- Email: [email protected]
Validation of thermal techniques for measurement of pelvic organ blood flows in the nonpregnant sheep: Comparison with transit-time ultrasonic and microsphere measurements of blood flow
Validation of thermal techniques for measurement of pel vie organ blood flows in the nonpregnant sheep: Comparison with transit-time ultrasonic and microsphere measurements of blood flow Nigel J. Randall, PhD, Richard W. Beard, MD, Ian A. Sutherland, PhD, Jorge P. Figueroa, MD, Cor J. Drost, PhD, and Peter W. Nathanielsz, MD, PhD London, England, and Ithaca, New York Data obtained from a thermal system capable of measuring changes in organ temperature as well as tissue thermal clearance in the uterus and vagina have been compared with blood flow measured continuously with a transit-time ultrasound volume-flow sensor placed around the common internal iliac artery and intermittently with radioactive microspheres in the chronically instrumented nonpregnant sheep. Temperature changes in both the uterus and the vagina correlated well with blood flow changes measured by both techniques after intravenous administration of estradiol or norepinephrine. Thermal clearance did not correlate well with blood flow in the vagina or uterus. These methods may have value in the investigation of blood flow patterns in various clinical situations such as the pelvic pain syndrome and early pregnancy. (AM J 0BSTET GYNECOL 1988;158:651-8.)
Key words: Pelvic blood flow, thermal techniques, estrogens and blood flow The use of a thermal system to measure blood flow in a perfused tissue has several potential advantages. Thermal probes are simple to construct and noninvasive to the surrounding tissue and blood flow can be measured continuously or by repetitive sampling. Although interest in the technique has spanned more than 50 years, the method presents several problems that so far have prevented its practical clinical use. We propose that this method could form the basis for a low-cost ambulatory system for monitoring significant features of pelvic blood flow in the human. In the present experiments we used the chronically instrumented nonpregnant sheep model to validate these thermal methods to measure pelvic blood flow changes. Thermal monitoring at the uterine and vaginal walls was compared with continuous ultrasound and intermittent microsphere flow techniques to determine the degree of correspondence of the three methods at the two pelvic sites. From the Department of Obstetrics and Gynaecology, St. Mary's Hospital Medical School, London, and the Laboratory for Pregnancy and Newborn Research, New York State College of Veterinary Medicine, Cornell University. This work was supported by Grant PCM-8303035 from the National Science Foundation and by the Harris-Birthright Fund for Research into Early Pregnancy. Received for publication December 17, 1986; revised July 27, 1987; accepted October 28, 1987. Reprint requests: Dr. Peter W. Nathanie/sz, MD, PhD, Laboratory for Pregnancy and Newborn Research, New York State College of Veterinary Medicine, 526 Veterinary Research Tower, Cornell University, Ithaca, NY 14853.
Material and methods Surgical procedures. Four nonpregnant Rambouillet-Columbia ewes were premedicated with ketamine hydrochloride (20 mg · Kg- 1) and glycopyrrolate (10 µg · kg- 1). Operation was conducted with the animal under halothane anesthesia. Catheters were inserted into the left ventricle and the femoral artery. In one sheep a thermistor was placed in the dorsal aorta to monitor aortic blood temperature during flow changes. The uterus and pelvic blood vessels were exposed via a midline abdominal incision. A twin thermistor probe was inserted into each uterine horn near the tip. The sheep were allowed to recover for at least 4 hours after operation. Continuous measurement of uterine blood flow. A 6 mm transit-time ultrasonic flow probe (Transonics Systems, Inc., Ithaca, N.Y.) 2 ·• was placed around the common inernal iliac artery.2 Blood flow measured through this probe represents supply to the uterine, vaginal, vesical, and other pelvic vascular beds. In any one sheep mean common iliac blood How averaged over I-minute intervals showed a peak-to-peak variability of only ± 10 ml· min- 1 • Thermistor probes. The uterine thermistor probes contained two 2.4-mm diameter precision thermistors (tolerance ± 0.2° C) (type ACC-001, Ametek, Rodan Division, 2905 Blue Star St., Anaheim, CA 92806) attached to a short length of polyvinyl chloride catheter. One thermistor was mounted at the catheter tip and the second placed into its side, 15 mm further back. 651
652
Randall et al.
March 1988
Am
J Obstet Gynecol
700
600
-
500
I. c
.E
:s:::
Whole Organ Flow by Micro-spheres
400 300
0
u:::: 200
100 0
I 2
0
3
Time (hours) Fig. l. Comparison of common iliac artery flow measured by ultrasonic flowmeter (recorded continuously and hence plotted over the full range of time with calibration bar at 600 ml · min - i shown) and total flow as determined by microspheres during the estradiol stimulations. The envelope of the ultrasound flow is ±SD while the range is shown for the microsphere results.
Common Iliac Artery Flow ml.min-I
AT~ Temp.
1
Al 1 Uterine
~Thermal
Conductance
AU~=
Conductance
0
2
3
4
5
Time After Estracliol Admiistration (hrs) Fig. 2. Changes in common internal iliac artery flow (upper trace), intrauterine tissue temperature (second trace), vaginal tissue temperature (third trace), uterine thermal conductance (fourth trace), and vaginal thermal conductance (bottom trace) after 50 µg of 1713-estradiol administered intravenously to a non pregnant ewe. Calibration for common iliac artery blood flow is 700 ml · min - 1• All other vertical scales are arbitrary.
