We investigated whether hypoxemia without acidemia affects ductus venosus (DV) blood velocity waveform pattern in sheep fetuses with intact placenta and whether worsening acidemia and impending fetal death are related to changes in DV velocimetry in fetuses with increased placental vascular resistance. A total of 34 fetuses were instrumented at 115–136/145 days of gestation. Placental embolization was performed in 22 fetuses on the fourth postoperative day, 24 h before the experiment. The control group was comprised of 12 fetuses with intact placenta. The experimental protocol consisted of fetal hypoxemia that was induced by replacing maternal inhaled oxygen with medical air. To further deteriorate fetal oxygenation and blood-gas status, uterine artery volume blood flow was reduced by maternal hypotension. Fetuses that underwent placental embolization were divided into two groups according to fetal outcome. Group 1 consisted of 12 fetuses that completed the experiment, and group 2 comprised 10 fetuses that died during the experiment. DV pulsatility index for veins (PIV) and fetal cardiac outputs (COs) were calculated. Placental volume blood flow, fetal blood pressures, and acid base and lactate values were monitored invasively. On the experimental day, the mean gestational age did not differ significantly between the groups. In groups 1 and 2, the baseline mean DV PIV and fetal COs were not statistically significantly different from the control group. In the control group, the DV PIV values increased significantly with hypoxemia. In groups 1 and 2, the DV PIV values did not change significantly, even with worsening acidemia and imminent fetal death in group 2. During the experiment, the fetal COs remained unchanged. We conclude that fetal hypoxemia increases the pulsatility of DV blood velocity waveform pattern. In fetuses with elevated placental vascular resistance, DV pulsatility does not increase further in the presence of severe and worsening fetal acidemia and impending fetal death.
- Doppler ultrasound
- placental insufficiency
- fetal physiology
fetal oxygen tension is an important regulator of ductus venosus (DV) diameter (14). Hypoxemia leads to the dilatation of DV and an increase in the volume blood flow shunting through it (14, 20). Studies on fetal sheep in early pregnancy have shown increased pulsatility in DV blood velocity waveform pattern during hypoxemia (13). Another mechanistic pathway leading to increased pulsatility in Doppler-derived DV blood velocity waveform can be a rise in end-diastolic pressure of the ventricles (11). This leads to elevated atrial and central venous pressures, resulting in decreased venous forward flow throughout diastole.
In human fetuses with placental insufficiency and intrauterine growth restriction, a significant negative correlation between umbilical venous pH and DV pulsatility index (PI) value for veins (PIV) has been found (12). In fetal hypoxemia and acidemia, there is a decrease in DV velocity during atrial contraction, and in severe cases, this velocity can be reduced to zero or there may even be a reversal of blood flow. It has also been shown that the combination of inferior vena cava, DV, and umbilical venous Doppler parameters correctly predict acid-base status in a significant proportion of intrauterine growth restricted neonates (3).
We have developed a sheep model of increased placental vascular resistance. In this experimental setting, we tested the hypothesis that acute and severe fetal acidemia significantly increases the pulsatility of DV blood velocity waveform pattern. Specifically, we asked the following questions: 1) Does fetal hypoxemia without acidemia lead to increased pulsatility in the DV blood velocity waveform pattern in fetuses with intact placental circulation, and 2) is worsening acidemia and impending fetal death related to changes in the DV blood velocity waveform pattern in fetuses with increased placental vascular resistance?
MATERIALS AND METHODS
Thirty-four sheep with singleton pregnancies were instrumented at 115–136 days (term, 145 days) of gestation (Table 1). All experiments were performed in accordance with the guidelines of the European Convention for the Protection of Vertebrate Animals used for Experimental and Other Scientific Purposes (1986) and in compliance with the European Union Directive 86/609/EEC (1997). The research protocol was approved by the Animal Care and Use Committee of the University of Oulu (Oulu, Finland).
The detailed surgical procedure has been described previously (9). Briefly, a laparotomy was performed under general anesthesia maintained with inhaled isoflurane-air-oxygen mixture. A 6-mm transit-time ultrasonic flow probe (Transonic Systems, Ithaca, NY) was secured around the uterine artery (UA), supplying the pregnant uterine horn for UA volume blood flow (QUtA) measurements. Thereafter, the fetal lower body was exteriorized through a hysterotomy, and 18-gauge polyurethane catheters were introduced into the descending aorta and the inferior vena cava for blood pressure monitoring and blood sampling. A 4-mm transit-time ultrasonic flow probe (Transonic Systems) was secured around the umbilical arteries for placental volume blood flow (QUA) measurement. After the replacement of amniotic fluid by saline, the surgical incisions were closed. The catheters and probes were exteriorized on the ewe's flank. During a 5-day recovery period, postoperative analgesia and antibiotic treatment were provided.
