Single right ventricles (SRV) are postulated to be disadvantaged compared with single left ventricles (SLV). We compared the evolution of SRV versus SLV function during infancy using conventional measures and speckle-tracking echocardiography (STE). We hypothesized that the SRV is mechanically disadvantaged during early infancy.
SRVs ( n = 32) were compared with SLVs ( n = 16) at the neonatal (presurgery) and pre-bidirectional cavopulmonary anastomosis (pre-BCPA) stages. Functional measures (fractional area change, indexed ventricular annular plane systolic excursion [iVAPSE], isovolumic acceleration [IVA], myocardial performance index, E and A velocities, tissue Doppler imaging annular velocities and STE-measured global longitudinal and circumferential strain, strain rate [SR], and early diastolic SR [EDSR]) were compared between SRV and SLV at each stage and between presurgery and pre-BCPA.
Compared with SLV, presurgery SRV had lower circumferential strain (−10.6% vs −16.5%; P = .0002) and EDSR (1.41%/sec vs 2.13%/sec; P = .001). Pre-BCPA SRV had decreased IVA (1.2 vs 2.1 m/sec 2 ; P = .006): longitudinal strain (−15.3% vs −19.1%; P = .001), SR (−0.97%/sec vs −1.53%/sec; P = .0001), EDSR (1.5%/sec vs 2.1%/sec; P = .001); circumferential strain (−10.6% vs −14.9%; P = .002), SR (−0.8%/sec vs −1.21%/sec; P = .0001), and EDSR (1.3%/sec vs 1.8%/sec; P = .009). SRV showed reduction of iVAPSE, IVA, s′, e′, a′ velocities, longitudinal strain, SR, EDSR, and circumferential SR ( P < .05) from presurgery to pre-BCPA, while circumferential strain was unchanged. SLV showed no significant change in these parameters during this interval.
The progressive reduction in SRV longitudinal and circumferential function suggests that SRV may have a mechanical disadvantage from birth and progressive impairment with age.
Survival of patients with a single morphological right ventricle (SRV) has been found to be poorer than those with a dominant single left ventricle (SLV). Tweddell et al. found a greater event-free survival in SLV versus SRV at 10 years after the Fontan completion, and a dominant right ventricle (RV) has been identified as a risk factor for mortality, particularly before the bidirectional cavopulmonary anastomosis (BCPA). In our previous study of hypoplastic left heart syndrome (HLHS) during early infancy, reduced RV contractility was observed between the first two stages of palliation. Our more recent cross-sectional study comparing SRV and SLV at pre-BCPA showed reduced contractility and reduced systolic and diastolic function in SRV, with the difference in diastolic function being more prominent after Fontan completion.
The objective of this study was to compare the evolution of functional changes in the morphologic SRV compared with the SLV from birth (presurgery) to just prior to the second stage of surgical palliation (pre-BCPA). We hypothesized that the systemic SRV would show evidence of relatively decreased ventricular function compared with the SLV from an early stage.
This was part of a larger study to examine ventricular function in patients with single ventricular physiology using echocardiography from 2008 to 2014 at the Stollery Children’s Hospital, Edmonton, Canada. There were 65 SRV and 17 SLV patients assessed during this period. The inclusion criteria were patients with functionally single ventricle who survived the first two stages of surgical palliation and had both presurgery and pre-BCPA functional echocardiograms. Exclusion criteria included patients who required transplantation or died before the BCPA, uncooperative patients with poor echocardiographic images, and patients with severe atrioventricular valve regurgitation since the newborn period. Sixteen SLV and 53 SRV patients fulfilled the inclusion criteria. Of the patients not included, one patient with SLV had pulmonary atresia with intact ventricular septum and coronary ostial atresia and died awaiting transplantation, two SRV did not undergo Norwood procedure at stage 1 palliation, and 10 SRV died prior to BCPA. The SRV deaths included one patient with Turner’s syndrome who underwent compassionate care and did not undergo Norwood; of the remaining nine deaths (all of whom underwent Norwood procedure), there were one with interstitial lung disease, one with a mitochondrial depletion disorder, one with severe tricuspid regurgitation, one with chromosomal abnormalities, one with prenatal RV dysfunction who developed low cardiac output post-Norwood and persistent dysfunction requiring extracorporeal membrane oxygenation and who did not recover function, and four interstage deaths after discharge from the hospital. We then matched the 16 eligible SLV patients in a 1:2 ratio by an observer (S.S.) blinded to the patients’ clinical course with 32 SRVs of similar age and weight. All enrolled patients underwent detailed echocardiographic scans during the neonatal period prior to the first palliative surgery (presurgery) and then prior to the BCPA (pre-BCPA), using a functional protocol optimized for the analysis of ventricular function. The demographic data collected included age, weight, and height at the time of the echocardiogram. The clinical data collected at each stage included heart rate, blood pressure, oxygen saturations, blood pH, and lactate at time of echocardiography when available. Parents gave informed consent to participate, and the study was approved by the Health Research Ethics Board at the University of Alberta.
