Use of Right Ventricular Free-Wall Strain in a Multivariable Estimate of Right Ventricular Arterial Coupling in Pediatric Pulmonary Arterial Hypertension

Abstract

Background:

Right ventricular-arterial coupling (RVAC) describes the relationship between right ventricular (RV) contractility and pulmonary vascular afterload. Non-invasive surrogates for RVAC using echocardiographic estimates of RV function, such as tricuspid annular plane systolic excursion (TAPSE), have been shown to correlate with invasively measured RVAC and predict clinical outcomes in pediatric pulmonary arterial hypertension (PAH). However, given the limitations of TAPSE at accurately estimating RV function in children, we hypothesized that a multivariable estimate of RVAC using RV Free-Wall Longitudinal Strain (RVFW-LS) may perform better than those utilizing TAPSE at predicting clinical outcomes.

Methods:

108 children from two institutions with PAH underwent hemodynamic catheterization with simultaneous echocardiography. In a retrospective analysis, hybrid (echo and invasive) RVAC metrics included TAPSE/pulmonary vascular resistance (PVRi) and RVFW-LS/PVRi. Non-invasive echocardiographic metrics were TAPSE/ echo-derived Pulmonary Artery Systolic Pressure (PASP) and RVFW-LS/PASP.

Results:

RVFW-LS correlated with PVRi (r= 0.315, P=0.01), though TAPSE did not (r= 0.058, P=0.64). PVRi, PASP, and RVAC metrics declined in patients with worse World Health Organization Functional Class (n=108), while TAPSE and RVFW-LS did not. PVRi, PASP, RVFW-LS/PVRi, TAPSE/PVRi, and RVFW-LS/PASP predicted the outcome variable of transplant or death (AUC 0.771 (p<0.001), 0.729 (p=0.004), 0.748 (p=0.002), 0.732 (p=0.009), and 0.714 (p=0.01), respectively), while TAPSE/PASP, RVFW-LS, and TAPSE did not (AUC 0.671, 0.603, and 0.525). In patients without a history of repaired congenital heart disease (CHD) (n=88), only RVFW-LS/PASP, PVRi, PASP, and RVFW-LS/PVRi predicted outcomes (AUC and 0.738 (p=0.002), 0.729 (p=0.01), 0.729 (p=0.01), 0.729 (p=0.015), respectively).

Conclusions:

In the pediatric population, baseline PVRi and echo-estimated PASP were strongly associated with adverse clinical outcomes, but TAPSE and RVFW-LS were not. Estimates of RVAC utilizing RVFW-LS were superior to those utilizing TAPSE – however only marginally additive to PASP and PVRi at predicting the adverse clinical outcome in patients without a history of repaired congenital heart disease.

Graphical Abstract

Introduction

The ability of the right ventricle (RV) to deliver an entire cardiac output into the pulmonary vascular circulation in an energy-efficient manner is known as right ventricular-arterial coupling (RVAC). RVAC is increasingly recognized as an important factor in patients with pulmonary arterial hypertension (PAH), where the load imposed on the RV is often severe and chronic. In patients with PAH, the initial rise in pulmonary vascular resistance is met with adaptive responses from the RV which helps maintain cardiac output. However, after prolonged exposure to a high afterload state, the RV progressively dilates and becomes fibrotic due to increased wall-stress¹. Mathematically, RVAC is expressed as the ratio between a marker of RV performance over pulmonary arterial (PA) afterload. The RV and PA are considered “uncoupled” when the RV is no longer able to provide adequate cardiac output in the face of increased afterload. This ultimately leads to clinically apparent RV failure, which is a common cause of mortality in pediatric patients with PAH².

The gold-standard for measuring RVAC is the ratio of end-systolic elastance to arterial elastance (Ees/Ea) from invasively-derived pressure volume loops³. Previously identified Ees/Ea thresholds have been identified and are thought to predict impending RV failure in adult patients with PAH⁴. However, this measurement is technically challenging and not routinely obtained in children, even those undergoing right heart catheterization. Therefore, there is keen interest in alternative assessment of RVAC⁵. Indeed, prior studies have found multiparametric variables which estimate RVAC provide important mechanistic insights into RV failure and improve prediction of clinical outcomes over echo parameters alone³. These multivariable parameters can either be hybrid (utilizing echo parameters of RV function and invasively-measured afterload), or fully non-invasive (utilizing echo measurements only).

