Abstract
Aims:
The clinical utility of pulmonary hypertension(PH) risk scores in non-group 1 PH with pulmonary vascular disease(PVD) remains unresolved.
Methods and results:
We utilized the prospective multicenter PVDOMICS cohort with group 2,3,4 or 5 PH related PVD and calculated group 1 PH risk scores[REVEAL 2.0, REVEAL Lite 2, French registry score and COMPERA 2]. The C-statistic to predict death was compared separately in i) precapillary PH groups 3/4/5, and ii) combined post and precapillary PH group 2. Exercise right heart catheterization reserve, ventricular interdependence and RV-PA coupling were compared across risk categories. Among 449 individuals with group 3/4/5 PH, the REVEAL 2.0 risk score had the highest C-statistic for predicting death (0.699 [95% CI 0.660–0.737], p<0.0001) with comparable performance using the simpler REVEAL Lite 2 score (0.695 [95% CI 0.656–0.734], p<0.0001). The French and COMPERA 2 risk scores were also predictive of mortality, but performance of both was statistically inferior to REVEAL 2.0 [C-statistic difference −0.072 [−0.123 to −0.020], p=0.006 and −0.043 [−0.067 to −0.018], p=0.0007 respectively]. RV function and RV-PA coupling measures were prognostic in isolation, but did not add incremental value to REVEAL(p>0.50 for all). Findings were similar in patients with group 2 PH(n=239). Stratification by REVEAL Lite 2 score non-invasively identified non-group 1 PH with more advanced PVD with worse exercise capacity, RV PA uncoupling, ventricular interdependence and impaired cardiac output reserve(p<0.05 for all).
Conclusion:
Non-invasive REVEAL risk predicts mortality in non-group 1 PH without incremental prognostic value from detailed RV function or RV-PA coupling assessment. Baseline REVEAL Lite 2 risk stratification non-invasively identifies greater pulmonary vascular dysfunction and right heart related exercise limitation, which may help guide patient selection for targeted pulmonary vascular therapies in non-group 1 PH.
Keywords: Pulmonary Hypertension, risk scores, REVEAL, exercise hemodynamics
Introduction
Patients with pulmonary hypertension (PH) have increased risk of mortality related to right heart failure. For PH primarily due to pulmonary vascular disease(PVD) (Group 1), pulmonary vasodilators can reduce mortality and improve functional status,1 with intensive upfront therapy providing incremental benefit.2 A number of prognostic scores have been developed in group 1 PH to discriminate risk of adverse outcomes and guide treatment intensity. These include REVEAL 2 (Registry to Evaluate Early and Long-Term PAH Disease Management),3 REVEAL Lite 2,4 the French model5 and COMPERA 2 (Comparative, Prospective Registry of Newly Initiated Therapies for Pulmonary Hypertension).6 Beyond group 1 PH, there is emerging data that these risk scores may also have prognostic value in other forms of PH with pulmonary vascular disease including lung disease (group 3), thromboembolic disease (group 4) or left heart disease with associated PVD (combined post and precapillary (Cpc) group 2 PH).7–9
However, despite inclusion of key prognostic markers such as 6 minute walk distance (6MWD), N-terminal pro-B type natriuretic peptide (NT-proBNP) and New York Heart Association (NYHA) functional class in these risk scores; measures of right ventricular (RV) function or RV-pulmonary artery (PA) coupling are not incorporated despite their known prognostic value in PH.10–13 The relevance of these non-invasive scores to dynamic hemodynamic reserve and exercise capacity in non-group 1 PH is also unknown. To address these knowledge gaps in non-group 1 PH, we utilized the prospective PVDOMICS (Redefining Pulmonary Hypertension through Pulmonary Vascular Disease Phenomics) cohort where participants underwent comprehensive assessment of right heart function and RV-PA coupling with careful follow-up.
Methods
Cohort description
The PVDOMICS clinical research network is a National Heart, Lung, and Blood Institute–funded, prospective, longitudinal cohort study (NCT02980887) that enrolled participants from November 30, 2016, to October 18, 2019 at seven centers across the US [Brigham and Women’s Hospital, Harvard Medical School, Columbia University Irving Medical Center, Weill Cornell Medical Center, Johns Hopkins Hospital, Mayo Clinic (Rochester), University of Arizona (Tucson), and Vanderbilt University Medical Center].14 The protocol was approved at each institution by the local Institutional Review Board, and informed consent was obtained from all participants. Enrolling centers recruited persons ≥18 years of age referred for right heart catheterization (RHC) for clinical purposes with known or suspected PH who were able to complete diagnostic testing. The primary analysis was focused on prognostic impact of PH risk scores in non-group 1 precapillary PH including lung disease (group 3), chronic thromboembolic pulmonary hypertension (group 4) and group 5 PH. We separately evaluated patients with group 2 PH with PVD. A separate cohort of recruited healthy controls were utilized for baseline comparisons.
Clinical evaluation
Participants underwent detailed baseline evaluation including review of their medical history, functional status, measurement of NT-proBNP, 6MWD, core-lab interpreted transthoracic echocardiography and cardiac magnetic resonance imaging (MRI) as previously described.15
All participants underwent resting RHC in the supine position with core-lab adjudication of pulmonary arterial wedge pressure (PAWP) and right atrial (RA) pressure tracings at end-expiration as previously described.16 As per the original PVDOMICS protocol15, the diagnosis of chronic thromboembolic group 4 PVD was based upon hemodynamic presence of mean PA pressure ≥25 mm Hg and PAWP≤15 mm Hg along with anatomic confirmation of chronic thromboembolic disease by either high probability ventilation perfusion scan, or low/intermediate probability ventilation perfusion scan with positive invasive pulmonary angiogram or CT angiogram, supportive pulmonary angiogram and >3 months of therapeutic anticoagulation. PVD related to group 3 or group 5 mechanisms was defined as the presence of associated lung or miscellaneous diseases with mean PA pressure ≥25 mm Hg and PAWP≤15 mm Hg. PVD related to left heart disease (Group 2 or Cpc PH) was defined as mean PA pressure≥25 mm Hg and PAWP>15 mm Hg, with either i) PVR>3 Wood units if cardiac output was>4 L/min or ii) diastolic pressure gradient>7 mm Hg or transpulmonary gradient>12 mm Hg if cardiac output was <4 L/min. Following resting assessment, a subset of patients underwent exercise RHC with simultaneous cardiopulmonary exercise testing based on center availability. To estimate the contribution of diastolic ventricular interaction from the right heart to PAWP elevation and cardiac output impairment, we calculated RA/PAWP ratio and left ventricular transmural pressure (LVTMP) [PAWP-RA pressure] where LVTMP reflects the true distending pressure to the LV.17 As ventricular interdependence and relative pericardial restraint worsens, pericardial pressure (and its surrogate RA) rise with reduction in LVTMP and increase in RA/PAWP ratio.18–20 Participants who did not undergo invasive cardiopulmonary exercise testing underwent a non-invasive cardiopulmonary exercise test.
Non-invasive assessment of right heart function and RV-PA coupling
All patients underwent prospective echocardiography with RV focused views as previously described.21 RV function was assessed using i) Tricuspid Annular Plane Systolic Excursion (TAPSE) from M mode analysis of lateral tricuspid annular longitudinal motion towards the apex ii) tissue doppler imaging of the lateral tricuspid valve annulus to estimate RV free wall systolic peak velocity (RV s’) iii) RV fractional area change (FAC) from RV end-diastolic and end-systolic areas and iv) RV free wall strain with normally negative strain values reported as positive integers for ease of reporting. Cardiac MRI assessment of RV ejection fraction from end-diastolic and end-systolic volumes was also performed as a corollary measure of RV systolic performance.
Pulmonary artery systolic pressure (PASP) was estimated from peak TR velocity using the simplified Bernoulli equation added to estimated RA pressure from the inferior vena cava (4 * peak TR velocity2 + estimated RA pressure). RV-PA coupling was then estimated using TAPSE/PASP10–13 along with analogous measures of RV-PA coupling using alternate metrics of RV function including FAC/PASP and RV s’/PASP as previously described22,23, along with RV strain based RV-PA coupling using RV free wall strain/PASP.
