Background
In severe right heart obstruction (RHO), redistribution of cardiac output to the left ventricle (LV) is well tolerated by the fetal circulation. Although the same should be true of severely regurgitant tricuspid valve disease (rTVD) with reduced or no output from the right ventricle, affected fetuses more frequently develop hydrops or suffer intrauterine demise. We hypothesized that right atrium (RA) function is altered in rTVD but not in RHO, which could contribute to differences in outcomes.
Methods
Multi-institutional retrospective review of fetal echocardiograms performed over a 10-year period on fetuses with rTVD (Ebstein’s anomaly, tricuspid valve dysplasia) or RHO (pulmonary atresia/intact ventricular septum, tricuspid atresia) and a healthy fetal control group. Offline velocity vector imaging and Doppler measurements of RA size and function and LV function were made.
Results
Thirty-four fetuses with rTVD, 40 with RHO, and 79 controls were compared. The rTVD fetuses had the largest RA size and lowest RA expansion index, fractional area of change, and RA indexed filling and emptying rates compared with fetuses with RHO and controls. The rTVD fetuses had the shortest LV ejection time and increased Tei index with a normal LV ejection fraction. RA dilation (odds ratio, 1.27; 95% CI, 1.05–1.54) and reduced indexed emptying rate (odds ratio, 2.49; 95% CI, 1.07–5.81) were associated with fetal or neonatal demise.
Conclusions
Fetal rTVD is characterized by more severe RA dilation and dysfunction compared with fetal RHO and control groups. RA dysfunction may be an important contributor to reduced ventricular filling and output, potentially playing a critical role in the worsened outcomes observed in fetal rTVD.
Highlights
- •
This is the first study to examine intrinsic right atrial function using velocity vector imaging in fetuses with and without right heart disease.
- •
Fetuses with regurgitant tricuspid valve disease, including Ebstein’s anomaly and tricuspid valve dysplasia, have evidence of impaired right atrial function characterized by increased global size, decreased emptying fraction and reservoir function, and reduced indexed emptying and filling rates.
- •
Impaired right atrial function in fetuses with regurgitant tricuspid valve disease may compromise fetal cardiac output, potentially contributing to the higher rate of poor outcomes in this population.
Fetal tricuspid valve disease with significant tricuspid valve regurgitation (rTVD), including both Ebstein’s anomaly and tricuspid valve dysplasia, is commonly associated with cardiovascular compromise that can lead to evolution of hydrops, sudden fetal demise, or hemodynamic instability in the neonatal period even in the absence of frank hydrops. Despite improvements in prenatal detection of rTVD, management of affected fetuses before and after birth remains an ongoing challenge, with an overall perinatal mortality reported as high as 83%. Even in the most recent surgical era, fetal mortality from rTVD remains high. Currently, the pathophysiology of fetal rTVD that contributes to the dismal outcome of these fetuses remains incompletely defined.
It has long been recognized that most structural heart defects are well tolerated by the fetal circulation. In cases of severe right heart obstruction (RHO) including tricuspid atresia and pulmonary atresia (PA)/intact ventricular septum, redistribution of systemic venous return across the patent foramen ovale to the left heart permits maintenance of the combined fetal cardiac output. Interestingly, although venous Doppler changes that suggest increased central venous pressure are common in fetal RHO, evolution of cardiovascular compromise is rare. While fetal rTVD represents right heart pathology and in the most severe cases requires redistribution of the entire cardiac preload to the left heart, these fetuses may be less able to maintain a normal cardiac output, resulting in hemodynamic instability before and after birth. This suggests a unique pathophysiology for fetal and neonatal rTVD. Past studies have identified features potentially attributable to the cardiovascular compromise of fetal rTVD including impaired left ventricular (LV) function and reduced LV preload and output. It has been postulated that altered LV filling and function in fetal and neonatal rTVD is a consequence of the mechanical influence of a dilated right atrium (RA) and ventricle (RV) and altered ventricular septal position.