The ratios of tricuspid plane annular systolic excursion (TAPSE) and indexed pulmonary vascular resistance (PVRi), or TAPSE with echo-estimated systolic pulmonary artery pressure (PASP) have been show to correlate with invasively-derived Ees/Ea in adults and are associated with worsening clinical status in pediatric PAH patients^5,6. TAPSE/PASP was included as an additional diagnostic tool in the recent European guidelines for diagnosis of PAH in adult patients⁷. There are also data to suggest TAPSE/PASP may help improve risk stratification in adults⁸. However, interest has turned toward utilizing RV free-wall longitudinal strain (RFVW-LS) instead of TAPSE for RVAC assessment. RVFW-LS is thought to provide a more accurate assessment of RV myocardial function than TAPSE, which can be spuriously normal due to tethering and basal RV rocking motion despite reduced RV contraction^9,10. Studies in adult PAH patients have recently validated RVFW-LS-based metrics to invasively measured Ees/Ea and found them to be superior to those using TAPSE in predicting clinical outcomes^11,12.

The use of TAPSE in pediatric PAH is particularly fraught with challenges, given that it is preserved late into the disease course and does not correlate well with clinical outcomes, especially in children who do not have a history of repaired congenital heart disease (CHD) ¹³. Therefore, TAPSE may not be a sufficiently sensitive measure of RV dysfunction in pediatric PAH secondary to idiopathic, heritable, or lung-disease related causes.

The aim of this study was to evaluate if hybrid and non-invasive multiparametric estimates of RVAC using RVFW-LS could better predict clinical outcomes than metrics using TAPSE in pediatric patients with PAH of various etiologies. We hypothesized that RVFW-LS/PVRi and RVFW-LS/PASP would be better than RVFW-LS alone at correlating with patient functional class given strain is still a load dependent variable, so accounting for RV afterload should improve its clinical relevance. Furthermore, because RVFW-LS should provide a more nuanced evaluation of RV mechanics, we proposed the RVFW-LS derived metrics would more accurately associate with clinical outcomes than the previously validated metrics of TAPSE/PVRi and TAPSE/PASP.

Methods

Study Population

All data from this study were available to the corresponding author who takes responsibility for its integrity. Children and adolescents with PAH secondary to a variety of etiologies from two centers - Children’s Hospital Colorado and the Hospital for Sick Children in Toronto - underwent simultaneous transthoracic echocardiograms during clinically-indicated right-heart catheterization between November 1^st 2008 and December 31^st 2021. This study was part of a larger prospective study on ventricular-ventricular interactions in patients with pulmonary arterial hypertension. We previously published data pertaining to left ventricular performance¹⁴, ventricular interdependence¹⁵, and eccentricity index¹⁶ from this cohort. For this analysis, we retrospectively analyzed 108 patients (18 of whom came from the Toronto cohort) who underwent standard invasive hemodynamic measurements performed under general anesthesia (Table 1).

Table 1.

Patient Characteristics

	N = 108
Age (yrs)	9 (4.2–15)
Female/Male	66 (61.1%)/42 (38.9%)
Weight (kg)	33.4 (17.9 – 54.9)
Etiology of Pulmonary Hypertension
Group 1 (Idiopathic)	54 (50%)
Group 1 (Heritable, CHD, CTD)	24 (22%)
Group 2 (Left-heart disease)	2 (1.9%)
Group 3 (Lung/hypoxia related)	12 (11.1%)
Group 4 (Chronic Thromboembolic)	2 (1.9%)
Group 5 (Multifactorial)	14 (13.0%)
WHO Class
I	30 (27%)
II	46 (42.6%)
III	14 (13%)
History of Repaired CHD	20 (18%)
Type of CHD
ASD	10 (50%)
VSD	3 (11%)
AVSD	3 (15%)
PDA	9 (45%)
Initial Cardiac Catheterization	29 (27%)
Right-to-left Shunt at Time of Cath	7 (6.5%)
History of prior positive response to vasoreactivity testing	21 (19.4%)
Need for Transplant (heart and/or lung)	5 (4.6%)
Death	7( 6.5%)
PAH Medications
None	8 (7.4%)
Oxygen	30 (27%)
PDE-5-Inhibitors	73 (68%)
Endothelin Receptor Antagonists	45 (42%)
Prostacyclin Analogs	28 (26%)
Calcium Channel Blockers	17 (16%)

