Author + information
- Received August 5, 2014
- Accepted August 13, 2014
- Published online November 11, 2014.
- Peter C. Frommelt, MD∗∗ (, )
- Eric Gerstenberger, MS†,
- James F. Cnota, MD‡,
- Meryl S. Cohen, MD§,
- Jessica Gorentz, RDCS∗,
- Kevin D. Hill, MD, MS‖,
- J. Blaine John, MD¶,
- Jami C. Levine, MD#,
- Jimmy Lu, MD∗∗,
- William T. Mahle, MD††,
- Rachel T. McCandless, MD‡‡,
- Luc Mertens, MD, PhD§§,
- Gail D. Pearson, MD, ScD‖‖,
- Carolyn Spencer, MD¶¶,
- Deepika Thacker, MD##,
- Ismee A. Williams, MD, MS∗∗∗,
- Pierre C. Wong, MD†††,
- Jane W. Newburger, MD, MPH#,
- Pediatric Heart Network Investigators
- ∗Medical College of Wisconsin, Milwaukee, Wisconsin
- †New England Research Institute, Watertown, Massachusetts
- ‡Cincinnati Children’s Hospital, Cincinnati, Ohio
- §The Children's Hospital of Philadelphia, Philadelphia, Pennsylvania
- ‖Duke University, Durham, North Carolina
- ¶Congenital Heart Institute of Florida/Pediatrix, Tampa, Florida
- #Boston Children’s Hospital, Boston, Massachusetts
- ∗∗University of Michigan, Ann Arbor, Michigan
- ††Emory University School of Medicine, Atlanta, Georgia
- ‡‡Primary Children’s Hospital, Salt Lake City, Utah
- §§Hospital for Sick Children, Toronto, Ontario, Canada
- ‖‖National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland
- ¶¶Medical University of South Carolina, Charleston, South Carolina
- ##Alfred I. DuPont Hospital for Children, Wilmington, Delaware
- ∗∗∗Columbia University Medical Center, New York, New York
- †††Children’s Hospital of Los Angeles, Los Angeles, California
- ↵∗Reprint requests and correspondence:
Dr. Peter C. Frommelt, Children’s Hospital of Wisconsin 9000 West Wisconsin Avenue MS 713, Milwaukee, Wisconsin 53226.
Background In children with single right ventricular (RV) anomalies, changes in RV size and function may be influenced by shunt type chosen at the time of the Norwood procedure.
Objectives The study sought to identify shunt-related differences in echocardiographic findings at 14 months and ≤6 months pre-Fontan in survivors of the Norwood procedure.
Methods We compared 2-dimensional and Doppler echocardiographic indices of RV size and function, neo-aortic and tricuspid valve annulus dimensions and function, and aortic size and patency at 14.1 ± 1.2 months and 33.6 ± 9.6 months in subjects randomized to a Norwood procedure using either the modified Blalock-Taussig shunt (MBTS) or right ventricle to pulmonary artery shunt (RVPAS).
Results Acceptable echocardiograms were available at both time points in 240 subjects (114 MBTS, 126 RVPAS). At 14 months, all indices were similar between shunt groups. From the 14-month to pre-Fontan echocardiogram, the MBTS group had stable indexed RV volumes and ejection fraction, while the RVPAS group had increased RV end-systolic volume (p = 0.004) and decreased right ventricular ejection fraction (RVEF) (p = 0.004). From 14 months to pre-Fontan, the treatment groups were similar with respect to decline in indexed neo-aortic valve area, >mild neo-aortic valve regurgitation (<5% at each time), indexed tricuspid valve area, and ≥moderate tricuspid valve regurgitation (<20% at each time).
