3D Echo Derived Right Ventricular Principal Surface Strain in Pulmonary Arterial Hypertension

Abstract

Background.

Traditional echocardiographic metrics of right ventricular (RV) function including tricuspid annular plane systolic excursion (TAPSE) and 2-dimensional (2D) strain are limited to the description of longitudinal systolic function. These metrics however fails to account for the complex, 3-dimensional (3D) deformation of the RV.

Methods.

3D echocardiograms (3DE) were obtained simultaneous during clinically indicated right heart catheterization (RHC). We determined the maximum principal surface strain (PS_Max) and angle (ϴ_Max) of RV surface deformation in PAH and control patients.

Results.

We compared 22 control patients to 37 patients with PAH, of whom 11 met hemodynamic criteria for right heart (RH) failure. Compared to 2D descriptors of RV function, PS_Max was significantly different between controls and PAH patients and between PAH patients with and without RH failure. ϴ_Max was oriented progressively longitudinally in PAH patients without RH failure compared PAH patients with RH failure [37.5(34.3 – 40.8)° vs. 34.3(32.1 – 36.2)°, p=0.042] and in PAH patients with worse NYHA functional class. 30-day outcomes were significantly different with an optimal cutoff of PS_Max of −21.4%, with a hazard ratio of 6.8 (95% CI 1.3 to 35.2, p=0.022).

Conclusion.

PS_Max is a robust marker of RH failure and provides prognostic value in PAH beyond conventional 2D descriptors of RH function. Progressive longitudinal deformation of ϴ_Max is associated with worse RH function and functional class.

New & Noteworthy:

Impaired PS_MAX is associated with worse RV systolic function and outcomes in PAH when compared to conventional 2D metrics of RV systolic function, whilst ϴ_Max is progressively longitudinally oriented in PAH patients with RH failure compared to those without RH failure, potentially reflecting an echocardiographic representation of maladaptive RV myofiber reorientation. PS_Max and ϴ_Max represents a powerful and concise way to describe RV systolic function that may prove useful in the care of PAH patients.

Introduction

Although it is well established that function of the right ventricle (RV) represents the single most important prognostic determinant in pulmonary arterial hypertension (PAH)(1), consensus on the best method(s) for describing RV-related outcomes in PAH clinical trials remains elusive(2). Given the advantages of portability, cost, and relative ease of performance, echocardiography is currently the mainstay investigational tool, with tricuspid annular plane excursion (TAPSE) and 2-dimensional (2D) strain commonly used to provide descriptive assessment of RV longitudinal assessment. TAPSE and 2D-strain metrics can be readily measured from standard transthoracic views, and impairments in TAPSE and 2D-strain metrics have both been associated with adverse clinical outcomes in patients with PAH(3),(4),(5) However, it is known that longitudinal assessment by 2D echocardiography incompletely describe RV systolic function(6).(7);(8). Additionally, mechanical strain is fundamentally 3-dimensional (3D) making it difficult to fully quantify the complex shape and structure of the RV using standard geometry (9). This is of particular relevance in pulmonary hypertensive disease processes, where in the context of altered hemodynamics conduction delays, and RV maladaptation, TAPSE and 2D strain assessments can be distorted leading to over- or even underestimation of RV systolic function(9).