Risk Score Calculation
REVEAL 2.0 and REVEAL Lite 2
The REVEAL 2.0 score was calculated as previously described3 based on assigning scores for being a male>60 years, glomerular filtration rate(GFR)<60 ml/min/1.73m2, NYHA class, systolic blood pressure <110 mm Hg, heart rate>96 beats/minute, 6MWD, NT-proBNP, pericardial effusion on echocardiogram, % predicted DLCO ≤40%, mean right atrial pressure and PVR. Information on all cause hospitalizations within 6 months were not available and given a value of 0. Since the score was applied to non-group 1 PH, the score for group 1 subgroup component was also assigned a value of 0. Missing values for other variables were given a score of 0 which imputes an intermediate risk for missing NT-proBNP, 6MWD and NYHA class. The overall REVEAL 2.0 score was calculated as the sum of individual components plus 6.3 In addition to the continuous scale version of REVEAL 2.0, we also separated patients into previously validated risk categories as low (≤6), intermediate (7–8) or high REVEAL 2.0 risk (≥9).3
The shortened REVEAL Lite 2 score4 was also computed using NT-proBNP, 6MWD, NYHA class, glomerular filtration rate(GFR)<60 ml/min/1.73m2, systolic blood pressure <110 mm Hg and heart rate>96 beats/minute. The resulting REVEAL Lite 2 score was calculated as the sum of individual component scores plus 6. Corresponding low (≤5), intermediate (6–7) and high (≥8) REVEAL Lite 2 score categories were computed as previously validated.4
French Risk Score
The cumulative number of low-risk criteria in the French risk score5 were calculated with low risk defined by the presence of i) NYHA functional class I or II, ii) 6MWD >440 m, iii) right atrial pressure <8 mmHg and iv) cardiac index ≥2.5 L/min/m2. The number of low risk features in the four component French Risk score were summed up with a final score ranging from 0–4 with higher scores implying lower risk. As opposed to the REVEAL 2 score, imputation of 0 assumes high risk for these categories which introduces error. Therefore, only patients with no missing data were included in this analysis similar to the original derivation cohort.5 Although higher French risk scores are associated with lower risk, for ease of presentation the inverse Hazard ratios are presented for the French risk score to be directionally concordant with other risk scores for relationship to death. This French risk score applied has not been previously studied in Group 3 or Group 4 PH.
COMPERA 2 Four Strata Model
The COMPERA 2 four strata model6 was calculated based on points assigned from 1 (low risk) to 4 (high risk) across 3 variables – NYHA class, 6MWD and NT-proBNP. With missing values, the score cannot be calculated as imputation of 0 or 1 assumes low risk for missing NYHA class, 6MWD or NT-proBNP which is not a reliable assumption. Therefore, similar to the original derivation cohort6 and consistent with calculation for the French score5, the COMPERA 2 score was only computed for individuals without missing data for the three components. Individual measured components were averaged and rounded to the nearest whole integer per participant as the final COMPERA 2 risk score. The resulting COMPERA 2 score could be either 1(low risk), 2 (intermediate-low risk), 3 (intermediate-high risk) or 4 (high risk).
Statistical Analysis
Differences across groups were compared using ANOVA, Kruskal Wallis test and chi square test as appropriate. Hemodynamic and functional differences between risk categories stratified by REVEAL Lite 2 score were tested with individual group comparisons by Tukey and Steel Dwass test as appropriate in the setting of multiple comparisons. Patients were followed for all-cause mortality using the log rank test from time of study enrollment based on baseline risk score across the different prognostic models. Cox proportional hazard models were used to determine the hazard ratio for death. For all tested models, the c-statistic was calculated as a measure of model discrimination that evaluates concordance between model predictions of survival time and actual survival time (1.0 indicating perfect concordance).24 Discrimination between models was determined by comparison of c-statistics between models and calculation of associated 95% confidence intervals. Analyses were performed with JMP, version 14.1.0 (JMP Statistical Discovery LLC), and BlueSky Statistics, version 10.3.1 (BlueSky Statistics LLC).
Results
Baseline characteristics
A total of 449 individuals had either WSPH group 3,4 or 5, and 239 had WSPH group 2 (Table 1). Compared to healthy controls (n=96), patients with PH were older and more obese with increased comorbidities including atrial fibrillation, hypertension, diabetes and valvular heart disease (more common in group 2), and with increased obstructive sleep apnea, prior pulmonary emboli and interstitial lung disease (more common in groups 3/4/5). NT-proBNP levels were highest in group 2 PH but were also higher in groups 3/4/5 PH compared to controls. Pulmonary function test abnormalities with reduced DLCO and ground glass opacities on chest computed tomography were common across PH groups. Consistent with cohort selection to have PVD, mean pulmonary vascular resistance was elevated in both PH groups with higher resting PAWP in those with group 2 PH. 57% of included group 2 PH patients with suspected PVD met current hemodynamic criteria for Cpc PH (mean PA>20, mean PAWP>15 and PVR>2 Wood units). Heart failure with preserved ejection fraction(HFpEF) was the most common cause in group 2 PH (76%). A subset of group 2 PH had heart failure with reduced ejection fraction (16%) and the overall group LV ejection fraction was slightly impaired in group 2 PH with higher E/e’ consistent with left heart dysfunction.
Table 1:
Baseline Characteristics of non-group 1 PH compared to healthy controls
| Healthy controls (n=96) | Groups 3/4/5 PH (n=449) | P value vs healthy controls | Group 2 PH (n=239) | P value vs healthy controls | |
|---|---|---|---|---|---|
| Age, years | 47.9 ± 14.3 | 61.9 ± 12.7 | <0.0001 | 65.8 ± 12.4 | <0.0001 |
| Female, % | 70 | 54 | 0.005 | 57 | 0.03 |
| Body mass index, kg/m2 | 27.6 ± 5.7 | 31.0 ± 7.8 | <0.0001 | 32.3 ± 8.9 | <0.0001 |
| Atrial fibrillation, % | 0 | 8 | 0.001 | 20 | <0.0001 |
| Hypertension, % | 15 | 51 | <0.0001 | 61 | <0.0001 |
| Diabetes, % | 4 | 29 | <0.0001 | 37 | <0.0001 |
| Prior Pulmonary emboli, % | 0 | 18 | <0.0001 | 13 | <0.0001 |
| Interstitial Lung Disease, % | 0 | 28 | <0.0001 | 14 | <0.0001 |
| HF with reduced EF, % | 0 | 6 | <0.0001 | 16 | <0.0001 |
| HF with preserved EF, % | 0 | 29 | <0.0001 | 76 | <0.0001 |
| Valvular heart disease, % | 0 | 7 | <0.0001 | 20 | <0.0001 |
| Known coronary disease, % | 1 | 11 | 0.0003 | 18 | <0.0001 |
| Obstructive sleep apnea, % | 0 | 40 | <0.0001 | 34 | <0.0001 |
| Rest/Exertional O2 use, % | 0 | 39/51 | <0.0001 | 24/34 | <0.0001 |
| Lab testing | |||||
| Hemoglobin, g/dl | 13.9 ± 1.3 | 13.5 ± 2.3 | 0.17 | 12.9 ± 2.1 | <0.0001 |
| GFR, ml/min/1.73m2 | 94.3 ± 16.5 | 71.8 ± 24.2 | <0.0001 | 65.7 ± 23.8 | <0.0001 |
| N-Terminal proBNP, pg/ml | 49 [25–76] | 360 [112–1426] | <0.0001 | 784 [242–2296] | <0.0001 |
| Pulmonary Structure/Function | |||||
| FEV1, % predicted | 100 ± 14 | 67 ± 22 | <0.0001 | 66 ± 20 | <0.0001 |
| FVC, % predicted | 104 ± 14 | 75 ± 21 | <0.0001 | 73 ± 18 | <0.0001 |
| DLCO, % predicted | 89 ± 16 | 47 ± 21 | <0.0001 | 49 ± 20 | <0.0001 |
| Ground glass opacities on CT, % | 1 | 45 | <0.0001 | 38 | <0.0001 |
| Hemodynamics | |||||
| Mean PAWP, mm Hg | - | 12.9 ± 6.0 | - | 18.0 ± 6.5 | - |
| Mean PA, mm Hg | - | 38.6 ± 12.0 | - | 40.9 ± 11.4 | - |
| Mean RA, mm Hg | - | 8.3 ± 5.1 | - | 11.1 ± 5.3 | - |
| PVR, Wood units | - | 5.5 ± 3.8 | - | 5.3 ± 3.5 | - |
| Cardiac index, L/min/m2 | - | 2.7 ± 0.9 | - | 2.4 ± 0.8 | - |
| PVR≥5 Wood units, % | - | 44 | - | 43 | - |
| Cardiac Structure/Function | |||||
| LV EF | 61.8 ± 5.5 | 60.4 ± 10.0 | 0.19 | 56.6 ± 12.6 | 0.0002 |
| E/e’ | 6.7 ± 2.5 | 9.3 ± 5.4 | <0.0001 | 13.2 ± 7.6 | <0.0001 |
| Pericardial effusion, % | 1 | 10 | 0.003 | 11 | 0.0009 |
| TAPSE, mm | 23.3 ± 4.2 | 18.5 ± 4.7 | <0.0001 | 17.1 ± 5.2 | <0.0001 |
| TAPSE/PASP | 1.03 ± 0.31 | 0.36 ± 0.17 | <0.0001 | 0.33 ± 0.16 | <0.0001 |
| RV GLS | 26.6 ± 6.4 | 18.9 ± 6.0 | <0.0001 | 18.6 ± 5.9 | <0.0001 |
| RV GLS/PASP | 1.19 ± 0.41 | 0.38 ± 0.21 | <0.0001 | 0.36 ± 0.17 | <0.0001 |
| RV s’, cm/s | 13.1 ± 2.0 | 11.3 ± 3.2 | <0.0001 | 10.5 ± 3.3 | <0.0001 |
| RV s’/PASP | 0.56 ± 0.15 | 0.22 ± 0.11 | <0.0001 | 0.20 ± 0.09 | <0.0001 |
| RV FAC | 44.5 ± 6.7 | 31.5 ± 9.6 | <0.0001 | 31.9 ± 8.8 | <0.0001 |
| RV FAC/PASP | 1.92 ± 0.55 | 0.64 ± 0.35 | <0.0001 | 0.62 ± 0.30 | <0.0001 |
| MRI RV EF | 56.0 ± 6.1 | 40.5 ± 11.8 | <0.0001 | 42.6 ± 11.9 | <0.0001 |
Values represent mean (standard deviation) or median (25th, 75th), TAPSE-Tricuspid Annular Plane Systolic Excursion, PASP-Pulmonary Artery Systolic Pressure, GLS- Global longitudinal strain , s’ -tricuspid valve annular systolic velocity, FEV1-Fraction of forced vital capacity expired in 1 second, FVC-Forced Vital capacity, DLCO – Diffusion Capacity for Carbon Monoxide, GFR-Glomerular Filtration Rate.
