Systemic sclerosis (SSC; scleroderma)–associated pulmonary arterial hypertension (PAH) is a major cause of mortality in SSc patients and represents an important diagnostic and therapeutic target. Our aims were to evaluate the relationship between echocardiogram-derived right-sided heart hemodynamics and gold standard right-sided heart catheterization (RHC) measurements in a scleroderma population and to investigate whether this relationship is modified by a subset of pulmonary hypertension.
We performed RHC and echocardiography on the same day, with pulmonary function testing in 21 consecutive subjects with scleroderma and precapillary pulmonary hypertension (mean ± SD age 57 ± 10 years, 81% women).
RHC measures, including pulmonary arterial systolic and mean pressure and pulmonary vascular resistance (PVR), correlated strongly with echocardiogram-derived data. RHC-derived PVR was negatively associated with right ventricular (RV) systolic performance, as measured by tricuspid annular plane systolic excursion (TAPSE; rho = −0.70, P < 0.001), tissue Doppler tricuspid s′ velocity (rho = −0.68, P = 0.002), and RV fractional area change (rho = −0.78, P < 0.001). Correlations with TAPSE and s′ velocity were strengthened when forced vital capacity %/diffusing capacity of the lung for carbon monoxide % ≥1.6 was used to identify pure PAH phenotypes in SSc. Bland-Altman analyses demonstrated strong agreement between RHC and echocardiogram-derived hemodynamic measures.
Our findings suggest that echocardiography may play a clinical role in identifying pulmonary hypertension and RV dysfunction noninvasively, particularly in a subset of SSc patients stratified by pulmonary function testing. This method may establish specific disease phenotypes with differential cardiovascular impact and prove useful as a marker of disease progression/risk stratification in SSC patients that warrants further investigation in larger cohorts.
Systemic sclerosis (SSc; scleroderma) is a systemic inflam matory connective tissue disorder associated with autoimmune activation and collagen overproduction that progressively leads to fibrosis of the skin, joints, and internal organs with a predilection for pulmonary, gastrointestinal, and renal involvement (). Specifically, scleroderma-associated pulmonary arterial hypertension (PAH) is a leading cause of mortality in this population and represents an important therapeutic target in clinical studies ([2, 3]). Precapillary pulmonary hypertension in scleroderma is characterized as either isolated PAH or pulmonary hypertension secondary to interstitial lung disease (ILD).
Right-sided heart catheterization (RHC) is the gold standard for the assessment of cardiopulmonary hemodynamics; however, its invasive nature limits its use in large population studies or longitudinal clinical trials that require repeat procedures ([4, 5]). Echocardiography provides a noninvasive tool not only for evaluating right-sided heart structure, such as chamber size and function, but also for comprehensive assessment of hemodynamics (). As such, transthoracic Doppler echocardiography is currently the recommended initial screening modality for the evaluation of PAH (). While the clinical role of echocardiography in SSc-associated PAH continues to be an area of significant interest, few studies have directly simultaneously compared echocardiography to RHC in this population ([5, 8, 9]).
The American College of Cardiology Foundation (ACCF)/American Heart Association (AHA) Expert Consensus Document on Pulmonary Hypertension identifies pulmonary vascular resistance (PVR) as a key measure integral to the evaluation and diagnosis of precapillary PAH (). Both right-sided heart hemodynamics and right ventricular (RV) function predict all-cause mortality in pulmonary hypertension ([5, 10, 11]). Noninvasive assessment of intracardiac pressures in patients with SSc-associated PAH, particularly in relation to right-sided heart functional measures, including RV tissue Doppler, fractional area change (FAC), and tricuspid annular plane systolic excursion (TAPSE), have not been well investigated. Therefore, the aims of this study were to 1) evaluate the relationship of echocardiogram-derived right-sided heart hemodynamics to invasive RHC measurements in a scleroderma population, and 2) investigate whether the association is modified by the etiology of pulmonary hypertension (with or without significant ILD).
Box 1. Significance & Innovations
The present study evaluated the assessment of pulmonary hypertension in patients with systemic sclerosis using noninvasive techniques with transthoracic echocardiography, including tricuspid annular plane systolic excursion, tricuspid annular systolic (s′) velocity, and right ventricular fractional area change. These correlated significantly with multiple hemodynamic measurements obtained via right-sided heart catheterization.
