- Email: [email protected]
Development of uterine artery compliance in pregnancy as detected by Doppler ultrasound
Monitoring of twin pregnancies
Volume 155 Number 5
fetal death in twin pregnancy," Rapid decrease of estriol level, but not human placental lactogen level, in one patient after fetal death may indicate the usefulness of estriol values for the monitoring of the surviving fetus in such a case.
REFERENCES 1. Batra S, Sjoberg ND, Aberg A. Human placental lactogen,
estradiol-Izjs, and progesterone levels in the third trimester and their respective values for detecting twin pregnancy. AMl OBSTET GYNECOL 1978;131:69-72. 2. Gerhard I, Runnebaum B. Hormonal evaluation in multiple pregnancies. Geburtschilfe Frauenheilkd 1980;40: 101-15. 3. Duff GB, Brown lB. Urinary oestriol excretion in twin pregnancies. 1 Obstet Gynaecol Br Commonw 1974; 81:695"700.
4. Daw E. Human placental lactogen and twin pregnancy. Lancet 1977;2:299-300. 5. Trapp M, Bohnet HG, Weise HC, Leidenberger FA. Referenzbereiche fur Serum- Konzentrationen des humanen plazentaren Laktogens und des Ostriols von Frauen mit Zwillingsschwangerschaften. Lab Med 1984;8: 194-7. 6. Enbom JA. Twin pregnancy with intrauterine death of one twin. AM 1 OBSTET GYNECQL 1985;152:424-9. 7. Lilford R], Obiekwe BC, Chard T. Maternal blood levels of human placental lactogen in the prediction of fetal growth retardation: choosing a cut-off point between normal and abnormal. Br J Obstet GynaecoI1983;90:511-5. 8. Litschgi M, Stucki D. Course of twin pregnancies after foetal death in utero. Z Geburtshilfe Perinatol 1980;184: 227-30. 9. Hanna lH, Hill 1M. Single intrauterine fetal demise in multiple gestation. Obstet Gynecol 1984;63: 126-9.
Development of uterine artery compliance in pregnancy as detected by Doppler ultrasound Harold Schulman, M.D., Adiel Fleischer, M.D., George Farmakides, M.D., Luis Bracero, M.D., Burton Rochelson, M.D., and Lawrence Grunfeld, M.D. Mineola, Stony Brook, and Bronx, New York Uterine artery velocimetry was performed by means of a continuous-wave Doppler ultrasound. Serial studles were done on 12 nonpregnant and 79 normal pregnant women. Measurements were made on both uterinearteriesand averaged. From the proliteratlve phase of the menstrual cycle through 40 weeks of pregnancy, four developmental phases are described. These changes are based on the calculation of the systolic-diastolic ratio and the disappearance of the early diastolic notch in the velocity wave. The ratio after 26 weeks averaged 2 ± 0.3. These data provide a foundation for the study of pregnancies in which these normal evolutionary changes do not occur. (AM J OesTET GVNECOL 1986;155:1031-6.)
Key words: Doppler, uterine artery, pregnancy, compliance, systolic/diastolic ratio
The development of adequate uterine blood flow is believed to be essential for normal fetal growth and nutrition and for the avoidance of maternal hypertensive complications. Until now there was no simple and reliable technique for assessing the maternal and fetal placental circulation. Doppler ultrasound instrumentation provides a noninvasive method of studying the status of various blood vessels by providing a spectrum analysis of the velocities of the moving red blood cells.1
From theDepartments of Obstetrics and Gynecology, Winthrop University Hospital, Mineola, State University of New York at Stony Brook, and AlbertEinstein College of Medicine, Bronx. Received for puhlication February 5, 1986; revised June 5, 1986; accepted June 9, 1986. Reprint requests: HaroldSchulman, M.D., Department of Obstetrics and Gynecology, Winthrop-University Hospital, 259 First St., Mineola, NY 11501.
In previous reports we detailed the reliability of umbilical velocimetry for the evaluation of normal growth and development of the fetus and the identification of intrauterine growth retardation.v" In this report we describe our initial findings when velocimetry was performed on the uterine arteries in normal pregnancy and nonpregnant women. The changes in the uterine circulation in pregnancy can be attributed to a number of factors. The principal influencing elements are the effect of the hormones estrogen and progesterone, the changes in blood volume, intravascular pressure, and the change in peripheral resistance created by the development of the intervillous spaces. The summation of these changes are a reflection of vessel compliance; hence we have chosen to describe the changes seen with velocimetry as changes in compliance! 1031
November 1986 Am J Obstet Gynecol
1032 Schulman et al.
13 12
SID Ratio Phase I Prolif-Secretory n First trimester m2nd trimester IInate 2nd trimester and Third trimester
11 10
9
· L
ANTERIOR DIVISION OF INTERN AL ILIAC ARTERY EXTERNAL ILIAC
_ _ Vaginal
b--'" Abdom inal
.2 8
o
IX
7
e6
VI
5 4
p< .05(vs20wks)
3 2 Fig. 1. Velocity wave patterns in the parauterine vessels. The arcuate vessels have the highest diastolic frequency. Time frame of each panel is 1.6 seconds .
Table I. Uterine artery velocity ratios during the menstrual cycle* Systolic!diastolic ratios (mean ± SD)
Cycle days 10-13 (n = 8)
23 26-29 (n
=
8)
12.9 ± 4.4 8.0 ± 2.9 7.2 ± 3.2
*** I<---IIL-4
.