Right ventricular function and RV-PA coupling
All measures of RV function were impaired in groups 3/4/5 or group 2 PH compared to controls – whether assessed by MRI (RV ejection fraction) or by echocardiography (TAPSE, RV s’, RV free wall strain and RV FAC) (Table 1). Similarly, all metrics of RV-PA coupling including TAPSE/PASP, RV s’/PASP, RV free wall strain/PASP and RV FAC/PASP were impaired in groups 3/4/5 or group 2 PH compared to healthy controls.
Distribution of risk scores
In group 3/4/5 PH, a total of 30% and 28% were classified as high risk by REVEAL Lite 2 and full REVEAL 2.0 scores respectively (n=449), while 53% and 53% respectively were classified as low risk. In contrast to REVEAL scores which could be calculated in all patients, the French risk score could be calculated in only 76% of patients with group 3/4/5 PH (n=343), of whom only 4% had low risk features across all components while 17% had no low risk features (highest risk). COMPERA 2 scores could be calculated in 78% of patients with group 3/4/5 PH (n=350) with 20% being low risk and 7% being at highest risk. The remaining patients represented intermediate risk.
In group 2 PH, A total of 40% and 35% were classified as high risk by REVEAL lite 2 and full REVEAL 2.0 scores respectively (n=239), while 39% and 40% respectively were classified as low risk. The French risk score could be calculated in 76% of patients with group 2 PH (n=181), of whom only 3% had low risk features across all components while 33% had no low risk features (highest risk). COMPERA 2 scores could be calculated in only 77% of patients with group 2 PH (n=184) with 14% being low risk and 10% being at highest risk.
Prognostic performance of PH risk scores in group 3/4/5 PH
Among patients with group 3/4/5 PH, the REVEAL 2.0 risk score had the highest c-statistic for predicting death (0.699 [95% CI 0.660–0.737], p<0.0001) and was used as reference for subsequent comparisons (Table 2). The REVEAL Lite 2 score despite using fewer input variables demonstrated comparable performance to the full REVEAL 2 score (0.695 [95% CI 0.656–0.734], p<0.0001) without statistically significant difference in c-statistics (c-statistic difference −0.004 [95% CI −0.006 to +0.007], p=0.51). (Figure 1) Categorization of REVEAL 2.0 or REVEAL Lite 2 scores as low, intermediate or high risk resulted in modest lost of discriminatory performance compared to the continuous variable version of REVEAL 2.0.
Table 2:
Survival analysis by risk scores in Group 3/4/5 PH
| HR [95% CI] | p value | C statistic [95% CI] | C statistic comparison [95% CI] | p value | |
|---|---|---|---|---|---|
| Risk scores | |||||
| REVEAL 2, per unit score (n=449) | 1.25 [1.19–1.31] | <0.0001 | 0.699 [0.660–0.737] | Reference | - |
| REVEAL 2, low/int/high risk groups (n=449) | 1.95 [1.64–2.31] | <0.0001 | 0.668 [0.630–0.705] | −0.031 [−0.046 to −0.015] | 0.0001 |
| REVEAL Lite 2, per unit score (n=449) | 1.29 [1.22–1.38] | <0.0001 | 0.695 [0.656–0.734] | +0.006 [−0.007 to +0.004] | 0.51 |
| REVEAL Lite 2, low/int/high risk groups (n=449) | 1.95 [1.64–2.32] | <0.0001 | 0.666 [0.627–0.705] | −0.033 [−0.051 to −0.015] | 0.0004 |
| French low risk score (inverted HR), per unit score (n=343) | 1.38 [1.15–1.66] | 0.0004 | 0.610 [0.560–0.660] | −0.072 [−0.123 to −0.020] | 0.006 |
| Compera 2, per unit score (n=350) | 1.83 [1.45–2.30] | <0.0001 | 0.648 [0.598–0.698] | −0.043 [−0.067 to −0.018] | 0.0007 |
| RV function and RV-PA coupling | |||||
| RV free wall strain, % (n=287) | 0.96 [0.93–0.99] | 0.006 | 0.597 [0.544–0.649] | −0.104 [−0.154 to −0.055] | <0.0001 |
| RV free wall strain/PASP, %/mm Hg (n=254) | 0.16 [0.05–0.46] | 0.0004 | 0.619 [0.564–0.673] | −0.100 [−0.152 to −0.048] | 0.0002 |
| TAPSE, mm (n=345) | 0.95 [0.91–0.98] | 0.005 | 0.583 [0.530–0.635] | −0.105 [−0.159 to −0.050] | 0.0002 |
| TAPSE/PASP, mm/mmHg (n=292) | 0.24 [0.06–0.82] | 0.03 | 0.595 [0.537–0.652] | −0.105 [−0.159 to −0.051] | 0.0001 |
| RV s’, cm/s (n=375) | 0.96 [0.91–1.01] | 0.15 | 0.555 [0.507–0.603] | −0.139 [−0.191 to −0.087] | <0.0001 |
| RV s’/PASP cm/s.mmHg (n=317) | 0.23 [0.03–1.41] | 0.12 | 0.584 [0.532–0.636] | −0.123 [−0.174 to −0.070] | <0.0001 |
| RV FAC, % (n=357) | 0.98 [0.96–0.99] | 0.011 | 0.591 [0.540–0.642] | −0.103 [−0.153 to −0.054] | <0.0001 |
| RV FAC/PASP, %/mm Hg (n=311) | 0.37 [0.20–0.66] | 0.0006 | 0.620 [0.566–0.674] | −0.090 [−0.143 to −0.038] | 0.0007 |
| MRI RV EF (n=261) | 0.97 [0.96–0.99] | 0.004 | 0.595 [0.531–0.660] | −0.095 [−0.159 to −0.032] | 0.003 |
Figure 1: Risk of death in group 3/4/5 PH by baseline PH score.
Increasing risk score categories by REVEAL2.0(A), REVEAL Lite 2(B), French Risk score(C) and COMPERA 2 scores(D) were associated with increased mortality
The French and COMPERA 2 risk scores were predictive of risk of death (0.610 [95% CI 0.560–0.660] for French score, 0.648 [95% CI 0.598–0.698] for COMPERA 2), but performance of both scores was statistically inferior to the REVEAL 2.0 score. The subset of patients classified as intermediate risk (n=256) with COMPERA 2, could be accurately risk re stratified as low risk by applying the REVEAL Lite 2 algorithm stratification [low REVEAL Lite 2 Risk HR 0.51(95% CI .031–0.84), p=0.009 vs intermediate REVEAL Lite 2 risk, HR 0.42 (95% CI 0.25–0.72) vs high REVEAL Lite 2 risk, p=0.002].
