; D, 10 mglhr; n = 8 for each group) from 0 to 4 hours. Data expressed as mean ± SEM change from values during I-hour control period. S1uuJ.ed area, 95% Confidence interval of control values.
unchanged during the vehicle infusion and increased in the furosemide-treated groups (p = 0.022). During the I, 5, and lO mglhr infusions, the respective increases were 23% ± 3% (p = 0.016), 15% ± 5% (p = 0.013), and 44% ± 13% (p = 0.005) above control. Maternal plasma effects. There were no differences among groups in the maternal plasma sodium or osmolality in response to the infusions. Maternal potassium underwent dose-dependent decreases with significant differences among treatment groups (p < 0.00001), with the changes being 0.1 ± 0.1, -0.1 ± 0.1, -0.3 ± 0.1, and -0.6 ± 0.1 mEqlL at 4 hours into the furosemide infusion in the 0, 1,5, and 10 mglhr groups, respectively. The maternal plasma chloride showed significant differences : among groups (p = 0.0002) with a maximal decrease of 2.6 mEqlL during the 10 mg/hr infusion when compared with the vehicle.
Comment Our study suggests that furosemide infusion is associated with dose-dependent cardiovascular and renal
effects in the ovine fetus; these effects lead to increases in amniotic fluid volume. The cardiovascular responses included elevations in fetal arterial pressure, which were significant during the 5 and 10 mglhr infusions. These elevations may reflect activation of the renin-angiotensin system, which has been previously shown in the fetus and adult", 13. 14 Lumbers and Stevens' noted that when a bolus of furosemide was given to fetuses made hypotensive with nitroprusside there was a significant rise in diastolic pressure. They also demonstrated that saralasin, a competitive antagonist of angiotensin II, inhibited the rise in blood pressure associated with furosemide. Studies of human beings have shown peripheral vascular responses to furosemide when given as a bolus, and these were related to renin release and elevated angiotensin II concentrations. 15 Furthermore, increases in renin activity associated with furosemide have been shown to be inhibited by indomethacin, suggesting prostaglandins are involved. 16 Thus the elevation in fetal arterial pressure response to furosemide was likely caused by the activation of the renin-angiotensin system, possibly mediated by prostaglandins.
266
Kelly, Moore, and Brace
January 1993 Am J Obstet Gynecol
200 150 r-I
o J..4
.j.J
c o u
100 50
I I • I I I I
L..-_ _....&...._ _---L_ _......JL..-_ _-'--_ _- ' -_ _- - ' _
~
o 250 200 150 100
:::~····~········1·········i···········,
150 100
-1
«.~
••-
+. . . . -..-.. . ~«.,««.<.
••-
........... , ..
o
1
2
3
4
5
TIME (hours) Fig. 5. Free water clearance responses to vehicle 0, n = 8) or furosemide infusions (B, 1; C, 5; D, 10 mglhr; n = 8 for each group) from 0 to 4 hours. Data expressed as mean ± SEM as percent of preinfusion period. Shaded area, 95% Confidence interval of preinfusion mean.
The drop in fetal heart rate (FHR) during the vehicle infusion reflects the nonnal diurnal variation we have previously reported. 17 The FHR in the groups receiving furosemide differed significantly from that in the vehicle infusion group in that furosemide prevented the decrease with time that occurred during the vehicle infusion. These relative increases in FHR in the animals receiving furosemide may represent a direct effect of angiotensin II on the fetal heart as previously shown. 18 Fetal venous pressure decreased in the 5 and 10 mglhr groups at the end of the infusion period. This decrease is consistent with the finding of decreased right atrial pressure noted in the adult human by Dikshit et al. 19 in response to furosemide and increased venous capacitance noted by Johnston et al. 20 They concluded that furosemide acted by dilating capacitance vessels as well by its tubular effects in the kidney. This effect of furosemide on vascular smooth muscle has also been observed in the rabbit, where there has been a decrease in resting tension. 21 The prolonged diuresis caused by furosemide also would tend to de-
crease fetal venous pressure resulting from decreases in intravascular volume. Fetal blood volume decreased in a dose-dependent manner. However, only a small fraction of the excess urine produced during furosemide infusion can be accounted for on the basis of blood volume loss. In the 10 mglhr group, the increased urine produced was 252 ml, and the blood volume decrease was 31. 2 ml. Therefore approximately 88% of the excess fluid lost through the kidneys must have been recruited from other sources. The fetal extravascular space was the likely source, although gastrointestinal and intramembranous absorption22 and transplacental flow also may have contributed. In addition to the cardiovascular observations there were several new findings regarding fetal urine flow. We found that a dose-dependent effect on urine flow appeared initially. By the end of the infusion period the urine flow rate in all groups receiving furosemide was similar, possibly because of either saturation of the tubular transport mechanism for furosemide or satura-
Volume 168 Number I. Part 1
tion at the site of action of furosemide in the proximal tubule lumen. Also, as the fetus becomes volume depleted furosemide may not be as effective in promoting further increases in free water clearance. Free water clearance may be dependent on the increased concentration of plasma proteins, thereby increasing protein osmotic pressure. This theory is consistent with our observation of return of free water clearance to normal despite the continuous furosemide infusion. The observed correlation between fetal arterial pressure and urine flow suggests that approximately 20% of the diuresis associated with furosemide infusion may be pressure related. Alternatively, the juxtaglomerular apparatus may sense a decrease in circulating volume caused by furosemide and increase renin output, resulting in an increase in fetal arterial pressure. The amniotic fluid index increased dose dependently in all groups receiving the furosemide, but not in the vehicle, which most likely reflects the increased urine output caused by furosemide, but changes in swallowing and lung secretion and intramembranous flow may have contributed. There were, however, no consistent changes in the amniotic fluid osmolality or electrolytes except at the highest furosemide infusion rate, when increases in sodium and chloride concentrations occurred. These increases are surprising in view of the large volumes of urine entering the amniotic space and suggest a large, yet undefined capacity to maintain amniotic composition at near-normal levels. We speculate that this is largely the result of the exchange of water and electrolytes across the intramembranous pathway.22 Maternal potassium and chloride changes appear to reflect the transplacental effect of furosemide. Furosemide has been shown to cross the human placenta!' hence the rationale for the "Lasix challenge test. "24 However, in subsequent sheep studies Chamberlain et a1.' used a furosemide assay sensitive to 100 ng/ml to show that there was no transfer of furosemide from the maternal to the fetal intravascular compartments. Similarly, Ross et a1. 28 found no evidence of transplacental effects of furosemide in the sheep. Although furosemide may not cross the placenta from the fetal to the maternal compartment, there are clear maternal effects that may reflect the maintenance of the concentration gradients of potassium and chloride across the placenta. For example, the fall in maternal chloride may have been the consequence of the reduction in fetal plasma chloride concentration. In summary, furosemide elicits a strong diuretic response in the ovine fetus and is associated with important dose-dependent cardiovascular effects. Fetal arterial pressure increases with higher doses, which may reflect activation of the renin-angiotensin system via prostaglandins. The fetal blood volume and venous
Fetal sheep responses to furosemide infusion
267
pressure falls significantly in animals receiving the diuretic. Amniotic fluid volume increases in response to the increased urine output, and changes in the maternal plasma probably reflect the maintenance of concentration gradients across the placenta. We thank Eddie Kwan for his technical assistance.
