Abstract
Aims
Imatinib mesylate, as add-on therapy in patients with pulmonary arterial hypertension (PAH) who remain inadequately treated despite receiving at least two PAH-specific drugs, improves exercise capacity and haemodynamics. We evaluated whether 24 weeks of add-on therapy with imatinib compared with placebo also improves right ventricular (RV) function assessed by echocardiography.
Methods and results
Echocardiograms were obtained at baseline, 12 weeks, and 24 weeks in 74 patients randomized to imatinib or placebo in the Imatinib in Pulmonary arterial hypertension, a Randomized Efficacy Study (IMPRES) trial. Right ventricular function was assessed by tissue Doppler tricuspid annular peak systolic velocity (TA S′), tricuspid annular plane systolic excursion (TAPSE), RV Tei index, and RV fractional area change. Between-treatment-group differences in the changes from baseline to week-24 were assessed using an ANCOVA with the last observation carried forward. At week-24 patients randomized to imatinib demonstrated greater improvements in TA S′ (1.6 ± 2.3 imatinib vs. 0.5 ± 2.4 cm/s placebo, P = 0.007) and RV Tei index (−0.11 ± 0.18 imatinib vs. 0.05 ± 0.18 placebo, P = 0.005) compared with placebo, but not in TAPSE (0.07 ± 0.44 imatinib vs. 0.03 ± 0.32 cm placebo, P = 0.08). Imatinib therapy was also associated with significant reduction in peak tricuspid regurgitation velocity, increase in LV size, and improvement in LV early diastolic relaxation velocity.
Conclusions
Among patients with advanced PAH who remain symptomatic on at least two PAH-specific drugs, treatment with imatinib compared with placebo is associated with significant improvements in echocardiographic measures of RV function, in addition to LV size and LV early diastolic relaxation.
Clinical trial registration
NCT00902174 (Clinicaltrials.gov).
Keywords: Pulmonary heart disease, Echocardiography, Pulmonary hypertension, Trials
Introduction
Pulmonary arterial hypertension (PAH) is associated with significant morbidity and mortality, particularly among patients with poor functional capacity despite disease-specific therapy.1 Elevated right atrial (RA) pressure, pulmonary vascular resistance (PVR), and low cardiac index (CI) are significant risk factors for increased mortality in these patients.2–4 Right ventricular (RV) failure is a key prognostic determinant in PAH. However, due to its complex geometry, RV function is challenging to assess non-invasively by echocardiography. Limited data are available regarding the impact of pharmacologic interventions known to improve exercise capacity, haemodynamics, and functional status on non-invasive echocardiographic measures of RV function. However, novel M-mode and tissue Doppler measures that do not require geometric assumptions regarding RV shape may offer a simplified and widely applicable method to assess RV function in PAH.5–7
The echocardiography sub-study of the Imatinib in Pulmonary arterial hypertension (PAH), a Randomized Efficacy Study (IMPRES), was an exploratory sub-study designed to evaluate the hypothesis that among adult patients with significant PAH, treatment with imatinib as an add-on drug will improve echocardiographic measures of RV function to a greater extent than placebo. Specifically, we tested whether 24 weeks of therapy with imatinib compared with placebo were associated with greater improvements in four prospectively selected echocardiographic measures of RV function: tissue Doppler tricuspid annular peak systolic velocity (TA S′), tricuspid annular plane systolic excursion (TAPSE), RV Tei index, and RV fractional area change (RVFAC). We further assessed the impact of imatinib therapy on measures of left ventricular (LV) size, LV systolic function, and LV diastolic function.
Methods
Patient population
The IMPRES trial demonstrated significant improvements in exercise capacity and haemodynamics with imatinib mesylate as add-on therapy compared with placebo.8 The IMPRES trial enrolled 202 patients at 71 sites in 14 countries on 3 continents. All patients provided written informed consent, which included consent for the echocardiographic analysis. Enrolled patients were at least 18 years of age, diagnosed with Pulmonary Hypertension Group 1, i.e. PAH, with decreased exercise capacity [defined as 6-min walk distance (6MWD) ≤450 m] and a PVR > 800 dynes s/cm5 despite therapy with two or more PAH-specific medications (endothelin receptor antagonists, phosphodiesterase 5 inhibitors, prostacyclin, or prostacyclin analogues). Patients with a prolonged QTc (>450 ms for males and >470 ms for females), syncope in the prior 3 months, or history of a bleeding disorder were excluded. Patients were randomized 1 : 1 to imatinib or placebo. Participation in the echocardiography sub-study was voluntary and offered to all participating sites. The 27 sites in 8 countries that chose to participate were asked to enroll all patients enrolled in the overall trial at that site (to minimize selection bias). Of 81 patients enrolled in the overall trial at these sites, 74 patients (91%) were also enrolled in the sub-study. For participating sites, echocardiograms were performed by a study-specific protocol at baseline, week-12, and week-24 (Figure 1).
Figure 1.
Disposition of patients in the IMPRES Echocardiographic sub-study.
Echocardiographic methods
All sonographers at participating sites underwent central training in the details of the echocardiographic views and techniques at study investigator meetings. Echocardiograms were sent on digital storage media to the core laboratory at the Brigham and Women's Hospital. All quantitative measures on study echocardiograms were performed by a single cardiologist at the core laboratory (P.C.) and independently confirmed by a staff echocardiographer (A.M.S.) with both readers blinded to clinical information and randomized treatment assignment. All measures were performed in triplicate. In accordance with American Society of Echocardiography (ASE) guidelines,9 TA S′ was measured using tissue Doppler and TAPSE was measured using M-mode from the lateral tricuspid annulus, RV Tei index was derived from tricuspid regurgitation duration and RV ejection time, and RVFAC was calculated as the percent change in RV cavity area from end-diastole to end-systole in the apical four-chamber view. Peak RV-to-RA systolic gradient was calculated as 4·[peak tricuspid regurgitation (TR) velocity].2 Left ventricular size was expressed as LV chamber area measured in the apical four-chamber view and as LV dimension measured from the parasternal long-axis view. Left ventricular volumes were derived according to the modified biplane Simpson's rule and used to calculate LV ejection fraction (EF) in the standard fashion.10 Systolic function was also quantified by septal and lateral mitral annular peak systolic velocity. Left ventricular diastolic function was assessed using the mitral annular peak early diastolic velocity (E′), measured from both the lateral and septal aspects of the mitral annulus. Mitral inflow velocity was assessed by pulsed wave Doppler from the apical four-chamber view in standard fashion.
