Fetal aortic balloon valvuloplasty (FAV) has shown promise in averting progression of midgestation aortic stenosis (AS) to hypoplastic left heart syndrome in a subset of patients. Patients who achieve biventricular circulation after FAV frequently have left ventricular (LV) diastolic dysfunction (DD). This study evaluates DD in fetuses with AS by comparing echocardiographic indices of LV diastolic function in fetuses underwent FAV (n = 20) with controls (n = 40) and evaluates for LV factors associated with DD in patients with FAV. We also compared pre-FAV and post-FAV DD variables (n = 16). Median gestational age (24 weeks, range 18 to 29 weeks) and fetal heart rate were similar between FAV and controls. Compared with controls, patients with FAV had universally abnormal LV diastolic parameters including fused mitral inflow E and A waves (p = 0.008), higher E velocity (p <0.001), shorter mitral inflow time (p = 0.001), lower LV lateral annulus E′ (p <0.001), septal E′ (p = 0.003), and higher E/E′ (p <0.001) than controls. Patients with FAV had abnormal right ventricular mechanics with higher tricuspid inflow E velocity (p <0.001) and shorter tricuspid inflow time (p = 0.03). Worse LV diastolic function (lower LV E′) was associated with higher endocardial fibroelastosis grade (r = 0.74, p <0.001), large LV volume (r = 0.55, p = 0.013), and sphericity (r = 0.58, p = 0.009) and with lower LV pressure by mitral regurgitation jet (r = −0.68, p <0.001). Post-FAV, fewer patients had fused mitral inflow E and A than pre-FAV (p = 0.05) and septal E′ was higher (=0.04). In conclusion, fetuses with midgestation AS have evidence of marked DD. Worse DD is associated with larger, more spherical LV, with more extensive endocardial fibroelastosis and lower LV pressure.
Background
Fetal aortic balloon valvuloplasty (FAV) has shown promise in altering in utero progression of aortic stenosis (AS) to hypoplastic left heart syndrome. The postnatal course, including size of the left heart structures, left ventricular (LV) function, and surgical management, is variable in patients who have undergone FAV. Recent studies have reported promising results, with upward of 35% to 40% of patients achieving a biventricular circulation. In children with other left heart obstructive lesions, including congenital AS and aortic coarctation, LV diastolic dysfunction (DD) is well described. Our previous work has shown DD in infants and children with biventricular circulation after FAV is common, more severe than isolated congenital AS and persists over medium-term follow-up. DD in this patient population is a challenge as there are few management options, and DD may limit some patient’s ability to tolerate biventricular circulation. Timing of onset of DD, risk factors for DD, and associations with other LV characteristics, including LV size, geometry, pressure load, and endocardial fibroelastosis (EFE), have not been well described. DD in patients with fetal AS has not been evaluated prenatally but presumably occurs early in gestation and may be related to several factors including pressure load on the LV leading to EFE and myocardial fibrosis, myocyte hyperplasia, decreased in utero flow through the LV resulting in altered myocardial compliance, and abnormal coronary blood flow. This study aims to evaluate DD in patients with fetal AS by comparing echocardiographic indices of LV diastolic function in fetuses with AS underwent FAV with controls and to evaluate for LV factors associated with DD in patients with FAV. Additionally, we compare pre-FAV and post-FAV diastolic function variables to evaluate for acute change.
Methods
We retrospectively reviewed the records of the last 20 consecutive patients who underwent technically successful FAV from January 2010 to October 2013. Patient selection, technical performance, and outcomes for FAV have been previously described. Echocardiograms from gestational age-matched control fetuses (n = 40) with no structural or functional heart disease were used as controls. Reasons for fetal echocardiogram in the control group included family history of heart disease (n = 28), increased nuchal thickness (n = 3), suboptimal cardiac views on obstetrical screening (n = 4), or other (n = 5). Fetal echocardiograms were excluded in cases in which maternal or fetal systemic disease that could impact LV systolic or diastolic function was present (including maternal diabetes, lupus erythematous, and chromosomal anomalies). The Committee for Clinical Investigation at the Children’s Hospital Boston approved the use of patient medical records for this retrospective review.
All patients with FAV had a pre-FAV echocardiogram on the day before FAV and a post-FAV echo (1 to 3 days post-FAV) included in the analysis. In the single case where 2 FAVs were performed, only the second FAV was included in the analysis for both pre-FAV and post-FAV analyses. All fetal echocardiograms were performed using commercially available ultrasound equipment (Philips iE33; Koninklijke Philips Electronics, N.V., Phillips Electronics Andover, Massachusetts) or Accuson Siemens Sequoia 512 (Siemens Medical Solutions USA, Inc., Malvern, Pennsylvania).