Conclusions Initial Norwood shunt type influences pre-Fontan RV remodeling during the second and third years of life in survivors with single RV anomalies, with greater RVEF deterioration after RVPAS. Encouragingly, other indices of RV function remain stable before Fontan regardless of shunt type. (Comparison of Two Types of Shunts in Infants with Single Ventricle Defect Undergoing Staged Reconstruction—Pediatric Heart Network; NCT00115934)
The Pediatric Heart Network (1) SVR (Single Ventricle Reconstruction) trial compared outcomes in 549 infants undergoing a Norwood procedure for single right ventricle anomalies randomized to either a modified Blalock-Taussig shunt (MBTS) or right ventricle to pulmonary artery shunt (RVPAS) at 15 North American centers (2). The trial’s primary finding was superior 1-year transplant-free survival in subjects who received a RVPAS compared with those who had an MBTS (3). However, initial shunt type for the Norwood procedure did not impact echocardiographic indices measured after shunt removal (4). Specifically, at 14 months, right ventricular (RV) systolic, diastolic, and global function; cardiac and vascular dimensions; neo-aortic and tricuspid annulus dimensions and valve function; and neo-aortic flow patterns all were similar for survivors with MBTS or RVPAS. Although significant interstage differences were noted in neo-aortic annular size and flow patterns between shunt types before stage II palliation, these were explained by the differing shunt physiologies.
On the basis of these results, we hypothesized that differences in echocardiographic findings between the shunt groups would emerge with longer-term follow-up. RV dysfunction related to the creation of a right ventriculotomy in the RVPAS group was of particular concern. The Pediatric Heart Network’s SVR Extension Study was undertaken to provide longitudinal surveillance of the cohort, including echocardiography within 6 months of Fontan palliation. In this analysis, we compared 2-dimensional (2D) and Doppler indices of RV size and function, neo-aortic and tricuspid valve (TV) annulus dimensions and function, and aortic size and patency to evaluate these changes for the second and third years of life for our cohort, picking up at 14 months when the original analysis left off and then conducting echocardiographic analysis again before the Fontan procedure. By investigating changes over time, we sought to identify echocardiographic features at 14 months that predict clinical outcomes.
Entry criteria for subjects in the SVR trial have been reported previously (2). In the current study, all SVR trial survivors who underwent a transthoracic echocardiogram within 6 months of planned Fontan palliation and had that study submitted to the SVR core echocardiography lab for interpretation were eligible for inclusion. We excluded subjects who underwent biventricular repair (n = 3), had their pre-Fontan echocardiogram outside the 6-month window before Fontan surgery (n = 2), or had no Fontan procedure scheduled at the time of this data collection (n = 16).
The SVR study design also has previously been published (2). Briefly, SVR is a randomized clinical trial of infants with a diagnosis of single, morphologically right ventricle anomaly undergoing a Norwood procedure. The primary outcome has been published (incidence of death or cardiac transplantation at 12 months after randomization) (3), as have secondary outcomes, including hospital morbidity and rate of serious adverse events through 12 months, and risk factors for mortality and cardiac transplantation (5–7). The secondary echocardiographic markers of outcome for the SVR trial previously reported included 14-month indices of RV function, cardiac and vascular dimensions, valve annulus dimensions and function, and neo-aortic flow patterns (4).
For the SVR Extension Study, an echocardiography core laboratory at the Medical College of Wisconsin reviewed 2D/Doppler echocardiograms performed at each clinical center within 6 months of a planned Fontan procedure. Core lab procedures for image analysis and data management have been previously described (4). Echocardiographic measures obtained from the imaging are summarized in Tables 1 and 2⇓⇓.
Summary descriptive statistics of echocardiographic indices are presented by intended shunt type for the Norwood procedure. Distributions of echocardiographic indices were examined with histograms. Echocardiographic indices were compared between shunt groups at the pre-Fontan study. In addition, we performed paired comparisons of changes in these indices from 14 months of age (which marked the end of the SVR study visit from previously published data) (4) and the pre-Fontan study between shunt groups when a protocol echocardiogram was available at both time periods. Continuous measures were compared between assigned shunt types with either Student or Welch’s t test or a Wilcoxon rank sum test as appropriate. Categorical measures were compared between shunt types with a Fisher exact test; ordinal measures were compared with a Fisher exact test and the Mantel-Haenszel test for trend. For paired analysis between values obtained at 14 months and the pre-Fontan study, the mean and standard deviation of the change over time in each of the echocardiographic measures were compared between shunt groups using a Student t test. Finally, echocardiographic indices obtained at 14 months were explored to assess whether they predicted transplant-free survival from the time of that study to Fontan surgery. The hazard ratios for death or heart transplant after the 14-month echocardiogram were assessed as a function of the echocardiographic measurement using Cox proportional hazards regression. The hazard ratio measures the association of the echocardiogram parameter with transplant-free survival, and the p value tests whether the hazard ratio is significantly different from 1 (no effect). Subjects who did not die or receive a heart transplant between the date of the 14-month echocardiogram and the date of Fontan surgery were censored at the time of their Fontan procedure. The p values were considered significant at <0.05.