REFERENCES
1. Barrett R], Rayburn WF, Barr M. Furosemide (Lasix) challenge test in assessing bilateral fetal hydronephrosis. AM] OBSTET GYNECOL 1983;147:846-7. 2. Chestnut DH, Pollack KL, Weiner CP, Robillard ]E, Thompson CS, DeBruyn CS. Does furosemide alter the hemodynamic response to rapid intravascular transfusion of the anemic fetal lamb? AM] OBSTET GYNECOL 1989;161: 1571-5. 3. Chamberlain PF, Cumming M, Torchia MG, Biehl D, Manning FA. Ovine fetal urine production following maternal intravenous furosemide administration. AM] OBSTET GVNECOL 1985;151:815-9. 4. Lumbers ER, Stevens AD. The effects offrusemide, saralasin and hypotension on fetal plasma renin activity and on fetal renal function.] Physiol (Lond) 1987;393:479-90. 5. Brace RA, Cheung CY. Fetal cardiovascular and endocrine responses to prolonged fetal hemorrhage. Am ] Physiol 1986;251 :R417-24. 6. Woo WC, Dupont C, Collinge ], Aranda ]V. Effects of furosemide in the newborn. Clin Pharmacol Ther 1978; 23:266-71. 7. Brace RA, Brittingham DS. Fetal vascular pressure and heart rate responses to nonlabor uterine contractions. Am ] Physiol 1986;251:R409-16. 8. Brace RA, Miner LK, Siderowf AD, Cheung CY. Fetal and adult urine flow and ANF responses to vascular volume loading. Am] Physiol 1988;255:R846-50. 9. Brace RA. Blood volume and its measurement in the chronically catheterized sheep fetus. Am] Physiol 1983; 244:H487-94. 10. Moore TR, Brace RA. Amniotic fluid index (AFI) in the term ovine pregnancy: a predictable relationship between AFI and amniotic fluid volume [Abstract 286]. In: Proceedings of the thirty-sixth annual meeting of the Society for Gynecologic Investigation, San Diego, California, March 15-18, 1989. 11. Gresham EL, Rankin ]HG. Makowski EL, Meschia G. Battaglia FC. An evaluation of fetal renal functIOn in chronic sheep preparation.] Clin Invest 1972;51:149-55. 12. Brace RA. Fetal blood volume, extracellular fluid and lymphatic function. In: Brace RA, Ross MG, Robillard ]E, eds. Reproductive and perinatal medicine. Volume 11: Fetal and neonatal body fluids: the scientific basis for clinical practice. New York: Perinatology Press, 1989:1-18. 13. Siegel SR, Leake RD, Weitzman RE, Fisher DA. Effects of furosemide and acute salt loading on vasopressin and renin secretion in the fetal lamb. Pediatr Res 1980; 14:86971. 14. Corsini WA, Hook ]B, Bailie MD. Control of renin secretion in the dog. Effects of furosemide on the vascular and macula densa receptors. Circ Res 1975;224:425-30. 15. Passmore AP, Whitehead EM, Johnston GD. Comparison of the acute renal and peripheral vascular responses to frusemide and bumetanide at low and high dose. Br] Clin Ph..rmacol 1989;27:305-12. 16. Mackay]G, Muir AL, Watson ML. Contribution of prostaglandins to the systemic and renal vascular response to fiusemide in normal man. Br] Clin Pharmacol 1984; 17: 513-9. 17. Lawler FH, Brace RA. Fetal and maternal arterial pres-
Kelly, Moore, and Brace
18. 19.
20. 21.
sures and heart rates: histograms, correlations and rhythms. Am J Physiol 1982;243:R433-44. Iwamoto HS, Rudolph AM. Effects of angiotensin II on the blood flow and its distribution in fetal lambs. Circ Res 1982;48: 183-9. Dikshit K, VydenJK, Forrester JS, ChatteIjee MB, Prakash R, Swan HJC. Renal and extrarenal hemodynamic effects of furosemide in congestive heart failure after acute myocardial infarction. N Engl J Med 1973;238: 1087-90. Johnston GD, Nicholls DP, Kondowe GB, Finch MB. Comparison of the acute vascular effects of frusemide and bumetanide. Br J Clin Pharmacol 1986;21:359-64. Andreasen F, Christensen JH. The effect of furosemide on
January 1993 Am J Obstet Gynecol
22. 23. 24. 25.
vascular smooth muscle is influenced by plasma protein. Pharmacol Toxicol 1988;63:324-6. Gilbert WM, Brace RA. The missing link in amniotic fluid volume regulation: intramembranous absorption. Obstet Gynecol 1989;74:748-53. Beermann B, Groschinsky-Grind M, Fahraeus L, Lindstrom B. Placental transfer of furosemide. Clin Pharmacol Ther 1978;24:560-2. Wladimiroff JW. Effect of frusemide on fetal urine production. Br J Obstet GynaecoI1975;82:221-4. Ross MG, Ervin MG, Leake RD. Failure of Lasix to induce a fetal diuresis. AM J OBSTET GYNECOL 1985; 152:1107.