A.'~·fi't't"it:l.f'=lt;l~ I
I
I
I
I
I
I
12 16 20 24 28 32 36 40
WEEKS OF GESTATION
Fig. 2. Uterine artery compliance during the menstrual cycle and throughout pregnancy. Means ± SD. Circles represent measurements through the vaginal forn ices and the triangles, measurements across the abdomen. Each value represents the average between the left and right artery. SID = Systolic/diastolic.
(p = 0.003)
*Four additional women were studied and excluded because of inability to isolate vessel.
Mat~rial
and methods Studies were done on 12 nonpregnant and 79 pregnant women. Approval for this work was obtained from both institutional research review committees. Informed consent was obtained from each woman studied. The data obtained were not used for clinical therapeutic decisions. Continuous-wave Doppler studies were carried out with a Sonicaid flow probe, which has a calculated power density of 6.5 mW/cm2 • An analysis of the returning frequency signal was performed on an Angioscan spectrum analyzer displayed on a screen. Continuous recordings were stored on a JVC audio unit. The transducer used in these studies was a 4 MHz probe that has a beam diameter of 12 mm . . The majority of these studies were done without real time imaging of the vessels under study. It is our thesis that successful examinations can be performed by pattern recognition. In the early phases of this study confirrnation of vessel patterns was obtained by duplex imaging or by in vivo measurements obtained by using a 10 MHz probe during cesarean section. It became clear from experience and team review of the recorded tapes that patterns of iliac, uterine, and arcuate vessels could be confidently separated. Other vessels encoun-
tered in the pelvis have patterns that show no diastolic frequency and are easily excluded. The range depth of the probe of 8 em excludes the possibility of reaching the ovarian or renal arteries. Velocimetry studies were done abdominally and vaginally. Initially vaginal studies were done in the belief that the proximity to the uterine vessels would allow easier and more accurate examinations. We have learned that in the early second trimester and onward, reliable studies can be obtained abdominally. The pregnant uterine artery can be identified by the presence of velocity during diastole. Efforts were made to study both uterine arteries and the umbilical artery. Vaginal studies were done by inserting the transducer into the forefinger of a rubber glove with coupling jelly. The glove and transducer were inserted into the vaginal fornix with the aid of lubricating jelly and directed along the paracervical area. Abdominal studies were done by directing the transducer into the parauterine area in the region of the lower uterine segment. The probe was gently rotated until the appropriate pattern was identified. (Recall that the angle of incidence must be less than 90 degrees in order to produce a satisfactory signal.) Examinations were carried out in the dorsal lithotomy position in nonpregnant women or the left or right tilt position in pregnant subjects. Waveforms were acceptable for measurement when they were dense and of equal amplitude. An effort was made to obtain the largest complexes possible. The Angioscan
Uterine artery compliance in pregnancy
Volume 155 Number 5
DAY 1/ 23
DAY 1/13
1033
DAY 1/ 29
MENSTRUAL CYCLE
Fig. 3. Uterine artery velocity waves during the menstrual cycle. Note increased diastolic frequency return in secretory phase.
EARLY PItECSA.'IlCY
Fig. 4. Uterine artery velocimetry in first trimester. Note that there is little difference from that seen in secretory phase.
monitoring screen shows a 1.6-second time frame. The duration of an individual study was 5 to 15 minutes. Successful imaging of the uterine vessels depended on the stage of pregnancy. From 14 weeks onward at least one vessel could be studied unless there was significant obesity. More difficulty was encountered in the first trimester, and there was a 10% failure rate. A minimum of four waveforms were measured on each side, and the values were averaged. The mean intraobserver and interobserver error between examinations was found to be 4% with a standard deviation of 2.3%. In the studies on the nonpregnant vessels, one third were considered unsatisfactory. Calculations were made by measuring the systolic! diastolic ratio. This computation circumvents the problem of not knowing the angle of insonation. When appropriate, significant differences were calculated by estimating multiples of standard deviations and by the Student t test.
VAG INAL
ABDOMINAL
RT
SID· 4 .2
sID · 2.9
sID · 2. 9
s ID· 2.4
LT
Results A continuous-wave Doppler system provided a spectrum analysis of all vessels that were crossed. Three characteristic vessels were seen in the parauterine area (Fig. 1). Those immediately adjacent to the uterus showed the lowest ratios and were believed to be the arcuates and the distal uterine branches. The proximal
Fig. 5. Abdominal versus vaginal measurements. The placental blood supply is predominantly on the left; thus the ratios are lower and the notching has disappeared. Note closeness of the ratios. Pregnancy of 24 weeks. SID = Systolic/diastolic.