Placental embolization was performed in 22 cases on the fourth postoperative day, using 45–150-μm microspheres (Contour Emboli, Target Therapeutics, Fremont, CA). A dry volume of 0.25-ml microspheres was suspended in 0.5 ml 20% albumin and diluted with 5 ml sterile saline solution. Bolus doses of 1 ml were injected via the femoral artery catheter into the descending aorta every 15 min until fetal arterial oxygen saturation decreased by ∼30% from the preembolization value (9).
Data collection was carried out on the fifth postoperative day and 24 h after placental embolization. General anesthesia was induced by propofol (4–7 mg/kg) and maintained with isoflurane (1–1.5%) in an oxygen-air mixture. Muscle relaxation was achieved by 20 mg rocuronium. Tidal volume, respiratory rate, and gas concentrations were adjusted to maintain normoventilation. Maternal jugular vein (Criticath SP5107H, Becton Dickinson, Sandy, UT) and descending aorta (16-gauge polyurethane catheter) were cannulated for blood pressure measurements. An epidural catheter was placed into the epidural space above the lumbosacral junction. During the experiment, Ringer solution was infused at a rate of 100 ml/h.
The animals were allowed to stabilize for 30 min before the baseline measurements were obtained. Thereafter, hypoxemia was induced by replacing maternal inhaled oxygen with medical air until maternal oxyhemoglobin saturation reached 80–90%. After a 30-min hypoxemia period, measurements during the hypoxemia phase were obtained. To further deteriorate fetal oxygenation and blood-gas status, QUtA was reduced by maternal hypotension, allowing maternal systolic blood pressure to decrease at least 30%. Maternal hypotension was achieved by administering 0.5% bupivacaine through an epidural catheter. After a 20–30-min period of maternal hypotension, the hypotension-phase measurements were obtained. At the end of the experiment, the ewe and the fetus were euthanized with an intravenous overdose of pentobarbital sodium, and fetal weights were determined. Data collection was performed at the end of the each phase, and for the final analysis, the measurements obtained at the baseline and the last phases were used.
Invasive data acquisition.
UtA and QUA
Ultrasonographic data acquisition.
Doppler ultrasonographic recordings (Acuson Sequoia 512, Mountain View, CA) from DV and UA were obtained. The lowest high-pass filter level was used. The mean values for DV PIV and UA PI were derived from three consecutive cardiac cycles. In addition, DV preload index [PLI; (systolic peak velocity − velocity during atrial contraction)/systolic peak velocity] and peak velocity index for veins [PVIV; (systolic peak velocity − velocity during atrial contraction)/diastolic peak velocity] were calculated (3). To calculate fetal cardiac outputs, pulmonary and aortic valve diameters were measured from frozen real-time images during systole by using the leading edge-to-leading edge method (19). Mean values of three separate measurements were used to calculate the cross-sectional area [cross-sectional area = π(diameter/2)2] of the valve. The calculation of the cross-sectional area of the valve was based on the assumption that the cross sections of the valves were circular. From the blood flow velocity waveforms across the pulmonary and aortic valves, time-velocity integrals were obtained by planimetry of the area underneath the Doppler spectrum. An angle < 15° between the vessel and Doppler beam was accepted for analysis. Volumetric blood flows (Q) were calculated (Q = cross-sectional area × time-velocity integral × fetal heart rate). Right and left ventricular cardiac outputs were determined, and their sum is the combined cardiac output. All the Doppler studies, which were videotaped for later analysis, were performed by one investigator (J. Räsänen).
Data were analyzed using SPSS for Windows version 15.0 (SPSS, Chicago, IL). Analysis of variance was used to evaluate differences between the groups (between-subjects P), changes in measurements during the protocol (within-subjects P), and differences in changes during the protocol between the groups (interaction P). Post hoc multiple comparisons between observed means in the three groups were performed using Scheffé's test. Possible differences between baseline and the last phase in the three groups were tested by using a paired sample t-test if the within-subject P reached statistical significance. The Spearman rank test was employed to show the relationship between the change (Δ) in DV PIV values and changes in fetal invasive parameters during the protocol. A two-tailed P value of <0.05 was used as the level of statistical significance. Values are presented as means (SD) and median (range).