In recognition of the limitations inherent to the 2D assessment of RV longitudinal systolic function, other approaches to describe the 3D shape and movement of the RV have been pursued. For example, the Right VentrIcular Separate wall motIon quantification (ReVISION) method was developed to decompose RV motion into longitudinal, radial, and anteroposterior wall displacement and has demonstrated prognostic value in a variety of heart diseases(10, 11) as well as nuanced differences in RV adaptation to surgical repair of the mitral valve(12) and congenital heart defects(13). In order to make the description of RV deformation even more granular but still relatively simple to report and interpret, Kundu P et al recently described an approach based upon established methods for quantifying deformation across the RV free wall in pediatric patients with idiopathic PAH and congenital heart disease associated PAH(14). Using this novel approach, the authors reported a change in strain orientation in their PAH population that likely reflects myocardial fiber reorientation and perhaps an imaging construct of RV remodeling and dysfunction(14). Conceptually, this method entails defining the maximum principal surface strain (PS_Max) and the maximum principal strain angle (ϴ_Max) to describe the magnitude and direction, respectively, of the computationally largest amount of surface deformation. Previous data indicate that ϴ_Max is related to RV myofiber orientation(15) suggesting particular relevance to RV remodeling in PAH. In PAH, unlike that observed in the normal RV, where the myofibers display a heterogenous orientation pattern with its superficial circumferential and deeper longitudinal arrayed fibers(16, 17), the sustained increase in RV afterload in PAH culminates in progressive longitudinal orientation of the RV myofibers(15, 18). In fact, there is evidence in PAH that suggests that this structural change is inefficient and maladaptive, as demonstrated by decreased RVEF(18) and increased myocardial stiffness in the longitudinal direction(15) potentially affecting RV diastolic function. However, whether 3D deformation across the RV surface quantified by PS_Max and ϴ_Max confers clinical and prognostic relevance in the context of PAH relative to current conventional 2D echocardiographic assessment of RV longitudinal systolic function remains unknown.

Accordingly, in the current study, we applied this concept to the free wall of the RV in patients with PAH and compared this metric to other currently used echocardiogram descriptors of RV function and clinical outcomes. we hypothesized that ϴ_Max would be oriented progressively longitudinally in PAH patients with progressive right heart (RH)failure.

Material and Methods

Overall Study Design and Objectives

The current study is a prospective cohort study, designed to assess the clinical and prognostic utility of 3D echocardiography-derived PS_Max and ϴ_Max of the RV in PAH. The primary aims were (a) to compare PS_Max with current, commonly clinically utilized 2D conventional echocardiogram descriptors of RV longitudinal systolic function, (b) to correlate PS_Max with clinical outcomes, and (c) correlate ϴ_Max with severity of disease, stratified as PAH patients with and without right heart (RH) failure. RH failure was defined as elevated right atrial pressure (>15 mmHg) or depressed cardiac index (<2 L/min/m²) in the absence of elevated pulmonary capillary wedge pressure(19, 20). Clinical outcomes were defined similar to prior PAH studies and includes (a) death (all cause), (b) hospitalization from worsening PAH, defined as (i) hospitalization from clinical conditions related to PAH or heart failure, including need for intravenous diuretics, (ii) lung or heart transplant or (iii) atrial septostomy, (c) initiation of prostacyclin therapy, or (d) disease progression, defined as (i) worsening NYHA functional class or (ii) signs and symptoms of heart failure(21, 22).

Study population

We enrolled consecutive patients who underwent clinically indicated diagnostic right heart catheterization (RHC) between November 2020 and February 2023. Our methods for RHC have been previously described(23, 24). Patients were consented for collection of comprehensive echocardiograms coincident to their clinically indicated RHC (Yale University Institutional Review Board: IRB 2000027616 and IRB 2000024783).

The study included patients with World Health Organization Group 1 PAH and control patients with normal 2D echocardiography and supine resting hemodynamics on RHC(25). Control patients represents those who underwent clinically indicated RHC as part of their invasive cardiopulmonary exercise testing for unexplained exertional intolerance and were ultimately found to have a primary peripheral limit to exercise in the setting of dysautonomia (i.e., no central cardiopulmonary limitation to exercise)(26). Patients with left ventricular systolic dysfunction (LVEF < 50%) and patients with either moderate to severe aortic or mitral valve stenosis or regurgitation were excluded from the study.

Echocardiography

Echocardiograms were performed by one of four trained diagnostic cardiac sonographers (F.Z., D.I., S.M., A.R.) using a Philips Epiq (Philips Healthcare, Andover, MA, USA) ultrasound system with dedicated 3D ultrasound probe. Standard 2D echocardiogram parameters were obtained to screen for left ventricular or valvular dysfunction. Standard 2D descriptors of RV function including TAPSE and Fractional Are Change (FAC) were obtained. 2D echocardiographic strain was measured using two widely used metrics from commercially available echocardiography software (TomTec Arena): RV free-wall longitudinal strain (RVFWSL), which measures the change in length of the RV free wall throughout the cardiac cycle as measured in the apical four-chamber view, and RV four-chamber longitudinal strain (RV4CSL), which also accounts for the change in length of the interventricular septum.