Values are given as medians (IQR) or n (%). CTD indicates connective tissue diseases; CHD, congenital heart disease; ASD, atrial septal defect; VSD, ventricular septal defect; AVSD, atrioventricular septal defect; PDA, patent ductus arteriosus; PDE-5, phosphodiesterase-5.

Right Heart Catheterization

Pressures were directly measured in the right atrium, RV, and pulmonary arteries. Here we report systolic pulmonary artery pressure (SPAP) as well as the mean pulmonary artery pressure (mPAP). Pulmonary capillary wedge pressure was used as a surrogate for left atrial pressure when the left atrial pressure was not directly accessed. Using the pulmonary blood flow (Qp) derived from either thermodilution (if no shunt was present) or by the Fick calculation, the systemic and indexed pulmonary vascular resistance (PVRi) were calculated.

Echocardiography

Using a General Electric Vivid 7 or E95 system (General Electric Healthcare), transthoracic echocardiographic images were obtained from standard imaging windows. With the catheter in the main pulmonary artery, tricuspid regurgitation jet peak velocity (TR Vmax), RVFW-LS, and TAPSE were measured from the apical four-chamber view. TAPSE Z-scores were calculated for patients under 18 years of age using published reference values from a large cohort of healthy children^17,18. For speckle-tracking echocardiography analysis (EchoPAC v.113, GE Healthcare), the RVFW mid-myocardial borders were traced, and peak values averaged over 2–4 consecutive beats. Accuracy of tracking was assessed visually and by EchoPAC. As long as 4 or greater segments tracked accurately, the strain curve was accepted. Segments with poor imaging quality or inadequate tracking were excluded. RVFW-LS Z-scores were calculated using previously published reference values in children ¹⁹. PASP was calculated from the peak velocity of the tricuspid regurgitation jet using the modified Bernoulli equation (4 x Vmax²), not accounting for right atrial pressure. At the time of making echocardiographic measurements, physicians were blinded to clinical variables. Clinical outcome data and World Health Organization (WHO) functional classes were obtained from the most recent clinic note at the time of cardiac catheterization. Hybrid RVAC metrics were defined as those that utilized echocardiographic estimates of RV function and invasively-measured RV afterload (TAPSE/PVRi and RVFW-LS/PVRi). Non-invasive metrics were defined as those utilizing echocardiographic metrics alone (TAPSE/PASP and RVFW-LS/PASP). Ratios utilizing RV strain were expressed as absolute values, for improved generalizability.

Statistical Analysis

Data were analyzed using SPSS (IBM Corp., version 23, Armonk, N.Y., USA). Spearman correlations were utilized to assess relationships between continuous variables. Medians were compared between WHO functional classes by Kruskal-Wallace H-test. Receiver Operating Characteristic (ROC) curves were generated for outcome prediction analysis. A combined adverse event outcome variable was created for patients with the following outcomes: death, heart transplant, lung transplant, or heart/lung transplant (n=12). The study was approved by the Institutional Review Board at both institutions and informed consent was obtained from the patient or legal guardian.

Results

Patient Characteristics:

Out of 108 patients (median age: 9.0 years; range 4 months – 23 years; 61% female), 54 (50%) were diagnosed with idiopathic PAH (Table 1). Twenty (18%) had a history of repaired CHD at the time of their simultaneous catheterization and echo; all were acyanotic lesions and included atrial septal defects (ASD) (n=10), ventricular septal defects (VSD) (n=3), atrioventricular septal defect (AVSD) (n=3), and patent ductus arteriosus (n=9). The remaining patients had a variety of etiologies including heritable PAH, chronic thromboembolic disease, underlying lung disease, PAH related to connective tissue disease, and multifactorial etiologies (often the above conditions and obstructive sleep apnea). Invasive hemodynamic data, echocardiographic measurements, as well as non-invasive RVAC metrics are presented in Table 2. There were no significant differences between RVAC metrics in patients from Colorado and Toronto, so to increase statistical power, these patients were analyzed together (Table S1)

Table 2.