Prognostic value of RV-PA coupling in group 3/4/5 PH
Among measures of RV function - RV free wall strain, TAPSE and FAC by echocardiography and MRI RV EF were all predictive of risk of death (Table 2). Similar to RV function measures, measures of RV-PA coupling using RV free wall strain, TAPSE and FAC indexed to PASP were all associated with mortality. In contrast, neither RV s’ nor RV s’/PASP was associated with risk of death.
Notably, although the above measures of RV function and RV-PA coupling were all associated with mortality in univariate analyses, they were each inferior to the multiparametric REVEAL 2 score at predicting mortality [C statistic difference ranging from −0.139 to −0.090, p<0.004 for all] (Table 2). When RV free wall strain, TAPSE, RV EF or measures of RV PA coupling using RV strain or TAPSE were incorporated into either the REVEAL 2.0 or REVEAL Lite 2 risk score, there was no incremental prognostic value from the addition of RV function or RV-PA coupling indices (Table S1). Results were unchanged when considering only patients with group 3/4 PH without group 5 PH in a sensitivity analysis (Table S2), or when considering group 3 or 4 PH separately (Table S3).
Prognostic value of PH risk scores and RV-PA coupling in group 2 PH with PVD
When applied to patients with group 2 PH, the REVEAL 2.0 risk score again demonstrated prognostic value with comparable performance using the simpler REVEAL Lite 2 score with less input variables (Table S4). The COMPERA 2 and French risk scores were also predictive of outcome in group 2 PH. Compared to REVEAL 2, COMPERA 2 was statistically less discriminatory (c-statistic comparison −0.046 [95% CI −0.083 to −0.009], p=0.02) and the French risk score tended to have lower discrimination although this was not statistically significant (c-statistic comparison −0.068 [−0.143 to +0.009], p=0.09). Measures of RV function and RV-PA coupling in group 2 PH provided comparable prognostic value as the multiparametric REVEAL 2 score (Table S4). But similar to group 3/4/5 PH there was no incremental prognostic value over the REVEAL 2 score with incorporation of RV function or RV-PA coupling in group 2 PH (Table S5).
Resting Hemodynamic and Functional consequences of increased REVEAL Lite 2 risk in group 3/4/5 PH
To evaluate clinical implications of risk stratifying patients with PH using simple non-invasive measures, we stratified patients into low, intermediate and high risk by REVEAL Lite 2 in group 3/4/5 PH. Despite not being included in the risk score calculation, RV free wall strain, RV functional metrics and RV-PA coupling by echo were progressively worse with increasing risk, along with greater MRI based RV and right atrial dilation and dysfunction (Figure 3) (Table 4). Pulmonary vascular disease markers including PVR and PA compliance were worse with increasing risk, along with lower DLCO and more pericardial effusions. Troponin levels were highest and body mass index and serum albumin lowest with higher REVEAL Lite 2 risk suggestive of more advanced disease. Quality of life by SF-36, Minnesota Living with Heart Failure and Emphasis-10 scores were all impaired with increased REVEAL Lite 2 risk.
Figure 3: Differences in resting right heart function and pulmonary vascular disease in group 3/4/5 PH stratified by REVEAL Lite 2 score.
Higher REVEAL Lite 2 risk categorization in group 3/4/5 PH was associated with worse RV free wall longitudinal strain (A) and RV-PA coupling by echocardiography (RV free wall strain/PA systolic pressure). Cardiac MRI also demonstrated worse RV and RA function with lower RV ejection fraction (C) and RA emptying fraction (D). Right heart catheterization demonstrated worse pulmonary vascular resistance (PVR) and PA compliance with increasing risk categorization.
Table 4:
Cardiopulmonary response and exercise hemodynamic reserve among group 3/4/5 PH stratified by REVEAL lite 2 risk
| REVEAL Lite 2 low risk gp 345 (n=237) | REVEAL Lite 2 intermediate risk gp 345 (n=88) | REVEAL Lite 2 high risk gp 345 (n=124) | p value | |
|---|---|---|---|---|
| Rest catheterization | ||||
| Mean PA, mm Hg (n=441) | 34.3 ± 10.2* | 38.9 ± 10.5* | 46.5 ± 12.2* | <0.0001 |
| Mean RA, mm Hg (n=441) | 6.5 ± 3.7* | 9.1 ± 4.8* | 11.1 ± 6.0* | <0.0001 |
| Mean PAWP, mm Hg (n=439) | 11.5 ± 5.3* | 14.9 ± 6.1 | 14.2 ± 6.6 | <0.0001 |
| Rest RA/PAWP, (n=438) | 0.59 ± 0.31 | 0.64 ± 0.48 | 0.92 ± 1.06* | <0.0001 |
| Rest LVTMP, mm Hg (n=438) | 4.9 ± 3.9 | 5.8 ± 4.9 | 3.3 ± 5.9* | 0.0003 |
| PVR, Wood units (n=428) | 4.3 ± 2.5 | 5.2 ± 2.8 | 8.2 ± 5.2* | <0.0001 |
| PA compliance, ml/mm Hg (n=432) | 2.8 ± 1.3* | 2.2 ± 1.4* | 1.5 ± 0.9* | <0.0001 |
| Rest CO, L/min (n=432) | 5.8 ± 1.7* | 5.2 ± 1.8 | 4.7 ± 1.8 | <0.0001 |
| Rest stroke volume, ml (n=432) | 80 ± 25* | 71 ± 30* | 60 ± 23* | <0.0001 |
| Exercise catheterization/CPET | ||||
| Peak watts, (n=280) | 60 ± 39* | 35 ± 26 | 30 ± 20 | <0.0001 |
| Peak mean PA, mm Hg (n=167) | 53.6 ± 13.5 | 56.3 ± 13.3 | 60 ± 14.5 | 0.06 |
| Peak RA, mm Hg (n=163) | 10.7 ± 6.5* | 17.2 ± 9.5 | 20.9 ± 9.0 | <0.0001 |
| Peak PAWP, mm Hg (n=157) | 17.0 ± 8.7 | 21.0 ± 9.9 | 20.4 ± 10.0 | 0.04 |
| Peak RA/PAWP, (n=153) | 0.64 ± 0.32* | 0.95 ± 0.64 | 1.19 ± 0.73 | <0.0001 |
| Peak LVTMP, mm Hg (n=153) | 6.2 ± 6.0* | 3.7 ± 9.0 | −0.4 ± 7.6 | <0.0001 |
| Peak PVR, Wood units (n=133) | 4.2 ± 2.3 | 5.2 ± 3.2 | 7.1 ± 4.2* | <0.0001 |
| Peak CO, L/min (n=141) | 9.6 ± 2.9* | 7.4 ± 2.6* | 5.6 ± 2.0* | <0.0001 |
| Peak Stroke volume, ml (n=138) | 84 ± 26* | 67 ± 25 | 57 ± 24 | <0.0001 |
| Peak VO2, ml/kg/min (n=266) | 12.1 ± 3.8* | 8.8 ± 2.1 | 7.8 ± 2.2 | <0.0001 |
| % predicted peak VO2, (n=266) | 57.3 ± 19.5* | 45.8 ± 13.7 | 40.9 ± 13.0 | <0.0001 |
| ΔCO/ ΔVO2 (n=127) | 6.1 ± 3.1 | 5.3 ± 3.6 | 3.5 ± 2.5L | 0.002 |
| Peak CO, % predicted (n=127) | 101 ± 52 | 88 ± 61 | 58 ± 42L | 0.002 |
| Peak O2 pulse, (n=262) | 8.7 ± 3.2* | 7.1 ± 2.9 | 6.5 ± 2.4 | <0.0001 |
| Peak Heart Rate, bpm (n=275) | 120 ± 21* | 108 ± 22 | 103 ± 22 | <0.0001 |
| Peak AVO2 difference, ml/dl (n=140) | 10.4 ± 3.0 | 10.3 ± 2.6 | 12.2 ± 3.2* | 0.02 |
| Peak systolic BP, mm Hg (n=272) | 160 ± 31 | 157 ± 35 | 142 ± 32L | 0.003 |
| Peak VE/VCO2 (n=250) | 40 ± 10 | 45 ± 12 | 50 ± 15L | 0.0003 |
p<0.05 vs all
p<0.05 vs no low risk
Exertional Hemodynamic consequences of increased REVEAL Lite 2 risk in group 3/4/5 PH
Peak exercise capacity (peak VO2) was markedly reduced with intermediate and high risk in group 3/4/5 PH, along with progressively worsening exercise cardiac output and stroke volume reserve and higher right atrial pressures during rest and exercise with ventilatory inefficiency (Figure 4) (Table 4). Despite having precapillary PH, there were higher pulmonary capillary wedge pressures (PACWP) with higher RA/PAWP and lower LV transmural pressures during exercise with intermediate and high REVEAL Lite 2 risk suggestive of greater ventricular interdependence and LV underfilling from the right heart compromising exertional cardiac output.