1034 Schulman at al.
November
Am
1986
J Obstet Gynecol
Table II. Uterine artery systolic/diastolic ratios weeks of gestation Weeks of gestation
10
8 Study
Mean ±SD
I
n
Mean ±SD
14
In
Abdominal Vaginal
9.6 ±3
3
8.3 ±4
4
Mean ±SD
2.4 ±0.1 4.1* ±0.95
I
n
6 4
Mean ±SD
1.8 ±0.6 2.8 ± 1.3
n
7 3
Mean ±SD
1.9 ±0.6 3.8 ±0.95
22
20
18
16
In 7 6
Mean ±SD
2.3 ±0.2 3 ±0.9
I
n
6 12
Mean ±SD
2.5 ±0.6 2.7 ± 1.5
24
I
n
8 7
Mean ±SD
2.2 ±0.4 2.7
I
n
8 3
*p < 0.05 from 10 weeks. tp < 0.01 from 14 weeks.
uterine branch seemed to have a higher systolic component, and this vessel was commonly seen during the vaginal studies. The external iliac is a characteristic high-resistance artery with no end-diastolic velocity and a prominent notch. On occasion a pelvic vein was also seen. Table I provides data on systolic/diastolic ratios in eight nonpregnant women. An additional four women in whom studies were attempted were considered unsatisfactory subjects because of the inability to obtain satisfactory scans. Vessel recognition is dependent on the identification of diastolic velocity. In these four women a diastolic component was not consistently present until the end of the secretory phase. This may indicate a deficiency of the methodology or that these women had a higher uterine artery resistance than the others. All 12 women were nulliparous. Ovulation was confirmed by plasma progesterone levels. The data are presented to demonstrate apparent trends but should be viewed with caution. It appears that during the proliferative phase the vessel remains in a state of high resistance, but the ratio declines in the secretory phase when there is a diastolic velocity. The ratios in the endsecretory phase were significantly lower than those seen in the late proliferative phase. Table II summarizes the measurements in the pregnant populations. All women in these studies had a normal pregnancy outcome. The upper part of the table represents the studies made across the abdomen; the lower part depicts those made through the vaginal fornices. The values represent the average of both the right and the left uterine artery. The studies done vaginallyhad higher mean values in gestations at <28 weeks. Fig. 2 graphically depicts the evolution of the systolic! diastolic ratio from the menstrual cycle through term pregnancy. The proliferative phase systolic!diastolic ratio was ~2.9 + 4.4 (SD) and was significantly different from the end-secretory phase measurements. No changes were seen in the first trimester. Early in the second trimester, the ratio rapidly declined to a level
of :%3, which is significantly different from the secretory first-trimester level of 8.3, p < 0.01. At 26 weeks, the vaginal and uterine studies were equal. The large standard deviation seen in the vaginal uterine artery studies before 30 weeks represents the discrepancies between the two arteries. It was not unusual for one artery to have a significantly lower ratio in the midtrimester. Presumably this represents a manifestation of one artery providing the predominant supply to the placenta. We have not been able to correlate this with placental site since such differences are seen in placentas implanted anteriorly, posteriorly, or fundally. There also appeared to be delayed artery compliance in some women (Figs. 3 to 5). Comment Uterine artery velocimetry appears to be able to identify whether a normal or abnormal circulation to the uterus is developing. When studies are extended through the menstrual cycle and throughout pregnancy, four distinctive phases can be identified. During the proliferative phase of the menstrual cycle the parauterine vessels appear similar to other arteries in the body; that is, there is a prominent notch and little or no end-diastolic velocity. Near the end of the secretory phase, vessel compliance changes begin and an end-diastolic component is usually present, yielding a systolic/diastolic ratio of approximately 8. Inasmuch as the end-diastolic component is a reflection of resistance distal to the point of measurement, it can be hypothesized that there is a dilatation and uncoiling of the spiral arteries and capillaries as is seen in a late secretory endometrium. When normal pregnancy occurs there are no changes seen during the first 10 weeks. There is persistent diastolic velocity and the ratio remains approximately 8. At the beginning of the second trimester there are dramatic changes. These are characterized by a sharp decrease in the ratio to a range of 3 to 2.6. At this point in gestation there are also differences when measurements are done abdominally versus vaginally. The ab-
Uterine artery compliance in pregnancy 1035
Volume 155 Number 5
Weeks of gestation
26 Mean ±SD
1.9 2.5t ±O.5
30
28
I
n
2
7
Mean ±SD
1.9 ±O.4 2.3 ±O.5
I
n
12
Mean ±SD
2.1
I
n 11
±O.3 4
2.2 ±O.4
Mean ±SD
2
I
n
21
±O.4
3
2.1
2
±O.3
dominal measurements are primarily those of the arcuate branches of the uterine artery whereas the vaginal studies are more likely coming from the proximal portion of the vessel. Thus the decrease in resistance as reflected in the velocity waveform is first seen in the arcuate vessels. The proximal uterine artery would represent the summation of all the distal branches and would be expected to be higher until the placental circulation is fully established. An important concomitant change that occurs during this phase is the disappearance of the early diastolic notch. These findings complement the studies of Brosens" and DeWolf et al.," who described the morphologic changes of the spiral arteries in the placental bed. In their studies they described a sequence of structural changes in the spiral arteries beginning in the decidua and followed by those in the myometrial portion of the vessels. These alterations in the spiral arteries result in "conversion of small muscular arteries to dilated, tortuous vessels whose walls are composed of a mixture of fibrinoid, hyalin, and fibrous