Intra-observer variability for measures of RV function was assessed in a subset of 20 studies, with the following results: TA S′: coefficient of variation 4.0%, bias 0.04 ± 0.42 m/s; TAPSE: coefficient of variation 1.6%, bias 0.2 ± 2.6 mm; RV Tei index: coefficient of variation 8.6%, bias 0.02 ± 0.06; RVFAC: coefficient of variation 17.4%, bias 1 ± 4%.
Statistical analysis
Continuous variables are presented as means and standard deviations unless otherwise specified. Comparison of baseline clinical and echocardiographic measures between randomized treatment arms was performed using a Fisher's exact test for categorical variables and a t-test for continuous variables. When continuous variables were not normally distributed, comparison was made using the Wilcoxon rank sum test. Two-sided P-values of <0.05 were considered significant. Sample size was allowed to float. Between-treatment-group differences in the changes in echocardiographic measures from baseline to week-24 were assessed using an ANCOVA with adjustment for the baseline measure using the last observation carried forward (LOCF). Between-treatment-group differences in the change from baseline to week-12, week-12 to week-24, and baseline to week-24 were also assessed using an ANCOVA with adjustment for baseline. In a sensitivity analysis, we repeated the analysis using a mixed-effects repeated-measures model which included terms for randomized treatment, visit, treatment-by-visit interaction, and baseline value of the measure of interest as covariates (Supplementary material online, Table S1). Where the treatment-by-visit interaction term was not statistically significant, this term was dropped and the treatment term effect estimate and P-value reported. Changes from baseline to week-12 and week-24 were also assessed using the mixed-effects repeated-measures model. Correlation between baseline echocardiographic measures of RV function and baseline haemodynamics, biomarkers, and 6MWD, and between changes in echocardiographic measures and changes in haemodynamics (as assessed by right heart catheterization), biomarkers, and 6MWD at weeks 12 and 24 were assessed by Spearman correlation. All analyses were performed using SAS version 9.2 and Stata version 11.2.
Results
Of 202 patients enrolled in the IMPRES trial, echocardiograms were available in 74 at baseline (37%; 35 placebo, 39 imatinib). Of these patients with baseline echocardiograms, follow-up studies were available in 62 at 12 weeks (84%; 33 placebo, 29 imatinib), and 59 at 24 weeks (80%; 28 placebo, 31 imatinib) (Figure 1). Compared with IMPRES patients who did not participate in the echocardiography sub-study, patients participating in the echo sub-study were more frequently white (87 vs. 66%, P = 0.002) and more frequently WHO functional class II vs. class III (class II: 35 vs. 19% respectively, class III: 57 vs. 75% respectively; P = 0.03). No significant differences were noted in age, gender, baseline 6MWD, or baseline haemodynamics. Compared with the 59 patients (80% of the 74 enrolled in the sub-study) with baseline echocardiograms who also had week-24 echocardiograms, the 15 patients (20%) who did not were more likely to have prematurely discontinued IMPRES due to an adverse event or a clinical worsening event (2 vs. 27%, P = 0.005); additionally, numerically more deaths (non-significant; 2 vs. 13%, P = 0.10) and PAH-related hospitalizations (non-significant; 10 vs. 27%, P = 0.11) were reported for these 15 patients. Nevertheless, the number of patients without week-24 echocardiographic data was evenly distributed between the placebo and imatinib treatment arms [7 (20%) vs. 8 (21%), respectively, P = 0.96].
Among patients participating in the echocardiography study, the average age was 50 years old and the majority of patients were female, white, and WHO functional class III. Haemodynamic measures at RHC were consistent with advanced PAH with severely elevated PVR and PAP, elevated RAP, reduced CI, and normal PCWP. No significant differences were observed in baseline clinical characteristics by treatment group (Table 1). A lower baseline PVR was observed in the imatinib group (P = 0.04). At baseline, patients randomized to imatinib had higher TA S′, higher TAPSE, higher RV outflow tract time velocity integral (RVOT VTI), larger LV end-diastolic area, and higher mitral annular septal E′ (Table 2). No other differences in baseline echocardiographic measures of RV or LV size or function were observed between treatment groups. Using LOCF, at week-24, patients randomized to imatinib demonstrated greater improvement in RV functional parameters of TA S′ and RV Tei index compared with patients randomized to placebo (Table 3). Imatinib therapy was also associated with reduction in TR jet velocity, increase in RVOT VTI, and decrease in the TR jet velocity/RVOT VTI ratio. No significant between-group differences were apparent in change in right heart size. No between-group differences in resolution of pericardial effusion (P = 0.61) or of RVOT Doppler notching (P = 0.12) was noted. Imatinib treatment was associated with an increase in LV end-diastolic dimension. No change in LV systolic function was noted between treatment groups, although imatinib was associated with a greater improvement in septal S′. Patients randomized to imatinib also demonstrated greater improvement in LV early diastolic relaxation, reflected in both septal E′ and lateral E′. The results of a sensitivity analysis using a mixed-effects repeated-measures model were concordant with the results of the LOCF analysis (Supplementary material online, Table S1). Similar findings were observed when analysis was restricted to only patients with measurements at baseline and week-24 (Supplementary material online, Table S2) or with patients with measurements at all three study time points (baseline, week-12, and week-24; Supplementary material online, Table S3).
Table 1.