Figure 4: Exercise hemodynamic reserve by baseline REVEAL Lite 2 score in group 3/4/5 PH.
Higher risk categories were associated with lower peak oxygen consumption (VO2) (A) and lower cardiac output reserve during exercise (B). There was also evidence of higher Right atrial pressure (RAP) (C) during exercise with exaggerated ventricular interdependence contributing to higher exertional PAWP (D,E) despite left ventricular underfilling with lower LV transmural pressure (F).
Sensitivity analyses performed separately in group 3 or group 4 PH demonstrated clinical, structural and hemodynamic stratification across REVEAL Lite 2 risk categories consistent with the overall study results. (Table S6–9)
Resting and exertional consequences of increased REVEAL Lite 2 risk in group 2 PH with PVD
Similarly, patients with group 2 PH related PVD with higher REVEAL Lite 2 risk also demonstrated greater pulmonary vascular disease, worse quality of life, troponin elevation, RV dysfunction and RV-PA uncoupling (Table S10). There was progressive worsening of exercise capacity and cardiac output reserve during exercise with higher RA pressures and comparable PAWP elevation during exercise (Table 5). However, the contribution of ventricular interdependence from the right heart to the measured PAWP during exercise was greater (higher RA/PAWP and lower LVTMP) with higher REVEAL Lite 2 risk scores.
Table 5:
Cardiopulmonary response and exercise hemodynamic reserve among group 2 PH stratified by REVEAL lite 2 risk
| REVEAL Lite 2 low risk gp 2 (n=96) | REVEAL Lite 2 intermediate risk gp 2 (n=59) | REVEAL Lite 2 high risk gp 2 (n=84) | p value | |
|---|---|---|---|---|
| Rest catheterization | ||||
| Mean PA, mm Hg (n=235) | 38.2 ± 12.8 | 40.7 ± 9.8 | 44.1 ± 10.0L | 0.002 |
| Mean RA, mm Hg (n=234) | 9.3 ± 4.9 | 11.3 ± 4.6 | 12.8 ± 5.6L | <0.0001 |
| Mean PAWP, mm Hg (n=233) | 17.0 ± 6.2 | 19.0 ± 6.6 | 18.5 ± 6.7 | 0.14 |
| Rest RA/PAWP, (n=232) | 0.55 [0.39–0.71] | 0.58 [0.45–0.73] | 0.70 [0.50–0.93]L | 0.001 |
| Rest LVTMP, mm Hg (n=232) | 7.6 ± 4.8 | 7.6 ± 5.8 | 5.7 ± 6.6 | 0.06 |
| PVR, Wood units (n=227) | 4.2 ± 2.6 | 5.4 ± 3.4 | 6.5 ± 4.1L | <0.0001 |
| PA compliance, ml/mm Hg (n=229) | 2.57 ± 1.22* | 2.05 ± 1.24* | 1.59 ± 0.78* | <0.0001 |
| Rest CO, L/min (n=229) | 5.4 ± 1.5* | 4.5 ± 1.5 | 4.5 ± 1.7 | <0.0001 |
| Rest stroke volume, ml (n=229) | 75.1 ± 23.2* | 64.5 ± 23.0 | 62.4 ± 23.9 | 0.0009 |
| Exercise catheterization/CPET | ||||
| Peak watts, (n=137) | 60.2 ± 36.2 | 51.4 ± 27.7 | 33.6 ± 21.4* | 0.0002 |
| Peak mean PA, mm Hg (n=76) | 56.3 ± 17.3 | 52.1 ± 10.3 | 54.5 ± 12.3 | 0.57 |
| Peak RA, mm Hg (n=73) | 16.4 ± 7.7 | 22.1 ± 9.1 | 21.8 ± 10.0 | 0.047 |
| Peak PAWP, mm Hg (n=74) | 25.3 ± 7.9 | 28.5 ± 8.7 | 25.6 ± 9.8 | 0.40 |
| Peak RA/PAWP, (n=71) | 0.61 [0.48–0.82] | 0.76 [0.57–1.00] | 0.87 [0.59–1.00]L | 0.02 |
| Peak LVTMP, mm Hg (n=71) | 9.5 ± 6.0 | 6.5 ± 8.1 | 3.8 ± 8.9L | 0.04 |
| Peak PVR, Wood units (n=61) | 4.2 ± 2.7 | 3.7 ± 1.9 | 5.2 ± 4.3 | 0.30 |
| Peak CO, L/min (n=61) | 8.3 ± 3.4 | 6.9 ± 3.0 | 5.5 ± 1.7L | 0.008 |
| Peak Stroke volume, ml (n=59) | 78.3 ± 31.8 | 65.3 ± 28.9 | 58.1 ± 24.2 | 0.08 |
| Peak VO2, ml/kg/min (n=131) | 11.3 ± 3.1* | 9.6 ± 3.3 | 8.2 ± 2.2 | <0.0001 |
| % predicted peak VO2, (n=131) | 58.9 ± 16.9* | 50.1 ± 19.5 | 45.2 ± 15.1 | 0.0006 |
| Peak O2 pulse, (n=129) | 9.0 ± 3.1 | 7.9 ± 3.2 | 6.9 ± 2.7L | 0.007 |
| Peak Heart Rate, bpm (n=135) | 115 ± 24 | 105 ± 24 | 104 ± 27 | 0.07 |
| Peak systolic BP, mm Hg (n=135) | 164 ± 35 | 154 ± 31 | 146 ± 27L | 0.03 |
| Peak VE/VCO2 (n=124) | 37.9 ± 9.6 | 43.4 ± 12.9 | 43.1 ± 11.0 | 0.03 |
p<0.05 vs all
p<0.05 vs no low risk
Discussion
In the current study, we evaluated the prognostic performance and clinical implications of existing group 1 PH risk scores in a prospectively recruited multicenter PVDOMICS cohort of non-group 1 PH where all participants underwent detailed phenotyping with measurement of right ventricular performance, RV-PA coupling and hemodynamic functional reserve. We demonstrate that non-invasive PH risk scores provide meaningful discrimination of prognosis even in patients with PH and PVD among groups 2–4 with the REVEAL 2 and REVEAL Lite 2 scores providing the highest statistical discrimination of risk. Since the REVEAL Lite 2 score requires only two lab values (NT-proBNP and serum creatinine), two vital signs (systolic blood pressure and heart rate), a history (NYHA functional class) and 6MWD without right heart catheterization or echocardiographic data, this score may provide the greatest balance between ease of clinical use, feasibility and prognostic accuracy. Stratifying non-group 1 PH patients with PVD by REVEAL Lite 2 risk was able to non-invasively identify individuals with more advanced PVD, right heart dysfunction, RV-PA uncoupling, and impaired cardiac output response during exercise with enhanced ventricular interdependence related to right heart disease. The ability of the REVEAL Lite 2 score to separate individuals by degree of baseline RV dysfunction resulted in no incremental prognostic value from detailed measurement of RV function or RV-PA coupling by all currently available measures including RV strain and cardiac MRI. These findings were consistent across patients with PVD related to left heart disease and non-group 1 precapillary PH (groups 3/4/5). These data therefore suggest that PH risk scores and particularly REVEAL Lite 2 may have clinical utility to predict prognosis and exercise hemodynamic reserve related to pulmonary vascular dysfunction across the entire spectrum of PH with PVD. By facilitating simple non-invasive identification of individuals with greatest pulmonary vascular related ventricular interdependence and exercise cardiac output impairment, baseline REVEAL Lite 2 stratification may serve as a useful trial enrichment criteria to enroll Cpc PH and non-group 1 precapillary PH patients with greatest potential to benefit from novel therapies targeting pulmonary vascular dysfunction, without the need for exercise right heart catheterization.