Analyses were performed in SAS version 9.3 (SAS Institute, Cary, North Carolina). To adjust for the effect of somatic growth and age on the linear, area, and volumetric dimensions of cardiac structures and Doppler flow patterns, z scores were used when normative data were available (8). The dimensions of the neo-aortic valve annulus were expressed as z scores using normal values for the dimensions of the native aortic valve in a normal population. These z scores are calculated with respect to body surface area (BSA), with the power of BSA determined by prior work examining linearity between echo measures and various functions of BSA (9).
All echocardiographic measures were performed by a single dedicated pediatric echocardiography technician and all measures were confirmed by a single experienced pediatric echocardiographer. To access reproducibility of echocardiogram measurements, approximately 5% of the echocardiograms at each time point were randomly selected to be read blindly a second time by the same echocardiographer. The intrarater intraclass correlations and percent agreement for continuous and dichotomous variables, respectively, were calculated. Agreement was good (>80%) for most variables; RV volume assessment had more variance, but this did not affect the consistency of right ventricular ejection fraction (RVEF) estimates (86% agreement).
Of the 262 SVR survivors who had echocardiographic studies submitted within 6 months of planned Fontan, 258 (98.5%; 125 MBTS and 133 RVPAS) were acceptable for core-lab analysis. The mean age at the time of the pre-Fontan echocardiogram was 33.6 ± 9.6 months (median 32.7 months; range 13.6 to 65.9 months). Among the 258 subjects, 240 (93%; 114 MBTS and 126 RVPAS) also had undergone a 14-month protocol echocardiogram at a mean of 14.1 ± 1.2 months (median 14.1 months, range 11 to 20.9 months); of the remaining 18 subjects, 16 had no 14-month echocardiogram and 2 had undergone their Fontan at the time of or before the 14-month study. For the 240 subjects with paired echocardiograms, the mean duration between the 14-month echocardiogram and the pre-Fontan echocardiogram was 19.6 ± 9.5 months (range: 1.5 to 51.9 months).
RV systolic size and function
At the pre-Fontan echocardiogram, RVEF was significantly higher in the MBTS subjects compared with RVPAS (45 ± 6% vs. 42 ± 7%; p = 0.007). All other values at this time point were similar between shunt types including RV indexed systolic volume, indexed systolic area, percent fractional area change, peak tricuspid annular systolic velocity, and isovolumic acceleration (Table 3).
In comparing the 14-month and pre-Fontan studies, we found no significant change in RVEF or systolic volume in the MBTS group. However, RV indexed end-systolic area increased and fractional area change decreased between the 2 time periods. All other systolic indices were similar (Table 4). In the RVPAS group, the RV indexed end-systolic area and end-systolic volume both increased while fractional area change and ejection fraction (44 ± 7% vs. 41 ± 7%; p = 0.004) both decreased over time (Table 5, Figure 1). This interval change in RVEF differed significantly between groups (–3.24 for the RVPAS vs. 0.99 for the MBTS; p = 0.009). No other significant changes in systolic indices were identified from 14 months to the pre-Fontan study for the RVPAS group.
RV diastolic size and function
Diastolic indices, including RV indexed end-diastolic volume, indexed end-diastolic area, tricuspid peak E and A velocities and E/A ratio, tricuspid annular E’ and A’ velocities, E/E’ ratio, and presence of pulmonary vein flow reversal during atrial systole were similar between shunt groups at the pre-Fontan study (Table 3). In addition, the presence of pulmonary vein reversal of flow with atrial systole was rarely identified in either group (6.5% in MBTS, 2% in RVPAS; p = 0.24). When compared with measures obtained at 14 months, there was no significant change in diastolic indices over time for the MBTS group (Table 4). In the RVPAS group, tricuspid peak E and A velocities decreased and the tricuspid annular E’ velocity increased from 14 months to the pre-Fontan study, which resulted in a significantly decreased E/E’ ratio at the pre-Fontan study, bringing this measure into the normal range (E/E’ ratio for age z score 3.0 ± 2.6 at 14 months vs. 1.7 ± 2.0 pre-Fontan). All other diastolic indices were stable between the 2 intervals for the RVPAS group (Table 5).