Baseline clinical and haemodynamic measures in the echocardiographic sub-study by treatment arm
| Placebo | Imatinib | P-value | |
|---|---|---|---|
| Age (years) | 48.5 ± 13.0 | 50.9 ± 13.7 | 0.44 |
| Female, n (%) | 31 (89%) | 30 (77%) | 0.23 |
| White, n (%) | 29 (83%) | 35 (90%) | 0.50 |
| WHO functional class, n (%) | 0.77 | ||
| II | 13 (37%) | 14 (36%) | |
| III | 20 (57%) | 21 (54%) | |
| IV | 2 (6%) | 4 (10%) | |
| 6-min walk distance (meters) | 340 ± 74 | 369 ± 71 | 0.09 |
| Body mass index | 25.5 ± 6.0 | 25.3 ± 5.5 | 0.88 |
| Mean right atrial pressure (mmHg) | 9 (6–12) | 10 (6–13) | 0.47a |
| Mean pulmonary artery pressure (mmHg) | 61 (52–71) | 60 (51–65) | 0.35a |
| Mean pulmonary capillary wedge pressure (mmHg) | 11 (6–13) | 10 (7–13) | 0.45a |
| Cardiac index (L/min/m2) | 2.0 (1.9–2.2) | 2.1 (1.9–2.5) | 0.30a |
| Pulmonary vascular resistance (dynes s/cm5) | 1200 (1016–1371) | 1010 (873–1304) | 0.04a |
| NT-proBNP (pg/mL) | 114 (50–199) | 81 (30–242) | 0.87a |
aComparison made using the Wilcoxon rank sum test.
Table 2.
Baseline echocardiographic measures by the treatment group
| Placebo |
Imatinib |
P-value | |||
|---|---|---|---|---|---|
| n | n | ||||
| RV function | |||||
| TA S′ (cm/s) | 35 | 10.08 ± 2.22 | 37 | 11.10 ± 2.60 | 0.08 |
| TAPSE (cm) | 29 | 1.60 ± 0.34 | 33 | 1.88 ± 0.37 | 0.003 |
| Tei index | 32 | 0.698 ± 0.217 | 34 | 0.713 ± 0.196 | 0.78 |
| RVFAC (%) | 33 | 23.5 ± 7.6 | 38 | 22.0 ± 7.8 | 0.44 |
| RV Doppler | |||||
| Peak TR velocity (m/s) | 32 | 4.61 ± 0.67 | 36 | 4.47 ± 0.46 | 0.32 |
| RVOT VTI (cm) | 32 | 16.5 ± 4.1 | 31 | 14.0 ± 4.1 | 0.02 |
| TR velocity/RVOT VTI ratio | 29 | 0.29 ± 0.08 | 28 | 0.36 ± 0.14 | 0.02 |
| Systolic notching | 31 | 24 (77%) | 34 | 22 (65%) | 0.29 |
| Right heart size | |||||
| RVEDA (cm2) | 33 | 35.1 ± 11.1 | 38 | 36.3 ± 11.4 | 0.66 |
| RA area (cm2) | 35 | 24.8 ± 7.1 | 39 | 25.5 ± 10.3 | 0.71 |
| RV/LV area ratio | 29 | 1.42 ± 0.38 | 36 | 1.51 ± 0.57 | 0.49 |
| Pericardial effusion | 34 | 20 (59%) | 35 | 20 (57%) | 0.89 |
| LV structure | |||||
| LVEDA–A4C (cm2) | 31 | 21.7 ± 3.3 | 35 | 24.0 ± 4.8 | 0.02 |
| LVEDD (cm) | 35 | 3.62 ± 0.58 | 38 | 3.70 ± 0.48 | 0.51 |
| Septal wall thickness (cm) | 35 | 1.00 ± 0.19 | 38 | 1.05 ± 0.23 | 0.30 |
| Posterior wall thickness (cm) | 35 | 1.01 ± 0.20 | 38 | 1.04 ± 0.18 | 0.50 |
| LV systolic function | |||||
| LVEF (%) | 31 | 58.4 ± 5.3 | 36 | 58.0 ± 6.6 | 0.80 |
| S′ lateral (cm/s) | 35 | 9.6 ± 2.4 | 33 | 9.2 ± 2.6 | 0.54 |
| S′ septal (cm/s) | 35 | 7.1 ± 1.5 | 35 | 7.9 ± 2.7 | 0.12 |
| LV diastolic function | |||||
| E′ lateral (cm/s) | 35 | 10.8 ± 3.1 | 34 | 10.5 ± 3.0 | 0.69 |
| E′ septal (cm/s) | 35 | 5.3 ± 2.0 | 36 | 6.4 ± 2.7 | 0.05 |
| E/E′ ratio | 33 | 6.9 ± 3.5 | 33 | 7.0 ± 2.4 | 0.85 |
Table 3.
Change in echocardiographic measures from baseline to 24 weeks by treatment arm using last observation carried forward
| Placebo |
Imatinib |
P for change | |||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Baseline | 24 weeks | Change | n | Baseline | 24 weeks | Change | ||
| Right ventricle function | |||||||||
| TA S′ (cm/s) | 33 | 10.2 ± 2.2 | 10.7 ± 2.5 | 0.5 ± 2.4 | 32 | 11.1 ± 2.7 | 12.7 ± 2.7 | 1.6 ± 2.3 | 0.007 |
| TAPSE (cm) | 28 | 1.59 ± 0.34 | 1.62 ± 0.41 | 0.03 ± 0.32 | 30 | 1.89 ± 0.36 | 1.96 ± 0.39 | 0.07 ± 0.44 | 0.08 |
| Tei index | 26 | 0.67 ± 0.19 | 0.72 ± 0.22 | 0.05 ± 0.18 | 25 | 0.74 ± 0.19 | 0.63 ± 0.24 | −0.11 ± 0.18 | 0.005 |
| RVFAC (%) | 31 | 23.8 ± 7.7 | 23.0 ± 7.9 | −0.8 ± 7.5 | 33 | 22.3 ± 7.6 | 24.5 ± 8.0 | 2.2 ± 9.0 | 0.23 |
| RV Doppler | |||||||||
| Peak TR velocity (m/s) | 29 | 4.64 ± 0.68 | 4.74 ± 0.70 | 0.09 ± 0.58 | 31 | 4.44 ± 0.45 | 4.27 ± 0.53 | −0.16 ± 0.57 | 0.02 |
| RVOT VTI (cm) | 29 | 16.3 ± 4.2 | 15.5 ± 5.0 | −0.8 ± 4.4 | 24 | 14.4 ± 4.3 | 17.5 ± 4.3 | 3.1 ± 3.4 | 0.004 |
| TR velocity/RVOT VTI ratio | 24 | 0.30 ± 0.08 | 0.35 ± 0.14 | 0.05 ± 0.11 | 20 | 0.35 ± 0.14 | 0.25 ± 0.08 | −0.10 ± 0.09 | <0.0001 |
| Right heart size | |||||||||
| RVEDA (cm2) | 31 | 33.5 ± 6.9 | 32.3 ± 8.4 | −1.2 ± 4.8 | 33 | 35.4 ± 10.4 | 34.6 ± 13.2 | −0.7 ± 6.3 | 0.83 |