We used two designs of vaginal probe inserted into the vagina to rest against the cervix, with the thermistors in contact with the vaginal wall. One used 2.4-mm diameter thermistors mounted on the rim of a 5 cm poly-
vinyl chloride pessary 15 mm apart. The other had two probes of the uterine type, but with single thermistors at the tip, mounted in a foam sponge 30 mm in diameter and 30 mm long. The thermistor tips stood out
Volume 158 Number 3, Pan I
Thermal technique to measure pelvic blood flow
653
Table I. Effect of estradiol administration (50 µg via jugular vein) on regional blood flow with radioactive microspheres Time of isotope injection I= 0 hr (ml· min- 1)
Mean
Tissue flow Endometrium Cervix Vagina Vulva Right mesothelii.im, tube Left mesothelium, tube Right horn, body Left horn, body Whole organ Tissue weight (gm) Endometrium Cervix Vagina Vulva Right mesothelium, tube Left mesothelium, tube Uterus, both horns Total
7.6 4.8 7.7 2.4 1.6 1.6 4.1 3.7 33 17.6 13.3 26.8 20.4 15.1 15.7 40.2 149
I
Mean fold increase
t = 1 hr (ml· min- 1)
Range
Mean
6-11 2-10 3-15 1-6 1-3 1-2 3-6 3-4 22-53
I
t = 2 hr
Range
Mean
109 48 52 6.6 6.7 7.2 45 48
101-117 28-57 48-56 5-8 4-9 6-9 41-49 46-50
156 67 57 15 12
323
288-358
I
For each sheep (2 hr/Ohr)
I
Range
Mean
67 65
129-205 64-68 51-66 12-16 10-14 10-13 56-75 55-76
22 23 12 8 8 8 18 18
12-31 7-38 5-22 4-12 6-12 5-9 13-25 14-25
450
387-533
15
8-20
II
Range
13.4-24.4 11.2-16.6 21.0-35.0 13.5-25.0 12.3-18.4 14.1-16.6 31.8-46.6 135-156
from the surface of the foam. This latter design proved much easier to insert and was mechanically more stable within the vagina. The foam, with its reduced thermal mass, gave better heat transfer characteristics and all data reported in this paper were obtained with this type of probe. To prevent ingress of body fluids between the thermistor and probe body containing the electrical connections, probes were dip-coated in a latex solution whose sticky film was cured with a dusting of powder. The thermistor probes were connected to an isothermal control unit. One of a pair of thermistors operated in a self-heating mode while the other monitored tissue temperature. The control unit maintained the heated thermistor core temperature at a constant increment above tissue temperature and provided outputs of tissue temperature, heated thermistor temperature, and the driving voltage to this thermistor for calculation of dissipated power. Data were fed into an Apple II-based data acquisition system.' Absolute thermal conductance of the heated probe in vivo was calculated while probe temperature was increased up to a 3° C increment. No significant effects on thermal conductance with probe temperature were observed. As reported by other authors there were variations in absolute thermal conductance between probes (4 to 10.5 mW· 0 c- 1) that were also observed in water-bath experiments. Thus, absolute values are meaningless. To obtain absolute or relative measurements, probes would need to be thermally identical and very accurately calibrated in tissue of the type to be studied. Thermistors were operated at a temperature elevation of 2.0° C.
The surface temperature of the encapsulating material of the thermistors is less than this." Microsphere measurement of regional blood flow to pelvic organs. In all four sheep regional blood flows were measured with radioactive microspheres 5· 6 before and after the administration of I 713-estradiol. Using similar microsphere-based measurement techniques, Rosenfeld et al. 7 found that flow in the uterine arteries accounted for approximately 80% of the total uterine blood flow while Fuller et al. 2 found this to vary between 66% and 81 % in pregnant sheep, depending on the period of gestation. All radioactive microsphere studies with tin 113, gadolinium 153, and scandium 46 were carried out during the final 1713-estradiol stimulation. Pharmacologic manipulation of pelvic blood flow. Intravenous bolus injection of 1713-estradiol (50 µg) was administered to induce increases in uterine blood flow in 14 experiments. Repeat injections of 1713-estradiol were given only after uterine blood flow had returned to basal levels. On eight other occasions, twice to each animal, injections of norepinephrine (I 00 µg) were given to produce a temporary flow decrease. Uterine blood flow had always returned to basal levels between experiments. The regional organ blood flows were measured at three time points: just before 1713-estradiol injection;· I hour after 1713-estradiol injection, when flow in the common iliac had approximately doubled; and at 2 to 2 1/2 hours after the injection, when the blood flow was at its peak. In one sheep readings were taken at 2 and 5 hours on the downward slope. Analytical methods. For the 1713-estradiol studies the
654
Randall et al.
March 1988
Am
J Obstet Gynecol
(a) 2 Hours
-''•'"'''····
..
1
T µ 1-
(b) 5 Hours
..................-.: ..·.; . _.. ,- ....... :~~==~=:·-:-·
l T
._;~~~·f"~,~~~~:r~:&rf~i~;; .·. . 1
200
400
600
Common Iliac Flow (ultrasound) ml. min.-1 Fig. 3. a, Scattergram of temperature change measured by the intrauterine probe as a function of common internal iliac blood flow in the first 2 hours after administration of 50 µg of 1713-estradiol intravenously to a nonpregnant ewe; b, the same scattergram but including all data points for 5 hours after administration of 50 µg of 1713-estradiol intravenously to a nonpregnant ewe. Arrows mark calibration of 0.1° C.
data acquisition system logged the mean of each output channel every 8 seconds. In order to follow the more rapid changes caused by norepinephrine the outputs were sampled every 2 se.conds. Correlation was performed by linear regression analysis. The slopes and linear correlation coefficients (r) from all the 1713estradiol studies in the four sheep were averaged, and each is presented as a mean ± SD unless otherwise stated. The number of experiments, n, varied from 10 to 12, depending on the site of measurement. In the microsphere experiments the regional blood flows represent the mean value from three separate sheep.
Results Blood flow measiirements with microspheres. There was a threefold to fourfold increase in blood flow through the common internal iliac artery during an 1713-estradiol stimulation as measured by the ultrasonic flowmeter (Fig. 1). Table I lists the mean and range of
regional flows to eight different sites. Basal blood flows were commonly restored after 10 hours. As demonstrated in Fig. 1, the total common iliac flow exceeded the total flow in the eight tissues measured by the microspheres (Table I) by approximately 150 ml · min- 1 both at rest and during 1713-estradiol stimulation. This difference reflects distribution of blood in the common iliac artery to tissues such as the bladder and the muscles of the pelvis, which were not counted for radioactivity. During stimulation by 1713-estradiol there is a redistribution of relative flow with a larger proportion of internal iliac blood flow going to the endometrium and cervix. Fig. 1 suggests that the blood flow to areas other than those in which regional blood flow was measured was not significantly altered by 1713-estradiol administration. Thermal measurements during blood flow changes. Fig. 2 shows a typical result of changes in thermal conductance (ilU) and tissue temperature (ilT) in the
Thermal technique to measure pelvic blood flow
Volume 158 Number 3, Part I
655
(a) 2 Hours
J_
-,
"j
u
~
~
(b) 5Hours
_l
-, 0
200
400
600
Common Iliac Flow (Ultrasound) ml. min.-1 Fig. 4. a, Scattergram of thermal conductance measured by the intrauterine probe as a function of common internal iliac blood flow in the first 2 hours after administration of 50 µ.g of 1713-estradiol intravenously to a nonpregnant ewe; b, the same scattergram but including all data points for 5 hours after administration of 50 µ.g of 1713-estradiol intravenously to a nonpregnant ewe. Arrows mark calibration of 0.2 mW· ° C _,
uterus and the vagina in relation to changes in common iliac artery blood flow. During the rising phase of the 17~-estradiol response there appeared to be a reasonable correlation between common iliac flow, thermal conductance, and tissue temperature changes. This was not true after the peak in flow. The flow responses were analyzed for both 2- and 5-hour periods after 17~-estradiol administration. Scattergrams of '1.T (Fig. 3, a and b) and '1.U (Fig. 4, a and b) against common iliac flow at 2 and 5 hours were generated on a microcomputer. The 5-hour scattergrams generally showed very little correlation, and therefore all further discussion will concentrate on the initial 2-hour period. The theoretical relationship for '1.T should be an inverse function of flow and that for '1.U should be an increasing exponential. However, within the 2-hour period the ranges of values observed in '1.T showed a marked linearity with flow. Tissue temperature (iiT). The temperature sensitivity with flow was slightly greater in the uterus than in the vagina. Mean slopes of - 0.13° ( ± 0.03°) C · 100 ml- 1 • min- 1 with a mean r value of -0.83 (±0.13) in the uterus and - 0.09° ( ± 0.05°) C · 100 ml · - ' · min - ' with a mean r value of - 0.90 ( ± 0.07) in the vagina
were found. Small standard deviations indicate that there is little spread in these values. As pointed out by Abrams et al."' the assumptions behind the use of tissue temperature as a measure of flow are that aortic blood temperature remains constant and that the metabolic rate of the uterus and vagina does not alter with flow. The sheep with the aortic thermistor in place enabled us to determine whether blood temperature changes occurred during estradiol administration. Aortic blood temperature was remarkably constant, with a variation of ± 0. I ° C over 5 hours or more after estradiol administration. This variability was about a quarter of that generally seen in the uterus. The rise in pelvic tissue temperature after the peak in common internal iliac flow (Fig. 2) strongly suggests the appearance of metabolic effects caused by estrogen, which do not contribute to the local temperature in the first 2 hours. The conclusion is drawn that during the period of no metabolic changes the temperature changes in both the uterus and the vagina give a good indication of the flow changes in the whole organ as measured by flow in the common iliac artery. Thermal conductance (iiU). The changes in '1.U are small with an average value of 0.15 ( ± 0.18)
656
Randall et al.