3D volumes were obtained for 3–5 cardiac cycles, including end-diastolic volume and end-systolic volume. The TomTec software utilizes 3D speckle and feature tracking at the subendocardial level to generate individual 3D models of the cavity of the RV. These 3D images were stored for offline use. Images were further analyzed on a desktop computer using TomTec 4D RV-Function 2.0 (TomTec Imaging Systems, Unterschleissheim, Germany).

Strain Computation

Mechanical strain is the change in length of any object – or part of it – and is generally expressed as a unitless percentage of the original length. Strain along a surface (here, the endocardial surface of the RV), is expressed as the percent change in the length and width of any segment; these are normal strains. (For the purposes of the methods described in this paper, change in thickness of the wall – radial strain – is ignored as it is mathematically small and is not available from the output of current 3D echocardiography software.) In addition to normal strains, the complete description of a change in shape also requires reporting shear strain. Strains are always reported relative to an axis in space, which must be defined. However, the axis can be conveniently chosen, in the case of the methods of the paper, the chosen axis is the principal axis, for which shear strains are zero. This leaves only the normal strains: RV PS_Max and RV PS_Min, and when paired with their angle of reference, ϴ_Max, fully and simply describe surface deformation. RV PS_Max is defined as the normal strain which represents the computationally largest strain over any surface. RV PS_Min is the smallest strain, which is always perpendicular to RV PS_Max; for this reason, RV PS_Min is not reported in this manuscript. Maximum principal angle (ϴ_Max) was defined as the angle between the local longitudinal direction and the direction of PS_Max; therefore, as this angle approaches 0 degrees, PS_Max is directed longitudinally, and as it approaches 90 degrees, it is directed circumferentially.

The technical methods of computation of RV PS_Max from echocardiogram-derived models is described in detail previously(14). Briefly, the steps for strain computation are: (i) project mesh coordinates from the 3D (global) model into 2D (local) in each element; (ii) compute local strain from the finite element analysis (FEA) strain-displacement relations; (iii) transform local strain back into global coordinates, both in the longitudinal-circumferential directions and as principal strains. Principal strains were found using the standard technique. Verification of the strain computation technique is reported elsewhere(14). The methods to select relevant regions of the model and to calculate strain and segmentation into the septal and free walls are also described in detail elsewhere(14, 27, 28). Briefly, segmentation relies on overall apex-base length along with local surface curvature to delineate the septum-free wall boundaries; the result of such segmentation is embedded in the RV mesh as two scalar coordinates corresponding to a polar angle (similar to latitude) and an azimuthal angle (similar to longitude) for each mesh vertex, which in turn are readily used to label each mesh cell as belonging to a specific region of the RV. Here, we report 3D surface strain values obtained from an area-weighted average of cells on the entire free wall.

The measurements of principal strain (by KH) and RV volumes using 3DE (by HTO) were performed by different investigators. The investigator who performed the principal strain analyses (KH) was blinded to the 3D echocardiogram RV volumes, RVEF, and RHC data. Although repeatability analysis has not been directly measured for 3D surface strain, we followed consensus guidelines on the acquisition and analysis of 3D echo(29), and our 2D metrics – which are also dependent on the generated 3D mesh – have undergone repeatability studies(30–32) demonstrating good-to-excellent intraclass correlation coefficients (ICC).

Statistical Analysis

All analyses were performed in R version 4.2 using the base, tidyverse, survival and cutpointr packages. Variables were checked for normality using visual inspection and Shapiro-Wilks. Comparisons across groups were done using one-way ANOVA for normally distributed and Kruskal-Wallis tests for non-normally distributed variables. Pairwise tests were done using t-tests for normally distributed and Wilcoxon tests for non-normally distributed variables with Holm-Bonferroni correction. Chi-squared tests were used for comparison of binary variables. Association between groups is plotted with linear weighted least squares regression lines and assessed using Pearson’s correlation coefficient (r). For survival analysis, the RV strain threshold that maximized the Youden Index was identified for composite outcomes at 30 days. Kaplan-Meir curves were generated using this cutoff value and the groups compared using the log-rank (Mantel-Cox) test. Significance for all statistical tests was based on an α-level of 0.05.

