Objective
The purpose of this study was to examine the presence and frequency of antegrade late diastolic arterial blood flow (ALDAF) in the fetus and to determine its contribution to cardiac output.
Study Design
We evaluated the presence of ALDAF in 457 fetal and 21 postnatal echocardiograms. The timing of ALDAF to the ventricular systolic Doppler recording (ALDAF-V) was compared with the mechanical atrioventricular interval and, in neonates, the electrical PR interval. Velocity time integrals of ALDAF and the ventricular systolic Doppler signals were measured, and the percent contribution of ALDAF was calculated.
Results
ALDAF was observed in 365 of 457 studies and included all <11 weeks’ gestations. Strong correlation between ALDAF-V, atrioventricular interval, and electrical PR interval suggests that ALDAF coincides with atrial contraction. ALDAF contributed substantially to cardiac output in early gestation with later decline.
Conclusion
ALDAF results from atrial contraction. Increasing gestational age results in less ALDAF, and reduced ALDAF contribution to cardiac output likely due to improved diastolic function.
Fetal echocardiography has provided insight into the human fetal circulation including the evolution of myocardial function, even from the first trimester. For instance, based on Doppler recordings, ventricular filling in the first trimester is of short duration, occurs largely through atrial systole, and is coupled with longer isovolumic relaxation and contraction times. After 10 weeks’ gestation, ventricular filling develops a biphasic pattern with an increasing contribution of early diastolic blood flow and significantly shorter isovolumic relaxation times with advancing gestational age. It has been suggested that this change in filling pattern is related to changes in diastolic function of the fetal myocardium. It also suggests that atrial function, through its influence on ventricular preload, may be very important for the maintenance of cardiac output in the normal fetus.
While performing routine early fetal echocardiograms, we recently discovered the presence of antegrade late diastolic arterial blood flow (ALDAF) during spectral Doppler sampling in the fetal great arteries ( Figure 1 ). ALDAF was demonstrated to occur just before the opening of the aortic and pulmonary valves at the end of diastole. To our knowledge, ALDAF has not been described previously in the human first trimester. Late diastolic forward blood flow in the fetal pulmonary artery has been observed previously in fetuses at >20 weeks’ gestation, but determination of its physiologic cause has not been elucidated. Late diastolic blood flow in the aorta has not been described previously in the human fetus; the changes in ALDAF occurrence, magnitude, and timing through gestation and in the neonatal period have not been investigated previously. In the postnatal setting, the Doppler finding of late diastolic antegrade blood flow in the pulmonary artery is a known feature of restrictive right ventricular physiologic condition. Generated by atrial contraction, this Doppler signal that has been observed in affected pediatric and adult patients is the result of the poorly compliant right ventricle acting as a passive conduit between the atrium and great vessel. As previously confirmed by invasive catheterization, the finding of ALDAF in the right ventricular outflow occurs when the right ventricular end-diastolic pressure equals or exceeds the diastolic pulmonary artery pressure.
In the present investigation, we sought to examine the presence and frequency of ALDAF in normal human fetuses over the course of gestation and to evaluate its contribution to cardiac output, which is a potential new role of atrial function in the fetus. We hypothesized that ALDAF represents blood flow during atrial contraction and that it occurs more frequently and contributes more substantially to ventricular output at earlier gestational ages as a likely consequence of the less compliant nature of the early fetal myocardium.
Materials and Methods
After research ethics board approval and informed consent from participants, we prospectively performed 457 fetal echocardiograms on 237 pregnant women with healthy pregnancies between 6 and 38 weeks’ gestation for the presence of ALDAF. All fetal echocardiograms that were included in this study were performed from February 2009 to May 2011 in the Fetal & Neonatal Cardiology Program of the University of Alberta in Edmonton, Canada. Initial fetal echocardiograms were performed between 6-10 weeks’ gestation with follow-up studies performed between 10-14 weeks’ gestation and again after 17 weeks’ gestation to confirm cardiac normalcy. For those women who were enrolled at >10 weeks’ gestation, only 2 fetal echocardiograms were performed: one between 10-14 weeks and then at >17 weeks’ gestation. All first-trimester echocardiograms at <12 weeks’ gestation included a short (5-10 minute) period of 2 dimensional imaging and very limited periods of Doppler interrogation (10 seconds of sampling per assessment) to reduce exposure of the early fetus to ultrasound waves. The lowest possible intensities of Doppler energy were used (thermal and mechanical indices of ≤1 in >95% of cases). Fetuses with identified structural or functional cardiac disease or extracardiac disease and fetuses in pregnancies that were complicated by maternal disease, including diabetes mellitus, which potentially could influence fetal and placental vascular function, were excluded.
All fetal echocardiograms were attempted transabdominally with additional transvaginal ultrasound imaging for the early examinations only when satisfactory transabdominal views could not be obtained. Both sagittal and axial cardiac planes were used to obtain optimal Doppler samples. We used an 8-6 MHz or a 5-1 MHz curved array transducer for transabdominal imaging and a 12-6 MHz endovaginal transducer for transvaginal imaging on a General Electric Voluson E8 imaging system (General Electric, Milwaukee, WI). Pulsed Doppler recordings that were performed on each fetus included (1) the main pulmonary artery with the pulsed Doppler sample placed just above the pulmonary valve leaflets, (2) the aorta with the pulsed Doppler sample placed just above the aortic valve leaflets, and (3) the mitral valve inflow and aortic outflow Doppler pattern with the pulsed Doppler sample placed in the left ventricular outflow tract below the level of the mitral valve leaflets. Low wall filters were used to optimize Doppler signals particularly at earlier gestational ages. When needed, color Doppler interrogation was applied before spectral Doppler sampling for improved visualization of the fetal vessels. An angle of <20 degrees between the vessel and pulsed Doppler sample was accepted for analysis.