RV global function
The myocardial performance index, as assessed by both blood flow Doppler and annular Doppler tissue imaging, was not significantly different between the shunt groups at the pre-Fontan study (Table 3). Additionally, we found no significant change in the index as calculated by either method from 14 months to the pre-Fontan study for either group (Tables 4 and 5).
Neo-aortic valve size and function
Indexed neo-aortic annular areas were similar in both shunt groups at the pre-Fontan study with marked neo-aortic annular dilation, as evidenced by mean z scores of 5.6 ± 2.5 for MBTS and 5.1 ± 3.1 for RVPAS (p = 0.22) (Table 3). Both shunt groups showed a significant decrease in indexed neo-aortic valve size from the 14-month to the pre-Fontan study, however, with an average change in z score of –1.1 for the MBTS and –1.3 for the RVPAS (Figure 2). This change over time was not significantly different between the shunt groups. The degree of neo-aortic regurgitation was similar in the 2 groups and did not vary significantly between time points. At the pre-Fontan study, 2.5% of the MBTS and 4.6% of the RVPAS cohort (p = 0.73) had moderate regurgitation and no subject had severe regurgitation.
TV size and function
Indexed tricuspid annular area and z score were similar in both groups at the pre-Fontan study, with a mean z score of 1.9 ± 2.2 for MBTS and 2.1 ± 2.0 for RVPAS (p = 0.38) (Table 3). Both groups demonstrated a significant decrease in indexed TV size from the 14-month to the pre-Fontan study. The degree of change over time was similar for both groups, with an average change in z score of –0.7 for the MBTS and –0.9 for the RVPAS (Figure 3). The degree of tricuspid regurgitation was comparable in the 2 groups, with 19% of the MBTS and 16% of the RVPAS cohort (p = 0.62) having moderate or severe regurgitation at the pre-Fontan study. The percentage of subjects with moderate or greater tricuspid regurgitation was stable when compared with the 14-month assessment; it was present in <20% in both groups at both intervals.
Aortic size and patency
The ascending aortic diameter was similar in both shunt groups at the pre-Fontan study with moderate ascending aortic dilation, as evidenced by mean z scores of 3.4 ± 2.3 for MBTS and 3.1 ± 2.4 for RVPAS (p = 0.31) (Table 3). As the ascending aorta was not quantified at the 14-month study, information about change over time was unavailable. We saw similarities in peak velocity through the distal aortic arch between the shunt groups at the pre-Fontan study (1.7 ± 0.5 for MBTS vs. 1.8 ± 0.6 m/s for RVPAS; p = 0.08). The peak velocity was also stable when compared with the 14-month assessment, with average peak velocities <2 m/s for both groups at both intervals (Tables 4 and 5).
Transplant-free survival to Fontan
Eighteen subjects died or required heart transplantation from the 14-month echocardiographic study to the Fontan procedure. In univariate analyses, a larger RV indexed end-systolic volume and area, lower RVEF, larger RV indexed diastolic volume and area, higher peak early diastolic tricuspid filling velocity, higher myocardial performance index (MPI), larger indexed tricuspid indexed annular area and z score for BSA, and at least moderate TV regurgitation were associated with an increased risk of transplant or death between the 14-month echocardiogram to Fontan palliation (Table 6). Neo-aortic valve size, degree of neo-aortic regurgitation, and aortic arch gradient at 14 months were not associated with transplant-free survival to Fontan.