The relationship between echocardiographic parameters and right-sided heart catheterization measurements was stratified by pulmonary function test parameters, specifically forced vital capacity %/diffusing capacity of the lung for carbon monoxide % (FVC%/DLco%), as an approach used to identify patients with isolated pulmonary arterial hypertension versus pulmonary hypertension secondary to interstitial lung disease that strengthened correlations.
A method to estimate pulmonary vascular resistance by echocardiography correlated robustly with assessments of pulmonary vascular resistance by right-sided heart catheterization, particularly when stratified by FVC%/DLco% ≥1.6. This provides a potential noninvasive approach to following pulmonary vascular resistance in patients with systemic sclerosis at risk for or with pulmonary hypertension to potentially guide management.
PATIENTS AND METHODS
The Early, Simple and Reliable Detection of Arterial Hypertension in Systemic Sclerosis (DETECT) Study was a prospective, observational, cross-sectional cohort study in scleroderma patients evaluating screening tests and the incidence of pulmonary hypertension. The study involved 62 centers in 18 countries, with longitudinal data collection over 3 years. This study included 21 patients enrolled at Boston University Medical Center. The Institutional Review Board at Boston University Medical Center approved the research protocol, and written informed consent was obtained from all participants prior to enrollment.
The inclusion criteria included age ≥18 years with a definite diagnosis of SSc by the American College of Rheumatology (ACR) criteria (), including all patients with any other connective tissue diseases who in parallel also met the ACR criteria for SSc; disease duration >3 years dated from onset of the first non–Raynaud's phenomenon feature; and diffusing capacity of the lung for carbon monoxide % (DLco%) <60% of predicted. Excluded from enrollment were subjects with a prior history of known pulmonary hypertension with mean pulmonary arterial pressure (mPAP) ≥25 mm Hg at rest or ≥30 mm Hg during exercise independent of pulmonary capillary wedge pressure (PCWP) prior to enrollment, subjects who had an RHC within the past 12 months, subjects receiving therapy considered definite PAH/pulmonary hypertension treatment for any indication within 6 weeks of enrollment and/or for a total of >6 weeks during the previous 12 months, subjects with forced vital capacity (FVC) <40%, subjects with an estimated glomerular filtration rate <40 ml/minute/1.73 m2, subjects with known prior PWCP >15 mm Hg, and subjects with previous evidence or a diagnosis of clinically relevant left-sided heart disease (including left ventricular ejection fraction [LVEF] <50%, significant diastolic dysfunction, significant valvulopathy, known significant coronary disease, uncontrolled blood pressure, hypertrophic cardiomyopathy, decompensated congestive heart failure, congenital heart disease, prior cardiac surgery, and pregnancy).
Two-dimensional transthoracic echocardiograms were performed using commercially available ultrasound machines (IE33, Phillips Medical Systems). All echocardiograms were performed in our Intersocietal Accreditation Commission in Echocardiography laboratory, in accordance with the American Society of Echocardiography (ASE) guidelines (), and by the same sonographer (BD) to minimize technical variability. No patients in this study were noted to have atrial fibrillation/arrhythmias. Left atrial (LA) diameter was determined from parasternal long-axis view and right atrial (RA) area and the RV were measured utilizing the 4-chamber apical view. LA volume was quantified using the area-length method, with measurements taken in the apical 4-chamber and 2-chamber views at ventricular end systole. LVEF was measured by the modified Simpson's method. RV FAC was determined using measurements of end-diastolic and end-systolic areas and applying the following formula: FAC% = 100 × end-diastolic area − end-systolic area/end-diastolic area (). The Doppler measurements included continuous wave Doppler through the tricuspid valve, utilizing the highest velocity obtained from multiple views to determine the peak tricuspid regurgitant velocity (TRV; meters/second). The modified Bernoulli equation was used for calculation of the RV systolic pressure, which, in the absence of RV outflow tract obstruction, determined pulmonary arterial systolic pressure (PASP). RA pressure was determined by inferior vena cava size and collapsibility and was assigned a value of 3, 8, or 