Results

A total of 37 patients with PAH and 22 control patients with normal resting, supine hemodynamics were prospectively enrolled. 11 PAH patients met the hemodynamic criteria for RH failure. Idiopathic PAH constituted 43% of PAH cases; the 57% remainder were associated with connective tissue disease (scleroderma 35%, systemic lupus erythematosus 5%, mixed connective tissue disease and polymyositis 5% respectively, Sjogren’s syndrome and undifferentiated connective tissue disease 3% respectively). Approximately 56% of patients were on phosphodiesterase type 5 inhibitors and 41% were on endothelin receptor antagonists (41%). Over a third were on inhaled or either continuous intravenous or subcutaneous prostacyclin therapies (37%). A small minority were on riociguat and selexipag (5% and 3%, respectively). The majority were on combination therapy, with 37% of patients on dual therapy and 19% of patients on triple therapy. Patients with PAH were on average older than controls (67±11 vs. 52±15 years, p=0.0002). Female sex, race, and body mass index were similar between the groups [Table 1]. N-terminal Prohormone B-type Natriuretic Peptide (NT-pro-BNP) value was higher in PAH patients [713(50 – 190) vs. 74 (149 – 2150), p=0.002). 7 PAH patients experienced the composite endpoint at 30-day (6 of whom required initiation of prostacyclin therapy and one required in-patient hospitalization for worsening heart failure symptoms).

Table 1:

Clinical characteristics of PAH patients and controls.

	Controls (n=22)	PAH patients (n=37)	p-value
Age	52 ± 15	67 ± 11	0.0002
Female sex	14 (65%)	27 (73%)	0.645
BMI kg/m2	24.1 (22.0 – 30.3)	26.8 (24.8 – 31.9)	0.097
Race			0.297
Non-Hispanic white	12 (55%)	28 (76%)
Non-Hispanic black	5 (23%)	6 (16%)
Hispanic	3 (14%)	2 (5%)
Pacific Islander	1 (5%)	0 (0%)
Prefer not to share	1 (5%)	1 (3%)
NT-pro-BNP	74 (50 – 190)	713 (149 – 2150)	0.002

Data presented as mean ± standard deviation unless otherwise specified. PAH – pulmonary arterial hypertension; BMI – body mass index; NT-pro-BNP - N-terminal Prohormone B-type Natriuretic Peptide.

Patients with PAH had higher mean pulmonary artery pressures (39±9 vs. 18±3 mmHg, p<0.0001) and pulmonary vascular resistances (5.9±2.9 vs. 1.3±0.6 WU, p<0.0001). Right atrial pressures, RV end-diastolic pressures and pulmonary capillary wedge pressures were similar between the groups. Cardiac output and index by thermodilution method were also similar, although stroke volume index by thermodilution method was lower in the patients with PAH than controls (37±9 vs. 45±10 mL, p=0.015) [Table 2].

Table 2:

Invasive hemodynamic and echocardiographic parameters.