The SVR trial provides a unique opportunity for longitudinal assessment of a large, high-risk cohort of children born with single right ventricle anomalies as they progress through staged surgical palliation. In previous work, initial shunt type used with the Norwood procedure was found to influence transplant-free survival at 1 year, with lower mortality in those subjects who received a RVPAS compared with a MBTS (3). However, this survival benefit appears to dissipate with time after the first year, so the best surgical strategy remains in question. In the current manuscript, we describe changes in echocardiographic indices of RV, neo-aortic, and TV size and function from completion of the initial trial (14 months, which marked the end of the primary SVR study visit) until subsequent Fontan surgery was performed.
Preservation of single RV function from the time of the Norwood onward is paramount for children undergoing staged palliation as the resultant physiology depends on well-maintained RV function to avoid chronic cardiovascular morbidities including heart failure, frequent hospitalizations, arrhythmias, and transplant, as well as early mortality. Here, we found that initial shunt type appears to influence RV remodeling during the second and third years of life before Fontan surgery with deterioration in RVEF in those with an RVPAS (Central Illustration). This change is associated with poorer transplant-free survival during the same time period for subjects with an RVPAS (10), raising concerns that this surgical strategy may result in progressive deleterious effects on RV systolic function that could have important implications for long-term morbidity and mortality. Although previous single-center studies have also shown improved early outcomes with the RVPAS compared with the MBTS (11–14), concerns about poorer qualitative (15,16) and quantitative (17) RV systolic function before and after the final stage Fontan operation have been described in patients who had the RVPAS. It has been hypothesized that the ventriculotomy required for the RVPAS may cause myocardial injury and/or scar with aneurysm formation, and that this injury may ultimately affect ventricular performance (18). If true, it is possible that the effects of this early myocardial injury take time to manifest in a way that echocardiography can measure. No difference was detected in RV fractional area change, but this method does not evaluate the outflow tract, which could potentially be the area most affected by the ventriculotomy scar. RVEF, as calculated here by the biplane pyramidal method (19), is likely a more sensitive tool for RV systolic function in this cohort.
Review of other echocardiographic indices that characterize ventricular function in this cohort did not provide a clear picture of RV failure in the RVPAS survivors before Fontan, which suggests that conclusions about the shunt-influenced health of the RV may require additional longitudinal follow-up. Indexed RV end-diastolic volumes and areas were stable and did not show the expected progressive chamber dilation of a failing ventricular pump in either shunt group between 14 months and the pre-Fontan studies. Global indices of function were also remarkably stable and similar for both shunt groups over time as assessed by the myocardial performance index, with the RV MPI (as calculated by inflow Doppler) well below (i.e., much better than) previously published values for children with single ventricle anomalies after staged palliation using the same MPI technique (20). Additionally, RV diastolic indices, which typically show the earliest pathologic changes with progressive ventricular dysfunction, are also stable with little evidence of progressive restrictive disease in either group as evidenced by the absence of subjects who exhibited late diastolic pulmonary venous reversal of flow. In fact, diastolic function appears to improve over time in the RVPAS survivors from the 14-month to the pre-Fontan evaluation, with a significantly decreased tricuspid E/E’ ratio that falls within 2 standard deviations of age-matched normal controls.
Neo-aortic and TV size and function also exhibited reassuring trends from 14 months to the pre-Fontan echocardiogram. Although the neo-aortic annulus remained markedly dilated compared with normative controls at both intervals, there was a significant decrease in indexed annular size and z score over time in both shunt groups. Despite the significant neo-aortic annular and ascending aortic dilation, more than mild valve regurgitation was rare (<5%) and similar at both intervals for the 2 shunt groups. Neo-aortic valve size and integrity will be important to track in this cohort over time, given reports of significant neo-aortic root dilation and neo-aortic valve regurgitation later in childhood after Norwood palliation with MBTS (21). In our cohort, it is similarly encouraging that indexed tricuspid annular size also decreased at the pre-Fontan study compared with the 14-month assessment in both shunt groups, nearly normalizing in size with z scores approximating 2. The incidence of moderate or greater tricuspid regurgitation was stable (<20% at both intervals for both groups), a finding that differs from a recently published report that identified worse TV function in the MBTS patients after Fontan (22).