To determine whether ALDAF is present in either great artery in the neonatal period and to confirm its timing in the cardiac cycle, we prospectively evaluated simultaneous transthoracic echocardiograms and electrocardiograms that had been performed on 21 healthy singleton newborn infants at a mean age of 2.9 days (range, 1–7 days). All newborn infants had an uncomplicated prenatal and perinatal history and were born at term. These neonates had not had previous fetal echocardiograms performed.
From digitally recorded echocardiographic images, the following images were evaluated off-line: mitral valve inflow and aortic outflow Doppler patterns, Doppler patterns in the ascending aorta, and Doppler patterns in the main pulmonary artery. For all studies in which ALDAF was present in one or both great arteries, the time interval from the onset of ALDAF to the onset of the respective great artery systolic Doppler signal (ALDAF-V interval) was measured ( Figure 2 , A). In the fetal studies, the mechanical atrioventricular time interval was measured with the use of the onset of atrial contraction of the mitral valve inflow to the left ventricular aortic outflow (MV-Ao) as an indirect marker of electrical atrial and ventricular activation ( Figure 2 , B). A comparison of each ALDAF-V time duration with MV-Ao time duration was performed. In the newborn infant studies, comparisons of ALDAF-V time interval with the MV-Ao time interval and the electrical PR time interval from a simultaneous electrocardiogram recording were performed to better define the timing of ALDAF and its relationship to mechanical and electrical events. In addition, the electrocardiogram was used to quantify the time delay from the start of the electrical PR interval to the start of the MV-Ao interval and the start of the ALDAF-V interval. All time intervals were measured on 3 cardiac cycles and averaged. Two investigators (L.W.H. and Y.Y.) performed all offline measurements and were blinded to subject identity and gestational age. Intraobserver and interobserver variability was evaluated on ALDAF-V time interval measurements with 15 randomly selected patient examinations. The examinations were reviewed independently by each observer, who was blinded to previous measurements. Intraobserver measurements were performed at least 2 weeks apart.
To determine the contribution of ALDAF to total ventricular output in the fetus, the velocity time integral (VTI) of ALDAF and the ventricular systolic Doppler signal in the great artery were measured. The relative contribution of ALDAF VTI to total forward VTI (sum of ALDAF and ventricular systolic blood flow) in all fetuses was calculated and expressed as a percentage.
The correlation between ALDAF-V and MV-Ao time durations was performed with a Pearson correlation coefficient with the use of a single measurement from each subject. Equivalence of ALDAF-V and MV-Ao time intervals was performed by a test of the mean of the paired differences in a simple regression model. When present, the changes in ALDAF-V and MV-Ao time durations with advancing gestational age were investigated with the use of piecewise linear regressions with knots placed at 11 and 20 weeks’ gestation. Parameters were estimated separately for the aorta and pulmonary artery measurements. All models accounted for repeated measures and the correlation because of multiple fetuses per pregnancy with the use of generalized estimating equations. Intraobserver and interobserver agreement and differences were assessed by the Bland-Altman method. A probability value of < .05 was considered statistically significant. All analyses were performed with SAS software (version 9.2; SAS Institute Inc., Cary, NC).
Results
Of the 237 pregnant women who were enrolled, 106 women (45%) had 1 fetal echocardiogram; 79 women (33%) had 2 fetal echocardiograms, and 52 women (22%) had ≥3 fetal echocardiograms. ALDAF was present in one or both ventricular outflows in 365 of 457 fetal studies (80%). It was identified in one or both great arteries in all fetuses <11 weeks’ gestation, and its presence continued with slight decline, in the aorta more so than pulmonary artery, with increasing gestational age ( Table ). Before 11 weeks’ gestation, clear identification of the interrogated great artery was difficult in most fetuses because of technical limitations; therefore, the outflow data were grouped. In the neonatal cohort, ALDAF was present in the pulmonary artery of all fetuses and was not observed in the aorta of any of the infants ( Figure 1 ).
| Weeks’ gestation | n | Aorta, n (%) | Pulmonary artery, n (%) |
|---|---|---|---|
| <11 | 72 | 72 (100) | – |
| 11-13 | 160 | 122 (76.3) | 117 (73.1) |
| 14-15 | 36 | 22 (61.1) | 29 (80.6) |
| 16-20 | 74 | 30 (40.5) | 36 (48.6) |
| >20 | 115 | 51 (44.4) | 70 (60.9) |
Timing of ALDAF before and after birth
In all fetuses in which ALDAF was present, the measured time durations of ALDAF-V and MV-Ao were significantly correlated (<11 weeks’ gestation: r = 0.94; P < .01; >11 weeks’ gestation: pulmonary artery, r = 0.70; P < .01; aorta, r = 0.78; P < .01; Figure 3 ). However, on average the MV-Ao time interval was consistently 4.7 ms (95% confidence interval [CI], 3.2–6.2) longer than the ALDAF-V time interval in fetuses less than 11 weeks’ gestation. Similarly, in fetuses at >11 weeks’ gestation, the MV-Ao time interval was consistently 11.73 msec (95% CI, 10.48–12.98) longer when compared with ALDAF-V time duration in the pulmonary artery and 9.27 msec (95% CI, 7.65–10.88) longer when compared with ALDAF-V time duration in the aorta ( Figure 4 ).