March 1988
Am
~l. ~ Ii:-~
8g:!!E....
J Obstet Gynccol
600~
400j
c
200
Q)
~
i~~
::::>-::::>
~
~
l 0·1
b
T
I-
~
Ill
1
~~ 0·5
-~~ T 5
a
-
0
Nor~Epi
I
1
I
2
I
3
I
4
I
5
I
6
I
7
>.'J--..I 10
Time (min.) After Norepinephrine Mministration Fig. 5. Changes in (a) uterine tissue temperature, (b) uterine thermal conductance, and (c) common internal iliac artery blood flow before and after the administration of norepinephrine (100 µ.g) intravenously to the nonpregnant sheep.
mW· 0 c- 1 • 100 ml- 1 • min- 1 with a mean r value of 0.57 ( ::+:: 0.29) m the uterus and 0.14 ( ::+:: 0.18) mW· 0 c- 1 • 100 ml-1 · min- 1 with a mean r value of 0.50 ( ::+:: 0.34) in the vagina. For a change of 100 ml · min - 1 , this represents only a 1o/c to 2o/c change in the total power dissipated in the thermistors. This is surprising in view of the large flow changes occurring in the endometrium and vaginal wall shown by the microspheres. The mean correlation coefficients of 0.57 for the uterus and 0.5 for the vagina are poor. On a number of occasions flow and thermal conductance moved in opposite directions to that expected, even within the initial 2-hour period but particularly on the downward phase of the l 7J3-estradiol response, where ilU sometimes continued to rise when the iliac artery flow was falling. Norepinephrine-induced flow changes. The response to a bolus injection of norepinephrine ( 100 µ,g) is seen in Fig. 5. The thermal conductance might be expected to follow this change as its time constant is short (approximately 10 seconds) and the vasoconstrictive effects should be immediate at the endometrial and vaginal walls. The organ, however, has considerable
thermal inertia and cannot be expected to show significant temperature changes during this short period of decreased How. The tissue temperature plot in Fig. 5, a, although not quantifiable for the reasons given above, did show a measurable and consistent increase during each period of decreased How. It is a highly dampened step response correlating poorly with common iliac artery How, probably a result of the rapidity of the blood How change. The plot of thermal conductance against ultrasound How in Fig. 5, b, appears to indicate a good response. The changes in ilU for the induced flow reductions are large when compared with the increases caused by 17J3-estradiol. Correspondence between uterine and vaginal temperature and thermal conductance. Table II shows the ratios of vaginal to uterine temperature and thermal conductance changes. The uterine tissue temperature changes during l 7J3-estradiol stimulations correlate well with those in the vagina. For the initial 2-hour periods, vaginal temperature changes are around 69% of the uterine changes. The short-term changes induced with norepinephrine gave a slightly higher cor-
Volume 158 Number 3, Part I
respondence of 79%. By contrast, thermal conductance changes measured in the uterus and the vagina do not follow each other particularly well.
Thermal technique to measure pelvic blood flow
Table II. Correspondence between data from uterus and vagina Ti.1.1ue temperature ilT
Comment
The overall objective of this work was to examine the validity of thermal techniques as a method of recording changes in pelvic blood flow and to determine the relationship between the vagina and the uterus as sites for measuring this blood flow. Our oqservations suggest that thermal cl_earance is a poor indicator of wholeorgan (pelvic) blood flow changes while organ tissue temperature changes appear far better. The temperature of the organ results from a balance between metabolic heating and cooling by the total blood flow and is largely independent of regional flow differences. As a result it is less dependent on probe positioning and is totally independent of probe geometry and construction, the only requirement being an accurate calibration. Aortic blood temperature has to be taken into account and was found to be very constant over the experimental periods. The presence of any metabolic changes in the tissue will also affect the rate of heat production and hence temperature. The poor correlation of thermal conductance with blood flow has to be explained. Thermal conductance is only measured in the vicinity of the probe and hence is markedly influenced by the local flow distribution. It is possible, though unlikely, that the pressure from a protruding thermistor itself may cause a change in local tissue perfusion and give rise to the poor correlation observed. Thermal conductance is also subject to differences in probe geometry, sensitivity, construction, and the type and nature of the tissue it is placed against. Local changes may occur with How and posture (affecting local blood volume). Thus, it is not surprising that thermal clearance measured in the vagina shows no better agreement than in the uterus. In contrast, the uterine temperature changes are replicated to a high degree in the vagina, although the changes are reduced to around 70% of those in the uterus. Therefore the vaginal temperature changes show the same qualitative agreement with changes in pelvic flow. This agreement may not hold in all situations, especially in the presence of pathologic states that may alter relationships of blood flow and tissue metabolism. The thermal clearance probes could be made more sensitive by increasing their surface area, improving good contact with the tissue, and increasing the surface temperature of the thermistor. These modifications would result in a large increase in the power requirements, which conflicts with the requirements for a portable long-term recording system. A probe covering a larger area of tissue would undoubtably lessen the sen-
657
Experiment
R
I
Thermal ro11d11rla11ff il C R
A
058 0.67 0.96 0.66 0.45 0.88 0.66
0.97 0.90 0.99 0.98 0.99 0.90 0.95
0.06 0.05 -0.49 0.63 0.41 0.49
~ean
0.69 0.16 0.99 0.81 0.60 0.99 0.79 0.58 0.75
0.95 U.04 0.86 0.99 0.78 0.99 0.75 0.76 0.78
0. I ~I 0.41 0.02
0.79 0.16
0.84 0.11
SD B
Mean SD
*
I 0.02 0.07 -0.32 0.83 0.86 0.56
* o.:~3
0.44 0.()2
0.5:~
0.3~1
-o.:rn
-0.49 0.62 -0.13 0.08
0.63 -0.02 0.04
*
*
0.15 (J.'.15
0.08
o.:rn
R = Ratio of vaginal change/uterine changes {:\) in the initial 2 hours after estradiol administration (50 µg intra1·enously) and (B) after I 00 µg of norepinephrine intravenously. Ratios of tissue temperature are significant at the I O'lc level while there is no significance for tissue thermal conductance ratios. r = Correlation coefficient for the regression line.