| RA area (cm2) | 33 | 24.2 ± 5.8 | 24.6 ± 7.8 | 0.4 ± 4.4 | 34 | 24.7 ± 8.8 | 23.0 ± 9.0 | −1.7 ± 4.9 | 0.07 |
| LV/RV area ratio | 24 | 0.68 ± 0.16 | 0.76 ± 0.22 | 0.07 ± 0.13 | 26 | 0.76 ± 0.28 | 0.89 ± 0.39 | 0.13 ± 0.19 | 0.25 |
| LV structure | |||||||||
| LVEDA–A4C (cm2) | 26 | 21.9 ± 3.4 | 22.3 ± 3.8 | 0.5 ± 3.4 | 27 | 24.1 ± 4.6 | 26.1 ± 5.8 | 2.0 ± 4.2 | 0.07 |
| LVEDD (cm) | 33 | 3.58 ± 0.57 | 3.40 ± 0.63 | −0.18 ± 0.44 | 32 | 3.70 ± 0.45 | 3.88 ± 0.43 | 0.18 ± 0.34 | 0.0001 |
| IVS (cm) | 33 | 0.99 ± 0.20 | 1.06 ± 0.23 | 0.07 ± 0.21 | 32 | 1.05 ± 0.22 | 1.14 ± 0.24 | 0.10 ± 0.19 | 0.33 |
| PW (cm) | 33 | 1.00 ± 0.21 | 1.06 ± 0.20 | 0.06 ± 0.22 | 32 | 1.03 ± 0.15 | 1.08 ± 0.16 | 0.05 ± 0.15 | 0.88 |
| Systolic function | |||||||||
| LVEF (%) | 28 | 58.2 ± 5.2 | 61.0 ± 7.4 | 2.3 ± 7.7 | 29 | 58.5 ± 7.0 | 61.0 ± 7.2 | 2.6 ± 7.7 | 0.83 |
| S′ lateral (cm/s) | 33 | 9.9 ± 2.5 | 9.6 ± 2.6 | −0.4 ± 2.1 | 26 | 9.0 ± 2.8 | 9.7 ± 1.8 | 0.6 ± 3.2 | 0.46 |
| S′ septal (cm/s) | 34 | 7.19 ± 1.53 | 7.03 ± 1.42 | −0.16 ± 1.55 | 27 | 8.03 ± 3.00 | 9.19 ± 2.32 | 1.16 ± 2.67 | <0.0001 |
| Diastolic function | |||||||||
| E′ lateral (cm/s) | 32 | 10.9 ± 3.2 | 10.0 ± 3.5 | −0.9 ± 2.4 | 29 | 10.3 ± 3.0 | 11.3 ± 3.9 | 1.0 ± 3.2 | 0.02 |
| E′ septal (cm/s) | 33 | 5.4 ± 2.0 | 5.3 ± 1.8 | −0.1 ± 1.9 | 30 | 6.4 ± 2.9 | 7.1 ± 2.5 | 0.7 ± 3.0 | 0.006 |
| E/E′ ratio | 29 | 6.7 ± 3.6 | 7.4 ± 3.6 | 0.7 ± 3.4 | 27 | 7.0 ± 2.4 | 7.1 ± 2.0 | 0.1 ± 1.9 | 0.50 |
P-values for between-group difference in changes were assessed using ANCOVA.
For both TA S′ and RV Tei index, significant improvements were apparent from baseline to week-12 with imatinib compared with placebo (Figure 2). The between-group differences in change in both measures from week-12 to week-24 were minimal, suggesting that the majority of the improvement seen at week-24 with imatinib was present by 12 weeks of treatment. Qualitatively similar findings were observed when the analysis was restricted to patients with measurements available at all three time points (Supplementary material online, Table S4).
Figure 2.
TA S′ (A) and RV Tei index (B) at baseline, 12 weeks, and 24 weeks by treatment assignment (imatinib vs. placebo). P-value is based on ANCOVA with last value carried forward.
As observed in the overall IMPRES trial, within the echo sub-study imatinib was associated with reduction in PVR and mean pulmonary artery pressure (mPAP), and increase in cardiac index (CI). At baseline, TA S′ was correlated with both invasive PVR (Spearman correlation coefficient −0.30, P = 0.009) and the PVR/SVR ratio (Spearman correlation coefficient −0.45, P = 0.0002). Baseline RV Tei index was not correlated with baseline invasive haemodynamics. The increase in TA S′ from baseline to week-24 correlated with change in mPAP (Spearman correlation coefficient −0.33, P = 0.02) (Figure 3). The change in RV Tei index over the same period also correlated with change in mPAP (Spearman correlation coefficient 0.35, P = 0.03), cardiac index (Spearman correlation coefficient −0.46, P = 0.002), and PVR (Spearman correlation coefficient 0.55, P = 0.0003).
Figure 3.
Scatter plots demonstrating the change in invasively measured mPAP (A, D), CI (B, E), and PVR (C, F) by change in TA S′ (A–C in blue) and by change in RV Tei index (D–F in red) from baseline to 24 weeks. Spearman correlation coefficient and associated P-values are given for the relationship between the invasive parameter and echocardiographic measure as continuous variables.
No difference in median baseline NT-proBNP was noted between treatment groups (Table 4). During follow-up, the change in NT-proBNP was correlated with the change in RV Tei index at week-12 (Spearman correlation 0.33, P = 0.04; n = 38) and at week-24 (Spearman correlation 0.43, P = 0.007; n = 39), but not with the other RV functional measures evaluated in the echo sub-study. As observed in the overall IMPRES trial, in the echo sub-study imatinib was associated with improvements in 6MWD at week-24 (Table 4). None of the RV functional measures correlated with 6MWD at baseline or with the change in 6MWD at week-24.