After their first year, children with single RV anomalies remain at risk for death or transplant before Fontan surgery. Not surprisingly, transplant-free survival to Fontan was associated with significantly smaller RV indexed end-diastolic and end-systolic volumes and areas, higher RV percent fractional area change and RVEF, lower myocardial performance index, smaller TV indexed annular area, and a lower incidence of moderate or worse TV regurgitation at 14 months. This suggests that patients with significantly larger RV volumes, decreased RV systolic and global functional indices, TV annular dilation, and significant TV regurgitation warrant closer surveillance.
Unlike for the left ventricle, 2D echocardiographic tools used to assess RV volume and systolic function remain limited due to the chamber’s complex geometry. Analysis of regional RV wall motion was not performed as part of this protocol, so the impact of focal scarring/dyskinesis was not specifically assessed. Careful training in protocol image acquisition was provided and reinforced at all sites with regular quality assurance feedback from the core lab to optimize appropriate image capture; consequently, >98% of the submitted studies were found to be acceptable in providing images that allowed data extraction. However, while data extraction was possible in nearly all studies, not all measurements were achievable in the majority of studies because of incomplete or inadequate image acquisition. The impact of this limitation in the characterization of RV function for this cohort is unclear but emphasizes the challenge of multi-institutional echocardiographic trials that require extensive data acquisition, particularly when obtaining those data from young children with complex heart disease.
Initial shunt type used with the Norwood procedure may influence RV remodeling during the second and third years of life before Fontan surgery in survivors with single RV anomalies, as evidenced by decreased RVEF after the RVPAS that is not seen after MBTS. Other echocardiographic measures of RV function, however, appear to remain stable in both shunt groups. In addition, neo-aortic and TV size decrease in both shunt groups over time, and both valves continue to function well for most survivors without progressive regurgitation before the Fontan. Factors associated with poorer transplant-free survival to the Fontan operation included echocardiographic evidence of markedly enlarged RV volumes, severely abnormal RV systolic and global indices of function, significant tricuspid annular dilation, and at least moderate tricuspid regurgitation at 14 months.
COMPETENCY IN MEDICAL KNOWLEDGE: Initial shunt type used for the Norwood procedure may influence RV remodeling during the second and third years of life before Fontan surgery in survivors with single RV anomalies, as evidenced by decreased RVEF after the RVPAS procedure that is not seen in those after use of the MBTS. Other echocardiographic measures of RV and valve function, however, appear to remain stable in both shunt groups.
COMPETENCY IN PATIENT CARE: The best choice of initial shunt type in infants requiring the Norwood procedure for single RV anomalies remains unclear on the basis of current data from the SVR trial.
TRANSLATIONAL OUTLOOK: Longitudinal follow-up of the SVR trial cohort is ongoing and will likely provide further insights into the long-term effect of initial shunt type on clinical outcome and echocardiographic indices of cardiac size and function.
The study was supported by U01 grants from the National Heart, Lung, and Blood Institute (HL068269, HL068270, HL068279, HL068281, HL068285, HL068292, HL068290, HL068288, HL085057, HL109781, HL109737). The contents of this work are solely the responsibility of the authors and do not necessarily represent the official views of the National Heart, Lung, and Blood Institute. Dr. Hill has received grant support from the Gilead Scholars in Cardiology Program. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- body surface area
- modified Blalock-Taussig shunt
- myocardial performance index
- right ventricular
- right ventricular ejection fraction
- right ventricle-to-pulmonary artery shunt
- single ventricle reconstruction
- tricuspid valve
- Received August 5, 2014.
- Accepted August 13, 2014.
- American College of Cardiology Foundation
- Ohye R.G.,
- Gaynor J.W.,
- Ghanayem N.S.,
- et al.
- Frommelt P.C.,
- Guey L.L.,
- Minich L.L.,
- et al.
- Hourihan M.,
- Colan S.D.,
- Wernovsky G.,
- et al.
- Sluysmans T.,
- Colan S.D.
- Newburger J.W.,
- Sleeper L.A.,
- Frommelt P.C.,
- et al.
- Pizarro C.,
- Malec E.,
- Maher K.O.,
- et al.
- Malec E.,
- Januszewska K.,
- Kolcz J.,
- Mroczek T.
- Cohen M.S.,
- Marino B.S.,
- McElhinney D.B.,
- et al.