s1ttv1ty to local regional How differences. Local flow changes possibly revealed by a thermal clearance technique may be of great interest in some applications and should not be considered to be of no value. Measurement of temperature alone suffers none of the above problems. In this species and under the present conditions we calculate that a change of 20 ml · min - ' in common internal iliac blood flow produced a change of at least 0.02° C in the vagina. For comparative changes, absolute accuracy is not required. For long-term monitoring further investigations are needed to determine whether pelvic and aortic temperature and pelvic flow remain constant over prolonged periods of time. In particular, changes may be provoked by alterations in metabolic state. The aortic temperature needs monitoring to ascertain its changes when a flow change is provoked other than by a drug. However, a short-term evoked response, from a posture change or emotional stress, where aortic temperature should remain constant, could be clinically useful in determining pelvic How abnormalities.'' This would be achieved by the simple, nontraumatic insertion of a vaginal sponge probe measuring only temperature. We would like to thank Mr. D. Wertheim for his help with the thermal recording equipment, Beth Elias for her help with the computerized data storage, Dr. E. R.
Randall et al.
March 1988
Am J Obstet Gynecol
Poore for his help with the data acquisition system facilities and other advice, and Transonic Systems, Inc., for the loan of the ultrasonic ftowmeter equipment.
7.
REFERENCES I. Figueroa JP, Mahan S, Poore ER, Nathanielsz PW. Char-
2. 3.
4. 5. 6.
acteristics and analysis of uterine electromyographic activity in pregnant sheep. AM J 0BSTET GYNECOL 1985; 151 :524-31. Fuller EO, Galletti PM, Takeuchi T. Major and collateral components of blood flow to pregnant sheep uterus. Am J Physiol l 975;229:279-85. Drost CJ. Vessel diameter-independent volume flow measurements using ultrasound. In: Proceedings of San Diego biomedical symposium. San Diego: San Diego Biomedical Society, 1978 vol 17:299-302. Burton RG, Gorewit RC. Ultrasonic flowmeter uses wide beam transit time technique. Med Electronics 1984; 15: 68-73. Heyman MA, Payne BD, Hoffman JIE, Rudolph AM. Blood How measurements with radionuclide-labelled particles. Prog Cardiovasc Dis l 977;20:55-70. Rudolph AM, Heyman MA. Methods of studying the cir-
8.
9.
10. 11.
culation of the fetus in utero. In: Nathanielsz PW, ed. Animal models in fetal medicine. Ithaca, New York: Perinatology Press, 1985 vol I: 1-5 7. Rosenfeld CR, Morriss FH, Makowski EI, Meschia G, Battaglia EC. Circulatory changes in the reproductive tissues of ewes during pregnancy. Gynecol Invest l 974;5:252-68. Coremansj, Vermarien A. A standardized apparatus for continuous control of local How in tissue using chronical sensor implant. In: Proceedings of the sixteenth international conference of medical and biological engineering and the seventh international conference of medical physics. Espoo, Finland: 1985. Meschia G. Circulation to female reproductiYe organs. In: Shepard JT, Abboud F:\f, eds.: Handbook of physiology-the cardioYascular system. Bethesda: American Physiological Society, 1983: vol 3 part 1:241-69. Abrams RM, Caton D, Clapp .JE, Barron DH. Thermal aspects of uterine blood flow in nonpregnanl sheep. A~I J 0BSTET GY:'\EC:OL 1970;108:919-24. Sutherland IA, Beard RW, Randall NJ, Reginald PW, Wertheim DFP. Postural changes in pelvic blood flow. In: Sheldon CD, Evens DH, Salvage R, eds. Obstetric and neonatal blood flow. London: Biological Engineering Society, 1987;12:77-83.
Effects of ritodrine hydrochloride on arteriovenous blood gas and shunt in healthy pregnant yellow baboons Gary D. V. Hankins, Major, USAF, MC, John C. Hauth, Colonel, USAF, MC (Ret), John H. Cissik, Colonel, USAF, BSC, and Thomas J. Kuehl, PhD Lackland Air Force Base and San Antonio, Texas Although the cardiovascular, renal, and metabolic effects of ritodrine hydrochloride have been extensively investigated, its effect on shunt and oxygen dynamics have not. To define the effects of ritodrine hydrochloride on these parameters, six pregnant baboons were exposed to both low (1.5 µg/kg/min) and high (5 µg/kg/min) concentrations of the drug and compared with six control animals not exposed. Ritodrine had no effect on arterial or mixed venous oxygen tension, the arteriovenous oxygen content difference, or the venous-arterial shunt fraction. The treated animals did manifest a significant decline in both arterial (p < 0.05) and mixed venous carbon dioxide (p < 0.01) pressures, with associated increases in the respective pH values (p < 0.001 and p < 0.01 ). Total pulmonary resistance also showed a significant fall during ritodrine treatment (p < 0.05). These data support a minimal, if any, impact of ritodrine hydrochloride on the blunting of normal hypoxic pulmonary vasoconstriction. Furthermore, these data are of potential use as a reference in the treatment of pregnant women who have sustained life-threatening insults from exposure to these agents. (AM J 0BSTET GYNECOL 1988;158:658-63.)