Two-dimensional echocardiography was performed using a Vivid 7 ultrasound platform (GE Medical Systems, Milwaukee, WI) using a 10- or 7-MHz probe with electrocardiographic recordings, with the images optimized for higher frame rates (mean 132 ± 35 Hz). Images were obtained according to our single ventricle protocol and included the parasternal short-axis view at the base for circumferential deformation and the equivalent apical four-chamber view for longitudinal deformation ( Figure 1 ).
Conventional Parameters of Ventricular Function
The dominant ventricle’s global systolic function was assessed using percentage of ventricular fractional area change (FAC) from the apical four-chamber view. Ventricular longitudinal function was assessed using ventricular annular plane systolic excursion indexed to body surface area (iVAPSE), color tissue Doppler imaging (TDI) s′ velocity of the dominant ventricular free wall, isovolumic acceleration (IVA), and the myocardial performance index (MPI) determined on color TDI. Ventricular diastolic function was assessed using the dominant ventricular inflow pulsed Doppler E and A velocities, color TDI e′ and a′ velocities of the dominant ventricular free wall, and the E/e′ ratio of the ventricular free wall.
Offline two-dimensional speckle-tracking echocardiographic (STE) images were analyzed using commercially available software (EchoPAC version 7.1; GE Medical Systems) as described elsewhere ( Figure 1 ). Measurement of longitudinal strain was made from the apical four-chamber image. The region of interest was drawn starting at the dominant ventricular free wall up to and including the ventricular septum and ending at the crest of the septum in both SRV and SLV. The nondominant ventricle was not included in longitudinal strain measurements. For circumferential strain, the region of interest was drawn around the basal free wall to the ventricular septum of the dominant ventricle if there was no ventricular septal defect (VSD) or only a small VSD. In four SLVs with a large VSD, the region of interest extended from the dominant ventricular wall to the hypoplastic ventricle’s free wall. The overall impact of including this small portion of the nondominant ventricle only accounted for one of the six segments analyzed and was unlikely to significantly affect the overall results. We measured circumferential and longitudinal peak strain (%), peak strain rate (SR, %/sec), myocardial dyssynchrony index (MDI; the standard deviation of time to peak strain of six segments normalized as a percentage of systole, %), early diastolic SR (EDSR, %/sec) as a surrogate for ventricular relaxation, and the longitudinal:circumferential strain ratio. We have previously reported a low interobserver variability, a small 1.96× within-subject SD, and high intraclass correlation coefficients for all the reported parameters, indicating good repeatability of these novel parameters in an anatomically similar cohort of SRV patients using the same equipment and analysis software.
All functional data are presented as medians with ranges. Wilcoxon matched-pairs signed rank test was used for comparisons between stages within each group. Comparisons between SV groups were performed using Mann-Whitney test at presurgery and pre-BCPA stage. P < .05 was considered statistically significant. Analyses were performed using GraphPad Prism version 6.0 (GraphPad Software, Inc., La Jolla, CA).
This study consisted of 48 infants with single ventricular physiology (32 SRV vs 16 SLV). The majority of patients with SRV had HLHS ( n = 28), while the rest had critical aortic stenosis ( n = 2) or right dominant atrioventricular septal defect ( n = 2). Tricuspid atresia ( n = 5), pulmonary atresia with intact ventricular septum ( n = 3), double inlet left ventricle (LV) ( n = 5), transposition of the great arteries with hypoplastic RV ( n = 1), congenitally corrected transposition with tricuspid atresia ( n = 2), and left dominant atrioventricular septal defect ( n = 1) were the morphologies in SLV patients.