15 mm Hg according to the ASE guidelines (). Pulsed-wave Doppler to assess peak E (early diastolic) and A (late diastolic) velocities was obtained using standard methodology, as previously described (). Tissue Doppler early (e′) and late (a′) diastolic velocities and systolic velocity (s′) were obtained on both mitral and tricuspid annular planes, respectively. Tissue Doppler measures the intrinsic myocardial velocities in a longitudinal fashion; thus, limitations can occur due to beam angle when it is not parallel to myocardial motion. The septal annulus was utilized for mitral measurements, and tricuspid measurements were obtained at the junction of the RV free wall and anterior leaflet. Calculation of the RV myocardial performance index, a global measurement of both systolic and diastolic function of the RV, was performed using tissue Doppler and defined as the ratio of tricuspid valve closure-opening time − ejection time/ejection time (). TAPSE was obtained by placing an M-mode cursor across the tricuspid annulus in a 4-chamber apical view and capturing the maximum annular movement during peak systole. The RV outflow tract velocity-time integral (VTIRVOT; cm) was obtained by placing a 1–2-mm pulsed wave Doppler sample volume in the proximal RV outflow tract just below the pulmonary valve in the parasternal short-axis view. PVR by echocardiography (PVRecho) was calculated by a previously validated method using the formula PVRecho = 0.1618 + 10.006 × TRV/VTIRVOT ().
RHC was performed using a standard protocol and no sedation was administered for the procedure. Except for 1 participant, all subjects underwent RHC and echocardiography on the same day within 1 hour. All catheterization procedures were performed by the same operator who was blinded to real-time echocardiographic data (HWF). Cardiac output was measured using the Fick method. The following cardiopulmonary measurements were recorded: RA pressure, RV pressure, PASP, mPAP, pulmonary artery diastolic pressure, and PCWP. PVR by catheterization (PVRcath) was calculated by the following equation: PVRcath = mPAP − PCWP/cardiac output ().
Pulmonary function evaluation
All patients underwent standard pulmonary function testing as entry into the study requiring total lung capacity (TLC) ≥40%. Measurements (reported as % predicted) used for the analyses included: forced expiratory volume in 1 second (FEV1), FVC, TLC, and DLco.
In addition, all patients underwent a high-resolution computed tomography scan for ILD evaluation. Patients were classified as having ILD based on the criteria used in the main DETECT Study, which was defined as PCWP ≤15 mm Hg and FVC <60%, or FVC 60–70% plus high-resolution computed tomography data not available or determined moderate–severe as assessed by radiology staff blinded to any echocardiogram or invasive hemodynamic data ().
Descriptive statistics are shown as the mean ± SD for continuous variables and the percentage for categorical variables. Relationships between echocardiographic and RHC variables were assessed by Spearman's rank correlation coefficient and linear regression. Agreement between echocardiographic and RHC parameters was analyzed using Bland-Altman plots. Analyses were further stratified by pulmonary function testing cut point ratios of FVC%/DLco% ≥1.6 or <1.6, previously reported to discriminate patients with PAH versus pulmonary hypertension secondary to ILD () and risk stratify in scleroderma ([19, 20]). FVC%/DLco% stratification was evaluated in all correlations; however, only statistically significant analyses were shown. A 2-sided P value less than 0.05 was accepted as statistically significant. Analyses and graphics were performed with SAS, version 9.2.
As shown in Table 1, subjects were mean ± SD ages 57 ± 10 years and were predominantly white women, consistent with the expected demographics of the population with scleroderma. All participants had normal renal function and none had diabetes mellitus. FVC and FEV1 were both slightly decreased with a normal ratio (mean ± SD 75.1 ± 19.5 and 75.4 ± 18.9, respectively). TLC was decreased and DLco was severely reduced at 42.4%. Five patients (23.8%) were classified as having ILD by computed tomography scan and the DETECT criteria, as previously described. The majority of patients had pulmonary hypertension and/or pulmonary hypertension secondary to ILD; however, 7 patients did not have a diagnosis of PAH based on the criterion of mPAP ≥25 mm Hg.
Values are the mean ± SD unless indicated otherwise. FEV1 = forced expiratory volume in 1 second; FVC = forced vital capacity.