Pregnancy-induced changes in the three-dimensional mechanical properties of pressurized rat uteroplacental (radial) arteries George Osol, PhD, and Marilyn Cipolla, BS Burlington, Vermont OBJECTIVE: The purpose of this study was to describe the effects of pregnancy on the size and three-dimensional mechanics of uterine radial arteries. STUDY DESIGN: Measurements of lumen diameter, wall thickness, and axial and radial distensibility were made in situ and in pressurized segments of excised vessels from nonpregnant (n = 29) and late-pregnant (days 19 to 21, n = 19) Sprague-Dawley rats. RESULTS: In the unpressurized state, the overall length of the radial artery segment of the arcade increased approximately 4.8 times during gestation. Lumen diameter increased by 60%, as did distensibility in both the radial and axial directions. However, there was no measurable change in the thickness of the vascular wall, which was muscular in appearance and comprised approximately two layers of circumferentially oriented smooth muscle, a relatively thick internal elastic lamina, and a well-defined endothelial layer. Cross-sectional area increased significantly during pregnancy (1.37 times at 50 mm Hg), as did the overall volume of the vascular wall (6.86 times at 50 mm Hg), primarily as a result of arterial growth in the longitudinal (axial) direction. CONCLUSIONS: Remodeling of the radial artery segment of the uterine vasculature clearly occurs during gestation, resulting in vessels that are of a larger caliber and are longer and more distensible in both the axial and radial directions. (AM J OBSTET GYNECOL 1993;168:268-74.)
Key words: Uteroplacental arteries, arterial mechanics, pregnancy, rats, vascular growth and remodeling Large increases in uterine blood flow occur during gestation in every species, including human beings. 1-4 Although the actual cellular mechanisms remain undefined, dramatic remodeling of the afferent uterine vasFrom the Division of Research, Department of Obstetncs and GynecoLogy, Umverslty of Vermont College of Medicine. Supported by American Heart Association grant-in-aid 901261. Received for publtcation February 6,1992; revised JuLy 21,1992; accepted JuLy 24, 1992. Reprint requests: George OsoL, PhD, Department of Obstetncs and GynecoLogy, Given But/ding, Room C-213, University of Vermont College of Medicme, BurLington, VT 05405. 6/1/41191
268
culature during gestation is clearly a significant contributing factor. Although pregnancy-induced increases in uterine artery diameter, length, or both have been reported in several species, including sheep, pigs, rats, and guinea pigs,5-9 most previous studies have focused on the main uterine artery or on changes in the overall geometry of the uterine arcade. 6 -8 . 10 In animals with hemochcrial placentas such as rodents, primates, and human bei'lgs, most of the vascular resistance is localized in the radial arteries, II. 12 a segment of the vasculature about which little direct information is available. 0002-9378/93 $1.00 + 0.20
17.
18. 19.
20.
21.
and classification of enzymes. New York: Academic Press, 1991. Quantin B, Murphy G, Breathnach R. Pump-1 cDNA codes for a protein with characteristics similar to those of classical collagenase family members. Biochemistry 1989; 28:5327-34. Woessner jFj, Taplin CJ. Purification and properties of a small latent matrix metalloproteinase of the rat uterus. j Bioi Chern 1988;263:16918-25. Rentrop MB, Knapp B, Winter H, Schweizer j. Aminoalkylsilane-treated glass slides as support for in situ hybridization of keratin cDNAs to frozen tissue sections under varying fixation and pretreatment conditions. Histochemj 1986;18:271-6. McDonnell SM, Navre M, Coffey Rj, Matrisian LM. Expression and localization of the matrix metalloproteinase PUMP-1 (MMP-7) in human gastric and colon carcinomas. Mol Carcin 1991;4:527-33. Muller D, Quantin B, Gesnel Me, Millon CR, Abecassisj,
January 1993 Am J Obstet Gynecol
22.
23.
24.
25.
Breathnach R. The collagenase gene family in humans consists of at least four members. Biochem j 1988;253: 187-92. Melton D, Krieg P, Rebagliati M, Maniatis T, Zinn K, Green M. Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promotor. Nucl Acids Res 1984; 12:7035-56. Cox KH, DeLeon DM, Angerer LM, Angerer Re. Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetric RNA probes. Dev Bioi 1984;101: 485-502. Angerer LM, Stoler MH, Angerer RC. In: Valentino KL, Eberwine jH, Barchas jD, eds. In situ hybridization with RNA probes: an annotated recipe. New York: Oxford University Press, 1987:42-70. Cathala G, Savouret j-F, Mendez B, et al. A method for isolation of intact, translationally active ribonucleic acid. DNA 1983;2:329-35.
Hemodynamic and fluid responses to furosemide infusion in the ovine fetus Thomas F. Kelly, MD, Thomas R. Moore, MD, and Robert A. Brace, PhD
San Diego, California OBJECTIVES: The direct effects of furosemide infusion have not been systematically examined in the fetus. Our objective was to explore the hemodynamic and urinary responses to a 4-hour infusion of furosemide into fetal sheep. STUDY DESIGN: Furosemide (0, 1, 5, or 10 mg/hr) was infused into the fetal inferior vena cava for 4 hours in 15 chronically catheterized near-term sheep. RESULTS: Relative to vehicle infusion, furosemide produced dose-dependent Increases in fetal arterial pressure (analysis of variance, p < 0.05 when comparing groups), fetal heart rate (p < 0.0001), urine flow (p < 0.0001), and urine osmolality, sodium, and chloride, concentrations (p < 0.0001). Concomitantly, there were dose-dependent decreases in fetoplacental blood volume, fetal plasma chloride (p < 0.0001) and fetal venous pressure (p < 0.05). The changes in urine flow rate and fetal arterial pressure were positively correlated (r = 0.46, P < 0.01), suggesting that the diuresis was mediated in part by fetal arterial pressure. The four-quadrant amniotic fluid index increased during the furosemide infusions but not during vehicle infusions (p = 0.022). CONCLUSIONS: Furosemide infuSion caused a marked dose-dependent diuresis in the ovine fetus that appears to be partly mediated by increases in vascular pressure. Although amniotic fluid volume increases and fetal blood volume decreases, the reduction in blood volume was small compared with the urine volume excreted. (AM J OBSTEr GVNECOL 1993;168:260-8.)