Key words: Hypoxic vasoconstriction, ritodrine, shunt
From the Department~ of Obstetrics and Gynecology and Clinical lnvestigatiII(' Facility, Wilford Hall United States Air Force Medical Centa and Southwest Foundation for Biomedical Research. The opiniow expressed in this article are those of the authors and not necessarily those of the United States Air Force or the Department of Defense. Received for publication July 30, 1987; revised November 4, 1987; accepted November 6, 1987. Reprint requests: Gary D. V. Hankins, Major, USAF, MC, Wilford Hall USAF Med Cen!SGHPG, Lackland AFB, TX 78236-5300.
658
13-Sympathomimetic agents, either alone or in combination with magnesium sulfate, remain the most frequently used tocolytic agents in the United States. The complications of such therapy, including pulmonary edema and maternal death, have recently been reviewed by both Benedetti' and Hankins.' While the literature is replete with data on the cardiovascular, renal, and metabolic effects of these drugs, 1. " little
Key words: Pelvic blood flow, thermal techniques, estrogens and blood flow The use of a thermal system to measure blood flow in a perfused tissue has several potential advantages. Thermal probes are simple to construct and noninvasive to the surrounding tissue and blood flow can be measured continuously or by repetitive sampling. Although interest in the technique has spanned more than 50 years, the method presents several problems that so far have prevented its practical clinical use. We propose that this method could form the basis for a low-cost ambulatory system for monitoring significant features of pelvic blood flow in the human. In the present experiments we used the chronically instrumented nonpregnant sheep model to validate these thermal methods to measure pelvic blood flow changes. Thermal monitoring at the uterine and vaginal walls was compared with continuous ultrasound and intermittent microsphere flow techniques to determine the degree of correspondence of the three methods at the two pelvic sites. From the Department of Obstetrics and Gynaecology, St. Mary's Hospital Medical School, London, and the Laboratory for Pregnancy and Newborn Research, New York State College of Veterinary Medicine, Cornell University. This work was supported by Grant PCM-8303035 from the National Science Foundation and by the Harris-Birthright Fund for Research into Early Pregnancy. Received for publication December 17, 1986; revised July 27, 1987; accepted October 28, 1987. Reprint requests: Dr. Peter W. Nathanie/sz, MD, PhD, Laboratory for Pregnancy and Newborn Research, New York State College of Veterinary Medicine, 526 Veterinary Research Tower, Cornell University, Ithaca, NY 14853.
Material and methods Surgical procedures. Four nonpregnant Rambouillet-Columbia ewes were premedicated with ketamine hydrochloride (20 mg · Kg- 1) and glycopyrrolate (10 µg · kg- 1). Operation was conducted with the animal under halothane anesthesia. Catheters were inserted into the left ventricle and the femoral artery. In one sheep a thermistor was placed in the dorsal aorta to monitor aortic blood temperature during flow changes. The uterus and pelvic blood vessels were exposed via a midline abdominal incision. A twin thermistor probe was inserted into each uterine horn near the tip. The sheep were allowed to recover for at least 4 hours after operation. Continuous measurement of uterine blood flow. A 6 mm transit-time ultrasonic flow probe (Transonics Systems, Inc., Ithaca, N.Y.) 2 ·• was placed around the common inernal iliac artery.2 Blood flow measured through this probe represents supply to the uterine, vaginal, vesical, and other pelvic vascular beds. In any one sheep mean common iliac blood How averaged over I-minute intervals showed a peak-to-peak variability of only ± 10 ml· min- 1 • Thermistor probes. The uterine thermistor probes contained two 2.4-mm diameter precision thermistors (tolerance ± 0.2° C) (type ACC-001, Ametek, Rodan Division, 2905 Blue Star St., Anaheim, CA 92806) attached to a short length of polyvinyl chloride catheter. One thermistor was mounted at the catheter tip and the second placed into its side, 15 mm further back. 651
652
Randall et al.
March 1988
Am
J Obstet Gynecol
700
600
-
500
I. c
.E
:s:::
Whole Organ Flow by Micro-spheres
400 300
0
u:::: 200
100 0
I 2
0
3
Time (hours) Fig. l. Comparison of common iliac artery flow measured by ultrasonic flowmeter (recorded continuously and hence plotted over the full range of time with calibration bar at 600 ml · min - i shown) and total flow as determined by microspheres during the estradiol stimulations. The envelope of the ultrasound flow is ±SD while the range is shown for the microsphere results.
Common Iliac Artery Flow ml.min-I
AT~ Temp.
1
Al 1 Uterine
~Thermal
Conductance
AU~=
Conductance
0
2
3
4
5
Time After Estracliol Admiistration (hrs) Fig. 2. Changes in common internal iliac artery flow (upper trace), intrauterine tissue temperature (second trace), vaginal tissue temperature (third trace), uterine thermal conductance (fourth trace), and vaginal thermal conductance (bottom trace) after 50 µg of 1713-estradiol administered intravenously to a non pregnant ewe. Calibration for common iliac artery blood flow is 700 ml · min - 1• All other vertical scales are arbitrary.
We used two designs of vaginal probe inserted into the vagina to rest against the cervix, with the thermistors in contact with the vaginal wall. One used 2.4-mm diameter thermistors mounted on the rim of a 5 cm poly-
vinyl chloride pessary 15 mm apart. The other had two probes of the uterine type, but with single thermistors at the tip, mounted in a foam sponge 30 mm in diameter and 30 mm long. The thermistor tips stood out
Volume 158 Number 3, Pan I
Thermal technique to measure pelvic blood flow
653
Table I. Effect of estradiol administration (50 µg via jugular vein) on regional blood flow with radioactive microspheres Time of isotope injection I= 0 hr (ml· min- 1)
Mean
Tissue flow Endometrium Cervix Vagina Vulva Right mesothelii.im, tube Left mesothelium, tube Right horn, body Left horn, body Whole organ Tissue weight (gm) Endometrium Cervix Vagina Vulva Right mesothelium, tube Left mesothelium, tube Uterus, both horns Total
7.6 4.8 7.7 2.4 1.6 1.6 4.1 3.7 33 17.6 13.3 26.8 20.4 15.1 15.7 40.2 149
I
Mean fold increase
t = 1 hr (ml· min- 1)
Range
Mean
6-11 2-10 3-15 1-6 1-3 1-2 3-6 3-4 22-53
I
t = 2 hr
Range
Mean
109 48 52 6.6 6.7 7.2 45 48
101-117 28-57 48-56 5-8 4-9 6-9 41-49 46-50
156 67 57 15 12
323
288-358
I
For each sheep (2 hr/Ohr)
I
Range
Mean
67 65
129-205 64-68 51-66 12-16 10-14 10-13 56-75 55-76
22 23 12 8 8 8 18 18
12-31 7-38 5-22 4-12 6-12 5-9 13-25 14-25
450
387-533
15
8-20
II
Range
13.4-24.4 11.2-16.6 21.0-35.0 13.5-25.0 12.3-18.4 14.1-16.6 31.8-46.6 135-156
from the surface of the foam. This latter design proved much easier to insert and was mechanically more stable within the vagina. The foam, with its reduced thermal mass, gave better heat transfer characteristics and all data reported in this paper were obtained with this type of probe. To prevent ingress of body fluids between the thermistor and probe body containing the electrical connections, probes were dip-coated in a latex solution whose sticky film was cured with a dusting of powder. The thermistor probes were connected to an isothermal control unit. One of a pair of thermistors operated in a self-heating mode while the other monitored tissue temperature. The control unit maintained the heated thermistor core temperature at a constant increment above tissue temperature and provided outputs of tissue temperature, heated thermistor temperature, and the driving voltage to this thermistor for calculation of dissipated power. Data were fed into an Apple II-based data acquisition system.' Absolute thermal conductance of the heated probe in vivo was calculated while probe temperature was increased up to a 3° C increment. No significant effects on thermal conductance with probe temperature were observed. As reported by other authors there were variations in absolute thermal conductance between probes (4 to 10.5 mW· 0 c- 1) that were also observed in water-bath experiments. Thus, absolute values are meaningless. To obtain absolute or relative measurements, probes would need to be thermally identical and very accurately calibrated in tissue of the type to be studied. Thermistors were operated at a temperature elevation of 2.0° C.