tissue, with remnants of smooth muscle and elastic tissue." We interpret the changes in the early second trimester as reflecting those that occur in the myometrial segments of the vessel. From 20 to 26 weeks the diastolic notch should completely disappear. One exception is when one uterine artery seems to be providing the bulk of the blood supply to the placenta. We do not have sufficient data to analyze the consequences of delayed compliance in one vessel. Fig. 6 provides an example of a woman in whom normal uterine artery compliance failed to develop in both vessels.' The appearance of the uterine velocity waveform is quite instructive. A vessel of this caliber would not be expected to be able to constrict when cut, thus providing visual evidence of the potential of a uterine artery hemorrhage and the importance of the myometrium in controlling postpartum bleeding. Campbell et al." measured uteroplacental blood flow by using a pulsed Doppler technique. These studies were done after the thirty-fifth week with the aid of linear-array ultrasound to direct the Doppler signal. In
Mean ±SD
I
n
2 ±O.3
26
2.2
6
±O.3
Mean ±SD
2
I
n
19
Mean ±SD
2
I
n
13
±O.4
±O.3
2
40
38
36
34
32
2
2 ±O.3
18
Mean ±SD
I
n
1.9 ±O.2
14
2
11
±O.2
addition they used a more complex calculation, which they called the "frequency index profile." In this measurement the entire velocity waveform is plotted as a percentage of the mean frequency. In a group of women with abnormal pregnancies, the waveform showed a persistent diastolic notch and a reduced enddiastolic velocity. 8 Their waveforms appear comparable to ours and the ratios should be similar. Trudinger et al." calculated systolic/diastolic ratios of the uterine artery waveform. From 28 to 42 weeks the mean value ranged from 1.6 to 1.4. Their 95% value was reported to be 1.8, perhaps because they measured subplacental vessels only. They suggested that elevated ratios were associated with abnormal umbilical artery waveforms, growth retardation, and hypertension. Velocity is a component of volume flow but the two are not synonymous. Velocity can be altered by the impedance and area of the vascular bed, but volume flow may be unchanged because of compensation in intra-vascular pressure or redistribution of the blood. Sequential studies on uterine blood flow in the human are not possible at present, but a number of such studies have been performed in sheep. Greiss'" found a dramatic change in absolute flow beginning on approximately the forty-fifth day of ovine gestation. This is the onset of the second trimester in sheep. The velocity wave data portrayed in this study demonstrate that dramatic changes in compliance also appear in the early second trimester in women. Our velocity flow data show a change in the late secretory phase of the menstrual cycle and little change until the second trimester. Volume flow studies in the sheep have suggested quite different patterns, such as increased flow during estrus and variable uterine blood flow during the period of embryogenesis. Our data represent an initial effort to outline velocity changes during the menstrual cycle. The variations encountered indicated that more refined methodology will be necessary to clarify these changes. The essence of the capacity of a pregnancy to develop normally is that there be adequate nutrients and oxygen. The principal vehicles are the supply lines, the uterine and umbilical arteries. In previous reports it
1036 Schulman at al.
November 1986 AmJ ObstetGynecol
velocity waveforms adds additional support to our understanding of placentation and provides a foundation for our understanding of the pathophysiology of various disorders in pregnancy.
,.r:
~.~:~\
I!&, '. ",
~.
~:.. ~ :-- _ .
...--- - -- - -
18 WElXS
-
-_ _- --.
24 \nElS
t'MS ILtCAL S/ 'J - 7. 4
) 7 ....EEKS NON-COMPL IAJrlCE
Fig. 6. Failure of compliance to develop. The lowerright is the umbilical artery velocity waveform. This woman had hypertension, superimposed preeclampsia, and fetal growth retardation. SiD = Systolic/diastolic.
was shown that the normal and abnormal umbilical arterial circulation can be identified with a high degree of reliability. The studies of the uterine arteries provide the second part of the equation. This new technology reveals by the twenty-fourth week of pregnancy whether adequate placentation has occurred. Hence the combined study of the uterine and umbilical artery
REFERENCES J. Gosling RG, Dunbar G, King DH, et al. The quantitative analysis of occlusive peripheral arterial disease by a noninvasive technique. Angiology 1971;22:52-5. 2. Schulman H, Fleischer A, Stern W, Farmakides G,Jagani N, Blattner P. Umbilical velocity wave ratios in human pregnancy. AM J OBSTET GYNECOL 1984; 148:985-90. 3. Fleischer A, Schulman H, Farmakides G, Bracero L. Umbilical velocity waveforms in intrauterine growth retardation. AM J OBSTET GYNECOL 1985; 151:502-5. 4. Hamilton WF, Dow P, eds. Handbook of physiology, sect 2, circulation vol 2. Washington DC: American Physiologic Society, 1963:820, 842. 5. Brosens 1. Morphological changes in the utero-placental bed in pregnancy hypertension. Clin Obstet Gynecol 1977;4:573. 6. DeWolf F, DeWolf Peeters C, Brosens 1. Ultrastructure of the spiral arteries in the human placental bed at the end of normal pregnancy. AM J OBSTET GYNECOL 1973;117: 833-48. 7. Fleischer A, Schulman H, Farmakides G, et al. Uterine artery Doppler velocimetry in pregnant women with hypertension. AMJ OBSTET GYNECOL 1986;154:806-13. 8. Campbell S, Griffin DR, Pearce JM, et al. New doppler technique for assessing uteroplacental blood flow. Lancet 1983;1:676-7. 9. Trudinger BJ, Giles WB, Cook CM. Flow velocity waveforms in the maternal uteroplacental and fetal umbilical circulations. AMJ OBSTET GYNECOL 1985;152:155-63. 10. Greiss PC. Uterine and placental blood flow. In: Depp R, Eschenbach DA, SciarraJJ, eds. Gynecology & Obstetrics. Philadelphia: Harper & Row, 1983, chap 62.