Left-sided heart variables collected from reports produced at the time of the pre-FAV study included end-diastolic volume (EDV), end-diastolic dimension, ejection fraction, aortic valve dimension, mitral valve dimension, and the z-scores for each of these variables. In our noninvasive imaging laboratory, LVEDV is calculated using the 5th/6th area-length formula using 2-dimensional volumetric measurements. LV sphericity was calculated as the ratio of short-axis area to long-axis length. Mitral regurgitation (MR) jet gradient and aortic valve Doppler maximum instantaneous gradient were recorded. MR jet gradient >20 mm Hg is a major criteria in our current selection process for FAV. EFE was qualitatively graded for all patients with FAV by a single observer (K.F.) using a previously described 4-point scoring system in half unit increments: 0, none; 1, mild (scattered echogenic sots with the LV); 2, moderate (noncontiguous echogenic patches throughout the LV); and 3, severe (contiguous echogenic lining of the LV). Previous studies have shown moderate intraobserver and interoberserver reproducibility using this scoring system.
Diastolic function assessment included pulsed-Doppler of the atrioventricular valve inflow and tissue Doppler imaging. All measurements of diastolic variables were retrospectively re-measured from images obtained at the time of the study. Conventional pulsed-Doppler indices of diastolic function, including peak early (E) and late (A) diastolic transmitral velocities, E:A ratio, and atrioventricular valve inflow duration were measured. To control for heart rate, mitral inflow was indexed as percentage of the cardiac cycle and as ratio of mitral to tricuspid inflow duration. Pulsed-wave tissue Doppler velocities were obtained from the lateral mitral annulus, the interventricular septum, and lateral tricuspid annulus from the apical 4-chamber view. Tissue Doppler tracings with an angle of interrogation ≤15° from parallel were considered technically adequate. Tissue Doppler measurements for each myocardial segment included peak systolic velocity (S′), peak early diastolic velocity (E′), and peak late diastolic velocity (A′). Only tracings that demonstrated a clear E′ were used. Each velocity was measured on 3 consecutive cardiac cycles, and the average of these values was used for the analysis. E′:A′ and E/E′ were calculated for all wall segments.
Median and range are used to express measures of central tendency and dispersion of data. Comparisons between FAV (pre-FAV echo) and control patients were evaluated using Mann-Whitney test. Comparisons between pre-FAV and post-FAV echo variables were performed using Wilcoxon signed-rank test. Associations between LV characteristics and LV diastolic function parameters were evaluated. Associations between normally distributed continuous variables were assessed using Pearson correlation coefficients. Associations between ordinal variables (EFE grade) and continuous variables were assessed using Spearmen correlation coefficients. Interobserver agreement for 2 independent observers measurement of LV E′ in 10 control fetuses, using Cronbach’s α test showed good agreement (=0.862). All statistical analysis were 2 sided, and type I error was controlled at a level of 0.05. Analyses were performed with SPSS (version 16.0; SPSS Inc., Chicago, Illinois).
Methods
We retrospectively reviewed the records of the last 20 consecutive patients who underwent technically successful FAV from January 2010 to October 2013. Patient selection, technical performance, and outcomes for FAV have been previously described. Echocardiograms from gestational age-matched control fetuses (n = 40) with no structural or functional heart disease were used as controls. Reasons for fetal echocardiogram in the control group included family history of heart disease (n = 28), increased nuchal thickness (n = 3), suboptimal cardiac views on obstetrical screening (n = 4), or other (n = 5). Fetal echocardiograms were excluded in cases in which maternal or fetal systemic disease that could impact LV systolic or diastolic function was present (including maternal diabetes, lupus erythematous, and chromosomal anomalies). The Committee for Clinical Investigation at the Children’s Hospital Boston approved the use of patient medical records for this retrospective review.
All patients with FAV had a pre-FAV echocardiogram on the day before FAV and a post-FAV echo (1 to 3 days post-FAV) included in the analysis. In the single case where 2 FAVs were performed, only the second FAV was included in the analysis for both pre-FAV and post-FAV analyses. All fetal echocardiograms were performed using commercially available ultrasound equipment (Philips iE33; Koninklijke Philips Electronics, N.V., Phillips Electronics Andover, Massachusetts) or Accuson Siemens Sequoia 512 (Siemens Medical Solutions USA, Inc., Malvern, Pennsylvania).