The surface temperature of the encapsulating material of the thermistors is less than this." Microsphere measurement of regional blood flow to pelvic organs. In all four sheep regional blood flows were measured with radioactive microspheres 5· 6 before and after the administration of I 713-estradiol. Using similar microsphere-based measurement techniques, Rosenfeld et al. 7 found that flow in the uterine arteries accounted for approximately 80% of the total uterine blood flow while Fuller et al. 2 found this to vary between 66% and 81 % in pregnant sheep, depending on the period of gestation. All radioactive microsphere studies with tin 113, gadolinium 153, and scandium 46 were carried out during the final 1713-estradiol stimulation. Pharmacologic manipulation of pelvic blood flow. Intravenous bolus injection of 1713-estradiol (50 µg) was administered to induce increases in uterine blood flow in 14 experiments. Repeat injections of 1713-estradiol were given only after uterine blood flow had returned to basal levels. On eight other occasions, twice to each animal, injections of norepinephrine (I 00 µg) were given to produce a temporary flow decrease. Uterine blood flow had always returned to basal levels between experiments. The regional organ blood flows were measured at three time points: just before 1713-estradiol injection;· I hour after 1713-estradiol injection, when flow in the common iliac had approximately doubled; and at 2 to 2 1/2 hours after the injection, when the blood flow was at its peak. In one sheep readings were taken at 2 and 5 hours on the downward slope. Analytical methods. For the 1713-estradiol studies the
654
Randall et al.
March 1988
Am
J Obstet Gynecol
(a) 2 Hours
-''•'"'''····
..
1
T µ 1-
(b) 5 Hours
..................-.: ..·.; . _.. ,- ....... :~~==~=:·-:-·
l T
._;~~~·f"~,~~~~:r~:&rf~i~;; .·. . 1
200
400
600
Common Iliac Flow (ultrasound) ml. min.-1 Fig. 3. a, Scattergram of temperature change measured by the intrauterine probe as a function of common internal iliac blood flow in the first 2 hours after administration of 50 µg of 1713-estradiol intravenously to a nonpregnant ewe; b, the same scattergram but including all data points for 5 hours after administration of 50 µg of 1713-estradiol intravenously to a nonpregnant ewe. Arrows mark calibration of 0.1° C.
data acquisition system logged the mean of each output channel every 8 seconds. In order to follow the more rapid changes caused by norepinephrine the outputs were sampled every 2 se.conds. Correlation was performed by linear regression analysis. The slopes and linear correlation coefficients (r) from all the 1713estradiol studies in the four sheep were averaged, and each is presented as a mean ± SD unless otherwise stated. The number of experiments, n, varied from 10 to 12, depending on the site of measurement. In the microsphere experiments the regional blood flows represent the mean value from three separate sheep.
Results Blood flow measiirements with microspheres. There was a threefold to fourfold increase in blood flow through the common internal iliac artery during an 1713-estradiol stimulation as measured by the ultrasonic flowmeter (Fig. 1). Table I lists the mean and range of
regional flows to eight different sites. Basal blood flows were commonly restored after 10 hours. As demonstrated in Fig. 1, the total common iliac flow exceeded the total flow in the eight tissues measured by the microspheres (Table I) by approximately 150 ml · min- 1 both at rest and during 1713-estradiol stimulation. This difference reflects distribution of blood in the common iliac artery to tissues such as the bladder and the muscles of the pelvis, which were not counted for radioactivity. During stimulation by 1713-estradiol there is a redistribution of relative flow with a larger proportion of internal iliac blood flow going to the endometrium and cervix. Fig. 1 suggests that the blood flow to areas other than those in which regional blood flow was measured was not significantly altered by 1713-estradiol administration. Thermal measurements during blood flow changes. Fig. 2 shows a typical result of changes in thermal conductance (ilU) and tissue temperature (ilT) in the
Thermal technique to measure pelvic blood flow
Volume 158 Number 3, Part I
655
(a) 2 Hours
J_
-,
"j
u
~
~
(b) 5Hours
_l
-, 0
200
400
600
Common Iliac Flow (Ultrasound) ml. min.-1 Fig. 4. a, Scattergram of thermal conductance measured by the intrauterine probe as a function of common internal iliac blood flow in the first 2 hours after administration of 50 µ.g of 1713-estradiol intravenously to a nonpregnant ewe; b, the same scattergram but including all data points for 5 hours after administration of 50 µ.g of 1713-estradiol intravenously to a nonpregnant ewe. Arrows mark calibration of 0.2 mW· ° C _,
uterus and the vagina in relation to changes in common iliac artery blood flow. During the rising phase of the 17~-estradiol response there appeared to be a reasonable correlation between common iliac flow, thermal conductance, and tissue temperature changes. This was not true after the peak in flow. The flow responses were analyzed for both 2- and 5-hour periods after 17~-estradiol administration. Scattergrams of '1.T (Fig. 3, a and b) and '1.U (Fig. 4, a and b) against common iliac flow at 2 and 5 hours were generated on a microcomputer. The 5-hour scattergrams generally showed very little correlation, and therefore all further discussion will concentrate on the initial 2-hour period. The theoretical relationship for '1.T should be an inverse function of flow and that for '1.U should be an increasing exponential. However, within the 2-hour period the ranges of values observed in '1.T showed a marked linearity with flow. Tissue temperature (iiT). The temperature sensitivity with flow was slightly greater in the uterus than in the vagina. Mean slopes of - 0.13° ( ± 0.03°) C · 100 ml- 1 • min- 1 with a mean r value of -0.83 (±0.13) in the uterus and - 0.09° ( ± 0.05°) C · 100 ml · - ' · min - ' with a mean r value of - 0.90 ( ± 0.07) in the vagina
were found. Small standard deviations indicate that there is little spread in these values. As pointed out by Abrams et al."' the assumptions behind the use of tissue temperature as a measure of flow are that aortic blood temperature remains constant and that the metabolic rate of the uterus and vagina does not alter with flow. The sheep with the aortic thermistor in place enabled us to determine whether blood temperature changes occurred during estradiol administration. Aortic blood temperature was remarkably constant, with a variation of ± 0. I ° C over 5 hours or more after estradiol administration. This variability was about a quarter of that generally seen in the uterus. The rise in pelvic tissue temperature after the peak in common internal iliac flow (Fig. 2) strongly suggests the appearance of metabolic effects caused by estrogen, which do not contribute to the local temperature in the first 2 hours. The conclusion is drawn that during the period of no metabolic changes the temperature changes in both the uterus and the vagina give a good indication of the flow changes in the whole organ as measured by flow in the common iliac artery. Thermal conductance (iiU). The changes in '1.U are small with an average value of 0.15 ( ± 0.18)
656
Randall et al.