Table 4.
Change in haemodynamics parameters (obtained by right heart catheterization), NT-proBNP, and 6-min walk distance by treatment arm
| Placebo |
Imatinib |
P for change | |||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Baseline | 24 weeks | Change | n | Baseline | 24 weeks | Change | ||
| Invasive haemodynamics | |||||||||
| mRAP (mmHg) | 32 | 8.5 (5.5,12.0) | 9.0 (6.5, 12.0) | 1.0 (−2.0, 4.0) | 28 | 8.5 (6.0, 12.0) | 7.5 (5.0, 11.5) | −1.0 (−3.5, 1.0) | 0.06a |
| mPAP (mmHg) | 33 | 59 (51, 71) | 58 (53, 68) | −1 (−6, 4) | 29 | 54 (50, 64) | 48 (43, 52) | −7 (−13, 0) | 0.02a |
| mPCWP (mmHg) | 31 | 10 (6, 13) | 9 (6, 12) | 0 (−2, 1) | 28 | 9.0 (6.5, 11.0) | 9.5 (7.0, 13.0) | 1 (−2, 4) | 0.19a |
| CI (L/min/m2) | 32 | 2.01 (1.89, 2.24) | 2.03 (1.75, 2.43) | 0.01 (−0.27, 0.23) | 29 | 2.08 (1.81, 2.35) | 2.58 (2.12, 3.10) | 0.52 (0.18, 0.99) | 0.0004a |
| PVR (dynes s/cm5) | 31 | 1161 (1006, 1350) | 1150 (887, 1446) | −11 (−223, 217) | 28 | 1008 (873, 1324) | 732 (588, 912) | −333 (−481, −229) | 0.0004a |
| Biomarkers | |||||||||
| NT-proBNP (pg/mL) | 35 | 106 (48, 195) | 111 (48, 196) | 6 (−9, 53) | 30 | 69 (30, 215) | 54 (22, 103) | −10 (−50, 4) | 0.14b,c |
| Exercise capacity | |||||||||
| 6MWD (m) | 36 | 343 ± 74 | 340 ± 103 | −3 ± 65 | 32 | 370 ± 75 | 400 ± 84 | 30 ± 37 | 0.02b |
aMeasurement was only performed at baseline and 24 weeks; comparison performed using non-parametric Wilcoxon rank sum test.
bAnalysis was performed using ANCOVA.
cAnalysis performed on log-transformed data.
Discussion
In this echocardiographic sub-study of a multicentre 24-week clinical trial of imatinib as add-on therapy vs. placebo in patients with advanced PAH who remained symptomatic on at least two PAH-specific drugs, imatinib was associated with improvements in RV functional measures of TA S′ and RV Tei index. Concordant with the invasive haemodynamic parameters, imatinib was associated with reduction in peak TR velocity, suggesting decreased pulmonary artery systolic pressure, an increase in the RVOT VTI, a measure related to RV stroke volume, and a decrease in the peak TR velocity to RVOT VTI ratio, a measure that has been correlated with PVR.11,12 Supportive of improved RV function and decreased RV afterload, imatinib was associated with increased LV size, reflected in an increase in LV end-diastolic dimension, and an increased LV early diastolic relaxation velocities. These changes in measures of both RV function and LV size and diastolic function were observed at the week-12 assessment, and persisted without significant change at week-24.
Right ventricular failure is associated with increased risk of mortality in PAH.1,3 Right ventricular function may continue to decline despite effective reduction in PVR in a subset of PAH patients, and these patients experience poor prognosis despite reduced PVR.13 The complex geometry of the RV, which is further distorted by hypertrophy and enlargement with progressive RV pressure overload, makes non-invasive assessment of RV function by echocardiography challenging. Indeed, RVFAC, an established risk factor for adverse outcomes in patients with LV dysfunction,14,15 has not been consistently related to risk for adverse outcomes in PAH.16–18 As in our study, prior clinical trials of pharmacotherapy known to improve exercise capacity and haemodynamics in PAH have not reliably detected associated improvements in RVFAC.5,6 Instead, indirect markers of RV overload—such as degree of septal flattening quantified by the LV eccentricity index and the presence of a pericardial effusion—have been more consistently associated with increased risk of death.16–19
Given the prominent contribution of longitudinal shortening to RV ejection, measures of RV systolic longitudinal displacement (TAPSE) and rate of displacement (TA S′) offer simplified measures of RV systolic function that do not rely on geometric assumptions of RV shape.20 TA S′ has been shown to correlate significantly with RV ejection fraction by MRI21–23 and with mPAP and PVR24 in heart failure and in chronic thromboemblic pulmonary embolism. To our knowledge, our study is the first to assess the impact of a PAH-specific drug shown to improve haemodynamics on these novel parameters of RV function. We observed significant improvement in TA S′, in addition to septal S′, with imatinib therapy. Baseline TA S′ correlated with baseline mPAP and PVR, and the improvement in TA S′ significantly correlated with the decrease in mPAP measured invasively.
We did not observe a significant improvement in TAPSE (P = 0.08). While reduced TAPSE is associated with increased risk of death in PAH,16,18,25 TAPSE has not been assessed in interventional studies in pulmonary hypertension.5–7 Therefore, the responsiveness of TAPSE to change in PVR and RV systolic function is not well established. The angle dependency of TAPSE in addition to the relatively small degrees of excursion being measured may limit its sensitivity. In addition, TAPSE may be significantly affected by LV apical motion, such that an undefiled and hyperdynamic LV may increase cardiac rocking and the TAPSE value.26 Relief of pulmonary vascular obstruction with improved LV filling may paradoxically result in lower TAPSE due to less hyperdynamic LV function and diminished rocking.
The RV Tei index is a Doppler-based measure of combined RV systolic and diastolic performance. Multiple studies have demonstrated a relationship between RV Tei index and risk of death in PAH18,27,28 and improvement in RV Tei index with therapies that improve haemodynamics.6,7 In our study, imatinib was similarly associated with significant improvements in RV Tei index. Possibly because it integrates information regarding both RV systolic and diastolic function, the change in RV Tei index demonstrated the strongest correlation with the changes in invasive haemodynamic parameters (PVR, CI) and NT-proBNP among the echocardiographic measures of RV function assessed.