Comparison between SRV and SLV at Each Stage of Surgical Palliation
Median age at echocardiogram and body weight in both SRVs and SLVs were similar presurgery. Clinical variables including heart rate, blood pressure, and oxygen saturation were also not different between SRVs and SLVs. The blood pH and serum lactate levels were mildly elevated in SRV patients prior to surgery; however, all patients were stable with normal blood pressure and perfusion ( Table 1 ). There were two SRV patients with moderate tricuspid regurgitation at this stage; the rest had none to mild regurgitation, and no patient had severe regurgitation.
|Stage||SRV median (range)||SLV median (range)||P value|
|Age at echocardiogram||Presurgery (days)||4 (0–31)||7 (0–44)||NS|
|Pre-BCPA (months)||5.2 (3.2–11.3)||5.1 (3.5–8)||NS|
|Weight (kg)||Presurgery||3.3 (2.1–4.1)||3.05 (2–4.3)||NS|
|Pre-BCPA||5.8 (3.7–8)||6.1 (4.6–7.9)||NS|
|Heart rate (beats/min)||Presurgery||160 (118–177)||150 (120–170)||NS|
|Pre-BCPA||115 (86–150)||120 (101–170)||NS|
|Systolic BP (mmHg)||Presurgery||67 (47–90)||71 (55–86)||NS|
|Pre-BCPA||87 (55–114)||88 (68–121)||NS|
|Diastolic BP (mmHg)||Presurgery||40 (31–57)||38 (28–50)||NS|
|Pre-BCPA||43 (20–65)||48 (29–71)||NS|
|Oxygen saturation (%)||Presurgery||90 (76–98)||83 (74–98)||NS|
|Pre-BCPA||78 (68–90)||81 (67–90)||NS|
|pH||Presurgery||7.39 (7.26–7.47)||7.34 (7.3–7.45)||.02|
|Serum lactate (mmol/L)||Presurgery||1.6 (0.8–2.9)||1.0 (0.5–2.2)||.02|
All patients underwent a first-stage surgical procedure. The modified Norwood procedure with Sano shunt was the first surgery in all SRVs and in five patients with SLVs who had hypoplastic aortas. One patient with SLV had bilateral branch pulmonary arterial banding with repair of total anomalous pulmonary venous return. The remaining SLVs underwent aortopulmonary shunt. During the interstage period, balloon dilatation of recoarctation occurred in nine patients with SRVs (28% of SRVs). Three patients with SRVs (9.4% of SRVs) were on afterload reduction (captopril or enalapril). In SRVs, five patients with lower grades of tricuspid regurgitation developed moderate regurgitation by this stage. The two patients with moderate tricuspid regurgitation presurgery both improved to a lower grade of regurgitation by pre-BCPA stage, and none had severe regurgitation. Only one patient in the SLV group had moderate mitral regurgitation at each stage.
Conventional measures of systolic and diastolic ventricular function and indexed end-diastolic area (EDA) and end-systolic area (ESA) did not significantly differ between SRVs and SLVs prior to the first surgery ( Table 2 ). However, circumferential strain ( P = .0002) and SR ( P = .02) were significantly reduced in SRVs compared with SLVs presurgery. Furthermore, circumferential EDSR was significantly lower in SRVs compared with in SLVs ( P = .001). The longitudinal:circumferential strain ratio was significantly higher in the SRVs compared with in the SLVs at this stage ( P = .002). There were no significant differences in longitudinal parameters or circumferential MDI between SRV and SLV at presurgery stage ( Table 2 ).
|SRV median (range)||SLV median (range)||P value|
|FAC (%)||38.8 (31.1–65.8)||41.6 (33.6–46.8)||NS|
|Indexed EDA (cm 2 )||46.3 (19.1–58.8)||48.9 (30.8–61)|
|Indexed ESA (cm 2 )||29.2 (7.4–39.8)||28.7 (17.5–36.2)|