Key words: Fetal, urinary output, amniotic fluid index, blood volume, sheep
From the Division of Perinatal Medicine, Department of ReproductIVe Medicine, University of California, San Diego. Supported by National Institutes ofHealth grant HD 23724 from the National Institute of Child Health and Human Development. Received for publication January 30, 1992; reVISed July 9, 1992; accepted July 16, 1992. Reprint requests: Robert A. Brace, PhD, Department of Reproductive Medicine, University of California, San Diego, LaJolla, CA 920930802. 6/1/41092
260
Furosemide is a diuretic that has been used in a variety of clinical situations during pregnancy, including evaluation of fetal urine production and human fetal intravascular transfusions. 1. 2 Although this drug has been studied in various adult settings, there are little data regarding the fetal responses. Prior studies reported that a bolus injection of furosemide into the ovine fetus produced increases in fetal urine produc0002-9378/93 $1.00
+ 0.20
Volume 168 Number I, Part 1
Fetal sheep responses to furosemide infusion
261
tion.'· 4 However, the mechanism of furosemide-induced diuresis, particularly the role of cardiovascular responses, has not been elucidated completely. Further, it is not known whether a continuous infusion of furosemide will produce a sustained diuresis in the fetus and a resultant expansion of amniotic fluid volume. Thus our study was designed to determine the fetal cardiovascular and urinary responses to a continuous infusion of furosemide.
Table I. Fetal and amniotic control values
Material and methods Animal preparation and maintenance. The protocol used in this study was approved by our institution's animal subjects committee. Fifteen time-dated, nearterm pregnant sheep were used. Their source, preparation, and maintenance have been previously described.' Gestational age at the time of surgery was 125 ± 1 (mean ± SE) days. In brief, the ewe was anesthetized with 1 gm intravenous thiopental sodium, and anesthesia was maintained with 1% to 2% halothane in oxygen. The maternal abdomen was prepared and draped, and a laparotomy was performed. The uterus was exposed and opened, and the fetal hind limbs were extracted. Vascular catheters were placed in the fetal inferior vena cava and descending aorta via the femoral vessels. The fetal urinary bladder was catheterized via a suprapubic incision. Additional catheters were attached to the fetal limbs for accessing the amniotic fluid. The hysterotomy and laparotomy incisions were closed, and the maternal inferior vena cava and descending aorta were catheterized. All catheters were tunneled subcutaneously and exited through a small incision in the left flank. The animals were maintained postoperatively for 5 days on prophylactic antibiotics. 5 All vascular catheters were flushed daily with heparin sodium 1000 Vlml (Elkins-Sinn, Cherry Hill, N.J.) to maintain patency. Experimentation began on postoperative day 5. At the time of experimentation the animals were given furosemide or control infusions randomly, and most had more than one dose given. No animal received the same dose twice. At least 48 hours elapsed between experiments on each animal. Thirty-two vehicle (normal saline solution, Abbott Laboratories, North Chicago) or furosemide infusions were performed in 15 animals. Experimental protocol. A I-hour control period was followed by a four-hour infusion of furosemide into the fetal vena cava at doses of 0, 1, 5, or to mgihr. These doses were chosen because of previous literature showing significant urinary effects in the neonate at 1 mg/kg.6 The infusion was maintained at 0.12 mVmin (Harvard Pump, Harvard Apparatus, South Natic, Mass.). Mter the infusion there was a I-hour recovery period. Experimental measures. Fetal arterial pressure,
Values are mean ± SEM; n = 32 for all, except n = 24 for amniotic fluid index.
Fetal arterial pressure (mm Hg) Fetal venous pressure (mm Hg) Fetal heart rate (beats/min) Fetal urine flow rate (mVmin) Fetal arterial pH Fetal arterial Pco2 (mm Hg) Fetal arterial P02 (mm Hg) Fetal hematocrit (%) Fetal plasma protein concentration (gm/dl) Amniotic fluid index (mm)
47.5 2.9 167 0.62 7.31 53.8 20.5 34.2 3.8 78
± ± ± ± ± ± ±
± ±
±
0.8 0.4 3 0.05 0.01 0.5 0.4 0.7 0.1 3
venous pressure, amniotic fluid pressure, heart rate, and urine flow rate were continuously recorded on a polygraph recorder (Gould, Cleveland), and 30-second averages were stored with an on-line computer (Hewlett-Packard, Eybens, France) as prevously described. 7 • 8 Fetal urine was continuously returned to the amniotic compartment. Vascular pressures were referenced to the standard zero pressure reference for the fetus by continuously subtracting amniotic fluid pressure with the computer. Samples of fetal arterial blood (1.5 ml), maternal arterial blood (1.0 ml), fetal urine (0.5 ml), and amniotic fluid (1.0 ml) were taken at 30 and 60 minutes into the control period and every 60 minutes thereafter. Fetal blood was replaced immediately with an equal volume of heparinized maternal blood collected before the study. Fetal blood was used for the measurement of pH, Pco 2 , P0 2 at 39S C (model 1306 pHlblood gas analyzer, Instrumentation Laboratory, Lexington, Mass.), hematocrit in triplicate,9 plasma osmolality by freezing point depression (Advanced Instruments Inc., Needham Heights, Mass.), protein concentration (TS meter, American Optical Scientific Instruments, Buffalo, N.Y.), and Na+, K+ and CI- concentrations (model 5 + 5 electrolyte analyzer, Nova, Waltham, Mass.). The changes in blood volume relative to control were derived by using the inverse of the hematocrit as previously described. 9 Osmolality and electrolytes were measured on maternal blood, fetal urine, and amniotic fluid. A four-quadrant amniotic fluid index was performed by using real-time ultrasonography (3.5 MHz linear-array scanner, Siemens, Germany) in 24 of the 32 experiments. lo The amniotic fluid index measurement was not always successful because of an occasional uncooperative animal. Animals were killed with a pentobarbital sodium euthanasia solution (100 mg/kg) after completion of the experiments. Fetuses were towel dried and weighed. Assuming a 3% fetal weight gain per day, II weights at the time of experimentation were calculated for those animals not killed immediately after the last experiment. Statistical analysis. Experiments were divided into
262
Kelly, Moore, and Brace
January 1993 Am J Obstet Gynecol
110
D
.009
.............. ............. ....... ........