Volume 155 Number 5
fetal death in twin pregnancy," Rapid decrease of estriol level, but not human placental lactogen level, in one patient after fetal death may indicate the usefulness of estriol values for the monitoring of the surviving fetus in such a case.
REFERENCES 1. Batra S, Sjoberg ND, Aberg A. Human placental lactogen,
estradiol-Izjs, and progesterone levels in the third trimester and their respective values for detecting twin pregnancy. AMl OBSTET GYNECOL 1978;131:69-72. 2. Gerhard I, Runnebaum B. Hormonal evaluation in multiple pregnancies. Geburtschilfe Frauenheilkd 1980;40: 101-15. 3. Duff GB, Brown lB. Urinary oestriol excretion in twin pregnancies. 1 Obstet Gynaecol Br Commonw 1974; 81:695"700.
4. Daw E. Human placental lactogen and twin pregnancy. Lancet 1977;2:299-300. 5. Trapp M, Bohnet HG, Weise HC, Leidenberger FA. Referenzbereiche fur Serum- Konzentrationen des humanen plazentaren Laktogens und des Ostriols von Frauen mit Zwillingsschwangerschaften. Lab Med 1984;8: 194-7. 6. Enbom JA. Twin pregnancy with intrauterine death of one twin. AM 1 OBSTET GYNECQL 1985;152:424-9. 7. Lilford R], Obiekwe BC, Chard T. Maternal blood levels of human placental lactogen in the prediction of fetal growth retardation: choosing a cut-off point between normal and abnormal. Br J Obstet GynaecoI1983;90:511-5. 8. Litschgi M, Stucki D. Course of twin pregnancies after foetal death in utero. Z Geburtshilfe Perinatol 1980;184: 227-30. 9. Hanna lH, Hill 1M. Single intrauterine fetal demise in multiple gestation. Obstet Gynecol 1984;63: 126-9.
Development of uterine artery compliance in pregnancy as detected by Doppler ultrasound Harold Schulman, M.D., Adiel Fleischer, M.D., George Farmakides, M.D., Luis Bracero, M.D., Burton Rochelson, M.D., and Lawrence Grunfeld, M.D. Mineola, Stony Brook, and Bronx, New York Uterine artery velocimetry was performed by means of a continuous-wave Doppler ultrasound. Serial studles were done on 12 nonpregnant and 79 normal pregnant women. Measurements were made on both uterinearteriesand averaged. From the proliteratlve phase of the menstrual cycle through 40 weeks of pregnancy, four developmental phases are described. These changes are based on the calculation of the systolic-diastolic ratio and the disappearance of the early diastolic notch in the velocity wave. The ratio after 26 weeks averaged 2 ± 0.3. These data provide a foundation for the study of pregnancies in which these normal evolutionary changes do not occur. (AM J OesTET GVNECOL 1986;155:1031-6.)
Key words: Doppler, uterine artery, pregnancy, compliance, systolic/diastolic ratio
The development of adequate uterine blood flow is believed to be essential for normal fetal growth and nutrition and for the avoidance of maternal hypertensive complications. Until now there was no simple and reliable technique for assessing the maternal and fetal placental circulation. Doppler ultrasound instrumentation provides a noninvasive method of studying the status of various blood vessels by providing a spectrum analysis of the velocities of the moving red blood cells.1
From theDepartments of Obstetrics and Gynecology, Winthrop University Hospital, Mineola, State University of New York at Stony Brook, and AlbertEinstein College of Medicine, Bronx. Received for puhlication February 5, 1986; revised June 5, 1986; accepted June 9, 1986. Reprint requests: HaroldSchulman, M.D., Department of Obstetrics and Gynecology, Winthrop-University Hospital, 259 First St., Mineola, NY 11501.
In previous reports we detailed the reliability of umbilical velocimetry for the evaluation of normal growth and development of the fetus and the identification of intrauterine growth retardation.v" In this report we describe our initial findings when velocimetry was performed on the uterine arteries in normal pregnancy and nonpregnant women. The changes in the uterine circulation in pregnancy can be attributed to a number of factors. The principal influencing elements are the effect of the hormones estrogen and progesterone, the changes in blood volume, intravascular pressure, and the change in peripheral resistance created by the development of the intervillous spaces. The summation of these changes are a reflection of vessel compliance; hence we have chosen to describe the changes seen with velocimetry as changes in compliance! 1031
November 1986 Am J Obstet Gynecol
1032 Schulman et al.
13 12
SID Ratio Phase I Prolif-Secretory n First trimester m2nd trimester IInate 2nd trimester and Third trimester
11 10
9
· L
ANTERIOR DIVISION OF INTERN AL ILIAC ARTERY EXTERNAL ILIAC
_ _ Vaginal
b--'" Abdom inal
.2 8
o
IX
7
e6
VI
5 4
p< .05(vs20wks)
3 2 Fig. 1. Velocity wave patterns in the parauterine vessels. The arcuate vessels have the highest diastolic frequency. Time frame of each panel is 1.6 seconds .
Table I. Uterine artery velocity ratios during the menstrual cycle* Systolic!diastolic ratios (mean ± SD)
Cycle days 10-13 (n = 8)
23 26-29 (n
=
8)
12.9 ± 4.4 8.0 ± 2.9 7.2 ± 3.2
*** I<---IIL-4
.