Left-sided heart variables collected from reports produced at the time of the pre-FAV study included end-diastolic volume (EDV), end-diastolic dimension, ejection fraction, aortic valve dimension, mitral valve dimension, and the z-scores for each of these variables. In our noninvasive imaging laboratory, LVEDV is calculated using the 5th/6th area-length formula using 2-dimensional volumetric measurements. LV sphericity was calculated as the ratio of short-axis area to long-axis length. Mitral regurgitation (MR) jet gradient and aortic valve Doppler maximum instantaneous gradient were recorded. MR jet gradient >20 mm Hg is a major criteria in our current selection process for FAV. EFE was qualitatively graded for all patients with FAV by a single observer (K.F.) using a previously described 4-point scoring system in half unit increments: 0, none; 1, mild (scattered echogenic sots with the LV); 2, moderate (noncontiguous echogenic patches throughout the LV); and 3, severe (contiguous echogenic lining of the LV). Previous studies have shown moderate intraobserver and interoberserver reproducibility using this scoring system.
Diastolic function assessment included pulsed-Doppler of the atrioventricular valve inflow and tissue Doppler imaging. All measurements of diastolic variables were retrospectively re-measured from images obtained at the time of the study. Conventional pulsed-Doppler indices of diastolic function, including peak early (E) and late (A) diastolic transmitral velocities, E:A ratio, and atrioventricular valve inflow duration were measured. To control for heart rate, mitral inflow was indexed as percentage of the cardiac cycle and as ratio of mitral to tricuspid inflow duration. Pulsed-wave tissue Doppler velocities were obtained from the lateral mitral annulus, the interventricular septum, and lateral tricuspid annulus from the apical 4-chamber view. Tissue Doppler tracings with an angle of interrogation ≤15° from parallel were considered technically adequate. Tissue Doppler measurements for each myocardial segment included peak systolic velocity (S′), peak early diastolic velocity (E′), and peak late diastolic velocity (A′). Only tracings that demonstrated a clear E′ were used. Each velocity was measured on 3 consecutive cardiac cycles, and the average of these values was used for the analysis. E′:A′ and E/E′ were calculated for all wall segments.
Median and range are used to express measures of central tendency and dispersion of data. Comparisons between FAV (pre-FAV echo) and control patients were evaluated using Mann-Whitney test. Comparisons between pre-FAV and post-FAV echo variables were performed using Wilcoxon signed-rank test. Associations between LV characteristics and LV diastolic function parameters were evaluated. Associations between normally distributed continuous variables were assessed using Pearson correlation coefficients. Associations between ordinal variables (EFE grade) and continuous variables were assessed using Spearmen correlation coefficients. Interobserver agreement for 2 independent observers measurement of LV E′ in 10 control fetuses, using Cronbach’s α test showed good agreement (=0.862). All statistical analysis were 2 sided, and type I error was controlled at a level of 0.05. Analyses were performed with SPSS (version 16.0; SPSS Inc., Chicago, Illinois).
Results
Median gestational age at FAV was 22 weeks (range 19 to 29) and was similar to controls (23 weeks [18 to 28], p = 0.87). One patient underwent FAV twice at 19 and 26 weeks. Pre-FAV left heart structural variables and z-scores are listed in Table 1 . Systolic function was severely depressed in 19 (95%) patients with FAV and moderately depressed in 1. All FAV patients had grade 1 or higher EFE with 15 (75%) having moderate or greater EFE. All patients had an estimate of LV pressure by MR jet (n = 16) and/or aortic valve gradient (n = 13). LV pressure-load was present in all patients with median MR jet gradient of 33 mm Hg and aortic valve gradient of 22 mm Hg with a wide range in degree of LV pressure elevation (MR jet range 20 to 70 mm Hg and aortic valve gradient range 10 to 58 mm Hg).
| FAV, n = 20 | |
|---|---|
| Aortic valve (mm) | 2.8 (2.4–4.0) |
| Aortic valve z-score | −2.8 (−3.9 to −1.2) |
| Mitral valve (mm) | 6.3 (3.6–8.6) |
| Mitral valve z-score | −0.7 (−2.8 to 3.5) |
| LV end-diastolic volume (ml) | 2.5 (0.4–8.2) |
| LV end-diastolic volume z-score | 1.9 (−1.9 to 5.1) |
| LV end-diastolic dimension (mm) | 1.2 (0.5–2.0) |
| LV end-diastolic dimension z-score | 1.2 (−2.6 to 6.2) |
| LV long-axis dimension (mm) | 1.8 (1.3–3.1) |
| LV long-axis dimension z-score | 0.5 (−1.7 to 4.8) |
| LV sphericity | 0.69 (0.31–0.81) |
| Ejection fraction (%) | 24 (5–41) |
| Mitral regurgitation jet gradient (mm Hg) | 33 (20–70) |
| Aortic valve gradient (mm Hg) | 22 (10–58) |
| Aortic regurgitation grade (0–4) | 0 (0–1) |
| Endocardial fibroelastosis grade (0–3) | 2 (1–3) |
| Endocardial fibroelastosis ≥moderate (n, %) | 15 (75%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