100
. ....... , ..... .
).)
'I-l
o
110
100hllo;.,i.....~
110
100
o
-1
1
2
3
4
5
TIME (hours) Fig. 1. Fetal arterial pressure responses to different infusions of furosemide (A, vehicle; B, 1; C, 5, and D, 10 mglhr; n = S for each dose) over 4 hours. Data expressed as mean ± SEM percent change from values averaged during I-hour preinfusion period. Shaded area, 95% Confidence interval of control values. Vertical dotted lines, Beginning and end of 4-hour infusion.
Table II. Fetal plasma, urine, amniotic fluid and maternal plasma control values Osmolality (mOsm/kg) Fetal plasma Maternal plasma Amniotic fluid Fetal urine
302 301 273 141
± 1 ± 1 ± 7 ± 7
Sodium
(mEq/L)
13S.9 144.3 115.0 32.2
± 0.3 0.3 ± 2.7 ± 3.2 ±
Potassium (mEq/L)
4.4 ± 0.1
4.3 ± 0.0
7.1 ± 0.6 7.0 ± O.S
Chloride (mEq/L)
10S.7 114.0 9S.5 30.7
± 0.4 ± 0.3 ± 2.4 ± 3.1
Values are mean ± SEM; n = 32.
four groups, with each group receiving either 1,5, or 10 mglhr infusions of furosemide or vehicle. Data are expressed as mean ± SEM, and are plotted as either a change from control or as a percentage of control to reduce interanimal variability. Control values were taken as the mean value during the 60-minute preinfusion period. These values were analyzed for statistical differences among the four treatment groups by using a three-way analysis of variance, with time, treatment, and animal being the three factors. Changes with time in each treatment group were analyzed with a two-way analysis of variance with repeated measures. Assessment
of significant differences in the responses to the four infusions were derived from the interaction term of a three-way analysis of variance. Statistical significance was taken as p ::s; 0.05. To determine whether a significant relationship existed between variables, linear regression was used on the average values from each animal during the 4-hour infusion period. Results
The mean fetal age at experimentation was 134 ± 1 days; mean fetal weight was 3.03 ± 0.10 kg. Furosemide was infused at 0.35 ± 0.02 (n = 8), 1.83 ± 0.11
Fetal sheep responses to furosemide infusion
Volume 168 Number I, Part 1
263
: p < 0.05
1
a
-1
-1
o
2
1
3
4
5
TIME (hours) Fig. 2. Fetal venous pressure changes in response to either vehicle and I mglhr (n = 16, dots) or 5 and 10 mglhr (n = 16, sqw.res) infusions from 0 to 4 hours. Data expressed as mean ± SEM changes from values during I-hour preinfusion period. VertICal dotted lmes, Beginning and end of infusion; shaded area, 95% confidence interval of control values.
(n = 8), or 3.37 ± 0.32 (n = 8) mglkglkr in the 1, 5, and 10 mglhr groups, respectively. Control values for the maternal and fetal variables during the I-hour preinfusion period are listed in Tables I and II. There were no significant differences during the control period among the four groups except for a small difference in fetal plasma sodium in which the vehicle group was significantly lower than the three groups that received furosemide (137.6 ± 0.4 vs 138.9 ± 0.6 mEqlL, respectively p = 0.03). Fetal cardiovascular effects. Fetal arterial pressure increased significantly during the 5 and 10 mg/hr infusions (p < 0.05 when comparing groups, Fig. 1). During the 10 mglhr infusion, the peak in arterial pressure occurred earlier and was more sustained (p < 0.009) than the peaks associated with the I or 5 mglhr infusions. Fetal venous pressure was unchanged during the vehicle or 1 mglhr furosemide infusions but decreased significantly during the 5 and 10 mglhr infusions (p < 0.05, Fig. 2), with a reduction approaching 1.0 mID Hg during the last 2 hours of the infusion. There were significant differences in the fetal heart rate responses to vehicle versus furosemide infusion. Fetal heart rate fell significantly during the vehicle infusion from a control mean of 167 ± 15 to 141 ± 5 beats/min (p < 0.00001) by the end of the infusion. In the groups receiving furosemide there was a significant
dose-dependent effect (p < 0.0001), with mean heart rate changing by - 5, - 3, and + 8 beats/min during the last 2 hours of the 1, 5, and 10 mg/hr infusions, respectively. Fetal blood volume effects. Marked dose-related decreases in fetal blood volume were significant in all groups receiving furosemide compared with vehicle infusions (p < 0.00001, Fig. 3 ). Blood volume decreased by 3.6% ± 0.9%, 7.3% ± 1.7%, and 8.9% ± 1.3% in the 1, 5, and 10 mglhr groups, respectively, representing an average blood volume loss of 12, 23, and 31 mI.12 Concomitantly, fetal plasma protein concentration increased significantly in all groups receiving furosemide (p < 0.00001) by 7.1% ± 1.3%, 11.0% ± 2.2%, and 15.7% ± 2.6% at 4 hours into the infusion of 1,5, and 10 mglhr of furosemide. Fetal urine flow effects. Urine flow rose significantly above control in all groups receiving furosemide (p < 0.00001, Fig. 4). Overall, the average increases in urine flow in the 1, 5, and 10 mg/hr groups during the 4-hour infusion period were 0.6 ± 0.2, 0.9 ± 0.2, and 1.1 ± 0.2 mllmin, respectively, above control, corresponding to excess excretion volumes of 141, 221, and 252 ml. During the 1 and 5 mglhr infusions the peak flow occurred at 120 minutes into the infusion, whereas the peak occurred at 60 minutes during the 10 mg/hr infusion. The urine flow change exhibited a significant linear
264
Kelly, Moore, and Brace
January 1993 Am J Obstet Gynecol
105
>-I
95
+I
90
0 1-1
c::
0
u
85
0
105
.....
.:P
D
100
'" ............ " ........ .01 ..........
...T
'I' ---- ----- ____ .1
T
.i.