The observed increase in LV size with imatinib likely reflects improved LV filling as a result of reduced PVR and improved RV function. Interestingly, we also observed significant improvement in LV early diastolic relaxation velocity with imatinib. This may be a result of increased LV filling. Alternatively, PAH has been associated with impairments in LV diastolic function, possibly as a result of the markedly enlarged RV adversely affecting ventricular coupling.29–31 However, to our knowledge, the concurrent improvements in haemodynamics, RV function, and LV diastolic function have not been previously described, particularly with the use of relatively novel parameters of diastolic function such as TDI mitral annular E′.
We are unable to quantify the relative contribution of afterload reduction vs. improved ‘intrinsic’ chamber function to the observed improvements in structural and functional measures associated with imatinib. Changes in TA S′ and RV Tei index correlated with changes in haemodynamic measures, suggesting a contribution of reduced afterload to the improvements noted in these measures. However, the correlations of RV functional measures with invasive haemodynamics were generally modest in magnitude, suggesting that alterations in load alone cannot fully account for the observed changes in RV functional measures. Other novel measures of RV function that may be less load-dependent, such as strain or strain rate, were not prospectively evaluated in this study as methods for their measurement were not well established at the time of study design.
This study has several limitations. Only a subset of patients enrolled in the overall trial participated in the echocardiography sub-study, potentially limiting the generalizability of our findings. However, among participating sites in the echo sub-study, 91% of patients enrolled in the overall trial were also enrolled in the sub-study, suggesting minimal selection bias at participating sites. In addition, patients in the sub-study were comparable with those not in the sub-study with regards to baseline 6MWD and haemodynamics, although a higher proportion of patients in the echo sub-study were white and WHO functional class II vs. class III. Fifteen of the 74 (20%) patients in the sub-study had missing echocardiographic data at week-24, although the numbers of patients without any follow-up echo data were evenly distributed between treatment arms. We addressed missing data by using a LOCF analysis, which is often considered to be more conservative than an observed case analysis.32,33 The validity of our approach is supported by two additional findings. First, the majority of the changes observed in both TA S′ and RV Tei index between treatment arms occurred by week-12 with minimal changes between week-12 and week-24. Second, performing the analysis using an observed case analysis (i.e. only individuals with values available at both baseline and week-24) produced similar results. Patients in the sub-study randomized to imatinib demonstrated higher TA S′, TAPSE, RVOT VTI, septal E′, and larger LV end-diastolic area, suggesting better RV function at baseline in the imatinib group. To mitigate the impact of these differences in baseline values, we employed an ANCOVA which adjusts for the baseline value in determining the significance of the between-group differences in the changes in echo measures.
Despite these limitations, this remains a sizeable echocardiographic study in PAH. All studies were performed by a uniform pre-specified echo protocol and analysed by a single blinded cardiologist at a core laboratory with good intra-observer repeatability. To our knowledge, this is the largest study to assess the impact of this novel agent in PAH on RV function, and the only randomized study in PAH to investigate the impact of a pharmacologic intervention on TA S′, TAPSE, and LV relaxation patterns using tissue Doppler imaging. These findings suggest that echocardiography can be effectively utilized to detect treatment-associated improvement in RV function in PAH trials. Additional studies, assessing the prognostic implications of changes in RV measures with intervention, are needed to determine the incremental value of these measures in therapeutic trials and clinical practice.
Conclusions
Among patients with advanced PAH who remained symptomatic despite treatment with at least two PAH-specific drugs, treatment with imatinib as an add-on drug compared with placebo is associated with significant improvements in echocardiographic Doppler and tissue Doppler measures of RV function, with associated improvements in LV size and LV early diastolic relaxation. These findings suggest that imatinib may be effective at attenuating the haemodynamic and the cardiac structural and functional changes associated with PAH. Further study is warranted to assess whether the observed changes in these non-invasive measures of RV function are prognostic for outcomes in PAH.
Supplementary material
Supplementary material is available at European Heart Journal online.
Funding
The IMPRES trials was sponsored by Novartis.
Conflict of interest: A.M.S. and S.D.S. have received research support from and have consulted for Novartis. R.J.B. and A.P. have served on advisory boards and steering committees for Novartis. D.Q. is an employee of Novartis.