Twelve fetuses with placental embolization completed the experimental protocol (group 1), and 10 fetuses with embolized placenta died during the experiment (group 2), 10–35 min after the last data collection. All the fetuses with intact placenta (n = 12) completed the experimental protocol (control group). On the day of the experiment, the mean gestational age in the control group was 125 (SD 5) and 123 (SD 7) and 124 (SD 1) days in groups 1 and 2, respectively.
Mean maternal weights were 73 (SD 7) and 70 (SD 19) kg in the control group and group 1. In group 2, the maternal weight of 56 (SD 5) kg was less than (P < 0.05) in the control group. Mean fetal weight was 2,597 (SD 704) g in the control group. In groups 1 [1,827 (SD 532) g] and 2 [1,316 (SD 95) g], fetal weights were less than in the control group. At baseline, fetuses in groups 1 and 2 had lower (P < 0.05) Po2 values than the control group fetuses. In group 2, pH values were also decreased (P < 0.05) compared with those in group 1 and the control group. Furthermore, in groups 1 and 2, lactate concentrations at baseline were higher (P < 0.05) than in the control group. In the control group, fetal pH values remained stable throughout the experiment, whereas in groups 1 and 2, pH values decreased significantly (Fig. 1). In addition, Po2 decreased and lactate concentrations increased significantly in each group during the experiment (Table 1).
In groups 1 and 2, weight-indexed QUA was lower (P < 0.05) and placental vascular resistance higher (P < 0.05) at baseline than in the control group. Fetal heart rate and arterial blood pressures did not differ significantly between the groups at baseline. However, in group 2, systemic venous pressure was higher (P < 0.05) than in the control group and group 1. During the experiment, weight-indexed QUA and placental vascular resistance did not result in a statistically significant change. In the control group, fetal systolic arterial and systemic venous pressures increased (P < 0.05), whereas in groups 1 and 2, they remained unchanged (Table 2).
Mean DV PIV values did not differ significantly between the groups at baseline. However, DV PLI and PVIV were significantly higher in group 2 than in the control group (Table 2). In groups 1 and 2, UA PI values were increased (P < 0.05) compared with those in the control group. In groups 1 and 2, weight-indexed cardiac outputs were not statistically, significantly different from those in the control group. In the control group, DV PIV and UA PI values increased significantly during the experiment. However, in groups 1 and 2, DV PIV and UA PI values did not change significantly (Fig. 2). No statistically significant changes were found in DV PLI and PVIV in any of the groups during the experiment (Table 2). A qualitative analysis of UA blood velocity waveform pattern demonstrated forward diastolic blood flow in every fetus during the experiment. In the DV blood velocity waveform profile, one fetus in group 2 developed reverse flow with atrial contraction during the last phase of the experiment. Weight-indexed cardiac outputs did not differ significantly from the baseline values during the experiment (Fig. 3). No significant correlation was found between ΔDV PIV and ΔpH (R = −0.207, P = 0.26) or ΔPo2 (R = −0.068, P = 0.72).
We found that in sheep fetuses with an intact placental circulation, fetal hypoxemia without acidemia led to a significant increase in the pulsatility of DV blood velocity waveform. Our finding is in agreement with published experimental studies (13, 22). It has been shown that a comparable degree of fetal hypoxemia with our study caused a profound and immediate distension of the DV inlet (14, 22). This is one of the mechanistic pathways to increase blood flow shunting across DV, thus optimizing the oxygen supply to the myocardium and brain. Hypoxemia-related dilatation of DV leads to decreased DV blood velocities, especially during atrial contraction, which in turn increases DV PIV values (13). A more profound decrease in DV blood flow velocity during the A wave has been suggested to be a consequence of enhanced atrial contractility during fetal hypoxemia. In addition, these fetuses demonstrated a significant increase in the systemic venous pressure during the experiment, which can also contribute to a rise in the pulsatility of the DV blood velocity waveform.
In fetuses with placental embolization and increased placental vascular resistance, Po2 values further decreased during the experiment. In addition, fetuses that died during the experiment had already significantly lower pH values at baseline, which decreased significantly during the experiment. However, DV PIV, PLI, and PVIV values remained similar to the baseline level, even 10–35 min before fetal death. In human fetuses with intrauterine growth restriction and abnormal UA Doppler findings, it has been shown that inferior vena cava, DV, and umbilical vein Doppler parameters were predictive of an umbilical artery pH < 7.20 (3). However, none of the venous Doppler parameters was significant predictors of severe (pH < 7.0 and/or base excess below −13) fetal metabolic compromise.