T
:P
c
d(>
100
< 0.00001
'!' .....01"'.~"" .... ~ ... Jo:"'....... '" 10:10.......':.. 10.10.. . . . .1"Jo: 10.• • •:~..r"'Io.Io. ••• ~
< 0.00001
rz:I
~
95
~
90
~
§ 0
~
I:Q
85 105
~ rz:I ~
100 95 105 100 95
: P < 0.003
B
~
. .-
............I.;.:··:.:.··cc,.·~ ... ,.,..oj>·"'·" ...... "'... .,..,.,. ..'. . ,."'...... ·,.'Iw,..j.:... •• ...
...-
rr ::1~;~. ~. A
:
H
-1
0
~
..
:tt:~m''T.::h;",,*:*..t*:"".'_ _j";- - "..: .
.w.'' l
•
t+ . . . . ·!"'...... ~ .......
1
2
3
4
5
TIME (hours) Fig. 3. Fetal blood volume responses to infusion of vehicle (A, n = 8) or furosemide infusions (B, 1; C, 5; D, 10 mglhr; n = 8 for each group) from 0 to 4 hours. Data expressed as mean ± SEM percent of mean value during I-hour preinfusion period. Shaded area, 95% Confidence interval of control values; vertical dotted lines, beginning and end of 4-hour infusion.
correlation with the fetal arterial pressure change in all animals during the furosemide infusions (r = 0.46, P < 0.01). Fetal urine composition effects. The diuresis was associated with dose-dependent increases in urinary osmolality to 58 ± II, 78 ± 12, and 103 ± 12 mOsm/kg above control in the groups receiving furosemide at 1, 5, and 10 mglhr (p < 0.00001). Free water clearance during the vehicle infusion rose significantly in all groups receiving furosemide (Fig. 5). The peak free water clearance was 51% ± 27%, 139% ± 38%, and 66% ± 37% above control in the 1, 5, and 10 mglhr groups, respectively. Urinary sodium and chloride concentrations increased significantly during the diuresis (p < 0.00001). The sodium excretion rate increased in all furosemide groups in a dose-dependent fashion (98 ± 32, 150 ± 23, and 224 ± 33 ILEqlmin). The pattern of urinary chloride excretion was similar.
Fetal blood and plasma effects. There were no significant differences in the fetal plasma osmolality responses to vehicle or furosemide infusion. The urinary losses of electrolytes were reflected in significant decreases in fetal plasma chloride when comparing groups (p < 0.00001). At the 0, 1, 5, and 10 mglhr doses, changes averaged +0.5% ± 0.5%, -1.4% ± 0.5%, -3.0% ± 0.4%, and -3.7% ± 0.4%, respectively. Fetal plasma potassium did not change significantly. Fetal plasma sodium remained unchanged during the vehicle and 1 and 5 mglhr infusions but fell during the 10 mglhr infusion by 1.0% ± 0.2% (p = 0.002). Fetal pH and P0 2 did not change during vehicle or furosemide infusions. Fetal Pco2 increased significantly during the 5 and 10 mglhr infusions when compared with the vehicle (p < 0.003), with peaks of 2.1 ± 0.6 and 2.6 ± 0.9 mm Hg occurring respectively in the postinfusion period. Amniotic fluid effects. The amniotic fluid index was
Volume 168 Number I , Part 1
Fetal sheep responses to furosemide infusion
265
2
:p
D
< 0.00001
1
c
1
< 0.00001
a hw.......~··· · ·················· ·· ·············.· .. ..... : . ... ... .. .. .• . B
1
1
~
A
a : •••• -1
< 0.0001
t.....:•••••:.....~.....l.·,..'.:·· ::
a
:
1
2
3
4
NS
5
TIME (hours) Fig. 4. Fetal urine flow rate responses to infusion of vehicle (A, n = 8) or furosemide infusions (B, 1; C, 5; D, 10 mglhr; n = 8 for each group) from 0 to 4 hours. Data expressed as mean ± SEM change from values during I-hour control period. S1uuJ.ed area, 95% Confidence interval of control values.
unchanged during the vehicle infusion and increased in the furosemide-treated groups (p = 0.022). During the I, 5, and lO mglhr infusions, the respective increases were 23% ± 3% (p = 0.016), 15% ± 5% (p = 0.013), and 44% ± 13% (p = 0.005) above control. Maternal plasma effects. There were no differences among groups in the maternal plasma sodium or osmolality in response to the infusions. Maternal potassium underwent dose-dependent decreases with significant differences among treatment groups (p < 0.00001), with the changes being 0.1 ± 0.1, -0.1 ± 0.1, -0.3 ± 0.1, and -0.6 ± 0.1 mEqlL at 4 hours into the furosemide infusion in the 0, 1,5, and 10 mglhr groups, respectively. The maternal plasma chloride showed significant differences : among groups (p = 0.0002) with a maximal decrease of 2.6 mEqlL during the 10 mg/hr infusion when compared with the vehicle.
Comment Our study suggests that furosemide infusion is associated with dose-dependent cardiovascular and renal
effects in the ovine fetus; these effects lead to increases in amniotic fluid volume. The cardiovascular responses included elevations in fetal arterial pressure, which were significant during the 5 and 10 mglhr infusions. These elevations may reflect activation of the renin-angiotensin system, which has been previously shown in the fetus and adult", 13. 14 Lumbers and Stevens' noted that when a bolus of furosemide was given to fetuses made hypotensive with nitroprusside there was a significant rise in diastolic pressure. They also demonstrated that saralasin, a competitive antagonist of angiotensin II, inhibited the rise in blood pressure associated with furosemide. Studies of human beings have shown peripheral vascular responses to furosemide when given as a bolus, and these were related to renin release and elevated angiotensin II concentrations. 15 Furthermore, increases in renin activity associated with furosemide have been shown to be inhibited by indomethacin, suggesting prostaglandins are involved. 16 Thus the elevation in fetal arterial pressure response to furosemide was likely caused by the activation of the renin-angiotensin system, possibly mediated by prostaglandins.
266
Kelly, Moore, and Brace
January 1993 Am J Obstet Gynecol
200 150 r-I
o J..4
.j.J
c o u
100 50
I I • I I I I
L..-_ _....&...._ _---L_ _......JL..-_ _-'--_ _- ' -_ _- - ' _
~
o 250 200 150 100
:::~····~········1·········i···········,
150 100
-1
«.~
••-
+. . . . -..-.. . ~«.,««.<.
