Survey of how pediatric cardiologists noninvasively evaluate patients with hypoplastic left heart syndrome

Comparison of Quantitative Versus Qualitative Assessment of Single Right Ventricular Function for the Prediction of Midterm Outcomes

Multiple studies have shown that quantitative evaluation of right ventricular (RV) systolic function in children with hypoplastic left heart syndrome (HLHS) is associated with outcomes. However, the most widely used method is qualitative, or subjective echocardiographic evaluation. Tricuspid annular displacement (TMAD) is a quantitative method and has been shown to be associated with transplant/death in a cohort of pediatric patients with HLHS. In this study, the same echocardiograms used in the quantitative TMAD study were qualitatively evaluated to see if the assessment correlated with midterm outcomes. We hypothesized that TMAD measures would outperform qualitative measurements. A previously published retrospective study of patients with systemic right ventricle demonstrated that TMAD measurements of function prior to the Glenn procedure were associated with midterm mortality/transplant (mean TMAD 12.4% for survivors and 10.0% for non-survivors/transplant, p = 0.03). Echocardiographic images used in that study were re-evaluated using qualitative assessment of function. A score between severely depressed (0) and supra-normal (7) was assigned independently by three cardiologists. A chart review included short-term clinical outcome measures in addition to midterm mortality/transplant. Spearman correlations and logistic regression were used to estimate the associations between function scores and clinical outcomes. Function scores were measured with acceptable inter- and intra-rater reliabilities in 47 patients who had echocardiograms used in the prior analysis before their bidirectional Glenn. The mean functional score was 5.3 ± 1.32 in survivors and 5.0 ± 1.02 in non-survivors/transplant. Subjective echocardiographic scores of function were not predictive of mortality (p = 0.23). Scores were correlated with length of cardiac intensive care unit stay (Spearman's rho =  - 0.31, p = 0.04), but not post-Glenn total length of hospital stay (p = 0.4). A sub-analysis was performed for each individual operator. Only one of three operators produced scores with a significant association with mortality/transplant (p = 0.01, p = 0.25, p = 0.22, respectively). Averaged subjective measurement of function by three pediatric cardiologists in children with single right ventricle prior to the Glenn procedure was not associated with midterm outcomes. Previous work based on the same echocardiograms showed that quantitative analysis was associated with midterm outcomes. This suggests quantitative analysis may be more useful in prognostication. Future studies could confirm these results and identify which quantitative methods are most helpful.

Systolic/diastolic ratio correlates with end diastolic pressures in pediatric patients with single right ventricles

BACKGROUND

Increased ventricular end-diastolic pressure (VEDP) is a known risk factor for morbidity and mortality in patients with single right ventricle (RV) physiology. Previous studies have shown mixed results correlating echocardiographic measurements with catheter-derived VEDP in this population. Goal of this study was to evaluate if echocardiographic systolic/diastolic ratio (S/D) correlated with VEDP.

METHODS

Patients with single RV physiology who underwent simultaneous echocardiography and catheterization were evaluated. Systolic and diastolic durations were measured using tricuspid inflow durations from Doppler analysis to calculate the S/D ratio. VEDP was obtained from the catheterization report.

RESULTS

Twenty-seven studies were performed on patients with single RV physiology. Median age at time of catheterization was 11.4 months (range, 0-132 months). Mean VEDP was 9.9 ± 4.5 mm Hg. S/D ratio was 1.8 ± 0.5. S/D ratio significantly correlated with VEDP (r = 0.63, P < .01). Optimum value of S/D ratio for discriminating between patients with high (>10 mm Hg) vs low EDP was found to be 1.9. High S/D ratio had an area under the curve of 0.82 (0.65, 1.0), with 75% sensitivity and 89% specificity for predicting elevated VEDP.

CONCLUSION

In patients with single RV physiology, S/D significantly correlated with VEDP. S/D ratio is a simple technique that may be useful in both estimating and discriminating between high and low VEDP in this complex patient population.

Right Ventricular Systolic Function Parameters in Hypoplastic Left Heart Syndrome

Assessment of the systolic function of the right ventricle (RV) in patients with hypoplastic left heart syndrome (HLHS) is important. The asymmetric shape and heavy trabeculations make accurate assessment of RV systolic function challenging. Novel measures of RV function could be helpful in distinguishing reduced versus preserved function in HLHS and may also be worse in HLHS with preserved function compared to normal controls. These novel methods offer promise, but research and clinical applicability is hindered as no cut-off values for normal function have been established. We performed a retrospective comparison of functional assessments from echocardiograms of HLHS patients with preserved and reduced RV function along with a control group of normal patients. Measures of function included fractional area change (FAC), tissue motion annular displacement of the tricuspid (TMAD-TV) and pulmonary valves (TMAD-PV), myocardial performance index (MPI), tricuspid tissue Doppler S' velocity, and RV global longitudinal strain (RVGLS). Comparisons were made between three groups: normal patients, HLHS with preserved function, and HLHS with reduced function defined as FAC < 35%. FAC was chosen as the reference as it is a historical standard. 41 HLHS patients were studied. Of these patients, 20 had HLHS with reduced function, and 21 had preserved function. They were compared with 27 age-matched, normal, controls. Comparison between HLHS and normal controls: in HLHS with preserved RV systolic function, compared to normal controls, tissue Doppler S', MPI, and TMAD-TV were all abnormal (all p < 0.05). RVGLS was not statistically different (20.5 ± 3.6% for normal vs. 17.9 ± 2.6% for HLHS with preserved function). TMAD-PV was similar between groups (16.1 ± 4.6% vs. 16.7 ± 5.1%). All measures were significantly worse (all p < 0.05) in the HLHS with reduced function group compared to normal controls. Comparison between HLHS preserved vs reduced function: in HLHS with reduced function defined by FAC < 35%, all measures were significantly worse compared to HLHS with preserved function (all p < 0.05). The cut-off values that correspond to a FAC of > 35% were 14.5% for TMAD-TV and 16% for RVGLS. All measures except RVGLS and TMAD-PV estimated worse function than controls even for HLHS with preserved function. Each of the functional measures was able to identify preserved vs reduced function in HLHS with FAC as the reference standard. Cut-off values between preserved and reduced function in HLHS were estimated for TMAD-TV and RVGLS based on a relatively small cohort. These cut-off values will aid in the research design of future studies.