The importance of atrial function on cardiac performance has been previously examined in both adult populations and animal models. The atria modulate ventricular filling through three phases: (1) a reservoir or expansion phase during systole, (2) a conduit phase during early diastole, and (3) an active contractile “booster” phase during late diastole. Abnormalities in atrial function are known to directly impact ventricular filling and, consequently, cardiac output. The role of the atria in ventricular performance and output may be even more important before birth. During fetal life, the RA, in particular, plays an integral role in the circulation, receiving blood from the systemic and umbilical veins and redistributing the venous return to both the RV and LV. However, the RA in rTVD is often massively dilated, while it is not in RHO and in normal fetuses. In the current investigation, we sought to determine whether there exists an inherent difference in RA function between fetuses with rTVD, those with critical RHO, and normal healthy fetuses. We hypothesized that in fetal rTVD, but not in RHO, intrinsic RA function is abnormal, which may result in ineffective blood flow redistribution, decreased LV filling and output, and increased RA and central venous pressures, ultimately contributing to worse outcomes.
Methods
Study Patients
Following receipt of ethics approval, we retrospectively identified from existing hospital databases all cases of rTVD diagnosed in utero between January 2001 and February 2010 at three major tertiary care institutions in North America: Stollery Children’s Hospital, Edmonton, Alberta, Canada; the University of California San Francisco Benioff Children’s Hospital, San Francisco, California; and the Children’s Hospital, Boston, Massachusetts. All cases of fetal cardiac anomalies were initially referred following abnormal obstetrical ultrasounds with concern for structural cardiac disease. We included fetuses with tricuspid valve dysplasia and Ebstein’s anomaly that were identified on prenatal ultrasound to have significant tricuspid valve regurgitation with apical displacement of the tricuspid valve septal leaflet or dysplastic leaflets. We also searched the existing fetal cardiac pathology databases of Stollery Children’s Hospital and University of California San Francisco Benioff Children’s Hospital to identify all encountered pregnancies with fetal RHO within the study period including fetuses with PA and intact ventricular septum, severe pulmonary stenosis with exclusively retrograde flow in the ductus arteriosus, and tricuspid atresia with normally related great vessels. Fetuses with rTVD or RHO associated with more complex heart disease or extracardiac anomalies were excluded. A cohort of normal fetuses from uncomplicated singleton pregnancies was identified for comparison purposes from the Stollery Children’s Hospital fetal database. Referral indications for the normal controls included suspicion of arrhythmia and family history of congenital heart disease. Prenatal and postnatal medical records were reviewed for clinical information, including perinatal outcome for both pathologies and control patients.
Image Acquisition
Detailed prenatal echocardiographic examinations, including anatomic assessment and Doppler interrogation, were performed with the following ultrasound systems: Siemens C512 (Mountain View, CA) and General Electric Voluson E8 imaging systems (General Electric, Milwaukee, WI). All data images were digitally recorded and stored as standard Digital Imaging and Communication in Medicine images for offline analysis. In cases where multiple evaluations of the same fetus were available, only the initial patient ultrasound study or the earliest serial study in which images for offline analysis were sufficient was analyzed. Fetuses were excluded if image format or quality was inadequate for offline velocity vector imaging (VVI) analysis.
Right Atrial Velocity Vector Imaging
RA functional parameters were measured using offline commercially available software (Syngo Velocity Vector Imaging, Siemens Medical Solutions, Malvern, PA). All digitally stored fetal echocardiogram images were reexamined, and VVI measures were performed offline by a single investigator (L.W.H.). The fetal cardiac cycle was determined using the anatomic M-mode by detecting fetal atrioventricular valve motion. Using a standard four-chamber view, a single-frame tracing of the RA endocardial border was performed to obtain a velocity vector profile of native atrial wall motion and area, with extrapolated volume data ( Figure 1 ). In fetuses with Ebstein’s anomaly and “atrialization” of a portion of the RV, the VVI tracing included only the native RA to the level of the tricuspid valve annulus and did not include the atrialized ventricular component as the ventricular interaction (atrialized RV and LV) often results in dyssynchronous motion. When septum primum was challenging to visualize in its entirety, the estimated plane of the atrial septum was used in the measurement. The software generated time-volume curves, providing data regarding the RA maximum and minimum calculated volumes as well as RA filling and emptying rates. To adjust for the differences in atrial volumes between gestations, both RA filling and RA emptying rates were indexed by dividing each value by the maximum RA volume, resulting in an indexed value measured in second −1 . As previously described, we calculated the RA emptying fraction, a measure of global atrial function, to be (RA max volume − RA min volume)/RA max volume and the atrial volume expansion index, a measure of atrial reservoir function, to be (RA max volume − RA min volume)/RA min volume. All parameters were measured and averaged over the three cardiac cycles. In 15 randomly selected patients, RA functional assessment by VVI was performed by a second independent investigator (N.S.K.) in a blinded fashion for evaluation of interobserver variability.