Potential utility of PH risk scores in non-group 1 PH
PH risk score stratification is currently endorsed by guidelines in group 1 PH25 to identify individuals at higher risk who may benefit from more aggressive upfront vasodilator therapy and more careful follow up to confirm clinical improvement. In recent years there has also been a substantial increase in therapeutic options for non-group 1 PH. Patients with group 3 PH related to interstitial lung disease derive benefit from inhaled treprostinil to improve functional status and reduce clinical worsening events.26 Although clinical response to therapy is heterogeneous, identifying individuals with greater ventricular interdependence from PVD and cardiac output impairment during exercise using non-invasive risk scores may help enrich for a cohort with increased likelihood to benefit from inhaled vasodilators to improve this physiology.22 There is also evidence that higher achieved inhaled dose may be associated with less clinical worsening suggesting potential for guiding treatment intensity based on risk, followed by early consideration of lung transplantation in eligible patients at high risk of death.27 For group 4 PH, riociguat, pulmonary endarterectomy and balloon angioplasty are all proven therapies.28–30 Identification of individuals with worse prognosis and greater likelihood of hemodynamic reserve limitations may help stratify patients for pretreatment with riociguat to lower baseline risk prior to mechanical intervention with balloon angioplasty, a strategy associated with decreased complications in the RACE trial.29 Baseline risk stratification may also help guide shared decision making about consideration of surgical pulmonary endarterectomy. Risk stratification of patients with left heart disease associated PVD may prompt more aggressive pharmacotherapy with heart failure with reduced ejection fraction; or sodium glucose co-transporter 2 inhibitors or semaglutide (with obese HFpEF), strategies that are associated with improved pulmonary hypertension, RV-PA coupling and outcomes.31–36 With persistent high risk stratification in group 2 PH despite these established therapies, whether novel therapies directed to the pulmonary vasculature and ventricular interdependence22 may be beneficial requires further study. As recently summarized37, optimal patient selection for clinical trials of novel therapies in PVD related to left heart disease remains challenging. By demonstrating that non-invasive REVEAL Lite 2 scores are predictive of PVD related exercise hemodynamic reserve and mortality, this score could potentially be utilized to more pragmatically identify group 2 PH patients likely to benefit from novel therapies addressing exertional left heart underfilling and cardiac output reserve related to PVD in clinical trials.
Relationship to prior literature
A recent registry study demonstrated that PH risk scores were prognostic across non-group 1 PH.7 Inherent to a registry analysis however, there was a large amount of missing data and the primary analysis utilized imputation approaches which may introduce error if data is not truly missing at random, but rather missing related to clinical worsening or patient status. Our findings in a prospectively recruited cohort are reassuring in its reproducibility of the registry findings showing prognostic value of all tested risk scores in non-group 1 PH, with redemonstration of the superior prognostic value of the REVEAL 2 risk scores. Together with prior studies demonstrating prognostic value of risk scores in group 4 PH8,9, these data add to the totality of evidence validating these PH risk models across all forms of PVD.
Importantly, through prospective echocardiography and MRI protocols we were able to evaluate the incremental value of RV function and RV-PA coupling demonstrating that although prognostic in isolation, these complex measures may not add incremental prognostic value in non-group 1 PH beyond simpler clinical measures incorporated in risk scores. Although prior studies have suggested incremental prognostic value from detailed RV function assessment in group 1 PH38, patients with group 1 PH have a more isolated pulmonary vascular-right heart pathological process, whereas non-group 1 PH patients by definition have 2 pathological processes with both pulmonary vascular disease and separate primary pathology (lung disease, left heart disease or chronic thromboemboli). Therefore, the competing risk from the primary pathological process in non-group 1 PH may limit the incremental prognostic value of RV function or RV-PA coupling measures. The REVEAL Lite 2 based stratification also appeared to indirectly capture information related to resting RV dysfunction and RV-PA coupling which may limit the incremental value of detailed resting RV assessment. Additionally, resting measures of RV performance may be uncoupled from dynamic RV myocardial reserve and diastolic impact of the RV on the left ventricle through ventricular interdependence. Differences in LV filling and performance as a result of this ventricular interdependence from the right heart likely contributed to worse cardiac output reserve and exercise capacity with higher REVEAL Lite 2 risk categories as previously shown in group 2 PH with PVD.18 Further study of the prognostic value of dynamic RV assessment, left-right heart diastolic interaction and exercise hemodynamic reserve to improve risk stratification and clinical decision making in non-group 1 PH beyond resting RV measurements are needed.
Strengths and Limitations
The present analysis has multiple strengths, including its multicenter representation, comprehensive, prospective evaluation of right heart function through detailed echocardiography, MRI and right heart catheterization with exercise with core laboratory-based interpretation to reduce variability. Despite prospective data collection, there were missing data for some key functional metrics required for the French and COMPERA risk scores. However, this represents clinical reality where all metrics may not be universally available at the point of clinical evaluation similar to prior registry studies, and the current study provides reasonable comparative estimates of the feasibility of calculating the various scores in clinical practice. Evaluation for undiagnosed chronic thromboembolic disease or coronary artery disease was not protocolized in all patients, but was instead performed at the discretion of the site principal investigator based on clinical suspicion. We did not have information on performance of subsequent surgical thromboendarterectomy or balloon angioplasty for group 4 PH which may confound baseline risk prediction requiring further study. Sensitivity analyses evaluating risk score performance in group 4 or 3 PH separately were consistent with the primary results suggesting that REVEAL based risk prediction was able to predict mortality across the spectrum of precapillary PVD, similar to a prior registry study.7 Hemodynamic reserve stratification by REVEAL score also demonstrated consistent differences in group 4 and group 3 patients separately, although further study of the interaction between subsequent thromboembolic treatment and REVEAL Lite 2 hemodynamic stratification in group 4 PH is needed. Our exercise hemodynamic associations were only assessed at baseline and although prior studies have suggested that riociguat can lower baseline REVEAL risk status in group 4 PH9, whether targeted chronic thromboembolic directed therapy is meaningfully associated with improved pulmonary vascular related hemodynamic reserve requires further study. The determination of Group 2, 3, 4 or 5 PH was made by integrated assessment by site PI with further adjudication as needed by a committee of PH physicians. Although such categorizations are inherently subjective, this approach is consistent with current clinical practice increasing generalizability of our findings in non-group 1 PH.
Conclusion
Among patients with non-group 1 PH and PVD of varying etiologies, available multiparametric PH risk scores allow prediction of future risk of death with the REVEAL 2 and REVEAL Lite 2 algorithm provided the greatest statistical discrimination. Baseline stratification of risk by the simple REVEAL Lite 2 algorithm identifies individuals with non-group 1 PH who have worse exercise capacity, ventricular interdependence, and cardiac output reserve related to PVD. Although RV dysfunction and RV-PA uncoupling by echocardiography and MRI are individually prognostic, they may not add incremental value to simpler clinical PH risk scores in non-group 1 PH. REVEAL Lite 2 stratification may therefore allow pragmatic identification of patients in clinical trials who are most likely to benefit from novel PVD targeted therapies in either Cpc PH or non-group 1 related precapillary PH, without the need for complex imaging based RV-PA coupling assessment or exercise hemodynamic evaluation.
Supplementary Material
Figure 2: Risk of death in group 2 PH by baseline PH score.