A.'~·fi't't"it:l.f'=lt;l~ I
I
I
I
I
I
I
12 16 20 24 28 32 36 40
WEEKS OF GESTATION
Fig. 2. Uterine artery compliance during the menstrual cycle and throughout pregnancy. Means ± SD. Circles represent measurements through the vaginal forn ices and the triangles, measurements across the abdomen. Each value represents the average between the left and right artery. SID = Systolic/diastolic.
(p = 0.003)
*Four additional women were studied and excluded because of inability to isolate vessel.
Mat~rial
and methods Studies were done on 12 nonpregnant and 79 pregnant women. Approval for this work was obtained from both institutional research review committees. Informed consent was obtained from each woman studied. The data obtained were not used for clinical therapeutic decisions. Continuous-wave Doppler studies were carried out with a Sonicaid flow probe, which has a calculated power density of 6.5 mW/cm2 • An analysis of the returning frequency signal was performed on an Angioscan spectrum analyzer displayed on a screen. Continuous recordings were stored on a JVC audio unit. The transducer used in these studies was a 4 MHz probe that has a beam diameter of 12 mm . . The majority of these studies were done without real time imaging of the vessels under study. It is our thesis that successful examinations can be performed by pattern recognition. In the early phases of this study confirrnation of vessel patterns was obtained by duplex imaging or by in vivo measurements obtained by using a 10 MHz probe during cesarean section. It became clear from experience and team review of the recorded tapes that patterns of iliac, uterine, and arcuate vessels could be confidently separated. Other vessels encoun-
tered in the pelvis have patterns that show no diastolic frequency and are easily excluded. The range depth of the probe of 8 em excludes the possibility of reaching the ovarian or renal arteries. Velocimetry studies were done abdominally and vaginally. Initially vaginal studies were done in the belief that the proximity to the uterine vessels would allow easier and more accurate examinations. We have learned that in the early second trimester and onward, reliable studies can be obtained abdominally. The pregnant uterine artery can be identified by the presence of velocity during diastole. Efforts were made to study both uterine arteries and the umbilical artery. Vaginal studies were done by inserting the transducer into the forefinger of a rubber glove with coupling jelly. The glove and transducer were inserted into the vaginal fornix with the aid of lubricating jelly and directed along the paracervical area. Abdominal studies were done by directing the transducer into the parauterine area in the region of the lower uterine segment. The probe was gently rotated until the appropriate pattern was identified. (Recall that the angle of incidence must be less than 90 degrees in order to produce a satisfactory signal.) Examinations were carried out in the dorsal lithotomy position in nonpregnant women or the left or right tilt position in pregnant subjects. Waveforms were acceptable for measurement when they were dense and of equal amplitude. An effort was made to obtain the largest complexes possible. The Angioscan
Uterine artery compliance in pregnancy
Volume 155 Number 5
DAY 1/ 23
DAY 1/13
1033
DAY 1/ 29
MENSTRUAL CYCLE
Fig. 3. Uterine artery velocity waves during the menstrual cycle. Note increased diastolic frequency return in secretory phase.
EARLY PItECSA.'IlCY
Fig. 4. Uterine artery velocimetry in first trimester. Note that there is little difference from that seen in secretory phase.
monitoring screen shows a 1.6-second time frame. The duration of an individual study was 5 to 15 minutes. Successful imaging of the uterine vessels depended on the stage of pregnancy. From 14 weeks onward at least one vessel could be studied unless there was significant obesity. More difficulty was encountered in the first trimester, and there was a 10% failure rate. A minimum of four waveforms were measured on each side, and the values were averaged. The mean intraobserver and interobserver error between examinations was found to be 4% with a standard deviation of 2.3%. In the studies on the nonpregnant vessels, one third were considered unsatisfactory. Calculations were made by measuring the systolic! diastolic ratio. This computation circumvents the problem of not knowing the angle of insonation. When appropriate, significant differences were calculated by estimating multiples of standard deviations and by the Student t test.
VAG INAL
ABDOMINAL
RT
SID· 4 .2
sID · 2.9
sID · 2. 9
s ID· 2.4
LT
Results A continuous-wave Doppler system provided a spectrum analysis of all vessels that were crossed. Three characteristic vessels were seen in the parauterine area (Fig. 1). Those immediately adjacent to the uterus showed the lowest ratios and were believed to be the arcuates and the distal uterine branches. The proximal
Fig. 5. Abdominal versus vaginal measurements. The placental blood supply is predominantly on the left; thus the ratios are lower and the notching has disappeared. Note closeness of the ratios. Pregnancy of 24 weeks. SID = Systolic/diastolic.