Anatomic and Doppler Parameters
Each selected echocardiographic study was retrospectively reviewed, and the following anatomic and functional findings were measured and collected: the grade of severity of tricuspid valve regurgitation (TR) as determined by color Doppler; presence or absence of patency of the pulmonary valve as determined by color Doppler blood flow characteristics in the main pulmonary artery; and LV ejection fraction as determined by single-plane Simpson’s method of discs as described elsewhere. TR was categorized as none, mild with a narrow vena contracta (≤1/3 the tricuspid valve annulus diameter), or moderate-severe with a wide color jet/vena contracta (>1/3 the tricuspid valve annulus diameter). Doppler measurements included mitral E and A wave velocities and calculated E/A ratio; systemic vein A wave reversal index as the ratio of inferior vena cava (IVC) retrograde A wave velocity time integral (VTI) over the forward flow VTI; LV Tei index; TR velocity; umbilical artery pulsatility index (UA PI); the presence of flow reversal in the ductus venosus during atrial systole; and the presence of umbilical venous pulsations. LV Tei index was calculated by dividing the combined LV isovolumic contraction (IVCT) and relaxation time (IVRT; time interval between the mitral valve closure and opening minus aortic ejection time [ET]) by the aortic ET. UA PI was calculated as (peak systolic velocity − end-diastolic velocity)/time-averaged mean velocity.
Statistical Analysis
Data were evaluated for normality using a Shapiro-Wilk test. Continuous variables were expressed as a median (interquartile range). Categorical variables were expressed as frequency (percentage). Comparisons between fetuses with rTVD, fetuses with RHO, and healthy normal fetuses used Kruskal-Wallis and χ 2 testing when indicated. The rTVD with pulmonary valve atresia (rTVD/PA) group was compared with the RHO group and the rTVD with antegrade pulmonary blood flow (rTVD/APF) group using Wilcoxon-Mann-Whitney and χ 2 testing as appropriate. Univariable logistic regression analysis was performed to identify factors associated with death (fetuses who were terminated were not included in these analyses). Intraclass correlation coefficients and 95% CIs were calculated to estimate the variability of repeated measurements. Data analysis was performed using Statistical Analysis Software (SAS version 9.3, SAS Inc., Cary, NC). A P value < .05 was considered significant.
Results
Clinical Characteristics
Thirty-four fetuses with rTVD and 40 with RHO were encountered in our three institutions during the study period. Findings in the fetuses with rTVD and RHO were compared with 79 singleton control fetuses of comparable gestational age (median 32.0 weeks vs 26.0 weeks vs 28.5 weeks, respectively; P = .09). Of the 34 fetuses with rTVD, 22 (64.7%) had Ebstein’s anomaly and 12 (35.3%) had tricuspid valve dysplasia. Gestational age at diagnosis ranged from 19 to 40 weeks. Moderate to severe TR was present in 26 (76.5%), mild TR was present in four (11.8%), and either anatomic or functional PA (rTVD/PA) was present in 16 (47.1%). Fetuses with RHO included 15 (37.5%) with tricuspid atresia and 25 (62.5%) with PA and intact ventricular septum. Gestational age at diagnosis ranged from 17 to 39 weeks. In the RHO group, moderate to severe TR was present in five (12.5%) and mild TR was present in six (15%).
VVI Parameters
rTVD versus RHO versus Controls
RA VVI parameters are summarized in Table 1 . Fetuses with rTVD had increased total maximum and minimum RA volumes compared with fetuses with RHO and controls ( P < .0001). Both RA maximum and minimum volumes in rTVD were significantly greater in late gestation compared with RHO and controls ( Figures 2 A and 2B). RA emptying fraction and fractional area change were decreased in rTVD and RHO fetuses compared with controls ( P < .0001; Figure 3 A). In rTVD fetuses, but not in those with RHO, indexed RA filling and emptying rates were also reduced compared with controls ( P < .0001; Figure 3 B). Finally, volume expansion and area expansion indices were reduced in rTVD fetuses compared with in RHO fetuses and controls ( P < .0001). Due to the small sample size, detailed analysis of the effect of TR on VVI parameters could not be performed.