Right Ventricular Longitudinal Shortening Before and After Stage I Surgical Palliation Correlates with Outcomes

Quantification of right ventricular function is difficult, but important, in patients with single ventricles. Tissue motion annular displacement (TMAD) is an echocardiographic tool that measures displacement of the tricuspid valve relative to the apex. We evaluated TMAD, lateral annular displacement (LAD), and fractional area change (FAC) for correlation with outcomes. We measured TMAD, LAD, FAC, and other variables that may affect prognosis in patients with single right ventricle physiology pre- and post-Stage I palliation and correlated them with outcomes up to the Glenn procedure. Intra- and inter-observer variability for TMAD measurements were 2.7% (1.2-3.5%) and 6.1% (3.3-8.1%), respectively. Sixty-six subjects met the inclusion criteria. Pre-Stage I TMAD was 13.7% (SD 3.9%). TMAD had a linear relationship with FAC (r2 = 0.76). There was a correlation between TMAD and hospital stay (p = 0.044) and ECMO/arrest (p = 0.024). LAD correlated with ECMO/arrest (p = 0.045) and mortality/transplant (p = 0.049). FAC correlated with in-hospital mortality (p = 0.028). Post-Stage I TMAD was 11.8% (SD 3.7%). TMAD, LAD, and FAC all correlated with in-hospital mortality and mortality/transplant. In multivariate models, TMAD was independently predictive of weight for age Z score pre-Glenn. TMAD, FAC, and LAD correlate with clinically significant outcomes after the first-stage palliation. TMAD correlated with more outcomes than FAC and was the only measure that was independently predictive of any outcome. TMAD is a reproducible measure of RV function in this population. TMAD has prognostic value before and after first-stage palliation and may outperform more traditional measures.

Right ventricular systolic function in hypoplastic left heart syndrome: A comparison of manual and automated software to measure fractional area change

BACKGROUND

Quantitative echocardiographic assessment of right ventricular function is important in children with hypoplastic left heart syndrome (HLHS). The aim of this study was to examine the repeatability of different echocardiographic techniques, both manual and automated, to measure fractional area change (FAC) in patients with HLHS and to correlate these measurements with magnetic resonance imaging (MRI)-derived ejection fraction (EF).

METHODS

Fifty-one children with HLHS underwent transthoracic echocardiography and cardiac MRI under the same general anesthetic as part of routine inter-stage assessment. FAC was measured from the apical four-chamber view using three different techniques: velocity vector imaging (VVI) (Syngo USWP 3.0; Siemens Healthineers), QLAB (Q-lab R 10.0; Philips Healthcare), and manual endocardial contour tracing (Xcelera, Philips Healthcare). Intra- and inter-observer variability was calculated using intra-class correlation coefficient (ICC). FAC was correlated with MRI EF calculated using a single standard method.

RESULTS

Fractional area change had a good correlation with MRI-derived EF with an R value for VVI, QLAB, and manual methods of .7, .6, and .4, respectively. Intra- and inter-observer variability for FAC was good for automated echocardiographic methods (ICC>.85) but worse for manual method particularly inter-observer variability of FAC and end-systolic area. Both automated techniques tended to produce higher FAC values compared with manual measurements (P<.001).

CONCLUSION

Automation improves the repeatability of FAC in HLHS. There are some differences between automated software in terms of correlation with MRI-derived EF. Measurement bias and wide limits of agreement mean that the same echocardiographic technique should be used during the follow-up of individual patients.

Hybrid Palliation for Ductal-Dependent Systemic Circulation

We reviewed our hybrid palliation experience for 91 neonates, with ductal-dependent systemic circulation, born between August 2007 and October 2015. For analysis, we stratified the 91 patients by a risk factor (RF) score and divided them into three groups: (1) high-risk two-functional ventricles (2V) median RF score of 3 (N = 20); (2) low-risk one-functional ventricle (1V) RF score 0-1 (N = 32); and (3) high-risk 1V RF score ≥2 (N = 39). Midterm survival (median 4 years) by group was: (1) 95 %, (2) 91 %, and (3) 15 %, (p = 0.001). In conclusion, hybrid palliation was associated with excellent midterm results for high-risk 2V and low-risk 1V patients with ductal-dependent systemic circulation. In contrast, high-risk 1V patients had significantly worse outcomes.