References
- 1.Benza RL, Miller DP, Gomberg-Maitland M, Frantz RP, Foreman AJ, Coffey CS, Frost A, Barst RJ, Badesch DB, Elliott CG, Liou TG, McGoon MD. Predicting survival in pulmonary arterial hypertension: insights from the registry to evaluate early and long-term pulmonary arterial hypertension disease management (REVEAL) Circulation. 2010;122:164–172. doi: 10.1161/CIRCULATIONAHA.109.898122. [DOI] [PubMed] [Google Scholar]
- 2.Kanemoto N. Natural history of pulmonary hemodynamics in primary pulmonary hypertension. Am Heart J. 1987;114:407–413. doi: 10.1016/0002-8703(87)90510-2. [DOI] [PubMed] [Google Scholar]
- 3.D'Alonzo GE, Barst RJ, Ayres AM, Bergofsky EH, Brundage BH, Detre KM, Fishman AP, Goldring RM, Groves BM, Kernis JT, Levy PS, Pietra GG, Reid LM, Reeves JT, Rich S, Vreim CE, Williams GW, Wu M. Survival in patients with primary pulmonary hypertension. Ann Intern Med. 1991;115:343–349. doi: 10.7326/0003-4819-115-5-343. [DOI] [PubMed] [Google Scholar]
- 4.Eysmann SB, Palevsky HI, Reichek N, Hackney K, Douglas PS. Two-dimensional and Doppler-echocardiographic and cardiac catheterization correlates of survival in primary pulmonary hypertension. Circulation. 1989;80:353–360. doi: 10.1161/01.cir.80.2.353. [DOI] [PubMed] [Google Scholar]
- 5.Hinderliter AL, Willis PW, Barst RJ, Rich S, Rubin LJ, Badesch DB, Groves BM, McGoon MD, Tapson VF, Bourge RC, Brundage BH, Koerner SK, Langleben D, Keller CA, Murali S, Uretsky BF, Koch G, Li S, Clayton LM, Jobsis MM, Blackburn SD, Crow JW, Long WA. Effects of long-term infusion of prostacyclin (Epoprostenol) on echocardiographic measures of right ventricular structure and function in primary pulmonary hypertension. Circulation. 1997;95:1479–1486. doi: 10.1161/01.cir.95.6.1479. [DOI] [PubMed] [Google Scholar]
- 6.Galie N, Hinderliter AL, Torbicki A, Fourme T, Simonneau G, Pulido T, Espinola-Zavaleta N, Rocchi G, Manes A, Frantz R, Kurzyna M, Nagueh SF, Barst R, Channick R, Dujardin K, Kronenberg A, Leconte I, Rainisio M, Rubin L. Effects of the oral endothelin-receptor antagonist Bosentan on echocardiographic and Doppler measures in patients with pulmonary hypertension. J Am Coll Cardiol. 2003;41:1380–1386. doi: 10.1016/s0735-1097(03)00121-9. [DOI] [PubMed] [Google Scholar]
- 7.Seyfarth HJ, Pankau H, Hammerschmidt S, Schauer J, Wirtz H, Winkler J. Bosentan improves exercise tolerance and Tei index in patients with pulmonary hypertension and prostanoid therapy. Chest. 2005;128:709–713. doi: 10.1378/chest.128.2.709. [DOI] [PubMed] [Google Scholar]
- 8.Hoeper MM, Barst RJ, Bourge RC, Feldman J, Frost AE, Galie N, Gomez-Sanchez MA, Grimminger F, Grunig E, Hassoun PM, Morrell NW, Peacock AJ, Satoh T, Simonneau G, Tapson VF, Torres F, Lawrence D, Quinn D, Ghofrani HA. Imatinib mesylate as add-on therapy for pulmonary arterial hypertension: Results of the randomized IMPRES study. Circulation. 2013 doi: 10.1161/CIRCULATIONAHA.112.000765. (epub 12 February 2013). [DOI] [PubMed] [Google Scholar]
- 9.Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, Solomon SD, Louie EK, Schiller NB. Guidelines for the echocardiographic assessment of the right heart in adults: A report from the American Society of Echocardiography. J Am Soc Echocardiogr. 2010;23:685–713. doi: 10.1016/j.echo.2010.05.010. [DOI] [PubMed] [Google Scholar]
- 10.Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, Picard MH, Roman MJ, Seward J, Shanewise JS, Solomon SD, Spencer KT, Sutton MS, Stewart WJ Chamber Quantification Writing Group; American Society of Echocardiography's Guidelines and Standards Committee; European Association of Echocardiography. Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr. 2005;18:1440–1463. doi: 10.1016/j.echo.2005.10.005. [DOI] [PubMed] [Google Scholar]
- 11.Abbas AE, Fortuin D, Schiler NB, Appleton CP, Moreno CA, Lester SJ. A simple method for noninvasive estimation of pulmonary vascular resistance. J Am Coll Cardiol. 2003;41:1021–1027. doi: 10.1016/s0735-1097(02)02973-x. [DOI] [PubMed] [Google Scholar]
- 12.Kouzu H, Nakatani S, Kyotani S, Kanzaki H, Nakanishi N, Kitakaze M. Noninvasive estimation of pulmonary vascular resistance by Doppler echocardiography in patients with pulmonary arterial hypertension. Am J Cardiol. 2009;103:872–876. doi: 10.1016/j.amjcard.2008.11.039. [DOI] [PubMed] [Google Scholar]
- 13.Van de Veerdonk MC, Kind T, Marcus JT, Mauritz GJ, Heymans MW, Bogaard HJ, Boonstra A, Marques KM, Westerhof N, Vonk-Noordegraaf A. Progressive right ventricular dysfunction in patients with pulmonary arterial hypertension responding to therapy. J Am Coll Cardiol. 2011;58:2511–2519. doi: 10.1016/j.jacc.2011.06.068. [DOI] [PubMed] [Google Scholar]
- 14.Zornoff LAM, Skali H, Pfeffer MA, St John Sutton M, Rouleau JL, Lamas GA, Plappert T, Rouleau JR, Moye LA, Lewis SJ, Braunwald E, Solomon SD. Right ventricular dysfunction and risk of heart failure and mortality after myocardial infarction. J Am Coll Cardiol. 2002;39:1450–1455. doi: 10.1016/s0735-1097(02)01804-1. [DOI] [PubMed] [Google Scholar]
- 15.Anavekar NS, Skali H, Bourgoun M, Ghali JK, Kober L, Maggioni AP, McMurray JJ, Velazquez E, Califf R, Pfeffer MA, Solomon SD. Usefulness of right ventricular fractional area change to predict death, heart failure, and stroke following myocardial infarction (from the VALIANT ECHO Study) Am J Cardiol. 2008;101:607–612. doi: 10.1016/j.amjcard.2007.09.115. [DOI] [PubMed] [Google Scholar]