	Controls (n=22)	PAH patients (n=37)	p-value
	Right heart catheterization data
RAP (mmHg)	7 (4 – 8)	7 (5 – 12)	0.112
RV EDP (mmHg)	10 (7 – 12)	12 (8 – 14)	0.114
mPAP (mmHg)	18 ± 3	39 ± 9	<0.0001
PAWP (mmHg)	11 ± 3	11 ± 3	0.949
Cardiac output TD, L/min	5.7 ± 1.4	5.2 ± 1.5	0.167
Cardiac index TD mL/m2	3.2 ± 0.6	2.9 ± 0.7	0.102
Stroke volume TD, mL	81 ± 24	69 ± 18	0.045
SV index TD, mL/m2	45 ± 10	37 ± 9	0.015
PVR, Woods Unit	1.3 ± 0.6	5.9 ± 2.9	<0.0001
MvO2, %	72 (69 – 74)	66 (54 – 69)	0.0003
	2D echocardiogram data
LV ejection fraction, %	58 ± 6	61 ± 6	0.09
Estimated RVSP, mmHg	29 ± 6	49 ± 18	0.0012
Peak TR velocity, m/s	2.5 (2.3 – 2.6)	2.9 (2.7 – 3.4)	0.0017
TAPSE, cm	2.4 (2.1 – 2.6)	1.8 (1.5 – 2.1)	0.011
TAPSE/RVSP (cm/mmHg)	0.08 (0.06 – 0.09)	0.03 (0.02 – 0.05)	0.001
RVFWSL, %	−22.3 (−25.3 – −18.5)	−15 (−17.2 – −12.6)	2.4e−5
RV4CSL, %	18.8 (−22.7 – −15.8)	−13.7 (−14.9 – −12)	0.0002
	3D echocardiogram data
RV EF, %	48.7 (42 – 52)	38 (33 – 42)	0.0004
PSMax strain, %	−32.3 (−37.3 – −28.8)	−23.6 (−28.5 – −21.0)	6.2e−7
PSMin strain, %	−10.0 (−11.6 – −8.9)	−8.1 (−9.8 – −5.9)	0.007
ϴMax	39.8 (38.1 – 42.2)	36.7 (34.2 – 40.4)	0.003

Data presented as mean ± standard deviation unless otherwise specified. PAH – pulmonary arterial hypertension; RAP – right atrial pressure; RV – right ventricular; EDP – end diastolic pressure; mPAP – mean pulmonary artery pressure; PAWP – PA wedge pressure; TD – thermodilution; PVR – pulmonary vascular resistance; MvO₂ – mixed venous oxygen; LV – left ventricular; RVSP – RV systolic pressure; TR – tricuspid regurgitation; TAPSE – tricuspid annulus systolic plane excursion; RVFWSL – RV free wall strain; RV4CSL – RV 4 chamber longitudinal strain; EF – ejection fraction; PS_Max – Maximum principal surface strain; PS_Min – Minimum principal surface strain; ϴMax – maximum principal angle.

As shown in Table 2, 2D echocardiographic parameters showed significant differences between PAH and control groups. TAPSE and TASPE/RVSP were significantly impaired in patients with PAH [1.8 (1.5 2.1) vs. 2.4 (2.1 – 2.6) cm, p=0.011)] [0.03 (0.02 – 0.05) vs. 0.08 (0.06 – 0.09)], respectively. 2D strain metrics were more impaired (i.e. less negative) than controls when measured both as RVFWSL [−15.0 (−17.2 - −12.6) vs. −22.3 (−25.3 - −18.5)]%, p=2.4e⁻⁵) and RV4CSL [−13.7(−14.9 – −12) vs. −18.8 (−22.7 - −15.8)%, p=0.0002)]. Both PS_Max and PS_Min were impaired (less negative) in PAH patients compared to controls [−23.6(−28.5 - −21.0) vs. −32.2 (−37.3 - −28.8)%, p = 6.2e-7] and [−8.1(−9.8 - −5.9) vs. −10.0 (−11.6 - −98.9)%, p = 0.007, respectively)]. Additionally, RVEF was lower in patients with PAH [(38.0 (33 – 42) vs. 48.7 (42 – 52)%, p=0.0004)] [Table 2]. There was a significant correlation between PS_Max and RVEF (r=0.95). There was moderate correlation between RV ejection fraction and TAPSE (r=0.55), TAPSE/RVSP (r=0.67), RVFWSL (r=0.66) and RV4CSL (r=0.67).