57.1 ± 9.5
Women, no. (%)
White, no. (%)
Past or current smoker, no. (%)
Hypertension, no. (%)
Diabetes mellitus, no. (%)
Coronary artery disease, no. (%)
Body mass index, kg/m2
28.8 ± 4.4
B-type natriuretic peptide, pg/ml
254.9 ± 500.0
1.1 ± 0.6
Interstitial lung disease, no. (%)
FEV1, % predicted
75.4 ± 18.9
FVC, % predicted
75.1 ± 19.5
1.0 ± 0.09
Total lung capacity, % predicted
73.2 ± 20.3
Diffusion capacity, % predicted
42.4 ± 11.3
FVC/diffusion capacity ratio
1.9 ± 0.6
Total lung capacity/diffusion capacity ratio
1.8 ± 0.7
Cardiac catheterization and echocardiographic parameters
As shown in Table 2, the mean ± SD RA pressure for the cohort was 8.2 ± 5.8 mm Hg. PCWP was uniformly within normal range (mean ± SD 9.6 ± 3.9 mm Hg) and cardiac output was preserved. RV systolic pressure and PASP were elevated without a gradient step up, thus excluding any RV outflow or pulmonic valve obstruction. Mean ± SD PVR by RHC was >2 Wood units in half of the study population (4.5 ± 3.7 Wood units). Echocardiographic parameters, including tissue Doppler and TAPSE data, are shown in Table 2. Two subjects had undetectable tricuspid regurgitant jet profiles and were excluded from the analyses. All participants exhibited LVEF ≥50%. The mean ± SD PASP by echocardiographic measurement was elevated (47 ± 21 mm Hg) and was essentially identical to values obtained by invasive measures. Mean basal RV systolic performance, as measured by tricuspid s′ velocity (mean ± SD 12.1 ± 3.4 cm/second) and TAPSE (mean ± SD 22.7 ± 5.4 mm), was within the normal range, based on ASE guidelines ().
Table 2. Hemodynamic and echocardiographic parameters (n = 21)*
Mean ± SD
RA = right atrial; RV = right ventricle; WU = Wood units; LA = left atrial; BSA = body surface area; MPI = myocardial performance index; VTI = velocity-time integral.
Correlation of catheterization and echocardiographic hemodynamics
As shown in Table 3, several echocardiographic measures correlated significantly with invasive hemodynamics, with the strongest correlation for PASP (rho = 0.92, P < 0.001). The relationship between echocardiogram- and RHC-derived PASP is shown in Figures 1A and B. Several newer right-sided heart echocardiographic variables were also closely associated with RHC measurements. TAPSE, a measure of systolic descent of the base of the RV free wall, correlated significantly with PASP, mPAP, and PVR. The association was particularly strong for PVR (rho = −0.90, P < 0.001) when analyses were performed in the subset of individuals with predominant PAH, as stratified by FVC%/DLco% ≥1.6 (Figure 2). An additional method of RV systolic function quantification using tricuspid s′ velocity also correlated significantly with invasive hemodynamic measures of PASP, mPAP, and PVR (Table 3). Again, both PASP and PVR associations with tricuspid s′ velocity strengthened when stratified by FVC%/DLco% ratio ≥1.6. RV end-diastolic area was only modestly related to PVRcath (rho = 0.46, P = 0.04); however, when stratified by FVC%/DLco% ≥1.6, the correlation strengthened (rho = 0.77, P = 0.004).
Table 3. Correlations between echocardiography and right-sided heart catheterization parameters (n = 21)*
Catheterization and echocardiography-derived measurements of PVR correlated significantly (rho = 0.75, P < 0.001). Consistent with other analyses, when stratified by FVC%/DLco% ≥1.6, this association markedly strengthened (rho = 0.87, P = 0.003), as shown in Figure 3A. Bland-Altman analysis for echocardiogram- and catheterization-derived PVR showed good agreement, as shown in Figure 3B. Additionally, as shown in Table 3, we demonstrated an inverse relationship between RV FAC and PVRcath (rho = −0.78, P < 0.001).