1034 Schulman at al.
November
Am
1986
J Obstet Gynecol
Table II. Uterine artery systolic/diastolic ratios weeks of gestation Weeks of gestation
10
8 Study
Mean ±SD
I
n
Mean ±SD
14
In
Abdominal Vaginal
9.6 ±3
3
8.3 ±4
4
Mean ±SD
2.4 ±0.1 4.1* ±0.95
I
n
6 4
Mean ±SD
1.8 ±0.6 2.8 ± 1.3
n
7 3
Mean ±SD
1.9 ±0.6 3.8 ±0.95
22
20
18
16
In 7 6
Mean ±SD
2.3 ±0.2 3 ±0.9
I
n
6 12
Mean ±SD
2.5 ±0.6 2.7 ± 1.5
24
I
n
8 7
Mean ±SD
2.2 ±0.4 2.7
I
n
8 3
*p < 0.05 from 10 weeks. tp < 0.01 from 14 weeks.
uterine branch seemed to have a higher systolic component, and this vessel was commonly seen during the vaginal studies. The external iliac is a characteristic high-resistance artery with no end-diastolic velocity and a prominent notch. On occasion a pelvic vein was also seen. Table I provides data on systolic/diastolic ratios in eight nonpregnant women. An additional four women in whom studies were attempted were considered unsatisfactory subjects because of the inability to obtain satisfactory scans. Vessel recognition is dependent on the identification of diastolic velocity. In these four women a diastolic component was not consistently present until the end of the secretory phase. This may indicate a deficiency of the methodology or that these women had a higher uterine artery resistance than the others. All 12 women were nulliparous. Ovulation was confirmed by plasma progesterone levels. The data are presented to demonstrate apparent trends but should be viewed with caution. It appears that during the proliferative phase the vessel remains in a state of high resistance, but the ratio declines in the secretory phase when there is a diastolic velocity. The ratios in the endsecretory phase were significantly lower than those seen in the late proliferative phase. Table II summarizes the measurements in the pregnant populations. All women in these studies had a normal pregnancy outcome. The upper part of the table represents the studies made across the abdomen; the lower part depicts those made through the vaginal fornices. The values represent the average of both the right and the left uterine artery. The studies done vaginallyhad higher mean values in gestations at <28 weeks. Fig. 2 graphically depicts the evolution of the systolic! diastolic ratio from the menstrual cycle through term pregnancy. The proliferative phase systolic!diastolic ratio was ~2.9 + 4.4 (SD) and was significantly different from the end-secretory phase measurements. No changes were seen in the first trimester. Early in the second trimester, the ratio rapidly declined to a level
of :%3, which is significantly different from the secretory first-trimester level of 8.3, p < 0.01. At 26 weeks, the vaginal and uterine studies were equal. The large standard deviation seen in the vaginal uterine artery studies before 30 weeks represents the discrepancies between the two arteries. It was not unusual for one artery to have a significantly lower ratio in the midtrimester. Presumably this represents a manifestation of one artery providing the predominant supply to the placenta. We have not been able to correlate this with placental site since such differences are seen in placentas implanted anteriorly, posteriorly, or fundally. There also appeared to be delayed artery compliance in some women (Figs. 3 to 5). Comment Uterine artery velocimetry appears to be able to identify whether a normal or abnormal circulation to the uterus is developing. When studies are extended through the menstrual cycle and throughout pregnancy, four distinctive phases can be identified. During the proliferative phase of the menstrual cycle the parauterine vessels appear similar to other arteries in the body; that is, there is a prominent notch and little or no end-diastolic velocity. Near the end of the secretory phase, vessel compliance changes begin and an end-diastolic component is usually present, yielding a systolic/diastolic ratio of approximately 8. Inasmuch as the end-diastolic component is a reflection of resistance distal to the point of measurement, it can be hypothesized that there is a dilatation and uncoiling of the spiral arteries and capillaries as is seen in a late secretory endometrium. When normal pregnancy occurs there are no changes seen during the first 10 weeks. There is persistent diastolic velocity and the ratio remains approximately 8. At the beginning of the second trimester there are dramatic changes. These are characterized by a sharp decrease in the ratio to a range of 3 to 2.6. At this point in gestation there are also differences when measurements are done abdominally versus vaginally. The ab-
Uterine artery compliance in pregnancy 1035
Volume 155 Number 5
Weeks of gestation
26 Mean ±SD
1.9 2.5t ±O.5
30
28
I
n
2
7
Mean ±SD
1.9 ±O.4 2.3 ±O.5
I
n
12
Mean ±SD
2.1
I
n 11
±O.3 4
2.2 ±O.4
Mean ±SD
2
I
n
21
±O.4
3
2.1
2
±O.3
dominal measurements are primarily those of the arcuate branches of the uterine artery whereas the vaginal studies are more likely coming from the proximal portion of the vessel. Thus the decrease in resistance as reflected in the velocity waveform is first seen in the arcuate vessels. The proximal uterine artery would represent the summation of all the distal branches and would be expected to be higher until the placental circulation is fully established. An important concomitant change that occurs during this phase is the disappearance of the early diastolic notch. These findings complement the studies of Brosens" and DeWolf et al.," who described the morphologic changes of the spiral arteries in the placental bed. In their studies they described a sequence of structural changes in the spiral arteries beginning in the decidua and followed by those in the myometrial portion of the vessels. These alterations in the spiral arteries result in "conversion of small muscular arteries to dilated, tortuous vessels whose walls are composed of a mixture of fibrinoid, hyalin, and fibrous