••-
........... , ..
o
1
2
3
4
5
TIME (hours) Fig. 5. Free water clearance responses to vehicle 0, n = 8) or furosemide infusions (B, 1; C, 5; D, 10 mglhr; n = 8 for each group) from 0 to 4 hours. Data expressed as mean ± SEM as percent of preinfusion period. Shaded area, 95% Confidence interval of preinfusion mean.
The drop in fetal heart rate (FHR) during the vehicle infusion reflects the nonnal diurnal variation we have previously reported. 17 The FHR in the groups receiving furosemide differed significantly from that in the vehicle infusion group in that furosemide prevented the decrease with time that occurred during the vehicle infusion. These relative increases in FHR in the animals receiving furosemide may represent a direct effect of angiotensin II on the fetal heart as previously shown. 18 Fetal venous pressure decreased in the 5 and 10 mglhr groups at the end of the infusion period. This decrease is consistent with the finding of decreased right atrial pressure noted in the adult human by Dikshit et al. 19 in response to furosemide and increased venous capacitance noted by Johnston et al. 20 They concluded that furosemide acted by dilating capacitance vessels as well by its tubular effects in the kidney. This effect of furosemide on vascular smooth muscle has also been observed in the rabbit, where there has been a decrease in resting tension. 21 The prolonged diuresis caused by furosemide also would tend to de-
crease fetal venous pressure resulting from decreases in intravascular volume. Fetal blood volume decreased in a dose-dependent manner. However, only a small fraction of the excess urine produced during furosemide infusion can be accounted for on the basis of blood volume loss. In the 10 mglhr group, the increased urine produced was 252 ml, and the blood volume decrease was 31. 2 ml. Therefore approximately 88% of the excess fluid lost through the kidneys must have been recruited from other sources. The fetal extravascular space was the likely source, although gastrointestinal and intramembranous absorption22 and transplacental flow also may have contributed. In addition to the cardiovascular observations there were several new findings regarding fetal urine flow. We found that a dose-dependent effect on urine flow appeared initially. By the end of the infusion period the urine flow rate in all groups receiving furosemide was similar, possibly because of either saturation of the tubular transport mechanism for furosemide or satura-
Volume 168 Number I. Part 1
tion at the site of action of furosemide in the proximal tubule lumen. Also, as the fetus becomes volume depleted furosemide may not be as effective in promoting further increases in free water clearance. Free water clearance may be dependent on the increased concentration of plasma proteins, thereby increasing protein osmotic pressure. This theory is consistent with our observation of return of free water clearance to normal despite the continuous furosemide infusion. The observed correlation between fetal arterial pressure and urine flow suggests that approximately 20% of the diuresis associated with furosemide infusion may be pressure related. Alternatively, the juxtaglomerular apparatus may sense a decrease in circulating volume caused by furosemide and increase renin output, resulting in an increase in fetal arterial pressure. The amniotic fluid index increased dose dependently in all groups receiving the furosemide, but not in the vehicle, which most likely reflects the increased urine output caused by furosemide, but changes in swallowing and lung secretion and intramembranous flow may have contributed. There were, however, no consistent changes in the amniotic fluid osmolality or electrolytes except at the highest furosemide infusion rate, when increases in sodium and chloride concentrations occurred. These increases are surprising in view of the large volumes of urine entering the amniotic space and suggest a large, yet undefined capacity to maintain amniotic composition at near-normal levels. We speculate that this is largely the result of the exchange of water and electrolytes across the intramembranous pathway.22 Maternal potassium and chloride changes appear to reflect the transplacental effect of furosemide. Furosemide has been shown to cross the human placenta!' hence the rationale for the "Lasix challenge test. "24 However, in subsequent sheep studies Chamberlain et a1.' used a furosemide assay sensitive to 100 ng/ml to show that there was no transfer of furosemide from the maternal to the fetal intravascular compartments. Similarly, Ross et a1. 28 found no evidence of transplacental effects of furosemide in the sheep. Although furosemide may not cross the placenta from the fetal to the maternal compartment, there are clear maternal effects that may reflect the maintenance of the concentration gradients of potassium and chloride across the placenta. For example, the fall in maternal chloride may have been the consequence of the reduction in fetal plasma chloride concentration. In summary, furosemide elicits a strong diuretic response in the ovine fetus and is associated with important dose-dependent cardiovascular effects. Fetal arterial pressure increases with higher doses, which may reflect activation of the renin-angiotensin system via prostaglandins. The fetal blood volume and venous
Fetal sheep responses to furosemide infusion
267
pressure falls significantly in animals receiving the diuretic. Amniotic fluid volume increases in response to the increased urine output, and changes in the maternal plasma probably reflect the maintenance of concentration gradients across the placenta. We thank Eddie Kwan for his technical assistance.