- 16.Ghio S, Klersy C, Magrini G, D'Armini AM, Scelsi L, Raineri C, Pasotti M, Serio A, Campana C, Vigano M. Prognostic relevance of the echocardiographic assessment of right ventricular function in patients with idiopathic pulmonary arterial hypertension. Int J Cardiol. 2010;140:272–278. doi: 10.1016/j.ijcard.2008.11.051. [DOI] [PubMed] [Google Scholar]
- 17.Raymond RJ, Hinderliter AL, Willis PW, Ralph D, Caldwell EJ, Williams W, Ettinger NA, Hill NS, Summer WR, de Boisblanc B, Schwartz T, Koch G, Clayton LM, Jobsis MM, Crow JW, Long W. Echocardiographic predictors of adverse outcomes in primary pulmonary hypertension. J Am Coll Cardiol. 2002;39:1214–1219. doi: 10.1016/s0735-1097(02)01744-8. [DOI] [PubMed] [Google Scholar]
- 18.Brierre G, Blot-Souletie N, Degano B, Tetu L, Bongard V, Carrie D. New echocardiographic prognostic factors for mortality in pulmonary arterial hypertension. Eur J Echocardiogr. 2010;11:516–522. doi: 10.1093/ejechocard/jeq011. [DOI] [PubMed] [Google Scholar]
- 19.Hinderliter AL, Willis PW, Long W, Clarke WR, Ralph D, Caldwell EJ, William W, Ettinger NA, Hill NS, Summer WR, de Boisblanc B, Koch G, Li S, Clayton LM, Jobsis MM, Crow JW. Frequency and prognostic significance of pericardial effusion in primary pulmonary hypertension. Am J Cardiol. 1999;84:481–484. doi: 10.1016/s0002-9149(99)00342-2. [DOI] [PubMed] [Google Scholar]
- 20.Horton KD, Meece RW, Hill JC. Assessment of the right ventricle by echocardiography: a primer for cardiac sonographers. J Am Soc Echocardiogr. 2009;22:776–792. doi: 10.1016/j.echo.2009.04.027. [DOI] [PubMed] [Google Scholar]
- 21.Moustapha A, Lim M, Saikia S, Kaushik V, Kang SH, Barasch E. Interrogation of the tricuspid annulus by Doppler tissue imaging in patients with chronic pulmonary hypertension: implications for the assessment of right ventricular systolic and diastolic function. Cardiology. 2001;95:101–104. doi: 10.1159/000047354. [DOI] [PubMed] [Google Scholar]
- 22.Meluzin J, Spinarova L, Bakala J, Toman J, Krejci J, Hude P, Kara T, Soucek M. Pulsed Doppler tissue imaging of the velocity of tricuspid annular systolic motion: a new, rapid, and non-invasive method of evaluating right ventricular systolic function. Eur Heart J. 2001;22:340–348. doi: 10.1053/euhj.2000.2296. [DOI] [PubMed] [Google Scholar]
- 23.Wang J, Prakasa K, Bomma C, Tandri H, Dalal D, James C, Tichnell C, Corretti M, Bluemke D, Calkins H, Abraham TP. Comparison of novel echocardiographic parameters of right ventricular function with ejection fraction by cardiac magnetic resonance. J Am Soc Echocardiogr. 2007;20:1058–1064. doi: 10.1016/j.echo.2007.01.038. [DOI] [PubMed] [Google Scholar]
- 24.Gurudevan SV, Malouf PJ, Kahn AM, Auger WR, Waltman TJ, Madani M, DeMaria AN, Blanchard DG. Noninvasive assessment of pulmonary vascular resistance using Doppler tissue imaging of the tricuspid annulus. J Am Soc Echocardiogr. 2007;20:1167–1171. doi: 10.1016/j.echo.2007.02.004. [DOI] [PubMed] [Google Scholar]
- 25.Forfia PR, Fisher MR, Mathai SC, Housten-Harris T, Hemnes AR, Borlaug BA, Chamera E, Corretti MC, Champion HC, Abraham TP, Girgis RE, Hassoun PM. Tricuspid annular displacement predicts survival in pulmonary hypertension. Am J Resp Crit Care Med. 2006;174:1034–1041. doi: 10.1164/rccm.200604-547OC. [DOI] [PubMed] [Google Scholar]
- 26.Giusca S, Dambrauskaite V, Scheurwegs C, D'hooge J, Claus P, Herbots L, Magro M, Rademakers F, Meyns B, Delcroix M, Voigt JU. Deformation imaging describes right ventricular function better then longitudinal displacement of the tricuspid ring. Heart. 2010;96:281–288. doi: 10.1136/hrt.2009.171728. [DOI] [PubMed] [Google Scholar]
- 27.Tei C, Dujardin KS, Hodge DO, Bailey KR, McGoon MD, Tajik AJ, Seward JB. Doppler echocardiographic index for assessment of global right ventricular function. J Am Soc Echocardiogr. 1996;9:838–847. doi: 10.1016/s0894-7317(96)90476-9. [DOI] [PubMed] [Google Scholar]
- 28.Yei TC, Dujardin KS, Tei C, Mahoney DW, McGoon MD, Seward JB. Value of a Doppler-derived index combining systolic and diastolic time intervals in predicting outcome in primary pulmonary hypertension. Am J Cardiol. 1998;81:1157–1161. doi: 10.1016/s0002-9149(98)00140-4. [DOI] [PubMed] [Google Scholar]
- 29.Ross J. Acute displacement of the diastolic pressure-volume curve of the left ventricle: role of the pericardium and the right ventricle. Circulation. 1979;59:32–37. doi: 10.1161/01.cir.59.1.32. [DOI] [PubMed] [Google Scholar]
- 30.Chang SM, Lin CC, Hsiao SH, Lee CY, Huang WC. Pulmonary hypertension and left heart function: insights from tissue Doppler imaging and myocardial performance index. Echocardiography. 2007;24:366–373. doi: 10.1111/j.1540-8175.2007.00405.x. [DOI] [PubMed] [Google Scholar]
- 31.Yilmaz R, Gencer M, Ceylan E, Demirbag R. Impact of chronic obstructive pulmonary disease with pulmonary hypertension on both left ventricular systolic and diastolic performance. J Am Soc Echocardiogr. 2005;18:873–881. doi: 10.1016/j.echo.2005.01.016. [DOI] [PubMed] [Google Scholar]
- 32.Gomberg-Maitland M, Dufton C, Oudiz RJ, Benza RL. Compelling evidence of long-term outcomes in pulmonary arterial hypertension? A clinical perspective. J Am Coll Cardiol. 2011;57:1053–1061. doi: 10.1016/j.jacc.2010.11.020. [DOI] [PubMed] [Google Scholar]
- 33.Oudiz RJ, Galie N, Olschewski H, Torres F, Frost A, Ghofrani HA, Badesch DB, McGoon MD, McLaughlin VV, Roecker EB, Harrison BC, Despain D, Dufton C, Rubin LJ. Long-term ambrisentan therapy for the treatment of pulmonary arterial hypertension. J Am Coll Cardiol. 2009;54:1971–1981. doi: 10.1016/j.jacc.2009.07.033. [DOI] [PubMed] [Google Scholar]