Catheter Derived Hemodynamics and Echocardiographic Variables

Invasive Hemodynamics
RV Systolic Pressure (mmHg)	54.0 (41.5 – 64.0)
RV End-diastolic Pressure (mmHg)	8.0 (6.0 – 9.0)
Systolic Pulmonary Artery Pressure (mmHg)	52.0 (40.0 – 64.3)
Mean Pulmonary Artery Pressure (mmHg)	31.5 (25.0 – 42.5)
Pulmonary Capillary Wedge Pressure (mmHg)	9.0 (7.0 – 10.0)
Indexed Pulmonary Vascular Resistance (Wu*m2)	6.5 (3.9 – 10.2)
Echo Metrics
LV Ejection Fraction (%)	70.0 (64.0 – 76.0)
LV Global Longitudinal Strain (%)	−17.8 (−16.2 – −20.0)
RV End-Diastolic Area (cm2)	13.6 (10.4 – 21.4)
Systolic Eccentricity Index	1.3 (1.1 – 1.5)
TAPSE (cm)	1.6 (1.4 – 2.0)
TAPSE Z-score	−2.2 (−3.8 – −0.32)
RVFW-LS (%)	−20.9 (−24.0 – −17.6)
TR-Peak Velocity (m/s)	3.3 (2.7 – 3.7)
Non-invasive RVAC Metrics
TAPSE/PVRi	0.25 (0.16 – 0.39)
TAPSE/PASP	0.45 (0.29 – 0.58)
RVFW-LS/PVRi	2.91 (1.9 – 4.9)
RVFW-LS/PASP	0.46 (0.33 – 0.70)

Values are given as medians (IQR) or n (%). TAPSE indicates Tricuspid Annular Systolic Plane Excursion; RVAC, Right Ventricular Arterial Coupling, RVFW-LS, Right Ventricular Free Wall Longitudinal Strain; PVRi, indexed pulmonary vascular resistance; PASP, echo-estimated pulmonary artery systolic pressure.

Hemodynamic Evaluation:

Correlation coefficients quantifying associations between echo metrics and invasive hemodynamics are shown in Figure 1. TAPSE and TAPSE Z-scores were not significantly correlated with mPAP (r =0.20, p =0.11; r=−0.06, p=0.63 respectively) or PVRi (r=0.14, p=0.26; r=−0.13, p =0.33 respectively). Patients with worse RVFW-LS tended to have higher mPAP (r=0.30, p =0.01) and PVRi (r =0.32, p=0.01). RVFW-LS Z scores trended similarly, but were not significantly correlated with either mPAP (r=0.22, p=0.08) or PVRi (r=0.24, p=0.06) (Figure S1). Higher TR Vmax was strongly associated with having higher mPAP (r=0.75 p <0.001) or PVRi (r=0.64 p <0.001).

Figure 1.

Correlations between echocardiographic metrics and invasive hemodynamics. Scatterplots show associations between echocardiographic metrics and invasively measured PVRi (left column) and mPAP (right column). r -values represent Spearman’s rho. **, P<0.001; *, P<0.05. TAPSE indicates Tricuspid Annular Systolic Plane Excursion; RVFW-LS, Right Ventricular Free Wall Longitudinal Strain; TR Vmax, maximum velocity of the tricuspid valve regurgitation jet; PVRi, indexed pulmonary vascular resistance; mPAP, mean pulmonary artery pressure.