In the present study, we carefully examined cardiopulmonary hemodynamics concurrently by cardiac catheterization and echocardiography in a scleroderma population with varying degrees of pulmonary vascular disease and demonstrated that several right-sided heart measures, including PASP, mPAP, and PVR, strongly correlated with each other using both methodologies. Additionally, we showed that novel measures of right-sided heart systolic function, specifically TAPSE, tricuspid s′ velocity, and RV FAC related well to invasively measured PASP and PVR. Many of these relationships were stronger in the subset of individuals with a phenotype more suspect of PAH, as defined by FVC%/DLco% ≥1.6.
Scleroderma-associated PAH is a leading cause of scleroderma-related deaths (), and elevated pulmonary pressures, particularly increased PVR, independently predict mortality in this population ([2, 3, 21, 22]). Hemodynamic data, including PASP, mPAP, and PVR usually obtained via RHC, are recommended for the definitive diagnosis and monitoring by the ACCF/AHA Expert Consensus Document on Pulmonary Hypertension (). Recent published data from the DETECT Study place echocardiographic measures as an important clinical component of the evidence-based algorithm for the detection of PAH (). In the present study, we provided evidence that clinically relevant right-sided heart pressure measurements can be reasonably well estimated noninvasively using echocardiography. While previous studies have reported modest relationships in RHC–echocardiogram correlations in SSc populations ([5, 8, 9]), the primary reasons for variability in findings may relate to technical issues, such as temporal delays, sometimes days, between the acquisition of Doppler and invasive measurements. In addition, inaccurate echocardiogram estimation and reliance on assumptions of RA pressure and/or utilization of technically poor-quality TR Doppler jet profiles that underestimate effective regurgitant velocities may also have contributed to decreased accuracy. As PASP remains one of the most used and emphasized parameters in clinical practice, we demonstrated that carefully performed echocardiograms can yield hemodynamic information comparable to invasively derived values. In particular, we found striking concordance in PASP determination by echocardiogram and catheterization. Our echocardiogram calculation of PVR was based on a derived formula from prior published data in 44 subjects of mixed etiology of pulmonary hypertension (). While in our study PVRecho correlated significantly with PVRcath, accuracy was less than that observed for PASP estimation. Frea and colleagues recently examined Doppler parameters in scleroderma patients without pulmonary hypertension and identified TRV/VTIRVOT ratio, a surrogate of PVR, as one of the strongest predictors for incident PAH in their cohort (). As such, despite some limitations, there is growing recognition that PVRecho provides clinical insight, and further studies are warranted to refine the accuracy of this noninvasive variable, specifically in patients with SSc.
Another novel aspect of our study involved the investigation of newer echocardiogram parameters of RV function in relation to RHC-derived hemodynamics. TAPSE is a simple method that quantifies the distance of RV systolic annular apical displacement using M-mode and represents longitudinal RV systolic function. TAPSE is validated against RV ejection fraction and fractional area shortening, and is recommended as a routine measure for the assessment of RV function (). When evaluated in various PAH cohorts, impaired TAPSE has been linked to reduced survival ([10, 11]). A recent study specifically examined this relationship in SSc-associated PAH patients, demonstrating a negative association between RHC PVR and TAPSE, consistent with our data, and TAPSE ≤1.7 cm conferred a 4-fold greater risk of death compared to patients with TAPSE >1.7 cm (). In our study, we demonstrated significant negative correlations between TAPSE and RHC measurements of PASP, mPAP, and PVR. The relationships were particularly striking in subjects with predominant PAH stratified by FVC%/DLco%, where the PVR–TAPSE correlation reached rho = −0.9 (P < 0.001). As such, TAPSE not only provided prognostic information, but also tracked closely with right-sided heart hemodynamic overload, and may prove useful as a noninvasive marker of disease progression and risk stratification in patients with SSc-associated PAH.