| rTVD ( n = 34) | RHO ( n = 40) | Control ( n = 79) | P value | |
|---|---|---|---|---|
| RA emptying fraction (%) | 40.3 (32.0–47.3) | 43.7 (39.3–54.5) | 52.3 (45.0–59.0) | <.0001 ∗ |
| Fractional area of change (%) | 32.3 (25.0–38.3) | 36.0 (32.2–44.3) | 44.7 (37.7–48.0) | <.0001 ∗ |
| RA indexed filling rate (sec −1 ) | 2.8 (1.9–3.4) | 3.0 (2.3–3.7) | 3.4 (2.8–3.9) | .0020 † |
| RA indexed emptying rate (sec −1 ) | −2.9 (−3.6 to −2.3) | −3.8 (−4.6 to −2.8) | −4.3 (−4.9 to −3.9) | <.0001 ∗ |
| RA maximum volume (mL) | 7.6 (3.0–13.5) | 1.6 (0.7–2.9) | 1.6 (0.6–2.8) | <.0001 ‡ |
| RA minimum volume (mL) | 5.0 (1.6–8.3) | 0.8 (0.3–1.6) | 0.7 (0.3–1.4) | <.0001 ‡ |
| RA maximum area (mm 2 ) | 4.5 (2.7–6.7) | 1.7 (0.9–2.3) | 1.6 (0.8–2.4) | <.0001 ‡ |
| RA minimum area (mm 2 ) | 3.0 (1.7–5.2) | 1.1 (0.5–1.5) | 0.9 (0.5–1.5) | <.0001 ‡ |
| Volume expansion index | 0.7 (0.5–0.9) | 0.8 (0.7–1.2) | 1.1 (0.8–1.5) | <.0001 † |
| Area expansion index | 0.5 (0.3–0.6) | 0.6 (0.5–0.8) | 0.8 (0.6–0.9) | <.0001 ∗ |
∗ All groups are significantly different from each other.
† The rTVD group is significantly different from the control group.
‡ The rTVD group is significantly different from the RHO and control groups.
rTVD/PA versus RHO
We performed subgroup comparisons for RA VVI function parameters between rTVD/PA and RHO ( Table 2 ). The rTVD/PA fetuses had increased RA volume (9.0 mL vs 1.6 mL; P < .0001), reduced RA indexed emptying rate (−2.6 sec −1 vs −3.8 sec −1 ; P = .0008), decreased emptying fraction (38.3% vs 43.7%; P = .0302), and decreased fractional area of change (30.7% vs 36.0%; P = .0161).
| rTVD with PA ( n = 16) | RHO ( n = 40) | P value | |
|---|---|---|---|
| RA emptying fraction (%) | 38.3 (32.0–44.0) | 43.7 (39.3–54.5) | .0302 |
| Fractional area of change (%) | 30.7 (24.0–37.3) | 36.0 (32.2–44.3) | .0161 |
| RA indexed filling rate (sec −1 ) | 2.6 (1.8–3.2) | 3.0 (2.3–3.7) | .10 |
| RA indexed emptying rate (sec −1 ) | −2.6 (−3.1 to −2.1) | −3.8 (−4.6 to −2.8) | .0008 |
| RA maximum volume (mL) | 9.0 (5.3–14.6) | 1.6 (0.7–2.9) | <.0001 |
| RA minimum volume (mL) | 6.1 (3.0–10.8) | 0.8 (0.3–1.6) | <.0001 |
| RA maximum area (mm 2 ) | 5.4 (3.5–7.3) | 1.7 (0.9–2.3) | <.0001 |
| RA minimum area (mm 2 ) | 3.8 (2.2–5.3) | 1.0 (0.5–1.5) | <.0001 |
| Volume expansion index | 0.6 (0.5–0.8) | 0.8 (0.7–1.2) | .0225 |
| Area expansion index | 0.4 (0.3–0.6) | 0.6 (0.5–0.8) | .0136 |
rTVD/PA versus rTVD/APF
Table 3 displays the RA VVI parameters for rTVD fetuses categorized by presence or absence of forward flow across the pulmonary valve. No significant differences in RA function parameters were demonstrated between the two subgroups.