Increasing risk score categories by REVEAL2.0(A), REVEAL Lite 2(B), French Risk score(C) and COMPERA 2 scores(D) were associated with increased mortality
Table 3:
Markers of clinical status and right heart remodeling in group 3/4/5 PH stratified by REVEAL lite 2 risk
| REVEAL Lite 2 low risk gp 345 (n=237) | REVEAL Lite 2 intermediate risk gp 345 (n=88) | REVEAL Lite 2 high risk gp 345 (n=124) | p value | |
|---|---|---|---|---|
| Age, years (n=449) | 58.3 ± 12.6* | 65.0 ± 12.4 | 66.6 ± 10.7 | <0.0001 |
| Body mass index, kg/m2(n=449) | 31.6 ± 8.3 | 32.0 ± 7.2 | 29.2 ± 7.0* | 0.008 |
| Diuretics, % (n=446) | 51 | 58 | 78 | <0.0001 |
| Rest O2 use, % (n=446) | 32 | 45 | 48 | 0.004 |
| DLCO, % predicted | 51.8 ± 21.7* | 43.5 ± 19.4 | 40.0 ± 20.6 | <0.0001 |
| Physical Component Summary SF-36 (n=431) | 36.6 ± 10.0* | 32.3 ± 8.5 | 31.9 ± 7.9 | <0.0001 |
| Minnesota Living with Heart Failure score (n=431) | 42.1 ± 27.4* | 53.3 ± 25.3 | 52.1 ± 26.8 | 0.0003 |
| EmPHAsis-10 score (n=429) | 25.3 ± 12.4* | 30.3 ± 11.4 | 31.8 ± 11.0 | <0.0001 |
| Albumin, g/dl (n=432) | 4.1 ± 0.4* | 3.9 ± 0.4 | 3.9 ± 0.5 | <0.0001 |
| Troponin, pg/ml (n=439) | 9.5 ± 5.5 | 10.9 ± 7.9 | 18.5 ± 22.4* | <0.0001 |
| Magnetic resonance imaging | ||||
| RV EF (n=261) | 43.3 ± 10.6 | 42.1 ± 12.0 | 32.9 ± 11.1* | <0.0001 |
| RV GLS (n=214) | 17.4 ± 4.8 | 17.4 ± 4.7 | 13.3 ± 4.3* | <0.0001 |
| RV end diastolic volume, ml (n=261) | 177 ± 73 | 171 ± 66 | 216 ± 63* | 0.0003 |
| RV mass, g (n=260) | 35.0 ± 14.9 | 33.8 ± 14.6 | 42.3 ± 16.1* | 0.002 |
| RA minimal volume, ml (n=262) | 43 ± 27 | 48 ± 33 | 89 ± 54* | <0.0001 |
| RA maximal volume, ml (n=262) | 69 ± 35 | 72 ± 38 | 116 ± 56* | <0.0001 |
| RA emptying fraction, % (n=262) | 38.8 ± 15.4 | 35.7 ± 17.5 | 26.0 ± 14.3* | <0.0001 |
| Echocardiography | ||||
| Echo RVSP, mm Hg (n=362) | 52.5 ± 17.8 | 57.5 ± 18.8 | 65.2 ± 19.7* | <0.0001 |
| Echo RA volume, ml (n=416) | 54 ± 31* | 66 ± 37* | 86 ± 39* | <0.0001 |
| RV end diastolic area, cm2 (n=362) | 25.7 ± 9.2 | 27.5 ± 9.0 | 30.1 ± 7.0L | 0.0002 |
| Tricuspid E/e’, (n=243) | 5.4 ± 2.1 | 6.1 ± 3.3 | 6.8 ± 3.5L | 0.008 |
| Pericardial effusion, % | 5 | 14 | 15 | 0.003 |
| RV GLS, % (n=287) | 20.9 ± 5.4* | 18.8 ± 5.7* | 15.2 ± 5.4* | <0.0001 |
| GLS/PASP, %/mm Hg (n=254) | 0.46 ± 0.23* | 0.37 ± 0.19* | 0.25 ± 0.13* | <0.0001 |
| TAPSE, mm (n=345) | 19.7 ± 4.4* | 18.1 ± 4.4* | 16.0 ± 4.7* | <0.0001 |
| TAPSE/PASP, mm/mm Hg (n=292) | 0.42 ± 0.17* | 0.35 ± 0.17* | 0.27 ± 0.12* | <0.0001 |
| RV s’, cm/s (n=375) | 12.1 ± 3.2* | 11.4 ± 3.0 | 9.9 ± 2.9 | <0.0001 |
| RV s’/PASP, cm/s.mmHg (n=317) | 0.25 ± 0.11* | 0.22 ± 0.12 | 0.17 ± 0.08 | <0.0001 |
| RV FAC, % (n=357) | 34.3 ± 8.9* | 32.3 ± 9.6 | 25.7 ± 8.5 | <0.0001 |
| RV FAC/PASP, %/mm Hg (n=311) | 0.76 ± 0.37* | 0.64 ± 0.33 | 0.44 ± 0.24 | <0.0001 |
p<0.05 vs all
p<0.05 vs no low risk
Acknowledgments
The authors thank the patients participating in the PVDOMICS network who agreed to participate in research, allowing for this study to be completed.
Support
The study was supported by grants from the NIH/NHLBI: U01 HL125218 (PI: E.B. Rosenzweig), U01 HL125205 (PI: R.P. Frantz), U01 HL125212 (PI: A.R. Hemnes), U01 HL125208 (PI: F.P. Rischard), U01 HL125175 (PI: P.M. Hassoun), U01 HL125215 (PI: J.A. Leopold), and U01 HL125177 (PI: G.J. Beck), and by the Pulmonary Hypertension Association. Dr Reddy is supported by NIH grant K23HL164901. Dr. Borlaug is supported by R01 HL128526, R01 HL162828, and U01 HL160226, from the NIH/NHLBI, and W81XWH2210245 from the US Department of Defense.
Disclosures
Dr. Reddy receives research grants from NHLBI, Sleep Number, Bayer Accelerated Pulmonary Hypertension Award, United Jenesis Award, Merck, and the Earl Wood Career development award from Mayo Clinic. Dr. Borlaug receives research support from the National Institutes of Health (NIH) and the United States Department of Defense, as well as research grant funding from AstraZeneca, Axon, GlaxoSmithKline, Medtronic, Mesoblast, Novo Nordisk, and Tenax Therapeutics. Dr. Borlaug has served as a consultant for Actelion, Amgen, Aria, Axon Therapies, BD, Boehringer Ingelheim, Cytokinetics, Edwards Lifesciences, Eli Lilly, Imbria, Janssen, Merck, Novo Nordisk, NGM, NXT, and VADovations, and is named inventor (US Patent no. 10,307,179) for the tools and approach for a minimally invasive pericardial modification procedure to treat heart failure. Dr Mukherjee is supported by National Scleroderma Foundation, NIH/NHLBI R01 - 90099095, Department of Defense
References
- 1.Barst RJ, Rubin LJ, Long WA, et al. A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. N Engl J Med. 1996;334(5):296–301. doi: 10.1056/NEJM199602013340504 [DOI] [PubMed] [Google Scholar]
- 2.Galie N, Barbera JA, Frost AE, et al. Initial Use of Ambrisentan plus Tadalafil in Pulmonary Arterial Hypertension. N Engl J Med. 2015;373(9):834–844. doi: 10.1056/NEJMoa1413687 [DOI] [PubMed] [Google Scholar]
- 3.Benza RL, Gomberg-Maitland M, Elliott CG, et al. Predicting Survival in Patients With Pulmonary Arterial Hypertension: The REVEAL Risk Score Calculator 2.0 and Comparison With ESC/ERS-Based Risk Assessment Strategies. Chest. 2019;156(2):323–337. doi: 10.1016/j.chest.2019.02.004 [DOI] [PubMed] [Google Scholar]
- 4.Benza RL, Kanwar MK, Raina A, et al. Development and Validation of an Abridged Version of the REVEAL 2.0 Risk Score Calculator, REVEAL Lite 2, for Use in Patients With Pulmonary Arterial Hypertension. Chest. 2021;159(1):337–346. doi: 10.1016/j.chest.2020.08.2069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Boucly A, Weatherald J, Savale L, et al. Risk assessment, prognosis and guideline implementation in pulmonary arterial hypertension. Eur Respir J. 2017;50(2):1700889. doi: 10.1183/13993003.00889-2017 [DOI] [PubMed] [Google Scholar]
- 6.Hoeper MM, Pausch C, Olsson KM, et al. COMPERA 2.0: a refined four-stratum risk assessment model for pulmonary arterial hypertension. Eur Respir J. 2022;60(1):2102311. doi: 10.1183/13993003.02311-2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yogeswaran A, Gall H, Fünderich M, et al. Comparison of contemporary risk scores in all groups of pulmonary hypertension - a PVRI GoDeep meta-registry analysis. Chest. Published online March 18, 2024:S0012–3692(24)00309-X. doi: 10.1016/j.chest.2024.03.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Humbert M, Farber HW, Ghofrani HA, et al. Risk assessment in pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension. Eur Respir J. 2019;53(6):1802004. doi: 10.1183/13993003.02004-2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Benza RL, Farber HW, Frost A, et al. REVEAL risk score in patients with chronic thromboembolic pulmonary hypertension receiving riociguat. J Heart Lung Transplant. 2018;37(7):836–843. doi: 10.1016/j.healun.2018.02.015 [DOI] [PubMed] [Google Scholar]
- 10.Tello K, Wan J, Dalmer A, et al. Validation of the Tricuspid Annular Plane Systolic Excursion/Systolic Pulmonary Artery Pressure Ratio for the Assessment of Right Ventricular-Arterial Coupling in Severe Pulmonary Hypertension. Circ Cardiovasc Imaging. 2019;12(9):e009047. doi: 10.1161/CIRCIMAGING.119.009047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Tello K, Axmann J, Ghofrani HA, et al. Relevance of the TAPSE/PASP ratio in pulmonary arterial hypertension. Int J Cardiol. 2018;266:229–235. doi: 10.1016/j.ijcard.2018.01.053 [DOI] [PubMed] [Google Scholar]
- 12.Richter MJ, Peters D, Ghofrani HA, et al. Evaluation and Prognostic Relevance of Right Ventricular-Arterial Coupling in Pulmonary Hypertension. Am J Respir Crit Care Med. 2020;201(1):116–119. doi: 10.1164/rccm.201906-1195LE [DOI] [PubMed] [Google Scholar]