Associations with Outcomes:

There was no significant difference in the median RVFW-LS or TAPSE between WHO classes (H(2) 5.2, p=0.074 and H(2) 2.0, p=0.373, respectively) (Figure 2). RVFW-LS Z-score and TAPSE Z-score were similarly unchanged across classes (Figure S2.) However, both PVRi and echo-estimated PASP worsened significantly across higher functional classes (H(2) 25.45 and H(2) 22.87; p<0.001 for both). All RVAC metrics (RVFW-LS/PVRi, TAPSE/PVRi, RVFW-LS/PASP, and TAPSE/PASP) decreased with worsening WHO class (H(2) 17.03, 20.1, 9.9, and 16.12, respectively; P<0.001 for all).

Figure 2.

Stratification of Echocardiographic and RVAC metrics across WHO functional classes. There were no significant differences between median RVFW-LS (a) or TAPSE (b) across WHO functional class. The median PVRi was significantly lower in patients with WHO FC 1 compared to FC 2 and FC 3, as it was lower in WHO FC 2 compared to FC3. The median echo estimated PASP was significantly lower in patients with WHO FC 1 and FC3, as well as between WHO FC 2 and FC3. The median RVFW-LS/PVRi (b) and TAPSE/PVRi (e) was significantly lower in patients with WHO FC 3 or WHO FC 2 compared to those with WHO FC 1. The median RVFW-LS/PASP (c) and TAPSE/PASP (f) was significantly lower in patients with WHO FC 3 compared to those with FC 1 and in those with WHO FC 2 compared to those with WHO FC 3. TAPSE indicates Tricuspid Annular Systolic Plane Excursion; RVFW-LS, Right Ventricular Free Wall Longitudinal Strain; PASP, Pulmonary Artery Systolic Pressure (estimated by echocardiogram); PVRi, indexed pulmonary vascular resistance; WHO, World Health Organization. “x” in the box denotes the mean, horizontal line indicates the median. Vertical bars denote interquartile range; **, P<0.001; *,P<0.05.

A multivariate ROC analysis was performed using isolated markers of pulmonary afterload and resistance (PVRI and echo-derived PASP), isolated markers of right ventricular function (TAPSE and RVFW-LS), and the coupling metrics (RVFW-LS/PVRi, TAPSE/PVRi, RVFW-LS/PASP, and TAPSE/PASP) with the clinical endpoint of death or need for transplant. PVRI, PASP, RVFW-LS/PVRi, TAPSE/PVRi, and RVFW-LS/PASP were all associated with the clinical endpoint. (AUC 0.771 (p<0.001), 0.729 (p=0.004), 0.748 (p=0.002), 0.732 (p=0.009), and 0.714 (p=0.014),respectively), while TAPSE/PASP, RVFW-LS, and TAPSE were not (AUC 0.671 (p=0.059), 0.603 (p=0.094), and 0.525 (p=0.813), respectively).

In a sub-analysis of patients without history of repaired CHD (n=88), a similar multivariate ROC analysis was performed. In this subset, RVFW-LS/PASP, PVRi, PASP, and RVFW-LS/PVRi were associated with the clinical endpoint of death or transplant (AUC and 0.738 (p=0.002), 0.729 (p=0.01), 0.729 (p=0.01), 0.729 (p=0.015), respectively), while TAPSE/PVRi, TAPSE/PASP, RVFW-LS and TAPSE were not (AUC 0.668 (p=0.128), 0.650 (0.140), 0.612 (p=0.238), and 0.433(p=0.561), respectively) (Figure 3b).

Figure 3.

Associations with outcomes. (a) Receiver Operating Characteristics (ROC) curves showing each variable’s association with an adverse outcome of death or need for transplant in all patients. (b) ROC curves showing each variable’s ability to predict and adverse outcome of death or need for transplant in patients with PAH not related to underlying CHD. **=P<0.005 *=P<0.05. P-values indicate significance for AUC of variables compared to reference line. RVFW-LS, Right Ventricular Free Wall Longitudinal Strain; PASP, Pulmonary Artery Systolic Pressure (as measured by echocardiogram); PVRi, Indexed Pulmonary Vascular Resistance; TAPSE, Tricuspid Annular Systolic Plane Excursion.