2D and 3D echocardiographic metrics of RV function were compared between 3 groups: 1) PAH patients with RH failure, 2) PAH patients without RH failure, and 3) controls. PS_Max was different between the 3 groups [RH failure: −21.0 (−28.6 - −21.9)% vs. no RH failure: −25.9 (−28.0 - −20.9)%vs. control: −32.2(−37.4 - −28.4)%, p=0.032] between RH failure and no RH failure, p= 6.4e⁻⁵ between no RH failure and control, and p=2.1e⁻⁷ between RH failure and control). All 2D metrics of RV function showed differences between control patients and PAH patients with RH failure. The 2D strain metrics (RVFWSL and RV4CSL) and TAPSE/RVSP, but not TAPSE, showed significant differences between control and PAH patients without RH failure. No 2D echocardiographic metrics, including TAPSE/RVSP showed significant differences between PAH patients with and without RH failure. RVEF was significantly different between the 3 groups controls 51.8 (46.1– 53.6); PH without RH failure 39.4% (33.3 – 45.2%), PH with RH failure (25.4–32.4) with respective p-values 0.0001 (control vs. no RV failure), 0.0003 (control vs. RH failure), 0.002 (RH failure vs PH with RH failure).

ϴ_Max was lower (more longitudinal) in RH failure [Figures 1 and 2]. ϴ_Max was more longitudinally oriented in PAH patients without RH failure compared to PAH patients with RH failure patients [37.5(34.3 – 40.8)° vs. 34.3(32.1 – 36.2)°, p=0.042]. ϴ_Max was significantly different between control and RH failure patients [39.8 (38.0 – 42.3)° vs. 34.3(32.1 – 36.2)°, p=0.0004] but not significantly different between controls and PAH patients without RH failure. ϴ_Max was also more longitudinally oriented in patients with NYHA functional class 3–4 symptoms when compared patients with NYHA functional class 1–2 symptoms [35.6(32.9 – 39.3)° vs. 39.5 (36.8 – 41.7)°, p=0.0007].

Figure 1:

Typical magnitude (PS_MAX, vector length and color) and directionality (ϴ_MAX, surface colormap and vector direction) of maximal principal strains of right ventricle (RV) free wall on (A) a control patient with normal resting right heart hemodynamics. All three patients demonstrate more longitudinally directed strains on either side of the RV free wall which converge into a more circumferential direction at the middle of the RV free wall, and (C) scleroderma associated pulmonary arterial hypertension (PAH) patient with right heart failure, (B) idiopathic PAH patient without right heart failure, The patient in (C) with right heart failure exhibits the lowest average ϴ_MAX compared to patient (B) and (A), with strains on each side of the RV free wall being nearly longitudinal (ϴ_MAX -> 0). The idiopathic PAH patient in (B) had a higher average ϴ_MAX compared to (C), while having similar overall strain magnitudes. The control patient in (A) demonstrated strain directions on each side of the RV free wall that were approximately 45° to vertical (ie., highest ϴ_MAX within the group).

Figure 2:

A: Comparison of ϴmax between PAH patients with and without right heart failure and controls. B: Comparison of ϴmax between PAH patients with NYHA class 1–2 vs. 3–4 symptoms. Deg: degree, NYHA: New York Heart Association Functional Class. Data presented as median and interquartile range.C: Kaplan-Meir curves showing percentage of patients without clinical event within 30 days of obtaining echocardiogram.

Area under the receiver operating curve for PS_Max as a predictor of composite outcomes was 0.738. Youden index identified −21.4% strain as the optimal cutoff for PS_Max with a sensitivity of 71%, specificity of 83% and accuracy of 81%. Estimated Cox proportional hazard ratio for the composite outcome for patients with PS_Max below this cutoff was 6.8 (95% CI 1.3 to 35.2, p=0.022). 30-day outcomes were significantly different between PAH patients above and below this strain value (p=0.02) (figure 2c). There was no significant difference in median RVEF for patients who experienced a 30-day clinical outcome [(32.4 (26.7 – 33.4) %] vs. those did not [(36.7 (31.3 – 44.4) %, p=0.21] in the current cohort.