In the present study, we found that fetuses were able to maintain their weight-indexed cardiac outputs during the experiment. Even the fetuses with worsening acidemia and impending fetal death kept their right, left, and combined cardiac outputs comparable with those at baseline. In addition, fetal arterial blood pressures did not change significantly in fetuses with placental embolization and increased placental vascular resistance. It is known that an increase in ventricular end-diastolic pressure can lead to a rise in right atrial and central venous pressures. Finally, this can lead to an abnormal DV blood velocity waveform pattern, especially during atrial contraction. Previously, we have demonstrated in our fetal sheep model that acute metabolic acidosis impaired myocardial contractility during the isovolumic phase and relaxation during the isovolumic and early filling phases of the cardiac cycle obtained by the tissue-Doppler method (2). However, despite severe acidemia, they were able to maintain their cardiac output. In addition, the results of the present study showed that in fetuses with worsening acidemia, fetal systemic venous pressure did not increase during the experiment. We propose that in fetal acute acidemia, the DV blood velocity waveform pattern becomes abnormal when fetal cardiac output fails or ventricular end-diastolic pressure increases, leading to a rise in systemic venous pressure. However, it appears that this pathophysiological process evolves very rapidly, and we were unable to predict it even 10–35 min before fetal death.
Previously, we have shown that in human pregnancies complicated by placental insufficiency and intrauterine fetal growth restriction, fetuses with biochemical evidence of cardiac dysfunction and myocardial cell damage were able to maintain their cardiac outputs, even in the presence of severely abnormal DV blood velocity waveform pattern with reversed blood flow during atrial contraction. In addition, these human fetuses showed no signs of acidemia at the time of delivery (17). On the other hand, in these pregnancies, placental insufficiency was severe with significantly elevated cardiac afterload. A rise in afterload can increase ventricular end-diastolic pressure, which leads to abnormal DV blood velocity waveform during atrial contraction. In our present experimental study, placental vascular resistance did not increase with statistical significance, even in fetuses that died during the experiment, suggesting that the ventricular afterload was not markedly altered. We propose that fetal acidemia alone does not significantly contribute to the development of abnormal DV blood velocity waveform pattern. This is also supported by the fact that the change in fetal pH did not correlate with the change in DV PIV value during the experiment. However, if fetal acidemia is related to a pregnancy complication that also disturbs fetal cardiac function leading to a rise in ventricular end-diastolic pressures, an increased pulsatility in DV blood velocity waveform pattern can be found in the presence of fetal acidemia.
In the present study, baseline DV PLI and PVIV values were higher in group 2 fetuses than in the control group. This could be a consequence of elevated systemic venous pressure. However, these indexes were not affected by worsening fetal acidemia and impending fetal death, suggesting that they are not more informative in the clinical assessment of the fetus than DV PIV.
DV is an important shunt in fetal circulation allowing oxygen rich blood to bypass the liver and to stream toward foramen ovale and left atrium and ventricle (4, 5, 7, 10, 15, 20). Our results demonstrate that fetal hypoxemia can increase pulsatility in DV blood velocity waveform pattern. However, it appears that it cannot recognize those ovine fetuses that will become acidemic and even die within a short time period. We propose that the development of abnormal DV blood flow pattern requires additional pathophysiological events that lead to increased ventricular end-diastolic and systemic venous pressures. A recognition of these pathways might better predict the development of an abnormal DV waveform pattern.
Our experimental sheep model contains limitations. The surgical procedures may constitute a significant stress to the examined fetuses. However, the 5-day recovery period after surgery should be enough for the recovery of fetal myocardial function (8). With regard to the use of general anesthesia, our previous data show that uterine and umbilical artery volume blood flows before and after the induction of general anesthesia are similar, suggesting conditions close to the physiological circulatory state (1). Although isoflurane may modify fetal cardiovascular regulation, newborn lambs under isoflurane anesthesia can increase cardiovascular performance during stress (6). Validation studies in fetal sheep have shown that invasive and Doppler echocardiographic volume flow calculations correlate well (21). In addition, the intraobserver variabilities of the Doppler ultrasonographic parameters of fetal sheep cardiovascular hemodynamics have been shown to be comparable with those in previous human fetal studies during the second half of gestation (16, 18).
We conclude that in this experimental sheep model, fetal hypoxemia increases the pulsatility of the DV blood velocity waveform pattern. In fetuses with elevated placental vascular resistance, the pulsatility of DV velocimetry does not increase further in the presence of severe and worsening fetal acidemia and impending fetal death.
This study was supported by the University Hospital of Oulu, Regional Health Authority of Northern Norway, University of South Florida, All Children's Hospital Foundations, The Sigrid Juselius Foundation, and The Academy of Finland.
No conflicts of interest are declared by the author(s).
- Copyright © 2010 the American Physiological Society