Discussion

Here we sought to compare two different echocardiographic markers of right ventricular function: TAPSE and RVFW-LS, in multivariable estimates of RVAC. However, our results suggest that at a given point in time, PVRi and echo-estimated PASP alone are superior to either RV function measurements in their association with clinical outcomes. Furthermore, correcting TAPSE or RVFW-LS for their pressure-load by indexing them to either PVRi or PASP, was non-superior to PVRi or PASP alone in predicting death, heart transplant, lung transplant, or heart/lung transplant. Of the RVAC metrics, variables utilizing RVFW-LS outperformed PVRI, PASP, and TAPSE-based metrics in patients without a history of repaired congenital heart disease.

PVRi, PASP, as well as both hybrid (RVFW-LS/PVRi and TAPSE/PVRi) and non-invasive parameters (RVFW-LS/PASP and TAPSE/PASP) were inversely related to a patient’s WHO functional class, while isolated echo metrics of RVFW-LS or TAPSE were not. These results are similar to a study in children which found that PVRi, PASP, TAPSE/PVRi, and TAPSE/PASP all had more pronounced changes across worsening functional classes compared to TAPSE alone⁶. It would seem that a child’s functional capabilities are less dependent on their absolute right ventricular function and more strongly related to the dynamic interplay between right ventricular function and pulmonary afterload, or at the very least that we do not yet have an ideal measure of right ventricular function.

When analyzing our entire cohort, regardless of PAH etiology, PVRi, PASP, RVFW-LS/PVRi, TAPSE/PVRi, and RVFW-LS/PASP moderately predicted the combined adverse outcome of death or transplant, but one was not necessarily better than the other. In contrast, TAPSE/PASP, and TAPSE and RVFW-LS in isolation were not associated with outcomes in this group. Prior studies in the adult population have reported similar findings, although with somewhat inconsistent use of PVR or PASP as independent variables in their analyses. Ünlü and colleagues found RVFW-LS/PASP to be superior to several other metrics including TAPSE and TAPSE/PASP at predicting need for transplant or death¹¹. However, PASP in isolation was not included in their univariate or multivariate models. In a large study of adult patients, Tello et.al. found TAPSE/PASP was independently associated with mortality in a large study of adult patients, even after including both TAPSE and PASP in multivariate regression analyses²⁰.

A potential source of variability that may explain the disparate findings between our work and those in the adult patients is etiologic heterogeneity of pediatric PAH. With this in mind, we analyzed a subset of patients (n=88) with PAH who did not have a history of repaired CHD. In this cohort, RVFW-LS/PASP performed slightly better than PVRi and PASP alone at predicting the combined outcome, but this did not reach statistical significance. Notably in this cohort, TAPSE-based RVAC metrics, TAPSE, and RVFW-LS in isolation were not significantly associated with the clinical end-point. This is consistent with prior reports that TAPSE is reduced in patients with a history of PAH related to CHD but is often maintained late into the disease course in those with PAH due to other etiologies¹³. This may also reflect the limitations of TAPSE in accurately assessing intrinsic contractility of the RV myocardium as opposed to translational motion of the basal free-wall. RVFW-LS, while not technically a direct measure of cardiac contractility, should provide a more accurate measure of myocardial performance and accounts for segments of the RV free-wall not accounted for by TAPSE²¹. Our results add to a growing body of literature that suggests RVFW-LS-derived coupling metrics may provide more valuable insights into the complex interactions between RV function and pulmonary afterload, but what remains unclear is if markers of afterload alone may have similar prognostic capability.

Ultimately, non-invasive RVAC coupling metrics attempt to better quantify how the RV adapts to the increase in pulmonary vascular afterload to aid in our long-term prognostication for patients. The relative weight of either right ventricular function or pulmonary vascular resistance on prognosis is likely modified by numerous other factors such as the patient’s age, disease etiology, and other medical comorbidities. Prior studies have primarily found that changes in right ventricular function were more strongly associated with adverse outcomes than changes in PVR²². Here, we observed that in children, the PVRi and echocardiographically estimated PASP were more strongly associated with patient outcomes than measures of right ventricular function alone. This is however, at a single point in time. Tracking changes in right ventricular function and RVAC and how they impact long-term clinical outcomes would require a larger, longitudinal cohort and are likely not reflected in this retrospective analysis at a single time-point.