tissue, with remnants of smooth muscle and elastic tissue." We interpret the changes in the early second trimester as reflecting those that occur in the myometrial segments of the vessel. From 20 to 26 weeks the diastolic notch should completely disappear. One exception is when one uterine artery seems to be providing the bulk of the blood supply to the placenta. We do not have sufficient data to analyze the consequences of delayed compliance in one vessel. Fig. 6 provides an example of a woman in whom normal uterine artery compliance failed to develop in both vessels.' The appearance of the uterine velocity waveform is quite instructive. A vessel of this caliber would not be expected to be able to constrict when cut, thus providing visual evidence of the potential of a uterine artery hemorrhage and the importance of the myometrium in controlling postpartum bleeding. Campbell et al." measured uteroplacental blood flow by using a pulsed Doppler technique. These studies were done after the thirty-fifth week with the aid of linear-array ultrasound to direct the Doppler signal. In
Mean ±SD
I
n
2 ±O.3
26
2.2
6
±O.3
Mean ±SD
2
I
n
19
Mean ±SD
2
I
n
13
±O.4
±O.3
2
40
38
36
34
32
2
2 ±O.3
18
Mean ±SD
I
n
1.9 ±O.2
14
2
11
±O.2
addition they used a more complex calculation, which they called the "frequency index profile." In this measurement the entire velocity waveform is plotted as a percentage of the mean frequency. In a group of women with abnormal pregnancies, the waveform showed a persistent diastolic notch and a reduced enddiastolic velocity. 8 Their waveforms appear comparable to ours and the ratios should be similar. Trudinger et al." calculated systolic/diastolic ratios of the uterine artery waveform. From 28 to 42 weeks the mean value ranged from 1.6 to 1.4. Their 95% value was reported to be 1.8, perhaps because they measured subplacental vessels only. They suggested that elevated ratios were associated with abnormal umbilical artery waveforms, growth retardation, and hypertension. Velocity is a component of volume flow but the two are not synonymous. Velocity can be altered by the impedance and area of the vascular bed, but volume flow may be unchanged because of compensation in intra-vascular pressure or redistribution of the blood. Sequential studies on uterine blood flow in the human are not possible at present, but a number of such studies have been performed in sheep. Greiss'" found a dramatic change in absolute flow beginning on approximately the forty-fifth day of ovine gestation. This is the onset of the second trimester in sheep. The velocity wave data portrayed in this study demonstrate that dramatic changes in compliance also appear in the early second trimester in women. Our velocity flow data show a change in the late secretory phase of the menstrual cycle and little change until the second trimester. Volume flow studies in the sheep have suggested quite different patterns, such as increased flow during estrus and variable uterine blood flow during the period of embryogenesis. Our data represent an initial effort to outline velocity changes during the menstrual cycle. The variations encountered indicated that more refined methodology will be necessary to clarify these changes. The essence of the capacity of a pregnancy to develop normally is that there be adequate nutrients and oxygen. The principal vehicles are the supply lines, the uterine and umbilical arteries. In previous reports it
1036 Schulman at al.
November 1986 AmJ ObstetGynecol
velocity waveforms adds additional support to our understanding of placentation and provides a foundation for our understanding of the pathophysiology of various disorders in pregnancy.
,.r:
~.~:~\
I!&, '. ",
~.
~:.. ~ :-- _ .
...--- - -- - -
18 WElXS
-
-_ _- --.
24 \nElS
t'MS ILtCAL S/ 'J - 7. 4
) 7 ....EEKS NON-COMPL IAJrlCE
Fig. 6. Failure of compliance to develop. The lowerright is the umbilical artery velocity waveform. This woman had hypertension, superimposed preeclampsia, and fetal growth retardation. SiD = Systolic/diastolic.
was shown that the normal and abnormal umbilical arterial circulation can be identified with a high degree of reliability. The studies of the uterine arteries provide the second part of the equation. This new technology reveals by the twenty-fourth week of pregnancy whether adequate placentation has occurred. Hence the combined study of the uterine and umbilical artery
REFERENCES J. Gosling RG, Dunbar G, King DH, et al. The quantitative analysis of occlusive peripheral arterial disease by a noninvasive technique. Angiology 1971;22:52-5. 2. Schulman H, Fleischer A, Stern W, Farmakides G,Jagani N, Blattner P. Umbilical velocity wave ratios in human pregnancy. AM J OBSTET GYNECOL 1984; 148:985-90. 3. Fleischer A, Schulman H, Farmakides G, Bracero L. Umbilical velocity waveforms in intrauterine growth retardation. AM J OBSTET GYNECOL 1985; 151:502-5. 4. Hamilton WF, Dow P, eds. Handbook of physiology, sect 2, circulation vol 2. Washington DC: American Physiologic Society, 1963:820, 842. 5. Brosens 1. Morphological changes in the utero-placental bed in pregnancy hypertension. Clin Obstet Gynecol 1977;4:573. 6. DeWolf F, DeWolf Peeters C, Brosens 1. Ultrastructure of the spiral arteries in the human placental bed at the end of normal pregnancy. AM J OBSTET GYNECOL 1973;117: 833-48. 7. Fleischer A, Schulman H, Farmakides G, et al. Uterine artery Doppler velocimetry in pregnant women with hypertension. AMJ OBSTET GYNECOL 1986;154:806-13. 8. Campbell S, Griffin DR, Pearce JM, et al. New doppler technique for assessing uteroplacental blood flow. Lancet 1983;1:676-7. 9. Trudinger BJ, Giles WB, Cook CM. Flow velocity waveforms in the maternal uteroplacental and fetal umbilical circulations. AMJ OBSTET GYNECOL 1985;152:155-63. 10. Greiss PC. Uterine and placental blood flow. In: Depp R, Eschenbach DA, SciarraJJ, eds. Gynecology & Obstetrics. Philadelphia: Harper & Row, 1983, chap 62.