Discussion

The current study demonstrates that PS_Max is a powerful descriptor of RV function and outcomes when compared to conventional TAPSE and 2D strain metrics in this cohort of PAH patients. PS_Max was significantly different between controls and PAH groups with all severities of RH failure. However, TAPSE – the most utilized descriptor of RV systolic function in clinical settings – was only significantly different between the extremes of controls and RH failure patients, while TAPSE/RVSP - a commonly used echocardiographic surrogate for RV-PA coupling did not distinguish between PAH patients with and without RH failure. In addition, the current study also demonstrated that PS_Max was more closely correlated to RVEF than TAPSE, TAPSE/RVSP, or 2D metrics of RV strain. While RVEF is a well-established descriptor of RV function(2, 33–35) and in current study differentiated between the 3 groups, PS_Max has the added benefit of providing directionality of deformation via ϴ_Max. Prior research has established that RV myofibers reorient longitudinally in PAH in response to increased afterload(15, 18). Our data corroborate this by demonstrating that ϴ_Max progressively orients longitudinally in PAH patients vs. controls and in PAH patients with vs. those without RH failure, consistent with previous studies(36). This longitudinal orientation, as demonstrated by ϴ_Max, may reflect an echocardiography representation of RV myofiber reorientation. Alteration in RV contractile pattern as a function of progressive RV dysfunction and hemodynamic parameters was also reported in a recent study by Rako et al(37). Using the ReVISION method, the authors demonstrated progressive alteration in RV contractile pattern as suggested by an initial reduction in longitudinal RVEF in patients with no or mild RV–PA uncoupling to a reduction in anteroposterior RVEF in patients with more advanced RV–PA uncoupling with accompanying worsening hemodynamic parameters. Results of this study by Rako et al however needs to be interpreted within the context of its limitations, including their small sample size, particularly amongst the advanced RV-PA uncoupling group and the heterogenous PH population studied, both of which limits the generalizability of their findings to specific PH group (37)

PS_Max and ϴ_Max were calculated solely from echocardiography data using a 3D ultrasound probe and software that are commercially available. Although cardiac MRI remains the gold standard for detailed assessment of the RV, echocardiography is cheaper, quicker, and easier for patients to tolerate. It is also portable, allowing its use in a wide variety of research and clinical settings(38). Unlike cardiac MRI, 3DE can be performed concurrently with right heart catheterization(39) as we demonstrated in the current study. In fact, while the absolute volumes derived from 3DE may underestimate cardiac MRI derived RV volumes, the RVEF derived from 3DE is comparable to that of the reference standard cardiac MRI(40). Taken together, 3DE represents a comparable alternative to cardiac MRI, particularly at the bedside and in the critical care setting(38, 39).

PS_Max confers several advantages over current conventional echocardiographic measures of RV systolic function. First, it accounts for all planes of 3D deformation of the RV as opposed to most metrics which are confined exclusively to the longitudinal plane. Second, the concise output, given as a single magnitude PS_Max and angle ϴ_Max, is simple to report and interpret. Third, ϴ_Max may be a surrogate for myofiber orientation, a known pathophysiologic mechanism of RH failure with potential implications on clinical outcomes and pharmacotherapeutic development.

Limitations

Study results need to be interpreted in the context of several limitations. First, although the described data were collected prospectively from consecutively enrolled patients with PAH, a larger, separate cohort is required to validate this method. Second, our PAH cohort consisted of patients who were already on established PAH-targeted pharmacotherapy. Whether treatment PAH specific therapy would have affected the PS_Max and ϴ_Max measurements is uncertain. Third, one third of the PAH patients had underlying scleroderma, which is known to be associated with an intrinsic cardiomyopathy independent of afterload. This may have affected the principal strain results and will be an important area of future research. Nonetheless, in the current study, our sub-analysis showed similar PS_Max (−24.6% vs. −23.5%, p=0.558) and ϴ_Max (35.7° vs. 37.8°, p=0.133) values between PAH patients with and without scleroderma. We finally note that while ICC is generally lower for regional measurements obtained from 3D echocardiogram, we believe the 3D strain used herein averaged from the entire free wall represents a more global measure of strain.

Conclusion

Impaired PS_Max is associated with worse RV systolic function and worse short-term clinical outcomes in PAH. Additionally, PS_Max is highly correlated with RVEF and has the additional advantage of providing an angle of directionality in 3D space (ϴ_Max). ϴ_Max in turn, progressively orients longitudinally as RH function and functional class worsens in PAH. Future studies incorporating larger number of patients across the different spectrums of pulmonary hypertension phenotypes is warranted to validate the current findings and examine the role PS_Max and ϴ_Max in PH disease progression and their respective roles predicting response to PAH pharmacotherapy.