REFERENCES
1. Barrett R], Rayburn WF, Barr M. Furosemide (Lasix) challenge test in assessing bilateral fetal hydronephrosis. AM] OBSTET GYNECOL 1983;147:846-7. 2. Chestnut DH, Pollack KL, Weiner CP, Robillard ]E, Thompson CS, DeBruyn CS. Does furosemide alter the hemodynamic response to rapid intravascular transfusion of the anemic fetal lamb? AM] OBSTET GYNECOL 1989;161: 1571-5. 3. Chamberlain PF, Cumming M, Torchia MG, Biehl D, Manning FA. Ovine fetal urine production following maternal intravenous furosemide administration. AM] OBSTET GVNECOL 1985;151:815-9. 4. Lumbers ER, Stevens AD. The effects offrusemide, saralasin and hypotension on fetal plasma renin activity and on fetal renal function.] Physiol (Lond) 1987;393:479-90. 5. Brace RA, Cheung CY. Fetal cardiovascular and endocrine responses to prolonged fetal hemorrhage. Am ] Physiol 1986;251 :R417-24. 6. Woo WC, Dupont C, Collinge ], Aranda ]V. Effects of furosemide in the newborn. Clin Pharmacol Ther 1978; 23:266-71. 7. Brace RA, Brittingham DS. Fetal vascular pressure and heart rate responses to nonlabor uterine contractions. Am ] Physiol 1986;251:R409-16. 8. Brace RA, Miner LK, Siderowf AD, Cheung CY. Fetal and adult urine flow and ANF responses to vascular volume loading. Am] Physiol 1988;255:R846-50. 9. Brace RA. Blood volume and its measurement in the chronically catheterized sheep fetus. Am] Physiol 1983; 244:H487-94. 10. Moore TR, Brace RA. Amniotic fluid index (AFI) in the term ovine pregnancy: a predictable relationship between AFI and amniotic fluid volume [Abstract 286]. In: Proceedings of the thirty-sixth annual meeting of the Society for Gynecologic Investigation, San Diego, California, March 15-18, 1989. 11. Gresham EL, Rankin ]HG. Makowski EL, Meschia G. Battaglia FC. An evaluation of fetal renal functIOn in chronic sheep preparation.] Clin Invest 1972;51:149-55. 12. Brace RA. Fetal blood volume, extracellular fluid and lymphatic function. In: Brace RA, Ross MG, Robillard ]E, eds. Reproductive and perinatal medicine. Volume 11: Fetal and neonatal body fluids: the scientific basis for clinical practice. New York: Perinatology Press, 1989:1-18. 13. Siegel SR, Leake RD, Weitzman RE, Fisher DA. Effects of furosemide and acute salt loading on vasopressin and renin secretion in the fetal lamb. Pediatr Res 1980; 14:86971. 14. Corsini WA, Hook ]B, Bailie MD. Control of renin secretion in the dog. Effects of furosemide on the vascular and macula densa receptors. Circ Res 1975;224:425-30. 15. Passmore AP, Whitehead EM, Johnston GD. Comparison of the acute renal and peripheral vascular responses to frusemide and bumetanide at low and high dose. Br] Clin Ph..rmacol 1989;27:305-12. 16. Mackay]G, Muir AL, Watson ML. Contribution of prostaglandins to the systemic and renal vascular response to fiusemide in normal man. Br] Clin Pharmacol 1984; 17: 513-9. 17. Lawler FH, Brace RA. Fetal and maternal arterial pres-
Kelly, Moore, and Brace
18. 19.
20. 21.
sures and heart rates: histograms, correlations and rhythms. Am J Physiol 1982;243:R433-44. Iwamoto HS, Rudolph AM. Effects of angiotensin II on the blood flow and its distribution in fetal lambs. Circ Res 1982;48: 183-9. Dikshit K, VydenJK, Forrester JS, ChatteIjee MB, Prakash R, Swan HJC. Renal and extrarenal hemodynamic effects of furosemide in congestive heart failure after acute myocardial infarction. N Engl J Med 1973;238: 1087-90. Johnston GD, Nicholls DP, Kondowe GB, Finch MB. Comparison of the acute vascular effects of frusemide and bumetanide. Br J Clin Pharmacol 1986;21:359-64. Andreasen F, Christensen JH. The effect of furosemide on
January 1993 Am J Obstet Gynecol
22. 23. 24. 25.
vascular smooth muscle is influenced by plasma protein. Pharmacol Toxicol 1988;63:324-6. Gilbert WM, Brace RA. The missing link in amniotic fluid volume regulation: intramembranous absorption. Obstet Gynecol 1989;74:748-53. Beermann B, Groschinsky-Grind M, Fahraeus L, Lindstrom B. Placental transfer of furosemide. Clin Pharmacol Ther 1978;24:560-2. Wladimiroff JW. Effect of frusemide on fetal urine production. Br J Obstet GynaecoI1975;82:221-4. Ross MG, Ervin MG, Leake RD. Failure of Lasix to induce a fetal diuresis. AM J OBSTET GYNECOL 1985; 152:1107.
Pregnancy-induced changes in the three-dimensional mechanical properties of pressurized rat uteroplacental (radial) arteries George Osol, PhD, and Marilyn Cipolla, BS Burlington, Vermont OBJECTIVE: The purpose of this study was to describe the effects of pregnancy on the size and three-dimensional mechanics of uterine radial arteries. STUDY DESIGN: Measurements of lumen diameter, wall thickness, and axial and radial distensibility were made in situ and in pressurized segments of excised vessels from nonpregnant (n = 29) and late-pregnant (days 19 to 21, n = 19) Sprague-Dawley rats. RESULTS: In the unpressurized state, the overall length of the radial artery segment of the arcade increased approximately 4.8 times during gestation. Lumen diameter increased by 60%, as did distensibility in both the radial and axial directions. However, there was no measurable change in the thickness of the vascular wall, which was muscular in appearance and comprised approximately two layers of circumferentially oriented smooth muscle, a relatively thick internal elastic lamina, and a well-defined endothelial layer. Cross-sectional area increased significantly during pregnancy (1.37 times at 50 mm Hg), as did the overall volume of the vascular wall (6.86 times at 50 mm Hg), primarily as a result of arterial growth in the longitudinal (axial) direction. CONCLUSIONS: Remodeling of the radial artery segment of the uterine vasculature clearly occurs during gestation, resulting in vessels that are of a larger caliber and are longer and more distensible in both the axial and radial directions. (AM J OBSTET GYNECOL 1993;168:268-74.)
Key words: Uteroplacental arteries, arterial mechanics, pregnancy, rats, vascular growth and remodeling Large increases in uterine blood flow occur during gestation in every species, including human beings. 1-4 Although the actual cellular mechanisms remain undefined, dramatic remodeling of the afferent uterine vasFrom the Division of Research, Department of Obstetncs and GynecoLogy, Umverslty of Vermont College of Medicine. Supported by American Heart Association grant-in-aid 901261. Received for publtcation February 6,1992; revised JuLy 21,1992; accepted JuLy 24, 1992. Reprint requests: George OsoL, PhD, Department of Obstetncs and GynecoLogy, Given But/ding, Room C-213, University of Vermont College of Medicme, BurLington, VT 05405. 6/1/41191
268
culature during gestation is clearly a significant contributing factor. Although pregnancy-induced increases in uterine artery diameter, length, or both have been reported in several species, including sheep, pigs, rats, and guinea pigs,5-9 most previous studies have focused on the main uterine artery or on changes in the overall geometry of the uterine arcade. 6 -8 . 10 In animals with hemochcrial placentas such as rodents, primates, and human bei'lgs, most of the vascular resistance is localized in the radial arteries, II. 12 a segment of the vasculature about which little direct information is available. 0002-9378/93 $1.00 + 0.20