March 1988
Am
~l. ~ Ii:-~
8g:!!E....
J Obstet Gynccol
600~
400j
c
200
Q)
~
i~~
::::>-::::>
~
~
l 0·1
b
T
I-
~
Ill
1
~~ 0·5
-~~ T 5
a
-
0
Nor~Epi
I
1
I
2
I
3
I
4
I
5
I
6
I
7
>.'J--..I 10
Time (min.) After Norepinephrine Mministration Fig. 5. Changes in (a) uterine tissue temperature, (b) uterine thermal conductance, and (c) common internal iliac artery blood flow before and after the administration of norepinephrine (100 µ.g) intravenously to the nonpregnant sheep.
mW· 0 c- 1 • 100 ml- 1 • min- 1 with a mean r value of 0.57 ( ::+:: 0.29) m the uterus and 0.14 ( ::+:: 0.18) mW· 0 c- 1 • 100 ml-1 · min- 1 with a mean r value of 0.50 ( ::+:: 0.34) in the vagina. For a change of 100 ml · min - 1 , this represents only a 1o/c to 2o/c change in the total power dissipated in the thermistors. This is surprising in view of the large flow changes occurring in the endometrium and vaginal wall shown by the microspheres. The mean correlation coefficients of 0.57 for the uterus and 0.5 for the vagina are poor. On a number of occasions flow and thermal conductance moved in opposite directions to that expected, even within the initial 2-hour period but particularly on the downward phase of the l 7J3-estradiol response, where ilU sometimes continued to rise when the iliac artery flow was falling. Norepinephrine-induced flow changes. The response to a bolus injection of norepinephrine ( 100 µ,g) is seen in Fig. 5. The thermal conductance might be expected to follow this change as its time constant is short (approximately 10 seconds) and the vasoconstrictive effects should be immediate at the endometrial and vaginal walls. The organ, however, has considerable
thermal inertia and cannot be expected to show significant temperature changes during this short period of decreased How. The tissue temperature plot in Fig. 5, a, although not quantifiable for the reasons given above, did show a measurable and consistent increase during each period of decreased How. It is a highly dampened step response correlating poorly with common iliac artery How, probably a result of the rapidity of the blood How change. The plot of thermal conductance against ultrasound How in Fig. 5, b, appears to indicate a good response. The changes in ilU for the induced flow reductions are large when compared with the increases caused by 17J3-estradiol. Correspondence between uterine and vaginal temperature and thermal conductance. Table II shows the ratios of vaginal to uterine temperature and thermal conductance changes. The uterine tissue temperature changes during l 7J3-estradiol stimulations correlate well with those in the vagina. For the initial 2-hour periods, vaginal temperature changes are around 69% of the uterine changes. The short-term changes induced with norepinephrine gave a slightly higher cor-
Volume 158 Number 3, Part I
respondence of 79%. By contrast, thermal conductance changes measured in the uterus and the vagina do not follow each other particularly well.
Thermal technique to measure pelvic blood flow
Table II. Correspondence between data from uterus and vagina Ti.1.1ue temperature ilT
Comment
The overall objective of this work was to examine the validity of thermal techniques as a method of recording changes in pelvic blood flow and to determine the relationship between the vagina and the uterus as sites for measuring this blood flow. Our oqservations suggest that thermal cl_earance is a poor indicator of wholeorgan (pelvic) blood flow changes while organ tissue temperature changes appear far better. The temperature of the organ results from a balance between metabolic heating and cooling by the total blood flow and is largely independent of regional flow differences. As a result it is less dependent on probe positioning and is totally independent of probe geometry and construction, the only requirement being an accurate calibration. Aortic blood temperature has to be taken into account and was found to be very constant over the experimental periods. The presence of any metabolic changes in the tissue will also affect the rate of heat production and hence temperature. The poor correlation of thermal conductance with blood flow has to be explained. Thermal conductance is only measured in the vicinity of the probe and hence is markedly influenced by the local flow distribution. It is possible, though unlikely, that the pressure from a protruding thermistor itself may cause a change in local tissue perfusion and give rise to the poor correlation observed. Thermal conductance is also subject to differences in probe geometry, sensitivity, construction, and the type and nature of the tissue it is placed against. Local changes may occur with How and posture (affecting local blood volume). Thus, it is not surprising that thermal clearance measured in the vagina shows no better agreement than in the uterus. In contrast, the uterine temperature changes are replicated to a high degree in the vagina, although the changes are reduced to around 70% of those in the uterus. Therefore the vaginal temperature changes show the same qualitative agreement with changes in pelvic flow. This agreement may not hold in all situations, especially in the presence of pathologic states that may alter relationships of blood flow and tissue metabolism. The thermal clearance probes could be made more sensitive by increasing their surface area, improving good contact with the tissue, and increasing the surface temperature of the thermistor. These modifications would result in a large increase in the power requirements, which conflicts with the requirements for a portable long-term recording system. A probe covering a larger area of tissue would undoubtably lessen the sen-
657
Experiment
R
I
Thermal ro11d11rla11ff il C R
A
058 0.67 0.96 0.66 0.45 0.88 0.66
0.97 0.90 0.99 0.98 0.99 0.90 0.95
0.06 0.05 -0.49 0.63 0.41 0.49
~ean
0.69 0.16 0.99 0.81 0.60 0.99 0.79 0.58 0.75
0.95 U.04 0.86 0.99 0.78 0.99 0.75 0.76 0.78
0. I ~I 0.41 0.02
0.79 0.16
0.84 0.11
SD B
Mean SD
*
I 0.02 0.07 -0.32 0.83 0.86 0.56
* o.:~3
0.44 0.()2
0.5:~
0.3~1
-o.:rn
-0.49 0.62 -0.13 0.08
0.63 -0.02 0.04
*
*
0.15 (J.'.15
0.08
o.:rn
R = Ratio of vaginal change/uterine changes {:\) in the initial 2 hours after estradiol administration (50 µg intra1·enously) and (B) after I 00 µg of norepinephrine intravenously. Ratios of tissue temperature are significant at the I O'lc level while there is no significance for tissue thermal conductance ratios. r = Correlation coefficient for the regression line.