Further underscoring the challenge of accurately assessing right ventricular function in pediatric patients, we found that in isolation, TAPSE (or TAPSE z-score) did not correlate with either invasively measure PVRi or mPAP. Notably, TAPSE was highly variable, particularly in patients with lower pulmonary vascular resistances. This is likely a limiting factor in its reliability to predict afterload in this patient population. RVFW-LS, on the other hand, demonstrated a relatively weak, but significant, positive correlation with these hemodynamic measures, such that patients with higher PVRi, tended to have lower RVFW-LS. While one would not necessarily expect a marker of RV function to “predict” a measure of afterload, these findings could suggest that RVFW-LS is more sensitive to changes in the right ventricular-arterial coupling relationship.

Limitations:

This study is limited by its retrospective nature and only using death, heart transplant, lung transplant, or heart/lung transplant as the clinical outcomes, which are relatively rare events in children. A validation cohort using similar measurements is needed to consolidate the prognostic value of our findings. These echocardiograms were also performed with patients under general anesthesia – albeit simultaneously with invasive hemodynamics to better document associations, which could limit generalizability to an unsedated population. It is also worth noting that both TAPSE and RV global longitudinal strain (RV-GLS) change with age – which is one of the great challenges in working with pediatric patients. Z-scores have been established for TAPSE and RVFW-LS) ^17,19. However, we noted minimal differences in the correlations with invasive hemodynamics, as well as in the stratification across WHO functional classes between the Z-scored and raw data. Furthermore, given that previously published literature on the topic has used uncorrected variables we elected not to use the Z-scores in our RVAC variables. Absolute numeric values for RVAC metrics in children should be interpreted with caution given these age-related differences, and should not be compared to cut-off values for event prediction that have been established in the adult literature. However, changes within individual patients over time or with additional therapies may prove to be the most applicable tool for these metrics, but further studies are needed to understand this longitudinal relationship better Another consideration is that RV-strain measures have been shown to vary by vendor, so this may limit our ability to compare absolute values against one another and is something clinicians should be cognizant of when looking at these metrics for individual patients.

Conclusions:

The relative prognostic significance of pulmonary afterload, right ventricular function, or some combination thereof, remains a challenging clinical question in the field of pulmonary arterial hypertension. We found that baseline PVRi or echocardiographically-estimated PASP were stronger predictors of clinical outcomes in children than estimates of right ventricular function alone, or to RVAC coupling metrics that combine the two variables.

However, our findings do suggest that RVFW-LS/PASP, more so that TAPSE/PASP, may yet be a useful tool in assessing the RV-PA coupling relationship in patients without a history of repaired congenital heart disease. Future studies should focus on the longitudinal assessment of these variables within individual patients as this may provide a more nuanced assessment of how the right ventricle adapts to a high afterload state which ultimately is the critical driver of long-term outcomes.

Clinical Perspective:

Right ventricular arterial coupling (RVAC) can be estimated using the ratio of an echocardiographic measure of right ventricular function (typically TAPSE) to a measure of pulmonary afterload. These metrics have become increasingly utilized in the evaluation of adult and pediatric patients with pulmonary arterial hypertension. However, in pediatric patients, TAPSE is thought to provide a limited evaluation of the actual underlying myocardial contractility. We therefore performed a two-center, retrospective, study to evaluate if echocardiographic derived surrogates of RVAC utilizing right ventricular free-wall longitudinal strain performed better than those utilizing TAPSE at predicting clinical outcomes. Our results reveal that the baseline PVRi and echo-estimated PASP were strongly associated with adverse clinical outcomes, whereas echocardiographic estimates of right ventricular function were not. Of the RVAC metrics – the only one to outperform PVRi or PASP was RVFW-LS/PASP in a subset of patients without a history of repaired congenital heart disease.

Non-Standard Abbreviations and Acronyms

RVAC

Right ventricular-arterial coupling

TAPSE

tricuspid annular plane systolic excursion

RVFW-LS

RV Free-Wall Longitudinal Strain

PAH

pulmonary arterial hypertension

PASP

Pulmonary arterial systolic pressure

PVRi

pulmonary vascular resistance

CHD

congenital heart disease