In this context, we examined another newer echocardiogram method that quantifies RV systolic performance, tricuspid annular systolic (s′) velocity, which has evolved as a simple and reproducible technique that uses pulsed-wave tissue Doppler to measure peak longitudinal velocity excursion of the tricuspid annular and/or basal RV free wall. This measure of myocardial tissue velocity has been validated for the assessment of RV systolic function in patients with pulmonary hypertension and correlates well with TAPSE. In SSc patients without pulmonary hypertension, tricuspid annular s′ velocity was lower compared with age- and sex-matched controls and predicted RV dysfunction (). In our study, tricuspid s′ velocity correlated significantly with PASP, mPAP, and PVR, suggesting that this parameter along with TAPSE provides information about the extent of RV dysfunction in pulmonary hypertension. A limitation of tricuspid annular Doppler relates to its beam angle dependency and that measurements must be acquired parallel to annular motion for accuracy. Last, we found RV FAC, another measure of RV systolic function, negatively correlated with RHC assessment. One small study of scleroderma in the literature evaluated the prognostic value of FAC and found that it did not predict PAH development ().
Scleroderma comprises 2 primary pulmonary vascular disease phenotypes, ILD with subsequent development of PAH or isolated PAH without significant ILD (more frequently associated with limited cutaneous SSc). Recent data have suggested that scleroderma with pulmonary hypertension resulting from ILD is associated with greater mortality risk, as compared to individuals with isolated PAH ([21, 26]). Disease pathophysiology in scleroderma may vary depending on the etiology of pulmonary hypertension, with differential impact on the pulmonary vasculature and the right side of the heart with concomitant pulmonary parenchymal involvement. In our study, correlative findings strengthened when specifically examined in subjects with predominant PAH based on FVC%/DLco% ≥1.6, a novel observation that has not been previously reported in echocardiographic or catheterization studies in scleroderma. Both FVC and DLco% decrease to a similar extent in ILD; however, in SSc-associated PAH, DLco% has been found to be disproportionately reduced compared with FVC%, thus leading to higher FVC%/ DLco% ratios ([3, 18]). DLco% has been shown in several studies to predict PAH in scleroderma; therefore, stratification by FVC%/DLco% likely increases specificity for disease subsets that may have distinct pathophysiologies. While our subset sample size was small, the correlations were robust and warrant further investigation in larger clinical investigations to validate and gain further understanding of the impact of scleroderma subtypes on cardiopulmonary function.
The present study has several limitations. First, the investigation involved a relatively small number of subjects in a single-center tertiary referral hospital. The small number of participants limited our ability to perform multivariate analysis and rigorous statistical modeling with adjustment for multiple comparisons; therefore, our findings are in part exploratory and hypothesis generating. In addition, subjects in our investigation were part of the DETECT Study and, owing to the inclusion criteria, may not be representative of the broader disease spectrum of SSc. However, a small sample size may be expected because scleroderma is a rare disease. We nevertheless demonstrated strong correlations between RHC- and echocardiogram-measured variables in this group of individuals, suggesting biologic connection of the described relationships. Second, we did not assess RV strain as an additional evolving measure of RV systolic function because it was beyond the technical capability of the image acquisition system. Third, pulmonary hypertension severity was generally in the moderate range, and further studies would be required to validate findings in more advanced disease. The findings are counterbalanced by the careful nature of our investigation, including single-blinded operators for both the RHC and echocardiogram procedures.
In conclusion, our findings suggest that echocardiography may be clinically useful in identifying and monitoring pulmonary hypertension and RV function noninvasively in patients with scleroderma. The utilization of echocardiography can aid in the screening and monitoring of PAH for progression and possibly for treatment effects in these patients, but at this time it would not replace RHC for the definitive diagnosis of PAH, which remains the gold standard for a precise and accurate assessment of pulmonary hemodynamics. Stratification by FVC%/DLco% strengthened our findings, prompting recognition that the cardiopulmonary impact of specific scleroderma disease subtypes warrants further investigation, as does the specific role of echocardiography in the screening and monitoring of scleroderma patients with and without pulmonary hypertension.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Gopal had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Gopal, Doldt, Finch, Simms, Farber, Gokce.
Acquisition of data. Gopal, Doldt, Finch, Simms, Farber, Gokce.
Analysis and interpretation of data. Gopal, Doldt, Finch, Simms, Farber, Gokce.
ROLE OF THE STUDY SPONSOR
Actelion Pharmaceuticals had no role in the study design or in the collection, analysis, or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by Actelion Pharmaceuticals.