s1ttv1ty to local regional How differences. Local flow changes possibly revealed by a thermal clearance technique may be of great interest in some applications and should not be considered to be of no value. Measurement of temperature alone suffers none of the above problems. In this species and under the present conditions we calculate that a change of 20 ml · min - ' in common internal iliac blood flow produced a change of at least 0.02° C in the vagina. For comparative changes, absolute accuracy is not required. For long-term monitoring further investigations are needed to determine whether pelvic and aortic temperature and pelvic flow remain constant over prolonged periods of time. In particular, changes may be provoked by alterations in metabolic state. The aortic temperature needs monitoring to ascertain its changes when a flow change is provoked other than by a drug. However, a short-term evoked response, from a posture change or emotional stress, where aortic temperature should remain constant, could be clinically useful in determining pelvic How abnormalities.'' This would be achieved by the simple, nontraumatic insertion of a vaginal sponge probe measuring only temperature. We would like to thank Mr. D. Wertheim for his help with the thermal recording equipment, Beth Elias for her help with the computerized data storage, Dr. E. R.
Randall et al.
March 1988
Am J Obstet Gynecol
Poore for his help with the data acquisition system facilities and other advice, and Transonic Systems, Inc., for the loan of the ultrasonic ftowmeter equipment.
7.
REFERENCES I. Figueroa JP, Mahan S, Poore ER, Nathanielsz PW. Char-
2. 3.
4. 5. 6.
acteristics and analysis of uterine electromyographic activity in pregnant sheep. AM J 0BSTET GYNECOL 1985; 151 :524-31. Fuller EO, Galletti PM, Takeuchi T. Major and collateral components of blood flow to pregnant sheep uterus. Am J Physiol l 975;229:279-85. Drost CJ. Vessel diameter-independent volume flow measurements using ultrasound. In: Proceedings of San Diego biomedical symposium. San Diego: San Diego Biomedical Society, 1978 vol 17:299-302. Burton RG, Gorewit RC. Ultrasonic flowmeter uses wide beam transit time technique. Med Electronics 1984; 15: 68-73. Heyman MA, Payne BD, Hoffman JIE, Rudolph AM. Blood How measurements with radionuclide-labelled particles. Prog Cardiovasc Dis l 977;20:55-70. Rudolph AM, Heyman MA. Methods of studying the cir-
8.
9.
10. 11.
culation of the fetus in utero. In: Nathanielsz PW, ed. Animal models in fetal medicine. Ithaca, New York: Perinatology Press, 1985 vol I: 1-5 7. Rosenfeld CR, Morriss FH, Makowski EI, Meschia G, Battaglia EC. Circulatory changes in the reproductive tissues of ewes during pregnancy. Gynecol Invest l 974;5:252-68. Coremansj, Vermarien A. A standardized apparatus for continuous control of local How in tissue using chronical sensor implant. In: Proceedings of the sixteenth international conference of medical and biological engineering and the seventh international conference of medical physics. Espoo, Finland: 1985. Meschia G. Circulation to female reproductiYe organs. In: Shepard JT, Abboud F:\f, eds.: Handbook of physiology-the cardioYascular system. Bethesda: American Physiological Society, 1983: vol 3 part 1:241-69. Abrams RM, Caton D, Clapp .JE, Barron DH. Thermal aspects of uterine blood flow in nonpregnanl sheep. A~I J 0BSTET GY:'\EC:OL 1970;108:919-24. Sutherland IA, Beard RW, Randall NJ, Reginald PW, Wertheim DFP. Postural changes in pelvic blood flow. In: Sheldon CD, Evens DH, Salvage R, eds. Obstetric and neonatal blood flow. London: Biological Engineering Society, 1987;12:77-83.
Effects of ritodrine hydrochloride on arteriovenous blood gas and shunt in healthy pregnant yellow baboons Gary D. V. Hankins, Major, USAF, MC, John C. Hauth, Colonel, USAF, MC (Ret), John H. Cissik, Colonel, USAF, BSC, and Thomas J. Kuehl, PhD Lackland Air Force Base and San Antonio, Texas Although the cardiovascular, renal, and metabolic effects of ritodrine hydrochloride have been extensively investigated, its effect on shunt and oxygen dynamics have not. To define the effects of ritodrine hydrochloride on these parameters, six pregnant baboons were exposed to both low (1.5 µg/kg/min) and high (5 µg/kg/min) concentrations of the drug and compared with six control animals not exposed. Ritodrine had no effect on arterial or mixed venous oxygen tension, the arteriovenous oxygen content difference, or the venous-arterial shunt fraction. The treated animals did manifest a significant decline in both arterial (p < 0.05) and mixed venous carbon dioxide (p < 0.01) pressures, with associated increases in the respective pH values (p < 0.001 and p < 0.01 ). Total pulmonary resistance also showed a significant fall during ritodrine treatment (p < 0.05). These data support a minimal, if any, impact of ritodrine hydrochloride on the blunting of normal hypoxic pulmonary vasoconstriction. Furthermore, these data are of potential use as a reference in the treatment of pregnant women who have sustained life-threatening insults from exposure to these agents. (AM J 0BSTET GYNECOL 1988;158:658-63.)
Key words: Hypoxic vasoconstriction, ritodrine, shunt
From the Department~ of Obstetrics and Gynecology and Clinical lnvestigatiII(' Facility, Wilford Hall United States Air Force Medical Centa and Southwest Foundation for Biomedical Research. The opiniow expressed in this article are those of the authors and not necessarily those of the United States Air Force or the Department of Defense. Received for publication July 30, 1987; revised November 4, 1987; accepted November 6, 1987. Reprint requests: Gary D. V. Hankins, Major, USAF, MC, Wilford Hall USAF Med Cen!SGHPG, Lackland AFB, TX 78236-5300.
658
13-Sympathomimetic agents, either alone or in combination with magnesium sulfate, remain the most frequently used tocolytic agents in the United States. The complications of such therapy, including pulmonary edema and maternal death, have recently been reviewed by both Benedetti' and Hankins.' While the literature is replete with data on the cardiovascular, renal, and metabolic effects of these drugs, 1. " little