| rTVD with PA ( n = 16) | rTVD with APF ( n = 18) | P value | |
|---|---|---|---|
| RA emptying fraction (%) | 38.3 (32.0–44.0) | 43.2 (29.8–51.8) | .29 |
| Fractional area of change (%) | 30.7 (24.0–37.3) | 36.0 (26.5–39.5) | .33 |
| RA indexed filling rate (sec −1 ) | 2.6 (1.8–3.2) | 3.1 (2.0–3.5) | .49 |
| RA indexed emptying rate (sec −1 ) | −2.6 (−3.1 to −2.1) | −3.3 (−4.3 to −2.5) | .08 |
| RA maximum volume (mL) | 9.0 (5.3–14.6) | 4.7 (2.4–10.3) | .06 |
| RA minimum volume (mL) | 6.1 (3.0–10.8) | 2.1 (1.2–6.3) | .0429 |
| RA maximum area (mm 2 ) | 5.4 (3.5–7.3) | 3.1 (2.3–5.8) | .06 |
| RA minimum area (mm 2 ) | 3.8 (2.2–5.3) | 2.0 (1.3–4.2) | .06 |
| Volume expansion index | 0.6 (0.5–0.8) | 0.8 (0.4–1.1) | .40 |
| Area expansion index | 0.4 (0.3–0.6) | 0.6 (0.4–0.7) | .29 |
Doppler and Function Parameters
Doppler parameters of all three groups are displayed in Table 4 . No differences were observed in heart rate or LV ejection fraction between rTVD, RHO, and control fetal groups. The LV Tei index was greatest in fetuses with rTVD, followed by fetuses with RHO, then controls during post hoc testing (0.61 vs 0.47 vs 0.41; P < .0001). While there was no difference in IVRT, the rTVD ET was significantly reduced ( P < .0004), and UA PI increased ( P < .0001) when compared with RHO and controls. Umbilical vein pulsations were also more frequently observed in rTVD compared with in RHO (40% vs 21.9%; P = .0013).
| rTVD ( n = 34) | RHO ( n = 40) | Control ( n = 79) | P value | |
|---|---|---|---|---|
| Heart rate (beats/min) | 136.4 (127.7–142.9) | 136.4 (130.4–142.9) | 142.9 (136.4–150.0) | .06 |
| IVC forward VTI (cm) | 8.4 (4.8–11.5) | 6.4 (4.4–8.8) | 5.7 (5.1–7.7) | .81 |
| IVC reverse VTI (cm) | 0.8 (0.6–1.6) | 1.5 (0.7–2.6) | 0.3 (0.2–0.6) | <.0001 ∗ |
| IVC A wave reversal ratio | 0.12 (0.08–0.27) | 0.20 (0.13–0.33) | 0.05 (0.02–0.10) | <.0001 † |
| Mitral valve E wave velocity (cm/sec) | 34.9 (32.0–45.8) | 35.0 (27.0–41.0) | 29.0 (24.0–36.0) | .0211 ‡ |
| Mitral valve A wave velocity (cm/sec) | 54.0 (41.0–58.0) | 62.5 (50.0–70.0) | 41.5 (36.0–50.0) | <.0001 ∗ |
| Mitral valve E to A wave ratio | 0.73 (0.56–0.93) | 0.54 (0.50–0.70) | 0.70 (0.64–0.77) | .0004 § |
| ET (sec) | 0.16 (0.15–0.17) | 0.18 (0.16–0.18) | 0.17 (0.16–0.19) | .0004 ‖ |
| IVCT (sec) | 0.04 (0.03–0.05) | 0.03 (0.02–0.04) | 0.02 (0–0.03) | <.0001 † |
| IVRT (sec) | 0.06 (0.05–0.06) | 0.05 (0.04–0.05) | 0.05 (0.04–0.06) | .09 |
| Umbilical venous pulsations present | 8/20 (40) | 7/32 (21.9) | 0 | .0013 |
| UA PI | 1.4 (1.3–1.5) | 1.2 (1.0–1.4) | 1.1 (0.9–1.3) | <.0001 ‖ |
| Ductus venosus A wave reversal present | 7/14 (50) | 16/25 (64) | 0 | <.0001 |
| TR velocity (cm/sec) | 258.5 (193.5–333.0) | 415.0 (363.0–475.0) | .0088 | |
| LV ejection fraction (%) | 57.1 (52.2–64.2) | 59.8 (54.8–64.2) | 58.0 (51.2–63.0) | .37 |
| LV Tei index | 0.61 (0.47–0.72) | 0.47 (0.35–0.56) | 0.41 (0.30–0.50) | <.0001 ∗ |
∗ All groups are significantly different from each other.
† Control group is significantly different from the rTVD and RHO groups.
‡ The rTVD group is significantly different from the control group.
§ The RHO group is significantly different from the rTVD and control groups.
‖ The rTVD group is significantly different from the RHO and control groups.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