|Indexed VAPSE (cm/m 2 )||4.5 (3.5–8.3)||4.7 (2.6–6.3)||NS|
|IVA (cm/sec 2 )||3.1 (0.8–7.1)||2.7 (0.7–5.9)||NS|
|s′ velocity (cm/sec)||6.7 (4–10.2)||5.4 (2.5–9.5)||NS|
|MPI||0.35 (0.15–0.69)||0.38 (0.19–0.92)||NS|
|Longitudinal strain (%)||−17.9 (−23.8 to −11.5)||−19.1 (−25.2 to −14.4)||NS|
|Longitudinal SR (%/sec)||−1.57 (−2.4 to −1.1)||−1.65 (−2.4 to −1.2)||NS|
|Longitudinal MDI (%)||17.6 (5.7–39.4)||15 (6.5–25.5)||NS|
|Circumferential strain (%)||−10.6 (−18.6 to −4.6)||−16.5 (−21.8 to −6.3)||.0002|
|Circumferential SR (%/sec)||−1.2 (−2.03 to −0.71)||−1.4 (−2.7 to −0.83)||.02|
|Circumferential MDI (%)||23 (10–40.3)||22 (14.1–44)||NS|
|Longitudinal to circumferential strain ratio||1.6 (0.97–5.1)||1.2 (0.88–2.4)||.002|
|E velocity (cm/sec)||98.5 (43–145)||70 (34–200)||NS|
|A velocity (cm/sec)||89 (21–139)||99.5 (49–30)||NS|
|E/A ratio||0.84 (0.45–5.2)||0.76 (0.36–3.1)||NS|
|e′ velocity (cm/sec)||8.5 (1.7–20)||7 (2.7–14.8)||NS|
|a′ velocity (cm/sec)||8 (1.7–12.1)||7.4 (3.1–14.6)||NS|
|E/e′ ratio||10.4 (5.3–30)||10.3 (3–37.3)||NS|
|Longitudinal EDSR (%/sec)||2.5 (0.4–4.5)||2.3 (1.2–4.2)||NS|
|Circumferential EDSR (%/sec)||1.41 (0.63–2.8)||2.13 (0.7–2.8)||.001|
There were no differences in demographic or clinical characteristics between SRVs and SLVs prior to BCPA ( Table 1 ). Comparisons at the pre-BCPA stage showed reduced IVA ( P = .006) and s′ velocity ( P = .01) in SRV compared with SLV, while no differences in FAC, iVAPSE, MPI, or conventional diastolic parameters were found. Indexed EDA and ESA did not differ between the SRV and SLV. STE-derived parameters, which were reduced in SRV compared with in SLV at the pre-BCPA stage, included longitudinal strain ( P = .001), SR ( P = .0001), and EDSR ( P = .003); and circumferential strain ( P = .002), SR ( P = .0001), and EDSR ( P = .009), while the longitudinal:circumferential strain ratio and both longitudinal and circumferential MDI did not differ ( Table 3 ).
|SRV median (range)||SLV median (range)||P value|
|FAC (%)||36.5 (26.2–56.3)||41.3 (31.1–49.6)||NS|
|Indexed EDA (cm 2 )||39.1 (20.6–67.6)||38.5 (22.5–47.8)|
|Indexed ESA (cm 2 )||25.8 (12.2–42.9)||23.4 (12.5–31.9)|
|Indexed VAPSE (cm/m 2 )||2.7 (1.8–4.4)||2.9 (2–4.2)||NS|
|IVA (cm/sec 2 )||1.2 (0.4–5.3)||2.1 (0.9–4)||.006|
|s′ velocity (cm/sec)||3.3 (1.3–6.1)||4.7 (1.1–5.8)||.01|
|MPI||0.35 (0.18–0.94)||0.42 (0.2–0.8)||NS|
|Longitudinal strain (%)||−15.3 (−21.2 to −5.9)||−19.1 (−24.9 to −11.2)||.001|
|Longitudinal SR (%/sec)||−0.97 (−2.1 to −0.3)||−1.53 (−2.3 to −0.6)||.0001|
|Longitudinal MDI (%)||16 (5.5–39.6)||13.9 (11.3–48.4)||NS|
|Circumferential strain (%)||−10.6 (−16.9 to −4.5)||−14.9 (−22.1 to −5.7)||.002|
|Circumferential SR (%/sec)||−0.8 (−1.38 to −0.39)||−1.21 (−2.3 to −0.76)||.0001|
|Circumferential MDI (%)||19.2 (3.3–48.8)||19.2 (10.3–46.4)||NS|
|Longitudinal to circumferential strain ratio||1.46 (0.92–2.9)||1.29 (0.86–3.8)||NS|
|E velocity (cm/sec)||78 (44–155)||71 (46–121)||NS|
|A velocity (cm/sec)||75 (30–157)||94.5 (38–121)||NS|
|E/A ratio||0.96 (0.6–3.3)||0.74 (0.4–3.3)||NS|
|e′ velocity (cm/sec)||5.4 (2.6–15.8)||7.2 (3.4–16.4)||NS|
|a′ velocity (cm/sec)||5.1 (0.4–13)||4.9 (0.9–13.1)||NS|
|E/e ratio||13.9 (4.6–37.6)||11.9 (5.8–18.7)||NS|
|Longitudinal EDSR (%/sec)||1.5 (0.7–4.6)||2.1 (1.3–4.3)||.001|
|Circumferential EDSR (%/sec)||1.31 (0.43–2.38)||1.8 (0.73–3.2)||.009|