- 13.Schmeisser A, Rauwolf T, Groscheck T, et al. Pressure-volume loop validation of TAPSE/PASP for right ventricular arterial coupling in heart failure with pulmonary hypertension. Eur Heart J Cardiovasc Imaging. 2021;22(2):168–176. doi: 10.1093/ehjci/jeaa285 [DOI] [PubMed] [Google Scholar]
- 14.Hemnes AR, Leopold JA, Radeva MK, et al. Clinical Characteristics and Transplant-Free Survival Across the Spectrum of Pulmonary Vascular Disease. J Am Coll Cardiol. 2022;80(7):697–718. doi: 10.1016/j.jacc.2022.05.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hemnes AR, Beck GJ, Newman JH, et al. PVDOMICS: A Multi-Center Study to Improve Understanding of Pulmonary Vascular Disease Through Phenomics. Circ Res. 2017;121(10):1136–1139. doi: 10.1161/CIRCRESAHA.117.311737 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tang WHW, Wilcox JD, Jacob MS, et al. Comprehensive Diagnostic Evaluation of Cardiovascular Physiology in Patients With Pulmonary Vascular Disease: Insights From the PVDOMICS Program. Circ Heart Fail. 2020;13(3):e006363. doi: 10.1161/CIRCHEARTFAILURE.119.006363 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Borlaug BA, Reddy YNV. The Role of the Pericardium in Heart Failure: Implications for Pathophysiology and Treatment. JACC Heart Fail. 2019;7(7):574–585. doi: 10.1016/j.jchf.2019.03.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Omote K, Sorimachi H, Obokata M, et al. Pulmonary vascular disease in pulmonary hypertension due to left heart disease: pathophysiologic implications. Eur Heart J. 2022;43(36):3417–3431. doi: 10.1093/eurheartj/ehac184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Reddy YNV, Obokata M, Verbrugge FH, Lin G, Borlaug BA. Atrial Dysfunction in Patients With Heart Failure With Preserved Ejection Fraction and Atrial Fibrillation. J Am Coll Cardiol. 2020;76(9):1051–1064. doi: 10.1016/j.jacc.2020.07.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Obokata M, Reddy YNV, Pislaru SV, Melenovsky V, Borlaug BA. Evidence Supporting the Existence of a Distinct Obese Phenotype of Heart Failure With Preserved Ejection Fraction. Circulation. 2017;136(1):6–19. doi: 10.1161/CIRCULATIONAHA.116.026807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Jellis CL, Park MM, Abidov A, et al. Comprehensive echocardiographic evaluation of the right heart in patients with pulmonary vascular diseases: the PVDOMICS experience. Eur Heart J Cardiovasc Imaging. 2022;23(7):958–969. doi: 10.1093/ehjci/jeab065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Reddy YNV, Obokata M, Koepp KE, Egbe AC, Wiley B, Borlaug BA. The β-Adrenergic Agonist Albuterol Improves Pulmonary Vascular Reserve in Heart Failure With Preserved Ejection Fraction. Circ Res. 2019;124(2):306–314. doi: 10.1161/CIRCRESAHA.118.313832 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Reddy YNV, Obokata M, Wiley B, et al. The haemodynamic basis of lung congestion during exercise in heart failure with preserved ejection fraction. Eur Heart J. 2019;40(45):3721–3730. doi: 10.1093/eurheartj/ehz713 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pencina MJ, D’Agostino RB. Evaluating Discrimination of Risk Prediction Models: The C Statistic. JAMA. 2015;314(10):1063–1064. doi: 10.1001/jama.2015.11082 [DOI] [PubMed] [Google Scholar]
- 25.Humbert M, Kovacs G, Hoeper MM, et al. 2022 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension. Eur Heart J. 2022;43(38):3618–3731. doi: 10.1093/eurheartj/ehac237 [DOI] [PubMed] [Google Scholar]
- 26.Waxman A, Restrepo-Jaramillo R, Thenappan T, et al. Inhaled Treprostinil in Pulmonary Hypertension Due to Interstitial Lung Disease. N Engl J Med. 2021;384(4):325–334. doi: 10.1056/NEJMoa2008470 [DOI] [PubMed] [Google Scholar]
- 27.Nathan SD, Deng C, King CS, et al. Inhaled Treprostinil Dosage in Pulmonary Hypertension Associated With Interstitial Lung Disease and Its Effects on Clinical Outcomes. Chest. 2023;163(2):398–406. doi: 10.1016/j.chest.2022.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ghofrani HA, D’Armini AM, Grimminger F, et al. Riociguat for the treatment of chronic thromboembolic pulmonary hypertension. N Engl J Med. 2013;369(4):319–329. doi: 10.1056/NEJMoa1209657 [DOI] [PubMed] [Google Scholar]
- 29.Jaïs X, Brenot P, Bouvaist H, et al. Balloon pulmonary angioplasty versus riociguat for the treatment of inoperable chronic thromboembolic pulmonary hypertension (RACE): a multicentre, phase 3, open-label, randomised controlled trial and ancillary follow-up study. Lancet Respir Med. Published online August 1, 2022:S2213–2600(22)00214–4. doi: 10.1016/S2213-2600(22)00214-4 [DOI] [PubMed] [Google Scholar]
- 30.Kawakami T, Matsubara H, Shinke T, et al. Balloon pulmonary angioplasty versus riociguat in inoperable chronic thromboembolic pulmonary hypertension (MR BPA): an open-label, randomised controlled trial. Lancet Respir Med. Published online August 1, 2022:S2213–2600(22)00171–0. doi: 10.1016/S2213-2600(22)00171-0 [DOI] [PubMed] [Google Scholar]
- 31.Borlaug BA, Reddy YNV, Braun A, et al. Cardiac and Metabolic Effects of Dapagliflozin in Heart Failure With Preserved Ejection Fraction: The CAMEO-DAPA Trial. Circulation. 2023;148(10):834–844. doi: 10.1161/CIRCULATIONAHA.123.065134 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Nassif ME, Qintar M, Windsor SL, et al. Empagliflozin Effects on Pulmonary Artery Pressure in Patients With Heart Failure: Results From the EMBRACE-HF Trial. Circulation. 2021;143(17):1673–1686. doi: 10.1161/CIRCULATIONAHA.120.052503 [DOI] [PubMed] [Google Scholar]
- 33.Reddy YNV, Anantha-Narayanan M, Obokata M, et al. Hemodynamic Effects of Weight Loss in Obesity: A Systematic Review and Meta-Analysis. JACC Heart Fail. 2019;7(8):678–687. doi: 10.1016/j.jchf.2019.04.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Butler J, Shah SJ, Petrie MC, et al. Semaglutide versus placebo in people with obesity-related heart failure with preserved ejection fraction: a pooled analysis of the STEP-HFpEF and STEP-HFpEF DM randomised trials. Lancet. Published online April 4, 2024:S0140–6736(24)00469–0. doi: 10.1016/S0140-6736(24)00469-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Brugts JJ, Radhoe SP, Clephas PRD, et al. Remote haemodynamic monitoring of pulmonary artery pressures in patients with chronic heart failure (MONITOR-HF): a randomised clinical trial. Lancet. 2023;401(10394):2113–2123. doi: 10.1016/S0140-6736(23)00923-6 [DOI] [PubMed] [Google Scholar]
- 36.Reddy YN, Carter RE, Sorimachi H, Omar M, Popovic D, Alogna A, Jensen MD, Borlaug BA. Effect of Dapagliflozin on Right Heart-Pulmonary Vascular Interaction in Heart Failure with preserved Ejection Fraction. JAMA Cardiology. 2024. [In Press]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Ameri P, Mercurio V, Pollesello P, et al. A roadmap for therapeutic discovery in pulmonary hypertension associated with left heart failure. A scientific statement of the Heart Failure Association (HFA) of the ESC and the ESC Working Group on Pulmonary Circulation & Right Ventricular Function. Eur J Heart Fail. 2024;26(4):707–729. doi: 10.1002/ejhf.3236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Badagliacca R, Ghio S, D’Alto M, et al. Relevance of Echocardiography-Derived Phenotyping in Pulmonary Arterial Hypertension Patients Treated with Initial Oral Combination Therapy. Am J Respir Crit Care Med. Published online May 31, 2024. doi: 10.1164/rccm.202402-0431LE [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.